Professional Documents
Culture Documents
Edited by
S. Ceccarelli
E.P. Guimares
E. Weltzien
The designations employed and the presentation of material in this information product do
not imply the expression of any opinion whatsoever on the part of the Food and Agriculture
Organization of the United Nations (FAO) concerning the legal or development status of any
country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers
or boundaries. The mention of specific companies or products of manufacturers, whether or not
these have been patented, does not imply that these have been endorsed or recommended by
FAO in preference to others of a similar nature that are not mentioned.
The views expressed in this information product are those of the authors and do not necessarily
reflect the views of the Food and Agriculture Organization of the United Nations.
ISBN 978-92-5-106382-8
All rights reserved. Reproduction and dissemination of material in this information product
for educational or other non-commercial purposes are authorized without any prior written
permission from the copyright holders provided the source is fully acknowledged. Reproduction
of material in this information product for resale or other commercial purposes is prohibited
without written permission of the copyright holders.
Applications for such permission should be addressed to the Chief, Electronic Publishing Policy
and Support Branch, Communication Division FAO Viale delle Terme di Caracalla, 00153
Rome, Italy or by e-mail to: copyright@fao.org
FAO 2009
iii
Contents
Foreword vii
Abbreviations and acronyms ix
Contributors xiii
Chapter 1
Crop domestication and the first plant breeders 1
Stan Cox
Chapter 2
Theory and application of plant breeding for quantitative traits 27
Javier Betrn, Jess Moreno-Gonzlez and Ignacio Romagosa
Chapter 3
Main stages of a plant breeding programme 63
Salvatore Ceccarelli
Chapter 4
Methodologies for priority setting 75
Eva Weltzien and Anja Christinck
Chapter 5
Methodologies for generating variability. Part 1: Use of genetic
resources in plant breeding 107
Bettina I.G. Haussmann and Heiko K. Parzies
Chapter 6
Methodologies for generating variability. Part 2: Selection of parents
and crossing strategies 129
John R. Witcombe and Daljit S. Virk
Chapter 7
Methodologies for generating variability. Part 3: The development of
base populations and their improvement by recurrent selection 139
John R. Witcombe
Chapter 8
Methodologies for generating variability. Part 4: Mutation techniques 159
Miroslaw Maluszynski, Iwona Szarejko, Chittaranjan R. Bhatia,
Karin Nichterlein and Pierre J.L. Lagoda
iv
Chapter 9
Selection methods. Part 1: Organizational aspects of a plant
breeding programme 195
Salvatore Ceccarelli
Chapter 10
Selection methods. Part 2: Pedigree method 223
Flavio Capettini
Chapter 11
Selection methods. Part 3: Hybrid breeding 229
Donald N. Duvick
Chapter 12
Selection methods. Part 4: Developing open-pollinated varieties
using recurrent selection methods 259
Fred Rattunde, Kirsten vom Brocke, Eva Weltzien and
Bettina I.G. Haussmann
Chapter 13
Selection methods. Part 5: Breeding clonally propagated crops 275
Wolfgang Grneberg, Robert Mwanga, Maria Andrade and Jorge Espinoza
Chapter 14
Breeding for quantitative variables. Part 1: Farmers and scientists
knowledge and practice in variety choice and plant selection 323
Daniela Soleri and David A. Cleveland
Chapter 15
Breeding for quantitative variables. Part 2: Breeding for durable
resistance to crop pests and diseases 367
Raoul A. Robinson
Chapter 16
Breeding for quantitative variables. Part 3: Breeding for resistance to
abiotic stresses 391
Stefania Grando and Salvatore Ceccarelli
Chapter 17
Breeding for quantitative variables. Part 4: Breeding for nutritional
quality traits 419
Yusuf Genc, Julia M. Humphries, Graham H. Lyons and Robin D. Graham
v
Chapter 18
Breeding for quantitative variables. Part 5: Breeding for yield
potential 449
Jos L. Araus, Gustavo A. Slafer, Matthew P. Reynolds and Conxita Royo
Chapter 19
Marker-assisted selection in theory and practice 479
Andrew R. Barr
Chapter 20
Coping with and exploiting genotype-by-environment interactions 519
Paolo Annicchiarico
Chapter 21
Variety release and policy options 565
Zewdie Bishaw and Anthony J.G. van Gastel
Chapter 22
Participatory seed diffusion: experiences from the field 589
Humberto Ros Labrada
Chapter 23
Towards new roles, responsibilities and rules: the case of
participatory plant breeding 613
Ronnie Vernooy with Pratap Shrestha, Salvatore Ceccarelli,
Humberto Ros Labrada, Yiching Song and Sally Humphries
Chapter 24
Breeders rights and IPR issues 629
Susanne Somersalo and John Dodds
Chapter 25
The impact of participatory plant breeding 649
Jacqueline A. Ashby
vii
Foreword
Participatory Plant Breeding (PPB) originated in the early 1980s as part of a movement
promoting the concept of participatory research, in response to criticisms of the failure
of post-green-revolution, experiment-station-based research to address the needs of
poor farmers in developing countries. Rooted in debate over the social consequences of
the narrow focus of the scientific type of research, PPB gained recognition as an activity
mostly promoted by social scientists and agronomists based in anti-establishment non-
governmental organizations (NGOs). In consequence, rather than being perceived from
the beginning as an additional option available to breeders, PPB for a long time had the
image of being one of two contrasting types of plant breeding, with PPB being more
socially correct than conventional plant breeding.
Even now, nearly thirty years later, this view is still common. Few professional
breeders accept that farmers can be full partners in a plant breeding programme, even
though everyone agrees that it was farmers that domesticated crops about 10 000 years
ago and, in some regions of the world, continued to modify and manipulate them to the
present day. Even before the re-discovery of Mendels laws of inheritance, the work of
a number of amateur breeders become an inspiration for Darwins theories. In several
respects, the relationship with farmers on which PPB is based is similar to the ways
in which plant breeders worked with producers in North America and Europe in the
early twentieth century. At that time it was commonplace for breeders to spend time
interacting with producers, and to test new materials collaboratively in farmers fields
in order to understand what producers considered to be desirable traits for an improved
variety. However, the combination of industrialization of agriculture and formal training
for plant breeders created a gap between breeders and farmers, a gap that was exported
to developing countries in the post-war era. As the profession of plant breeding lost the
habit of interacting closely with producers, concern for how to address farmers needs and
constraints fell by the wayside. PPB revived this as a central issue, because by the late 1970s
it was increasingly evident in developing countries that post-green-revolution improved
varieties were too often failing to satisfy farmer requirements and were being shunned.
Today there is widespread recognition that the conventional package of new varieties
and external inputs, while successful in the more favourable production areas, has often
failed to benefit small-scale farmers in marginal areas. As a result, the vital role of PPB as an
additional strategy is better understood. Experience has taught that PPB is complementary
to conventional plant breeding rather than an alternative type of plant breeding. Demand
for a complementary approach has expanded considerably because of pressure to ensure
the relevance of research to poor farmers and their diverse agricultural systems, and because
PPB allows selection for the specific adaptation required for such a diversity of target
environments. Today, about 80 participatory breeding programmes are known worldwide,
involving various institutions and various crops. In 2000, an international review of plant
viii
Contributors
CHAPTER 1
Stan Cox
2 Plant breeding and farmer participation
TABLE 1.1
Species domesticated in each of eight world regions, with approximate age of the oldest evidence
of domestication
Region Species Common name Age of the oldest evidence of
domestication (years BPE)
NOTES: (1) Wendorf et al. (1992) found archaeological evidence that wild millet and sorghum were being used in the Sahel
8000 years before present. The sorghum specimens showed evidence that they were in the process of domestication.
(2) Denham et al., 2003. (3) Independently domesticated in Mesoamerica as well. Species listed are among the worlds 20
most widely grown crops, on a land-area basis (FAO, 2005). Information is from Sauer (1993) unless otherwise indicated.
places and plants that best illustrate the abundance, annuality, lower seed dormancy,
important features of domestication. diploidy, harvestability and relative ease of
seed dehulling.
1.2 SELECTION AMONG SPECIES A common characteristic among crop
There is little doubt that certain species ancestors was their weediness: their tenden-
were pre-adapted (Zohary, 1984) for cy to thrive in disturbed, fertile soils like
domestication. Either entire populations or those associated with human habitation.
individual plants within populations had to The circumstances of domestication are, of
attract the attention of humans before they course, different for every species. In some
could be manipulated. With exceptions, places, people started out by harvesting
plants or populations that exhibited conveniently large stands of annual grasses;
unusually large or numerous edible parts; in others, variations on the so-called rub-
self-pollination (in sexually propagated bish heap theme were at work (Hawkes,
species); ease of propagation (in vegetatively 1969). Many crop ancestors were just as
propagated species); or delayed seed responsible for seeking out humans and
dispersal (e.g. chickpea: Ladizinsky, 1979) human-made environments as were peo-
caught the eyes of early cultivators. Bar- ple for tracking down the plants. Indeed,
Yosef and Kislev (1989) listed characteristics according to Hawkes (1969), it must have
of certain wild cereals (relative to other seemed little short of miraculous to find
wild plant species) that attracted early west that plants needed for food sprang up by
Asian domesticators: larger grain, local their very huts and paths.
4 Plant breeding and farmer participation
In west Asia, however, those destined increase seed yield and produce perennial
to become the first agriculturists tended grain crops (DeHaan, Van Tassel and Cox,
to make their homes near reliable water 2005), but only if the right combination of
sources, whereas they gathered wild grains breeding objectives is established.
from stands that were often some distance When we think of how many civi-
away (Willcox, 2005). Also relying on lizations built on annual cropping have
the west Asian domestication experience, fallen not to the sword but to the plough
Abbo et al. (2005) labelled the rubbish-heap (Hillel, 1991; Lowdermilk, 1953) and the
hypothesis environmental determinism that soil degradation that continues to haunt
tends to underestimate the role of human agriculture today, we can only lament the
initiative in the Neolithic transition. fact that the domesticators did not focus
One thing is certain: the original domes- more on erosion-resistant perennial species.
ticators did not adopt just any species that Apparently, ancient gatherers did utilize
showed up at their doorstep. Then, as now, the seed of perennial species as food. Weiss
people had strong ideas about the useful- et al. (2004) identified charred seeds from
ness of some plant species and the unac- 3 perennial and 12 annual species of small-
ceptability of others. Plants with the most grained grasses that people were consuming
to offer were domesticated long ago, while 23 000 years ago at a site in what is now
others that were sufficiently weedy, but less Israel. Bohrer (1972) discussed traditional
desirable, repeatedly presented themselves methods of harvesting seed from assorted
to humans, only to be ignored or targeted perennial grasses in Poland, Mongolia and
for eradication (Hawkes, 1969). North America. Harlan (1989a) listed a
Prehistoric people gathered and ate foods wide range of perennial grasses that people
from a huge range of plant species, but once living south of the Sahara have harvested
they began domesticating, it was annual for food. Perennial lymegrass (Leymus are-
plants that they transformed. Among the narius) was probably cultivated by Vikings
staple crops in Table 1.1 that yield edible before barley reached Scandinavia (Griffin
reproductive biomass, the banana is the lone and Rowlett, 1981). Yet no domesticated
herbaceous perennial. Herbaceous, grain- perennial grain species were handed down
producing, perennial species are not to be to us by the first plant breeders.
found at all among the worlds crops plants Perennials did not compete well with
(Cox et al., 2002). Herbaceous perennials annuals in disturbed soil and would not have
generally produce less seed in a season than followed people back to the fertile, churned
do annuals. Also, rapid climatic change soil around their dwellings; if some plants
across the Asian continent at the end of did happen to make their way there, they
the Pleistocene dramatically increased the would have been overwhelmed by repeated
availability of those annual, seed-producing disturbance and competition from weedy
species that attracted the attention of culti- annuals. More importantly for Neolithic
vators (Whyte, 1977). The difference in seed domesticators, farming and plant breeding
production between annuals and perennials were one and the same activity. As a result,
is a result of contrasting selection pres- they inevitably carried plant populations
sures during the two groups evolutionary rapidly through sexual cycles, thereby ful-
histories. Selection pressure applied in yet filling an essential requirement of gene-fre-
a different direction by plant breeders can quency change. Perennial plants re-growing
Crop domestication and the first plant breeders 5
from vegetative structures would have been contrast to annual domesticates from that
much more vigorous than either volun- region (Zohary and Spiegel-Roy, 1975).
teer seedlings or intentionally sown plants; Of course, early farmers also practised
therefore, even if people tried to cultivate selection in vegetatively propagated herba-
perennials, they would have felt little incen- ceous species. As with woody species, they
tive to sow new generations from seed. selected clones with desirable character-
As we shall see, the act of sowing har- istics often the results of unusual muta-
vested seed applied strong selection pres- tions and distributed them far and wide.
sure. Selection for non-shattering was Occasional hybridization or somatic muta-
strengthened when people began tilling tion fuelled some continuing selection; for
new land year after year to sow their example, spontaneous yam (Dioscorea spp.)
seed, perhaps as a part of shifting cultiva- clones selected for cultivation by present-
tion to avoid build-up of non-domesticated day farmers in Benin either are wild or are
weeds (Hillman and Davies, 1990). Stands hybrids between cultivars and wild yams
of perennial plants on undisturbed land (Scarcelli et al., 2006; Mignouna and Dansi,
would have been much less vulnerable to 2003). But with only rare sexual recombi-
weeds, much more poorly adapted to shift- nation, there was little opportunity for the
ing cultivation, and therefore less suscepti- degree of domestication seen in grain crops
ble to domestication. One harvest method (Zohary, 2004).
that spurred selection for seed retention in The earliest plant breeders dispropor-
the annual cerealsuprooting of the plant tionate attention to seed-propagated annual
(Bohrer, 1972; Hillman and Davies, 1990) plants has been replicated by most modern
is very difficult with most perennials. students of plant domestication. That pref-
Woody perennials of the Mediterranean erence will be evident in the range of exam-
and west Asiaincluding olive (Olea euro- ples on which the following sections draw.
paea), grape (Vitis vinifera), fig (Ficus car-
ica) and date (Phoenix dactylifera)were 1.3 INITIAL SELECTION WITHIN SPECIES
domesticated in the same region as cereals, It is widely recognized that crops were
but by descendants of the first plant breed- not domesticated simply through gather-
ers, several millennia after agriculture had ing or cultivation. Even the most intensive
been well established (Zohary and Spiegel- harvesting of cereals does not apply suf-
Roy, 1975). Fruit-producing trees and vines ficient selection pressure to domesticate
did not have to compete with annual coun- a crop fully. Intentional sowing, in con-
terparts for humans attention. They were trast, applies strong, unconscious selection
vegetatively propagated, and, even today, pressure (Zohary, 2004). Alleles for non-
most sexual progeny derived from them shattering, lack of dormancy, reproductive
are not only economically worthless, but determinacy and increased fertility of for-
often regress towards the mean found in merly sterile florets are all favoured by the
spontaneous populations, showing striking sowing-harvesting-sowing cycle (Harlan,
resemblance to the wild form (Zohary, De Wet and Price, 1973).
1984). The lack of far-reaching genetic In the west Asia of 10 000 years ago,
changes in Mediterranean tree crops is also wild cereals grew naturally in large fields
manifested in their failure to spread very of near-monoculture, but they were not a
far beyond their original climatic range, in food source that could simply be browsed
6 Plant breeding and farmer participation
at ones convenience. The time between full season. At that point, they write, farmers
ripening and total loss of seed through shat- gathering their first seed stocks from wild
tering was only a week or two, and with hot stands will have been totally unaware of the
dry weather, the period was shortened to existence of these tough-rachised mutant
two or three days (Zohary, 1969). Gatherers forms, and they would have remained
would have needed to be as timely in their oblivious of them as long as the crop stayed
harvest as todays farmers, but the harvest in its essentially wild state.
season was lengthened somewhat by dif- Beating spikes or panicles into a basket
ferences in time of maturity among differ- is the most time-efficient way to harvest
ent cereal species and by elevation differ- wild grain crops (Hillman and Davies,
ences in the hilly Levant. Staggered harvests 1990), but it does not apply selection pres-
would have allowed people to amass large sure for non-shattering. Harlan (1967)
quantities of grain with a relatively long famously collected wild cereals at the rate
shelf-life. At the heart of the wild cereals of 1 kg/hr by hand-stripping of spikes,
native range, people could obtain reliable but that method would not select effec-
harvests from naturally re-seeded stands; it tively against shattering either (Hillman
is therefore most likely that the west Asian and Davies, 1990). Sickling or uprooting
grain crops were first domesticated at the ripe or partially ripe crops does apply
fringes of their progenitors distributions selection pressure, because it shakes loose
(Harlan and Zohary, 1966). It was there some seed from wild-type plants, seed that
that people would have found intentional is lost to the harvester. Hillman and Davies
sowing most helpful in maintaining stands (1990) found experimentally that a consist-
of their proto-crops. At the same time, ently low 40 percent of wild-type seed was
Willcox (2005) emphasized the patchiness recovered by sickling or uprooting. Under
of wild wheat stands throughout the area those conditions, selection would strongly
where emmer wheat was domesticated. favour genes for non-shattering.
People may have felt some incentive to sow In their simulations, such strong selec-
seed, thereby initiating domestication, in tion intensity, combined with the high
any productive localities in that area where degree of self-pollination typical of wheat
wild wheat was not already growing. and barley, would have resulted in com-
A study by Hillman and Davies (1990) plete fixation of a recessive non-shattering
deserves to be discussed at some length, gene within 20 to 30 harvest seasons, if peo-
because it takes into account many of the ple sowed seed each year on virgin land.
factors that affect methods and rates of They further predicted that even if early
domestication in grain crops. They started farmers inadvertently relaxed the selection
by calculating that the rare, recessive muta- pressure by harvesting less fully ripened
tions for non-shattering that were necessary plants or repeatedly sowing on the same
for domestication of the west Asian cereals land, domestication would have been com-
were likely to have appeared once every 5 pleted within two to four centuries. It is no
to 20 years in a typical-sized plot tended wonder that we know so little about the
by an early cultivator. In predominantly mechanics of domestication, according to
self-pollinating wheat and barley, plants Hillman and Davies (1990). If it came and
homozygous for recessive non-shattering went as quickly as they envisioned it, the
alleles would have appeared the following process was unlikely to be preserved on
Crop domestication and the first plant breeders 7
most Mesolithic or Neolithic [archaeologi- and (3) it may simply have become custom-
cal] sites as a recognizable progression. ary during a series of wet summers when
Having assumed in their analysis that ini- wild cereals did not shatter as readily and the
tial domestication was entirely unconscious, beating method of harvest was inadequate.
Hillman and Davies (1990) then demonstrat- When wild cereals of west Asia shatter,
ed that even if Neolithic farmers had practised their morphologically distinct basal spikelet
intentional selection, they could not have remains attached to the culm. That spikelet
greatly speeded up the process. With con- would have been recovered by harvesters
scious selection, people could have done no who sickled or uprooted plants, but not
better than halve the length of time required by those who gathered already-shattered
for domestication, because they could have spikelets from the ground. Basal spikelets
started selecting only when the mutants were might also have been left behind by hand-
frequent enough to be obvious, perhaps at stripping, but that technique requires that
a frequency of 1 to 5 percent of the stand. grain be harvested before it is fully ripe,
By that point, the frequency of mutants had to avoid loss through shattering. Among
already passed through a lag phase and was wild barley and wild emmer remains from
poised for a rapid increase in frequency, even four archaeological sites greater than 11 000
under unconscious selection. years old, Kislev, Weiss and Hartmann
What if, because of a thunderstorm (2004) found no basal spikelets and a
or perhaps an excessive delay in harvest, miniscule number of unripe grains. These
the only intact spikes from which new observations, they maintained, point to
seed stocks could be recovered were those ground collecting as the original harvest
of mutants? Could domestication have method among pre-agricultural people of
occurred in a single season? Hillman and the region. The authors experimented with
Davies (1990) discounted this possibility, ground collection, finding that at any time
based on variation in ripening time and during the regions rainless summer they
the likelihood that birds or other ani- could pick up large clumps of spikelets by
mals would find the isolated spikes before grasping the upward-pointing awns.
humans did. Nevertheless, any environ- Kislev, Weiss and Hatmann (2004) rea-
mental factor that hastened shattering could soned that after the first autumn rains,
have increased the selection pressure and ground gatherers would have noticed seed-
speeded up domestication. lings sprouting from spikelets, and that
Hillman and Daviess argument begs the sight would have inspired them to sow a
question of why early cultivators resorted portion of their harvested seed. Of course,
to sickling or uprooting, if beating is the sowing of ground-collected seed would
most time-efficient harvest method for wild have selected not against but for shattering.
cereals. They suggested three reasons that Kislev, Weiss and Hatmann (2004) do not
sickling or uprooting apparently was pre- speculate on how the transition to sow-
ferred at some point: (1) it recovered more ing of non-shattered seed occurred, but a
seed per unit land area (which, as people scenario based on their results comes to
became more settled, may have become a mind. In collecting seed from the ground,
more important criterion than seed quantity people would have been moving slowly
per unit time); (2) it permitted utilization of through stands of wild cereals long after
the straw for fire-lighting and brick-making; full ripening. Any tough-rachised mutant
8 Plant breeding and farmer participation
with its spike still intact atop the culm may Zohary (1989) forcefully rejected
have attracted their interest, and they may Ladizinskys model, arguing that legume and
well have collected it for sowing in a special cereal domestication followed very similar
plot; if that happened, it would have been a paths, starting with cultivation and sowing
very early case of intentional breeding. of the wild progenitors. He maintained that
Using lentil (Lens culinaris) as a model, wild lentils can produce not ten, but rather
Ladizinsky (1987, 1993) showed how domes- 40 to 70 seeds per plant when well tended in
tication of west Asian legumes might have fertile soil; therefore, people might well have
followed a sequence different from that of found sowing to be worthwhile. Ladizinsky
cereals. He noted that wild lentil (L. orienta- (1989a) responded that the fields of early,
lis) plants are tiny, requiring that an estimated inexperienced cultivators would not have
10 000 plants be gathered in order to obtain been very conducive to high yields, and
one kilogram of clean grain. Therefore, len- that conditions would have been more like
tils could not have been a major part of the those encountered by wild legume stands
gatherers diet, as were cereals, which could than those in Zoharys (1989) tilled, weeded
be gathered much more quickly (Harlan, and well fertilized experiments.
1967). Furthermore, Ladizinsky argued, Some researchers have concluded that
there would have been no incentive for sow- domestication was a rapid process in the
ing; an incipient lentil farmer would have crops they have studied, certainly when
had to sow their entire harvest simply to compared with evolution through natural
produce another crop of equal size. That is selection. Harter et al. (2004) estimated that
because each wild lentil plant produces only in sunflower, genetic composition of the
about ten seeds, of which only one seed on domesticates has changed at least 50-fold
average will germinate the first year, given faster than the wild populations since they
the seeds strong dormancy. diverged. Wang et al. (1999) calculated that
Lentils and perhaps other pulses differed it took approximately 300 to 1 000 years
from cereals, argued Ladizinsky (1987, to completely fix the crucial domestication
1993), in that at least partial domestication gene tb1 that telescopes the lateral branches
had to precede sowing. Through intensive in maize. Other studies indicate a some-
harvesting, people would have drastically what slower process. Jaenicke-Despres et
curtailed natural reseeding, thereby leav- al. (2003) found that as far back as 4 400
ing fields more open to fast-germinating years ago, modern mutant alleles of the
mutants and selecting against seed dorman- genes tb1, pbf (prolamin box binding fac-
cy. Once dormancy was largely eliminated tor) and su1 (starch debranching, which
and people were able to sow seed to good affects tortilla quality) were common. But
effect, selection pressure for indehiscent, that was almost 2 000 years after the date of
non-shattering pods would have been feasi- the oldest known archaeological evidence
ble. But traditional harvesters in southwest of maize domestication. Based on archaeo-
Asia uproot lentil plants before full maturi- logical evidence from northern Syrian Arab
ty, then sun-dry and thresh thema process Republic and southeastern Turkey, Tanno
that largely avoids shattering. If that was the and Willcox (2006) argued that wild cereals
harvest method in Neolithic times, selection could have been cultivated for over 10 000
for non-shattering would have been much years before the emergence of domestic
weaker in legumes than in cereals. varieties, partly because Neolithic cultiva-
Crop domestication and the first plant breeders 9
tors may have taken care to harvest grain 1.4 THE DOMESTICATION BOTTLENECK
before any of it began shattering. That AND GENE FLOW
would have reduced the selection pressure The number of domestication events experi-
on alleles for non-shattering. Fuller (2007) enced by individual species has long been a
argued that during the domestication of favourite topic of debate among researchers.
rice, einkorn and barley, selection for grain Blumler (1992) and Zohary (1999) have
size proceeded faster than selection for non- argued that multiple domestications within a
shattering, but that grain-size increases were species have happened only rarely. They
much slower in pearl millet and leguminous pointed out that genetic variation is much
crops. Surveying the archaeological data, greater in most wild progenitors than in
he found significant grain-size increases in derived domesticates. They also noted the
Asian cereals within a matter of centuries, rarity of parallel domestication in related
a result, he reasoned, of the advantage large taxa above the species level. For example,
seeds had when early cultivators sowed people selected einkorn wheat (Triticum
them deeply in tilled soil. In contrast, he monococcum), pea (Pisum sativum:
concluded, shattering was not fully elimi- Ladizinsky, 1989b), emmer wheat, maize
nated for 1 000 to 2 000 years. and chickpea from their wild ancestors while
Gepts (2002) concluded that models leaving sympatric, phenotypically similar,
based on a few genes can estimate only closely related species undomesticated.
the minimum duration of the domestica- Matsuoka et al. (2002) detected a single
tion process, whereas archaeological data domestication event in maize by analysing
provide a reality check. Physical remains microsatellite variation. Based on amplified
often indicate that domestication took fragment length polymorphism (AFLP)
much longer than would be predicted by variation, Heun et al. (1997) concluded
genetic models. that einkorn was domesticated only once,
Whether farmers transformation of in southeastern Turkey, but that result has
various wild plants into crops went quick- been challenged on archaeological and cli-
ly or slowly, it was not always permanent. matic grounds (Hole, 1998; Jones, Allaby
False starts on the road to domestication and Brown, 1998). Willcox (2005) sum-
may have been common. At sites in west marized archaeological evidence indicating
Asia and North America, groups of peo- that einkorn, emmer and barley all experi-
ple practised relatively intense cultivation enced multiple domestications.
of wild progenitors, and even partially Noting that evidence for single versus
domesticated some species before even- multiple domestication events in Andean
tually abandoning them; those orphaned crops such as amaranth and peppers is
plant populations did not contribute to inconclusive, Blumler (1992) cited several
the founding gene pools of todays crops factors that render it seldom if ever possible
(Weiss, Kislev and Hartmann, 2006). In to rule out multiple independent invention:
one dramatic example of that phenom- the progenitor species may have diversified
enon, domestic rye may have arisen 10 000 after domestication of the crop; loci used
years ago in the Syrian Arab Republic and in comparing the wild and cultivated types
Anatolia, only to disappear for several may be linked to loci affecting traits of
millennia before being re-domesticated in domestication or ecological adaptation; or
Anatolia and Europe (Willcox, 2005). sampling by researchers may be unknowingly
10 Plant breeding and farmer participation
biased. In a simulation study, Allaby and ple had a chance to distribute it over a large
Brown (2003) showed that analyses relying geographical area.
on anonymous genetic markers might The founder effect often occurred when
provide seemingly conclusive evidence that domestication depended upon rare mutants,
a species was domesticated through a single but it was most severe when natural amphip-
event when it was in fact domesticated more loids (doubled interspecific hybrids) were
than once. domesticated. A rare amphiploid taken
The people of South America and those under human care, as was bread wheat,
of Mesoamerica probably took the common would have represented a gene pool consist-
bean through two separate domestications ing of a single plantthe tightest possible
(Sauer, 1993). Xu et al. (2002) concluded, genetic bottleneck (Cox, 1998).
on the basis of chloroplast DNA variation, Tenaillon et al. (2004) found that loss of
that the soybean had a polyphyletic origin, diversity in maize relative to teosinte was
but cluster analysis of nuclear random only 20 percent for putatively neutral loci,
amplified polymorphic DNA (RAPD) compared with 65 percent for loci affected
markers indicated that local differentia- by selection for traits of domestication.
tion of soybean occurred in farmers fields They estimated that the bottleneck that
after domestication was complete (Xu and caused this mild contraction of variability
Gai, 2003). In any case, the soybean passed had a ratio of population size to duration
through a very tight domestication bottle- ranging from approximately 2 to 5. That
neck (Hyten et al., 2006). Barley is unusual is, the bottleneck population might have
among the west Asian cereals in harbour- consisted of 10 000 plants over 2 000 gen-
ing a high level of genetic polymorphism. erations, or perhaps 2 000 plants over 1 000
Ladizinsky (1998) concluded that early generations. Based on data from the Adh-1
cultivators must have selected at least 100 locus, Eyre-Walker et al. (1998) estimated a
non-shattering mutant plants in order to bottleneck size/duration ratio for maize of
capture the level of variability seen in bar- approximately 2; assuming that domestica-
ley. Because the crop is highly self-pollinat- tion took 300 yearssimilar to the dura-
ed, post-domestication gene flow from its tion estimated for einkorn wheatthey
wild progenitor Hordeum spontaneum can- envisioned a bottleneck population of only
not have accounted for the high degree of 600 plants.
variability that is evident today (Ladizinsky Sunflower apparently went through
and Genizi, 2001). a substantial domestication bottleneck,
Whatever the initial number of domes- with inbreeding levels of Native American
tication events, it is clear that because of landraces varying from 0.3 to 0.5 (Harter et
genetic drift the diversity of most crop al., 2004). Abbo, Berger and Turner (2003)
species is low compared with that of their counted three successive bottlenecks that
wild ancestors. Drift results from a genetic tightly restricted the genetic variability of
bottleneck, usually at the point of initial the chickpea crop from its earliest days
domesticationthe well known founder onward: the highly restricted distribution
effect (Ladizinsky, 1985). A bottleneck of its wild ancestor Cicer reticulatum; the
could also be caused by some later event, founder effect resulting from domestica-
but generally would have to occur very tion; and an early shift by west Asian farm-
early in the history of the crop, before peo- ers from autumn to spring sowing of chick-
Crop domestication and the first plant breeders 11
pea (to avoid crop loss due to the Ascochyta often an important source of new variation
blight disease). That shift required selection in crops (Harlan, De Wet and Price, 1973;
of plants without a vernalization require- Small, 1984). People tend to remove from
ment. This third bottleneck, which, they a field those weedy hybrids that do not
argue, occurred early in the crops history, suit their needs, and those weeds tend to
affected chickpea uniquely among the major be less competitive in the natural environ-
west Asian crops. However, it reminds us ment as well. However, when weeds man-
that many species may have passed through aged to backcross to crop plants, their less
bottlenecks caused by intense, early farmer- weedy-looking progeny might well have
directed selection for traits other than seed escaped the early cultivators hand or hoe,
non-dispersal and lack of dormancy. remaining in the domesticated population
Haudry et al. (2007) found that domes- and exchanging genes with it. Weeds often
ticated emmer wheat showed a 70 percent migrate over larger areas than domesticates
loss of nucleotide diversity relative to its and jump from one domesticated popula-
progenitor Triticum dicoccoides. Durum tion to another, exchanging genes along the
wheat, derived by further selection from way (Small, 1984).
emmer, showed an additional diversity loss, Sang and Ge (2007) attempted to rec-
for a total loss of 84 percent. Bread wheats oncile seemingly contradictory evidence
diversity unexpectedly showed only a regarding the origin of the two rice subspe-
69 percent loss relative to T. dicoccoides, cies indica and japonica by showing that the
suggesting extensive introgression from current genetic situation could have arisen
tetraploid wheats during the 8 000 years from either one or two initial domestica-
since the origin of bread wheat. tions, followed by gene flow from the two
Finally, we should take note of a much potential wild progenitors or between the
more recent, possibly catastrophic, bot- partially domesticated subspecies, or both.
tleneck. Clement (1999) documented 138 It follows, they wrote, that introgression
Amazonian plant speciesthe bulk of them practised by modern plant breeding pro-
either fruits, nuts or vegetablesthat were grammes is, in effect, the continuation of
in an advanced state of domestication at domestication.
the time of the first contact with Europeans Weeds unrelated to the crop have at times
five centuries ago. Because these species enticed humans to adopt and domesticate
had become to some extent dependent on them as secondary crops. The ancestors of
humans for their propagation, Clement oats (Avena sativa) and rye (Secale cereale),
maintains that the cataclysmic post-1492 for example, caught the eyes of cultivators
loss of 90 to 95 percent of the areas human while growing as weeds in European wheat
population resulted in an approximate and barley fields (Holden, 1976).
90 percent loss of genetic diversity in plant Through analysis of microsatellites,
species then under cultivation. Matsuoka et al. (2002) determined that the
Introgressive hybridization between genetic diversity of maize was expanded
domesticates and their wild or weedy rela- greatly by introgression from teosinte.
tives has often expanded genetic diversity, Wilkes (1977) found maize farmers in the
counteracting the effects of the domesti- Nobogame Valley of Mexico encouraging
cation bottleneck. Hybridization among the growth of teosinte near and even
domestic, weedy and wild populations is within their maize fields. They told Wilkes
12 Plant breeding and farmer participation
that the teosinte germplasm makes kernels contrast between bread wheat and grain
more flinty and stronger. Nobogame sorghum is instructive (Cox and Wood,
was the only area in which Wilkes found 1999). Hexaploid bread wheat may well
hybridization intentionally fostered, and, have originated from only one or two
interestingly, it was the only place where hybrid plants with genomic constitution
the flowering times of maize and teosinte ABD (Cox, 1998; Haudry et al., 2007).
were somewhat synchronized. In other The tetraploid ancestor (AB) had experi-
areas, people weeded out teosinte, but, enced only limited introgression from dip-
at least in Chalco, they fed it to cattle as loid plants, mostly of the A-genome spe-
fodder, then inadvertently returned its seed cies. Subsequent gene flow from AB into
to the field when applying manure. It is ABD wheat plants occurred to some extent
possible that such mechanisms also played (Haudry et al., 2007), but gene flow from
a part in the introgression of teosinte the extremely diverse D-genome donor
genes into maize in the early phases of Aegilops tauschii into bread wheat was
domestication. either non-existent or extremely rare until
Gene flow into crops has been impor- it was done by twentieth-century plant
tant in crop evolution, but there is a much breeders (Cox, 1998). Therefore, through-
larger flow in the opposite direction: from out the entire bread wheat species, there
the domesticate into the wild form. That is limited genetic variability in the A and
would probably have been the case in B genomes, while its D genome contains
Neolithic grain fields as well. Migration only a tiny fraction of the diversity found
of large amounts of wild pollen into fields in Aegilops tauschii (Reif et al., 2005).
of self-pollinated crops was limited, and In contrast, people of Africa have
because pollen from the wild conveyed always grown grain and fodder sorghum
dominant genes for shattering, hybrid in areas where the crop comes into close
progeny were not likely to be collected or contact and interbreeds with wild sor-
planted by farmers (Ladizinsky, 1985). At ghum races (Doggett and Majisu, 1968).
the same time, there is much evidence that They probably domesticated sorghum in
genes regularly migrated out of fields and various, widespread locales on multiple
into wild populations (Ladizinsky, 1985; occasions, after which it was exposed to a
Harlan, De Wet and Price, 1973). Many continuous inflow of variability from the
studies have estimated hybridization rates wild and weedy gene pools. As a result,
by looking for crop-specific alleles in pop- grain sorghum today harbours vastly more
ulations of the crops wild relatives grow- genetic diversity than does bread wheat
ing at various distances from cultivated (Cox and Wood, 1999).
fields. They generally find surprisingly
high rates, even hundreds of metres away 1.5 GENETIC CONSEQUENCES OF
(Ellstrand, 2003). SELECTION
Differences among crop species in the Van Raamsdonk (1995) proposed that most
sizes of their founding populations and domesticated crops were developed through
subsequent opportunities for gene inflow one of four genetic models (Table 1.2). The
from the wild have profoundly affected models differ in the role of ploidy and
the levels of genetic diversity available the degrees and mechanisms of reproduc-
to present-day plant breeders. Here, the tive isolation. Differences in genetic and
Crop domestication and the first plant breeders 13
Reproductive isolation between a Soybean, common bean, chickpea, Selection for self-pollination and
diploid domesticate and its diploid lentil, cowpea (Vigna unguiculata), against weedy hybrids; fostering of
wild ancestor is caused by internal lettuce (Lactuca sativa), citrus fruits genetic drift
barriers, post-zygotic barriers, (Citrus spp.)
external reproductive barriers or
apomixis.
Development of crop-weed- Maize, rice, barley, grape, sorghum, Adoption of weeds that invade
wild complexes in which genetic pearl millet, foxtail millet (Setaria cultivated land; toleration or
information is exchanged more italica), radish (Raphanus sativus), encouragement of weeds that can
or less freely among diploid beet (Beta spp.), chili (Capsicum spp.), backcross to less wild cultigens
domesticates and their sexually quinoa (Chenopodium quinoa)
compatible wild progenitors.
One or more rounds of hybridization Cotton, sweet potato, groundnut, Selection at the polyploidy level
and polyploidization occur among tobacco (Nicotiana spp.), cucumber
wild species prior to domestication. (Cucumis spp.), coconut (Cocos
nucifera), alfalfa (Medicago sativa)
Interspecific hybridization involving Bread wheat, potato, banana, coffee Bringing formerly isolated plant
at least one domesticated species (Coffea arabica), yam (Dioscorea spp.) populations into contact; selection
is followed by polyploidization. and propagation of rare amphiploid
Resultant amphiploids are plant(s) found in or near cultivated
reproductively isolated. fields.
In some cases, domestication Sugar cane, oat, Brassica spp., tomato
occurs through a combination of (Lycopersicon esculentum)
mechanisms from more than one of
the above models.
14 Plant breeding and farmer participation
(Doebley, Gaut and Smith, 2006). Whatever selection in vegetatively propagated species
the nature of their mutations, alleles initial- allowed, or even encouraged, variations in
ly selected by domesticators often showed chromosomal number and structure, dis-
the simplest modes of inheritance. Many rupting reproductive development to vary-
genes governing traits of domestication are ing extents (Zohary, 2004).
recessive or additive, and would have been In a simulation study, Le Thierry
expressed more strongly among the prog- dEnnequin et al. (1999) predicted that to
eny of plants that tended to self-pollinate fix a full complement of alleles for domes-
most frequently. An increased tendency to tication, either linkage among loci or a
inbreed may also have been an indirect result significant degree of reproductive isolation
of selection for higher grain yield; self-pol- is essential. By their models, in predomi-
lination ensures seed and fruit development, nantly self-pollinating species subject to
especially if the new crop was transported little migration, people easily fixed alleles at
out of the range of its natural pollinators. unlinked loci through selection; however,
Inbreeding also leads to greater within- in species with a high degree of out-cross-
line uniformity, but it is hard to imagine ing, human selection favoured blocks of
uniformity being a direct selection criterion linked domestication genes.
for early domesticators, as it would have Empirical experiments have demonstrat-
required that they plant out the progeny ed that linkage among domestication loci
of individual plants in separate plots. It is common, regardless of breeding system
is almost certain that they practised mass (Paterson, 2002). In crosses between pearl
selection, not progeny testing. But genes millet and its wild progenitor Pennisetum
promoting self-pollination might have been mollissimum, Poncet et al. (1998, 2000, 2002)
favoured in very small populations main- found linkage among genes affecting spike
tained in isolation. Such isolation could charactersimportant components of the
have resulted from individual preferences, domestication syndromebut not among
or perhaps community customs, such as genes affecting vegetative characters or total
a belief in parts of Guatemala that plants grain yield. Burke et al. (2002) mapped 78
should be grown only from seed produced quantitative trait loci (QTLs) affecting 18
on the same plot of ground (Pickersgill, traits in a cross between sunflower and its
1969). Colour coding (Wilkes, 1989) based conspecific wild progenitor. The domestica-
on endosperm pigmentation may have tion-associated loci were spread across 15 of
helped farmers maintain small, genetically 17 linkage groups, but were highly clustered
isolated maize populations. within those groups. Both pearl millet and
Strong selection to reinforce inbreeding sunflower are highly cross-pollinated. In
did not occur in crops that were propa- rice, a selfing species, QTLs affecting domes-
gated vegetatively; in them, self-incompat- tication traits also tended to be clustered in
ibility and out-crossing remained common linkage groups (Cai and Morishima, 2000).
(Zohary, 2004; Rick, 1988). Through clonal Wright et al. (2005) found that 2 to 4 per-
propagation, cultivators could produce cent of the genes in maize have probably
large, genetically desirable populations. undergone artificial selection. Much of that
In contrast to seed-propagated species, in selection, especially for the genes involved
which human selection for improved grain in plant growth and auxin response that are
harvests also reinforced meiotic stability, responsible for the dramatic differences in
Crop domestication and the first plant breeders 15
plant morphology between teosinte and favoured during periods of rapid evolu-
maize, appears to have occurred during ini- tionary change, of which domestication is
tial domestication. Those growth-pattern an extreme example. Ross-Ibarra found no
genes were clustered, whereas genes affect- support for the alternative possibility: that
ing amino acid composition were not. species with higher recombination rates are
In wild progenitors, significant numbers pre-adapted to domestication.
of agronomically beneficial alleles are often Even under domestication, the recom-
embedded in linkage blocks with other, bination rate is under stabilizing rather
deleterious, genes. Such desirable alleles than unidirectional selection, because the
tended to be left behind during domestica- same high rates that help break up repul-
tion. For example, in a tetraploid wheat sion linkages also speed up the decay of
population, Peng et al. (2003) found 24 per- co-adapted gene complexes (Dobzhansky,
cent of positive QTL effects to be coming 1970). Indeed, Ross-Ibarras comparison of
from the wild Triticum dicoccoides parent. crop and wild species provided evidence for
By breaking up such linkage blocks, mod- selection against excessive recombination.
ern-day breeders can utilize genes that were There are, of course, other mechanisms for
hidden from early domesticators. maintaining favourable multilocus combi-
Gepts (2002), surveying studies of nations, including paracentric inversions
domestication traits in maize, pearl millet, (Dobzhansky, 1970) and self-pollination
common bean and rice, found an aver- (Clegg, Allard and Kahler, 1972).
age of 2.2 to 5.3 loci per trait. Those loci
accounted for only about 50 percent of the 1.6 INTENTIONAL SELECTION
total variation per trait, and loci affecting Although crops were domesticated through
all traits were spread among 3 to 5 linkage largely unintentional selection, there is lit-
groups per species, indicating rather diffuse tle doubt that the domesticators quickly
genetic control. Paterson (2002) found sim- became aware of their own ability to change
ilar patterns in the QTL-mapping literature the phenotypic composition of their crops
on sorghum, rice, maize and tomato. He from generation to generation. Genetic
concluded that loci with larger statistical modification, once initiated, spread in ever-
effects were probably biologically signifi- widening ripples through plant genomes.
cant as well, because they occurred in simi- Sowing spurred unconscious selection for
lar genomic regions in different crop spe- traits like non-shattering; changes caused
cies (Paterson, 2002; Paterson et al., 1995). by unconscious selection prompted observ-
During domestication, people may have ant farmers to practise intentional selection;
unknowingly favoured plants or popula- and intentional selection for one trait often
tions with a higher inherent rate of recom- affected other traits as well, through linkage
bination per unit of physical chromosomal and pleiotropy. Studies of a grain-quality
length. A comprehensive survey showed trait in rice show that human selection at a
that mean numbers of chiasmata per biva- single locus can exert very strong selection
lent were significantly higher in 46 crop pressure on a large chromosomal region
species than in 150 wild species (Ross- surrounding it, causing a so-called selec-
Ibarra, 2004). This result was in accord tive sweep that can affect other traits much
with theory, the bulk of which predicts more strongly than would natural selection
that an increased rate of recombination is (Olsen et al., 2006).
16 Plant breeding and farmer participation
From the dawn of agriculture until affected others; for example, selection for
the twentieth century, farmers acted as larger spikes in the cereals produced wider
plant breeders, working almost exclusively leaves and thicker culms as well.
through mass selection; that is, by ensuring Smartt (1969) catalogued the many
that some individual plants made a pro- traits for which early domesticators applied
portionately greater genetic contribution selection pressure in species of Phaseolus:
to the following generation than did oth- a reduced number of lateral branches (to
ers. Natural out-crossing would have been avoid excessive tangling in fields where
frequent enough, even in highly self-pol- beans were meant to climb maize plants);
linating species, to generate useful genetic more robust leaves and stems; larger flow-
recombinants. Early plant breeders worked ers; increased photoperiod sensitivity;
without the benefits of progeny testing increased pod and seed size; greater per-
or replication, both of which can enhance meability of the testa; and reduced pod
gain from selection, but they had two dehiscence. However, in examining four
other important factors working in their cultivated species, he found that not all of
favour: time and ecosystems. Even small those traits were affected in every species.
gene-frequency changes from year to year Chang (1976a, b) noted a similarly
translated into large improvements when increased size of vegetative organs and
they continued over vast numbers of grow- kernels in rice, along with a more exten-
ing seasons. And plant populations upon sive root system; higher tillering capacity;
which people exerted gradual selection in synchronization of tillering; more pani-
a particular locality, through the full range cle branches; a longer grain-filling period;
of weather conditions and pest, pathogen, tolerance to non-flooded conditions; and
weed and intercrop populations that the loss of pigmentation. However, increases
locality had to offer, were bound to be in kernel size and harvest index associ-
resilient and reliable food producers. ated with domestication of rice were less
When people applied direct selection than those in most other cereals (Cook
pressure for some traits, whether inten- and Evans, 1983). In several species of chili
tional or unconscious, they put indirect (Capsicum), people rejected erect-fruited
selection pressure on others. For example, wild plants in favour of mutants with
attached glumes increase seed dormancy, pendant fruits, which were hidden under
so selection for non-dormancy may have the foliage canopy and therefore protected
increased the frequency of free-threshing from bird damage (Pickersgill, 1969).
plants. Deep sowing may have favoured Maize is often recognized as a crop that
larger-seeded genotypes (Fuller, 2007), underwent some of the most remarkable
which, in turn, would have had lower morphological changes during domestica-
grain protein concentrations via dilution. tion, but, as in most crops, the most obvi-
Selection for greater allocation of resources ous transformation was in its reproductive
to reproductive growth (higher harvest structures. In the words of Iltis (2000),
index) could have increased susceptibil- Cover the ears, and it sometimes takes a
ity to pests (Rick, 1988). Because plant specialist to tell teosinte from maize But
parts growing from the same meristematic compare a many-rowed, 1000-grained ear
regions exhibit allometric growth, selection of maize to a 2-rowed, 5-to-12-grained
to increase the size of one organ generally ear of teosinte and be perplexed! How
Crop domestication and the first plant breeders 17
could such a massive, useful monster be niques, all of which were certain to reveal
derived from such a tiny, fragile, inedible, genetic variation in the crops upon which
useless mouse? they were practised.
Perhaps just as surprising is the find- Human selection for nutritional qual-
ing that morphological differences between ity of crop domesticates occurred in the
maize and its wild ancestor are under rela- context of other crops that were evolv-
tively simple genetic control (Doebley and ing simultaneously. The most commonly
Stec, 1993). cited example is the complementarity of
Maize is not the only species whose amino acid profiles in cereals and legumes.
reproductive structures evolved into mon- Selection among and within species was
strosities under the guiding hand of early a matter of health, even life and death.
breeders. For example, pearl millets wild Indeed, Wilkes (1989) declared an ethno-
ancestor has heads measuring no more than botanical rule, stating that when crops are
10 cm in length, but from it, early breed- consumed and not sold, a reasonable level of
ers selected cultivars with heads up to 2 m nutritional adequacy has evolved and been
long (Harlan, 1989b). In bringing about the maintained. Neither the single-minded
visually dramatic domestication of the sun- selection for high grain yield per unit area
flower, Native Americans selected for the nor the pursuit of high-lysine maize would
fusion of many smaller heads into fewer, have occurred to a Mesoamerican farmer of
larger ones. People worldwide selected 3 000 years ago.
for often dramatically larger reproductive Plant breeding requires differential phe-
structures in vegetable and fruit crops. notypic expression. For example, people
Plant breeding theory, as well as obser- could not venture very deeply into the
vation of crop domesticates, tells us that domestication and improvement of a species
the first breeders had their biggest impact as a food source if its consumption always
on traits that (i) were of the most intense resulted in serious illness or death. Indeed,
interest to the people who used the plants the process by which the sweet almond was
for food; (ii) were under relatively simple derived from its cyanogenic ancestor is still
genetic control; and (iii) had a relatively shrouded in mystery (Ladizinsky, 1999).
high heritability on a single-plant or single- People could begin selecting for lower tox-
propagule basis. Therefore, humans altered icity once they accomplished at least par-
the appearance and food quality of the har- tial breakdown of toxins through cooking.
vested product more rapidly than they did Other strategies were developed farther
traits such as yield per unit area. Contrasting back in the human family tree. Geophagy
intentional selection with the unconscious consumption of claysis practiced by at
selection that preceded and paralleled it, least eight primate species (Johns, 1989).
Harlan, De Wet and Price (1973) wrote: People commonly eat clay along with wild
Deliberate selection adds new dimensions potatoes (Johns, 1986) and yams (Irvine,
to the process [of domestication]. Human 1952) to de-toxify them, and the prac-
selection may be more intense and abso- tice might have provided latitude for early
lute and is often biologically capricious or domesticators to distinguish among dif-
even whimsical. ferent degrees of bitterness without falling
They went on to list a bewildering too ill too often. Once foods were rendered
array of food products and processing tech- edible via such practices, selection for lower
18 Plant breeding and farmer participation
toxicity might have been furthered simply be consumed by rival species. Reducing
through dilution, as people selected for the mean toxicity to a safer level allowed
greater root or tuber size (Johns, 1989). them to detect and exploit genetic variation
In the potato, there is a remarkable within species.
coincidence between toxic thresholds and Toxins aside, the simplification of diet
human capacity for detection. The plants that followed the expansion of agriculture
most common glycoalkaloid is toxic in appears in itself to have caused a decline in
concentrations above 200 ppm (Johns and overall human health (Kates, 1994). Gepts
Keen, 1986), and tubers with a concentra- (2002) even implies that had regulatory
tion of greater than 140 ppm are considered agencies existed in Neolithic times, domes-
unpleasantly bitter by North Americans ticated plants might well have failed to
(Sinden and Deahl, 1976). In contrast, the receive approval!
Aymara Indians of the Andes classify pota- Selection for food quality involved more
toes with concentrations above a range of than nutritional considerations. Where
200 to 380 ppm as bitter (Johns and Keen, muscle and fuel power were resources not
1986). Because several wild and cultivated to be squandered, genotypes that produced
Solanum species are crucial sources of calo- food with lower energy requirements for
ries in the Andes, the Aymara and other processing and cooking may have been
indigenous people may have developed more highly valued. For example, Harlan
a taste for somewhat riskier genotypes. (1989b) described how modern cultivators
Selection for improved nutritional qual- in Mali select sorghum heads with softer
ity can also work against improvement of grains for ease of pounding, but also keep
other traits. For example, potato popula- hard-seeded, more insect-resistant types,
tions selected for lower glycoalkaloid con- for longer-term storage.
centrations had lower resistance to potato In some cases, people may have utilized
leafhopper (Sanford et al., 1992). the progenitor of a crop for one pur-
In a seeming paradox, cyanogenesis (the pose only to find, once they became more
production of poisonous hydrocyanic acid) familiar with the species, that it possessed
is more common in crop plants than in the one or more other traits that warranted
plant kingdom as a whole. Jones (1998) its full domestication. For example, many
noted that 16 of the worlds 24 leading crop East Asian plants may have been used for
species (by total production) are cyano- medicinal purposes before being domesti-
genic in some plant part(s) at some stage cated for food production (Chang, 1970).
of growth. Cyanogenesis, Jones observed, Bohrer (1972) maintained that the wild
is an important mechanism of resistance to grasses that eventually gave rise to cereal
pests. People looking to become cultiva- crops were originally cut or uprooted for
tors, given a wide range of plant species use as animal fodder. However, Hillman
from which to choose, would probably and Davies (1990) disputed that idea, argu-
have been attracted to plants that had not ing that at the time and place of west Asian
already been damaged or largely consumed crop domestication there were no domestic
by other species. Having the unique abil- cattle and few domestic sheep or goats. The
ity to eliminate cyanogenic glycosides by squash (Cucurbita pepo) may have been
grinding, steeping and cooking, humans domesticated first for its seed, or for its
took advantage of plants that could not hard gourds to be used as containers; once
Crop domestication and the first plant breeders 19
fleshy vegetable genotypes were select- ...injects desirous human agents into the
ed, people may have stopped growing the account, a palliative for the stern food
gourd types to prevent the appearance of crises and population pressures that
bitter squashes through cross-pollination haunt our angst-driven prehistories. How
(Heiser, 1989). charming it would be to have a snack-
Iltis (2000) concluded that teosinte was and-party crowd, hassled by only an occa-
first grown by Mesoamericans for its green sional aggrandizer or two, at the base of
shoots and sugary pith and not for its grain, the Neolithic!
which remained enclosed in a hard fruit- The initial domestication of crops
case. Later, through increased contact with prompted expansion of farming into new
teosinte as a snack or vegetable, an alert environments, where people continued
cultivator may have noticed an extremely selection under different conditions, while
rare, grain-liberating mutanton possi- perhaps repeating the domestication proc-
bly a single occasionthus kicking off the ess with new species. Although the ability
process of maize domestication. to accumulate a large excess of grain during
Amplifying Iltiss hypothesis, Smalley a brief harvest season provided, in itself, a
and Blake (2003) suggested and then strong incentive to settle in one locality for
defended a possible sequence of events by at least a good part of the year [as Flannery
which teosinte domestication proceeded: (1969) asked regarding a hypothetical com-
(1) people began casually harvesting and munity of Neolithic gatherers, after all,
chewing the sweet stalks and shoots of where could they go with an estimated met-
Zea plants; (2) they found that they could ric tonne of clean wheat?], people eventu-
extract more juice by mechanical mashing; ally and inevitably migrated. The ability to
(3) to preserve the juice, they adopted fer- take with them a food source that doubled
mentation techniques already in use with as the means of sowing future crops allowed
other species; (4) they spread maize far and people to expand agriculture into previous-
wide, as a new resource for making alcohol- ly unsettled areas, where the crops encoun-
ic beverages; and, finally, (5) to expand Zea tered new selection pressures and the people
cultivation, they began sowing harvested encountered new species of plants.
seed. Once that sequence proceeded as far Abandoned fields created by early
as step (5)along with the discovery of the shifting cultivation in tropical forests may
free-kernel mutantdomestication of Zea have provided environments in which
mays as a grain crop would have followed useful wild plants could survive and
quickly. But the time between its very first grow unusually well, possibly to become
utilization by chewing and its full domesti- domesticates themselves (Piperno, 1989).
cation as a grain may have been as long as Barleys early maturity allowed farming at
2 500 years (Smalley and Blake, 2003). very high altitudes; pearl millets drought-
Perhaps too often, researchers tend to hardiness extended agriculture into parts of
portray the era of crop domestication as India and Africa that receive 200 mm or less
one of constant struggle against scarcity of annual rainfall; and maize brought more
and hardship. DeBoer (2003) commented people into the sparsely populated, mid-
that the possibility of people first having altitude hill country of India and Pakistan
utilized maize for sweet and fermented (Harlan, 1972). However, once settled in
products. new environments, thanks to a reliable
20 Plant breeding and farmer participation
staple crop, people have not always sought Keen observation and use of genetic var-
out additional species for domestication; iation in plant species has been a hallmark
rather, monocultures are common on the of societies that depend directly on those
fringes of agriculture (Harlan, 1972). plants, whether the people in those societies
were hunter-gatherers, the originators of
1.7 CONCLUSIONS agriculture, or todays subsistence farmers.
In recent decades, institutional plant breed- As the millennia have passed, knowledge has
ers have come to realize the importance expanded and methods have evolved, but
of integrating breeding methodology with that thread remains intact. Todays institu-
farmers knowledge. Doing so has benefits tional plant breeders also benefit from that
for breederswhose selection goals become accumulated knowledge. Although modern
more embedded in the real worldand breeders methodologies are often very dif-
for farmers, who come to appreciate better ferent, they are rooted firmly in the past.
their own ability to change gene frequen- They also utilize that major part of the first
cies of their crops in favourable directions. plant breeders unwritten knowledge that
This would appear to bring us full circle, survives in code, the genetic code of the
to a time like that of agricultures earliest plants themselves.
days, when breeding and farming were Had the original crop domesticators been
fully integrated. But todays agriculturalists familiar with the principles of genetics, the
also have ten millennia worth of hard-won crop species that they handed down to his-
farming and breeding knowledge on which tory might have been even more profoundly
they can draw by working together. transformed. Had they understood the haz-
The first plant breeders lived in pre- ards of genetic erosion or pest and pathogen
historic times, so they left us no direct epidemics, they might have domesticated
accounts of the methods they used to a wider range of species and avoided the
domesticate and improve crops. As we have genetic bottlenecks that restricted variation
seen, many of our hypotheses about their in many crops from the very beginning.
activities are influenced by our knowledge And could they have foreseen the devastat-
of the methodologies that farmer-breed- ing consequences of soil erosion and water
ers have used in historic times. That is no contamination under long-term annual
accident. By extrapolating recent methods cropping (Cox et al., 2006), they might have
back to the origin of agriculture, we are mounted an effort to domesticate resource-
acknowledging a 10 000-year-long, unbro- efficient perennial food crops.
ken thread of skills and knowledge that Nevertheless, that relative handful of
is derived from growing plants for food people was responsible for the most impor-
while simultaneously breeding them for the tant turning point humanity has yet experi-
future. Nevertheless, we should not forget enced, laying the foundation for the material
that by coming to rely largely on domesti- and cultural world that surrounds us today.
cated plants and animals, we humans have But in the evolution of agriculture, it has
also lost vast amounts of knowledge of not been the case that superior knowledge
other species and ecosystems; there is much and techniques continuously replace inferior
that we could re-learn from hunter-gath- ones. Knowledge survives from every era, all
erer societies of the present, the recent past the way back to the origin of crops (and even
and the days before agriculture. well before), so that farmers, plant breeders
Crop domestication and the first plant breeders 21
and all others who work in agriculture can L.H. 2002. Genetic analysis of sunflower
draw upon it in the years ahead. domestication. Genetics, 161(3): 12571267.
As it has turned out, the first plant Cai, H.W. & Morishima, H. 2000. QTL clus-
breeders brought about changes in our own ters reflect character associations in wild
species that equal any they achieved with and cultivated rice. Theoretical and Applied
plants, and the plant breeding traditions Genetics, 104(8): 12171228.
they established have brought humanity, Chang, K. 1970. The beginnings of agriculture
only in the past century, to a point at which in the Far East. Antiquity, 44(1): 175185.
we can study why and how they carried Chang, T.T. 1976a. The rice cultures.
off their revolution, and learn from the Philosophical Transactions of the Royal
answers. Society, London, Series B, 275(936): 143
157.
ACKNOWLEDGEMENT Chang, T.T. 1976b. The origin, evolution,
I am indebted to Sheila Cox for her assis- cultivation, dissemination, and diversifica-
tance in reviewing and interpreting litera- tion of Asian and African rices. Euphytica,
ture on the genetics of domestication. 25(2): 425441.
Clegg, M.T., Allard, R.W. & Kahler, A.L. 1972.
REFERENCES Is the gene the unit of selection? Evidence
Abbo, S., Berger, J. & Turner, N.C. 2003. from two experimental plant populations.
Evolution of cultivated chickpea: Four Proceedings of the National Academy of
bottlenecks limit diversity and constrain Sciences of the United States of America,
adaptation. Functional Plant Biology, 69(9): 2474-2478.
30(10): 10811087. Clement, C.R. 1999. 1492 and the loss of
Abbo, S., Gopher, A., Rubin, B., & Lev- Amazonian crop genetic resources. I. The
Yadun, S. 2005. On the origin of Near relation between domestication and human
Eastern founder crops and the dump-heap population decline. Economic Botany,
hypothesis. Genetic Resources and Crop 53(2): 188202.
Evolution, 52(5): 491495. Cook, M.G. & Evans, L.T. 1983. Some physi-
Allaby, R.G. & Brown, T.A. 2003. AFLP data ological aspects of the domestication and
and the origins of domesticated crops. improvement of rice (Oryza spp.) Field
Genome, 46(3): 448453. Crops Research, 6(3): 219238.
Bar-Yosef, O. & Kislev, M. 1989. Early farm- Cox, T.S. 1998. Deepening the wheat gene pool.
ing communities in the Jordan Valley. In Journal of Crop Production, 1(1): 125.
D.R. Harris & G.C. Hillman, eds. Foraging Cox, T.S. & Wood, D. 1999. The nature and
and Farming: the Evolution of Plant role of crop biodiversity. In D. Wood
Exploitation, pp. 632642. London, Unwin & J.M. Lenne, eds. Agrobiodiversity:
Hyman. 733 p. Characterization, Utilization, and
Blumler, M.A. 1992. Independent inventionism Management, pp. 3559. Wallingford, UK,
and recent genetic evidence on plant domes- CAB International. 490 p.
tication. Economic Botany, 46(1): 98111. Cox, T.S., Bender, M., Picone, C., Van Tassel,
Bohrer, V.L. 1972. On the relation of harvest D.L., Holland, J.B., Brummer, C.E., Zoeller,
methods to early agriculture in the Near B.E., Paterson, A.H. & Jackson, W. 2002.
East. Economic Botany, 26(2): 145155. Breeding perennial grain crops. Critical
Burke, J.M., Tang, S., Knapp, S.J. & Rieseberg, Reviews in Plant Science, 21(2): 5991.
22 Plant breeding and farmer participation
Cox, T.S., Glover, J.G., Van Tassel, D.L., Cox, effects of early domestication in Iran and the
C.M. & DeHaan, L.R. 2006. Prospects Near East. In P.J. Ucko & G.W. Dimbleby,
for developing perennial grain crops. eds. The Domestication and Exploitation of
BioScience, 56: 649659. Plants and Animals, pp. 73100. London,
DeBoer, W.R. 2003. Comment on J. Smalley Duckworth. 581 p.
& M. Blake, Sweet beginnings: stalk sugar Fuller, D.Q. 2007. Contrasting patterns in crop
and the domestication of maize. Current domestication and domestication rates:
Anthropology, 44(5): 690691. recent archaeobotanical insights from the
DeHaan, L.R., Van Tassel, D.L. & Cox, T.S. Old World. Annals of Botany, 100(5): 903
2005. Perennial grain crops: a synthesis of 924.
ecology and plant breeding. Renewable Gepts, P. 2002. A comparison between crop
Agriculture and Food Systems, 20(1): 514. domestication, classical plant breeding,
Denham, T.P., Haberle, S., Lentfer, C., and genetic engineering. Crop Science,
Fullagar, R., Field, J., Therin, M., Porch, N. 42(6): 17801790.
& Winsborough, B. 2003. Origins of agri- Griffin, L.C. & Rowlett, R.M. 1981. A lost
culture at Kuk swamp in the highlands of Viking cereal grain. Journal of Ethnobiology,
New Guinea. Science, 301(5630): 189193. 1(1): 200207.
Dobzhansky, T. 1970. Genetics of the Harlan, J.R. 1967. A wild wheat harvest in
Evolutionary Process. New York, USA, Turkey. Archaeology, 20(2): 197201.
Columbia University Press. 505 p. Harlan, J.R. 1972. Crops that extend the range
Doebley, J.F., Gaut, B.S. & Smith, B.D. 2006. of agricultural settlement. In P.J. Ucko, R.
The molecular genetics of crop domestica- Tringham & G.W. Dimbleby, eds. Man,
tion. Cell, 127(7): 13091321. Settlement and Urbanism, pp. 239243.
Doebley, J. & Stec, A. 1993. Inheritance of the London: Duckworth. 979 p.
morphological differences between maize Harlan, J.R. 1989a. Wild-grass seed harvesting in
and teosinte: comparison of results for two the Sahara and Sub-Sahara of Africa. In D.R.
F2 populations. Genetics, 134(2): 559570. Harris & G.C. Hillman, eds. Foraging and
Doggett, H. & B.N. Majisu. 1968. Disruptive Farming: the Evolution of Plant Exploitation,
selection in crop development. Heredity, pp. 7998. London, Unwin Hyman. 733 p.
23(1): 123. Harlan, J.R. 1989b. The tropical African cere-
Ellstrand, N.C. 2003. Current knowledge als. In D.R. Harris & G.C. Hillman, eds.
of gene flow in plants: implications for Foraging and Farming: the Evolution of
transgene flow. Philosophical Transactions Plant Exploitation, pp. 335-343. London,
of the Royal Society, London, Series B, Unwin Hyman. 733 p.
358(1434): 11631170. Harlan, J.R., De Wet, J.M.J. & Price, E.G.
Eyre-Walker, A., Gaut, R.L., Hilton, H., 1973. Comparative evolution in cereals.
Feldman, D.L. & Gaut, B.S. 1998. Evolution, 27(3): 311325.
Investigation of the bottleneck leading to Harlan, J. & Zohary, D. 1966. Distribution
the domestication of maize. Proceedings of wild wheats and barley. Science,
of the National Academy of Sciences of the 153: 10741079.
United States of America, 95(8): 44414446. Harris, D.R. & Hillman, G.C. eds. 1989.
FAO. 2005. FAO Statistical Databases (available Foraging and Farming: the Evolution
at http://faostat.fao.org/faostat/servlet/). of Plant Exploitation. London, Unwin
Flannery, K.V. 1969. Origins and ecological Hyman. 733 p.
Crop domestication and the first plant breeders 23
Harter, A.V., Gardner, K.A., Falush, D., Shoemaker, R.C., & Cregan, P.B. 2006.
Lentz, D.L., Bye, R.A. & Rieseberg, L.H. Impacts of genetic bottlenecks on soy-
2004. Origin of extant domesticated sun- bean genome diversity. Proceedings of the
flowers in eastern North America. Nature, National Academy of Sciences of the United
430(6996): 201205. States of America, 103(45): 1666616671.
Haudry A., Cenci, A., Ravel, C., Bataillon, T., Iltis, H.H. 2000. Homeotic sexual transloca-
Brunel, D., Poncet, C., Hochu, I., Poirier, tions and the origin of maize (Zea mays,
S., Santoni, S., Glmin, S. & David, J. Poaceae): a new look at an old problem.
2007. Grinding up wheat: a massive loss Economic Botany, 54(1): 742.
of nucleotide diversity since domestica- Irvine, F.R. 1952. Supplementary and emer-
tion. Molecular Biology and Evolution, gency food plants of West Africa. Economic
24(7): 15061517. Botany, 36(1): 8499.
Hawkes, J.G. 1969. The ecological back- Jaenicke-Despres, V., Buckler, E.S., Smith,
ground of plant domestication. In P.J. B.D., Gilbert, M.T.P., Cooper, A., Doebley,
Ucko & G.W. Dimbleby, eds. The J. & Paabo, S. 2003. Early allelic selection in
Domestication and Exploitation of maize as revealed by ancient DNA. Science,
Plants and Animals, pp. 1920. London, 302(5648): 12061208.
Duckworth. 581 p. Johns, T. 1986. Detoxification function of
Heiser, C.B. Jr. 1989. Domestication of geophagy and domestication of the potato.
Cucurbitaceae: Cucurbita and Lagenaria. Journal of Chemical Ecology, 12(3): 635646.
In D.R. Harris & G.C. Hillman, eds. Johns, T. 1989. A chemical-ecological model
Foraging and Farming: the Evolution of of root and tuber domestication in the
Plant Exploitation, pp. 471480. London, Andes. In D.R. Harris & G.C. Hillman,
Unwin Hyman. 733 p. eds. Foraging and Farming: the Evolution
Heun, M., Schafer-Pregl, R., Klawan, D., of Plant Exploitation, pp. 504519. London,
Castagna, R., Accerbi, M., Borghi, B. & Unwin Hyman. 733 p.
Salamini, F. 1997. Site of einkorn wheat Johns, T. & Keen, S.L. 1986. Taste evaluation
domestication identified by DNA finger- of potato glycoalkaloids by the Aymara:
printing. Science, 278(5341): 13121314. A case study in human chemical ecology.
Hillel, D. 1991. Out of the Earth: Civilization Human Ecology, 14(4): 437452.
and the Life of the Soil. Berkeley, USA, Jones, D. 1998. Why are so many food plants
University of California Press. cyanogenic? Phytochemistry, 47(2): 155
Hillman, G.C. & Davies, M.S. 1990. 162.
Domestication rates in wild-type wheats Jones, M.K., Allaby, R.G. & Brown, T.A.
and barley under primitive cultivation. 1998. Wheat domestication. Science,
Biological Journal of the Linnean Society, 279(5349): 302
39(1): 3978. Kates, R. 1994. Sustaining life on the earth.
Holden, J.H.W. 1976. Oats. pp. 8690, in Scientific American, 271(4): 114122.
N.W. Simmonds (editor). Evolution of Crop Kislev, M.E., Weiss, E. & Hartmann, A. 2004.
Plants. New York, USA, Longman. 381 p. Impetus for sowing and the beginning of
Hole, F. 1998. Wheat domestication. Science, agriculture: ground collecting of wild cere-
279(5349): 302. als. Proceedings of the National Academy
Hyten, D.L., Song, Q., Zhu, Y., Choi, I.-Y., of Sciences of the United States of America,
Nelson, R.L., Costa, J.M., Specht, J.E., 101(9): 26922695.
24 Plant breeding and farmer participation
Konishi, S., Izawa, T., Lin, S.Y., Ebana, K., Lowdermilk, W.C. 1953. Conquest of the land
Fukuta, Y., Sasaki, T. & Yano, M. 2006. An through seven thousand years. Agricultural
SNP caused loss of seed shattering during Information Bulletin, 99(1): 130.
rice domestication. Science, 312: 13921396. Matsuoka, Y., Vigouroux, Y., Goodman, M.,
Ladizinsky, G. 1979. Seed dispersal in relation Sanchez, J., Buckler, E. & Doebley, J. 2002.
to the domestication of Middle East leg- A single domestication for maize shown
umes. Economic Botany, 33: 284289. by multilocus microsatellite genotyping.
Ladizinsky, G. 1985. Founder effect in Proceedings of the National Academy of
crop-plant evolution. Economic Botany, Sciences of the United States of America,
39(2): 191199. 99(9): 60806084.
Ladizinsky, G. 1987. Pulse domestication before Mignouna, H.D. & Dansi, A. 2003. Yam
cultivation. Economic Botany, 41(1): 6065. (Dioscorea spp.) domestication by the Nago
Ladizinsky, G. 1989a. Pulse domestication: fact and Fon ethnic groups in Benin. Genetic
and fiction. Economic Botany, 43(1): 131 Resources and Crop Evolution, 50(5): 519
132. 528.
Ladizinsky, G. 1989b. Origin and domesti- Olsen, K.M., Caicedo, A.L., Polato, N.,
cation of the Southwest Asian grain leg- McClung, A., McCouch, S., & Purugannan,
umes. In D.R. Harris & G.C. Hillman, eds. M.D. 2006. Selection under domestication:
Foraging and Farming: the Evolution of evidence for a sweep in the rice Waxy
Plant Exploitation, pp. 374389. London, genomic region. Genetics, 173(2): 975983.
Unwin Hyman. 733 p. Paterson, A.H. 2002. What has QTL mapping
Ladizinsky, G. 1993. Lentil domestication: taught us about plant domestication? New
on the quality of evidence and arguments. Phytologist, 154(3): 591608.
Economic Botany, 47(1): 6064. Paterson, A.H., Lin, Y., Liu, S. & Chang,
Ladizinsky, G. 1998. How many tough-rachis Y. 1995. Convergent domestication of
mutants gave rise to domesticated bar- cereal crops by independent mutations
ley? Genetic Resources and Crop Evolution, at corresponding genetic loci. Science,
45(5): 411414. 269(5231): 17141718.
Ladizinsky, G. 1999. On the origin of almond. Peng, J., Ronin, Y., Fahima, T., Roder, M.S., Li,
Genetic Resources and Crop Evolution, Y., Nevo, E. & Korol, A. 2003. Domestication
46(2): 143147. quantitative trait loci in Triticum dicoc-
Ladizinsky, G. & A. Genizi. 2001. Could early coides, the progenitor of wheat. Proceedings
gene flow have created similar allozyme- of the National Academy of Sciences of the
gene frequencies in cultivated and wild bar- United States of America, 100: 24892494.
ley? Genetic Resources and Crop Evolution, Pickersgill, B. 1969. The domestication of chili
48(1): 101104. peppers. In P.J. Ucko & G.W. Dimbleby,
Le Thierry dEnnequin, M., Toupance, B., eds. The Domestication and Exploitation of
Robert, T., Godelle, B. & Gouyon, P.H. Plants and Animals, pp. 443-450. London,
1999. Plant domestication: a model for Duckworth. 581 p.
studying the selection of linkage. Journal of Piperno, D.R. 1989. Non-affluent foragers:
Evolutionary Biology, 12(6): 11381147. resource availability, seasonal shortages, and
Li, C., Zhou, A. & Sang, T. 2006. Rice domes- the emergence of agriculture in Panamanian
tication by reduced shattering. Science, tropical forests. In D.R. Harris & G.C.
311: 19361939. Hillman, eds. Foraging and Farming: the
Crop domestication and the first plant breeders 25
Evolution of Plant Exploitation, pp. 538 Crop Plants: A Selected Roster. Boca Raton,
554. London, Unwin Hyman,. 733 p. USA, CRC Press.
Poncet, V., Lamy, F., Enjalbert, J., Joly, H., Scarcelli, N., Tostain, S., Mariac, C., Agbangla,
Sarr, A. & Robert, T. 1998. Genetic analysis C., Da, O., Berthaud, J. & Pham, J.-L.
of the domestication syndrome in pearl 2006. Genetic nature of yams (Dioscorea
millet (Pennisetum glaucum L., Poaceae): spp.) domesticated by farmers in Benin
Inheritance of the major characters. (West Africa). Genetic Resources and Crop
Heredity, 81(6): 648658. Evolution, 53: 121130.
Poncet, V., Lamy, F., Devos, K., Gale, M., Sarr, A. Sinden, S.L. & Deahl, K.L. 1976. Effect of gly-
& Robert, T. 2000. Genetic control of domes- coalkaloids and phenolics on potato flavor.
tication traits in pearl millet (Pennisetum Journal of Food Science, 41(3): 520523.
glaucum L., Poaceae). Theoretical and Small, E. 1984. Hybridization in the domes-
Applied Genetics, 100(2): 147159. ticated-weed-wild complex. In W.F Grant,
Poncet V., Martel E., Allouis S., Devos ed. Plant Biosystematics, pp. 195210. New
M., Lamy F., Sarr A. & Robert T. 2002. York, USA, Academic Press. 674 p.
Comparative analysis of QTLs affecting Smalley, J. & Blake, M. 2003. Sweet beginnings:
domestication traits between two domesti- stalk sugar and the domestication of maize.
cated wild pearl millet (Pennisetum glau- Current Anthropology, 44(5): 675703.
cum L., Poaceae) crosses. Theoretical and Smartt, J. 1969. Evolution of American
Applied Genetics, 104(6): 965975. Phaseolus beans under domestication. In
Reif, J.C., Zhang, P., Dreisigacker, S., Warburton, P.J. Ucko & G.W. Dimbleby, eds. The
M.L., van Ginkel, M., Hoisington, D., Bohn, Domestication and Exploitation of Plants
M. & Melchinger, A.E. 2005. Wheat genetic and Animals, pp. 451-462. London,
diversity trends during domestication and Duckworth. 581 p.
breeding. Theoretical and Applied Genetics, Tanno, K. & Willcox, G. 2006. How fast
110: 859864. was wild wheat domesticated? Science,
Rick, C.M. 1988. Evolution of mating systems 311(5769): 1886.
in cultivated plants. In L.D. Gottlieb & S.K. Tenaillon, M.I., URen, J., Tenaillon, O. &
Jain, eds. Plant Evolutionary Biology, New Gaut, B.S. 2004. Selection versus demog-
York, USA, Chapman and Hall. 414 p. raphy: a multilocus investigation of the
Ross-Ibarra, J. 2004. The evolution of recom- domestication process in maize. Molecular
bination under domestication: A test of Biology and Evolution, 21(7): 12141225.
two hypotheses. American Naturalist, Thompson, K.F. 1976. Cabbages, kales, etc.
163(1): 105111. In N.W. Simmonds, ed. Evolution of
Sanford, L.L., Deahl, K.L., Sinden, S.L. & Crop Plants, pp. 4952. New York, USA,
Ladd, T.L. 1992. Glycoalkaloid contents of Longman.
tubers from Solanum tuberosum popula- van Raamsdonk, L.W.D. 1995. The cytological
tions selected for potato leafhopper resist- and genetical mechanisms of plant domes-
ance. American Potato Journal, 69(11): 693 tication exemplified by four crop models.
703. Botanical Review, 61(4): 367399.
Sang, T. & Ge, S. 2007. The puzzle of rice Wang, R.-L., Stec, A., Hey, J., Lukens, L.
domestication. Journal of Integrated Plant & Doebley, J. 1999. The limits of selec-
Biology, 49(6): 760768. tion during maize domestication. Nature,
Sauer, J.D. 1993. Historical Geography of 398(6724): 236239.
26 Plant breeding and farmer participation
Weiss E., Kislev, M.E. & Hartmann, A. 2006. ple origins of cultivated soybean. Theoretical
Autonomous cultivation before domestica- and Applied Genetics, 105(5): 645653.
tion. Science, 312(5780): 16081610. Xu, D.H. & Gai, J.Y. 2003. Genetic diversity
Weiss, E., Wetterstrom, W., Nadel, D. & of wild and cultivated soybeans growing in
Bar-Yosef, B. 2004. The broad spectrum China revealed by RAPD analysis. Plant
revisited: Evidence from plant remains. Breeding, 122: 503-506.
Proceedings of the National Academy of Zeder, M.A., Bradley, D.G., Emshwiller, E.
Sciences of the United States of America, & Smith D.B. eds. 2006. Documenting
101: 95519555. Domestication: New Genetic and
Wendorf, F., Close, A.E., Schild, R., Wasylikowa, Archaeological Paradigms. Berkeley, USA,
K., Housley, R.A., Harlan, J.R. & Krolish, University of California Press. 375 p.
H. 1992. Saharan exploitation of plants 8000 Zohary, D. 1969. The progenitors of wheat
years BP. Nature, 359(6397): 721724. and barley in relation to domestication
Whyte, R.O. 1977. The botanical Neolithic and agricultural dispersal in the Old
revolution. Human Ecology, 5(3): 209222. World. In P.J. Ucko & G.W. Dimbleby,
Wilkes, H.G. 1977. Hybridization of maize eds. The Domestication and Exploitation
and teosinte in Mexico and Guatemala of Plants and Animals, pp. 4766. London,
and the improvement of maize. Economic Duckworth. 581 p.
Botany, 31(3): 254293. Zohary, D. 1984. Modes of evolution in plants
Wilkes, H.G. 1989. Maize: domestication, racial under domestication. In W.F Grant, ed.
evolution, and spread. In D.R. Harris & Plant Biosystematics, pp. 579586. New
G.C. Hillman, eds. Foraging and Farming: York, USA, Academic Press. 674 p.
the Evolution of Plant Exploitation, pp. 440 Zohary, D. 1989. Pulse domestication and
455. London, Unwin Hyman. 733 p. cereal domestication: how different are
Willcox, G. 2005. The distribution, natural hab- they? Economic Botany, 43(1): 3134.
itats and availability of wild cereals in rela- Zohary, D. 1999. Monophyletic vs. polyphylet-
tion to their domestication in the Near East: ic origin of the crops on which agriculture
multiple events, multiple centres. Vegetation was founded in the Near East. Genetic
History and Archaeobotany, 14: 534541. Resources and Crop Evolution, 46(2): 133
Wright, S.I., Bi, I.V., Schroeder, S.G., Yamasaki, 142.
M., Doebley, J.F., McMullen, M.D. & Gaut, Zohary, D. 2004. Unconscious selection
B.S. 2005. The effects of artificial selection and the evolution of domesticated plants.
on the maize genome. Science, 308: 1310 Economic Botany, 58(1): 510.
1314. Zohary, D. & Spiegel-Roy, P. 1975. Beginnings
Xu H., Abe, J., Gai, Y. & Shimamoto, Y. 2002. of fruit growing in the Old World. Science,
Diversity of chloroplast DNA SSRs in wild 187(4174): 319327.
and cultivated soybeans: evidence for multi-
27
CHAPTER 2
money (Doorenbos, 1954). The German (3) Transportation of animals from one
city of Quedlinburg became known from country to another could be worthwhile,
the middle of the nineteenth century with provided that the introduced stock was
the establishment of profitable seed com- carefully bred in every subsequent genera-
panies breeding vegetables and flowers. tion; (4) Selective breeding was a powerful
The Vilmorin Company, a commercial agent of change, even for creating new
label still in operation, was established in breeds; (5) The more inbred a strain was,
1743 in France, when Philippe-Victoire the more likely to pass the selected traits to
de Vilmorin, a horticulturist and Pierre its progeny; (6) Both sexes contributed to
dAndrieux, a seed collector and botanist heredity, and either could be prepotent, and
of Louis XV, set up the boutique Andrieux thus characteristics could be transferred to
and Vilmorin in Paris (www.vilmorin- the opposite sex; (7) Carefully controlled
clause.com). One of the members of the progeny testing was the most efficient way
family, Louis de Vilmorin, a contemporary to evaluate an individuals hereditary prop-
of Mendel, introduced two key techniques erties; (8) Selective breeding was applied to
in modern plant breeding: first, progeny single traits or groups of traits; (9) Visible
testing, as a alternative way to assess the traits could indicate hidden properties; and
value of a given individual based on the (10) The value of crossing, as an adjunct to
phenotype of its offspring rather than just selection, was still a matter of controversy,
its own; and, second, indirect selection, although the first generation of a cross was
whereby sucrose yield in the beet root was becoming accepted for its hybrid vigour.
measured by means of a refractometer.
Continuous breeding boosted sugar con- 2.1.2 The onset of Quantitative
tent from 56 percent to 20 percent in just Genetics: the Mendelians vs.
a few decades. Biometricians debate, and the Neo-
Parallel to plant breeding, by the end of Darwinian synthesis
the eighteenth century, animal breeding had The theory of plant breeding rests on the
developed intensively, based on practical work of two most influential biologists,
experience. The great success worldwide of Charles Darwin (18091882) and Gregor
the breeders of the Spanish Merino sheep Mendel (18221884), and the passionate
by the leading pioneers in this field such as debate that took place between their fol-
Robert Bakewell (17251795) proved that lowers at the beginning of the twentieth
breeding was an empirical endeavour in century.
which ideas arising out of observation were Darwin published in 1859 The Origin
ahead of academic knowledge. Wood and of Species, in which he persuasively dem-
Orel (2001) summarized the ten intuitive onstrated that evolution had occurred. He
principles and practices used at that time, then elaborated the Natural Selection
most of them inadvertently compatible hypothesis to explain the evolutionary
with Mendelian inheritance: (1) The intrin- process. He backed his observations in
sic nature of an animal (its breed and blood) nature with the gains of artificial selection
was the most critical factor determining achieved in both animal and plant breeding.
its form and quality; (2) Good husband- He strongly supported gradual changes act-
ry (mainly diet and housing) were essen- ing over time, rather than discontinuous or
tial for maximizing their intrinsic quality; abrupt changes. However, he lacked a con-
30 Plant breeding and farmer participation
FIGURE 2.1
Key experimental evidence in the Mendelians vs. Biometricians debate
C
(A) Johannsens Pure-Line Theory (1903): heritable (selection between lines) and non-heritable (within a given line)
variation for seed weight in beans. (B) Nilsson-Ehles Multiple Factor Hypothesis (1909): F2 segregation ratios for
wheat cultivars differing in two alleles at one, two and three loci controlling seed colour. (C) Easts experiment on
the corolla length in Nicotiana longiflora (1916): the apparent continuous variation in the F2 could be modelled by
the superimposition of environmental effects (estimated to be approx. 18 mm) and genetic effects determined by a
reduced number of independent genes (approx. 5 mm per allelic substitution considering five loci with two alleles at
each locus).
Theory and application of plant breeding for quantitative traits 31
vincing theory for the origin of variation ing the sudden production of new species.
and a proper theory of inheritance. Darwin This theory was fervently accepted by the
assumed the prevailing model of the blend- Mendelians of the early twentieth century,
ing inherence as observed in progenies led by De Vries and William Bateson. They
of interbred animal crosses by which the considered the selective value of small vari-
attributes of an individual were the result ations negligible and believed genetic vari-
of merging or blending their parents char- ation could not be explained statistically.
acteristics (a complete fusion of parental On the contrary, biometricians such as
and maternal particles called gemmules). Francis Galtons disciples, guided by Karl
Jenkin (cited by Griffing, 1994) soon real- Pearson and Raphael Weldon, developed
ized that under blending inheritance, the statistical techniques for describing and
variation in the offspring would be halved analysing relationships between relatives
after each generation of random mating with respect to continuous variation. They
and, thus, variation within any population strongly supported Darwinism and rejected
would be quickly exhausted: if Mendelism as just a series of simple naive
general principles. As Griffing (1994) men-
xm + x f then tions, a profound controversy developed
xo =
2 between the two groups, augmented by the
xm + x f 1 2 personalities of the scientists involved.
2(xo ) = 2 ( ) 1
= ( x m ) + 2 ( x f ) = 2 ( x p )
2 4 2 The core of the debate was whether
continuous variation observed for
where xo, xm,, xf,, and xp, represent the off- metric characters could be reconciled
spring, male, female and parental values, with the discrete Mendelian factors and
respectively. New variability should have their inheritance. Mendels laws and
to be generated to maintain the level of biometrical methods were recognized
variation within the population. as complementary after a series of key
Mendels laws of segregation clearly plant experiments carried out in the first
established that any trait is determined decades of the twentieth century in what
by a pair of factors, gametes containing Griffing (1994) wisely called the Era of
just one of the two factors taken at ran- demystification and reconciliation. Two
dom. The law of independent assortment important questions were answered: first,
establishes that factors from the parents What are the basic causes of continuous
independently combine in the offspring. variation? and, second, What is the nature
Mendel supplemented Darwins Natural of the genotypic variation?
Selection because a direct consequence of Johannsen (1903) addressed the first
his laws was that genetic variation could question with his so-called Pure Line
be preserved through time. Whereas genes Theory. He studied the seed size in a
according to Mendel were conserved over heterogeneous bean variety, Princess, a
the generations, gemmules received from self-pollinated species that was a mixture
any parent according to the blending the- of pure lines. He proved that continuous
ory were physically lost by being merged selection within a pure-line did not trans-
together. late into larger grains, but genetic advances
Hugo De Vries, contrary to Darwinism, could be achieved upon individual selection
proposed the mutations theory for describ- within a mixture of lines (Figure 2.1A). He
32 Plant breeding and farmer participation
distinguished heritable from non-herita- the era of the 1920s was blessed
ble variability and proposed the concept by having three of the right scientists
of Genotype and Phenotype, which led (Fisher, Haldane and Wright) with spe-
to the formulation of the central equa- cial interest and abilities in the right
tion in breeding: Phenotype = Genotype + area (mathematical biology) at the right
Environment. time.
The Multiple Factor Hypothesis was They adopted Mendelian inheritance to
demonstrated by Nilson-Ehle in Sweden describe in mathematical terms the basic
in 1909, answering Griffings second ques- principles of Darwinian evolution (and,
tion. Crossing different wheat varieties thus, both natural and artificial selection).
of red and white seed colour, he found Their developments constitute the core
distinct intensities of red to white kernel of the plant breeding theory described
3:1, 15:1 and 63:1 segregation ratios in the below.
F2 (Figure 2.1B), depending on the red As a key component of genetic inher-
variety used. He postulated that two alleles itance, Morgan, Sturtevant, Bridges and
at each of three independent loci control- Muller clearly demonstrated, working with
led the trait which increasingly showed a Drosophila during the 1920s, the linear
continuous phenotypic distribution. East order of the genes in the chromosomes. By
(1916) published the final experiment that 1926, once Morgans book, The Theory of
clearly brought together Mendelians and the Gene, was published, it was generally
Biometricians. He studied the length of recognized that genetic maps indicating the
the corolla of progenies of two differ- position and order could be theoretically
ent strains of Nicotiana longiflora, which constructed for any organism (Stebbins,
seemed to display blending inheritance. 1994). It has not been until recently, with
The apparent continuous variation in the the advent of molecular markers, that this
F2 could be explained by the superimposi- asseveration has been fully developed. It is
tion of environmental effects (estimated to also worthwhile mentioning that the con-
be approx. 18 mm) on the corolla length cepts of marker-assisted selection (MAS)
determined by a reduced number of inde- and quantitative trait loci (QTL) can also
pendent genes (approx. 5 mm per allelic be traced to original studies in beans by Sax
substitution considering five loci with two in 1923 or in tomato by Lindstrom (1926).
alleles at each locus) (Figure 2.1C). Through the use of mathematical mod-
The crucial work for the development els, quantitative genetics studies the genetic
of the Theory of Plant Breeding was pub- architecture and heritability of plant traits,
lished by Fisher in 1918. He introduced the the genetic relationship among them, and
term variance and used its additive prop- the interaction between genotypes and
erties to partition the phenotypic variance environments. Quantitative genetics thus
into its components according to a genetic provides the foundation for the design
model. He provided the theoretical frame- and utilization of breeding approaches
work for the final settlement between to improve crops. In the next section,
Biometricians and Mendelians and estab- we describe the nature of plant traits and
lished the basis of quantitative genetic genetic phenomena affecting their expres-
theory, on which both animal and plant sion, ways to estimate heritability and
breeding rest. As Griffing says (1994), subsequent response to selection.
Theory and application of plant breeding for quantitative traits 33
of quantitative traits 1 3
Each genotype may have a norm of reaction 2 9
3 27
(i.e. a range of phenotypes) instead of a single
5 243
unique expression. Therefore, the same gen- 10 59049
otypes can have different phenotypes, based
on the environmental influence (Figure 2.2). N 3n
34 Plant breeding and farmer participation
TABLE 2.2
Phenotypic values, genotypes and number of plant breeding. First, the consideration of
genotypes per phenotypic value (assuming two an allele as favourable or unfavourable
loci with two alleles each (A/a and B/b) with may depend on the genotype at other loci.
alleles A and B adding one unit to the final
Therefore there are favourable and unfa-
phenotypic value)
vourable combinations of alleles that breed-
Genotype Phenotypic Number of
Value genotypes per ers select for. Epistasis affects the average
phenotype
effects of alleles and dominance deviations
AABB 4 1
AaBB, AABb 3 2 and, consequently, the additive and domi-
AAbb, AaBb, aaBB 2 3 nance genetic variances (see Section 2.2.2). If
Aabb, aaBb 1 2 epistasis is strong, there can be more herit-
Aabb 0 1 able variance within lines during selfing and
line development than expected. Second,
traits. Additive effects are the average effect epistasis reduces the correlation between
of alleles, which are associated with the the expression of quantitative traits of early
number of copies of each allele. Dominance and later selfing-generations. With the pres-
effects originate from the interaction ence of epistasis, early generation testing
between (among) alleles at the same locus. and selection is expected to be less effective
For a single locus in a diploid species, the than delaying selection until later genera-
comparison between heterozygous loci (e.g. tions when epistatic effects (e.g. additive
A1A2) with the parental homozygous loci additive epistatic effects) are fixed within
(e.g. A1A1 and A2A2) determines the level lines. Third, epistasis contributes to hetero-
of dominance: no dominance (i.e. A1A2 = sis and inbreeding depression, although in
(A1A1 + A2A2)/2), partial dominance (i.e. different manners. While hybrids and popu-
A1A2 is between (A1A1 + A2A2)/2 and A1A1 lation-cross cultivars can exploit all forms
or A2A2), complete dominance (i.e. A1A1 of epistasis (additive additive, additive
= A1A2 when allele A1 is dominant over dominant and dominant dominant), only
A2), and overdominance (i.e. A1A2 > A1A1 dominance and dominance dominance
or A2A2). Quantitative traits are therefore epistasis contribute to inbreeding depres-
regulated by many genes having diverse sion (Holland, 2001). Inbreeding depression
types of genetic effects. More details about is heterosis in reverse only when epistasis is
genetic effects and their influence in genetic absent, but not in the presence of epistasis.
variation and breeding are given by Falconer In summary, continuous variation for
and Mackay (1996) and Bernardo (2002). a quantitative trait is the result of the
effect of multiple genetic factors and the
The expression of individual genes is often environment. Hence, many genotypes can
modified by the expression of other genes have the same phenotype, and the same
(i.e. epistasis) genotype can have different phenotypes.
Epistasis is the interaction between alleles The number of loci affecting a quantitative
from different genes (i.e. interloci or non- trait, their genomic position, their effects in
allelic genetic interaction) (Holland, 2001). the final phenotype, the interactions among
For two loci, epistasis is the failure of a them, their regulation and the effect of the
gene replacement at one locus to remain the environment in gene expression is largely
same when a gene is replaced at the other unknown for most of quantitative traits
locus. Epistasis has strong consequences in (Bernardo, 2002).
Theory and application of plant breeding for quantitative traits 35
2.2.2 Means, variances and estimates are used to assess heritability and
correlations for quantitative traits expected response to selection.
Genotypic value (G) is the value of a The characterization of genetic prop-
genotype for the trait under consideration. erties of quantitative traits (loci effects,
Classically in quantitative genetics, G has intra-locus gene action, epistasis, pleiot-
been divided into additive, dominance and ropy, linkage, allele frequencies, environ-
epistatic effects: G = A + D + I (Falconer mental influence in gene expression, etc.)
and Mackay, 1996). The variation among has been pursued through the study of
genotypic values in a breeding population observable phenotypic properties (mean,
is the genotypic variance (G2). The breed- variance, resemblance between relatives,
ing value of one individual assesses its correlation among traits, response to selec-
usefulness in selection. It is determined by tion, inbreeding depression and heterosis).
the mean of its progeny and is associated More recently, advances in plant genom-
with additive effects. Breeding value in one ics and molecular biology and physiology
individual is twice the mean deviation of its are contributing to better understand the
outcrossed progeny from the population genetic architecture of quantitative traits
mean, which is equal to the sum of the aver- and facilitating the connection between
age effects of the alleles it carries. The vari- genetic and phenotypic properties (Cooper,
ation among breeding values is attributed Podlich and Smith, 2004).
to the additive effects of genes and is called Observable phenotypic properties can
additive genetic variance ( A2 ). Dominance be estimated statistically. For two traits (X
deviation, D, is the difference between the and Y) measured in several experimental
genotypic value (G) and the breeding value units (i.e. n genotypes), means, variances,
(A) for a given genotype in the absence of covariance and correlation can be estimated
epistasis. The dominance deviations are due as follows:
to within-locus interaction between the dif- Means:
1 n 1 n
ferent alleles. Variation among genotypes X xi Y = yi
for dominance deviations is the dominance n 11 n 1=1
genetic variance (D2). Finally, the vari- Variance:
ation associated with differences among 1 n 2 1 n 2
genotypes for epistatic interactions is the x2
xi X
xi nX 2
n 1 i 1 n 1 i 1
epistatic variance (I2).
Genetic variance is the sum of the addi- Covariance:
tive, dominance and epistatic genetic vari- 1 n 1 n
ance components: G2 = A2 + D2 + I2 XY xi X yi Y n 1
n 1 i 1
xi y i nXY
i 1
selected subpopulations from broad-based and selection intensities, and the breeding
populations (e.g. pearl millet). Average het- approaches employed.
erozygosity is much greater in cross-polli- When considering allele and genotypic
nated species than in self-pollinated species. frequencies at several loci together there
Selfing induces inbreeding depression while is another type of disequilibrium, known
cross-pollination induces hybrid vigour. as gametic phase or linkage disequilib-
The degrees of inbreeding depression and rium (LD). LD is defined as the non-
hybrid vigour are greater in cross-polli- random association of alleles at different
nated crops than in self-pollinated crops. loci (Bernardo, 2002). For two loci, it
Most of the scenarios in quantitative is measured as the difference between
genetics are studied in populations the observed gamete frequencies and the
of genotypes. These populations can product of the frequencies of the corre-
be characterized by their genotype and sponding alleles:
allele frequencies. In large, random-
mating populations the gene frequencies D = p ( Ai B j ) p ( Ai ) p ( B j )
and the genotype frequencies have a
simple relationship ( pij = pi p j ) and are where p( A B ) is the frequency for the AiBj
i j
2.2.5 The environment and its Selection response in the target environ-
interaction with genotypes in plant ment can be expressed as:
breeding
The environment affects the expression of RT = G ihs rG
quantitative traits, and different environ-
ments can affect genotypes differently (See where G = genetic standard deviation, i =
Chapter 20). Phenotypic values are classi- selection intensity, hs = square root of
cally divided into genotypic (G), environ- heritability in the selection environment,
mental (E) and genotype environmental and rG = genetic correlation between the
interaction (GE) effects: P = G + E + GE. selection and target environments (adapted
Likewise, phenotypic variance is divided from Bnziger and Cooper, 2001). Thus
into genotypic, environmental and GE the effectiveness of a selection environment
variance components: P2 G2 E2 GE2 . In is determined by the heritability of the
breeding, several genotypes are commonly traits(s) under selection in that environ-
evaluated in several environments. When ment, and the genetic correlation between
genotypes tested differ in their relative the performance in the selection environ-
performance across environments there is ment and the target environment (i.e. indi-
GE, which can affect response to selec- rect selection theory, see Section 2.2.8).
tion. Non-crossover interaction, i.e. where When the heritability is low or the cor-
rank of genotypes does not change across relation is low or negative, or both, little
environments, does not have any effect in progress in the target environment can
selection as the best and worst genotypes be expected regardless of who does the
are the same in all locations. Crossover selection (farmers or breeders). Multi-
interaction, where the rank of genotypes environment trials conducted at a large
changes across environments, has strong number of sites to adequately sample the
consequences in breeding as best and worst target environment maximizes the corre-
genotypes in different environments are not lation between target and selection envi-
the same (Romagosa and Fox, 1993). There ronments, facilitates selection of broadly
are different breeding strategies that deal adapted hybrids or varieties, and exploita-
with this issue (Annicchiarico, 2002, and tion of GE. Weighted selection strategies,
see Chapter 20 in this volume). Definition where individual trials found more relevant
of target and selection environments is crit- to the target environments are given more
ical when allocating resources in a breeding emphasis than less relevant trials, can be
programme. In this process, several deci- used (Podlich, Cooper and Basford, 1999).
sions have to be made regarding selection Increasing the number of environments, as
for broad versus specific or local adapta- the number of entries decreases during the
tion; selection on farm versus research breeding process, evaluating in environ-
station; and selection under optimal ver- ments that disclose genetic variation for
sus stressed conditions (Atlin, Cooper and the traits under selection, and combining
Bjrnstad, 2001). Cultivars can perform the understanding of the genetic control of
well under a wide range of environments target traits and of the target environments,
(broad adaptation) or under specific grow- are important components of successful
ing conditions (narrow, specific or local breeding strategies. As we will see later
adaptation) (Ceccarelli, 1989). in the book, Participatory Plant Breeding
40 Plant breeding and farmer participation
makes it easier to implement the first and Holland, Nyquist and Cervantes-Martinez,
the third of these strategies. 2003): H = h2b = G2/ P2. Broad-sense
heritability can be estimated from standard
2.2.6 Heritability analysis of variances. For example, in the
Heritability is the relative importance case of genotypes evaluated across several
of genetic and non-genetic factors in the environments, the corresponding analysis
expression of phenotypic differences of variance and heritability estimate can be
among genotypes in a population (Fehr, illustrated in Table 2.4.
1987). There are two basic types of herit- Genotypic and phenotypic variance
ability: broad-sense heritability and nar- components can also be estimated using
row-sense heritability (Holland, Nyquist Restricted Maximum Likelihood meth-
and Cervantes-Martinez, 2003; Nyquist, ods (Holland, Nyquist and Cervantes-
1991). Martinez, 2003).
Heritability in the broad sense (H) is Heritability in the narrow sense (h2)
the proportion of the phenotypic vari- is the proportion of phenotypic variance
ance of family means that is due to all among individuals in a population that is
genetic effects (Falconer and Mackay, 1996; due to heritable genetic effects (Nyquist,
TABLE 2.4
Analysis of variance and broad sense heritability estimates in the case of a series of g genotypes
evaluated across e environments
Source of variation Degrees of freedom Mean Squares Expected Mean Squares
Error (g-1)(r-1)e
MSE e2
Notes:
Total phenotypic variance: Var (Yijk ) = P2 = e2 + GE
2
+ G2
e2 GE
2
MS G
Phenotypic variance of genotypic means: Var (Yijk ) = P2 = + + G2 =
re e re
Genotypic variance = G2 = ( MS G MS GE ) / re
G2
Heritability on individual experimental unit basis: H i = hbi2 =
+ GE
2
G
2
+ e2
TABLE 2.5
Relatives, their covariance and regression or selection), faster, and in a low number of
correlation values in terms of narrow sense environments. In contrast, traits with low
heritability heritabilities require selection on a family
Relatives Covariance Regression (b) basis and in a greater number of environ-
(w/o epistasis) or correlation
(t) ments to determine breeding values of gen-
otypes. Heritability estimates for the same
Parent Offspring 2A b = h2
trait are variable (i.e. heritability of a trait
Midparent - Offspring 2A b = h2 is not a fixed value) and their magnitude
Half-sibs 2A t = h2 depends on several factors (Fehr, 1987):
Environment: it is important to have
Full-sibs 2A+ 2D t h2
adequate samples of environments from
the target population of environments.
TABLE 2.6
Additive and non-additive variance components In addition, estimates for genetic variance
in different generations should be free of GE variance.
Generation A2 D2 AD2 E2 Reference population: the amount of
P1 0 0 0 1 genetic variation and inbreeding present
P2 0 0 0 1 in the population affects heritability
F1 0 0 0 1 estimates. Higher inbreeding levels are
F2 1 1 0 1 associated with higher genetic variances
BC11 1 -1 1
and therefore with higher estimates of
BC12 1 1 1
Notes: A2 = (2 2F2 - 2BC11 - 2BC12);
heritability.
D2 = ( 2BC11 + 2BC12 - 2F2 - 2P1/P2/F1); Sample of genotypes evaluated: genotypes
AD = ( 2BC12 - 2BC11)
used to estimate heritabilities in one
population should be chosen at random.
1991; Holland, Nyquist and Cervantes- If the sample is not a representative
Martinez, 2003): h2n = A2/ P2. Narrow- random sample (e.g. selected genotypes),
sense heritability can be estimated from the ratio between genetic and phenotypic
variance components or from parent-off- variation is called Repeatability (Fehr,
spring regression. In both cases, genetic 1987). Repeatability estimates in a single
relationships among relatives (lineal (par- environment provide a measure of how
ent-offspring) or collateral (full- or half- much of the variation is genetic and
sibs)) are used (Table 2.5). The estima- therefore is a measure of the degree of
tion of additive and non-additive variance precision of data and the ability to detect
components is conducted through linear significant differences among genotypes.
models and proper mating designs (e.g. Method of estimation: heritability of a
North Carolina I, II and III, Hallauer and quantitative trait can be computed by
Miranda, 1988) or using information from several methods (variance components,
different generations (Kearsey and Pooni, parent-offspring regression, etc.) and her-
1998) (Table 2.6). itability estimates can differ among them
Heritability is used to estimate expected (e.g. heritability on a family basis is great-
response to selection and to choose the best er than on a plant basis; see Table 2.4).
breeding approach to improve the target Heritability estimates calculated on the
trait(s). Traits with high heritabilities can basis of selection unit are preferred to
be selected on a single-plant basis (e.g. mass estimated expected response to selection.
42 Plant breeding and farmer participation
Generation or progenies: different prog- genetic gain formula can be expressed as:
enies exploit different proportions of
additive and dominance variances. Inbred ic G2 ic G2
G = =
progenies have greater heritabilities than y P2 y 2 2
e
+ GE + G2
full-sib and half-sib families. re e
Allele frequencies: heritability is affect-
ed by allele frequencies. Therefore any where i = standardized selection differen-
change in allele frequencies (selection, tial, c = parental control (see below), y =
genetic drift, mutation, migration) could seasons per cycle, r = number of replica-
change heritability values for the same tions per environment, e = number of
2
trait and reference population. environments, and GE 2
, GE
2
and e are as
Heritability of a trait can be estimated defined earlier. The variance components
using the amount of genetic gain that is are estimated from the analysis of vari-
realized by selection within a population ance or mixed model solutions (Hallauer
(Falconer and Mackay, 1996). This is known and Miranda, 1988; Holland, Nyquist and
as realized heritability and can be estimates Cervantes-Martinez, 2003).
a posteriori as : h2=R/S, where R = response Increased heritability results in more
to selection and S = effective selection dif- effective selection. In the genetic gain
ferential applied in selection. formula, the value of the numerator can
be increased by increasing the selection
2.2.7 Response to selection intensity, parental control or genetic vari-
The theoretical response to selection can be ance. The value of the denominator can
defined as G = S h2, where S is the selec- be decreased by decreasing the number
tion differential (the difference between of seasons per cycle or the phenotypic
the mean of the selected individuals and variance. Response to selection is greater
the mean of the whole population) and h2 when a lower proportion of individuals or
the heritability of the target trait(s). S is families are selected. However, by decreas-
determined by the intensity of selection (i), ing the proportion selected (e.g. increasing
which is the number of genotypes selected selection intensity) the effective population
relative to the total number under evalua- size is reduced, thus increasing the possible
tion. Intensity of selection is the standard- occurrence of genetic drift or inbreeding.
ized selection differential: Desirable genes can be lost due to genetic
S drift. Because these effects can be detri-
i
P mental for future gains, it is recommended
to select a proportion of at least 20 percent.
where P is the square root of the pheno- Parental control, c, can be increased by
typic variance (Figure 2.4) recombining genotypes where both sources
Alternative selection methods differ in of gametes originated in selected genotypes
the types of progenies evaluated and recom- (c = 1). If the male gametes are coming from
bined, and the seasons required per cycle. unselected genotypes, then c is 0.5. Hence,
Hence, Eberhart (1970) incorporated the selection before pollination, where only
number of years (y) and the parental con- selected genotypes contribute to the next
trol (c) into the prediction formula. After generation, is more effective than selection
additional elaboration, a general expected after pollination, where non-selected geno-
Theory and application of plant breeding for quantitative traits 43
FIGURE 2.4
Schematic representation of selection differential and response to selection
Offspring
R
p = Sel/Pop
Parent
S
types can contribute to the next generation. the number of environments affects selec-
Further increase of c can be accomplished tion response more than increasing the
by inter-mating selfed progenies of selected number of replications. However, the envi-
genotypes (c = 2). Genetic variance can be ronments have to be representative of the
increased by selecting parents with high target area because if the environments are
genetic diversity, increasing the degree of very different, GE
2
can increase substantially
inbreeding before evaluation, increasing the and thereby reduce selection gain. This is
recombination between cycles or using dif- the case when, for example, the relative
ferent types of progenies. Different types of importance of stress factors differs between
progenies expressand therefore exploit selection and target environment (Betrn,
different proportion of genetic variances. Bnziger and Menz, 2004; Cooper et al.,
The theoretical proportion of A2 of total 2006).
G2 is 0.25 for half-sib progenies, 0.5 for Proper field experimental design and
full sib progenies and 1.0 for S1 progenies. analysis (see also Chapters 3 and 20 in
In addition to the type of progeny, differ- this volume), field stratification, number of
ent populations have different proportions plants measured, uniform soils and treat-
of genetic variance components, and there- ments, and reliable data collection and
fore different gains can be observed and processing reduce the error term and subse-
expected. quently increase the gain. Several field exper-
The number of seasons required to com- imental designs are used in plant breeding.
plete a cycle can be reduced by using The most common are randomized com-
off-season nurseries at lower latitudes or plete block, incomplete lattices and, more
in the opposite hemisphere, or by using recently, row-and-column designs among
greenhouses and growth chambers. The replicated designs, and augmented designs
phenotypic variance can be decreased by among unreplicated designs. Optimal
increasing the number of replications and experimental designs and field layout are
environments, or using improved statistical useful in reducing the error, increasing
design and analysis techniques that reduce the precision of mean estimates, and con-
the error variance and more accurately sequently improving selection (Gilmour,
estimate progeny performance. Increasing Cullis and Verbyla, 1997). Data processing
44 Plant breeding and farmer participation
and analysis has improved in recent years, (e.g. reduced anthesissilking interval under
with more powerful computer hardware drought is associated with increased yield in
and software addressing GE interaction maize (Edmeades, Bolaos and Chapman,
of multi-environment trials, in-trial spatial 1997)). Correlations among traits can be
variation, correlation among traits, vari- due to pleiotropy (same loci affect both
ance components estimation, prediction of traits), linkage/linkage disequilibrium (dif-
genetic parameters, QTL mapping, MAS, ferent loci affect the traits but these loci are
etc. (van Eeuwijk et al., 2005; Romagosa, linked together), or environmental effects.
van Eeuwijk and Thomas, 2008). Linkage in coupling will cause positive cor-
The use of artificial inoculation with relation and negative correlation in repul-
pathogens and infestation with major pests sion. If the environment affects both traits
has increased display of genetic variation in the same way (e.g. plant height and grain
and the heritability and gain for host plant yield in maize), it can create a positive cor-
resistance. Similarly, the use of managed relation. Different types of correlations
drought, nitrogen or low-pH stress envi- can be calculated: phenotypic correlations
ronments has increased heritability and (rP) are calculated using phenotypic values;
selection gain for abiotic stress tolerance genotypic correlations (rG) are calculated
(Bnziger and Cooper, 2001). using genotypic values; and additive genetic
Recent developments in biotechnology correlations (rA) (also known as genetic cor-
mean that effectiveness of selection can be relations) are calculated using breeding val-
increased by the use of molecular tools. For ues (Falconer and Mackay, 1996). Genetic
example, molecular markers can be used correlations among traits can change with
to: (1) increase selection intensity while selection as a consequence of change in
selecting genotypes with different genetic allelic frequencies.
backgrounds to maintain genetic variance; If the direct selection response for trait
(2) select before pollination genotypes with X is RX = ihX AX , the correlated response in
the desired allele composition for markers trait Y is defined as CRY = ihX hy rA PY , where
associated with traits of interest; (3) con- hx and hy are the square roots of herit-
duct selection in environments not repre- abilities for traits X and Y, respectively; rA
sentative of the target area (e.g. off-season is the genetic correlation between traits X
nurseries); and (4) conduct MAS per se or and Y; i is the selection intensity; and P is
in combination with phenotypic selection. the phenotypic variance. Indirect selection
for trait Y can improve trait X. Therefore
2.2.8 Correlated response to selection the relative efficiency or merit of indirect
and indirect selection selection can be compared with the direct
Several relevant traits are often considered selection for trait X as follows:
simultaneously in plant breeding, particu- CR X iY hY rA AX iY hY rA
larly when selection is done by farmers. The = =
RX i X hX AX i X hX
relationship among them determines breed-
ing strategies and response to selection. The (Falconer and Mackay, 1996)
association among desirable traits can be Indirect selection can be more effec-
negative (e.g. increasing grain yield is asso- tive than direct selection when this ratio
ciated with lower protein content in maize is >1 in cases where secondary traits show
(Duvick and Cassman, 1999)), or positive greater heritabilities than the primary trait
Theory and application of plant breeding for quantitative traits 45
and high correlations between secondary 2.3 KEY EXPERIMENTS FOR RESPONSE
and target traits are present, or if greater TO SELECTION
selection intensities can be applied to the Selection experiments have been designed
secondary trait (e.g. easier to screen in to make the improvement of quantitative
big populations). If selection intensity traits more efficient, trying to maximize
in both traits were considered the same, as much as possible the additive effects, as
indirect selection would be superior to well as to gather genes with complementary
direct selection when rAhy is greater than dominant and epistatic effects in genotype
hx. Ideally, a secondary trait should be crosses and synthetic varieties. There are
associated genetically with the target many examples that show the efficacy of
trait, highly heritable, easy and fast to these experiments. Below, we discuss the
measure, non-destructive, stable over the results of some selection experiments for
measurement period, and/or observable complex traits published in the last years,
at or before flowering so that undesir- which supposedly reveal the genetic archi-
able parents are not crossed (Edmeades, tecture of quantitative traits. Experiments
Bolaos and Chapman, 1997). are divided into short-term, mid-term and
Selection indices combine informa- long-term recurrent selection experiments.
tion from different traits with the goal We are focusing this section on recurrent
of selecting genotypes with the highest selection experiments because they reveal
aggregate breeding value. Many breeders better than any other kind of selection
and especially farmers have an ideotype method the role of the hidden genetic fac-
in mind when applying selection, which tors responsible for performance and behav-
is a rather subjective application of a iour of quantitative traits. Experiments that
selection index. More objective selec- combine selection and subsequent recom-
tion indices are linear combinations of bination of selected genotypes permit the
observable trait values that maximize best exploration of the intricate assortment
the expected genetic gain in an aggregate of both major genes and genes with small
breeding value. Baker (1986) and Lin effect for the trait studied.
(1978) have reviewed the theory of selec- Recurrent or cyclic selection is generally
tion indices and their application to plant applied to a population, where loci respon-
breeding. sible for a quantitative trait are segregating
MAS (i.e. selection of genotypes based with different gene effect and at different
on molecular markers associated with allele frequency, except for the case of an
target traits) is also a form of indirect F2 population derived from the cross of
selection. MAS for quantitative traits two inbred lines, in which allele frequen-
can be used in situations where pheno- cies for all segregating loci are 0.5. The aim
typic selection can not be conducted of recurrent selection programmes is to
(e.g. off-season nurseries), when target increase progressively, cycle after cycle, the
traits have low heritability and there are frequency of the favourable alleles respon-
tight linkage between QTLs and mark- sible for the performance of the trait under
ers, and a high proportion of additive selection (see later Chapters). The gen-
variance is explained by the markers eral scheme of each selection cycle involves
(Lande and Thompson, 1990; Hospital three steps: (1) creation of a family struc-
and Charcosset, 1997). ture from selection units; (2) evaluation
46 Plant breeding and farmer participation
of familial test units in replicated trials in criteria: (1) enough information is available
different environments, i.e. progeny test- to rely on the accuracy of results concern-
ing; and (3) recombination of genotypes ing genetic improvement; (2) comparison
related to the families selected (recombina- among different selection methods is pos-
tion units) in the evaluation trial (Moreno- sible; and (3) additional information is pro-
Gonzlez and Cubero, 1993). vided to help us to interpret the structure of
The above general scheme may be modi- genetic and environmental factors involved
fied for different methods of selection, in complex traits. Short-term, mid-term
increasing or reducing the number of gen- and long-term selections are arbitrarily
erations. For example, the three steps are designed here as those experiments that
combined in one generation per cycle in have undergone <6, between 6 and 20, and
the recurrent mass selection method, and in >20 selection cycles, respectively.
two generations for the half-sib or full-sib
family selection method as first introduced 2.3.1 Short-term recurrent selection
by Vilmorin in 1856 for sugar beet. In the experiments. The case of the BS11
case of mass selection, the selection units population
are individual plants, rather than families. A maize population, BS11 from Iowa State
In the case of half-sib and full-sib selection, University, USA, has been studied for sev-
creation of new families and recombination eral factors affecting the process of selection
of selected families is the same operation, for complex traits in short-term recurrent
thus both steps are done in the same sea- selection experiments (Weyhrich, Lamkey
son, while family evaluation is carried out and Hallauer, 1998a, b; Guzman and
on the next generation. In contrast, any of Lamkey, 1999, 2000). These authors focus
the basic steps may involve more than one on two useful concepts for plant breeders,
generation. For example, (1) creation of a which are worth comment: (1) comparisons
family structure in the S2 family selection of the response of the same population to
method requires two generations; (2) evalu- different intra-population and inter-popu-
ation trials may be conducted during more lation methods of selection; and (2) the role
than one year either to reduce the effect of of the effective population size in the selec-
genotype year interaction or to conduct tion response and the genetic variability of
selection in multiple stages, such as geno- the population.
types selected at low selection intensity in Six intra-population methodsfull-
the first year may be re-evaluated at high- sib, half-sib, mass, modified ear-to-row, S1
er selection intensities in following years progeny and S2 progeny selectionsand
(Piper and Fehr, 1987); and (3) selected one inter-population method, reciprocal
genotypes may be recombined during more full-sib recurrent selection (RFSRS), were
than one generation to break up repulsion applied to the same BS11 maize population
linkages among favourable alleles. and compared for grain yield and other
Many recurrent selection experiments traits during five selection cycles, except
have been reported during the last three or for half-sib selection, for which only four
four decades in the literature, which show cycles were completed (Weyhrich, Lamkey
genetic improvement for the selected trait and Hallauer, 1998a). General descriptions
in several crops. We choose here to discuss of the above and other schemes of selec-
some of these experiments based on three tion methods may be found in Moreno-
Theory and application of plant breeding for quantitative traits 47
TABLE 2.7
Grain yield genetic gains (%) averaged over cycles for different recurrent selection methods in the
maize population BS11
Selection method Selection cycles (No.) Genetic gain per cycle (%)
Averaged over population
Population per se
testcrosses
Full-sib 5 1.4 1.6
Half-sib 4 1.6 2.1
Mass 5 0.6 0.5
Modified ear-to-row 5 3.6 2.7
RFSRS 5 2.6 2.6
S1-progeny 5 1.9 1.6
S2-progeny 5 4.5 3.3
Notes: = reciprocal full-sib recurrent selection.
Source: adapted from Weyhrich, Lamkey and Hallauer, 1998a.
48 Plant breeding and farmer participation
TABLE 2.8
Genetic gains (%) averaged over cycles for different recurrent selection methods in the populations
per se and population crosses, and the change in genetic variance through selection cycles
Crop and Trait Selection Selected Selection Genetic gain Change References
method source cycles (No.) per cycle (%) in genetic
population Population Averaged variance
per se population over cycles
crosses
Oat yield F4.6 lines Univ. 7 2.2 No change De Koeyer and
Minnesota Stuthman, 1998
Spring wheat F3 lines Univ. North 8 4.5 No change Wiersma et al., 2001
kernel weight Dakota
Maize yield RFSRS BS10 5 3.2 2.5 Hallauer, 1984
RFSRS BS11 5 2.9 2.5
Maize yield RFSRS BS10 7 2.0 Rodriguez and
RFSRS BS11 7 0.8 Hallauer, 1988
Maize yield RFSRS BS10 8 3.0 6.5 Eyherabide and
RFSRS BS11 8 1.6 6.5 Hallauer, 1991
Maize yield RFSRS BS10 10 2.7 1.6 No change Frank and Hallauer,
RFSRS BS11 10 2.3 1.6 No change 1999
and crops is shown in Table 2.8. Most of to St. Paul, Minnesota, United States of
the experiments are for maize but two are America, after the fourth cycle (Wiersma et
reported for other crops. al., 2001). The selection scheme was similar
to that described above for oat, but the
Self-fertilized crop species number of evaluated and selected lines, the
A recurrent selection programme to increase number of inter-crosses and the generation
grain yield in oat has been carried out at the used for the derived lines were different.
University of Minnesota, United States of Kernel size increased linearly at 4.5 percent
America, since 1968. The selection method per cycle over eight cycles (Table 2.8), with
comprises three steps: (1) selection of 21 F4.6 an indirect increase in flour yield. Results
lines out of 630 tested; (2) creation of 63 F1 indicated no clear trend towards a decrease
single crosses by intercrossing the 21 selected in genetic variance. Results suggested that
lines, each with another six different lines; the trait is controlled by several genes with
and (3) derivation of 10 lines from each F4 small effects (Wiersma et al., 2001).
(De Koeyer and Stuthman, 1998). The linear The two above experiments are examples
regression response to selection over seven of the efficiency of recurrent selection for
cycles was 2.2 percent per cycle relative complex traits in self-fertilized crop spe-
to the original population (Table 2.8). The cies. Other reports of recurrent selection
results indicated that there has not been sig- in autogamous plant species have also been
nificant change in the estimates of the genetic published. Nine cycles of recurrent selec-
variance through the seven selection cycles tion for groat-oil content in oat produced
(De Koeyer and Stuthman, 1998). a linear increase in the groat (caryopsis) oil
A recurrent selection programme to content of oat at a rate of 6.5 percent per
increase kernel weight in spring wheat was cycle and non-decrease in the genetic vari-
initiated at Fargo, North Dakota, United ation (Frey and Holland, 1999). It seems
States of America, in 1967, and moved that additive effects were predominant in
Theory and application of plant breeding for quantitative traits 49
self-fertilized crops. Soybean and cotton variance and predicted genetic gain than
populations have also undergone recurrent the other populations involved in the study
selections with similar schemes to those (Russell, Blackburn and Lamkey, 1992).
described above (Piper and Feher, 1987; Inbred A632 was the tester of selected
Miller and Rawlings, 1967). population BS21(A632HI). It seems that
the elite inbred H99 might have masked
Inbred versus population tester in dominant, favourable alleles present in the
reciprocal recurrent selection experiments BS22 population. Comstock (1979), using
Reciprocal recurrent selection (RRS), first quantitative genetics theory, compared
designed by Comstock, Robinson and population improvement with both types
Harvey (1949), tries to alter two different of testers (i.e. inbred vs. population) based
genetically populations to improve their on change in allele frequency. He con-
cross mean. The original method consists cluded that the inbred line tester was not
of the following steps: (1) individual plants superior to the reciprocal population tester.
from two populations, A and B, are selfed Furthermore, the population tester might
and at the same time crossed to 3 to 5 be superior to the inbred tester in some
random plants from the reciprocal female situations, especially if overdominance and
tester population, B and A, respectively; multiple peak epistasis are present in the
(2) selection in each population is based on populations. Likewise, Moreno-Gonzlez
the performance of the testcross half-sib and Grossman (1976) demonstrated that
families; (3) remnant seed from the selected the theoretical genetic gain of the popula-
S1 families are mated at random within A tion cross was higher when a low-yielding
and B to form new cycles of the A and B population (i.e. smaller allele frequencies
populations. Russell and Eberhart (1975) for the segregating loci) was used as popu-
proposed a modification of RRS (MRRS), lation tester.
suggesting to use as tester of population A The expected change in allele frequency
an inbred line derived from or related to B, in the selected population (pA) (Moreno-
instead of the population B itself; recipro- Gonzlez and Grossman, 1976; Falconer,
cally, the tester of B should be an inbred 1981) will be
line derived from or related to A.
ip A (1 p A )[a + (1 2 pT )d ]
A programme to compare MRRS and p A =
2 p
RRS using the maize populations BS21 and
BS22 was initiated in the maize breeding where i is the selection intensity; pA and
programme at Iowa State University in 1975 pT are the frequencies of more favourable
(Russell, Blackburn and Lamkey, 1992). allele in the selected and tester populations,
After three cycles of selection, the popula- respectively; a and d are the additive and
tions per se and the population cross of the dominance effects; and p is the phenotypic
MRRS method showed less genetic response standard deviation of testcross or half-
than populations selected under the RRS sib families. The numerator of the above
method (Russell, Blackburn and Lamkey, expression is expected to have the same
1992). It should be specially noted that the value both when the tester is the recipro-
improved population BS22(H99HI), which cal population B and when it is an inbred
uses inbred H99 as tester in the MRRS line randomly derived from B; however, p
method, had less genetic response, genetic is larger when the tester is an inbred line
50 Plant breeding and farmer participation
(Comstock, 1979). Thus pA is not expect- ulation hybrid (Table 2.8; Hallauer, 1984;
ed to be higher for the inbred tester than for Rodriguez and Hallauer, 1988; Eyherabide
the reciprocal population tester. and Hallauer, 1991; Frank and Hallauer,
The same populations, BS21 and BS22, 1999). The direct selection response for
and breeding selection methods, RRS grain yield per cycle in the cross between
and MRRS, were also evaluated after the the two populations was significant for all
sixth cycle (Menz Rademacher, Hallauer reported studies of evaluation, but varied
and Russell, 1999). Results from this study among the different studies, being 2.5, 6.5
were essentially similar to that of Russell, and 1.6 percent per cycle when the first
Blackburn and Lamkey (1992) for cycle 5, 8 and 10 selection cycles were evalu-
three. The grain yield response of the popu- ated, respectively. Sampling of populations
lation cross, BS21 BS22, was higher for and different years and environments of
RRS (4.4 percent per cycle) than for MRRS evaluation might account for these differ-
(1.6 percent per cycle), and also for the test- ences. The indirect selection response of
cross direct responses of the MRRS method the populations per se was similar in the
(2.8 and 1.6 percent per cycle for BS21 four studies reported (Table 2.8), and it was
A632 and BS22 H99, respectively) (Menz consistently higher in BS10 than in BS11.
Rademacher, Hallauer and Russell, 1999). It seems that selection for the population
It seems that the MRRS was less efficient hybrid increased the allele frequency of
than RRS for increasing the population cross favourable alleles in BS10 itself at a higher
BS21 BS22. Efficiency of MRRS depends rate than in BS11. No significant differ-
on the choice of the tester, which is related to ences between cycle 0 and 10 were found
the type of gene action involved in the com- for the grain yield genetic variances of
plementary alleles between the tester and the the populations per se and the population
selection populations. If a favourable allele hybrid (Frank and Hallauer, 1999). This
with complete dominance is fixed in the suggests that further response to selection
inbred tester, frequency of this allele is not should be expected in the next cycles of
expected to increase in the selection popula- RFSRS applied to BS10 and BS11.
tion, whereas this frequency will increase if
the allele is segregating in a population used Over fifty years of reciprocal recurrent
as tester. Thus, a limitation exists in terms selection experiments
of increasing the frequency of dominant A RRS programme has been conducted by
favourable alleles in the selection population the Cooperative Federal-State maize breed-
when they are fixed in the inbred tester. ing programme at Iowa State University
with the synthetic maize populations BSSS
Reciprocal full-sib recurrent selection and BSCB1 since 1949 (Keeratinijakal and
experiments Lamkey, 1993a), the year when the RRS
The RFSRS method was designed for maize method was first published (Comstock,
yield selection by Hallauer and Eberhart Robinson and Harvey, 1949). Several
(1970), and has been applied to the BS10 reports on the genetic improvement in
and BS11 maize populations at Iowa State grain yield and other traits have been
University. This method has proven to be published through all the history of the
very efficient in increasing the genetic gain RRS programme (Eberhart, Debela and
of both the populations per se and the pop- Hallauer, 1973; Smith, 1983; Oyervides-
Theory and application of plant breeding for quantitative traits 51
TABLE 2.9
Average genetic gains (%) of grain yield over cycles for the reciprocal recurrent selection (RRS)
method applied to the maize populations BSSS and BSCB1 in the populations per se, population
cross and population topcrosses
Selected Cycle No. Genetic gain per cycle (%) Change in References
population genetic variances
Population Population Population
per se cross topcross
BSSS 5 -0.1 4.1 1.2 Eberhart, Debela and
BSCB1 5 1.0 4.1 0.3 Hallauer, 1973
BSSS 7 2.2 4.3 Smith, 1983
BSCB1 7 0.7 4.3
BSSS 8 1.9 1.4 Oyervides-Garcia and
Hallauer, 1986
BSSS 10 1.3 Rodriguez and
BSCB1 10 -1.6 Hallauer, 1988
BSSS 9 6.1 Increase in additive Betrn and Hallauer,
BSCB1 9 6.1 and no change 1996a, b)
in dominance
genetic variances of
population cross
BSSS 11 1.7 6.9 2.8 Keeratinijakal and
BSCB1 11 1.9 6.9 3.9 Lamkey, 1993a
BSSS 11 2.6 No change in additive Holthaus and Lamkey,
and reduction in 1995a
dominance variance
Source: adapted from references.
Garcia and Hallauer 1986; Rodriguez Part of the lower selection response in
and Hallauer, 1988; Helms, Hallauer and the populations per se compared to the
Smith, 1989; Keeratinijakal and Lamkey, population cross can be attributed to the
1993a,b; Holthaus and Lamkey, 1995a,b; genetic drift caused by restricted effec-
Betrn and Hallauer, 1996a,b). Some of tive population size during the process of
the results for grain yield are summarized selection. The inbreeding coefficient pro-
in Table 2.9. The direct response to selec- gressively increases in the populations per
tion in the population cross was reported se when the number of selected lines used
to be very effective in four independ- for recombination has been small (i.e. <20)
ent studies, 4.1 percent per cycle (Penny during several selection cycles. The expect-
and Eberhart, 1971), 4.3 percent per cycle ed inbreeding coefficient of a diploid popu-
(Smith, 1983), 6.1 percent per cycle (Betrn lation reproduced with a finite number of
and Hallauer, 1996a) and 6.9 percent per selected individuals during t generations
cycle (Keeratinijakal and Lamkey, 1993a), can be estimated as:
when the first 5, 7, 9 and 11 selection cycles i t 1
were evaluated, respectively. However the Ft 1 (1 F0 ) (1 )
i 1 2Ni
observed indirect response to selection was
much smaller in the populations per se than where F0 and Ft are the inbreeding coef-
in the population cross, ranging from -1.6 ficients of the original population and after
percent per cycle in BSCB1 (Rodrigez and t generations of selection, respectively; and
Hallauer, 1988) to 2.6 percent per cycle in Ni is the number of individuals selected at
BSSS (Holthaus and Lamkey, 1995a). generation i.
52 Plant breeding and farmer participation
If F0 = 0, and Ni is constant (i.e. N) for the populations BSSS and BSCB1 after
all generations, the above equation becomes 12 cycles of RRS (Labate, Lamkey and
the known formula: Woodman, 1999). Allele frequency changes
1 t in 28 loci, out of 82, could not be explained
Ft = 1 (1 ) by genetic drift, thus it should be attributed
2N
to selection. The within-population expect-
In this case, if N =10 and t = 14, then F14 = ed herezogosity decreased, while the inter-
0.512, which is larger than after one genera- population component of genetic variation
tion of selfing. Thus, inbreeding depression increased (Labate, Lamkey and Woodman,
may become evident in these selected popu- 1999). Also, another study was conducted
lations if dominance effects controlling the to investigate the genetic variation of pro-
trait under selection (i.e. grain yield) were genitor inbred lines used to synthesize pop-
important. When responses were adjusted ulations BSSS and BSCB1, as well as elite
for effects of genetic drift, the improve- inbred lines derived from advanced cycles
ments of the populations per se were similar of selection (Hagdorn et al., 2003). A larger
to those of the population cross in the study genetic distance was found between the
of Keeratinijakal and Lamkey (1993b), and BSSS and BSCB1 groups of lines derived
larger than in Smiths study (1983). In addi- from advanced cycles than between the
tion, additive and dominance effects were group of lines derived from cycle 0. Thus,
found in BSSS, but only dominance effects these studies confirm the success of RRS in
were important in BSCB1. The presence of increasing the genetic diversity of the two
important dominance effects may explain populations, which was one of the objec-
the inbreeding depression of the popula- tives of the method. Also, these results
tions per se. reinforce the hypothesis of the complemen-
For grain yield, the additive genetic vari- tary effects of heterozygous genes in the
ance of population BSSS did not decrease population cross, and the hybrids between
after 11 cycles of RRS, whereas the domi- elite inbred lines derived from the two
nance genetic variance was reduced, and reciprocal populations.
the heritability estimates increased in BSSS
when cycle 11 was compared to cycle 0 2.3.3 Long-term recurrent selection
(Holtaus and Lamkey, 1995a). Likewise for experiments
grain yield, the additive genetic variance Mass selection experiment for ear length in
of the population cross BSSS BSCB1 maize
increased after nine cycles of RRS, with A mass selection programme started at
no reduction in the dominance genetic Iowa State University in 1963 to select for
variance, and an increase in the heritabil- divergent ear length (i.e. short and long
ity estimates (Betrn and Hallauer, 1996a). ears) in the maize population BSLE (Lpez-
Therefore, it seems that the response of Reynoso and Hallauer, 1998). A modi-
these populations to RRS will continue in fied scheme of mass selection proposed by
the next cycles of selection. Gardner (1961), called grid selection, was
Restriction fragment length polymor- used in this experiment. The scheme sub-
phism (RFLP) loci were used to deter- divides the selection field into 100 plots,
mine changes in allele frequency, expected each with 40 plants, among which three
heterozygosity, and genetic variation in plants per plot (i.e. 7.5 percent selection
Theory and application of plant breeding for quantitative traits 53
pressure) were selected for long and short An increase in number of ears per plant
ear. Each selection cycle was completed in brought a decrease in other ear traits affect-
one year. The linear selection response was ing yield, such as ear length, ear diameter
1.4 percent per cycle for longer ears, and and kernel size, revealing that grain yield
1.9 percent per cycle for shorter ears over is a very complex trait and individual selec-
27 selection cycles. The rate of the inbreed- tion for one of the yield component does
ing depression in selfed populations rela- not necessarily have a significant effect on
tive to unselfed was about 18 percent for the yield trait.
longer ears and remained constant during
the process of selection, whereas this rate Over a century of an Illinois long-term
gradually reduced for shorter ears as gen- selection experiment for oil and protein
erations advanced, from 18.2 percent in the content in the maize kernel
original population up to 2.9 percent in the The Illinois long-term selection experiment
24th cycle (Lpez-Reynoso and Hallauer, begun in 1896, analysing percentage oil and
1998). If the genetic effects of some loci are protein in 163 ears of the maize open-polli-
assumed to be dominant for longer ears, nated variety Burrs White (Dudley, 1977).
then the recessive alleles of these loci would The 24 ears highest in protein, the 12 ears
be more easily fixed for shorter ears in the lowest in protein, the 24 ears highest in oil,
course of selection. Thus, no alleles with and the 12 ears lowest in oil were selected
dominant effects would be segregating in to initiate the Illinois high protein (IHP),
the selected population for short ears after Illinois low protein (ILP), Illinois high oil
cycle 24, which would explain the absence (IHO) and Illinois low oil (ILO) strains,
of inbreeding depression in the short-ear respectively (Dudley and Lambert, 1992).
population in comparison with the long-ear Results and studies of this experiment have
population. Genetic variance for ear length been published in many reports (Leng,
still remained in both populations after 1962; Dudley and Lambert, 1969; Dudley,
selection cycle 24, although the heritability Lambert and Alexander, 1974; Moreno-
of the trait was reduced. Gonzlez, Dudley and Lambert, 1975;
Dudley, 1977; Dudley, Lambert and de la
Mass selection experiment for prolificacy Roche, 1977; Dudley and Lambert, 1992;
in maize Dudley, 1994; Sughroue and Rochefort,
A mass selection programme was initiated 1994; Moose, Dudley and Rochefort, 2004;
at the University of Wisconsin-Madison in Laurie et al., 2004; Dudley and Lambert,
1971 to select for prolificacy (i.e. increase 2004; Clark et al., 2006; Dudley, 2007, 2008).
in number of ears per plant) in the maize So far, 106 cycles of recurrent selection have
open-pollinated population GG(MP) (de been carried out on these strains (Dudley,
Leon and Coors, 2002). The selection 2007). The schemes of the selection meth-
response of the programme was an increase ods were slightly modified through genera-
of 0.14 and 0.03 ears per plant per cycle at tions. Mass selection was used during gen-
low and high plant densities (15 000 and erations 09; ear-to row during generations
73 000 plants/ha), respectively, during 24 1025; and half-sib selection for percent oil
selection cycles. Most of the response was and mass selection for percent protein dur-
achieved between cycles 18 and 24. Indirect ing generations 26 to date (Dudley, 1977;
responses were also sensitive to selection. Dudley and Lambert, 1992).
54 Plant breeding and farmer participation
This experiment reveals the power of ods. Genetic variance still remains in the
long-term recurrent selection for achieving IHO and IHP strains, because selection
progressively significant genetic changes progress continued at the same rate in the
in quantitative traits, such as protein and two strains from generations 76 to 90. In
kernel oil content, through more than one addition, significant genetic variances were
hundred selection generations. Responses directly estimated in the strains in the 65th
to selection for the high and low strains of generation (Dudley and Lambert, 1969).
oil and protein in the maize kernel were Besides the potential breeding benefit
very important in the long run. Populations of developing maize strains for high and
changed from 4.6 percent oil to 19.6 percent low oil and protein, this experiment has
in IHO, and to 0.5 percent in ILO after 87 been a good tool to look into the intricate
generations. At generation 89, selection for complexity of the genetic architecture of a
ILO was discontinued because a biological quantitative trait. The experiment has also
limit had been reached in terms of maintain- been a test bench to check quantitative
ing seed viability of the ILO strain (Dudley, theory and to estimate some genetic param-
Lambert and Alexander, 1974; Dudley and eters otherwise difficult to compute. The
Lambert, 1992). For protein, populations question that arises is: What we can learn
changed from 10.9 percent to 32.5 percent from this long-term selection experiment?
in IHP and to 4.2 percent in ILP after 90 1. Recurrent selection was able to achieve a
generations. Progress in the ILO and ILP steady increase in the selection traits over
lines had ceased in the latter generations more than one hundred generations.
(Dudley, Lambert and Alexander, 1974; Mild selection (20 percent selection
Dudley and Lambert, 1992). In terms of pressure) and adequate effective
additive genetic standard deviations (a) of population size (>20) should have had
the populations, the genetic gains became a favourable effect in achieving this
huge, being 22 a for IHO and 26 a for selection progress. The experiment has
IHP (Dudley and Lambert, 1992). These been a text-book example of selection.
deviations are placed in the extreme tail of Recombination of selected genotypes
the normal distributions, with probabilities created new genotypes distributed
less than 10-105 and 10-145 for 22 a and 26 a, around new displaced means generation
respectively. Thus, assuming that recombi- after generation. Thus, recombination
nation of favourable loci could be gathered and selection were the key points for
at random in a unique genotype in one reaching this huge genetic gain (i.e. 22 a
generation, it would be necessary to grow and 26 a), which otherwise would not
an unimaginable number of maize plants be possible to attain.
(10105 and 10145 plants) to find one plant 2. The genetic variance is not still
with the same oil and protein percentages exhausted. Reverse selection strains
as IHO and IHP, respectively. If no more (i.e. RHO, RLO, RHP, and RLP) and
than 1014 maize plants are currently grown switch-back selection from the reverse
in the world every year, it would be neces- strain (SHO) clearly indicate that loci
sary to wait for at least 1091 and 10131 years are still segregating in the populations
to find at random the IHO and IHP strains and selection can be conducted in the
that have been selected in only 100 years by desired direction at the same or higher
applying simple recurrent selection meth- rate as before.
Theory and application of plant breeding for quantitative traits 55
selection experiments. Likewise, heritabil- tion strains, and also may help explain
ity estimates were frequently no lower in the continued success of commercial corn
the most advanced selected populations. breeding.
Either the favourable alleles are hard to
fix, or mutant alleles with small effects 2.4 SUMMARY AND CONCLUSIONS
may naturally arise in the populations and We have described the evolution of selec-
subsequently may be captured during the tion and plant breeding theory, the con-
selection process. Small effective popula- ceptual basis of key components of plant
tion size is not an impediment for short- breeding approaches, and empirical exam-
term selection, but it might limit selection ples of selection. The contribution of plant
in the long term. Inbreeding depression breeding to the improvement of crop pro-
of selected population may be caused by duction and quality has been enormous.
small population size during the process of Critical in this contribution has been the
selection in traits where dominant favour- implementation of more efficient breeding
able alleles are important. The inbreeding approaches supported by developments in
depression disappears when the selected quantitative genetics. Plant breeding has
populations are crossed to others. evolved and will continue to evolve, adopt-
Epistasis is well documented and has ing classical and modern technologies (e.g.
been recognized as a not rare phenomenon family selection, recurrent selection, multi-
in Mendelian qualitative traits that are con- location evaluation, off-season nurseries,
trolled by two or three loci. Thus, epistasis biotechnology) to increase efficiency of
should be also expected in more complex selection, to adjust to new environmental
traits. Analysis of generation means has conditions and variable demands for crop
been frequently used to look for epistasis in utilization, and to implement sustainable
quantitative traits. The magnitude of epista- production systems. Plant breeders and
sis effects detected was small relative to geneticists will continue the search for and
the additive and dominance effects in most adoption of the more effective methods to
of the studies, although some traits and develop, identify and evaluate cultivars that
elite crosses showed up significant epista- can contribute to superior agronomic per-
sis (Moreno-Gonzlez and Dudley, 1981; formance and sustainable profit.
Melchinger, 1987; Melchinger, Schmidt
and Geiger, 1988; Lamkey, Schnicker and REFERENCES
Melchinger, 1995; Hinze and Lamkey, Alonso de Herrera, G. 1513. Agricultura
2003). Either the favourable and unfavour- General. Madrid.
able epistatic effects involving the trait can- Annicchiarico, P. 2002. Genotype
cel out through the genome, or the genetic Environment Interactions Challenges
models or the statistical methods used are and Opportunities for Plant Breeding and
not powerful enough to separate epistasis Cultivar Recommendations. FAO Plant
from other effects. Dudley (2008) analyzed Production and Protection Papers, No. 174.
500 S2 lines derived from the crosses IHO 126 p. (available at http://www.fao.org/
ILO and IHP ILP, using SNP molecular docrep/005/y4391e/y4391e00.htm; accessed
markers. He reported that epistasis could 11 Feb. 2009)
contribute to the long continued response Atlin, G.N., Cooper, M. & Bjrnstad, . 2001.
to selection in the Illinois long-term selec- A comparison of formal and participatory
Theory and application of plant breeding for quantitative traits 57
breeding approaches using selection theory. and specific combining ability. Agronomy
Euphytica, 122: 463475. Journal, 41: 360367.
Baker, R.J. 1986. Selection indices in plant Cooper, M., Podlich, D. & Smith, O.S. 2004.
breeding. Boca Raton, USA, CRC Press. Complex trait genetics and gene-to-phe-
Bnziger, M. & Cooper, M. 2001. Breeding notype models. In T. Fischer et al. New
for low input conditions and consequences directions for a diverse planet. Paper 3.2.
for participatory plant breeding: examples Proceedings of the 4th International Crop
from tropical maize and wheat. Euphytica, Science Congress, Brisbane, Australia, 26
122: 503519. September1 October 2004. Published on
Bernardo, R. 2002. Breeding for quantitative CD-ROM (available at: http://www.crop-
traits in plants. Woodbury, USA, Stemma science.org.au/icsc2004/).
Press. Cooper, M., van Eeuwijk, F.A., Chapman,
Betrn, F.J. & Hallauer, A.R. 1996a. S.C., Podlich D.W. & Lffler, C. 2006.
Characterization of inter-popula- Genotype-by-environment interactions
tion genetic variability in three hybrid under water-limited conditions. In J.-M.
maize populations. Journal of Heredity, Ribaut, ed. Drought adaptation in cereals,
87(4): 319328. pp. 5195. Binghampton, USA, Haworth
Betrn, F.J. & Hallauer, A.R. 1996b. Hybrid Press.
improvement after reciprocal recurrent Crow, J.F. & Kimura, M. 1970. An introduction
selection in BSSS and BSCB1 maize popu- to population genetics theory. New York,
lations. Maydica, 41: 2533. USA, Harper&Row.
Betrn, F.J., Bnziger, M. & Menz, M. 2004. Darrow, G.M. 1966. The strawberry: history,
Corn Breeding. In W. Smith, F.J. Betrn & breeding, and physiology. New York, USA,
E. Runge, eds. Corn: origin, history, tech- Holt, Rinehart and Winston.
nology, and production. New York, USA, De Koeyer, D.L. & Stuthman, D.D. 1998.
John Wiley & Sons. Continued response through seven cycles
Buckler, E.S. & Thornsberry, J. 2002. Plant of recurrent selection for grain yield in oat
molecular diversity and applications (Avena sativa). Euphytica, 104: 6772.
to genomics. Current Opinion in Plant De Leon, N. & Coors, J.G. 2002. Twenty-four
Biology, 5: 107111. cycles of mass selection for prolificacy in
Ceccarelli, S. 1989. Wide adaptation. How the Golden Glow maize population. Crop
wide? Euphytica, 40: 197205. Science, 42: 325333.
Clark, D., Dudley, J.W., Rochefort, T.R. & Doorenbos, J. 1954 Notes on the history
LeDeaux, J.R. 2006. Genetic analysis of of bulb breeding in the Netherlands.
corn kernel composition in the random Euphytica, 3: 118.
mated 10 generations of the cross of gen- Dudley, J.W. 1977. Seventy-six generations
erations 70 of IHO ILO. Crop Science, of selection for oil and protein percentage
46: 807819. in maize. In E. Pollack, O. Kempthorne
Comstock, R.E. 1979. Inbred lines versus the & T.B. Bailey Jr, eds. Proceedings of the
populations as testers in reciprocal recur- International Conference on Quantitative
rent selection. Crop Science, 19: 881887. Genetics, 1621 August 1976. Ames, USA,
Comstock, R.E., Robinson, H.F. & Harvey, Iowa State University Press.
P.H. 1949. A breeding procedure designed Dudley, J.W. 1982. Theory for transfer of alle-
to make maximum use of both general les. Crop Science, 22: 631637.
58 Plant breeding and farmer participation
Dudley, J.W. 1984. A method for identifying East, E.M. 1916. Studies on size inheritance in
populations containing alleles not present Nicotiana. Genetics, 1: 164176
in elite germplasm. Crop Science, 24: 1053 Eberhart, S.A. 1970. Factors affecting efficien-
1054. cies of breeding methods. African Soils,
Dudley, J.W. 1994. Linkage disequilibrium 15: 669680.
in crosses between Illinois maize strains Eberhart, S.A., Debela, S. & Hallauer, A.R.
divergently selected for protein percent- 1973. Reciprocal recurrent selection in the
ages. Theoretical and Applied Genetics, BSSS and BSCB1 maize populations and
87: 10161020. half-sib selection in BSSS. Crop Science,
Dudley, J.W. 2007. From means to QTL: the 13: 451456.
Illinois long-term selection experiment as Edmeades, G.O., Bolaos, J. & Chapman,
a case study in quantitative genetics. Crop S.C. 1997. Value of secondary traits in
Science, 47: S-20S-31. selecting for drought tolerance in tropical
Dudley, J.W. 2008. Epistatic interactions in maize. In G.O. Edmeades, M. Bnziger,
crosses of Illinois High Oil Illinois Low H.R. Mickelson & C.B. Pea-Valdivia, eds.
Oil and of Illinois High Protein Illinois Developing Drought and Low-N Tolerant
Low Protein corn strains. Crop Science, Maize. Proceedings of a Symposium, 2529
48: 5968. March 1996. CIMMYT, El Batan, Mxico.
Dudley, J.W. & Lambert, R.J. 1969. Genetic Eyherabide, G.H. & Hallauer, A.R. 1991.
variability after 65 generations of selection Reciprocal full-sib recurrent selection in
in Illinois high oil, low oil, high protein and maize. I. Direct and indirect responses.
low protein strains of Zea mays L. Crop Crop Science, 31: 952959.
Science, 9: 179181. Falconer, D.S. 1981. Introduction to quantita-
Dudley, J.W. & Lambert, R.J. 1992. Ninety tive genetics. New York, USA, Longman
generations of selection for oil and protein Inc.
in maize. Maydica, 37: 8187. Falconer, D.S. & Mackay, T.F.C. 1996.
Dudley, J.W. & Lambert, R.J. 2004. 100 gen- Introduction to quantitative genetics. Essex,
erations of selection for oil and protein in UK, Longman.
corn. Plant Breeding Review, 24: 79110. Fehr, W.R. 1987. Principles of cultivar devel-
Dudley, J.W., Lambert, R.J. & Alexander, D.E. opment. Volume 1. Theory and Technique.
1974. Seventy generations of selection for New York, USA, Macmillan Publishing
oil and protein concentration in the maize Company.
kernel. In J.W. Dudley, ed. Seventy genera- Fisher, R.A. 1918. Correlations between rela-
tions for oil and protein in maize. Madison, tives on the supposition of Mendelian inher-
USA, Crop Science Society of America. itance. Transactions of the Royal Society of
Dudley, J.W., Lambert, R.J. & de la Roche, I.A. Edinburgh, 52: 339433.
1977. Genetic analysis of corn strains diver- Frank, T.E. & Hallauer, A.R. 1999. Inter- and
gently selected for percent oil and protein. intra-population genetic variances after ten
Crop Science, 17: 111117. cycles of reciprocal full-sib recurrent selec-
Duvick, D.N. & Cassman, K.G. 1999. Post- tion in the BS10 and BS11 synthetic maize
green revolution trends in yield potential populations. Maydica, 33: 924.
of temperate maize in the north-central Frey, K.J. & Holland, J.B. 1999. Nine cycles of
United States. Crop Science, 39: 1622 recurrent selection for increased groat-oil
1630. content in oat. Crop Science, 39: 16361641.
Theory and application of plant breeding for quantitative traits 59
Gardner, C.O. 1961. An evaluation of effects Hinze, L.L. & Lamkey, K.R. 2003. Absence
of mass selection and seed irradiation with of epistasis for grain yield in elite maize
thermal neutrons on yield of corn. Crop hybrids. Crop Science, 43: 4656.
Science, 1: 241245. Holland, J.B. 2001. Epistasis and plant breed-
Gilmour, A.R., Cullis, B.R. & Verbyla, A.P. ing. In J. Janick, ed. Plant Breeding Reviews,
1997. Accounting for natural and extrane- vol. 21. John Wiley & Sons, Inc.
ous variation in the analysis of field experi- Holland, J.B., Nyquist, W.E. & Cervantes-
ments. Journal of Agricultural Biological Martinez, C.T. 2003. Estimating and
and Environmental Statistics, 2: 269293. interpreting heritability for plant breed-
Griffing, B. 1994. Historical perspectives on ing: an update. In J. Janick, ed. Plant
contributions of quantitative genetics to Breeding Reviews, vol. 22. John Wiley &
plant breeding. In K.J. Frey, ed. Historical Sons, Inc.
perspectives in plant science. Ames, USA, Holthaus, J.F. & Lamkey, K.R. 1995a.
Iowa State University Press. Population means and genetic variances
Guzman, P.S. & Lamkey, K.R. 1999. Predicted in selected an unselected Iowa Stiff Stalk
gains from recurrent selection in the BS11 synthetic maize populations. Crop Science,
maize population. Maydica, 44: 93-99. 35: 15811589.
Guzman, P.S. & Lamkey, K.R. 2000. Effective Holthaus, J.F. & Lamkey, K.R. 1995b.
population and genetic variability in the Response to selection and changes in genet-
BS11 maize population. Crop Science, ic parameters for 13 plant and ear traits in
40: 338346. two maize recurrent selection programs.
Hagdorn, S., Lamkey, K.R., Frisch, M., Maydica, 40: 357370.
Guimaraes, P.E.O. & Melchinger, A.E. 2003. Horner E.S., Magloire, E, & Morera, J.A. 1989.
Molecular genetic diversity among progeni- Comparison of selection for S2 vs. tescross
tors and derived elite inbred lines of BSS performance for population improvement
and BSCB1. Crop Science, 43: 474482. in maize. Crop Science, 29: 868874.
Haldane, J.B.S. 1924. A mathematical theory Hospital, F. & Charcosset, A. 1997. Marker-
of natural and artificial selection. Part I. assisted introgression of quantitative trait
Transactions of the Cambridge Philosophical loci. Genetics, 147: 14691485.
Society, 23: 1941. Johannsen, W. 1903. ber Erblichkeit in
Hallauer, A.R. 1984. Reciprocal full-sib selec- Populationen und reinen Linien. Jena,
tion in maize. Crop Science, 24: 755759. Germany, Fischer.
Hallauer, A.R. & Miranda Fo, J.B. 1988. Kearsey, M.J. & Pooni, H.S. 1998. The geneti-
Quantitative genetics in maize breed- cal analysis of quantitative trait. UK, Stanley
ing. 2nd edition. Ames, USA, Iowa State Thornes (Publishers) Ltd.
University Press. Keeratinijakal, V. & Lamkey, K.R. 1993a.
Hallauer, A.R. & Eberhart, S.A. 1970. Reciprocal Responses to reciprocal recurrent selection
full-sib selection. Crop Science, 10: 315316. in BSSS and BSCB1 maize populations.
Harlan, J.R. 1975. Crops and Man. Madison, Crop Science, 33: 7377.
USA, American Society of Agronomy. Keeratinijakal, V. & Lamkey, K.R. 1993b.
Helms, T.C., Hallauer, A.R. & Smith, O.S. Genetic effects associated with recip-
1989. Genetic drift and selection evaluated rocal recurrent selection in BSSS and
from recurrent selection programs in maize. BSCB1 maize populations. Crop Science,
Crop Science, 29(3): 602607. 33: 7882.
60 Plant breeding and farmer participation
Kempthorne, O. 1969. An introduction to mass selection for ear length in maize. Crop
genetic statistics. Ames, USA, Iowa State Science, 38: 10991107.
University Press. Mackay, I. & Powell, W. 2007. Methods for
Labate, J.A., Lamkey, K.R. & Woodman, W.L. linkage disequilibrium mapping in crops.
1999. Temporal changes in allele frequen- Trends in Plant Science, 12: 5763.
cies in two reciprocally selected maize pop- Melchinger, A.E. 1987. Expectations of means
ulations. Theoretical and Applied Genetics, and variances of testcrosses produced from
99: 11661178. F2 and backcross individuals and their selfed
Ladizinsky, G. 1998. Plant evolution under progeny. Heredity, 59: 105115.
domestication. Dordrecht, Netherlands, Melchinger, A.E., Schmidt, W. & Geiger, H.H.
Kluwer Academic Publishers. 1988. Comparison of testcrosses from F2
Lamkey, K.R. 1992. Fifty years of recurrent and first backcross population in maize.
selection in the Iowa Stiff Stalk synthetic Crop Science, 28:743-748.
maize population. Maydica, 32: 1928. Menz Rademacher, M.A., Hallauer, A.R. &
Lamkey, K.R., Schnicker, B.J. & Melchinger, Russell, W.A. 1999. Comparative response
A.E. 1995. Epistasis in an elite maize hybrid of two reciprocal selection methods in BS21
and choice of generation for inbred line and BS22 maize populations. Crop Science,
development. Crop Science, 35: 12721281. 39: 8997.
Lande, R. & Thompson, R. 1990. Efficiency Miller, P.A. & Rawlings, J.O. 1967. Selection
of marker-assisted selection in the improve- for increasing lint yield and correlated
ment of quantitative traits. Genetics, responses in upland cotton, Gossypium hir-
124: 743756. sutum L. Crop Science, 7: 637640.
Laurie, C.C., Chasalow, S.D., LeDeaux, J.R., Moose, S.P., Dudley, J.W. & Rochefort, T.R.
McCarroll, R., Bush, D., Hauge, B., Lai, C., 2004. Maize selection passes the centu-
Lark, D., Rochefort, T.R. & Dudley, J.W. ry mark: a unique resource for 21st cen-
2004. The genetic response to long-term tury genomics. Trends in Plant Science,
artificial selection for oil concentration in the 9(7): 358364.
maize kernel. Genetics, 168: 21412155. Moreno-Gonzlez, J. & Cubero, J.I. 1993.
Lee, M. 1995. DNA markers and plant breed- Selection strategies and choice of breed-
ing programs. Advances in Agronomy, ing methods. In M.D. Hayward, N.O.
55: 265344. Bosemark & I. Romagosa, eds. Plant
Leng, E.R. 1962. Selection reversal in strains breeding principles and prospects. London,
of corn previously long-term selected Chapman & Hall.
for chemical composition. Crop Science, Moreno-Gonzlez, J. & Dudley, J.W. 1981.
2: 167170. Epistasis in related and unrelated maize
Lin, C.Y. 1978. Index selection for genet- hybrids determined by three methods. Crop
ic improvement of quantitative charac- Science, 21: 644651.
ters. Theoretical and Applied Genetics, Moreno-Gonzlez, J. & Grossman, M. 1976.
52: 4956. Theoretical modifications of reciprocal
Lindstrom, E.W. 1926. Linked inheritance recurrent selection. Genetics, 84: 95111.
in tomato. Iowa State College Journal of Moreno-Gonzlez, J., Dudley, J.W. & Lambert,
Science, 1: 313 R.J. 1975. A Design III study of linkage dis-
Lpez-Reynoso, J. de J. & Hallauer, A.R. equilibrium for percent oil in maize. Crop
1998. Twenty-seven cycles of divergent Science, 15: 840843.
Theory and application of plant breeding for quantitative traits 61
Morgan, T.H. 1926. The Theory of the Gene. Russell, W.A., Blackburn, D.J. & Lamkey,
New Haven, USA, Yale University Press. K.R. 1992. Evaluation of modified
Nilsson-Ehle, H. 1909. Kreuzungsunter- reciprocal recurrent selection proce-
suchungen an Hafer und Weizen. Acta dure for maize improvement. Maydica,
Universitas Lund, Ser. II, 5: 1122. 37: 6167.
Nyquist, W.E. 1991. Estimation of heritabil- Snchez-Monge, E. 2002. Diccionario de
ity and prediction of selection response in plantas de inters agrcola. Ministerio
plant populations. Critical Reviews in Plant de Agricultura, Pesca y Alimentacin,
Science, 10: 235322. Madrid.
Oyervides-Garcia, M. & Hallauer, A.R. 1986. Sax, K. 1923. The association of size dif-
Selection-induced differences among strains ferences with seed coat pattern and pig-
of Iowa Stiff Stalk Synthetic maize. Crop mentation in Phaseolus vulgaris. Genetics,
Science, 26: 506511. 8: 552560.
Penny L.H. & Eberhart, S.A. 1971. Twenty Smith, O.S. 1983. Evaluation of recurrent
years of reciprocal recurrent selection with selection in BSSS, BSCB1, and BS13 maize
two synthetic varieties of maize (Zea mays populations. Crop Science, 23: 3540.
L.). Crop Science, 11: 900903. Stebbins, G.L. 1994 Biological revolutions
Piper, T.E. & Fehr, W.R. 1987. Yield improve- of thought during the twentieth century.
ment in a soybean population by utilizing In K.J. Frey, ed. Historical perspectives
alternative strategies of recurrent selection. in plant science. Ames, USA, Iowa State
Crop Science, 27: 172178. University Press.
Podlich, D.W., Cooper, M. & Basford, K.E. Sughroue, J.R. & Rochefort, T.R. 1994.
1999. Computer simulation of a selection Restriction fragment length polymor-
strategy to accommodate genotype-envi- phism differences among Illinois long-
ronment interactions in a wheat recur- term selection strains. Theoretical and
rent selection programme. Plant Breeding, Applied Genetics, 87: 916924.
118: 1728. van Eeuwijk, F.A., Malosetti, M., Yin, X.,
Rodriguez, O.A. & Hallauer, A.R. 1988. Effects Struik, P.C. & Stam, P. 2005. Statistical
of recurrent selection in corn populations. models for genotype by environment data:
Crop Science, 28: 796800. from conventional ANOVA models to
Romagosa, I. & Fox, P.N. 1993. Genotype eco-physiological QTL models. Australian
environment interaction and adaptation. Journal of Agricultural Research, 56: 883
In M.D. Hayward, N.O. Bosemark & I. 894.
Romagosa, eds. Plant breeding: principles Weyhrich, R.A., Lamkey, K.R. & Hallauer,
and prospects. London, Chapman & Hall. A.R. 1998a. Responses to seven methods
Romagosa, I., van Eeuwijk, F. & Thomas, of selection in the BS11 maize population.
W.T.B. 2009. Statistical analyses of geno- Crop Science, 38: 308321.
type by environment data. In M.J. Carena, Weyhrich, R.A., Lamkey, K.R. & Hallauer,
ed. Cereals. [Handbook of Plant Breeding A.R. 1998b. Effective population size and
Series, Vol. 3]. Berlin, Springer Verlag. response to S1-progeny selection in the
Russell, W.A. & Eberhart, S.A. 1975. Hybrid BS11 maize population. Crop Science,
performance of selected maize lines from 38: 11491158.
reciprocal recurrent and testcross selec- Wiersma, J.J., Bush, R.H., Fulcher, G.G. &
tion program. Crop Science, 15: 14. Hareland, G.A. 2001. Recurrent selection
62 Plant breeding and farmer participation
for kernel weight in spring wheat. Crop Oxford, UK, Oxford University Press.
Science, 41: 9991005. Wright, S. 1921. Systems of matings. Genetics,
Wood, R.J. & Orel, V. 2001. Genetic prehisto- 6: 111178.
ry in sheep breeding: a prelude to Mendel.
63
CHAPTER 3
Salvatore Ceccarelli
64 Plant breeding and farmer participation
Parental material for specific traits can generated. In general, a breeding pro-
also be identified through specific screen- gramme with a regional perspective (i.e.
ing activities. A typical example is resist- serving a single country or state) per-
ance to biotic stresses, for which a routine forms between 150 and 200 crosses per
parallel activity in a breeding programme is season on average. The number can go
to screen germplasm either under artificial up to several thousand in international
inoculation or in hot spots, i.e. in a loca- breeding programmes. On the issue
tion with a very high incidence of natural of the number of parents, Witcombe
infection or pest challenge. Another exam- and Virk (2001) have proposed the use
ple is the case of quality characteristics, of a low number of crosses (see also
in which germplasm is screened for the Section 6.5.2 (Chapter 6) for further
presence or amount, or both, of specific details on their theory).
compounds associated with a given food, When the objective of the breeding
feed or processing quality. programme is to produce hybrids, the
Parental material can also be character- variability is generated through methods of
ized for the adaptation to different environ- population improvement based on recurrent
ments or countries, and therefore crosses selection or through crosses between elite
can be made with the objective of producing inbred lines (see Section 11.5 in Chapter 11
breeding material with general adaptation for further details).
to particular environments or countries. In a number of cross-pollinated species,
Examples are screening under controlled hybrid vigour is exploited by developing
conditions for resistance to salinity, cold, synthetic varieties through intercrossing a
drought, or for micro-elements such as number of genotypes of known superior
boron, manganese or aluminium (Marshall, combining ability, i.e. genotypes that are
1991; Karakousis et al., 2003). known to give superior hybrid perform-
Backcrossingeven though it is often ance when crossed in all combinations. In
described as a breeding methodactually contrast to hybrids, the seed of synthetic
refers to a type of cross to generate varieties can be used for succeeding sea-
variability in a fashion that increases the sons, and for this reason synthetic varieties
frequency of desirable combination of are common in crops, such as forage crops,
traits, i.e. most of the traits from the where cost precludes the development or
recurrent parent and one or a few from the use of hybrid varieties.
non-recurrent parent.
A number of molecular tools are avail- 3.2.2 Exploiting existing variability
able today to assist the breeder in the It is widely recognized that landraces and
selection of parents and in subsequent stag- wild relatives harbour large amounts of
es, and they are described in Section 19.5 genetic variability, which, in the case of
(Chapter 19). self-pollinated crops, is readily available,
The number of crosses made by a as landraces and wild relatives are largely
breeding programme depends on vari- composed of different homozygotes. In
ous factors, such as the number of dif- these cases, the first step is represented by
ferent objectives, the resources available the collection of single spikes or plants,
and the breeding method used, which while their evaluation represents the second
affects the number of breeding lines stage of the breeding programme.
66 Plant breeding and farmer participation
selection can be practised between crosses in In most of the cases where the yield test-
the target environment, and within crosses ing is conducted partly on station and partly
on station for traits with high heritability. in farmers fields, the testing on research sta-
Single-seed descent (described in tions usually covers a period of three years.
Chapter 10) is one of the best ways of Most commonly, the number of breeding
exploiting the variability within a superior lines entering the testing stage is progres-
cross and is fully compatible with a sively reduced by discarding those that per-
participatory programme as long as farmers formed below a given standard. The most
contribute to the identification of the commonly used agronomic trait used to
superior cross. promote or discard breeding material during
the testing stage of a breeding programme is
3.3.2 Cross-pollinated crops grain yield. However, when the objectives of
The methods of exploiting the variability a breeding programme include, for example,
generated in the first stage in cross-pol- quality characteristics and disease resistance,
linated crops vary according to the final traits such as seed size or reaction to diseases
product that the programme aims to pro- complement grain yield.
duce (a hybrid, an open-pollinated variety, The testing of experimental cultivars
a synthetic variety, etc.). has a number of methodological and philo-
When the objective of the breeding pro- sophical issues. The methodological issues
gramme is to produce hybrids, the variabil- include field plot techniques, design of
ity generated through methods of popula- trials (replicated vs. unreplicated trials),
tion improvement is exploited through the analysis of variety trials, and the organiza-
development of inbred lines (the methods tion and structure of Multi-Environment
to do that are described in Chapter 11), Trials (METs). The philosophical issue is
which are then crossed to produce the primarily whether the testing of experi-
commercial hybrids. In the case of hybrid mental cultivars should be conducted in an
production, farmers could be involved both optimum climatic and agronomic environ-
in the second step by contributing to the ment or should be conducted in the target
evaluation of the inbred lines, and in the environment. When selection is conducted
third step by participating in the evaluation in the target environment, the breeding pro-
of the hybrids. gramme has to decide how many and which
target environments to serve (Chapter 20).
3.4 TESTING OF EXPERIMENTAL One of the basic principles to apply
CULTIVARS in implementing the third stage of a plant
Testing of potential cultivars is the last stage breeding programme is that field trials are
of a breeding programme, which eventually expensive, and therefore the breeder should
ends with a new variety recommended for always find an optimal compromise between
cultivation. the number and the size of the trials, their
Usually, this stage takes place partly precision, and the amount and the relevance
on research stations and partly in farmers of the information generated. An ideal sys-
fields. However, there are exceptions, the tem of testing of experimental cultivars is a
best known being the breeding programmes system that has the capacity to self-monitor
in Australia where all the yield testing actu- its efficiency, effectiveness and relevance,
ally takes place in farmers fields. and has flexibility to allow changes.
68 Plant breeding and farmer participation
TABLE 3.1
Efficiency of selection for grain yield as affected by plot size during the testing stage
Selection criterion in the No. of lines Plot size No. and percentage of lines outyielding the
previous season best local check
TABLE 3.2
Average yield (kg/ha) and coefficient of variation in barley yield trials conducted in two locations in
northern Syrian Arab Republic as lattice designs
Trial No. Tel Hadya (352.6 mm rainfall) Bouider (243.6 mm rainfall)
nation prior to establishing a trial on a dry variability can be kept at levels compara-
site if there is reason to believe that some ble with those of well-managed research
entries will not germinate completely. stations with high average yields. As an
A series of check entries, spaced at example of what could be achieved by the
regular intervals throughout unreplicated package of plot techniques described in
progeny trials, is essential to compensate this section and the use of the experimental
for the effects of soil variability. Plot data designs described in the next section in
are expressed relative to the performance of controlling environmental variability, we
the check, adjusted for the physical distance compared the coefficient of variation of
between the nearest check plots and the nine trials grown on a stress site (average
plot in question. The check genotypes must yield = 0.95 t/ha) and on a non-stress site
always include the farmers cultivar(s), lines (average yield = 3.19 t/ha) (Table 3.2). At
that are well known to the breeder, and the yield levels where breeders usually do not
best lines previously identified by the breed- work because the coefficient of variation
ing programme. A progeny trial should be is too large, the lattice design was capable,
arranged in the field to ensure that check in most of the trials, of keeping it within
entries will not all be in the same columns; acceptable limits.
rather, they should form a grid which will A common source of error in conduct-
provide a visual impression of the uniform- ing yield trials is an uneven plot length
ity (or the variability) of the field. resulting from trimming plots to eliminate
When all these techniques are used on the edge effect. An effective way of ensur-
sites with low yield potential due, for ing a uniform plot length is to spray a her-
example, to moisture stress, environmental bicide using booms placed on a rigid arm
70 Plant breeding and farmer participation
at a distance equivalent to the desired plot Unfortunately, despite the greater effi-
length. ciency of lattice designs, of generalized
lattices (Patterson and Hunter, 1983) and
3.5.2 Design of trials of neighbour analysis (Cullis and Gleeson,
The choice of an appropriate experimental 1989), the randomized complete block
design is another important decision affect- design is still dominant in many breed-
ing the precision of a trial. ing programmes in developing countries,
One of the first issues in discussing the particularly in those conditions where an
design of a trial is the alternative of repli- increase in trial efficiency is most needed.
cated vs. unreplicated trials. In the testing The use of generalized lattice designs
stage of a breeding programme, which typi- (Patterson and Williams, 1976) combines
cally goes through three steps, the number good error control with flexibility in the
of breeding lines being tested is usually in numbers of treatments required. A promis-
the range of several hundreds or even a few ing extension of this design that removes
thousand in the first step, reducing to usu- both row and column effects has been
ally less than 50 in the third. described by Patterson and Robinson
The use of replicated designs with several (1989). Nearest-neighbour analysis (e.g.
hundred or a few thousand lines implies the Wilkinson et al., 1983; Hinz, 1987) have
use of large areas, which makes the control of been used extensively in Australia to
experimental error more problematic. In this remove the effects of gradients of moisture
step the number of experimental units (plots) stress within replicated and unreplicated
per breeding line is limited by the amount of trials (Marshall, 1987).
seed, which is usually small at this stage. Spatial variability is a reality in field
The combination of limited amounts trials and a proportion of this is account-
of seed and of a large number of lines ed for as inter-block variability by using
has made popular the use of unreplicated block (complete or incomplete) designs.
designs with systematic checks and with a However, a large amount of spatial vari-
row and column arrangement of the plot, ability still remains unaccounted for, and
which makes it possible to use spatial this may lead to erroneous conclusions. To
analysis (Singh et al., 2003). further capture this unaccounted-for vari-
One improvement over the unreplicated ation (which is mainly due to intra-block
design with systematic checks is a partially variation), yield data from variety yield
replicated design where only a certain per- trials can be analysed using various spatial
centage of entries is replicated while still models. Singh et al. (2003) showed that spa-
maintaining the systematic (also called grid) tial models add considerable value to trials;
checks. In addition to the expected higher the best spatial models gave efficiency
precision, the design responds to a frequent values of over 330 percent in winter-sown
problem in the initial stages of testing, that chickpea, 140 percent in lentil and 150 per-
is the different amount of seed available for cent in barley trials. Furthermore, the use
different entries. of these best models resulted in a change
In replicated yield trials, where geno- in the ranking of genotypes (on the basis
types under test may number 200500, of mean yield), which therefore resulted in
improved statistical designs can lead to a different set of genotypes being selected
important increases in trial efficiency. for high yield. It is recommended that
Main stages of a plant breeding programme 71
(1) incomplete block designs be used in When METs are conducted in farmers
variety trials; (2) the Akaike Information fields, which is common in some conven-
Criterion (developed by Akaike under the tional breeding programmes (in Australia,
name of An Information Criterion (AIC) for example), the breeder may face organi-
in 1971 and proposed in Akaike (1974) as a zational issues different from those in a
measure of the goodness of fit of an esti- research station (Chapter 9). METs planted
mated statistical model) be used to select in farmers fields are not yet participatory
the best spatial model; and (3) genotypes plant breeding because, as is the case for
be selected after the use of this model. the Australian breeding programmes, farm-
The selected model would most effectively ers only make land available against the
account for spatial variability in the field payment of a rent and do not participate in
trials, improve selection of the most desir- decisions related to the selection of breed-
able genotypes, and therefore improve the ing material.
efficiency of breeding programmes. One important aspect of a MET is the abil-
Additional information on this issue, ity to subdivide Genotype Environment
also with regard to useful software, is given interaction into Genotype Location (GL)
in Chapter 20. and Genotype Year (GY) interactions.
Distinguishing these interactions is impor-
3.6 MULTI-ENVIRONMENT TRIALS tant because the two differ in importance
The last stage of a plant breeding pro- with regard to plant breeding. While GY
gramme aims also at understanding the interactions are largely unpredictable, GL
reaction of the breeding material to a mul- interactions can be predicted. In the case of
titude of environments, i.e. locations, years, GL interaction, therefore, it is essential to
and possibly different types of agronom- assess both its magnitude (relative to G) and
ic management. During this process, the its repeatability over time. Such an assess-
amount of breeding material is progres- ment allows the target population of envi-
sively reduced, and the number of locations ronments to be divided into subsets in a way
and the number of replications per location that ensures that there is a high degree of
is progressively increased. In the first level repeatability or consistency (Kempthorne,
of yield testing the general tendency is to 1952) of GL interaction between subsets
have as many locations as possible (the and a low degree of repeatability within
main limiting factor being the amount of subsets. Eventually this leads to the identi-
seed available), sacrificing number of repli- fication of high-yielding, stable genotypes
cations per location. Because of the limited adapted either specifically or widely, and to
precision of these trials, most breeders will clusters of response-similar environments,
mostly do negative selection, i.e. discarding which can contribute to defining a selection
the obviously inferior breeding material. strategy and to locating a small number of
As the material is advanced, the precision optimal selection sites for future breeding.
of the trials also increases. At the end of These topics are extensively discussed in
the testing phase, and for the surviving Chapter 20.
breeding material, data are available from The use of biplots as means to graphi-
a number of locations and years, which cally display genotype adaptation patterns
can be analysed with one of the techniques and environment similarity for GE inter-
described in Chapter 20. action effects is discussed in Chapter 20.
72 Plant breeding and farmer participation
FIGURE 3.1
Biplots of farmers preferences in barley grown in two villages in Yemen
(6 and 14 varieties, respectively)
2.0 3.2
6
SY
2.4 BY4
1.5 9 13
TLN SY
KW
BY
1.0 1.6
1
WS SEED
0.5 4 0.8 3
PC 2
9
MS
PC 2
PH 2
0.0 5 WS 0.0 GY
10 6
MS
-0.5 -0.8 1 11
5
GY 14
-1.0 -1.6 12 SPL DM
3
HI DH
-1.5 -2.4
7
2
-2.0 SEED HI PH
-1.5 -1 0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.2 -2.4 -1.6 -0.8 0.0 0.8 1.6 2.4 3.2 4.0
PC 1 -PC 1
-
Bit Al Wali Yarim
Key: PH = plant height; KW = 1000-kernel weight; SEED = number of kernels/spike; BY, GY and SY = biological, grain and
straw yields, respectively; HI = harvest index; DH = days to heading; DM = days to maturity; TLN = tiller number; MS = male
selection; WS = female selection.
Biplots, such as those envisaged by Yan and wide angle between the vectors for farmer
Tinker (2005), can also be useful in PPB selection and those for plant height suggest-
programmes to assess (1) the traits used by ing that farmers in Yarim selected for short
farmers as selection criteria; (2) the consist- plants and for early heading and maturity, as
ency of selection criteria across environ- indicated by the direction of the vectors of
ments; and (3) whether locations that show these two characters opposite to the vectors
repeatable G Y interactions and that for farmers selection.
therefore could be lumped together as one Figure 3.2 shows an example of differ-
subregion (or mega-environment), actually ences in locations clustering depending on
differ in farmer preferences. whether grain yield or farmers preferences
An example of using the biplot to assess are used. On the basis of grain yield, four
which traits are actually used by farmers locations, namely MO, GWF, KH and
as selection criteria and the consistency of RAE, are closely correlated; they are likely
selection across environments in given in to represent a similar environment as they
Figure 3.1. A narrow angle between the vec- discriminate similarly among genotypes.
tors for farmer selection (MS for the males If this is repeatable over time, then it can
and WS for females) and the vector for a be argued that any one of the four will be
plant character indicates a strong preference sufficient to represent that given macro-
for that given character. This is the case environment, thus leading to a considerable
of plant height in Bit Al Wali, where both saving in resources. However, only 2 of
men and women selected for tall plants, and the 4 locations (MO and KH) are closely
to a slightly less extent for grain yield. It correlated also for farmer preference, while
can be noted that in the case of the second GWF and RAE are independent from each
village, the preferences were reversed, the other and weakly correlated with the pre-
Main stages of a plant breeding programme 73
FIGURE 3.2
Biplots of grain yield and farmers preferences for 182 barley genotypes grown in
six farmers fields and one research station in Jordan (not all genotypes are shown)
2.5 2.5 66 12
PC 2
115 138 29 25
39 44 156
107 173 30 166 15327 155
0.0 153 RAE
150 0.0 19
16 142 93
172 50 129
KH 26 45 22 178 72 180
124 132
-2.5 124 154 90 110 95 111
GWS
MOGWF
176 51 167 112
-2.5 139
119 37 134169 131
RAE
94 163 142 130 92112
-5.0 104 107 173
-5.0 174 2 56
156
109
21
28 -7.5 181
RAW
98 125
-7.5 121 -10.0
140
157 99 RA 63
-10.0
-7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 12.5 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 12.515.0
PC 1 PC 1
Grain yield Farmer selection
GWF = Gweer farmer field; GWS = Gweer research station; RA = Rabba; KH = Khanasri; MO = Mohay; RAE = Ramtha East; and
RAW = Ramtha West.
vious two. Therefore, if these patterns are tantly, it is a linear process. In contrast, in
repeatable over time, it would be advisable PPB farmers participate in selection when
to only drop either MO or KH. genetic variability is at or near its maxi-
mum, such as selection between or within
3.7 CONCLUSION early segregating populations. Also, con-
In a typical breeding programme trary to PVS, PPB is a cyclic process.
it is possible to recognize three Ultimately, a participatory plant breed-
distinct stages: the creation of genetic ing programme can use and benefit from
variability; selection of the desirable the use of the most advanced experimental
gene combinations; and the final testing designs and analytical tools.
of these desirable gene combinations.
In a truly participatory breeding pro- REFERENCES
gramme, farmers participate in all the three Al Yassin, A., Grando, S., Kafawin, O., Tell, A.
stages. When farmers participate in only & Ceccarelli, S. 2005. Heritability estimates
the last stage (as often is the case), it is in contrasting environments as influenced
more appropriate to talk of participatory by the adaptation level of barley germplasm.
variety selection (PVS). There are impor- Annals of Applied Biology, 147: 235244.
tant conceptual differences between the Akaike, H. 1974. A new look at the statistical
two. PVS is selection among (usually only model identification. IEEE Transactions on
a few) finished or nearly finished varieties, Automatic Control, 19(6): 716723.
such as when farmers choose from on-farm Blum, A. 1988. Plant breeding for stress envi-
variety trials, which are the very last stage ronments. Boca Raton, USA, CRC Press.
of a breeding programme, and, very impor- Comadran, J., Russell, J.R., van Eeuwijk,
74 Plant breeding and farmer participation
F., Ceccarelli, S., Grando, S., Stanca, efficiency of incomplete block designs in
A.M., Francia, E., Pecchioni, N., Akar, T., National List and Recommended List cereal
Al-Yassin, A., Benbelkacem, A., Choumane, variety trials. Journal of Agricultural Science
W., Ouabbou, H., Bort, J., Araus, J.-L., (Cambridge), 101: 427433.
Pswarayi, A., Romagosa, I., Hackett, C.A. Patterson, H.D. & Williams, E.R. 1976. A
& Thomas, W.T.B. 2008. Mapping adapta- new class of resolvable incomplete block
tion of barley to droughted environments. designs. Biometrika, 63: 8392.
Euphytica, 161: 3545. Patterson, H.D. & Robinson, D.L. 1989. Row-
Cullis, B.R. & Gleeson, A.C. 1989. Efficiency and-column designs with two replicates.
of neighbour analysis for replicated variety Journal of Agricultural Science, 112: 7377.
trials in Australia. Journal of Agricultural Rosenow, D.T. 1987. Breeding sorghum for
Science (Cambridge), 113: 233239. drought resistance. pp. 8389, in J.M.
Fischer, R.A. 1981. Optimizing the use of Menyonga, T. Bezuneh and A. Youdeowei
water and nitrogen through breeding of (editors). Food Grain Production in Semi-
crops. Plant and Soil, 58: 249278. Arid Africa. Ouagadougou.
Hinz, P.N. 1987. Nearest-neighbor analysis Schnell, F.W. 1982. A synoptic study of the
in practice. Iowa State Journal of Research, methods and categories of plant breeding.
62: 199217. Zeitschrift fur Pflanzenzchtung, 89: 118.
Karakousis, A., Barr, A.R., Kretschmer, J.M., Singh, M., Malhotra, R.S., Ceccarelli, S.,
Manning, S., Jefferies, S.P., Chalmers, Sarker, A., Grando, S. & Erskine, W. 2003.
K.J., Islam, A.K.M. & Langridge, P. 2003. Spatial variability models to improve dry-
Mapping and QTL analysis of the barley land field trials. Experimental Agriculture,
population Clipper Sahara. Australian 39: 110.
Journal of Agricultural Research, 54(11- Wilkinson, G.M., Eckert, S.R., Hancock, T.W.
12): 11371140. & Mayo, O. 1983. Nearest neighbour (NN)
Kempthorne, O. 1952. Design and Analysis analysis of field experiments (with dis-
of Experiments. New York, USA, John cussion). Journal of the Royal Statistical
Wiley. Society, Series B, 45: 151211.
Marshall, D.R. 1987. Australian plant breed- Witcombe, J.R. & Virk, D.S. 2001. Number of
ing strategies for rainfed areas. In J.P. crosses and population size for participa-
Srivastava, E. Porceddu, E. Acevedo & S. tory and classical plant breeding. Euphytica,
Varma, eds. Drought Tolerance in Winter 122: 451462.
Cereals, pp. 8999. Chichester, UK, Wiley Yan. W. & Tinker, N.A. 2005 An integrated
& Sons. biplot analysis system for displaying, inter-
Marshall, D.R. 1991. Alternative approaches preting, and exploring genotype environ-
and perspectives in breeding for higher ment interaction. Crop Science, 45: 1004
yields. Field Crops Research, 26: 171190. 1016.
Patterson, H.D. & Hunter, E.H. 1983. The
75
CHAPTER 4
(or crop varieties) have outstanding nutri- ment of farmers; and overcoming typical
tional qualities, are well adapted to mar- limitations of science in the development
ginal conditions and low input farming, or contextall these factors are the potential
open up possibilities for income generation advantages of participatory plant breeding
(IPGRI/GFU/MSSRF, 2005). (Ashby and Sperling, 1995; Weltzien et al.,
Thus, breeding programmes today often 2003).
have to be designed in a manner different PPB includes all approaches to genetic
from the past. To meet the above-men- plant improvement involving close farmer
tioned new goals, they tend to be less researcher collaboration. The term particu-
centralized, more targeted towards spe- larly refers to active involvement of farmers
cific user groups and often use differ- in at least one of the stages of a plant breed-
ent germplasm. However, this is not all. ing programme, including setting objec-
To obtain impacts beyond a very local tives, generating variability, selecting and
scale, approaches have to be developed that testing, as well as seed production and
address large geographical areas while at distribution.
the same time respecting agro-ecological This active involvement of farmers can
and socio-cultural differences. This usu- take different forms. Farmer participation
ally requires cooperation among different can be consultative, if farmers are inter-
organizations that work at different scales, viewed on agro-ecological issues, or on the
and often have diverse agendas and back- performance of test varieties. More active
grounds. Consequently, methodologies for forms of farmer participation include, for
priority setting have to be adapted for such example, trial management, selection, pri-
cooperation to make the process transpar- ority setting and the development of action
ent and acceptable for all stakeholders. plans, or the overall management and
The management of social cooperation, implementation of the project (Farnworth
learning and decision-making processes is, and Jiggins, 2003; Lilja and Ashby, 1999).
as such, new for most plant breeders and Which degree of farmer participation is
their institutions. However, experiences appropriate and in which phase of a breed-
exist from other disciplines, particularly ing programme depends largely on the
social and economic sciences; here one can goals of the programme, as well as the type
build on fundamental expertise in the fields of improvements needed, and it is thus also
of knowledge systems, communication, an issue for priority setting (see Section 4.5
social learning and management (Leeuwis, below, under Roles and Responsibilities).
2004; Manktelow, 2003).
4.3 PRIORITY SETTING AS AN
4.2 PARTICIPATORY PLANT BREEDING ITERATIVE PROCESS
The concept of participatory plant breed- Setting priorities is an important part of
ing (PPB) emerged in the late 1980s as part professional plant breeding work. Time and
of a general development in participatory resources are usually limited, and they have
research methodologies during that period. to be allocated in a rational way in order to
Increased user orientation and more effi- reach the goals of the breeding programme.
cient allocation of research funds; higher Thus, considering issues and methodologies
adoption rates; a close relation to local for priority setting is a necessary step for
cultures, knowledge and skills; empower- any plant breeding programme, irrespec-
Methodologies for priority setting 79
tive of the degree of farmer participation target group of farmers. Closely linked to
or the institutional setting. However, little this are priority traits to be used as selec-
has been reported to date on methodologies tion criteria. To achieve good progress
for priority setting in plant breeding pro- from selection, the germplasm base must
grammes. Resource allocation, primarily be chosen appropriately, based on pro-
during the phase of testing experimental found knowledge of the available options.
cultivars, has been researched intensely, It is also important to discuss what type of
usually based on models for maximizing variety might be the most appropriate for
genetic gain, thus focusing on one or two achieving the project or programme goals.
key traits (e.g. Cooper and Byth, 1996). Part of this issue is also to address the ques-
We regard priority setting as an iterative tion of intravarietal diversity: how much of
and progressive process that will be it would be beneficial or necessary, and for
considered at many stages during a plant which traits. An issue that is often left until
breeding programme, not only in the project activities are planned is the identification of
planning phase. It is often not possible to key roles and responsibilities of partners.
anticipate all the options that may emerge in However, since different options for shar-
the course of the research process. Priority ing responsibilities between partners have
setting methodologies should therefore a major impact on some of the goals, it is
become part of the regular project work, in important to consider them from the outset
a way that allows adjustments and further of the breeding programme. The following
development of goals and priorities as the sections explain in more detail how the dif-
project work evolves. ferent issues for priority setting for a breed-
In the following sections we will look ing programme relate to the overarching
at issues for priority setting first, and then goals in specific situations.
suggest methods and communication tools
that could help to achieve a transparent 4.4.1 Goals as a basis for priority
process and productive outcomes. setting
All breeding programmes have at least
4.4 ISSUES FOR PRIORITY SETTING one goal related to improving production,
Clear priorities need to be set for a number such as yield, yield stability or a higher
of issues. The goals are the guiding princi- product value. Many PPBs have additional
ples for priority setting in any project of a goals, such as the conservation of local
defined duration, scale and scope. At the diversity, skill building and empowerment
same time, the goals themselves are also an of farmers, policy and regulatory changes,
issue for priority setting, as complex, con- increasing research efficiency, or benefits to
flicting or too general goals are not likely to specific users. Many of these goals tend to
be reached through technical plant breed- be implicit and depend on the institutional
ing work alone. background and on the history of the
For plant breeding programmes, it is breeding project.
vital to define the target group(s) and the Each organization and institution has
target environment(s), i.e. production con- their own implicit goals that are not always
ditions under which the newly identified easily communicated. Thus close interac-
varieties should perform better than exist- tion, exchange visits, and joint planning
ing cultivars, and the specific needs of the workshops that are held variously in the
80 Plant breeding and farmer participation
FIGURE 4.1
Issues for priority setting in a plant breeding programme in relation to the overall goals
and based on the assessment of target groups and environments, production systems,
seed management, and farmers needs and preferences
Improve
productivity
& generate Farmer
Conserve empowerment
& use income
& skill
biodiversity building
Achieve
benets for Change
specic groups policies
of farmers Issues for priority setting:
Target groups & environments
Germplasm base
Selection criteria
Improve
health &
Types of variety
Roles & responsibilities
...
nutrition
different partners workplace (e.g. research may even expect other benefits from the
station, village, trading place) are impor- cooperation, which cannot be fulfilled by a
tant to achieve a mutual understanding of breeding programme. Such general aims of
the different partners perspectives. It is the people could perhaps better be addressed
also understood that the relative impor- by activities other than plant breeding, or
tance of the different goals may change as by establishing partnerships with marketing
the project and, foremost, the partnership organizations or food processing companies
advances and evolves. (and including their specific goals into the
If the project work involves close breeding programme).
interaction between farmers or farmer
organizations and researchers, it is particu- From goals to priority setting
larly important to clarify the goals from The goals have been described as the guid-
the project planning phase. For many ing principles for priority setting. At the
farmers, it is not easy to understand what same time, the priority setting process
scientists do and how research is organized. builds on understanding the present situ-
As a consequence, they may be tempted to ation, anticipated changes, and farmers
overestimate the direct effects of the research needs. A detailed analysis of the production
on yield or income generation, or they environment is required, including existing
Methodologies for priority setting 81
varieties and how they are used by farmers, with limited resources and under marginal
their preferences and relevant resources agro-climatic conditions. Ceccarelli,
(i.e. local knowledge, skills, germplasm). Grando and Booth (1996) and Ceccarelli
In particular, it is necessary to identify the et al. (2000) have shown theoretically
major constraints to production increases and practically that interactions between
and income generation. Participatory meth- genotype and environment can be positively
ods for such situation analysis have been exploited if the selection is done in the
described in detail by Christinck, Weltzien target environment, e.g. farmers fields.
and Hoffmann (2005). An open dialogue Farmers as well as scientists successfully
in the course of which all partners evaluate selected populations or experimental lines
potential options and obstacles for future that produced better under the farmers
breeding activities could then follow (see conditions than other varieties grown
Section 4.6 of this chapter). This approach previously by those farmers. Experiences
is graphically summarized in Figure 4.1. of other research groups, with various
crops in differing natural and socio-cultural
4.4.2 Target groups and target environments, support this understanding
environments (Goyal, Joshi and Witcombe, 2001;
Identifying the target environment and Mekbib, 1997; Sperling, Loevinsohn and
target group in view of the overall project Ntabomvura, 1993; Weltzien et al., 2003).
goals is generally among the first strategic Narrow adaptation to specific conditions,
decisions to be taken in a plant breeding leading to the selection of many different
project. We therefore suggest a few subjects cultivars for various conditions and
for consideration, which refer to agro-eco- purposes, is often regarded as an advantage
logical as well as socio-economic factors. of the PPB approach: it serves specific
needs of farmers and enhances the level of
Broad versus narrow adaptation, and the agrobiodiversity in farmers fields (Sperling,
impact of PPB Loevinsohn and Ntabomvura, 1993; Joshi
The issue that certain plant types or varie- and Witcombe, 2001).
ties may perform differently in different However, a possible criticism regarding
environments is called genotype by envi- decentralized plant breeding programmes
ronment interaction by plant breeders. In could be that, due to the focus on specific,
general, most plant breeders tend to give often marginal environments, and only the
preference to those populations that per- local importance of the varieties developed,
form well under a wide range of conditions; their impact remains insignificant. Only a
this ability of plant populations is known as very few farmers who produce mainly for
broad adaptation. their own subsistence and modest require-
Broadly adapted varieties are also the ments would profit from the activities,
prime matter of interest for seed companies, and this would never justify the breeding
as the potential profit from the entire release efforts, let alone the cost of official variety
and multiplication business is usually release and seed multiplication.
related to the scale of distribution. However, At the same time, there are also cases
these varieties, if tested on research stations where varieties developed through PPB
in multi-locational trials, may fail under programmes are not necessarily so nar-
the conditions of poor farmers working rowly adapted. In Nepal, for example,
82 Plant breeding and farmer participation
FIGURE 4.2
Functions of crops within a farming system
Human nutrition
- quality
- quantity
- security
Medicinal value
- for humans
Animal fodder - for domestic animals
- quality
- quantity
- security
Cultural functions
CROP 1
- crops as part of cultural identity
- religious importance
Functions within the
cropping system (offerings, special dishes)
Crop 1 is a typical multipurpose crop with high importance for most functions (except marketing), whereas crop 2 is a
food crop important for nutrition and marketing, but not for other functions mentioned
Source: Christinck, Dhamotharan and Weltzien, 2005.
Various tools for entering into dialogue may also be related to culinary preferences,
with farmers on variety characterization such as taste, usefulness for certain preferred
and use have been proposed by Christinck, dishes, to useful by-products (i.e. construc-
Weltzien and Dhamotharan (2005). tion material) or to market requirements,
Furthermore, understanding farmers own e.g. grain colour and shape. In most cases
seed selection and the underlying criteria these criteria must meet a certain threshold
will give us important keys for the types of level of acceptability.
improvements farmers are looking for. Experience from PPB projects has shown
Some selection criteria are largely deter- that farmers often select for many criteria
mined by the requirements of adaptation to simultaneously, and in this way can indirect-
the target environment, e.g. flowering date, ly achieve considerable yield increase. This
resistances to specific pests and diseases, seems to be mainly related to the farmers
or to abiotic stresses such as soil acidity. ability to anticipate the performance of certain
Other selection criteria are determined by plant types under specific conditions that are
the technologies farmers are using, such as well known to them (Sperling, Loevinsohn
ease of harvesting, transportability, manual and Ntabomvura, 1993; Christinck, vom
threshing, or by the requirements of the Brocke and Weltzien, 2000).
farming system, e.g. mixed cropping and However, for professional plant breeders,
fodder use. Furthermore, selection criteria a detailed evaluation of each and every trait
86 Plant breeding and farmer participation
that might be important for farmers will often thought of as a preference: something
lead to a dead end. Resources for testing visual, qualitative and not really associated
and evaluating new germplasm or breeding with productivity or adaptation. Most of
material are always limited. The more criteria these traits could actually be incorporated
that are included in a selection programme, into existing varieties by backcrossing, if a
the less effort can be spent on each of them, source for the desired trait, i.e. a gene, exists
and thus less progress tends to be obtained in the breeders collection.
from selection. Thus, a guiding principle Complex traits have a low heritability
in the choice of selection criteria should be because their expression is highly influ-
to keep them to the minimum necessary. enced by environmental factors, i.e. the
The more focused and clear the targets for conditions in which the variety is grown.
selection, the greater are the chances of Many of these traits also tend to show size-
achieving them. We find here an excellent able amounts of genotype environment
option for cooperation between farmers interactions, i.e. the expression of a trait
and scientists in a breeding programme. in specific varieties depends on the condi-
Farmers can more efficiently select those tions in which the trait is being evaluated
materials that are overall compatible (see Chapter 20). One example would be a
with their situation, farming system and variety which responds well to fertilizer; its
marketing requirements and preferences, yield under high fertility conditions could
whereas scientists can be most effective in be higher than that of a local variety, where-
assembling appropriate germplasm with the as the local variety would outperform this
traits desired by the farmer and in selection variety under low soil fertility conditions.
for a limited number of critical traits. This example shows clearly that identifying
yielding ability as a key preferred trait is of
Heritability of traits and environmental little relevance. However, what is impor-
adaptation tant is the specification for which kind of
Formally trained plant breeders tend to growing conditions a higher yield perform-
classify traits by the complexity of their ance is being sought by farmers. This type
genetic control. They differentiate highly of specification is necessary for most of
heritable traits with simple genetic control the complex, productivity-related, traits,
from genetically complex traits with low as their assessment cannot be dissociated
heritability, along a continuum of increas- from the conditions under which they are
ing complexity, and thus decreasing genetic evaluated.
control or heritability (see Chapter 2). Another example of a selection criterion,
Highly heritable traits with simple which is often high on farmers lists of
genetic control tend to be mostly descriptive preferences, but usually very difficult
traits, such as colours of the grain or other to assess, is drought tolerance. The first
plant parts, hairiness, key aspects of crop problem is that a trait like drought tolerance
duration or flowering date, plant height may mean very different things to farmers,
and some types of disease resistance. While to crop physiologists or to breeders,
some of these traits are key factors for the and would thus entail very different
adaptation of a variety, such as flowering ways of assessing it, from physiological
date or disease resistance, many others measurements of drought response at the
are more related to what is intuitively biochemical, plant tissue, plant organ or
Methodologies for priority setting 87
of diversity available for selection largely such, contribute to raising awareness about
determines the success of the selection pro- the usefulness of this diversity among
gramme. This is particularly important in participating farmers and scientists, and
view of the first guiding principle for pri- thus increase the chances for future use
oritizing selection criteria, namely to keep of this germplasm. However, the goal of
criteria to the minimum necessary. increasing biodiversity in farmers fields
does not necessarily require a focus on local
Using local germplasm as breeding parent: and traditional germplasm. Particularly in
a way to increase the acceptability and those regions where a major part of the local
adaptation of new varieties diversity is already lost, a plant breeding
One basic approach for keeping the number programme could also be based on material
of selection criteria to a minimum is to from elsewhere, showing enough diversity
identify base germplasm that already has in traits that have been identified as useful
most of the traits expressed at the threshold for the target group of farmers.
level or above, but is variable for the major Any adoption of new varieties by farm-
trait targeted for improvement. Many PPB ers will change the portfolio of varieties
programmes have been very successful in available in a village community. This may
this respect, because they did use the local provide interesting new options for some
germplasm and farmers knowledge of it farmers, and possibly disadvantages for
for this purpose. By using local germplasm, others. Such developments can often not
most of the traits for adaptation and use are be anticipated fully. Unintended (negative)
already expressed at this threshold level, outcomes for some farmers can be reduced
and the novel germplasm can be chosen to by ensuring the multiplication and access
introduce new variability specifically for to seed of the original varieties, for exam-
improving one or two key traits, e.g. reduc- ple through strengthening seed exchange
ing the period from planting to flowering, networks, institutionalizing seed fairs or
or increasing yielding ability, or stover qual- community seed banks.
ity, or resistance to a major pest or disease.
4.4.5 Types of variety
Plant breeding and biodiversity What type of variety will be developed in
conservation the course of a plant breeding programme
The choice of germplasm is also a key issue has important implications with regard to
for achieving goals related to biodiversity the biodiversity in farmers fields and to
conservation. If used successfully in plant the options farmers have to use seed of this
breeding programmes, there are much variety for re-sowing, selling, exchange and
better chances of endangered germplasm their own breeding activities. These aspects
being preserved, compared with other touch the overarching goals of the breed-
approaches focusing on conservation per se. ing programme, and are thus important
If diversity conservation is a primary goal for consideration in the process of priority
of a plant breeding programme, a very good setting.
understanding of the nature and functions
of this diversity for the target group needs Variety types and agrobiodiversity
to be achieved. Assessing local diversity Varieties can have very different genetic
in a participatory research process can, as structures; they can differ in the degree
Methodologies for priority setting 89
FIGURE 4.3
Degree of heterogeneity and heterozygosity in different types of varieties
Degree of heter ozygosity
Variety mixture s
Pure line varietie s (Landraces of self-pollinating crops)
FIGURE 4.4
Formal and informal dimensions of seed systems and how they may interact; varieties
from the formal system may enter the local systems and vice versa
National/
Genebank international
food markets
Farm
household
Commercial Public
Consumption
breeding breeding
Cultivation
Public Planting
seed
sector
Commercial
seed
Relief
sector
agencies
Variety types and farmers access to seed commercially, or via national agricultural
The seed channels farmers use for sourc- research systems (Louwaars, 1994).
ing their seed are normally grouped into The informal system embraces most of
two broad seed systems: the formal and the ways in which farmers themselves pro-
the informal seed systems. The latter is also duce, disseminate and obtain seed: directly
sometimes termed the local, traditional or from their own harvest; through barter
farmer seed system (see Figure 4.4). among friends, neighbours and relatives;
The formal seed system involves a chain and through local grain markets or traders.
of activities that lead to clear products: i.e. The same general steps take place in the
certified seed of verified varieties. Thus, informal system as in the formal, but they
the chain usually starts with plant breeding take place as integral parts of farmers rou-
in research institutions or commercial tine grain production rather than as sepa-
companies, and results in varieties or rate activities. Also, rather than be moni-
hybrids intended for formal variety release. tored or controlled by government regula-
Formal regulations aim to maintain varietal tions, informal seed sector production is
identity and purity, as well as to guarantee guided by local technical knowledge and
physical, physiological and sanitary standards, and by local social structures and
quality. Seed marketing takes place through norms, including market forces (McGuire,
officially recognized seed outlets, either 2001). Varieties may be landraces or mixed
Methodologies for priority setting 91
races, or improved varieties that have made system often builds on traditional rules and
their way into the local system. forms of cooperation in village communi-
Perhaps because of their ability to meet ties, including cooperation among different
local needs and preferences, informal chan- wealth and ethnic groups. Thus, detailed
nels provide most of the seed that small knowledge of the seed systems and how
farmers use: it is estimated that somewhere they are related to different groups of farm-
between 80 and 90 percent of total seed ers is required for developing such strate-
sown originates from the informal system, gies (Sperling and Christinck, 2005).
although this varies a lot between different The type of variety that will be devel-
countries and regions, as well as for different oped, and how it can be reproduced and
crops. A formal seed system does not exist maintained by farmers, is thus a very
in practice for many local crops or varieties important consideration for a breeding pro-
of minor economic importance, whereas it gramme, particularly in situations where
is particularly important in regions where the formal system alone cannot serve the
hybrid maize is grown. The relative impor- target groups of farmers.
tance of the formal and informal seed systems
also much depends on the seed legislation of 4.5 ROLES AND RESPONSIBILITIES OF
the respective country. Very restrictive seed PARTNERS
laws have practically abolished the informal 4.5.1 Cooperation between different
seed system in some countries, whereas in organizations and stakeholders
others the legislative framework allows for Plant breeding is increasingly being done as
the co-existence of both systems. a partnership among different stakeholders:
Professional plant breeders are usually individuals, groups, organizations who
members of formal institutions (public or share an interest in using and improving
private), so that formal channels of seed crops. It is thus clear that the discussion
production and dissemination are the nor- about roles and responsibilities of the
mal route through which newly developed different partners is at the heart of such
varieties find their way to farmers fields. plant breeding projects, and is thus a critical
However, the formal and the informal sys- issue in the priority-setting process.
tems have both comparative advantages The history of a project (who took the
and disadvantages for variety diffusion, initiative and for what interest?) appears
and often address different client groups. to play an important role in this regard. It
Considering these differences could form makes a difference whether one organization
part of an active strategy for effective vari- initiated the project and organized the
ety diffusion in relation to the goals of the major part of the resources, and then
breeding programme. For example, the sought potential partners, or whether
informal seed system has various advan- it was a joint initiative from the outset.
tages for poor farmers, as the seed price is The present structure of international
usually lower and the modes of payment agricultural research, particularly with
flexible. If poor farmers access to new vari- regard to funding and accountability,
eties is a goal of the breeding programme, potentially poses problems for cooperative
variety diffusion through the informal sys- research that involves very different types
tem could be a good option for reaching of institutions. This is due to the large
this goal. At the same time, the informal differences between organizations regarding
92 Plant breeding and farmer participation
their access to external funding, and the The farmers often play a rather more con-
fact that the institution that successfully sultative role, giving input into variety
acquires funds is usually alone accountable evaluation, prioritization of selection crite-
towards the donors, which often impedes ria, and the necessary insights required for
a real sharing of project responsibilities focusing the project. However, as partners
among the partners (Kolanoski, 2003). gain experience, and the scale at which the
Notwithstanding, for the process of project operates increases, projects tend to
priority setting, it appears recommend- develop towards a strengthened role for
able to look deeper into the key skills and farmers or their organizations, especially
resources (material and non-material) each in terms of selection decisions and variety
partner or partner organization has to offer evaluation.
for reaching the identified project goals, for If farmers, especially a farmer organiza-
example, with regard to several issues: tion, initiate a plant breeding project, it tends
Overall project management, including to be clear that they seek specific support or
decision-making processes, monitoring input from scientists to find solutions to
and evaluation, reporting, public rela- problems already well identified. In addi-
tions work at different levels, fund acqui- tion to specific technical support, scientists
sition and management. can make contributions to building farmers
Planning and implementation of practical skills with respect to obtaining new germ-
project activities, such as trial manage- plasm; crop biology or physiology; specific
ment and data analysis, or seed produc- plant breeding activities, such as crossing;
tion and dissemination. variety evaluation; and interpretation of
Training and skill-building activities. results. In such situations, it is clear that the
On this basis, contracts between the role of scientists is primarily a consultative
various institutions could be negotiated, one, while key decisions are taken by the
which include tasks and duties with regard farmers or their organizations.
to the project, as well as the distribution of In situations where farmers are not
funds and resources among the partners. well organized, but project partners have
Furthermore, a pre-agreed procedure for identified farmer empowerment and skill
mediation or a conciliation board should building as a project goal, the project may
be foreseen in view of future cases of disa- invest major resources in the establishment
greement that might crop up between the of farmer organizations, committees or
partners. groups, which can then manage more of
the key breeding activities, and over time
4.5.2 Cooperation between farmers become the primary decision-makers, as
and scientists their skills and organizations grow. In such
In projects initiated by formal-sector a scenario, the role of the researchers may
breeding programmes, which are mostly change considerably over time, especially in
concerned with the traditional goals of terms of the management of trials, such as
breeding programmes, such as productivity decisions about which materials to continue
increases and possibly changes in policies with or to abandon, or which priorities for
for variety release or seed diffusion, most of selection to add to the project. Usually
the decision-making about the project tends these changes are also accompanied by a
to be initially in the hands of the scientists. change in the scale of the project. There
Methodologies for priority setting 93
4.6 PRACTICAL METHODS FOR 4.6.2 Identify the relevant issues for
PRIORITY SETTING priority setting
Priority setting for plant breeding The critical issues for priority setting in
programmes is, as such, not much different a plant breeding programme have been
94 Plant breeding and farmer participation
outlined in the first part of this chapter required to effectively reach the goals of the
(see Figure 4.1). All these issues need to breeding programme. In general, it will
be addressed by any plant breeding pro- have to include the following issues:
gramme, but there may not always be agro-ecological conditions;
viable alternatives to chose from. Besides, socio-economic conditions, including
the goals of a breeding programme can marketing of crop-based products;
change over time, reflecting the particular the farming system, actual processes of
context or situation; thus, priorities need to change and main limitations;
be reviewed regularly. farmers use of varieties and their seed
For the purpose of identifying relevant management;
options for the key issues, it may be help- seed system analysis; and
ful to examine the chances of success with specific varietal needs and preferences of
regard to each of the goals. This could be the target group(s).
done during a planning workshop, or also The situation analysis could include the
in the form of an e-mail discussion for those following steps:
partners who are using this communication 1. Review secondary sources.
technology. Furthermore, it is likely that 2. Consult local experts, key people
new options and insights emerge in the with good knowledge of the poten-
course of the practical project activities. tial target area(s).
Therefore, the process of priority setting 3. Visit potential target areas and con-
should be implemented in such a way that sult farmers belonging to different
insights and challenges can be addressed social and wealth groups.
at regular intervals, and then be integrated 4. Structure and compile the informa-
into previous concepts. tion for further planning.
Experience gained in a number of PPB
4.6.3 Situation analysis projects has shown that participatory
Realistic new options or technologies communication tools, such as semi-
require a good knowledge of the situation structured or informal interviews, focus-
under which they are intended to function, group discussions, wealth ranking, transect
including the needs and preferences of the walks, time lines, mapping, classification
potential users. Client-orientation is a key and ranking exercises, can be extremely
concept in the general economy, and increas- useful for providing a good basis for
ingly also in plant breeding (Witcombe et further planning. The particular strength
al., 2005). In the past, client-orientation was of such communication tools is that they
sometimes under-developed in plant breed- facilitate direct dialogue between farmers
ing, particularly as far as resource-poor and researchers, and can help to develop a
farmers in marginal areas were concerned. common understanding of the situation, as
A basic understanding of the complexity well as of the main constraints and needs.
of farming systems in such situations, as Practical guidelines for conducting such
well as their dependency on environmental a situation analysis, particularly for plant
adaptation and biodiversity, has now been breeding projects, have been suggested
developing, mainly since the mid-1990s. by Christinck, Weltzien and Hoffmann
The situation analysis for a plant breed- (2005). Furthermore, much inspiration
ing programme should focus on those issues can be gained from general guides and
Methodologies for priority setting 95
BOX 4.3
Web sites on participatory research methods
Sources of information and training materials are listed below. We concentrate here on those
publications that are available via the Internet, often for free download.
1. The Web sites of FAO (www.fao.org) and the World Bank (www.worldbank.org) con-
tain sections on publications for download and/or purchase (search for "participation"
or "PRA").
2. Further publications may be found via the online bookshop of UNEP (United Nations
Environment Programme) www.earthprint.com in the section on Participation and
training.
3. An introductory guide to participatory learning approaches can be down-
loaded free of cost from the GTZ homepage: Schnhuth, M. & Kievelitz, U.
1994. Participatory Learning Approaches. Rapid rural appraisal, Participatory
Appraisal an introductory guide. http://www2.gtz.de/dokumente/bib/95-0930.pdf
Other language versions (Spanish, French) are available upon request.
More specific publications on participatory research and learning are accessible
for download from: http://www.gtz.de/de/themen/uebergreifende-themen/partizipa-
tion/15201.htm (Accessed 12 September 2008).
4. Participatory Learning and Action (formerly PLA Notes) is a series on Participatory
Learning and Action (Methods and Approaches), accessible through the IIED homepage
(International Institute for Environment and Development, London, UK): http://www.
iied.org/NR/agbioliv/pla_notes/about.html#a (Accessed 12 September 2008).
5. The Programme for Participatory Research and Gender Analysis (PRGA) has a Web site
with a series of publications and resources, including a listing of cases for participatory
plant breeding. (www.prgaprogramme.org )
6. Reading University, UK, maintains a Web site with training materials and resources
focusing on the statistical analysis of data from participatory research activities: http://
www.reading.ac.uk/ssc/workareas/participation.html (Accessed 12 September 2008).
formal methods (Bellon and Reeves, 2002; their decision first, and then explain the
Abbeyasekera, 2002). reasons for their choice. Implicit reasons
can thus be made explicit and transparent.
4.6.4 Develop concrete new options More refined tools, which can consider
Developing new options for varieties several criteria simultaneously, may be used
requires creativity and good knowledge of once the key criteria are agreed.
the conditions under which a new variety
will have to function. It also requires good 4.6.6 Tools for farmer participation in
knowledge of the available diversity of the priority-setting process
the crop. Similarly, it may require detailed In this last section, we present a series of
understanding of options for new crop uses, tools that have been used successfully in
and for marketing of crop products, possi- one or more of the steps of the priority-
bly new ones. Traditionally, plant breeders setting process outlined above. Some may
have done this based on their own under- be used only for one specific step in the
standing of the farmers reality, especially decision-making process; others may apply
as many of the early private plant breeders to several of the steps. Many of the tools
were farmers themselves. Nowadays, when have been successfully used with farmers
plant breeders work on a national, regional for the identification of critical selection
or international scale, the development of criteria. The tools we choose to describe
new options for variety development, and are primarily those that can be used with a
seed distribution requires working creative- wide variety of partners, specifically with
ly with farmers and other project partners farmers, but also with those who may have
from various institutions and disciplines. very little time, may not be literate, but may
This is usually a continuing process, and have a profound knowledge of their culture
thus the project or programme should be and crop related issues. Many of the tools
organized in such a way that regular reviews are described in more detail and with more
of alternative new options can take place. examples in other sources, sometimes in
other contexts. Some good source materials
4.6.5 Making decisions among various are cited and listed. In most instances,
options considering the goals one would apply not only a single tool,
Making choices between the different options but several; it is advisable to vary the
needs to be forward looking, based on the tools for different steps of the priority-
identified project goals, and on chances for setting process, and also for the purpose of
success. Different stakeholders and partners verifying and increasing the reliability of
will have different perspectives, and thus previous results and hypotheses.
their choices and preferences for specific
options will vary. Hence it is important Facilitated discussions on goals, issues and
that the process of making decisions among criteria
an array of options is transparent, and Invite all relevant project partners to a
that the roles and responsibilities of the meeting on discussing goals for a new plant
different partners in the decision-making breeding programme. As the outcomes will
process are agreed. Ranking exercises are possibly depend on the circle of persons
ideal tools for taking decisions based on invited, the invitation list should be carefully
transparent criteria. Participants may make thought out. Furthermore, particularly if
Methodologies for priority setting 97
farmers are involved, the language, the breeding programme, particularly if many
general setting and the working style partners are involved.
(are all participants literate?) should be The meeting could then finish by priori-
considered with awareness. tizing the suggested goals, such as through
Depending on the number of participants, a simple ranking or scoring exercise (see
there are various options for facilitating below).
such a meeting. One option would be that In any case, such discussions on goals
the participants from each organization should be regarded as preliminary results.
are asked to prepare a short presentation, Many goals are not easily expressed
which would include a sort of problem and are closely related to individual or
analysis based on their own experience and culture-specific values. Moreover, goals
viewpoints, and should propose goals and may evolve in the course of the project
priorities. After the presentation, the main activities. It is thus recommended that this
goals mentioned in the presentation would discussion be repeated later, for example
be documented on a board. In this manner, after completing the situation analysis (see
there would be a preliminary list of goals Section 4.6.3, above), and particularly in
at the end, which could then be further view of the question of whether the goals
discussed. are really relevant for the target group.
Another way would be to start with a Regular discussions on goals and indicators,
brain-storming session or open discussion for example at the beginning of each new
on goals, and to document the proposed working phase, or in a general planning
goals on a board for further discussion. meeting, can be rewarding if a good facilitator
There should then be time to discuss helps to ensure productive outcomes.
these goals in more detail and clarify what
they imply. Very often, it helps if the SWOT analysis
participants are asked what kind of indica- A discussion about the overall goals and
tors they would suggest as a measurement more specific priorities involving key actors
of whether the future project activities or stakeholders can be structured in the for-
would be successful or not in reaching mat of an analysis of the present situation
these goals. Such indicators could thus of the crop under discussion and the devel-
also be useful for future monitoring and opment of future varietal options. A stra-
evaluation meetings. tegic planning tool for this type of analysis
It is of particular importance to identify is SWOT analysis, a structured discussion
potentially conflicting goals, or utopian on Strengths, Weaknesses, Opportunities
goals. In such cases, the group could try and Threats. This discussion could be held
to weigh up different goals, or to make as part of a project planning workshop, for
utopian goals more realistic and situation- example on the topic: Farmers groundnut
specific. In general, it is of course much varieties for the dry areas of Senegal, or
easier to reach a few clear goals with high any other crop and region.
priority on the agenda of all participants, The participants, either individually
than a long list of potentially conflicting or in small groups, are first asked to
goals. At the same time, the discussion think about the strengths of the situation
of goals can anticipate many problems under discussion. The results should be
that might occur in the course of a plant documented on a board or piece of paper
98 Plant breeding and farmer participation
(for later presentation to the whole group). pebbles, paper pieces, adhesive dots, etc.,
In the following steps, the participants and are asked to put their counter next to
also discuss weaknesses, opportunities and those goals with the highest priority for
threats. The results should be documented them. The goals should be well understood
visually on a board, and could then serve for this exercise, and the rules explained
as a starting point for discussion on goals carefully. Generally, each participant should
and priorities of a breeding programme (see have fewer counters than goals, so that a
also Weltzien, 2005). real decision has to be taken. It should be
clarified whether it is allowed to assemble
Recurrent feedback discussions all counters at one goal, the one perceived to
Successful project work depends on good be more important than any other, or if only
interaction between partners, e.g. research- one counter can be placed for each goal. In
ers from various institutions, farmers, and this manner, you will obtain a clear result
extension or NGO personnel. Feedback within a relatively short timea result on
discussions during which the different which further discussions can be based.
partners openly exchange their views and
experiences with specific project activities Ideal variety
should be held at regular intervals. These Invite a small group of participants, prefer-
discussions about what worked well, or ably 2 to 4, with whom you have already
which problems or opportunities arose, are discussed variety trials or the importance
the basis for reviewing the project priorities of specific traits in particular. Larger groups
in an evolving partnership between very could split up into separate working groups
different types of organization. While there and later present their results to the whole
may not necessarily be a fixed framework group. Invite each participant to think
for such discussions, they are instrumental about what a really good variety of the
in refining project priorities and in the evo- crop on which you are working could look
lution of the overall goals of a project and like, referring to the previous discussions
a partnership. Participatory Monitoring you have had. Focus group discussion,
and Evaluation (PM&E) would be a more where different groups represent farmers
institutionalized way of conducting such with differing backgrounds, farming situa-
feedback discussions (Germann, Gohl and tions, gender, ethnic groups, etc., can reveal
Schwarz, 1996). underlying differing needs.
Ask the participants to think about all
Simple scoring exercises the characters that a good variety of mil-
If you wish to set priorities among a number let, cowpea, etc., should have, to be use-
of possible goals, criteria, problems or ful for them. The traits mentioned by the
issues in a formal way, simple scoring exer- participants should be written on cards,
cises can be applied. This requires that a or the participants should find symbols
tentative list of goals and criteria is already for visual representation; the cards should
established. then be placed vertically in a column. Make
These goals should be written on a board sure that everybody contributes and that
or be represented visually in some form all the important traits are mentioned. In
(graphically or as text). All participants get the course of the exercise, you may also
a predefined number of counters, such as suggest some trait(s) if you are particularly
Methodologies for priority setting 99
interested in sparking off a discussion on Scenarios are only useful if the farmers
the relative importance of some new traits. reality is reasonably well understood. If
Once all the traits have been identified, the options or choices presented to farmers
you can then ask the farmers to discuss are not realistic, the responses cannot be
the importance of the trait for a new vari- expected to be realistic either.
ety that would be better than the existing
ones. To indicate the level of importance Example 1: Seed shop exercise
of each trait the farmers could distribute a The scenario is that the farmer who has no
fixed total number of tokens between the seed of this crop at the time of sowing enters
traits they (and you) have mentioned. The into a seed shop and has to choose among a
more important a trait, the more tokens it set of varieties with different properties.
receives. Traits that are not required should For this purpose, seed of different varie-
get no tokens, and can be eliminated. ties, local and introduced, is displayed in
It is best to facilitate this discussion in the shop, so that the farmers can see and
such a way that the participants primarily touch the seed. Variety names, plant sam-
discuss among themselves about each trait; ples or drawings of the plant type can pro-
for example, how early the ideal variety vide additional information. If you really
should be, or how much grain yield in rela- plan to give the seed to the farmers after
tion to stover yield they think would be the exercise, small packages in sufficient
useful. The difficulty is to try to keep the number should be prepared.
discussion within the realm of biological The farmers are asked to enter the shop
reality, i.e. not only grain yield, increased one by one, take their decision and leave
10-fold with half the growth duration of the shop; an interview on the reasons for
existing varieties. their choice will be conducted immediately
after leaving the shop.
Create scenarios The rules of the exercise should be made
Scenarios can be used to find out whether very clear at the beginning, particularly
certain concrete new options are attractive concerning questions such as whether the
for the target group(s) of farmers. This farmers will really get seed of the preferred
approach is particularly useful in the case of variety, how much, at what time (in the
complex or interrelated trait combinations. shop or afterwards) and from whom. Such
For this purpose, we need seed and plant rules potentially influence the result. They
material in which these new trait combina- should be carefully considered beforehand
tions are already expressed (i.e. exotic or and then announced very clearly to the par-
experimental varieties). ticipating farmers.
By simulating a situation in which farm-
ers have to take a decision between various Example 2: Simulating plant selection in
complex options, immediately followed a field
by an interview about the reasons, then The scenario here is that a farmer selects
important criteria and trade-offs may be plants from a field. This is very close to
revealed. Furthermore, this is also a way to the farmers reality in most cases. A further
study whether and why people belonging advantage of this scenario is that many dif-
to different groups take different decisions ferent traits, which may be relevant for the
regarding the proposed options. adaptation to specific conditions, different
100 Plant breeding and farmer participation
lines. A similar approach can also be pur- Discussions with farmers about their
sued with farmers. scoring will lead to a better understanding
Scores indicate a certain level of per- of selection criteria, preferences of specific
formance or expression of a trait. For user groups or for target growing condi-
example, the early vigour of varieties could tions, market demands, etc.
be assessed using a score, where 5 indicates
that a variety is extremely vigorous, 4 = Other tools used for priority setting
very vigorous, 3 = vigorous, 2 = less vigor- The tools described above are explained in
ous and 1 = not vigorous, or weak. Thus more detail in various training manuals and
scoring applies a fixed scale, as a tool for handbooks (Box 4.4) for farmer participatory
assessing potentially a large number of new rural appraisals. Economists tend to use
varieties or other options.
There is a fundamental difference
between scores and ranks, which can have BOX 4.4
far-reaching implications. For example, Training materials and books
ranking puts varieties in the order of per- on participatory research
formance or expression of a specific trait. methodologies in plant breeding
The best variety could actually have a fairly projects
poor performance, if all the other varieties
are still worse. The differences between 1. Bellon, M.R. & Reeves, J. 2002.
varieties could be very small, but they may Quantitative analysis of data from
lead to different ranks. Ranks do not have participatory methods in plant breeding.
an underlying scale, and thus quantitative Mexico, CIMMYT.
analysis is more difficult. Ranking can only 2 Christinck, A., Weltzien E. &
be done meaningfully with a small set of Hoffmann, V. 2005. Setting breeding
varieties (not more than seven) (Coe, 2002; objectives and developing seed
Weltzien and Christinck, 2005). systems with farmers. A handbook
Discussions on the reasons for giving for practical use in participatory plant
a particular score to a variety will reveal breeding projects. Margraf Publishers,
the underlying criteria. It is furthermore Weikersheim, Germany, and CTA,
possible to compare the scores given Wageningen, Netherlands.
by different groups of farmers (gender 3. IPRA & CIAT. 1991. Farmer evaluations
groups, people from different villages, of technology: Methodology for open-
etc.). ended evaluation. Instructional Unit
Practically, scoring exercises can be real- No. 1. IPRA, CIAT, Cali, Colombia.
ized in the field in various ways. Literate 4. Guerrero, M.P., Ashby, J.A. & Gracia, T.
participants can enter scores (= numbers) 1993. Farmer evaluations of technology:
in a previously prepared evaluation form. Preference ranking. Instructional Unit
Alternatively, one can use counters (stones, No. 2. IPRA, CIAT, Cali, Colombia.
pebbles, paper pieces), which have to be 5. Cleveland, D.A. & Soleri, D. 2002.
put into a basket, box or bag near the Farmers, scientists and plant breeding:
scored plot. More detailed descriptions Integrating knowledge and practice.
and examples can be found in Weltzien and Wallingford, UK, CABI.
Christinck (2005).
102 Plant breeding and farmer participation
other tools, such as Decision Trees, Grid paper. Statistical Services Centre, University
Analysis or Hedonic Pricing Models, for of Reading, UK (available at http://www.
priority setting and the identification of reading.ac.uk/ssc/workareas/participation/
specific selection criteria. These tools have How_to_generate_stats_and_influence_
rarely been applied specifically to plant policy.pdf; accessed 26 Dec. 2008).
breeding programmes, with the important Becker, H. 1993. Pflanzenzchtung. Stuttgart,
exception of the hedonic pricing model, Germany, Eugen Ulmer Verlag.
which has been used in a number of instances Bellon, M.R. & Reeves, J. 2002. Quantitative
(e.g. Dalton, 2004; Faye et al., 2004). These analysis of data from participatory methods
quantitative analytical tools can also be in plant breeding. Mexico, CIMMYT.
used to analyse data from specifically set Ceccarelli, S., Grando, S. & Booth, R.H. 1996.
up scenarios, or from ranking or scoring International breeding programmes and
exercises. More examples for combining resource-poor farmers: Crop improvement
qualitative and quantitative tools can be in difficult environments. In P. Eyzaguirre
found in Bellon and Reeves (2002) or & M. Iwanaga, eds. Participatory plant
Barahona and Levy (2002). breeding, pp. 99116. Proceedings of a
workshop on participatory plant breeding,
REFERENCES CITED Wageningen, Netherlands, 2629 July 1995.
Abbeyasekera, S. 2002. Quantitative analysis Rome, IPGRI.
approaches to qualitative data: why, when Ceccarelli, S., Grando, S., Tutwiler, R., Baha,
and how. Paper presented at the Conference J., Martini, A.M., Salahieh, H., Goodchild,
on Qualitative and Quantitative Methods A. & Michael, M. 2000. A methodological
in Development Research. Centre for study on participatory barley breeding. I.
Development Studies, University of Wales, Selection phase. Euphytica, 111(2): 91104.
Swansea, UK, 12 July 2002 (available at Christinck, A. & Weltzien, E. 2005. Identifying
http://www.reading.ac.uk/ssc/workar- target regions and target groups. In
eas/participation/Quantitative_analy- Christinck, Weltzien and Hoffmann, 2005,
sis_approaches_to_qualitative_data.pdf; q.v.
accessed 24 December 2008). Christinck, A., Dhamotharan M. & Weltzien,
Ashby, J. & Sperling, L. 1995. Institutionalizing E. 2005. Characterizing the production sys-
participatory, client-driven research and tem and its anticipated changes with farm-
technology development in agriculture. ers. In Christinck, Weltzien and Hoffmann,
Development and Change, 26: 753770. 2005a, q.v.
Atlin G., Paris, T. & Courtois, B. 2002. Sources Christinck, A., vom Brocke, K. & Weltzien,
of variation in participatory varietal selec- E. 2000. What is a variety: Investigating
tion trials with rainfed rice: Implications for farmers concepts as a base for participa-
design of mother-baby trial networks. In tory plant breeding in Rajasthan, India. In
M.R. Bellon & J. Reeves, eds. Quantitative International Agricultural Research. A con-
analysis of data from participatory meth- tribution to crisis prevention. Conference
ods in plant breeding, pp. 3643. Mexico, on Tropical Agriculture and Forestry,
CIMMYT. University of Hohenheim, Stuttgart,
Barahona, C. & Levy, S. 2002. How to gen- Germany, 1112 October 2000 (avail-
erate statistics and influence policy using able at http://www.prgaprogram.org/ppb_
participatory methods in research. Working resources/ppb_christinck_00.pdf)
Methodologies for priority setting 103
Christinck, A., Weltzien, E. & Dhamotharan, analysis. PRGA Programme and CIAT,
M. 2005. Understanding farmers seed man- Cali, Colombia.
agement strategies. In Christinck, Weltzien Faye, M.D., Jooste, A., Loewenberg-Deboer, J.
and Hoffmann, 2005, q.v. & Fulton, J. 2004. The influence of cowpea
Christinck, A., Weltzien E. & Hoffmann, V. characteristic on cowpea prices in Senegal.
2005. Setting breeding objectives and devel- AGREKON, 43: 418425.
oping seed systems with farmers. A hand- Germann, D., Gohl, E. & Schwarz, B. 1996.
book for practical use in participatory plant Participatory impact monitoring. Four vol-
breeding projects. Weikersheim, Germany, umes: (1) Group-based impact monitor-
Margraf Publishers, and Wageningen, ing; (2) NGO-based impact monitoring;
Netherlands, CTA. (3) Application examples; and (4) The con-
Coe, R. 2002. Analyzing Ranking and Rating cept of participatory impact monitoring.
data from participatory on-farm trials. GATE/GTZ, Braunschweig, Germany.
pp. 4465, in M.R. Bellon and J. Reeves, Goyal, S.N., Joshi, A. & Witcombe, J.R. 2001.
eds. Quantitative analysis of data from Participatory crop improvement in maize
participatory methods in plant breeding. in Gujarat, India. In An exchange of expe-
Mexico, CIMMYT. riences from South and South East Asia,
Cooper, M. & Byth, D.E. 1996. Understanding pp. 237241. Proceedings of the interna-
plant adaptation to achieve systematic tional symposium on participatory plant
applied crop improvement a fundamental breeding and participatory plant genetic
challenge. pp. 523, in M. Cooper and G.L. resource enhancement, Pokhara, Nepal, 15
Hammer, eds. Plant adaptation and crop May 2000. PRGA Programme and CIAT,
improvement. Wallingford, UK, CABI, for Cali, Colombia.
ICRISAT and IRRI. Guerrero, M.P., Ashby, J.A. & Gracia, T.
Cooper, M., Rajatasereekul, S., Immark, S., 1993. Farmer evaluations of Technology:
Fukai, S. & Basnayake, J. 1999. Rainfed low- Preference Ranking. Instructional Unit
land rice breeding strategies for northeast No. 2. IPRA, CIAT, Cali, Colombia.
Thailand. I. Genotypic variation, and geno- HarvestPlus. 2007. Web site. See http://www.
type environment interactions for grain harvestplus.org/about.html (Accessed 16
yield. Field Crops Research, 64: 131151. December 2008).
Dalton, T.J. 2004. A household hedonic model Haussmann, B.I.G., Parzies, H.K., Presterl,
of rice traits: economic values from farm- T., Sui, Z. & Miedaner, T. 2004. Plant
ers in West Africa. Agricultural Economics, genetic resources in crop improvement.
31: 149159. Plant Genetic Resources, 2(1): 321.
FAO. No date. FAOSTAT data from www. ILEIA [Centre for Information on Low
fao.org/faostat (Accessed May 2006). External Input and Sustainable Agriculture].
FAO. 1996. Global Plan of Action for the 2000. Special issue: Grassroots innovation.
Conservation and Sustainable Use of LEISA Magazine, 16(2).
Plant Genetic Resources and the Leipzig IPGRI/GFU/MSSRF. 2005. UN Millennium
Declaration. Rome (available at http:// Development Goals Five years later
www.fao.org/ag/AGP/AGPS/Pgrfa/pdf/ Agricultural Biodiversity and Elimination of
gpaeng.pdf; accessed 4 February 2009). Hunger and Poverty. The Chennai Platform
Farnworth, C.R. & Jiggins, J. 2003. for Action. Rome, IPGRI, and Chennai,
Participatory plant breeding and gender India, MSSRF (available at: http://www.
104 Plant breeding and farmer participation
improvement, pp. 383402. Wallingford, Weltzien, E., Smith, M.E., Meitzner, L.S.
UK, CABI, for ICRISAT and IRRI. & Sperling, L. 2003. Technical and insti-
van Oosterom, E.J., Weltzien, E., Yadav, O.P. tutional issues in participatory plant
& Bidinger, F.R. 2006. Grain Yield compo- breeding from the perspective of for-
nents of pearl millet under optimum condi- mal plant breeding. A global analysis of
tions can be used to identify germplasm issues, results, and current experience. PPB
with adaptation to arid zones. Field Crops Monograph No. 1. PRGA Programme,
Research, 96: 407421. Cali, Colombia.
Weltzien, E. 2005. Research on priority setting Witcombe, J.R., Joshi, K.D., Gyawali, S., Musa,
with farmers. In Christinck, Weltzien and A.M., Johansen, C., Virk, D.S. & Sthapit,
Hoffmann, 2005, q.v. B.R. 2005. Participatory plant breeding is
Weltzien, E. & Christinck, A. 2005. Identifying better described as highly client-oriented
farmers needs and preferences for varieties plant breeding. I. Four indicators of client-
and specific traits. In Christinck, Weltzien orientation in plant breeding. Experimental
and Hoffmann, 2005, q.v. Agriculture, 41: 299319.
107
CHAPTER 5
Methodologies for
generating variability.
Part 1: Use of genetic resources
in plant breeding
distinguishable,
Decreasing Diversity
Improved
wide adaptation)
cultivars
resources in participatory plant breeding,
Variety development,
release and seed production such as management of diversified popu-
Adapted breeding populations lations and their potential contribution to
Evaluation, pre -breeding in situ PGR conservation; the use of lan-
Materials in the collections draces as genetic resources for adaptation
Discovery and collection to stress environments, climate variability
Available diversity of the crop species in and ex situ
and climate change; and to better serve
Source: FAO, 1996 farmers and end-users diverse needs.
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 109
FIGURE 5.3
The extended gene pool concept for classification of PGR
Secondary
gene pool Related species, that can be
crossed only using special
Tertiary methods (e.g. embryo rescue,
gene pool protoplast fusion)
Source: modified from Becker, 1993, and Harlan and de Wet, 1971.
110 Plant breeding and farmer participation
FIGURE 5.4
The concept of core collection
Collection containing
200 accessions in the
primary, secondary
and tertiary gene pool
Core consisting of
20 accessions
FIGURE 5.5
Some key clauses of the Standard Material Transfer Agreement (SMTA)
The Recipient may utilize and conserve the material for research,
breeding or training purposes .
The Recipient shall not claim any intellectual property or other rights
that limit the facilitated access to the Material provided under this
Agreement, or its genetic parts or components.
FIGURE 5.6
Overview over methods for using PGRs in plant breeding
FIGURE 5.7
Classical back-crossing for transfer of qualitative traits from a donor into a recipient
F 1 = BC 0 0.5
Selected BC 1 0.75
Selected BC 2 0.875
Selected BC 3 0.938
FIGURE 5.8
The concept of dynamic gene pool management
Selection cycle
Selection in in
cycle the
the
Genetic sub -populations
resources C0 C1 ... Cn
Sum of sub -
E1 Cold ... populations =
E2 Drought ... Mass reservoir
Diversied E3 Poor soil ... of adaptation
base E4 Disease ...
population Each sub -
E5 Flooding ...
population =
E6 Irrigation ... potential cultivar
E7 High input ... and source of
variation for
Division into Cultivation of subpopulations in specic characters
subpopulations contrasting environments (E)
on-farm and on-station, recurrent
selection (nature, farmer, breeder)
Contributes to in situ conservation of genetic resources!
(Simmonds, 1993; Cooper, Spillane and materials, and selection for yield traits
Hodgkin, 2001) and is also understood as carried out in the F2 (50% exotic genome)
a means of in situ maintenance of PGR or BC1 (25% exotic) generation. The
(Figure 5.8). Examples are the barley optimal percentage of the exotic genome
(Hordeum vulgare) composite cross of the genetic resource (100%, 50% or
developed at Davis, California, United 25%) in a breeders population depends
States of America (Cooper, Spillane and on the overall objective; time available
Hodgkin, 2001), dynamic gene pool and finances; the level of adaptation of the
management in wheat (Goldringer et al., genetic resource; and the yield difference
2001); pearl millet (Pennisetum glaucum) between the genetic resource and the actual
composite populations developed in Africa breeding population. Direct adaptation of
(Niangado, 2001); and the development the PGR takes usually longer than selection
of locally adapted farm cultivars for in F2 or BC1 (due to lack of adaptation
ecological agriculture in Europe (Mller, of the PGR) but will result in materials
1989). that are genetically quite different from
In the short term, genetic enhancement the actual breeding materials, which can
of breeding materials, genetic resources be an advantage. Selection in BC1 may be
are selected for desirable agronomic traits preferred over selection in an F2 population
and yield performance, but not for the if the PGR is highly unadapted to the target
highest degree of genetic diversity. They are environment. At the same time, selection
intercrossed, recombined and then selected in the F2 population is expected to reveal a
for adaptation to the target environment. higher genetic variance, a component of the
To speed up the process, selected PGR expected gain from selection (Bridges and
may also be crossed with the breeding Gardner, 1987).
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 115
exploitation of valuable QTLs from PGR terms of the heterozygosity, i.e. the average
by advanced backcross QTL analysis number of heterozygous gene loci.
to combine QTL analysis with the At the population level, protein markers
development of superior genotypes or by and DNA markers are commonly used to
marker-assisted, controlled introgression calculate, among others, (i) allelic diversity
of PGR into breeding materials through or allelic richness (A; the mean number
the development of introgression libraries; of alleles per locus); (ii) percentage of
and polymorphic loci (P; the mean proportion
association studies to mine directly the of polymorphic loci); (iii) Neis average
allelic diversity of PGR collections and gene diversity (He; which denotes the
to identify those alleles that are beneficial probability that two randomly chosen
for important agronomic traits. alleles at a certain locus from a population
are different. It is the generalized form of
5.7.1 Diversity assessment expected heterozygosity assuming Hardy-
For an efficient diversity assessment, Weinberg Equilibrium and thus often
molecular markers ideally need to be abbreviated as He); and (iv) Shannons
selectively neutral, highly polymorphic, index of diversity (H), which is widely used
co-dominant, well dispersed throughout in ecology but also applied to population
the genome, and cost- and labour-efficient genetics (Lowe, Harris and Ashton, 2004).
(Bretting and Widerlechner, 1995). Genetic With the employment of DNA point
markers complying with these requirements mutations, such as single nucleotide
are protein markers (i.e. iso-enzymes) polymorphisms (SNPs) and small DNA
and DNA markers such as Restriction Insertion/Deletions (InDels) as markers
Fragment Length Polymorphisms (RFLPs) for diversity studies, a number of indices
and Microsatellites or Simple Sequence have been put forward for variants of a
Repeats (SSRs). Because the development certain DNA sequence in a population.
of the latter two marker types requires prior These are (i) the number of polymorphic
knowledge of DNA sequences, a number of (segregating) sites (S); (ii) total number of
universal, dominant molecular marker types mutations (Eta); (iii) number of haplotypes
such as Random Amplified Polymorphic (h); (iv) haplotype (gene) diversity (Hd);
DNA (RAPD) and Amplified Fragment (v) nucleotide diversity (Pi; the average
Length Polymorphisms (AFLPs) have also number of nucleotide differences per
been employed in PGR diversity studies. site between two sequences; Nei, 1987);
However, the latter are not suitable for (vi) nucleotide diversity (Pi (JC); the
assessing factors such as mating behaviour average number of nucleotide substitutions
or heterozygosity of the germplasm. per site between two sequences (Lynch
Generally, genetic diversity can be and Crease, 1990, cited by Rozas et al.,
measured on three levels: in individual 2003); (vii) Watterson estimator Theta
plants, within populations (intrapopulation) (Watterson, 1975); on a base-pair basis it
and between populations (interpopulation), can be interpreted as 4N for an autosomal
while populations are considered as groups gene of a diploid organism, where N and
of randomly interbreeding individuals of are the effective population size and
one species. The diversity of individual the mutation rate per nucleotide site per
plants is most commonly characterized in generation, respectively); and (viii) average
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 117
be expected in the hybrid resulting from a backcrossing with the same proportion of
cross between them (Melchinger, Coors and the recurrent parent genome (Ragot et al.,
Pandey, 1999; Reif et al., 2003a, b). Yet, the 1995; Frisch, Bohn and Melchinger, 1999).
effect on heterosis and hybrid performance MAS for multigenic, quantitative
needs to be distinguished, since high traits at first requires the identification of
heterosis does not necessarily mean high the genomic regions (QTLs) that affect
hybrid yield. Recent studies have shown that the trait of interest. In classical QTL
the correlation between diversity measures mapping, a segregating population (e.g.
and hybrid performance gets stronger when F2, F3 or recombinant inbred population)
the markers used for diversity assessment is developed from two inbred lines. This
are linked to performance QTLs, rather mapping population is evaluated for the
than from using neutral markers (Vuylsteke, trait(s) of interest. Simultaneously, the
1999; Vuylsteke, Kuiper and Stam, 2000; population is genotyped with a number of
Jordan et al., 2003). markers and a genetic map is constructed
from the marker data. In the final QTL
5.7.2 Genetic mapping and marker- analysis, data is analysed for co-segregation
assisted selection of particular markers with the trait of
Marker-assisted selection (MAS) can help interest. QTL analysis is then followed by
(i) to select individuals carrying molecu- transfer of favourable QTL alleles into elite
lar markers that are linked to the trait of materials via pure MAS or MAS combined
interest, instead of performing extensive with phenotypic selection.
phenotypic tests (foreground selection); However, for complex, quantitative
and (ii) to reduce undesired parts of the traits, the efficiency of QTL mapping and
donor genome, including the linkage drag MAS is contested. There are a number
(background selection). Foreground selec- of risks that can render MAS inefficient.
tion requires a tight linkage between the For example, there may be no selection
trait of interest and its flanking markers for gain because of: unreliable QTL estimates
which one is selecting. Background selec- (too few QTLs, with highly over-estimated
tion necessitates genotyping with a larger effects); QTLs not being expressed in
number of markers, which cover the whole new genetic backgrounds; recombination
genome. between marker and QTL; unfavourable
MAS has proven efficient for the transfer alleles of other genes linked to good
of simply inherited qualitative traits from QTL alleles; or too-high costs for marker
genetic resources into elite materials using analyses. It is therefore essential to use
backcrossing procedures. It is particularly large mapping populations; genotype the
useful for traits that are recessive, that can mapping population with good genome
be assessed only after flowering or that coverage; assess phenotypic values in
are very difficult and expensive to assess. multi-environment field trials; cross-
By using a combination of foreground validate the gained data; verify QTL effects,
and background selection, the transfer of using independent population samples,
a monogenic trait from a genetic resource near-isogenic lines or different genetic
into a breeding line may be completed backgrounds; ensure close linkage between
within three to four generations, instead marker and QTL, and verify the linkage by
of the usual six generations of classical a phenotypic test in all 3 or 4 generations;
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 119
increase the marker density around the QTL and selfing generations are needed to derive
to allow reduction of the linkage drag; and QTL-bearing near-isogenic lines (QTL-
to optimize individual procedures while NILs). These carry recurrent parent alleles
taking into account economic parameters. throughout their genome except for the
For quantitative traits, where many loci specific target QTL (Tanksley and Nelson,
of minor effects are responsible, it is very 1996). QTL-NILs can be used to verify
difficult to obtain reliable, unbiased QTL observed QTL effects as well as commer-
estimates (e.g. Beavis, 1998; Melchinger, cial lines improved for one or more quan-
Utz and Schoen, 1998; Utz, Melchinger titative traits compared with the original
and Schn, 2000). Prospects for MAS are recurrent elite line.
therefore more promising for traits that are In contrast to the AB-QTL method,
determined by few QTLs with large effects Eshed and Zamir (1994, 1995) suggested
(Melchinger, 1990). the approach of establishing a population
of NILs such that the donor chromosome
5.7.3 Advanced backcross QTL analysis segments are evenly distributed over the
and introgression libraries whole recipient genome. Ideally, the total
QTL analysis can also be performed in genome of the exotic donor is comprised
backcross generations derived from crosses in the established set of NILs (Figure 5.9).
of exotic PGR with elite materials. The This NIL population, termed an introgres-
Advanced Backcross QTL Analysis (AB- sion library, consists of a set of lines, each
QTL; Tanksley and Nelson, 1996) com- carrying a single marker-defined donor
bines QTL analysis with the development chromosome segment introgressed from
of superior genotypes and has been shown an agriculturally unadapted source into
to be particularly useful for a trait transfer the background of an elite variety (Zamir,
from poorly adapted germplasm. AB-QTL 2001).
is therefore of special importance in the use The procedure of establishing an intro-
of PGR for crop improvement. The starting gression library implies systematic transfer
point is a segregating generation of a cross of donor chromosome segments from a
between an exotic parent and an elite line PGR (donor) into an elite line (recur-
that is analysed with as many molecular rent parent) by marker-aided backcrossing.
markers as possible. QTL mapping proce- Additional self-pollination and marker-
dure is delayed until one of the advanced based selection lead to NILs homozygous
backcross generations (BC2) when lines at donor chromosome segments. Such
or testcrosses are evaluated across environ- NILs differ from the elite line by only a
ments. small, defined chromosomal segment, and
To date, the AB-QTL strategy has been phenotypic differences between a line in
applied in several crops, including toma- the library and the nearly isogenic elite line
to, rice and barley (Tanksley et al., 1996; are associated with the single donor chro-
Fulton et al., 1997, 2000; Bernacchi et al., mosome segment (imi et al., 2003).
1998; Xiao et al., 1996, 1998; Moncada et Both introgression library and AB-QTL
al., 2001; Pillen, Zacharias and Lon, 2003, approaches provide a valuable opportunity
von Korff et al., 2008). Once favourable to extract quantitative trait alleles for mod-
QTL alleles from an exotic donor are iden- ern crop varieties from exotic PGR. Their
tified, one or two additional backcrossing main advantage is that the exotic genome is
120 Plant breeding and farmer participation
FIGURE 5.9
Graphical illustration of an introgression library
Chromosome
1 2 3 4 5 6 7
Individual lines (N=39)
introgressed into the elite line only as small, pers. comm.). The major advantages of
well defined donor chromosome segments. association studies over classical QTL map-
This reduces unfavourable effects that often ping experiments is that no segregating
impede the use of PGR in practical breed- population has to be established from two
ing programmes. inbred lines, and that the results are not
limited to the specific mapping population
5.7.4 Association studies and direct but can cover the full allelic variation avail-
allele selection able in natural or breeding populations or
Increased insight into the molecular organi- gene bank accessions (Jannink, Bink and
zation and sequence of plant genomes has Jansen, 2001; Jannink and Walsh, 2002).
led to new methods to mine directly the Associations between DNA sequence
allelic diversity of PGR. The aim of such polymorphisms and phenotypic trait
studies is to associate sequence polymor- variation can occur either when the
phisms within genes or across genomes polymorphisms are directly responsible
with phenotypic variants to detect superior for the functional differences between the
alleles affecting agronomically important alleles of the respective genes, or when the
traits. Such valuable alleles detected within analysed polymorphisms are in linkage
germplasm collections can subsequently disequilibrium (LD) with the functional
be transferred to elite breeding materi- alleles. LD is defined as a non-random
als via marker-assisted backcrossing using association of alleles at different loci within
allele-specific markers (direct allele selec- a population (Falconer and Mackay, 1996).
tion; Sorrells and Wilson, 1997) or marker- The basic idea of association mapping
assisted recurrent selection (D. Hoisington, can be investigated using two strategies.
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 121
FIGURE 5.10
Steps in an association study using the candidate-gene approach
One approach is first to identify candi- level and structure of LD. Low levels of LD
date genes (i.e. from available databases or would be favourable for high resolution fine
gene expression studies) and to re-sequence mapping within candidate genes, but limit
those candidate genes in plants derived from the feasibility of genome-wide association
diverse germplasm accessions (Figure 5.10). studies. A first attempt to use the genome-
The maize gene dwarf8, a candidate gene wide approach in plants was reported for
for flowering time and plant height, was Beta vulgaris subsp. maritima using 440
used by Thornsberry et al. (2001) in a first AFLP markers in 106 individual plants
association study with a crop species. They from four natural populations (Hansen
sequenced dwarf8 in a representative set of et al., 2001). Two markers were detected
92 inbred lines and found polymorphisms showing significant association with the
within the gene to be strongly associ- bolting gene, which is responsible for the
ated with flowering time. This group of vernalization requirement.
researchers also developed a software suite, Population structure in germplasm col-
TASSEL, (http://www.maizegenetics.net/ lections, which may be unknown to the
bioinformatics/index.htm) for analysing researcher, can cause spurious associations.
LD and for performing association map- Statistical methods were developed by
ping in populations of inbred lines. Pritchard, Stephens and Donnelly (2000)
A second approach is to analyse a set of and Falush, Stephens and Pritchard (2003)
randomly chosen molecular markers, evenly to detect such population structures using
distributed across the genome. If such a few molecular markers evenly spread
markers are in LD with the genes controlling across the genome. Removing the effects of
the trait variation, one will also detect a population structure increases the power
significant association. The practicability of the association study to detect useful
of this approach strongly depends on the markers.
122 Plant breeding and farmer participation
5.8 THE USE OF GENETIC RESOURCES pressures of a target region (see Figure 5.8
IN PARTICIPATORY PLANT BREEDING above). Natural and recurrent selection by
Genetic resources can be used in a number farmers and breeders will act on the distrib-
of ways in participatory plant breeding uted material and lead to the development
programmes. of new subpopulations that can be excellent
sources of variation for specific adaptation
Participatory improvement of diversified and farmer-preferred traits, as well as new
populations and potential contribution to trait combinations (via recombination) not
in situ conservation of PGR previously available. Such a dynamic gene
Farmer-participatory improvement of pool approach provides the best opportu-
diversified populations combines in situ nity to offer a wide diversity of material to
conservation with genetic improvement of the wide diversity of farmers for effective
PGR to meet farmers diverse needs as participatory plant breeding (Weltzien et
well as the challenges of adaptation to al., 2000).
site-specific conditions, climatic variability
and climate change. In a first step, farmers Use of landraces as genetic resources for
may evaluate a range of diverse varieties or specific adaptation to stress environments,
germplasm accessions of the target crop and climate variability and climate change,
chose accessions that carry traits of interest and to better serve farmers and end-users
to them. The diversified base population diverse needs
will then be built through crossing and Breeding for wide adaptation has been
recombining the farmer-selected materials. found to be inappropriate for extreme stress
Representative seed lots of targeted base environments, because of cross-over geno-
populations will be distributed to farmers type environment interactions appearing
in contrasting sites with specific selection at low yield levels (e.g. Simmonds, 1991;
FIGURE 5.11
Example of the utility of local landraces as source of adaptation to extreme growing conditions
140
120
100 Extreme stress
80
60 Landrace
40
Improved
20
cultivar
0
Jodhpur 1997 Mandore 1998
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 123
Lowe, A., Harris, S. & Ashton, P. 2004. Production. pp. 147157. Wallingford, UK,
Ecological Genetics. Design, Analysis CABI Publishing, in cooperation with FAO
and Application. Oxford, UK, Blackwell and IPGRI.
Publishing. Pillen, K., Zacharias, A. & Lon, J. 2003.
Melchinger, A.E. 1990. Use of molecular Advanced backcross QTL analysis in bar-
markers in breeding for oligogenic disease ley (Hordeum vulgare L.). Theoretical and
resistance. Plant Breeding, 104: 119. Applied Genetics, 107: 340352.
Melchinger, A.E., Utz, H.F. & Schoen, C.C. Pritchard, J.K., Stephens, M. & Donnelly,
1998. Quantitative trait locus (QTL) map- P. 2000. Inference of population structure
ping using different testers and independ- using multilocus genotype data. Genetics,
ent population samples in maize reveals 155: 945959.
low power of QTL detection and large Ragot, M., Sisco, P.H., Hoisington, D.A.
bias in estimates of QTL effects. Genetics, & Stuber, C.W. 1995. Molecular-marker-
149(1): 383403. mediated characterization of favorable
Melchinger, A.E., Coors, J.G. & Pandey, S. exotic alleles at quantitative trait loci in
1999. Genetic diversity and heterosis. In maize. Crop Science, 35: 13061315.
The genetics and exploitation of heterosis in Reif, J.C., Melchinger, A.E. & Frisch, M. 2005.
crops. pp. 99118. Proceedings of an interna- Genetical and mathematical properties of
tional symposium, CIMMYT, Mexico City, similarity and dissimilarity coefficients
Mexico, 1722 August 1997. CIMMYT, applied in plant breeding and seed bank
Mexico. management. Crop Science, 45: 17.
Mohammadi, S.A. & Prasanna, B.M. 2003. Reif, J.C., Melchinger, A.E., Xia, X.C.,
Analysis of genetic diversity in crop plants Warburton, M.L., Hoisington, D.A., Vasal,
Salient statistical tools and considerations. S.K., Beck, D., Bohn, M. & Frisch, M.
Crop Science, 43: 12351248. 2003a. Use of SSRs for establishing heterot-
Moncada, P.P., Martnez, C.P., Borrero, J., ic groups in subtropical maize. Theoretical
Chatel, M., Gauch, H. Jr, Guimaraes, and Applied Genetics, 107: 947957.
E., Tohme, J. & McCouch, S.R. 2001. Reif, J.C., Melchinger, A.E., Xia, X.C.,
Quantitative trait loci for yield and yield Warburton, M.L., Hoisington, D.A., Vasal,
components in an Oryza sativa Oryza S.K., Srinivasan, G., Bohn, M. & Frisch, M.
rufipogon BC2F2 population evaluated in 2003b. Genetic distance based on simple
an upland environment. Theoretical and sequence repeats and heterosis in tropical
Applied Genetics, 102: 4152. maize populations. Crop Science, 43: 1275
Mller, K.J. 1989. Die Entwicklung von 1282.
Basispopulationen bei Getreide fr biol- Rozas, J., Sanchez-Delbarrio, J.C., Messeguer,
ogisch-dynamische Betriebe. Zeitschrift X. & Rozas, R. 2003. DnaSP, DNA poly-
Lebendige Erde, 5(89): 364367. morphism analyses by the coalescent and
Nei, M. 1987. Molecular Evolutionary other methods. Bioinformatics, 19: 2496
Genetics. New York, USA, Columbia Univ. 2497.
Press. Snchez-Monge, E. 2002. Diccionario de
Niangado, O. 2001. The state of millet diver- plantas de inters agrcola. Ministerio de
sity and its use in West Africa. In H.D. Agricultura, Pesca y Alimentacin, Madrid.
Cooper, C. Spillane & T. Hodgkin, eds. imi, D., Presterl, T., Seitz, G. & Geiger,
Broadening the Genetic Base of Crop H.H. 2003. Comparing methods for inte-
Methodologies for generating variability. Part 1: Use of genetic resources in plant breeding 127
grating exotic germplasm into European explained by quantitative trait loci deter-
forage maize breeding programs. Crop mined from experimental data in maize
Science, 43: 19521959. using cross validation and validation with
Simmonds, N.W. 1991. Selection for local independent samples. Genetics, 154: 1839
adaptation in a plant breeding programme. 1849.
Theoretical and Applied Genetics, 82: 363 Van Hintum, T.J.L., Brown, A.H.D., Spillane,
367. C. & Hodgkin, T. 2000. Core collections of
Simmonds, N.W. 1993. Introgression and plant genetic resources. IPGRI Technical
incorporation. Strategies for the use of Bulletin, No. 3: 548.
crop genetic resources. Biological Reviews, vom Brocke, K., Christinck, A., Presterl, T.,
68: 539562. Weltzien-Rattunde, E. & Geiger, H.H.
Sorrells, M.E. & Wilson, W.A. 1997. Direct 2002a. Effects of farmers seed manage-
classification and selection of superior alle- ment on performance and adaptation of
les for crop improvement. Crop Science, pearl millet in Rajasthan, India. Euphytica,
37: 691697. 130: 267280.
Swaminathan, M.S. 2002. The past, present vom Brocke, K., Presterl, T., Christinck, A.,
and future contributions of farmers to Weltzien-Rattunde, E. & Geiger, H.H.
the conservation and development of 2002b. Farmers seed management prac-
genetic diversity. In J.M.M. Engels, V.R. tices open up new base populations for
Rao, A.H.D. Brown & M.T. Jackson, pearl millet breeding in a semi-arid zone of
eds. Managing Plant Genetic Diversity, India. Plant Breeding, 121: 3642.
pp. 2331. Rome, IPGRI, and Oxford, von Korff, M., Wang, H., Lon, J. & Pillen, K.
UK, CABI Publishing. 2008. AB-QTL analysis in spring barley:
Tanksley, S.D. & Nelson, J.C. 1996. Advanced III. Identification of exotic alleles for the
backcross QTL analysis: a method for the improvement of malting quality in spring
simultaneous discovery and transfer of barley (H. vulgare ssp. spontaneum).
valuable QTLs from unadapted germ- Molecular Breeding, 21(1): 8193.
plasm into elite breeding lines. Theoretical Vuylsteke, M.J.R. 1999. Genetic analysis
and Applied Genetics, 92: 191203. of maize by using the AFLP method.
Tanksley, S.D., Grandillo, S., Fulton, T.M., PhD Thesis, University of Wageningen,
Zamir, D., Eshed, Y., Petiard, V., Lopez, Netherlands.
J. & Beck-Bunn, T. 1996. Advanced back- Vuylsteke, M., Kuiper, M. & Stam, P. 2000.
cross QTL analysis in a cross between an Chromosomal regions involved in hybrid
elite processing line of tomato and its wild performance and heterosis: their AFLP-
relative L. pimpinellifolium. Theoretical based identification and practical use in
and Applied Genetics, 92: 213224. prediction models. Heredity, 85(3): 208
Thornsberry, J.M., Goodman, M.M., Doebley, 218.
J., Kresovich, S., Nielsen, D. & Buckler Watterson, G.A. 1975. On the number of
IV, E.S. 2001. Dwarf8 polymorphisms segregating sites in genetical models with-
associate with variation in flowering time. out recombination. Theoretical Population
Nature Genetics, 28: 286289. Biology, 7(2): 256276.
Utz, H.F., Melchinger, A.E. & Schn, C.C. Weltzien, E., Smith, M.E., Meitzner, L. &
2000. Bias and sampling error of the esti- Sperling, L. 2000. Technical and institu-
mated proportion of genotypic variance tional issues in participatory plant breed-
128 Plant breeding and farmer participation
ing from the perspective of formal plant Xiao, J., Grandillo, S., Ahn, S.N., McCouch,
breeding. A global analysis of issues, results S.R., Tanksley, S.D., Li, J. & Yuan, L.
and current experience. Working Document 1996. Genes from wild rice improve yield.
No. 3. CGIAR Systemwide Program Nature, 384: 223224.
on Participatory Research and Gender Xiao, J., Li, J., Grandillo, S., Ahn, S., Yuan,
Analysis for Technology Development and L., Tanksley, S.D. & McCouch, S.R. 1998.
Institutional innovation. 106 p. Identification of trait-improving quantita-
Witcombe, J.R. 2001. The impact of decentral- tive trait loci alleles from a wild rice rela-
ized and participatory plant breeding on the tive, Oryza rufipogon. Genetics, 150: 899
genetic base of crops. In H.D. Cooper, C. 909.
Spillane & T. Hodgkin, eds. Broadening the Zamir, D. 2001. Improving plant breed-
Genetic Base of Crop Production, pp. 407 ing with exotic genetic libraries. Nature
417. Wallingford, UK, CABI Publishing, Reviews Genetics, 2: 983989.
in cooperation with FAO and IPGRI.
129
CHAPTER 6
Methodologies for
generating variability
Part 2: Selection of parents
and crossing strategies
theory is clear that the optimum popula- number of crosses rather than increasing
tion size of any cross has to be large if the population size of each cross in order to
desirable transgressive segregants are to be minimize the risk of missing the favourable
obtained for traits that involve several loci plant. They found that each of the m
(Allard, 1999). However, most breeding crosses should be represented by only one
programmes use much smaller population F2 plant (n = 1) to give the lowest risk of
sizes than theory dictates to accommodate failure. However, this did not consider all
the many crosses that are made. of the possible scenarios. The probability of
success can increase to 1 long before n falls
6.2.1 A re-examination of the theory to 1. This happens when:
on the optimum cross number the probability of a cross succeeding
In theoretical determinations of the opti- (P1) is high, but still much lower than
mal number of crosses (m) and popula- what we have achieved in our few-cross
tion size (n) per cross, given a limit of K breeding programmes;
plants, two approaches have been used: the probability of a plant succeeding (P2)
either (i) minimizing the risk of excluding is also reasonably high; and
superior genotypes, or (ii) maximizing the K is large (Figure 6.1).
response to selection. In the second approach, the magnitude
Using the first approach, Yonezawa and of the response to selection among and
Yamagata (1978), Wricke and Weber (1986) within crosses is considered (Baker, 1984;
and Hehn (1996) suggested increasing the Wricke and Weber, 1986; Hehn, 1996;
FIGURE 6.1
The probability of success (left hand y axis) with a varying number of plants per cross
(right hand y axis log scale) for 1 to 600 crosses
1 100 000
Probability (P)
Number of plants per cross (n)
0.8
Probabaility of success
10 000
0.6
0.4
1 000
Plants per cross (n)
0.2
0 100
0 100 200 300 400 500 600
Cross number (m)
P1 (probability of cross succeeding) = 0.03, P2 (probability of a plant succeeding) = 0.005 and K (total size of
population) = 100 000 in all cases.
132 Plant breeding and farmer participation
FIGURE 6.2
Effect on the probability of success of varying the number of crosses (m) with four constant
total population sizes (K)
0.9
0.8
0.7
Probability of success
0.6
0
0 20 40 60 80 100 120 140 160 180 200
No. of crosses (m)
P1 (probability of cross succeeding) = 0.02, and P2 (probability of a plant succeeding) = 0.001 in all cases. Point A is at 6
crosses with K = 100 000 and point B is at 200 crosses with k = 5000.
Bernardo 2003) and fewer crosses where P2 are not extreme (Figure 6.2). Moreover,
n is always >1 maximizes the chance of when values of K are greater than this, e.g.
success. For example, Baker (1984) con- 100 000, then adding crosses improves the
cluded that for a total of 2 000 families probability of success very little after about
the maximum response is found with 50 15 crosses are made (Figure 6.2).
to 100 crosses having 40 to 20 families. Even though the model of Yonezawa
Therefore, Bernardo (2003) concluded that and Yamagata (1978) has been used to argue
the two approaches give conflicting results. the case for more crosses, the argument for
However, we have found that by examining increasing K is just as powerful. A few crosses
the response to increasing the cross number where K is large have a much lower risk over
(m) the conflict is not as great as it might many crosses when K is small. For example,
first appear. If the values of K are high, then 6 crosses (point A in Figure 6.2) where K =
adding more crosses is not very effective in 100 000 are less risky than 200 crosses (point
increasing the probability of success, and a B in Figure 6.2) where K = 5 000.
fewer-cross strategy with a larger popula- The relative costs of making crosses and
tion size would be more cost effective (see growing plants are ignored in the model
below for considerations of cost). Using the of Yonezawa and Yamagata (1978). Hence,
model of Yonezawa and Yamagata (1978), with one cross or many crosses K remains
the rate of increase in the probability of the same, even though it is easier and
success by adding more crosses declines cheaper to make one cross and grow 10 000
dramatically after about 50 to 100 when plants from it than make 10 000 crosses and
values of K are 50 000 and values of P1 and only grow one plant from each. This applies
Methodologies for generating variability. Part 2: Selection of parents and crossing strategies 133
to all other values, such as 200 crosses with single generation. However, the efficacy
50 plants each compared with 100 crosses of these methods can only be evaluated by
with 100 plants each. Not only is there the progeny tests.
additional cost of making more crosses, but In the second category of methods,
there is an additional cost in record keep- parents are evaluated on the basis of the
ing, planting and labelling the more crosses performance of their progeny. Such tests
there are within any given K. require time as at least two generations
In all of these models we have not con- of plants need to be grown and evaluated
sidered the ability of breeders to choose to determine means and variances. The
superior crosses rather than making random evaluation can be of combinations of F1, F2
ones. If the declining probability of success and later generations (Allard, 1956; Busch,
of each cross is consideredthe first choice Hanke and Frohberg, 1974; Cox and
of a breeder should be better and more Frey, 1984) or the evaluation of progeny
carefully considered than the hundredth produced from mating designs such as
then the optimum number of crosses falls. diallel and line tester (Lupton, 1965). The
This is simply a quantitative and realistic relationship between predicted and actual
extension of the argument that if the best progeny performance provides empirical
cross is known then only one needs to be evidence of the value of selecting parents
made. The optimum number of crosses by these methods, but the results of such
also becomes smaller as more resources are experiments provide no consensus.
spent on evaluating parents and more time Other authors have concluded that
is spent on choosing crosses. gathering experimental data to estimate
the value of possible crosses using progeny
6.3 CHOICE OF PARENTS tests demands so many resources that it
6.3.1 Selection of parents is unwarranted (Wricke and Weber, 1986;
Little consensus exists among plant breeders Lupton, 1965). We would agree with this
on how best to choose parents for crosses as participatory methods provide simpler
that will produce high yielding progenies and effective methods of choosing parents
and it remains a debated issue (Qualset, 1979; by using genotypic performance per se as a
Baker, 1984). The strategies for selection of prediction of parental value (Baenziger and
parents for desired progeny performance Peterson, 1992). This is done without using
fall into two categories: methods based on the formal quantitative analysis described
parental performance per se, or methods above for the first category of methods.
that assess the value of parents estimated Much information is already available for
from progeny performance. The first genotypes that have already been adopted
category includes selection based on: mid- by farmers. If a breeders knowledge of such
parental values; divergence coefficients genotypes can allow the prediction of the
among parents (Murphy, Cox and Rodgers, more useful cross combinations, then only
1986); character complementation or the a few would need to be made, each with
geometric approach (Grafius, 1964, 1965; a higher probability of success. We have
Lupton, 1965); and multivariate analysis demonstrated in practical, participatory
and parental distances (Bhatt, 1970; breeding programmes that this is the case
Pederson, 1981). These methods have in rice (Joshi et al., 2007; Virk et al., 2003)
the advantage that they use data from a and that the equivalent in maize to a few
134 Plant breeding and farmer participation
crossesa single composite populationis ments and hence be selected for its high
also effective (Witcombe, Joshi and Goyal, yield potential and multiple disease and
2003; Virk et al., 2005). pest resistance. Client-oriented, participa-
The breeder can predict that a cross tory methods then benefit from classical
is more likely to give rise to desirable breeding for adaptation to favourable envi-
segregants when the parents have comple- ronments.
mentary attributes suitable for the target Another participatory method
environment. It is advantageous if they conceptually closely related to PVS is to
are also unrelated, to increase the extent of exploit current varietal adoption. D.N.
possible transgressive segregation. A few- Duvick (pers. comm.) has pointed out how
cross, participatory strategy emphasizes the farmers do a tremendous amount of selection
role of the plant breeder in the evaluation for maize breeders in the United States of
of introduced and collected germplasm as America, because the inbred-line parents
potential parents and the collective skills of the most successful cultivars are used as
of farmers and breeders in selection, rather parents in breeding programmes. This is
than emphasizing the skill of the breeder in participatory researchthose cultivars have
selecting superior genotypes from within been grown over thousands of locations for
many crosses. several years providing a multilocational
In participatory, i.e. decentralized, testing system far beyond the capacity of
breeding at least one of the parents should a formal breeding programme. In our rice
be adapted to the target environment and breeding programmes in Nepal, widely
have traits that farmers like. Participatory adopted varieties such as CH 45 and Sabitri
varietal selection (PVS) efficiently identifies have been used as parents. Although extent
locally adapted parents: a range of germ- of current adoption is clearly a useful
plasm can be evaluated by farmers in their criteria for parental selection, in many
own fields that can include local landraces, marginal areas PVS will sometimes quickly
recommended cultivars and introduced and cheaply identify varieties superior to
varieties. A variety selected by PVS is an those that farmers are currently growing,
ideal parent since it has local adaptation e.g. Joshi and Witcombe (1996).
and traits that farmers prefer. Witcombe et Landraces may sometimes offer specific
al. (1996) suggested several types of crosses characteristics that are preferred by farm-
following PVS: a variety selected by farm- ers. However, in our breeding programmes
ers in the PVS trials is crossed with either in rice, using landraces as parents has been
a local landrace, another variety selected less productive than using high-yielding
by participatory methods, or an exotic modern varieties. Perhaps this is unsurpris-
variety. This allows crosses to be made ing, since improved varieties often have
both between adapted adapted parents or higher yields, and are more disease resist-
between adapted unadapted parents. Of ant than landraces. It makes no sense to use
these options, an adapted variety identified landraces just because they are landraces
by PVS crossed with an unadapted exotic (Wood and Lenn, 1997), but rather to
variety will often have the greatest genetic use them only when identified as having
dissimilarity. When breeding for marginal superior attributes. The argument that a
environments, the exotic variety can have landrace is locally adapted and has post-
adaptation to more favourable environ- harvest qualities that farmers appreciate is
Methodologies for generating variability. Part 2: Selection of parents and crossing strategies 135
insufficient when the best PVS variety also larger F2 population from which several
has these traits. generations of large, early-generation, bulk
The strategy for choosing parents may populations were derived before lines were
differ little across target environments and produced (Witcombe and Virk, 2001). Thus
the scale of the breeding programme. More we avoided using resources on selection in
favourable environments are highly diverse the early generations, when it is less effi-
and participatory approaches are also need- cient for low-heritability traits compared
ed for them (Witcombe, 1999). In breed- with selection in later ones (Fahim et al.,
ing for favoured mega-environments, the 1988). Instead, we concentrated resources
same crosses are used to cover several or on selecting in later generations when high-
many countries, but this process is better er between-line genetic variance increased
decentralized by matching crosses to target efficiency (Kearsey and Pooni, 1996). Mass
countries (Ceccarelli et al., 1994). selection in the advanced bulk popula-
tions produced rice varieties as uniform as
6.4 EVIDENCE AND CONCLUSIONS those from line selection (Virk, Steele and
We have used a few-cross approach in Witcombe, 2007).
our client-oriented breeding (COB) pro- There is further evidence that the few-
grammes (Witcombe et al., 2005). Using few cross approach is effective. Another rice
crosses was effective; we made only three breeding programme in Nepal funded by
crosses in our breeding programme for rice the International Plant Genetic Resources
targeted at Nepal and India by 1998, two Institute (IPGRI, now Bioversity
of which were clearly successful, compared International) has also relied on only a
with a success rate of <1 percent in classical few crosses. One parent was always a local
breeding programmes (Witcombe and Virk, landrace because landrace utilization was an
2001). The first cross we made, between the objective of the programme. Even with this
tall upland rice variety Kalinga III and the constraint on the choice of parents, of only
dwarf-statured lowland IR64, produced 8 crosses, 4 have resulted in varieties that are
successful varieties, including three that had in the release or pre-release stage (Gyawali,
been released by 2008: two in India (Virk unpublished). At the Africa Rice Center
et al., 2003) and one in Nepal (Gyawali et (WARDA; formerly the West Africa Rice
al., 2006). The third cross we made, Radha Development Association), a wide-cross
32/Kalinga III, produced Judi 582 and Judi breeding programme between Oryza sati-
572 that have been adopted in Bangladesh va and O. glaberrima placed considerable
(Joshi et al., 2007). The second cross we effort on choosing the parents of the crosses
made, Kalinga III/IR36, also produced high (Jones et al., 1997). Only eight parents of
yielding lines, but these were not promoted glaberrima and five of sativa were chosen
as they were inferior in grain quality to on the basis of their best combination of
those from the IR64 cross. traits, and only seven of the crosses set
A fewer-cross strategy greatly simplifies seed. All of the seven New Rice for Africa
the breeding scheme and saves resourc- (NERICA) varieties that were released in
es. Some resources were re-allocated to 2000 (WARDA, 2006) were from just one
increase the probability of obtaining desir- of these crosses, a success rate of 14 percent.
able segregants (and reduce the risk from As was the case for our crosses, this is a con-
using only a few crosses) by using a much siderable improvement over normal success
136 Plant breeding and farmer participation
rates and our experience suggests that this Allard, R.W. 1999. Principles of Plant Breeding.
was due to the great attention paid to choos- Second Edition. New York, USA, John
ing parents, necessitated by the high cost of Wiley and Sons.
making these wide crosses. In maize, the Baenziger, P.S. & Peterson, C.J. 1992. Genetic
parallel of a few-cross approach is to make variation: Its origin and use for breeding
only a single composite population, and we self-pollinated species. In H.T. Stalker and
tested this in western and eastern India. Two J.P. Murphy, eds. Plant breeding in the
populations were made, one for each region, 1990s. pp. 6992. Wallingford, UK, CABI.
and both have produced a released variety Baker, R.J. 1984. Quantitative genetic princi-
(Witcombe, Joshi and Goyal, 2003; Virk et ples of plant breeding. In J.P. Gustafson, ed.
al., 2005). Gene manipulation in plant improvement
What if all breeders used only a few I, pp. 147176. New York, USA, Plenum
crosses? This would restrict the amount of Press.
germplasm used in crosses but not restrict Bernardo, R. 2003. Parental selection,
the amount used in successful crosses. In number of breeding populations, and size
conventional programmes, although many of each population in inbred develop-
crosses are made, most produce neither ment. Theoretical and Applied Genetics,
released varieties nor progeny that could 107: 12521256.
be used in crosses to eventually produce a Bhatt, G.M. 1970. Multivariate analysis
released variety. However, the exceptions approach to selection of parents for hybrid-
are valuable: for example, IR64 has an ization aiming at yield improvement in
extremely complex parentage with 20 self-pollinated crops. Australian Journal of
original farmer varieties from eight countries Agricultural Research, 21: 17.
as parents (IRRI, 1985). Clearly, not all of Busch, R.H., Hanke, J.C. & Frohberg, R.C.
them would have previously been released 1974. Evaluation of crosses among high
varieties or parents of released varieties. To and low yielding parents of spring wheat
deliberately broaden the genetic base of (Triticum aestivum L.) and bulk predic-
crops, more crosses have to be made. This tion of line performance. Crop Science,
is particularly so when it involves parents 14: 4750.
about which little is known or, as is the case Ceccarelli, S., Erskine, W., Hamblin, J. &
for little-grown landraces or wild relatives, Grando, S. 1994. Genotype environment
when performance per se gives poor interaction and international breeding
indications that the parents are valuable. programmes. Experimental Agriculture,
Using few crosses is certainly suitable for 30: 177187.
breeding that is entirely targeted at rapidly Cox, D.J. & Frey, K.J. 1984. Combining ability
producing varieties for client farmers in and the selection of parents for interspecific
national programmes that have limited oat matings. Crop Science, 24: 963967.
resources. The few crosses allow better Fahim, M., Dhanapala M.P., Senadhira, D. &
market orientation and increase efficiency. Lawrence, M.J. 1988. Quantitative genetics
of rice II. A comparison of the efficien-
REFERENCES cy of four breeding methods. Field Crops
Allard, R.W. 1956. Estimation of prepotency Research, 55: 257266.
from lima bean diallel cross data. Agronomy Grafius, J.E. 1964. A geometry for plant breed-
Journal, 48: 537543. ing. Crop Science, 4: 241246.
Methodologies for generating variability. Part 2: Selection of parents and crossing strategies 137
Grafius, J.E. 1965. A geometry for plant Kearsey, M.J. & Pooni, H.S. 1996. The genetical
breeding. Research Bulletin, Michigan analysis of quantitative traits. Cheltenham,
Agricultural Experiment Station, No. 7. UK, Stanley Thornes.
Gyawali, S., Witcombe, J.R., Joshi, K.D., Devkota, Lupton, F.G.H. 1965. Studies in the breed-
K.P., Tripathi, M., Subedi, M., Sunuwar, S., ing of self pollinating cereals. 5. Use of
Chaudhary, B.N., Chaudhary, D., Adhikari, the incomplete diallel in wheat breeding.
N.P., Akhtar, T., Yadav, R.B., Khatiwada, S.P., Euphytica, 14: 331352.
Subedi. D. & Mishra, H.D. 2006. A proposal Murphy, J.P., Cox, T.S. & Rodgers, D.M. 1986.
for the release of rice variety Barkhe 3004 (a Cluster analysis of red winter wheat culti-
rice variety developed using client-oriented vars based upon coefficients of parentage.
breeding (COB) approaches). LI-BIRD, Crop Science, 26: 672676.
Pokhara, Nepal; NRRP, Hardinath, Nepal; Pederson, D.G. 1981. A least-squares method
Jaskelo Youth Club, Chitwan, Nepal; CAZS- for choosing the best relative proportions
NR, University of Wales, Bangor, UK. See when intercrossing cultivars. Euphytica,
www.dfid-psp.org 30: 153160.
Hehn, M. 1996. Optimum number of crosses Qualset, C.O. 1979. Mendelian genetics of
and progeny per cross in breeding self fer- quantitative characters with reference
tilizing crops. I. General approach and first to adaptation and breeding of wheat.
numerical results. Euphytica, 91: 365374. In S. Ramanujam, ed. Proceedings 5th
IPGRI [International Plant Genetic Resources International Wheat Genetics Symposium.
Institute]. 1996. Proceedings of the IDRC/ Vol. 2, pp. 577590. Indian Society of
IPGRI Workshop on Participatory Plant Genetics and Plant Breeding, New Delhi.
Breeding. International Agricultural Virk, D.S., Steele, K.A. & Witcombe, J.R. 2007.
Research Centre, Wageningen, Netherlands. Mass and line selection can produce equally
2830 July 1996. Edited by P. Eyzaguirre & uniform rice varieties. Field Crops Research,
M. Iwanaga. IPGRI, Rome. 100: 341347.
IRRI [International Rice Research Institute]. Virk, D.S., Singh, D.N., Prasad, S.C., Gangwar,
1985. Parentage of IRRI crosses IR1 J.S. & Witcombe, J.R. 2003. Collaborative
IR50,000. Los Baos, Philippines, IRRI. and consultative participatory plant breed-
Jones, M.P., Dingkuhn, M., Aluko, G.K. & ing of rice for the rainfed uplands of eastern
Semon, M. 1997. Interspecific Oryza sativa India. Euphytica, 132: 95108.
(L.) O. glaberrima (Steud.) progenies Virk, D.S., Chakraborty, M., Ghosh, J., Prasad,
in upland rice improvement. Euphytica, S.C. & Witcombe, J.R. 2005. Increasing
94: 237246. the client-orientation of maize breeding
Joshi, A. & Witcombe, J.R. 1996. Farmer par- using farmer participation in eastern India.
ticipatory crop improvement. II. participa- Experimental Agriculture, 41: 413426.
tory varietal selection, a case study in India. WARDA [Africa Rice Center]. 2006. Upland
Experimental Agriculture, 32: 461477. NERICA. Nerica for Africa. (available at
Joshi, K.D., Musa, A.M., Johansen, C., Gyawali, www.warda.org/warda/uplandnerica.asp;
S., Harris, D. & Witcombe, J.R. 2007. accessed 27 Dec. 2008).
Highly client-oriented breeding, using local Witcombe, J.R. 1999. Do farmer-participa-
preferences and selection, produces widely tory methods apply more to high potential
adapted rice varieties. Field Crops Research, areas than to marginal ones? Outlook on
100: 107116. Agriculture, 28: 4350.
138 Plant breeding and farmer participation
Witcombe, J.R. & Virk, D.S. 2001. Number of better described as highly client-oriented
crosses and population size for participa- plant breeding. I. Four indicators of client-
tory and classical plant breeding. Euphytica, orientation in plant breeding. Experimental
122: 451462. Agriculture, 41: 299319.
Witcombe, J.R., Joshi, A. & Goyal, S.N. 2003. Wood, D. & Lenn, J.M. 1997. The conserva-
Participatory plant breeding in maize: tion of agrobiodiversity on-farm: question-
A case study from Gujarat. Euphytica, ing the emerging paradigm. Biodiversity
130: 413422. and Conservation, 6: 109129.
Witcombe, J.R., Joshi, A., Joshi, K.D. & Wricke, G. & Weber, W.E. 1986. Quantitative
Sthapit, B.R. 1996. Farmer participatory genetics and selection in plant breeding.
crop improvement. I: Varietal selection Berlin, Germany, and New York, USA, W.
and breeding methods and their impact de Gruyter.
on biodiversity. Experimental Agriculture, Yonezawa, K. & Yamagata, H. 1978. On the
32: 445460. number and size of cross combinations in
Witcombe, J.R., Joshi, K.D., Gyawali, S., Musa, a breeding programme of self-fertilizing
A.M., Johansen, C., Virk, D.S. & Sthapit, crops. Euphytica, 27: 113116.
B.R. 2005. Participatory plant breeding is
139
CHAPTER 7
Methodologies for
generating variability
Part 3: The development of base
populations and their improvement
by recurrent selection
John R. Witcombe
140 Plant breeding and farmer participation
Gardner and Eberhart, 1966) are usually crossing locally adapted landraces with
not warranted. This is particularly true for complementary, high yielding parents
client-oriented approaches where often the (modern varieties) was demonstrated (vom
breeding is targeted at farmers who have Brocke et al., 2002, 2003). The introgression
had limited benefits from modern plant of modern varieties into landraces increased
breeding. The base population then needs genetic diversity, led to broader adaptation
to be developed from a viewpoint of speed than either landraces or modern varieties
and simplicity, because the task is not the alone, and under high rainfall conditions
difficult one of improving upon a recent still yielded as much as modern varieties.
modern variety but to improve upon a lan- In maize, in western India (Witcombe,
drace or an obsolete variety. Joshi and Goyal, 2003) and in eastern India
The biggest gains will be made when the (Virk et al., 2005) the traits that farmers
population has a high initial yield and a high wanted in new maize varieties (white grain
genetic variance. For decentralized breeding, (endosperm), with early maturity and with
we have found that the recombination of tolerance to the most common abiotic
locally adapted varieties with high-yielding constraints of drought encountered in these
exotic varieties can produce this desirable regions) were determined from interviews
combination. High genetic variance can be and from the results of participatory trials.
achieved by crossing unrelated germplasm, More detailed requirements emerged during
such as white- and yellow-endosperm mate- the course of these programmes, such as
rial (Witcombe, Joshi and Goyal, 2003; Virk a need for high cob-placement to avoid
et al., 2005). Tiwari (2001) tried several damage by jackals. In Africa, Bnziger and
types of crosses, and the one that involved Cooper (2001) have targeted the two major
both local and improved germplasm and constraints identified by farmers, namely
white- and yellow-endosperm types was drought and the ability to yield well even
the most successful. under low nitrogen conditions. In pearl
In client-oriented breeding programmes millet, a survey of farmers (ICRISAT, 1987)
that are decentralized to a given target region, in Maharashtra showed the importance that
a high proportion of the base population farmers placed on large individual grain size
parents should be adapted to it and have and early maturity. A breeding programme
traits that local farmers like (Witcombe et was based on the creation of a composite
al., 1996). To increase diversity, less-well- with bold grains and earliness (the Bold
adapted parents should also be included Seeded Early Composite). This produced
for their complementary attributes in order highly acceptable cultivars for farmers in
to produce desirable segregants. When both India and several African countries
breeding for marginal environments, these (ICRISAT, 1997).
complementary parents can be adapted Participatory varietal selection is a very
to more favourable environments and efficient way of identifying locally adapted
are chosen for high yield potential and parents with traits that meet specific client
superior disease and pest resistance. They needs. A closely related method, because it
are likely to be unrelated to the locally also relies on farmer acceptance of varieties,
adapted parents, thereby increasing the is to exploit knowledge of current varietal
genetic variance of the population. In pearl adoption. Duvick (2002) points out that
millet, the value of such an approach of farmers do a tremendous amount of selection
142 Plant breeding and farmer participation
for maize breeders in the United States of 7.2.4 Making the initial cycle (C0) bulk
America, because the inbred-line parents General considerations
of the most successful cultivars are used as Adequate genetic recombination between
parents in breeding programmes. However, the parents of a composite will produce a
although current adoption is clearly a diverse range of recombinants for the first
helpful criterion for parental selection, in cycle of recurrent selection. How early this
more marginal agricultural areas it may occurs will depend on how inbred the parents
be less useful. There may be considerably are. When the parents are inbred, the third
better varieties than those that farmers are generation of random mating is equivalent
currently growing and these can be quickly to the F2 in an inbreeding crop in that it is
and cheaply identified by PVS (e.g. Joshi the first in which transgressive segregation
and Witcombe, 1996). can occur. In this case a third generation of
random mating is needed before selection
7.2.3 Population size should commence. Transgressive segregants
Population size has to be large to provide occur earlier if the parents are heterozygous,
a reasonable probability of finding rare open-pollinated varieties. In this case, mass
or infrequent desirable transgressive selection after a single generation of random
segregants. In cross-pollinated crops, small mating can be expected to result in a genetic
population sizes cause significant inbreeding advance, but possibly at the expense of an
depression, so population sizes need to be early reduction in genetic diversity that
sufficiently large to avoid this; an effective reduces the potential for long-term gains. A
population size of more than 500 plants third generation of random mating remains
in each generation is sufficiently large to desirable before any progeny testing is
avoid significant inbreeding depression started.
(this is calculated from a standard formula,
e.g. Falconer, 1981; see also Chapter 2 in Using maternal ancestry to aid
this volume). The resources available also recombination
dictate the size of each population. The In crops where the occurrence of natu-
more base populations that are created ral selfing is low, the simplest method of
and improved, the fewer the resources recombination is to allow random mating
that can be devoted to each one. In the in a bulk grown from equal amounts of
breeding of open-pollinated crops, creating parental seed. However, the maintenance
and improving even a single population still of some form of population structure is
requires many resources. The strategy used desirable as it allows a visual estimate to be
in our client-oriented breeding programmes made of the extent of recombination. When
has therefore been to minimize the use of parental numbers are not too high, this can
resources by improving a single population be done by maintaining sub-bulks derived
for each group of target clients (Witcombe, from the individual parents of the base
Joshi and Goyal, 2003; Virk et al., 2005). population and growing them in an isolated
Not only does this reduce the resources plot (Figure 7.1). To aid randomness of
required, but it more carefully focuses recombination a pollinator bulk is used that
the base population to an identified target is made from aliquots of seed of the original
group of clients. It has proven to be a parents. The pollinator bulk is preferably
successful strategy. also planted on the borders of the isolated
Methodologies for generating variability. Part 3: The development of base populations 143
FIGURE 7.1
Planting layout design for hills in maize. Detasselled plants in grey
6 3 2 5 8 2 1 7 1 5 1 2 4 2 8 5 7 8 6 3
4 9 4 6 4 8 9 8 9 3 7 1 8 1 6 1 6 1 4 9
7 8 2 9 7 5 4 5 6 8 4 7 1 3 5 8 5 2 7 8
2 7 1 5 1 3 7 3 4 5 6 2 3 8 7 3 1 7 2 7
5 4 6 2 9 7 3 1 8 6 3 9 7 9 2 9 4 3 5 4
1 9 2 8 2 9 8 9 5 7 6 4 8 5 3 6 2 3 1 9
9 2 3 9 5 6 3 2 1 2 9 3 2 3 8 4 1 7 9 2
3 8 2 7 3 4 9 5 6 9 2 1 5 2 1 7 8 1 3 8
8 5 4 3 4 8 4 6 2 1 7 3 4 7 9 2 9 5 8 5
plot. When, for example, 50 percent of the in trials for evaluating the progress made
area is planted with the pollinator bulk, the by selection.
entries and the bulk are planted in alter-
nate ridges and beds so that every entry Using forced crossing in the first
is surrounded by the pollinator bulk. In generation of random mating
the second and subsequent random mating When there are few parents, it is much
generations, the pollinator is either recre- more efficient to employ diallel crossing
ated by bulking aliquots of seed taken from in the first generation of random mating.
the open-pollinated entries, or is advanced This greatly increases randomness of
through the generations by harvesting its mating by avoiding the high proportion
open-pollinated seed (Figure 7.1). Random of sibbing within the parental entries that
mating is repeated until the entries lose all occurs under natural random mating. It
or most of their identity relative to each also avoids the need for an isolated plot.
other and the pollinator bulk. In maize, When there are too many parents to make
unlike in pearl millet, the entry rows and a complete diallel, a half diallel can be used,
the pollinator bulk can be made to cross by or they can be crossed in, for example, a
detasselling the entry rows. In this case, the partial diallel using systematically selected
pollinator bulk is best made up afresh each crosses, or a randomly made partial diallel
time from aliquots of the female entries. (designs for such random crossings are
If necessary, only the entry rows can be discussed below). If there are not too many
planted, either to reduce the land require- crosses in the diallel, then progress of extent
ment or because there is insufficient seed to of recombination can be assessed, in the
sow both the entry rows and the pollinator second and subsequent random matings,
bulk. The extent of recombination is then by planting the entry rows according to
assessed by the between-entry phenotypic the maternal parents of the cross that were
differences. made in the first generation.
When recombination appears complete,
equal amounts of seed are taken from the Hill designs in maize
entry rows to make the C0 bulk of the In maize, the prevention of selfing is simply
composite. A portion of the C0 bulk should done by detasselling plants. This can lead
be retained for use as the base population to very effective methods of increasing
144 Plant breeding and farmer participation
ers in selection among the varieties that are 7.3.1 Mass selection methods
produced by the programme. This process General considerations
is commonly called participatory varietal It seems to be an unavoidable fact that most
selection (PVS) and involves farmers testing plant breeders dislike simple methods. In
material in their own fields (Witcombe et al., self-pollinated crops most breeders employ
1996, 2005). The methods used involve some elaborate forms of line selection whereas,
form of mother-and-baby design (Snapp, when experimental comparisons are made,
1999), where all of the entries are tested in it is the simplest methods, such as bulk-
relatively simple designs in the mother trials population breeding or single-seed-descent
and subsets or individual entries are tested methods, that show the highest efficiencies
in baby trials. In this system, researchers (e.g. Fahim et al., 1998, in rice). The same
and farmers evaluate the varieties. Particular may apply to outbreeding crops where sim-
attention is paid to the perceptions of the ple mass selection is rarely used, perhaps
farmers for a range of traits and their overall because it is considered to be an under-
preferences among the varieties. exploitation of the breeders knowledge
and skills. However, in outbreeding crops
7.3 POPULATION IMPROVEMENT there is some justification for more elabo-
METHODS REQUIRING ONE YEAR PER rate methods because evidence shows that
CYCLE greater genetic gains per year can be made
Mass selection, modified mass selection, with progeny testing. Nonetheless, as can
half-sib family selection and full-sib family be seen below, significant genetic progress
selection can be completed in a single year can be made from mass selection, and that
in locations where two generations can be progress requires the smallest investment in
grown in the field in a year (Figure 7.5). resources per unit of gain.
FIGURE 7.5
Four selection schemes that can be completed in a single year with two generations per year.
(o.p. = open pollinated).
Because of this cost effectiveness and mass selection in the New Elite Composite
because of its simplicity, mass selection (NELC) to produce a high-yielding open-
methods are attractive. They are particu- pollinated variety, ICMV 155, that was
larly useful for decentralized breeding in released for cultivation in India.
difficult environments where complex
breeding methods are beyond locally avail- Maize
able capacity. Hallauer and Miranda (1988) report many
Mass selection is most effective when studies on mass selection in maize in the
the initial base population yield is high and 1960s and 1970s. The responses to mass
there is high genetic variance. Given these selection were high given the simplicity
circumstances, high rates of genetic gain of the method, but few of the experiments
can be achieved in response to selection. were conducted over more than three cycles.
As was discussed earlier, in the develop- Examples of longer-term experiments are
ment of base populations, a combination of gains in grain yield of 19.1 percent per
locally adapted material with high yielding cycle over ten cycles (Genter, 1976) and
exotic genotypes can produce this desirable gains of 2.1 percent per cycle over six cycles
combination. (Lonnquist, Cota and Gardner, 1966). More
Mass selection can also be made more recent studies on mass selection are few.
attractive by simple modifications. The Weyhrich, Lamkey and Hallauer (1998)
most commonly suggested modification report on a comparison of selection methods
is stratified mass selection, but personal in a population. Over ten cycles of mass
experience has not shown this to be either selection an average gain of 0.6 percent per
simple to do or effective. Instead, alterna- cycle was achieved, the lowest of the gains
tive improvements to mass selection are per cycle in the various selection methods
considered after a brief consideration of the (from mass selection to S2). However, it was
evidence of its effectiveness in pearl millet superior to half-sib family selection in gains
and maize. per year. It was also the most superior of
the methods in terms of cost-effectiveness;
7.3.2 Evidence of the effectiveness of the costs per unit of gain were the lowest
mass selection and the returns on investment the highest.
Pearl millet
Mass selection can be effective in improving 7.4 IMPROVEMENTS OVER SIMPLE
pearl millet (Govil, Pokhriyal and Murty, MASS SELECTION
1986; Rattunde, Singh and Witcombe, 1989; 7.4.1 Gridded mass selection do grids
Singh et al., 1988). Rattunde, Singh and really help?
Witcombe (1989) showed in four compos- Various authors have suggested improve-
ites that heritabilities from single plants ments over mass selection. Gardner (1961)
were appreciable. Averaged across com- developed an improved method of mass
posites, the estimated heritabilities for 19 selectiongridded (or stratified) mass selec-
traits varied from 0.29 for yield, 0.45 for tionand demonstrated its effectiveness in
flowering time, to 0.64 for panicle length, improving grain yield in maize. Burton
indicating that mass selection will be effec- (1974) reported on recurrent restricted phe-
tive for many important traits. Singh et al. notypic selection in Pensacola Bahiagrass
(1994) demonstrated a significant gain from (Paspsalum notatum) that differs from mass
150 Plant breeding and farmer participation
selection by having five restrictions, the at a more than adequate seed rate and then
two most important being stratification thinned to a uniform spacing.
(grids) as well as the control of pollination
so that both male and female parents are Realistic selection differentials
selected (see below). In many mass selection schemes, very
Rattunde, Singh and Witcombe (1989) high selection pressures are applied by
examined heritability values in four pearl selecting, for example, the 100 best plants
millet composites, using non-stratified and from 10 000 (e.g. Weyhrich, Lamkey and
stratified data for many traits. Stratification Hallauer, 1998). Instead, we apply strong
was not worthwhile since the improve- roguing (removal of undesirable plants).
ments in heritability with stratification were Although the selection differential applied
low and erratic. For stratification to be will be lower, it will be more reliable; when
effective, uniform conditions are required only a few phenotypically best plants are
within the strata, with a gradient in one or selected the risk increases that a high pro-
more environmental variables across them. portion of the selected plants are mistakes
In practice, such conditions seem to occur because they happen to be in environmen-
rarely in typical pearl millet experimental tally better situations.
fields, which tend to be either uniform or
have random variation. Experience of mass Moving grids
selection in maize has shown that this was a The removal of the undesirable or poor
common problem in this crop as well. plants is done by the breeder walking
between alternate rows and removing plants
7.4.2 Improving mass selection by in the rows on either side that are inferior
simple modifications to selection to their neighbours. This can be likened to
procedure a form of moving grid where selection is
Discarding poor areas of the plot done on the relative performance of plants
When the principles of client-oriented in a small area and better accounts for ran-
breeding are followed, only one base popu- dom (patchy) variation than grids.
lation is improved per target domain. Given
the small number of base populations, this Avoiding selfing
allows sufficient resources to grow them in In mass selection in maize, the selected
large plots. Hence, it is possible to remove plants are allowed to random mate with the
all of the plants before flowering from rest of the population. However, any maize
patches and field margins where the crop breeder looking at the ear of a single white-
has grown poorly and still leave a large endospermed maize plant grown amongst
population. many yellow-endospermed plants (or vice
versa) cannot fail to be surprised by how
Equal plant spacing numerous are the grains with the maternal
Equal spacing of plants eliminates an impor- grain colour that result from selfing. Even
tant source of environmental variability and though reports in the literature report self-
improves the between-plant heritability. In ing rates of about 5 percent, this is not
maize, the population is hill planted (two based on any extensive experimental data.
plants per hill) and thinned to one plant per The proportion of selfing may be consider-
hill. In pearl millet, the crop can be sown ably higher, particularly when wind speeds
Methodologies for generating variability. Part 3: The development of base populations 151
Off season
ing means that the mass selection is more Recombination
Plant S1 lines as as a bulk.
Mass select during
efficient. Even with only 5 percent out- generation random-mating
in isolated plot
crossing, 5 percent of the plants have to be
removed for inferior phenotypic perform-
Main season
Plant half-sib families
Selng
ance because of inbreeding, without any generation as a bulk.
Self and mass select.
genetic gain being made, and with higher
selfing rates the problem increases.
Selfing is avoided by detasselling 50 per-
The scheme takes one year per cycle with two cropping
cent of the population and after mass selec- seasons per year
tion (which is done on all of the population
of plants whether detasselled or not) only
the ears from the detasselled plant contrib- generation, higher selection efficiency is
ute to the next generation. The detasselling achieved because the between-plant herit-
does not add greatly to the labour involved ability between S1 plants is higher than
as selection, which has to be done fre- between S0 plants.
quently to eliminate poor pollinator plants Of practical interest is the question of in
as early as possible, can be combined with which generation is selection the most effi-
detasselling. Alternatively, unskilled work- cient? Predicted gain with varying degrees
ers can be employed to do the detasselling. of pollination control can be estimated
from the following equation (Hallauer and
Selection before pollen shed Miranda, 1988):
Frequent, early selection, rather than selec- G = icA2//A2+D2+E2
tion just at maturity, means that many where: i is the selection intensity or stand-
plants can be eliminated before pollen shed. ardized selection differential; and c is a
Hence, selection is exerted not only on the coefficient where c = 0.5 (no control over
female parent but to a lesser extent on the pollination and selection after pollination)
male parents as well (so c > 0.5 but < 1 in or c = 1 (control over both male and female
the equation below). gametes as in the case of selfing). The
One form of modified mass selection is remaining terms refer to the square roots of
to employ alternate selfing and recombina- additive (2A), dominance (2D) and envi-
tion generations (Figure 7.6). This is feasi- ronmental (2E) variances. Selfing controls
ble in pearl millet as selfing only involves the source of both male and female gametes
bagging panicles, whereas controlled cross- (c = 1). In simple mass selection, the selec-
ing between the tassel and silk is required in tion takes place after pollination, so selec-
maize. The major advantages of employing tion is only on the female parent (c = 0.5).
alternate selfing generations is control of Hence, controlling pollination by selfing
pollination in the selfing generation, which always doubles the selection efficiency over
doubles the efficiency of mass selection as simple mass selection.
it is applied to both the male and the female The efficiency of selection in the S1 bulk
parents. In the subsequent recombination is increased over simple mass selection by
152 Plant breeding and farmer participation
Falconer, D.S. 1981. Introduction to & Witcombe, J.R. 1995. Mapping quantita-
Quantitative Genetics. 2nd edition. London tive trait loci for downy mildew resistance
and New York, USA. Longman Group in pearl millet. Theoretical and Applied
Limited. See pp. 214218. Genetics, 91: 448456.
Fischer, K.S., Fukai, S., Lafitte, R. & McLaren, Joshi, A. & Witcombe, J.R. 1996. Farmer partic-
G. 2003. Know your target environment. In ipatory crop improvement. II. Participatory
K.S. Fischer, R. Lafitte, S. Fukai, G. Atlin varietal selection, a case study in India.
& B. Hardy, eds. Breeding rice for drought- Experimental Agriculture, 32: 461477.
prone environments, pp. 511. Los Baos, Kapoor, R.L., Singh, F., Yadav, H.P. & Dass,
Philippines, IRRI. S. 1983. Recurrent selections in pearl millet.
Gardner, C.O. 1961. An evaluation of the Haryana Agricultural University Journal of
effects of mass selection ands seed irradia- Research, 13: 424428.
tion with thermal neutrons on the yield of Kelley, T.G., Rao, P.P., Weltzien, R.E. &
corn. Crop Science, 1: 241245. Purohit, M.L. 1996. Adoption of improved
Gardner, C.O. & Eberhart, S.A. 1966. Analysis cultivars of pearl milet in an arid environ-
and interpretation of the variety cross dial- ment: Straw yield and quality considera-
lel and related populations. Biometrics, tions in Western Rajasthan. Experimental
22: 439452. Agriculture, 32: 161171.
Genter, C.F. 1976. Mass selection in a com- Khadr, F.H. 1977. Recurrent selection in pearl
posite of intercrosses of Mexican landraces millet Pennisetum typhoides in Northern
of maize. Crop Science, 16: 556558. Nigeria. Zeitschrift fur Pflanzenzchtung,
Geiger, H.H. 1985. Hybrid breeding in 79: 145153.
rye. Proceedings of the cereal section Lonnquist, J.H., Cota, O. & Gardner, C.O.
on rye. Svalf, Sweden. Part 1: 237265. 1966. Effect of mass selection and thermal
EUCARPIA. neutron radiation on genetic variances in a
Govil, J.N., Pokhriyal, S.C. & Murty, B.R. 1982. variety of corn (Zea mays L.). Crop Science,
Full-sib and reciprocal recurrent selection 6: 330332.
in relation to pearl millet improvement. Monyo, E.S., Ipinge, S.A., Heinrich, G.M. &
Theoretical Applied Genetics, 62: 2530. Chinhema, E. 2000. Participatory breeding:
Govil, J.N., Pokhriyal, S.C. & Murty, B.R. 1986. Does it make a difference? Lessons from
Effect of recurrent mass selection on grain Namibian Pearl Millet farmers. In N. Lilja,
yield in pearl millet. Crop Improvement, J.A. Ashby & L. Sperling, eds. Assessing the
13: 6770. impact of participatory research and gen-
Hallauer, A.R. & Miranda, F.O. 1988. der analysis. Cali, Colombia, International
Quantitative genetics in maize breed- Center for Tropical Agriculture (CIAT).
ing. 2nd edition. Iowa, USA, Iowa State Rattunde, H.F, Singh, P. & Witcombe, J.R.
University Press. 1989. Feasibility of mass selection in pearl
ICRISAT [International Crops Research millet. Crop Science, 29: 14231427.
Institute for the Semi-Arid Tropics]. 1987. Schipprack, W. 1992. Estimation of popula-
ICRISAT Annual Report 1986. ICRISAT, tion parameters and optimization of recur-
India. See pp. 116118. rent selection in pearl millet. PhD Thesis.
ICRISAT. 1997. Southern and Eastern Africa University of Hohenheim, Germany.
Region Highlights 1996. ICRISAT, India. Seitz, G. 1989. Experimentelle und theo-
Jones, E.S., Liu, C.J., Gale, M.D., Hash, C.T. reitische untersuchungen zur neziehung
Methodologies for generating variability. Part 3: The development of base populations 157
CHAPTER 8
Methodologies for
generating variability
Part 4: Mutation techniques
with fertility of bread wheat cultivars with Nijisseiki of Japanese pear was developed
semi-dwarfness genes Rht-B1 and Rht- with chronic irradiation in the gamma field
D1 in temperate climates, the mutant line in Japan. The cultivar is more resistant to
Krasnodarskii karlik with gene Rht11 has Black spot disease and needs only one or
become more widely used in cross breeding two applications of fungicides per season.
programmes in Europe and Australia. The additional annual income for growers
is almost US$ 30 million (Ahloowalia,
Cotton Maluszynski and Nichterlein, 2004).
In cotton, mutagenic treatment with gamma
rays led to the development of two very 8.3 MUTAGENIC TREATMENT
important cultivars NIAB 78 in Pakistan 8.3.1 Radiation
and Lumian No. 1 in China. NIAB 78 was The FAO/IAEA Database on Officially
released in 1983 in Pakistan and during Released Mutant Varieties (MVD, no
the following ten years it doubled cot- date) indicates that radiation, especially
ton production - contributing more than gamma rays, has been the most often used
US$ 3.0 billion. The added income to cotton treatment for inducing mutations of crop
growers due to the cultivation of this culti- plants. The reason for this is the simplicity
var from the year of its release onwards has of the treatment rather than to any higher
been estimated at US$ 486 million. NIAB efficiency of mutation induction. Seeds
78, due to its wide adaptability, tolerance or other organs of the plant have to be
to heat and escape from bollworm attack delivered for irradiation to a nuclear centre.
because of early maturity, saved the textile Such centres have been established in most
industry of Pakistan, which was threatened countries. Nuclear centres are usually
by reduced cotton production. Annual cul- supervised by the Ministry of Energy or
tivation of the high yielding mutant cul- directly by the Prime Ministers Office.
tivar Lumian No. 1 exceeded one million They will inform the plant breeder which
hectares by the late 1980s. It was the most centre is providing a seed irradiation
widely grown cotton cultivar in China. service. Mutagenic treatment is free of
charge in most developing countries. Dry
Vegetatively propagated crops seeds (M0 generation), disease free, with
Important results have also been obtained good germination ability and about 1213
in breeding vegetatively propagated crops. percent moisture content, can be sent for
Two outstanding examples are mutant irradiation by regular mail. As the process
cultivars of grapefruit and of Japanese pear. of acute irradiation is very short, they
Budwood of grapefruit mutant cv. Star should be returned to the breeder in the
Ruby irradiated with thermal neutrons led same way by mail, within one or two
to the release of the Rio Red cultivar. This weeks. This is possible because irradiated
mutant cultivar was released in 1984 in seeds (M1 generation) or other plant organs
Texas. It is seedless and has red flesh, red are not radioactive and can be used directly
and stable juice colour, and good yield. The for sowing or kept refrigerated awaiting
fruits of these cultivars are known under the proper sowing period, even for a few
the trademark Rio Star. They are grown months, depending on the crop species. A
on 75 percent of the grapefruit-producing free-of-charge seed irradiation service is also
area in Texas. The mutant cultivar Gold provided by the FAO/IAEA Agriculture
Methodologies for generating variability. Part 4: Mutation techniques 163
Laboratory, Plant Breeding Unit, A-2444 The mutagenic action of a chemical muta-
Seibersdorf, Austria (<Official.Mail@iaea. gen induces somatic and genetic effects in a
org>). Any plant breeder can send seeds treated cell, tissue or organ. After treatment
for gamma ray irradiation to this address. of seeds, only unrepaired damage to the
The scientists working there can advise on DNA in initial cells of the sporogenic layer
the dose of gamma rays or apply the dose (germline cells) are transferred as mutations
requested by the breeder. They will also to the next generation. Other mutations
adjust seed moisture content if necessary. in somatic cells of the embryo, including
Physical mutagens are also very use- mitotic chromosomal aberrations, together
ful for inducing mutations in vegetatively with toxic action of a mutagen on all com-
propagated crops and in in vitro cultures. ponents of cytosol, affect plant growth and
Cuttings, immature spikes or Petri dishes development, and are called the somatic
with explants, calli, somatic embryos or effect of the mutagen.
microspores are often subjects of irradiation The steps generally followed in muta-
by ultraviolet (UV), gamma or X-rays. genic treatment of seeds with chemical
mutagens are:
8.3.2 Chemical mutagenesis pre-soaking in distilled water;
The use of chemical mutagens is also very pre-treatment rinsing in tap water;
simple and can be done in any biological treatment with the mutagen;
laboratory with basic equipment. However, post-treatment rinsing in tap water; and
it should be kept in mind that most chemi- drying (if necessary) on a filter paper.
cal mutagens are also strong carcinogens. All steps of mutagenic treatment should
For this reason, all steps of mutagenic be done using glass beakers to avoid any
treatment should be carried out wearing interaction of chemical mutagens with
gloves and under a Biohazard flow-hood. even trace quantities of metallic cations or
These safety conditions are not necessary other active reagents. Seeds for each dose
for treatment with sodium azide, which is of mutagenic treatment (M0 generation)
a very powerful mutagen, but only for a and for the untreated controlusually
limited number of species, including bar- the parent varietyare put into beakers
ley, rice, maize, oat, sorghum, sesame, jute that are visibly labelled with the applied
and soybean. Numerous chemical muta- concentration of mutagen.
gens have been successfully used for crop As dry seeds are usually used for treat-
improvement (Table 8.1). ment, pre-soaking in distilled water should
TABLE 8.1
Chemical mutagens used most commonly in plant mutagenesis
Name Abbreviation Molecular weight1
Ethyleneimine EI 43.07
Dimethyl sulfate DMS 126.13
Diethyl sulfate dES (DES) 154.19
Ethyl methanesulphonate EMS 124.20
N-ethyl-N-nitrosourea ENU (ENH) 117.11
N-methyl-N-nitrosourea MNU (MNH) 103.08
N-methyl-N1-nitro-N-nitrosoguanidine MNNG 147.09
Sodium azide NaN3 65.01
FIGURE 8.1
Schema of induced mutations project with description of mutated generations
f
c
e M2
M1 g
b d
M3
Mutagenic treatment h
a
M0 M4
i
Notes: a) seed before mutagenic treatment M0 generation; b) the same seed after mutagenic treatment M1 generation;
c) M1 generation in the field; d) individual M1 plant with visible chimeric chlorophyll sectors; e) spike of M1 plant with M2
seeds; f) M2 chlorophyll mutants at seedling stage; g) mutant selected in M2 generation and its parent variety;
h) M3 generation; i) small plots of selected mutants in M4 generation.
166 Plant breeding and farmer participation
FIGURE 8.2
Structural chromosomal aberrations in anaphase of root meristems
A B
C D
E F
AB Vicia faba and CD Hordeum vulgare after maleic hydrazide treatment; EF Vicia faba after MNU treatment;
a) bridge; b) two fragments; c) bridge and two fragments; d) fragment and delayed chromosome; e) bridge and
numerous fragments: f) interphase with micronuclei (courtesy of Dr. J. Maluszynska).
Methodologies for generating variability. Part 4: Mutation techniques 167
lering, flower and fruit shape and colour, tion of undefined deleterious mutations
but also resistance to herbicides, allows the from a mutated genotype. Significant yield
screening of very large populations under improvement of the selected recombinant
field conditions. What is usually not pos- with the mutated phenotype in relation to
sible or too costly is selection using any the original mutant is the best illustration
laboratory technique. that the elimination of deleterious muta-
In summary, the chosen doses for muta- tions has been achieved, at least partly.
genic treatment should be relatively low For the breeder without experience in
for the improvement of a parent mate- the use of mutation techniques, the most
rial with favoured genetic background. It difficult problem is to identify, in practice,
should be noted that mutagenic treatment a proper dose for mutagenic treatment,
generates mutations in many other genes keeping in mind the considerations dis-
in the genome of each cell, not only at the cussed above. There are several laboratory
desired locus. One result of using too high tests to help define a critical dose for both
a dose is a high frequency of desired muta- physical and chemical mutagens. The term
tions, unfortunately also accompanied by a critical dose implies the dose of a mutagen
high frequency of deleterious mutations in beyond which the somatic effects in the M1
other important loci. The deleterious muta- generation are too high. In sexually propa-
tions negatively influence the agronomic gated crops, doses of LD50 and above are
value of selected mutants. Such mutants definitely considered to be critical doses in
selected in M2 or M3 generations usually current approaches to induced mutations
have problems of low fertility, late matu- for breeding purposes.
rity or susceptibility to stresses, i.e. the The simplest and cheapest laboratory
characters that directly influence yielding test, suitable for the evaluation of somatic
capacity. The use of high doses of muta- effects of mutagenic treatment in mono-
gens, in the era of application of induced and dicotyledonous crop species with any
mutations termed as mutation breeding, size of seeds, is a pot test for the measure-
was the reason that many programmes ment of emergence and seedling growth
in the 1960s and 1970s did not yield the reduction. To perform this test, ceramic or
expected results. A dose causing 50 percent plastic pots, 1822 cm diameter, are all that
lethality (LD50) was often suggested as the are needed. The pots can be replaced by
optimal for breeding programmes, result- any other plastic container of a similar vol-
ing in too high a frequency of deleterious ume and size. Metal and wood containers
mutations. Nevertheless, it should be noted should be avoided as they can release ions
that low doses of mutagen decrease the or active reagents influencing germination
frequency of mutations and in consequence and seedling growth. For very small seeds,
a larger population of M2 or M3 is neces- such as poppy, much smaller pots or con-
sary to find the most desirable mutants tainers can be used. The pots are two-thirds
in a promising genetic background. It is a filled (about 4 cm below the top of the pot)
good breeding practice to cross a selected with garden soil. One hundred seeds are
mutant with its parent variety and select sown on the soil surface of each pot and
desired recombinant from the segregating are covered by a few centimetres of sand.
F2 generation. This approach, known as The depth of the covering layer depends on
the cleaning method, helps in elimina- the size of the seeds. Smaller seeds should
Methodologies for generating variability. Part 4: Mutation techniques 169
Percentage
60
ent cultivar), should be sown in three pots,
50
composing three replications. They should 40
be watered according to the normal practice 30
for the crop. All emerged seedlings should 20
10
be counted in each pot when the number
0
of seedlings in the control treatment is sta- 0.75 1.50 2.25 3.0 3.75
the most important criterion, followed by and replaced by mechanical means as some
the breeding objective of the programme. active components of herbicides, often also
It should also be considered that genes mutagenic, can influence the growth and
controlling other, often useful, characters development of injured plants. Sowing at
can mutate with lower frequency than double spacing in rows is often applied to
the genes controlling chlorophyll. After avoid competition between M1 plants with
choosing the dose, it is very important to different levels of somatic effects. Good
calculate the size of the M1 generation. In tillering will also allow exploitation of all
this step, the knowledge of survival and mutations from different initial cells. When
fertility reduction for a particular dose a high dose of mutagen is used, a significant
is necessary. These parameters help to delay in maturity should be expected.
calculate how many plants will survive As a multicellular tissue was the subject
after mutagenic treatment and how many of mutagenic treatment and a different
seeds will be harvested as a result of the spectrum of mutations was induced in
reduction in fertility in surviving plants. As each cell, the tissues developing from the
the pilot experiment was performed under embryo carry cells with differently modi-
the climatic conditions of one season, it fied genomes and are chimeric. Induced
should be also considered that the level genetic polymorphism among initial cells
of somatic effects under the conditions of of the sporogenic layer influences the seg-
the next season could be different. For this regation ratio in the M2 generation. The
reason, is very advisable, for the large-scale mutations in cells of somatic tissues are not
experiment, to grow the M1 generation transferred to the next sexual generation.
not only with the selected dose, but also Some morphological mutations in somatic
with slightly lower and higher doses to tissues, such as chlorophyll defects, are
be sure that even under different climatic often visible and they clearly illustrate
conditions it will be possible to collect the chimeric structure of the M1 plant
enough mutated seeds for the desired size of (Figure 8.5). The appearance of chlorophyll
the M2 generation. It is also a good practice defects is a good indicator of genetic action
to have a small plot of the parent cultivar in of the mutagen. Somatic tissue chimeras
the close neighborhood to have a control are a valuable source of genetic variation in
comparator for the reaction of plants to breeding of vegetatively propagated crops.
growing conditions and management. The method of harvesting M1 is a key
issue in exploiting induced genetic variation
8.4.2 Handling mutated generations in sexually propagated crops. The method
and mutant selection applied reflects various factors, including
M1 generation the biology of reproduction of a particular
Mutagen-treated seeds should be sown in crop species, the cost of labour and of the
fertile soil and grown under good manage- selection method, the objective of breeding,
ment practices, including the use of fertiliz- and the possible size of the M2 generation.
ers. It is very important to maintain the M1 Different methods can be chosen, ranging
crop at the proper soil moisture level, as from the collection of only one seed per
plants with somatic effects are much more plant, to bulk harvesting of entire M1 plots
sensitive to stresses, especially drought. for development of the M2 population.
The use of herbicides should be avoided From a theoretical point of view, the best
172 Plant breeding and farmer participation
FIGURE 8.5
Chlorophyll defects in M1 generation after MNU treatment
Notes: a) xantha sector in barley; b) albina and xantha sectors in maize; c) large albina sector in narrow lupine; d)
xantha sector in faba bean; e) xantha sector in pea; f) albina, xantha and viridis sectors in backwheat.
way is to harvest separately and thresh the a rather unrealistic approach in practice.
spikes, pods or fruits from each individual More often, in breeding programmes, a few
M1 plant and sow seeds using a spike-, pod- spikes, pods or fruits from each M1 plant
or fruit-to-row method. However, this is are harvested, threshed and sown together.
Methodologies for generating variability. Part 4: Mutation techniques 173
can be carried out in vitro, the haploid cells It should be noted that the doses of muta-
or embryos provide an extremely large gens applied to cells in vitro should be at
mutagenized population, increasing the least an order of magnitude lower than the
probability of identifying a rare mutation doses used for seed treatment. In Brassica
event. The advantages of combining mutation napus, the irradiation doses used for muta-
techniques with DH systems are apparent tion induction in isolated microspore cul-
only when an efficient procedure of DH tures ranged from 5 to 15 Gy for gamma
production is available for a particular crop. rays, 10 to 40 Gy for X-rays, and a dose
The recent progress in developing effective rate of 33 erg mm-2 s-1 for 10 to 60 s with
protocols for DH production, especially UV rays. In microspore cultures of barley
through isolated microspore culture, has treated with sodium azide, the mutagenic
made possible the application of mass- treatment lasted only 1 hour, and the con-
scale in vitro mutagenesis and selection centration of the applied mutagen has not
methods in major cereals: wheat, barley and exceeded 1-4 M.
maize, similar to the previous achievements Another approach to mutagenic treat-
in oilseed rape. However, DH protocols ment with the use of a DH system relies on
are difficult to directly transfer between using M1 plants derived from mutagenized
laboratories, and the success in producing seeds as donors for haploid production.
a sufficient number of DH plants depends In this method, seeds are treated with the
very much on the breeders ability to grow doses of physical or chemical mutagens
high quality donor plants. used in conventional seed mutagenesis.
There are two main approaches for the Treatment of dormant seeds instead of hap-
use of haploid systems for mutant produc- loid cells in culture allows for application
tion. Most often, mutagenic treatment is of much higher doses of mutagens, which
applied to haploid cells (microspores) or provide the higher frequency of mutations
organs containing haploid cells (anthers, in the DH population. Avoiding the somat-
spikes, panicles or flower buds) at, or ic effects of mutagenic treatment on in vitro
before, in vitro culture. Mutagenic treat- regeneration ability is another advantage
ment of isolated microspores with gamma, of this procedure in comparison with the
X-ray or UV radiation proved to be an treatment of microspores in culture. Use
efficient method for mutation induction of M1 plants as donors for anther culture
in oilseed rape. Chemical mutagens can has been successfully demonstrated in bar-
also be applied to haploid cells in vitro, ley and rice. In Peruvian barley cultivars,
but the mutagenized cultures are more dif- numerous DH mutants were produced
ficult to handle because of the requirement from anther culture of M1 plants developed
for extensive washing in order to remove by treatment with MNU and sodium azide.
the mutagen residues. The application of In rice, anther culture of gamma-irradiated
mutagenic treatment to the haploid cells or M1 plants resulted in the development of
tissues in in vitro culture usually drastically a short-duration upland rice mutant line,
decreases their regeneration ability. For this which in Myanmar matured 19 days earlier
reason it is advisable to perform a muta- than the parent variety.
gen sensitivity test for the haploid cell or The mutagenic treatment of haploid cells
embryo survival and regeneration capacity can be followed by selection applied at
before setting up a large-scale experiment. the in vitro stage. If the selected trait is
Methodologies for generating variability. Part 4: Mutation techniques 175
ment with two or three doses of mutagen. now become the most popular system for
M1 plants are harvested individually. Taking mutation detection. The advantage of this
into consideration the chimeric structure of procedure is that it eliminates the number
M1 plants, seeds from different spikes, pani- of false-positive mutation identifications, as
cles or pods are often threshed separately. both cleaved DNA fragments are labelled
Screening for mutations is performed with different fluorescent dyes. With two-
in the M2 population. Depending on the colour imaging, a true mutation has two
programme objective, M2 populations con- mutant bands below the wild-type band
sisting of several hundreds to 20 000 indi- in the same lane. The sum of the length of
viduals are created. To prevent redundancy, the mutant bands in a lane must equal the
usually one M2 plant from each selfed M1 length of the wild-type band in order for
individual or from one M1 spike, panicle the mutation to be confirmed.
or pod is sampled. DNA from each M2 Once the mutation is detected in a DNA
seedling is isolated separately and the M2 pool, the target gene region in all M2 indi-
plant is grown to maturity. Each M2 plant viduals comprising this pool is sequenced
is harvested individually and M3 seeds are and the M2 plant carrying the mutation is
carefully stored. identified. M3 progeny of the identified
Screening for mutations in a target mutant are then used for phenotypic evalu-
gene is based on detecting heteroduplexes ation of the mutated trait.
between the wild-type and mutant-DNA
fragments in the pooled DNA samples. 8.5.1 Case study: Development of
In the first step, polymerase chain reac- waxy mutants in bread and durum
tion (PCR) amplification of the targeted wheat (Slade et al., 2005)
genome region is performed using DNA Development of cereal varieties with waxy
pooled from 5 to 8 M2 individuals as a tem- starch, composed almost entirely of amylo-
plate. Next, the amplified DNA fragment pectine with little or no amylose, has been
is denatured and re-annealed, which allows one of the most important objectives of
formation of heteroduplexes between the commercial plant breeding. Waxy starch
wild-type and mutant DNA strands. Many wheat has broad potential commercial uses
approaches for detecting the mismatched in the food, paper and adhesive industries
sites within the heteroduplexes have been for making better quality products. Despite
tried, such as denaturing gradient gel elec- many breeding efforts over the last two
trophoresis (DGGE), temperature gradient decades, there are no wheat varieties with
gel electrophoresis (TGGE), denaturating fully waxy starch. Using the TILLING
high-performance liquid chromatogra- approach for mutation generation and dis-
phy (DHPLC) or cleavage by the spe- covery, the research team from Anavah Inc.,
cific endonuclease. DHPLC was the first USA, induced and identified 246 mutated
high throughput technology applied for waxy alleles in two elite wheat varieties
the detection of heteroduplexes with the through only one experiment.
mismatched sites. However, cleavage of the
mismatches with CelI endonuclease (the Development of TILLING libraries
novel plant enzyme isolated from celery), Seeds of bread wheat (Triticum aestivum)
followed by the analysis of DNA fragments and durum wheat (T. turgidum subsp.
on denaturating polyacrylamide gels has durum) were treated with two doses of
Methodologies for generating variability. Part 4: Mutation techniques 177
EMS. Elite commercial cultivars were Allelic series of mutations in the target loci
used as the genetic material for mutation In total, 246 new alleles at three waxy loci
induction. For the bread wheat cv. were identified in a population of 1 920 M2
Express, 0.75 percent and 1.2 percent EMS individuals used in the survey: 196 alleles in
treatment was applied, while for the durum hexaploid and 50 in tetraploid wheat. The
cv. Kronos the EMS concentrations were majority of mutations were G to A, or C to
0.75 percent and 1 percent. Each treatment T transitions. Among the identified changes
lasted 18 hours and was preceded by a there were 84 missense, 3 truncation and 5
short (4 minutes) vacuum infiltration (pre- splice junction mutations. These new alle-
soaking) with H2O. After treatment, M1 les encode waxy enzymes ranging in their
seeds were washed for 4 to 8 hours and activity from near wild-type to almost zero.
sown to produce M1 plants, which were A null mutant containing mutations in all
allowed to self-pollinate and were harvested three waxy homologues, highly desirable
individually. TILLING populations of for wheat starch improvement, was also
about 10 000 M2 individuals of hexaploid isolated. As the authors (Slade et al., 2005)
wheat and about 8 000 M2 plants of durum pointed out, the series of alleles created
wheat were created. To prevent redundancy, through TILLING in one experiment rep-
only a single M2 progeny from each M1 resent more genetic diversity than had been
selfed plant was used in the study. described in the preceding 25 years.
Screening for mutations at the waxy loci 8.6. MUTATION TECHNIQUES FOR THE
The target sequences in the bread wheat IMPROVEMENT OF MAJOR CROPS
involved 2 114 kb at the Wx-A1 locus and 8.6.1 Induced mutations in cereals
1 345 kb of the Wx-D1 locus. In the durum Cereals, of all crop species, have most often
wheat, 1 232 kb sequence of the Wx-A1 and been the subject of improvement through
487 kb of the Wx-B1 genes were screened. the use of mutation techniques. In addition
DNA from 1 152 bread wheat and 768 to major cereals such as rice, barley, wheat
durum wheat M2 seedlings was used for and maize, other species, including some
mutation screening. Most of bread wheat exotic ones, have also been the subject of
individuals (768 M2) and all M2 durum mutagenic treatment in several countries
plants analysed came from treatment with (Table 8.2). The majority of more than
0.75 percent EMS. DNA isolated from 1 000 mutant cultivars were obtained after
individual M2 plants was pooled two to radiation, especially gamma ray treatment,
sixfold. The PCR amplification was car- and directly released. However, the ease of
ried out with the use of specific primers mutagenic treatment rather than the kind
labelled with fluorescent dyes IRD700 and of mutagenic specificity determined the use
IRD800. The PCR products were digested of radiation rather than chemical mutagens.
with the CelI enzyme, denatured and re- Additionally, the bio-hazardous nature
annealed. The samples were then separated of most mutagenic components prevents
on denaturating polyacrylamide gel using their use in simply equipped laboratories.
the LI-COR2 DNA sequencer. Images Nevertheless, numerous valuable mutants
were analysed visually for the presence of have been obtained with the use of MNU,
the cleaved DNA fragments indicating a EMS and sodium azide. Another tendency,
mutation in the target region. observed more recently, is to transfer the
178 Plant breeding and farmer participation
TABLE 8.2
Species and the number of officially released cereal mutant cultivars (FAO/IAEA Mutant Varieties
Database actualized)
Crop species Common name No. of released mutant cultivars No. of countries
with released
Total Direct Cross mutant cultivars
TABLE 8.3
Doses of physical and chemical mutagens used for seed treatment in cereals
even if three doses of mutagen have been suggested to postpone selection to the M3
applied. Treatment of 2 0004 000 seeds per generation, which will allow selection to
dose should be sufficient to obtain a large be carried out on a progeny basis. This is
M2/M3 generation. However, the size of the especially important if selection for biotic and
M2 depends on the method of seed collection abiotic stress tolerance or resistance is done
from the M1 generation. If M1 has been set under field conditions in the stress-prone
up in three doses, the breeder can choose area. In this approach, as many as possible,
a population with a significant number randomly chosen, M2 plants are harvested,
of seeds during the harvest. They should threshed and sown on a plant-progeny basis.
consider the level of somatic effects, such as This method is more laborious but increases
mature plant growth reduction, and fertility the probability of selecting numerous plants
and survival reduction. The final decision homozygous for a desired character and
on which dose seeds should be chosen for helps avoid selecting false-positive mutants.
the M2 generation can be taken based on the To avoid having to thresh individual heads,
genetic effect of the mutagen. As M1 plants the most laborious part of this method,
are usually harvested individually (pulled interesting modifications were successfully
with roots), the spikes or panicles can be used for the selection of semi-dwarf barley
collected from a few hundred plants and and salt-tolerant rice. In both cases, heads
planted, without threshing, in boxes for from each M2 plant were harvested and
evaluation of the frequency of chlorophyll sownwithout threshingthe next season
mutants. Depending on the objective of directly in the field. In the case of rice,
breeding, and considering somatic and the panicles were sown in a saline area.
genetic parameters of mutagenic treatment, Numerous surviving lines, with mutants
the breeder can choose from which dose to homozygous for these characters, could be
take the seeds for the M2 generation. Plants directly selected, in addition to segregating
from the other two doses can be threshed progenies. Several hundred thousand M2
and seed stored as a reserve of treated plants could be characterized in the M3
material. progenies using this approach.
In classical work, one or a few spikes Generally, both M2 and M3 generation
or panicles from each individual M1 plant are grown using normal agronomic prac-
are treated separately: collected, threshed tices. Any modification in plant cultivation
and sown as plant progenies, head-to- depends mainly on the selection method. If
row method. Various modifications of the selection of desired mutants was initi-
this approach have been applied, such as ated in the M2, the seeds from selected plant
harvesting single heads from each M1 plant (M3) are sown on a progeny basis to check
in bulk and sowing seeds in rows after mutant homozygosity.
threshing. Should these approaches be too
laborious, mechanical harvesting in bulk, 8.6.2 Case study: Development of
threshing and sowing can be also used. If malting barley cultivar Diamant
selection is initiated in the M2 generation, (Source: Based on FAO/IAEA, 1977)
i.e. on a single-plant basis, seeds are usually The variety Valicky was chosen for
sown double spaced. However, taking mutagenic treatment to improve lodging
into consideration the low occurrence of resistance and yield. This cultivar with high
recessives in the M2 generation, it is strongly malting quality was first released (under the
Methodologies for generating variability. Part 4: Mutation techniques 181
TABLE 8.4
Most frequent and other characters altered in legume crops
Character most Yield, plant type, erect habit, dwarfness, branching habit, phytomass (biological) yield, leaf
frequently modified size and shape, flowering time, maturity, flower colour, number of flowers, pod and fruit
characters (size, length, number of seeds per pod or fruit, non-shattering pods), seed and
kernel size, seed coat colour
Other characters Day length insensitivity in pigeon pea, mungbean and sesbania
modified
Afila-type mutant in pea resulting in modification of leaflets into tendrils which facilitates
mechanical harvesting of green peas
Terminal inflorescence in pigeon pea and faba bean
Higher shelling percentage, thick or thin pod cover, harvest index, seed dormancy in
groundnut
Cotyledon colour in dry seeds
Cold and drought tolerance in soybean and pea
Resistance to bacterial, fungal, and viral diseases in several crops, and insect resistance in
some
Lodging resistance
Drought resistance
Superior nutritive value and protein content in cowpea and pea
Fodder quality in lupin
Nodulation mutants with hypernodulation, nitrate-tolerant nodulation, no nodulation and
ineffective nodules have been isolated in specially designed experiments
TABLE 8.5
Successful gamma or X-ray doses for dry seed exposure in legume crops1
Crop species Common name Successful dose range (Gy)
Arachis hypogaea Groundnut 150250 (11)
Cajanus cajan Pigeon pea 160 (1)
Cicer arietinum Chickpea 100200 (5)
Dolichos lablab Hyacinth bean 240 (1)
Glycine max Soybean 100200 (20)
Phaseolus vulgaris Common bean 100200 (5)
Pisum sativum Pea 100200 (5)
Vigna radiata Mungbean 100200 (7)
Notes: (1) Successful doses defined as the doses that led to the development, registration and release of a mutant cultivar
directly without using the mutant as a parent in crossbreeding. The number of released cultivars is in brackets.
that produces aerial nodules on the stem, dose that led to the development, registra-
was identified when the crop was grown tion and release of mutant cultivars directly
in the off season. The mutant produces tall without resorting to cross breedingare
plants with large phytomass and N-fixing given in Table 8.5. Exposures of seeds to
nodules irrespective of planting time. 100200 Gy, except for faba bean, resulted
in 49 out of 111 legume cultivars developed
Mutagenic treatment as direct mutants. Chlorophyll mutated
Gamma ray or X-ray exposure of the dry sectors appearing on the first true leaves
seeds is the most convenient method for after seed germination in leguminous plants
creating genetic variability in legume spe- can be used to monitor the effect of radia-
cies. Successful dose rangesdefined as the tion and chemical mutagens.
Methodologies for generating variability. Part 4: Mutation techniques 183
TABLE 8.6
Method for isolation and induction of mutations for use in breeding of grain legume crops
Generation Operations
M1 Expose 5000 to 10 000 seeds of the best available cultivar to 100200 Gy of gamma or X-rays.
Plant the M1 generation following normal cultivation practices.1
Harvest the first five pods, or all the pods, from each of the M1 plants.
M3 Grow the selected mutants as single-plant progenies and check for the segregation of the desired
trait. Continue selection for the desired trait on single-plant basis. Uniform, non segregating mutant
progenies, if any, can be bulked at this stage to hasten the breeding cycle.
M4 Evaluate the expression of the selected trait and yield of the bulk lines in comparison with the
parent as well as with the best check cultivar. Record observations on all agronomic parameters,
disease and insect resistance in comparison with the parent cultivar. Repeat the procedure for the
single-plant selections as outlined above.
M5 onwards Follow normal plant breeding procedures with the selected progenies. Evaluate the selected lines
at more than one location. Enter one or two of the best lines at a time in national and regional
evaluation trials, or to local farmers in participatory breeding programmes. Initiate seed multiplication
to meet the demand of the mandatory trials for official approval and release of the cultivar.
Notes: (1) Outcrossing is increased in the M1 plants due to pollen sterility induced by mutagenic treatments. It is desirable
to grow the M1 plants in isolation for facultative crosspollinating species. Selfing of the M1 plants is essential for genetic
experiments.
mutants with alteration in plant entire state, and also for Karnataka
type, yield components and seed State in India.
size. Three large seed size and
At the time of official release as TAU-1,
other putative mutants identified.
150 kg of Breeder's, 3 560 kg of Foundation
Large-seed mutants were obtained
and 1 440 kg of Certified seed were avail-
following 300 Gy exposure.
able. Foundation seed production reached
1976 All mutants were progeny tested,
40 tonne during 199495. Two crops per
and further selection of single
year were grown for experimental work.
plants continued in the M3 and
M4 generations. True breeding lines
8.6.5 Induced mutations in oil and
were isolated, three with large seed
fibre crops
size and 65 for other characters.
Mutation techniques have been used for
1977-78 Large seed size mutants evaluated
improvement of annual oil crops and crops
in yield trials with parent No. 55
providing bast and seed fibres (Table 8.7).
and a newly approved cultivar T-9.
Breeding objectives for using induced muta-
Large seed size mutant yields were
tions are similar to other crops, although
superior to the parent No. 55, but
for oil crops there are unique objectives for
less than T-9 with still smaller seed
the modification of oil quality (Table 8.8).
size (hundred seed weight <4 g).
Most of the oilseed mutant-derived culti-
1979 Large seed size mutant lines UM-
vars released have been developed directly
196 and UM-201 hybridized to
from mutants, but 50 percent of ground-
T-9.
nut and 67 percent of linseed mutation-
1980-81 F1 and F2 populations were grown,
derived varieties were developed from
and selections made for large seed
size, T-9 plant type and yield TABLE 8.7
components, and advanced to F3 Oil crops and fibre plants improved through
and F4. induced mutations
1982-84 Selection 80-7 was found to give Latin name Common name
the highest yield in on-station
Oil crops
trials. Arachis hypogaea Groundnut
1985-87 Line 80-7 was entered in the Brassica campestris Turnip rape
evaluation trials of the Punjabrao Brassica juncea Indian mustard
Agricultural University, and Brassica napus Rapeseed
Euphorbia fulgens Euphorbia
subsequently in the trials of the
Glycine max Soybean
Coordinated Pulses Improvement Helianthus annuus Sunflower
Programme of the Indian Council Linum usitatissimum Linseed
of Agricultural Research. 80-7 gave Ricinus communis Castor bean
24 percent and higher yield over Papaver somniferum Opium poppy
Sesamum indicum Sesame
No. 55, and 9 percent over T-9, the
national check cultivar. Its mean Fibre plants
hundred seed weight was 4.8 to Boehmeria nivea Ramie
Corchorus capsulari Jute
5.0 g. It was first released for the
Corchorus olitorius Tossa jute
Vidarbha region of Maharashtra Gossypium sp. Cotton
State, India, in 1987, and later for the Linum usitatissimum Flax
Methodologies for generating variability. Part 4: Mutation techniques 185
TABLE 8.8
Most frequent and other characters altered with induced mutations in oil and fibre crops
Oil crops Fibre crops
crosses with mutants. In fibre crops, 75 were made using crosses with EMS-induced
percent of the varieties have been directly low-linolenic-acid mutants and radiation-
developed, mainly through treatment with induced high-oleic-acid mutants. The
gamma radiation or X-rays, and 25 percent expansion of canola cultivation in Canada and
through crosses with mutants. For some of Europe is primarily due to its modified fatty
these varieties, in both oilseeds and fibre acid composition, i.e. low erucic acid, low
crops, remarkable economic gains have linolenic acid and high oleic acid, combined
been reported. In breeding of oilseed rape with a low content of glucosinolates, which
and sunflower, spontaneous and induced improved the nutritional quality of the
mutants have been used in combination oil for human nutrition, processing and
with conventional breeding methods to storage, and the meal for livestock feed.
modify oil composition and increase yield. Microspore mutagenesis has also been used
Traditional oilseed rape has high in combination with in vitro screening for
erucic acid content in the oil and high the development of B. napus tolerant to the
glucosinolate levels in the meal, and both herbicides imidazoline and chlorosulfuron.
components are nutritionally undesirable. In Canada, Australia and Europe, many
They have been reduced by breeders after Brassica varieties with a modified oil-profile
identifying and using spontaneous mutants are based on mutant germplasm (Bhatia,
in the development of canola with less than Nichterlein and Maluszynski, 1999).
2 percent erucic acid and less than 30 M/g A number of mutant varieties with
of aliphatic glucosinolates in the meal. This changed oil composition have also been
process was facilitated by the development developed in other species. In sunflower,
of analytical methods for quality assessment after mutagenic treatment with dES, the
suitable for screening individual plants or mutant cultivar Pervenets with high oleic
single seeds of large breeding populations. and low linolenic acid content was developed,
Canola quality has been developed for and widely used for hybrid production with
Brassica campestris (turnip rape), Brassica high oleic acid content in the United States
napus (oilseed rape) and Brassica juncea of America and Europe, to produce oil with
(Indian mustard). Further improvements the high oxidative stability preferred for
186 Plant breeding and farmer participation
food processing (e.g. frying). The oil with Pakistan is planted to this cultivar. The new
high oleic acid content has steadily gained cultivar NIAB Karishma, released in 1996,
market acceptance, leading to increasing and derived from a cross of the mutant
cultivation areas of high-oleic sunflower cultivar NIAB-86 with an American strain
cultivars. In linseed, a doubled mutant has W 83-29, is early maturing, has improved
been developed after crossing two individu- heat tolerance, high yield potential and has
ally selected mutants with reduced linolenic been cultivated on 486 000 ha.
acid obtained after EMS treatment, and used
in further crosses, resulting in the release of Mutagenic treatment
a number of low-linolenic-linseed cultivars Both ionizing radiation (gamma rays or
in Australia and Canada. X-rays) and chemical treatments have been
In cotton, the development of NIAB- applied to dry, dormant seeds of oil and
78, a gamma-ray induced, high yielding fibre crops. The successful dose ranges for
mutant cultivar, released in 1983, had great radiation, defined as the dose that led to
economic impact in Pakistan. Developed the development, registration and release
at the National Institute of Agricultural of varieties derived directly from mutants
Botany, NIAB-78 had a marked influence without using mutant crosses are given in
on sustaining the textile industry of Table 8.9. The doses given should be con-
Pakistan, and contributed to the national sidered as an orientation, because tolerance
economy in several ways. The variety, to seed irradiation can differ between varie-
ideal for a cotton-wheat rotation, had a ties. The recently released variety Abasin-
shorter stature, determinate growth habit, 95 was developed after gamma ray treat-
tolerance to heat and escaped bollworm ment of the Canadian variety Tower using
attack due to its early maturity. Within a much higher dose, 1 400 Gy, than for
five years of release, its cultivation doubled the development of most of the other
cotton production in Pakistan, and even oilseed rape varieties. Chemical mutagens
after 14 years of its first release, nearly were much less used; however, they led to
25 percent of the area under cotton in the development of important commercial
TABLE 8.9
Successful gamma or X-ray doses for dry seed exposure in oil seeds and fibre crops
Crop species Successful dose range (Gy)
Oil crops
Arachis hypogaea (groundnut) 150250 (11)
Brassica juncea (Indian mustard) 700
Brassica napus (rapeseed) 600800
Glycine max (soybean) 100200 (20)
Linum usitatissum (linseed) 100 (1)
Papaver somniferum (opium poppy) 50 (1)
Ricinus communis (castor bean) 400 (1)
Sesamum indicum (sesame) 100200 (5)
Fibre crops
Corchorus capsularis (jute) 250 (1)
Gossypium spp. (cotton) 200400 (5)
Notes: Successful doses defined as the doses which led to the development, registration and release of mutant variety
directly without using the mutant as a parent in cross breeding. The number of released cultivars is in brackets.
Methodologies for generating variability. Part 4: Mutation techniques 187
varieties with altered oil composition, e.g. position, the identification of mutants can
high-oleic-acid sunflower (dMS), low-lino- be done using half-seed methods, screen-
lenic oilseed rape (EMS) and low-linolenic ing M2 seeds harvested from M1 plants.
linseed, the so-called linola (EMS). For traits such as tolerance to biotic and
abiotic stresses, it is recommended to do
Growing and handling of the early the mutant screening not on a single-plant
generations level in the M2 but in M2 progeny rows
The general procedures for growing and (M3). Mutants (M4) with low agronomic
handling the early generations of oil and performance, but valuable mutant traits,
fibre crops are outlined in Table 8.10. For can be improved through backcrossing to
the detection of plants with altered oil com- the parent or by other crosses.
TABLE 8.10
Method for isolation and induction of mutations for use in breeding of oil and fibre crops
Generation Operations
M1 Expose 5 000 to 10 000 seeds of the best available variety or homozygote line to gamma or X-rays
Plant the M1 generation under optimal conditions to produce M2 seeds, following normal cultivation
practices, growing them in isolation or bagging before flowering.
Harvest the first two to five pods, capsules, bolls or heads; the M1 plants can be harvested
individually if the M2 will be grown in progeny rows, or in bulk if the M2 will be grown in bulk.
M2 For traits expressed and visible at the single-plant level: grow the M2 population as plant-
progeny rows (3050 plants from each M1 plant), and remaining seed can be stored for sowing in
subsequent seasons.
Look for all morphological and physiological changes in each M2 plant from the seedling stage to
harvest, and harvest the selected plants individually.
For traits expressed at the seed level (e.g. oil composition in oil seeds): cut part of the M2 seed
(linseed, sunflower) or germinate M2 seed (Brassica spp.) and cut one cotyledon for fatty acid
analysis, then continue cultivating the seed or seedlings with desired fatty acid composition.
For traits not expressed at the single-plant level: multiply seed of each M2 plant, and harvest single
plants for row evaluation in M3.
M3 Grow the selected mutants as single plant progenies and check for the segregation of the desired trait.
Continue selection for the desired trait on a single-plant basis.
Uniform, non segregating mutant progenies, if any, can be bulked at this stage to speed the
breeding cycle.
Select on an M2 plant-progeny basis for traits such as resistance to stresses and quality (not reliably
expressed on a single-plant basis).
M4 Evaluate the expression of the selected trait and yield of the bulk lines in comparison with the
parent as well as the best check cultivar.
Record observations on all agronomic parameters, disease and pest resistance compared with the
parent variety.
Repeat the procedure for the single-plant selections, as outlined above,
Mutants with valuable traits but undesirable characters should be backcrossed with parent or
used in crosses.
M5 generation Follow normal plant breeding procedures with the selected progenies: preliminary yield trials,
and beyond multilocation evaluation of mutants or lines derived from crosses with mutants, submission of one
or two of the best lines, at a time, for national or regional evaluation trials, or to local farmers in
participatory breeding programmes.
Initiate seed multiplication to meet the demand of the mandatory trials for official approval and
release of the variety.
188 Plant breeding and farmer participation
8.6.6 Oil crop Case study 1: traditional oil with good drying properties,
Development of the high yielding as a result of the advent of plastic paints. A
oilseed rape variety Abasin-95 programme to change the oil quality from
(Nuclear Institute for Food and industrial (high linolenic acid content) to
Agriculture, Peshawar, Pakistan) edible (low linolenic acid content) was con-
The objectives were to improve productivity ceived and proved to successful. Various
and resistance or tolerance to biotic and stepssee Table 8.11applied before and
abiotic stresses. during the programme illustrate well the
1988 10 00015 000 dry seeds of general principles that should be followed
oilseed rape (Brassica napus) cv. in the use of induced mutations in improve-
Tower with 10 percent moisture ment of industrial crops.
were irradiated with 1 000, 1 200
and 1 400 Gy gamma rays (60C), 8.6.8 Mutation induction enhanced
and planted directly in the field breeding of asexually or vegetatively
in isolation, as M1. At maturity, propagated crops
four pods from every primary According to the FAO/IAEA Mutant
branch were harvested and seeds Variety Database (MVD, no date), about
were bulked on a dose basis. three-quarters of all released mutant-
198990 M2 population was grown in derived cultivars are sexually propagated
rows and after every 20 rows species, while only a quarter are asexually or
the parent variety was included vegetatively propagated (AVP) crops such
for comparison. Individual M2 as ornamentals, trees and shrubs, fruits, root
plants were selected on the basis and tuber crops, and sugar cane. Mutation
of their short stature, early matu- induction shows its most promising aspects
rity, heavy bearing, long pods, in AVP crops compared with cross-breeding
more grains per pod, bold seed, methods due to its ability to change only a
stress tolerance or a combina- very few characters of an otherwise good
tion. The mutant RM-152-2 cultivar without significantly altering the
(1 400 Gy treatment) and other remaining, and often unique, genotype.
mutants were selected, and fur- Coupled to biotechnologies such as somatic
ther advanced to M5. embryogenesis or micropropagation,
1995 RM-152-2 officially released as a mutation induction might be considered
new cultivar: Abasin-95. an obvious means of conventional plant
breeding, and as a possible shortcut for
Two generations a year were grown from inducing desired genetic alterations in
M2 to M4 to speed up the breeding process asexually propagated cultivars. Obviously,
(winter in Peshawar; summer in Kaghan). mutation induction is the only means for
producing genetic variability in vegetatively
8.6.7 Oil crop Case study 2: Improving propagated sterile crops and in obligate
nutrition value. apomicts (Broertjes and van Harten, 1988).
Another case study describes the re-ori- As good mutation practice, the cultivar
entation of breeding objectives for linseed to be mutagenized is generally chosen for
and use of induced mutations as a response its outstanding agronomic performance and
to shrinking traditional markets for the good adaptation, and the least number of
Methodologies for generating variability. Part 4: Mutation techniques 189
TABLE 8.11
Breeding linseed (Linum usitatissimum) varieties for new uses with altered oil quality
Step Description of breeding procedure
Screening of crop germplasm Previous screening for low linolenic acid in linseed germplasm, but none was
found.
Screening of wild relatives Screening in wild species for low linolenic genotypes, and some were found,
but crossing with cultivated linseed failed.
Mutagenic treatment of linseed Seeds of locally adapted variety Glenelg were treated with gamma rays (300
to 900 Gy) and two doses of EMS (0.3 percent and 0.4 percent, 2 h at 2oC and
2 h at 20oC
Screening method development A rapid, efficient half-seed screening method was developed using a colour
reaction.
Mutant confirmation Reserve half-seeds of those with positive reaction were planted, selected and
progeny tested; most proved false positive.
Genetic study Two independent recessive mutants were induced by 0.4 percent EMS; both
showed reduced linolenic acid contents (from 46 reduced to 28 percent
linolenic acid) and proved to be at different loci; recombination of the two
mutated genes.
Identification of double mutant Identification of double mutant Zero from recombination of the two single
mutants, with only 2 percent linolenic acid (C18:3), and from 19 to 63 percent
increased linoleic acid (C18:2).
Backcrossing of mutants to Initial mutants showed low yield, and were backcrossed to parent Glenelg to
parent eliminate undesirable mutations. After four BCs, some lines exceeded parent.
Crossing of Zero with other Zero was crossed with a number of varieties and breeding lines, allowing
varieties selection of high yielding recombinants with Zero quality.
Release of varieties with altered Release of low linolenic acid (Linola TM) varieties Wallaga (1992), Eyre
oil quality (1992), Argyle (1994) in Australia; Coniston (1992) and Derwent (1992) in
UK; and Linola 947 (1993) and Linola 989 (1993) in Canada.
Sources: From data presented by A.G. Green (CSIRO, Australia) and co-authors in numerous publications during 19842003.
TABLE 8.12
Radiation doses and plant material used for the induction of somatic mutations in AVP crops
Genus or crop Plant material treated Dose
Ornamentals
Achimenes Detached leaves 30 Gy
Alstroemeria Rhizomes 46 Gy
Azalea Rooted cuttings 1020 Gy
Begonia Detached leaves 1525 Gy
Buddleia Plants 2030 Gy
Canna Rhizomes 1030 Gy
Chrysanthemum Rooted cuttings 1520 Gy; 6121012n(th)/cm2
Clematis Rooted cuttings 25 Gy
Conifers Rooted cuttings 0.55 Gy
Cosmos Rooted cuttings 20 Gy
Crocus Dormant bulbs, directly after harvest 1015 Gy
Dahlia Freshly harvested tubers 1525 Gy
Dianthus Rooted cuttings 4060 Gy
Dianthus Unrooted cuttings (base shielded) 80100 Gy
Endymion Detached leaves 15 Gy
Euphorbia Rooted cuttings 3050 Gy
Forsythia Rooted cuttings 4080 Gy
Gladioulus Dormant corms (2n) 40 Gy
Hyacinthus Bulbs, before-wounding basis 25 Gy
Iris Freshly harvested corms 10 Gy
Kalancho Detached leaves 1520 Gy
Laburnum Plants 2030 Gy
Lilium Bulb scales 2.5 Gy
Malus Just-grafted plants 2030 Gy
Muscari Detached leaves 1015 Gy
Narcissus Dormant bulbs, directly after harvest 510 Gy
Ornithogalum Detached leaves 510 Gy
Potentilla Rooted cuttings 6080 Gy
Prunus Just-grafted plants 2030 Gy
Rhododendron Rooted cuttings 3050 Gy
Roses Budding wood 2040 Gy
Dormant plants 40100 Gy
Saintpaulia Detached leaves 3040 Gy
Scilla Detached leaves 15 Gy
Streptocarpus Detached leaves 30 Gy
Syringa Plants 30 Gy
Tulip Dormant bulbs, directly after harvest 35 Gy
Fruit crops
Apple Grafts 3040 Gy
471012 n(th)/cm2
Banana Corms 2550 Gy
Blackberry Young dormant plants 6080 Gy
Blackcurrant Dormant woody cuttings 30 Gy
Cherry Grafts 2030 Gy
471012n(th)/cm2
Grape Dormant buds 1030 Gy
Lemon tree Cuttings 2070 Gy
Orange tree Dormant scions 50 Gy
Methodologies for generating variability. Part 4: Mutation techniques 191
TABLE 8.12
Continued
Genus or crop Plant material treated Dose
Peach Summer buds 1040 Gy
Pear Grafts 4050 Gy
Plum (European) Dormant scions 4060 Gy
Raspberry Spring suckers 10 Gy
Strawberry Young runner plant 150250 Gy
can be started, the mutated cells should based on in vitro plant material translates
participate in the formation of a shoot or into an important gain in efficiency.
plant. If some type of adventitious bud
technique has been used, many of the Mutagenic treatment
mutants will be solid and early selection is The most common mutagen used with AVP
possible. Work on irradiated multicellular crops is radiation. Bulky material, like bulbs,
apices necessitates the mutated sector size stolons or scions for grafting, is difficult to
to be increased in order to quickly obtain treat in a reproducible way with chemicals.
completely homohistont tissue, or at least The chemical mutagen must penetrate to
stable periclinal chimeras. These conditions the meristematic zones, and the excess of
can be reached in not fewer than 3 to 4 chemical has to be removed after treat-
vegetative propagation cycles. The devel- ment. The procedures for acute or chronic
opment of advanced in vitro techniques, irradiation are rather simple, with good
and the extension of these techniques to repeatability and high mutation frequency.
otherwise neglected species, has created All types of ionizing radiation are effective;
new potentials and opportunities to induce in practice, however, it is likely that only an
mutations in any AVP crop. In vitro plant X-ray machine or a gamma-ray source (e.g.
material, such as apical meristems, adventi- 60Co) are available. The dose to be applied
tious or axillary buds, embryogenic calli, depends on the radiosensitivity of the spe-
micro-cuttings, cell suspensions or pro- cies, cultivar, plant development stage and
toplasts, compared with in vivo starting also of the plant part to be treated. Plant
material, allows the treatment of larger parts that have developed new, adventi-
populations in less space, and plantlets are tious roots and shoots, e.g. unrooted cut-
maintained in a controlled, disease-free tings or freshly detached leaves, are more
environment, facilitating the recovery of sensitive than plant parts with existing root
mutants. Avoidance or dissociation of chi- and shoot meristems. Thus it is difficult to
mera and rapid clonal propagation of useful predict the dose that would be efficient in
mutants, as well as the production of a large any new mutation experiment, even for the
number of plants for further evaluation same cultivar. The best practice is empiri-
192 Plant breeding and farmer participation
TABLE 8.13
Actual in vitro Musa mutation induction enhanced breeding process using shoot-tip culture and
selection in the field
Steps Time Additional information
(months)
Sucker from the field 1986 ; cv. Grande Naine (AAA) ; ITC collection
T0
Shoot tip culture
Radiosensitivity test T0 + 6 19871988; FAO/IAEA Agriculture and Biotechnology Laboratory,
Seibersdorf, Austria
60Co irradiation: M1 T0 + 12
Micropropagation M1V1 T0 + 13 Radiosensitivity tests on a minimum of 200 shoot-tips
Micropropagation M1V2 T0 + 14 Irradiation of a minimum of 2000 shoot tips with an LD50 dose of -rays
Micropropagation M1V3 T0 + 15
Rooting M1V4 T0 + 16
Acclimatization to soil T0 + 17 19881990; line GN-60A; putative early flowering mutant; glasshouse
of the FAO/IAEA Agriculture and Biotechnology Laboratory, Seibersdorf,
Field Selection T0 + 24 Austria
Micropropagation of desired 19901993; vegetative progeny sent to Honduras, Australia, South Africa
T0 + 60
plants and Malaysia for field-testing under commercial plantation conditions.
Not all the plants were demonstrating earliness. The chimeric constitution
Multilocation trials T0 + 84 of the original plant could explain this behaviour as only progenies
deriving from the mutated sector gave rise to the putative early mutant.
September 1993, in Malaysia, only the early flowering plants obtained
in the field were tissue cultured again to produce about 2000 plants for
commercial evaluation in the United Plantations Bhd.
cal. Based on empirical data gathered over is greatly reduced, if not impossible. Which
more than forty years, some dose ranges dose level should or can be applied depends
may be inferred as guidance for a test series on the crop, the type of propagation avail-
(Table 8.12). For vegetatively propagated able, the numbers that can be handled and
crops, doses less than LD50 are usually the selection method.
employed. A moderate dose that permits
good growth and propagation of the mate- 8.6.9 Case study: Banana mutation
rial is to be preferred. Too many mutations induction using in vitro plant material
per cell may be induced at high dose levels, Based on information from N. Roux on
with the risk that a favourable mutation his own and F. Novaks experiments at the
is accompanied by one or more that are FAO/IAEA Agriculture and Biotechnology
unfavourable. In AVP crops, it is very Laboratory, Seibersdorf, Austria.
difficultif not impossibleto separate 1986 Initiation of shoot-tip culture.
favourable mutations from unfavourable Suckers from the field of cultivar
ones occurring in the same cells because Grande Naine (AAA), ITC col-
recombination through crossing or selfing lection.
Methodologies for generating variability. Part 4: Mutation techniques 193
CHAPTER 9
Selection methods
Part 1: Organizational aspects of a
plant breeding programme
Salvatore Ceccarelli
196 Plant breeding and farmer participation
organizational issues limit the freedom of the these reasons, some breeding programmes
breeder in terms of breeding strategies (for do extensive evaluation, selection and testing
example, the choice of testing the breeding in farmers fields. Even though the choice of
material under a given level of inputs). The evaluating and testing the breeding material
only advantage of following a commercial in farmers fields has nothing to do with
crop on station is that a commercial crop is participatory plant breeding (PPB) (it could
expected to leave the land highly uniform. possibly be considered participatory variety
However, this is not necessarily true because, selection PVS), it is expected to have
for example, the uneven application of inputs a number of advantages over on-station
can actually create additional sources of evaluation, selection and testing. These
uncontrollable variation. advantages derive from exposing the
A much better type of organization breeding material to a multitude of target
is when each breeder is allocated two to production areas at an early stage of the
three times the amount of land needed for breeding programme, assessing the response
the breeding trials and nurseries in a given of different breeding material to a range
cropping season to be able to manage, of different soil types, soil depths, rainfall,
according to their breeding philosophies agronomic management, etc.
and strategies, not only the portion of land The theoretical framework for discussing
allocated to current trials and nurseries, response to decentralized selection and more
but to manage (in terms of rotations and generally the optimum environment for
inputs) also the land that will host trials and selection, was developed by Falconer (1952,
nurseries in one or two years. This improves 1981), who demonstrated that selection in
considerably the situation compared with the target environment is almost invariably
the organization described earlier, but it is more efficient than indirect selection, i.e.
not without negative aspects. The breeder selection in a different environment. This
has often to produce and store, or to has been confirmed by the theoretical
purchase, the seed for the cover crops to work of Rosielle and Hamblin (1981) and
precede the trials, and has to supervise the Simmonds (1991), and validated by data
agronomic operations to make sure that the from numerous experiments, reviewed by
complex of rotation and management under Ceccarelli (1996).
which the cover crop is grown represents At the same time, the superior efficiency
exactly the conditions under which they of selecting in target environments has
intend to test the breeding material. also been disputed by several scientists.
Even when breeders have full control of However, in the majority of cases (such as
the land and of the agronomic operations, the Atlin, McRae and Lu, 2000; Rizza et al.,
situation on the research station will never 2004; Dodig et al., 2008) this was based on
be able to represent the variable agronomic data from a narrow range of yield levels (see
situations under which the crop is actually also Chapters 2 and 20).
grown by farmers. Often, particularly in It is important to specify that all
developing countries, farmers grow the same breeding programmes have some degree of
crop in a number of different rotations and decentralization in the sense that, sooner or
with different levels of inputs depending on later, the breeding material is tested outside
the environment and on the wealth of the the research station. However, we restrict
farmers and their access to the market. For the use of the term decentralized breeding
198 Plant breeding and farmer participation
support staff member, but shared with the before leaving the field, it is possible to
breeder(s) (see also the section below). quickly check the data through sorting
and ranking, and immediately correct
9.2.4 Data capture, storage and typing mistakes;
analysis data can be collected in the field under a
The traditional manner of organizing data wider range of climatic conditions than
capturing is the manual recording through with field books;
field books. Field books can be produced data analysis can immediately follow data
using specialized software tools, such as collection, thus providing an additional
AGROBASE (www.agronomix.mb.ca), means of checking for errors in data entry
Excel or databases such as Access. while the crop is still in the field; and
Manual capturing of data has a number of use of memory cards enables one to keep
disadvantages, including: at hand in the field all the relevant infor-
the preparation of field books is time mation concerning all trials and nurseries
consuming; in a large breeding programme.
note taking is weather dependent (field At the end of the season it is always pos-
books are very difficult to use on windy sible to produce a printout of all the files
or wet days); and to maintain a hard copy of all the data.
the data are handled twice, being written Safe data storage is a major issue in
in the field book first and entered in plant breeding programmes. Examples of
the computer later, thus increasing the strategies that can be used to reduce to a
probability of manual errors; and minimum the risk of data loss are frequent
the time required for data entry delays backup; storage of data in external disk
statistical analysis, usually until after har- drives; and storage of data in at least one
vesting, hence reducing the possibility of computer never connected with networks
detecting errors by examining the results or the Internet to reduce the probability of
of an analysis conducted immediately introducing viruses.
after the data are collected.
Today data capturing can be easily done 9.2.5 Farmer participation
electronically using palmtops (there are Farmer participation in a CBP (farmers
very many types available on the market) visiting and selecting from trials on-station)
or specifically designed devices, which are is not included as a section under PPB
usually much more expensive. The file, because farmers being invited to a station to
which will normally be printed as a field select between lines, hybrids or clones do
book when data capture is by hand, is not have a chance to develop any sense of
downloaded into the main memory or in ownership of the material they select, which
the flash card (recommended) of a palmtop is usually associated with participation in a
(they usually handle a variety of file types, cyclic, as opposed to linear, process (see also
depending on the brand), which can then be later). Therefore, as noted earlier, farmer
taken to the field to enter data. Electronic participation in a CBP is more akin to PVS
data capture has a number of advantages: than to participatory plant breeding.
data are entered manually only once As practical experience confirms, farm-
and then transferred electronically to the er selection is environment-dependent
main computer for analysis; (Ceccarelli et al., 2000), so farmers par-
200 Plant breeding and farmer participation
ticipating in on-station selection should be sion, and this explains the reluctance with
invited from areas that are similar to the which several plant breeders, particularly
research station in terms of the climatic, those in the developing countries, operate
agronomic and agricultural systems when away from their research stations.
the breeding programme has only one sta- Within a research station, all the opera-
tion. When the breeding programme has a tions associated with running a breeding
number of stations simulating a number of programme are shared by staff belonging to
different environments, different groups the same institution and having daily inter-
of farmers should be invited to different action (which does not necessarily make
research stations, unless this activity is part things easier). When a number of stages
of a research activity involving all farmers are transferred outside the research station,
making selections on all stations for the a number of operations can, and actually
sake of comparison. should, be shared with staff belonging to
As farmer visits are usually a single other institutions or to out-posted staff of
event, the organizational issues involve the same institution, or a combination.
transportation of the farmers to the station Depending on the presence or absence
and back, and possibly accommodation for of a strong extension service, and of the
those farmers coming from far away. The structure of the research institute respon-
logistics can be simplified by inviting only sible for the plant breeding, a number of
nearby farmers. different scenarios are possible.
Farmers visits to research stations are a In the case of countries with a strong
typical activity of the research programmes extension service and the presence of
(not only breeding) in developed countries, regional (or subregional or provincial)
where logistic problems are much reduced research centres with infrastructure such
as farmers are usually able to travel to and as offices, computer facilities, agricultural
from the station by themselves. equipment (including plot machinery), a
If the visit aims at farmers selecting DBP could be organized based on the fol-
finished or nearly finished varieties, other lowing principles:
organizational issues include whether the The scientist(s) at the institutes head-
farmers give a formal score to the breeding quarters are responsible for the prepa-
material; whether and how the breeders use rations of trials (choice of entries, plot
farmers preferences as an additional selec- size, experimental design, and having the
tion criterion; whether and how the materi- seed in envelopes ready for planting), the
al selected is made available to the farmers; preparation of field books (or electronic
and what follow up activity will occur. files for electronic capture of field data),
the preparation of draft field maps with
9.3 DECENTRALIZED BREEDING possible alternatives for the layout of the
PROGRAMMES trials, and the shipment of trials with all
Transferring a breeding programme to out- the detailed instructions for planting and
side a research station almost always implies note taking.
losing some degree of control of a number At the headquarters there will be a central
of steps and operations. This is often asso- database where all the information gen-
ciated with the perception that less control erated in the breeding programme is kept.
by scientists is associated with less preci- Information generated in the regional
Selection methods. Part 1: Organizational aspects of a plant breeding programme 201
9.4.1 Setting criteria to identify target breeding. In fact there is no reason why the
environments and target users approach should be confined to work with
A decentralized participatory plant low-income farmers. Basically, when done
breeding programme may lose a great deal properly, PPB is an approach that, even if
of its potential effectiveness if the sample applied in a variety of modes, merges the
of both environments and users in which technical knowledge of the scientists with
the programme is implemented do not the knowledge of the farmers, which is
represent both the target environments historically based on millennia spent in
and the target users. In order to do that, domesticating wild plants and adapting the
setting the criteria for identification of the resulting crops to a multitude of different
target environments and users is a critically environments. Therefore, in principle, PPB
important step. can apply equally well even in situations of
In setting the criteria, it is useful also market-oriented agriculture in favourable
to assign priorities to the different cate- environments.
gories of environments and users so that, The main criteria to identify farmers can
depending on the resources available to the be grouped in three broad categories:
programme, environments and users can be Farmer characteristics These include lan-
added or discontinued on the basis of prior- guage, ethnicity, caste, age, gender, income,
ities established in an ideal context. education, market relations or orienta-
The most obvious criterion for the tion, membership of farmer organizations
choice of the target physical environments, (unions or cooperatives), and relationships
is their representativeness of the major pro- among groups within the same commu-
duction areas for a given crop (or for the nity and between communities.
crops covered by the programme) in terms Farmer expertise This includes the need to
of climatic conditions (temperature, rain- understand whether farmers are already
fall, elevation), agronomic practices, soil practising some types of plant improve-
types, landscape, etc. The criteria for the ment, as this is essential in the choice of
choice of the socio-economic environments the breeding methodology (see below).
are closely interconnected with those of the In some communities, e.g. Eritrea, spe-
target users. Therefore the programme has cific individuals have specific respon-
to decide whether to work for all the vari- sibilities in relation to crop and variety
ous socio-economic environments present introduction (Soleri et al., 2002).
in the target area, or to privilege the most Farmer needs These include the needs
difficult environments where farmers have of different groups, their perception of
fewer opportunities for market access and risk and hence the type of variety they
where most of the agricultural products consider more appropriate in term of
are used within farms or within the com- stability and yield (Soleri et al., 2002),
munity, or to work only for the most and the need for special quality attributes
favourable, high potential, environments. either for feed or for food, or both. These
It has been argued that PPB has evolved include also the farmers understanding
mainly to address the difficulties of poor of production limitations with reference
farmers in developing countries (Ashby to the use of fertilizers, appropriate rota-
and Lilja, 2004) which have been largely tions and irrigation. It is also important
bypassed by the products of conventional to understand farmers needs in terms
204 Plant breeding and farmer participation
of seed supply, because it makes a large sible to give a cookbook formula for what
difference whether the farmer predomi- works better. In general, if some groups
nantly use their own seed (or the seed or individuals tend to be discriminated
of their neighbours), or usually buy seed against, it may be appropriate to have sepa-
from the formal sector. rate meetings with different social, gender,
age or wealth groups.
9.4.2 Identification of the target In the process of identifications of users,
environment and users it is very important to clarify (i) what plant
Once the criteria are set, the actual process breeding can offer and how long it can take;
of identification needs the involvement of (ii) what sort of commitment in land, time
partners who have a very good knowledge and labour is required from the farmer;
of both the environment and the users. (iii) what is the risk for the farmer and
These are typically the staff of the extension how this can be compensated for (in-kind
service or the staff of the outlying research compensation vs. money), and (iv) what
stations. The first step is to set meetings the overall benefits are that the farmers can
with all the stakeholders with the objective expect if everything goes well.
of identifying partners and locations. In these meetings it is also essential to
In this phase there are some potential understand what sources of seed farmers
biases that can affect the success of PPB. use for the various crops, to anticipate
Key decisions affecting the participatory which type of change the participatory
programme are (i) whether to seek individ- programme might introduce, and to make
ual or group participation; (ii) whether the sure that farmers are aware and prepared to
participants should be experts (germplasm absorb the changes.
experts are farmers who regularly experi- The organizational issues of the choice of
ment with varieties, are able to recognize sites are both at the macro- (identification
important intra- as well as inter-varietal dif- of villages or locations within a country or a
ferences, and who target specific varieties region) and at the micro-level (identification
to different micro-niches) or whether they of the field within a village for planting the
should represent the wider community; trial(s). The participation of farmers in the
and (iii) whether equity should be the main identification of the fields is unavoidable
objective in the identification of the users. because it is associated with the relevance
Meetings with all different typologies of of the results and with the issues of who
farmers may be inappropriate without a participates and who benefits: it is at this
proper knowledge of the power relation- point that small-scale farmers run the risk
ships within the community. This usually of being excluded as active participants
leads to a few farmers monopolizing all the because their land is not large enough to
discussions reducing the possibilities for host trials in addition to the farmers crop.
others to express their views. This danger As we will see later, it is possible to find
varies greatly with the culture: in some experimental designs that allow distributing
cultures, women are not even allowed to relatively large number of entries in small
attend meetings; in others, they can partici- blocks, each planted in a different farmers
pate with a passive role; and in others they field.
can participate freely and with the same An additional organizational issue in the
rights as the men. Therefore, it is not pos- choice of the sites, which is associated with
Selection methods. Part 1: Organizational aspects of a plant breeding programme 205
the issue of the breeding philosophy, is tion. Narrow and wide are relative terms;
whether they should be sufficiently repre- therefore, for an international breeding
sentative to allow some degree of extrapola- programme, a widely adapted variety is
tion of the results to other sites, or whether a variety performing well in a number of
the priority should be to meet farmers countries, while for a national breeding
needs to target micro-niches. In practice, it programme it is a variety performing well
is advisable that sites do represent the range in several locations within the country,
of environmental and agronomic conditions while, ultimately, to a farmers it means
in which the crop is grown, because this is a variety performing well across crop-
known to have a major effect on farmers ping seasons without too much concern
selection (Ceccarelli et al., 2000, 2003). whether it performs well elsewhere.
Participatory breeding programmes It is difficult to reach an optimal alloca-
are often seen exclusively as programmes tion of resources regarding to the number
leading to niche varieties, adapted to only of sites and to the number of farmers at
a restricted complex of environmental and each site. As we will see later, it is possible
social characteristics (see also Chapter 4). to organize a PPB programme in such a way
This is not necessarily true, as the type of that GE interaction, and more specifically
adaptation (narrow or wide) of the varieties Genotype Location (GL) and Genotype
emerging from a PPB programme is largely Years within Locations (GY(L)) will
dependent upon the nature of the locations eventually optimize the overall structure,
and the users. If the locations covered by the at least from a biological point of view. This
programme represent a mix of favourable aspect is covered in depth in Chapter 20.
and unfavourable growing conditions, it
may be expected that the more uniform 9.4.3 Type of participation
environmental conditions that generally Several scientists (Biggs and Farrington,
characterize favourable environments will led 1991; Pretty, 1994; Lilja and Ashby, 1999a,
to the selection of the same varieties across a b; Ashby and Lilja, 2004) discriminate
number of locations (widely adapted in a among different types or modes of partici-
geographical sense), assuming that farmers pation, which are not necessarily mutually
preferences are also homogenous across the exclusive, although there may be trade-offs
same locations. In the more unfavourable among the impacts of the different types.
conditions, one can expect that more location- Based on two groups of decision-makers,
specific varieties (narrowly adapted) will be namely scientists, which includes research
selected. Eventually, even if the selection is programmes and extension agencies, and
conducted independently in each of many farmers, which refers to the intended users
locations, giving the impression that selection of the participatory breeding programme
is for specific adaptation, the process will not varieties, PPB is categorized by Ashby and
discard a truly widely adapted genotype if Lilja (2004) as:
such a genotype does exist in the breeding Consultative Scientists make the decisions
material (Ceccarelli, 1989). Therefore a PPB alone, but with organized communication
programme easily results in a mixture of with farmers. Decisions are not made
widely and narrowly adapted varieties. with farmers nor delegated to them.
What is discussed above also depends on Collaborative Decision-making authority
the definition of wide and narrow adapta- is shared between farmers and scientists
206 Plant breeding and farmer participation
considering whether and how farmers are regardless of whether and which skills
handling genetic diversity. The rationale is farmers have in handling genetic variation,
as follows. As described in Chapter 3, the can not ensure true participation, as farmers
generation of variability is the first step of will be confronted with methodologies
any breeding programme, conventional or they can not relate to anything with which
participatory, followed by the utilization they are familiar.
of variability and eventually the testing of In addition to the examples given earlier,
the prospective varieties. In a number of breeding methods may differ for the same
countries, farmers do use genetic diversity crop within the same country. Using Africa
either as a specialized activity within the as an example, barley is grown in Ethiopia
community, or as an individual initiative and Eritrea both as food and feed (largely
(Chapter 22). landraces) and also for malt production for
For example, in Eritrea it is common for local breweries. While population methods
farmers to select individual heads within can well be used in the first case, pedigree
a wheat or a barley plot, plant them as breeding is suitable in the second.
head rows in a small portion of their field, An issue related with the choice of the
decide whether to bulk one or more rows breeding method is how much breeding
and start testing the bulk in the field of material farmers can handle. This is a
other farmers, initially on a small scale controversial issue, and several scientists
and gradually on a larger area. One of the believe that farmers can only handle a
most widely grown wheat varieties in the very limited number of genotypes and
country has been developed starting from therefore, implicitly, believe that the only
a small seed sample bought by an expert form of participation is PVS. If true, this
farmer in a local seed market and planted will make it impossible to implement true
initially as spaced plants. In Nepal, before PPB programmes, because plant breeding
harvesting the crop, farmers growing the needs to start from a sufficiently large
old barley landraces habitually collected sample of genetically variable material.
a sample of heads representing all the dif- Field experience shows that when dis-
ferent morphological types present in the cussing the number of genotypes farmers
field to produce the seed to be planted in can handle, it is very dangerous to make
the following cropping season. In contrast, assumptions before discussing the issue
in the Syrian Arab Republic and in many with the farmers.
other countries in the Near East and North The choice of the breeding method also
Africa, the selection unit is a plot, and depends of the genetic structure of the final
excessive heterogeneity within a plot not product, i.e. pure lines, mixtures, hybrids
only is not exploited, but is also considered or open pollinated varieties. It is important
undesirable. to note that farmers can change the type
These three examples indicate that, even of final product originally planned by the
within the same crop, a participatory breeding breeder. For example, in Syria, where, in
programme has to use different breeding the case of self pollinated crops, the formal
methods, at least at the beginning of the system only accepts pure lines for release,
programme, to ensure full participation. It farmers do not mind adopting bulks as
is obvious that a blanket approach, based on long as they are not too heterogeneous. In
the same breeding method used everywhere the case of barley, we also have the example
208 Plant breeding and farmer participation
of one farmer testing the advantage of a plant breeding programme, farmers control
mixture of a 6-row genotype, adapted to the crossing programme by selecting the
high rainfall and lodging resistant, with a best entries, which are usually the parents
2-row genotype adapted to low rainfall and of the following cycle, as discussed earlier
lodging susceptible. Similarly in Egypt, we under choice of parental material.
found that farmers plant a mixture of all Eventually, a breeder planning to start
the lines selected one year earlier (Grando, a PPB programme is faced with the issue
pers. comm.). of whether the breeding method used in a
In principle, all breeding methods can non-participatory programme needs to be
be employed in PPB, keeping in mind that changed. While there are breeding methods
participatory does not mean that ALL that are easier to fit into a participatory
the breeding material has to be planted context, a breeder does not have necessarily
in farmers fields. Several examples of to change the breeding method, given what
different breeding methods used in actual was said earlier about fitting the method
participatory breeding programmes can be to whatever type of breeding farmers are
found in Almekinders and Hardon (2006). already doing. Here, we might add that,
Given that plant breeding is a cyclic like other aspects of PPB, the methodol-
process (see Figure 9.2), one organizational ogy can also evolve as new farmer skills
issue that is often debated is the stage of emerge. Several examples can be found in
the plant breeding programme at which Almekinders and Hardon (2006).
participation should start. This issue in
effect makes the difference between PPB 9.4.6 Management of trials in farmers
and participatory variety selection (PVS; fields
see Chapter 3, section 3.7), where the par- The organizational issues of implementing
ticipation of farmers takes place during the trials in farmers fields differ considerably
third stage of the breeding process, after the from those in a research station, and are
genetic variability available at the beginning more similar to those of a DBP. However,
of the cycle has beenusuallydrastically they diverge significantly from a DBP when
reduced. We believe that farmer participa- farmers take full responsibility for planting
tion should, at a minimum, coincide with and harvesting.
the second stage of a breeding programme, The first differences are issues such the
possibly when the genetic variability is still choice of the actual portion of land on
at or near its maximum. There are examples which to plant the trial, the total number
of PPB programmes where farmers can start of plots in each trial, the type of controls
as early as making crosses, such as the PPB (check varieties), the plot size, the seed
rice programmes in Bhutan and Viet Nam rate, the distance between rows, the dates
(SEARICE, 2003), which does not neces- of planting and harvesting: all these have
sarily imply emasculation and manual pol- to be discussed with each community in
lination, but, for example, mixing different each location. It is not simply a matter of
genotypes or cultivars of cross-pollinated courtesy. Farmers interest in the trial is
crops to facilitate intercrossing. Even when directly proportional to their participation
they do not manually make the crosses, in in its design and management. The inability
a PPB programme that runs over cycles of of the scientists to accommodate farmers
selection and recombination like any other requirements may lead to a total lack
Selection methods. Part 1: Organizational aspects of a plant breeding programme 209
of interest by the farmers. For example, diversity within the crop, and in this case
in the case of barley in the Syrian Arab our suggestion is to start with a wide array
Republic, farmers believe that seed rate is of genotypes representing as wide range
extremely important. Whether this belief of diversity as possible. But there are cases
is correct or not is immaterial, because if where farmers have previous experience
the scientists use the seed rate they believe with various type of germplasm and they
right, farmers may even refuse to carry on may feel very strongly concerning one or
selection. Therefore, in the participatory more types of specific germplasm type.
barley breeding programme in the Syrian For example, in he Syrian Arab Republic,
Arab Republic, for example, we are using as farmers grow two landraces: one with black
many as eight different seed rates, ranging seed, which is grown predominantly in
from 50 kg/ha to 250 kg/ha. As this is dry areas, and one with white seed, which
believed by the farmers to have a major is grown predominantly in wetter areas.
effect on barley yields, an important side- Farmers feel very strongly about the seed
activity would be the visits of farmers to colour and therefore in the participatory
locations where a very different seed rate is barley breeding programme in the Syrian
used; this might be the best way to generate Arab Republic we make available different
an interest in testing alternative seed rates. initial genetic material in the two areas. The
One fundamental principle in discussing issue of the type of genetic material covers
organizational issues with farmers com- also the issue of the checks. The checks
munities is to pose and justify the problem, have the dual purpose of providing an
not to present a solution. The solution estimate of error variance (for example, in
should come from the community, and if unreplicated trials with systematic checks)
the community or the individual farmers and to provide a comparison for farmers
are not prepared to solve the problem, a during selection. The ideal solution is to
possible solution can be offered, but only have a well adapted variety to fit both pur-
as a suggestion. poses, and if the choice of the check(s) is
The choice of land, which in a CBP left, as it should be, to the farmers, this is
usually depends on the farm manager, in usually their choice.
the case of a DBP (whether participatory The issue of managing the equipment
or not) has to be agreed on by the farmer. in a PPB programme is similar to a DBP.
It has to represent a suitable rotation and If the country has a network of research
a good uniformity (this should be checked stations each with its own equipment, it
the year before, together with the past is obviously more economical that each
history of the field). The size required by station uses its own equipment for all the
the trial may not match that allocated by field operations. Where machinery has to
the farmer to that specific rotation. In this be moved from one central research station
case, the extra land has to be planted by the to all the trials sites, the number of sites
scientists with a cover crop using a variety and of trials has to be adjusted to allow all
chosen by the farmer. the necessary operations to be performed
The type of genetic material to be used in time. Usually farmers are extremely
in the programme needs to be discussed concerned about planting and harvesting at
with farmers. Initially, the scientists may the right time, and if the choice is between
find that farmers are not aware of the having several locations and being late in
210 Plant breeding and farmer participation
both planting and harvesting in some of marginal and dry areas where yield per
them, it is advisable to reduce them to a unit of land is low. Therefore, no gaps
number that can be managed properly. should be left between plots, as is common
The issue of timely harvesting, in the case practice on research stations to facilitate
of completely mechanized crops, can be the identification of plots, and the alleys
solved by estimating yield through a hand- should also be planted. To facilitate farmers
harvested sample of the plots. This has during selection, and to avoid seed mixture
the additional advantage of estimating the if the seed from the trial is to be used the
total biological yield, a character of major following year, the first and last rows of
importance in many developing countries. the plot can be harvested by hand shortly
The farmers can then harvest by combine before selection and harvesting. Similarly,
whatever is left in the field. This of course the alleys can be mechanically slashed or
assumes that the seed requirements for the hand harvested to facilitate moving across
following year are satisfied. The need for the field and harvesting. The second
timely planting and harvesting makes it principle is to lay out the trial in a fashion
much easier to organize a PPB programme that it occupies a piece of land of regular
in countries or for crops where both plant- shape, because this facilitates the handling
ing and harvesting is done by hand. In this of the rest of the land by the farmer.
case, the scientists can limit themselves to The management of trials residues (bor-
the preparation of the trials, visit each site ders, fillers around trials, border rows and
to show the trial layout, leave the envelopes what is left of a plot after taking samples) is
or the bags properly numbered, and let the an important organizational issue because it
farmers do the planting themselves, as it is a potential source of dispute. As a general
happens in a PPB programme for barley principle, as in many other organizational
in Iran. issues in PPB, this needs to be discussed
The issue of managing the equipment in in advance with farmers, justifying why
a situation of fully mechanized operations the handling of experimental plots is dif-
can also be addresses by empowering ferent from the handling of a field planted
farmers to conduct trials. This often poses for large-scale production, underlining the
technical challenges, because commercial need to generate information to use later in
drills and combines are not suitable for selection, and the need for as much preci-
planting experimental plots. sion and accuracy as possible to obtain
Finally, two additional issues in managing correct estimates of the genotypic values of
trials in farmers fields concern the physical the breeding material (the scientists do not
layout of the trials, and the management of necessarily have to use these terms when
crop residues, border rows and leftovers (in discussing with farmers!). As mentioned
the case of sampling). earlier, the guiding principle is to justify
In arranging the trials on the ground, and pose the problem, and involve farmers
two principles are important: the first is in the process of finding the most mutually
that no land should be left uncultivated. In suitable solution.
many farming communities in developing
countries, leaving even a few square metres 9.4.7 Farmer selection
of land uncultivated is considered almost An organizational issue peculiar to partici-
a crime, and this is particularly true in patory breeding programmes is the selec-
Selection methods. Part 1: Organizational aspects of a plant breeding programme 211
tion done by the farmers. This is one of the to be discussed with farmers during the
most important operations (and one that planning of the programme because it has
makes the breeding programme participa- implications for the amount of time farmers
tory). It is also one of the activities that, if need to allocate to selection and on the
done properly, can generate a strong sense total number of experimental units (plots
of ownership, and enhance farmer skills as or plants) farmers can handle. It also has
far as knowledge of the genetic material is implications for the degree of involvement
concerned. of the scientists where some of the traits
As for other organizational issues, it is that are important to the farmers need to
impossible to give general recommendations, be measured.
because the baseline can be very different in The choice of the ideal time for selection is
different communities. One of the extreme highly individual: some farmers prefer to visit
situations is represented by communities the field often during the cropping season,
where there is only a vague notion that while others, particularly in unpredictable
different varieties do exist, but farmers have environments, claim that only shortly before
had only sporadic contacts with scientists, harvesting is it possible to assess the real
and these contacts have been mostly of the value of the breeding material. Farmers may
type I am here to tell you what do; you also change their preferences in relation
do it, and I will come back to check if you to both when and how to select. Farmers
did it well!. In these communities, farmers who were used to an organized selection
often ignore the sexual reproduction in day, whereby all the farmers assembled at
plants and therefore the diversity itself a meeting point and visited and scored the
within a crop is surrounded by an aura of various trials, subsequently demanded to
mystery. The other extreme is represented do the selection by themselves on a date
by communities who already have a solid convenient to them. In fact, it is obvious
experience in breeding and experimenting. that while the first way of organizing the
Most of our experience has been with selection favours exchange of ideas among
the first type of situation, which is not the participants, it also implies fixing a
necessarily the most difficult, but is certainly date in advance that later may be no longer
the one in which PPB takes more time to convenient to some participants. The second
develop and one in which PPB evolves from solution has the advantage of allowing many
consultative to collegial (as defined under more farmers to do the selection as they are
9.4.3). Therefore we will illustrate some free to choose when to do it. This obviously
general principles that we followed with requires that the scoring sheets be made
the first type of situation, and how these available ahead of time.
principles need to be modified in the case The scoring method using by farmers
of the second situation. We will consider in during selection is another organizational
particular two aspects of farmers selection, issue, and like many others, the starting
namely when to select and how to select. point can be different in different countries
The timing of selection depends and in different communities within the
strongly on the crop and its uses, on the same country. In some communities, some
environment and on the traits farmers farmers are used to score different entities
consider important. This is a typical aspect based on merit or value; in others there is no
of the overall activity, and one which needs previous experience. The example of scoring
212 Plant breeding and farmer participation
school homework is often useful. For some developing new varieties of a given crop in
farmers, it is easier to use words representing a broader context.
different categories such as undesirable,
acceptable, good, very good and 9.4.9 Note taking, data management
excellent, which later be translated into a and analysis
numerical scale. With time, and particularly The trials conducted in a participatory
with those farmers participating regularly in breeding programme need to generate the
the selection session(s), the scoring method same quantity of information and of the
may change, particularly when farmers within same quality as those in a conventional
the same community use different methods, programme, for two reasons: first, because
and farmers will eventually converge towards the information has to be used to take the
a common scoring method. final decision of which material to promote
When scoring implies writing (words, and which material to discard, and, second,
symbols or numbers) there is risk of because the information can be later used
excluding farmers unable to read or write. in the phase of variety release. We learned
The problem can be solved by flanking that in addition to the visual selection,
the farmers who need assistance with a farmers may want to have access to some
researcher, an extension staff member or quantitative data to reach a final decision.
another farmer; this requires additional This is an additional issue to discuss at the
organizational arrangements, particularly in onset of the programme because if this is
remote areas. In those cases, the ideal solution required by the farmers, the trials have
is to make the communities capable of to be organized in such a way as to allow
organizing themselves as much as possible. collecting data on the traits considered
Other methods of scoring breeding important by farmers, analysing the results
material include the identifications of the with appropriate statistical analysis, and
best entries with ribbons of different col- reporting the results in a format that makes
ours (depending on the category). the information fully accessible to farmers.
Collecting field data may go beyond the
9.4.8 Visits to farmers time, the facilities and the expertise of the
In a participatory breeding programme farmers, but this is a possibility that can not
it is very important to maintain contacts be ruled out a priori. However, as in most
with farmers beyond and besides specific similar cases, the issue needs to be discussed
scientific activities. These courtesy visits with the farmers so that it is becoming almost
are not only instrumental in building a service that the scientists provide them.
and maintain good human relationships As for the analysis of data from par-
between scientists and farmers by bridging ticipatory breeding trials, the reader could
gaps, but are an incredibly fertile reciprocal refer to Bellon and Reeves (2002), who
source of information. Often farmers like present a wide range of analytical tools, and
to converse on issues not directly related to Chapter 20 in this volume.
with the specific participatory programme,
but related to the multitude of challenges 9.4.10 Managing the transition phase
that farmers, particularly those in marginal In this section we will consider the
agricultural environments, continually face. organizational issues faced by breeders
This helps scientists to put the issue of who decide to migrate from a CBP to a
Selection methods. Part 1: Organizational aspects of a plant breeding programme 213
FIGURE 9.2
Conventional plant breeding is a cyclic process that takes place largely within
one or more research stations (left) with the breeder taking all decisions;
participatory plant breeding is the same process, but takes place mostly in farmers
fields (right) and the decisions are taken jointly by farmers and breeders
Research
Research station station
Segregating
Segregating
populations
populations
Farmers
Breeder Crosses and
Crosses
Breeder
On-station On-farm
On-farm yield trials yield trials
yield trials
Farmers
elds
Farmers
elds
decentralized and participatory breeding We will discuss two scenarios, which are
programme. We will not consider the the most common in breeding programmes
case of transforming a decentralized non- in developing countries. The first scenario
participatory breeding programmes into a concerns breeding programmes that do not
participatory programme because this only generate genetic variability, but in which
require solving the organizational issues the base germplasm is introduced from
associated with farmer participation. other institutions (generally international
In general, the problem is to transfer a breeding programmes) and sometimes
cyclic process taking place largely within include locally collected germplasm and
one or more research stations (Figure 9.2, wild relatives. The second scenario is fully-
left) to farmers fields (Figure 9.2, right), fledged breeding programmes with all the
and to change the process of decision- steps described in Chapter 3, and which
making in the way discussed earlier. The may include acquisition through germplasm
general organizational issues in managing collection and molecular breeding.
the change is that, because it is unwise to
get rid of breeding material, the transfer of First scenario: breeding programmes with
the programme to farmers fields should only the selection and testing stages
start from the first step that the breeder We will examine the organizational issues in
intends to transfer and implies that, till managing the transition of a plant breeding
the transfer is completed, the CBP and the such as the one shown in Figure 9.3, which
DPBP will coexist. This should be clearer represents a situation common to several
from the examples given later. countries. In such a breeding programme,
214 Plant breeding and farmer participation
FIGURE 9.4
Steps to modify a conventional plant breeding programme
that relies on introduced breeding material as the source of genetic
variability into a decentralized-participatory breeding programme
Seed Increase Initial yield trials Seed Increase Initial yield trials
Year 1 Year 2
International Nurseries
International Nurseries PPB Programme
Seed Increase
Initial yield trials
Seed Increase Seed Increase Initial yield trials
Preliminary yield trials
Seed Increase Seed Increase Preliminary yield trials
Advanced yield trials
Seed Increase Advanced yield trials
On-farm trials
Variety Release
Maintenance breeding
(G0 G3)
Variety Release
Maintenance breeding
(G0 G3)
Year 3 Year 4
In each box the conventional program is on the left and the participatory on the right
station; therefore, in the third year, all the breeding material can be tested for impor-
three categories of trials will have migrated tant pests and diseases, while multiplying
into the PPB programme, while only seed the seed for the initial yield trials, as the
multiplication is conducted on station. screening can continue in suitable locations
In the fourth year there is no more need (plastic houses, hot-spot sites) during all
to plant the on-farm trials because all the testing and selection stages, using part of
trials have already been conducted on farm, the seed kept for increases.
and if the data are considered sufficient, Also, the lines from the advanced yield
and there is material worth releasing, the trials that are candidates for release can be
procedure for variety release can be initi- used as parent in a crossing programme.
ated, while the promising lines are further As we are discussing the case of a breeding
multiplied. programme without crossing programme,
A number of activities can be conducted these lines could be sent to the institution(s)
on station in parallel with the participatory supplying the incoming breeding material
programme. For example, the incoming with a request that they use them in
216 Plant breeding and farmer participation
FIGURE 9.5
Schematic representation of a conventional plant breeding programme of a
self-pollinated crop based on a classical pedigree method (left): all the phases before
the On-farm Verification trials are conducted on research station.
On the right, the same programme conducted in a participatory mode
0 Crosses 0 Crosses
F1 F1
1 1
2 F2 2 F2
14-15 RELEASE
Selection methods. Part 1: Organizational aspects of a plant breeding programme 217
FIGURE 9.6
Schematic representation of a participatory plant breeding programme of a
self-pollinated crop based on a bulk-pedigree method that allows selection between
crosses in farmers fields through yield testing in three successive cropping seasons (left)
and selection within selected crosses on station (right)
0 Crosses
1 F1
2 F2
The pure lines derived from selection within populations will be yield tested through steps 3 to 6.
promoted to the F5 only if farmers select farmers could exploit the relative advan-
the corresponding F4 bulks. The process is tages of the two partners, i.e. the extraor-
repeated in the F5 and the resulting families, dinary ability of institutions to generate
after one generation of increase, return as F7 variability, and their continuity, versus the
in the yield-testing phase. Therefore, when extraordinary ability of farmers to extract
the model is fully implemented, the breeding what can improve their livelihood from
material that is yield tested includes new that variability under their conditions.
bulks as well as pure lines extracted from We have already given an example of
the best bulks of the previous cycle. the technical aspects of institutionalizing
The method has a number of advan- in the section Managing the transition
tages: (i) the participation of farmers can be phase. However, the major issue with the
introduced very early in the overall process. institutionalization of PPB is to make
The method can actually start with the F2 this method acceptable to national and
if the amount of seed available is sufficient; international research institutes as being the
(ii) during the pure-line selection, it is pos- way in which plant breeding is conducted
sible to screen for biotic stresses, quality by the institute.
traits or other traits important to farmers Unfortunately, the several cases of both
and with high heritability, using conven- successful and unsuccessful institutionali-
tional or molecular approaches; and (iii) the zation of PPB do not allow drawing a
method can also be used solely for selection general lesson or a general methodology on
between crosses in those cases where the how to obtain institutional recognition of
system of variety release is not too strict in PPB as an approach that effectively com-
terms of uniformity. bines the development of improved varie-
While the aspects of managing the tran- ties with development objectives aiming
sition phase have been discussed with spe- at alleviating rural poverty and improving
cific reference to self-pollinated crops, the local food production (Almekinders and
concepts underlying the process are equally Hardon, 2006).
applicable to cross-pollinated crops. One good example to illustrate how
difficult it is to understand what influences
9.4.11 Institutionalization of policy and managers of agricultural research
participatory plant breeding is the experience at ICARDA based on
Institutionalizing PPB (i.e. mainstreaming the work done in nine countries (Algeria,
and scaling-up) must be one of the main Egypt, Eritrea, Iran, Jordan, Morocco, the
objectives when setting up a participatory Syrian Arab Republic, Tunisia and Yemen)
breeding programme. This is because it is with a number of crops, with one in com-
very unlikely that individual, small-scale mon (barley), by the same team of scientists
PPB projects, even though very successful with a similar methodology. This work
at local level, will ever determine impact yielded contrasting results in terms of insti-
at national level in terms of production tutionalization, ranging from institutional
increase, for example, even if this is not rejection, as in the Syrian Arab Republic
the only impact expected from a PPB and Egypt, where the programme continues
programme. At the same time, only col- as a direct ICARDA-farmers collaboration,
laboration between the institutions that and Tunisia, where, at the end of a special
have responsibility for plant breeding and project, all the activities ended, to full insti-
Selection methods. Part 1: Organizational aspects of a plant breeding programme 219
TABLE 9.1
Status of nine PPB programmes conducted by the same Institution (ICARDA) in nine different
countries and date of Institutionalization
Country Crop(s) Date started Date ended Date of institutionalization
Key to crops: B= barley; L = lentil, C = chickpea, F = faba bean, DW = durum wheat, BW = bread wheat.
Date is that at which the programme was fully supported by Government institutions either financially or ideologically (see
text for details). N/A indicates not institutionalized.
seminars and university lectures, and more Canada; the Governments of Italy,
importantly the adoption by farmers of a Switzerland, Algeria and Iran; BMZ,
number of varieties (not even considered Germany; DANIDA, Denmark; and the
for release) with large production increases, System-Wide Programme for Participatory
there is still institutional opposition to even Research and Gender Analysis of the
consider PPB as a complementary approach CGIAR PRGA Programme is gratefully
to conventional breeding. acknowledged.
In contrast, and after only two years
and with no travelling workshop, no REFERENCES
training, no scientific papers, and only a Almekinders, C. & Hardon, J., eds. 2006.
few seminars and meetings, there has been Bringing farmers back into breeding.
a complete uptake of the methodology by Experiences with participatory plant breed-
at least one institution in Iran and a very ing and challenges for institutionalisation.
similar reaction was observed from the two Agromisa Special 5. Agromisa, Wageningen,
leading institutions in Algeria (INRA and Netherlands. 140 p.
ITGC) after only three years. Ashby, J.A. & Lilja, N. 2004. Participatory
The facts that one crop is common research: Does it work? Evidence from
to all cases, that in those countries with Participatory Plant Breeding. In New
contrasting institutional reactions the crop directions for a diverse planet. Proceeding
is used mostly as animal feed, and that of the 4th International Crop Science
the farmers belong to the same culture, Congress, Brisbane, Queensland, Australia,
make it even more difficult to interpret the 26 September1 October 2004.
different attitudes of policy-makers and Atlin, G.N., McRae, K.B. & Lu, X. 2000.
scientists. Genotype region interaction for two-
It appears as if ultimately the success row barley yield in Canada. Crop Science,
in institutionalizing PPB depends on indi- 40: 16.
viduals, on their attitude to innovations Bellon, M.R. & Reeves, J. 2002. Quantitative
and on the power relationships within the analysis of data from participatory methods
institution. In our experience, the institu- in plant breeding. Mexico, D.F., CIMMYT.
tionalization has been enormously facili- Biggs, S.D. & Farrington, J. 1991. Agricultural
tated by the presence of a person within an research and the rural poor: a review of
institution with sufficient moral authority social science analysis. Ottawa, International
to influence all the others. Development Research Centre (IDRC).
139 p.
ACKNOWLEDGEMENTS Ceccarelli, S. 1989. Wide adaptation. How
The author wishes to thank the staff of the wide? Euphytica, 40: 197205.
institutions, the NGOs and the farmers Ceccarelli, S. 1996. Positive interpretation of
in Algeria, Egypt, Eritrea, Iran, Jordan, genotype by environment interactions in
Morocco, the Syrian Arab Republic, relation to sustainability and biodiversity.
Tunisia and Yemen who have given their In M. Cooper & G.L. Hammers, eds. Plant
support and have generously shared their adaptation and crop improvement, pp. 467
knowledge and ideas. 486. Wallingford, UK, CAB International;
The financial support of the International Andra Pradesh, India, ICRISAT; Manila,
Development Research Centre (IDRC), IRRI.
Selection methods. Part 1: Organizational aspects of a plant breeding programme 221
Ceccarelli, S., Grando, S., Tutwiler, R., Baha, Technical and institutional issues in partici-
J., Martini, A.M., Salahieh, H., Goodchild, patory plant breeding done from the perspec-
A. & Michael, M. 2000. A methodological tive of farmer plant breeding. Working Paper
study on participatory barley breeding. I . No. 2. CGIAR System-Wide Programme
Selection phase. Euphytica, 111: 91104. on Participatory Research and Gender
Ceccarelli, S., Grando, S., Singh, M., Michael, Analysis for Technology Development and
M., Shikho, A., Al Issa, M., Al Saleh, A., Institutional Innovation.
Kaleonjy, G., Al Ghanem, S.M., Al Hasan, Pretty, J. 1994. Alternative systems of inquiry
A.L., Dalla, H., Basha, S., & Basha, T. 2003. for sustainable agriculture. IDS Bulletin,
A methodological study on participatory 25: 3748.
barley breeding. II. Response to selection. Rizza, F., Badeck, F.W., Cattivelli, L., Lidestri,
Euphytica, 133: 185200. O., Di Fonzo, N. & Stanca, A.M. 2004. Use
Ceccarelli, S. & Grando, S. 2007. Decentralized- of a water stress index to identify barley
Participatory Plant Breeding: An exam- genotypes adapted to rainfed and irrigated
ple of demand-driven research. Euphytica, conditions. Crop Science, 44: 21272137.
155: 349360. Rosielle, A.A. & Hamblin, J. 1981. Theoretical
Dodig, D., Zoric, M., Knezevic, D., King, S.R. aspects of selection for yield in stress and
& Surlan-Momirovic, G. 2008. Genotype non-stress environments. Crop Science,
environment interaction for wheat yield 21: 943946.
in different drought stress conditions and Schnell, F.W. 1982. A synoptic study of the
agronomic traits suitable for selection. methods and categories of plant breeding.
Australian Journal of Agricultural Research, Zeitschrift fr Pflanzenzchtung, 89: 1-18.
59: 536545. SEARICE. 2003. Strengthening farmers agri-
Falconer, D.S. 1952. The problem of environ- cultural diversity management systems. In
ment and selection. American Naturalist, CIPUPWARD. Conservation and sustaina-
86: 293298. ble use of agricultural biodiversity: a source-
Falconer, D.S. 1981. Introduction to quantita- book, pp. 599607. Los Baos, Philippines.
tive genetics. 2nd ed. London, Longman Simmonds, N.W. 1984. Decentralized selec-
Group. tion. Sugar Cane, 6: 8-10.
Lilja, N. & Ashby, J.A. 1999a. Types of par- Simmonds, N.W. 1991. Selection for local
ticipatory research based on locus of adaptation in a plant breeding programme.
decision making. Consultative Group Theoretical and Applied Genetics, 82: 363
on International Agricultural Research 367.
(CGIAR), Participatory Research and Soleri, D., Cleveland, D.A., Smith, S.E.,
Gender Analysis (PRGA). PRGA Working Ceccarelli, S., Grando, S., Rana, R.B., Rijal,
Document No. 6. D. & Labrada, H.R. 2002. Understanding
Lilja, N. & Ashby, J.A. 1999b. Types of gender farmers knowledge as the basis for collabo-
analysis in natural resource management ration with plant breeders: methodologi-
and plant breeding. Consultative Group cal development and examples from ongo-
on International Agricultural Research ing research in Mexico, Syria, Cuba and
(CGIAR), Participatory Research and Nepal. In D.A. Cleveland & D. Soleri, eds.
Gender Analysis (PRGA). PRGA Working Farmers, scientists and plant breeding: inte-
Document No. 8. grating knowledge and practice, pp. 19-60.
McGuire, S., Manicad, G. & Sperling, L. 1999. Wallingford, UK, CABI.
222 Plant breeding and farmer participation
Weltzien, E., Smith, M.E., Sperling, L. & Weltzien, E., Smith, M.E., Meitzner, L.S. &
Meitzner, L.S. 2000. Technical and institu- Sperling, L. 2003. Technical and institu-
tional issues in participatory plant breeding tional issues in participatory plant breeding
done from the perspective of formal plant - from the perspective of formal plant breed-
breeding. Working Paper No. 3. CGIAR ing. A global analysis of issues, results, and
Systemwide Programme on Participatory current experience. PPB Monograph No. 1.
Research and Gender Analysis for PRGA Programme, Cali, Colombia.
Technology Development and Institutional
Innovation.
223
CHAPTER 10
Selection methods.
Part 2: Pedigree method
Flavio Capettini
224 Plant breeding and farmer participation
TABLE 10.1
10.1 INTRODUCTION Example of a programme following the
Since the beginnings of agriculture more Pedigree Method
than 10 000 years ago, crop producers Growing Cycle Procedure
have been striving to obtain better crops. 1 Plant 250 5000 F2 plants.
Different procedures have been followed Select individual plants (300)
to reach their objectives, with various 2 Grow F2:3 rows (300)
results. Modern plant breeders, after the Select best rows (50)
re-discovery of Mendels laws just over Select best plants in rows (23)
a century ago, tried to introduce new 3 Grow F3:4 lines (150)
time. Although different formal methods Select best plants in rows (2)
4 Grow F4:5 lines (80)
have been described, there is the general
Select best families (15)
impression that modifications are more
Select best rows in families (23)
the common rule in breeding programmes
Select best plants in rows (1)
rather than the strict methods described in 5 Grow F5:6 lines (20)
books. Ultimately, every applied breeder Select best families (15)
has their own approach and priorities to Harvest best rows in bulk
reach their objectives, using a package of 6 Testing of F5:7 lines in two
breeding tools, available knowledge and locations, one rep per loc.
resources. This package should be matched 7 Multi-location
environment, which include the mechanics Harvest heads from yield plots in one
replication to seed purification.
of institute where the work is carried out, 8 Multi-location testing.
the crop, the agro-ecological target area, Grow 50-100 F8:9 head-rows of lines
and the socio-economic parameters of the still in test to produce pure seed.
FIGURE 10.1
Diagrammatic representation of the pedigree method
Season Procedure
1 Plant F2 seed
Select individual plants
2 Grow individual rows w o x z v
Select best rows w o x z v
Select best F3 plants w Discard x z v
w o x z v
o
z z
3 Grow individual rows w w w x x x z z v v v
Select best families w w w x x x Discard v v v
Select best rows w w w x x x z z v v v
Select best F4 plants w w w x x x z z v v v
x x
4 Grow individual rows w w x x v v v
Select best families w w Discard v v v
Select best rows w w x x v v v
Select best F5 plants w w x x v v v
v v
5 Grow individual rows w w
Select best families w w v v
Harvest best rows in bulk w w v v
w w v v
w v
6 Extensive testing of w v
F5-derived lines w v
in 1891, where single plants were selected when homozygote lines are developed. This
from an existing cultivar or a landrace is because other methods are less labour
population. Newman (1912) described how (therefore resource) intensive and easier
selected plants were laid out in the field as to apply in earlier generations. Generally,
plant-row, ear-row or head-row plots. The selection starts with an F2 population
Vilmorin Company in France developed and goes on until homogeneous lines are
the system independently, where it was developed. In the next cycle, the best F2:3
termed the Vilmorin method of selection lines are selected, followed by the selection
(Fehr, 1987a). of the desirable F3 plants within each line.
In all subsequent generations selection is
10.2.2 Implementation performed on the most desirable families
The method can be initiated or stopped at any first, followed by the most desirable lines
generation during the inbreeding process. within that family, to finally select the best
When combined with mass or bulk selection, plants within each line, which are harvested
it is generally applied in generations closer to individually (Figure 10.1).
226 Plant breeding and farmer participation
TABLE 10.2
Expected additive genetic variability among and individuals from dominant homozygous.
within lines during inbreeding process without The same can occur with the heterosis
any selection expressed by heterozygous individuals,
Generation Additive Genetic Variability slowing the homozygosity process (Fehr,
of lines
Among Lines Within Lines 1987a).
F2:3 1 1/2
F3:4 1.5 1/4 10.3 PROS AND CONS OF THE
F4:5 1.75 1/8
PEDIGREE METHOD
F5:6 1.875 1/16
10.3.1 Advantages of the pedigree
method
Although breeding is a numbers r The art of breeding can be extensively
game the greater the numbers managed, exercised. The breeder can effectively
the higher the probability of finding the decide in the field the shape of the
right or a better combination of genes breeding programme and see the effect
the size of the programme will always be in every generation. Inferior genotypes
limited by the resources available. This can be discarded early in the process,
would include the level of the breeders and allowing the use of higher volumes for
support personnel expertise the institution early generations in the programme, and
is able to hire, land, hardware, technology retaining only the good ones for the later,
and time. At the beginning of each season, expensive, replicated experiments stage.
the breeder should decide how many r Different locations and environments can
progeny rows they will be able to grow for be used in each growing cycle, allowing
all selection generations. selection for different traits not expressed
The pedigree method also relies heavily everywhere. Populations can be replicated
on the expected genetic variability in each in hot spots, i.e. environments where the
generation, and that will determine the desired selection traits are expressed at a
number of selections made (Table 10.2) maximum, to increase the probability of
(Fehr, 1987a). The genetic variability selecting for specific traits.
expected within each line is at a maximum r The breeder can manage the amount of
at the F2:3 generation, decreasing by half in generic variance they want to keep within
each following generation. and between families, as well as the num-
ber of families.
10.2.3 Genetic background r Different qualitative and quantitative traits
The most important factor to be considered can be selected at the same time in dif-
is the genetic variability present among and ferent generations, including traits being
within lines during the inbreeding process expressed at plant as well as grain level.
(Table 10.1). Additive epistasis is generally
much smaller than the additive portion 10.3.2 Disadvantages of the pedigree
of total genetic variability. Variability method
associated with dominance and dominant r Cannot be utilized in environments where
epistasis cannot be utilized in inbred lines. genetic variability for the characters of
Dominance can complicate the selection of interest is not expressed. If one cannot
homogeneous and homozygous lines by not use off-season nurseries, there will be an
allowing differentiation of heterozygous associated increase in the length of time
Selection methods. Part 2: Pedigree Method 227
for cultivar development compared with of the pedigree method was that only one
other methods of breeding. generation could be grown each year. By
r Considerable record keeping. separating inbreeding from selection, the
r An experienced person must do the first process could be accelerated until
selection (at least we flatter ourselves in homozygosity was reached. Working with
thinking so). wheat, he suggested that the number of
r Requires more land and labour than progeny to be grown from a plant be one
other methods of inbreeding. or two, growing two generations in the
Besides the points above, there are ques- greenhouse in each autumn and winter,
tions concerning applied breeding pro- and one generation in the field during
grammes, especially those in the private the summer. With this method, the F6
sector. Questions about why the extensive generation can be reached in two years,
data recording is needed, if the plants compared with the five years needed with
that are kept in the programme are there the Pedigree Method. Once homozygosity
because they showed superiority anyway. is achieved, a large number of lines can be
Pedigree becomes difficult to use for quan- tested for the desired traits.
titative traits, especially those with lower The harvest of one seed from each plant
heritability. Despite the named disadvan- during inbreeding was first described by
tages, that are more due to the different Johnson and Bernard in 1962 for soybeans,
circumstances and priorities, this is the and the first time this method was termed
method that resulted in significant increases single-seed descent was by Brim (1966),
in the genetic gain in breeding programmes who considered it to be a modified pedigree
early in the twentieth century. The criteria method.
would depend on how narrow the cross The usual procedure is to harvest a sin-
is, the heritability of the traits, costs of the gle seed from each plant in a population,
evaluation, worth of the trait and resources bulk the harvested seeds and plant the bulk
available. Modifications were carried out in the next generation. The procedure can
once the system was understood, and it be started in the F2 and continued until the
helped to acquire deeper knowledge of desired level of homozygosity is achieved.
the high number of genes acting in plant The number of plants that will be needed
development. in the last generation of inbreeding should
be decided and the number of initial plants
10.4 SINGLE-SEED DESCENT METHOD should be calculated backwards to the F2
Single-seed descent is a method to generation, taking into account the expect-
rapidly advance generations of inbreeding ed losses due to lack of germination.
populations before starting the evaluation of In his book, Fehr (1987b) also describes
individual lines, which is frequently used in modifications of the single-seed descent
conjunction with the pedigree method. The method, such as a multiple-seed procedure
concept was proposed by Goulden in 1941. and a single-hill procedure. The objectives
He noted that a breeding programme can are the same: to obtain rapid generation
be divided into two stages: the development advance. The methods are well suited for
of pure lines from segregating populations; use in greenhouses and winter nurseries,
and selection of desired pure lines from where genotypes perform differently from
among those produced. The disadvantage their area of adaptation.
228 Plant breeding and farmer participation
10.5.1 Advantages of the single-seed r The amount of seed available at the end
descent procedures of the inbreeding process is reduced,
r They are easy to manage and speed up needing additional growing cycles just
the inbreeding process as no special labo- to multiply seed, thus delaying the total
ratories or techniques are needed in com- process.
parison with other methods, e.g. double
haploid production. REFERENCES
r Procedures are well suited for environ- Brim, C.A. 1966. A modified pedigree meth-
ments where otherwise only one genera- od of selection in soybeans. Crop Science,
tion per year can be grown in the field. 6: 220.
r Can be less expensive than field selection Fehr, W.R. 1987a. Principles of Cultivar
where the costs of land and land manage- Development, Theory and Technique. USA,
ment are high. Macmillan Publishing Co. 332 p.
Fehr, W.R. 1987b. Pedigree Method. In W.R.
10.5.2 Disadvantages Fehr, ed. Principles of Cultivar Development.
r The size of the population should be Vol. 1 Theory and Technique. USA, Mac
adjusted for germination losses. Graw-Hill Inc.
r All the F2 plants may not be present in Newman, L.H. 1912. Plant Breeding in
the line evaluation due to germination Scandinavia. Canadian Seed Growers
losses, decreasing the genetic variance. Association, Ottawa.
229
CHAPTER 11
Selection methods
Part 3: Hybrid breeding
Donald N. Duvick*
* deceased
230 Plant breeding and farmer participation
Many countries plant a high percentage of Hybrid cotton makes up about 15 percent
[sorghum] hybrids. In the USA, Mexico, of the Chinese acreage and 45 percent of
South America, Australia and South Africa the acreage in India. The hybrid cot-
the percentage of hybrids is close to 100 ton in China appears to be increasing. A
percent. In Indias rainy season it is close to small amount of F1 intraspecific and inter-
8090 percent hybrid. In the post-rainy [sea- specific hybrids are grown in the USA.
son] they plant an OPV almost 100 per- The hybrids are produced by hand, prob-
cent of which is used for local consumption ably in India and marketed by an Israeli
Only in many sorghum growing areas of company.
Africa have hybrids not taken off . (W. Meredith, pers. comm., 2005)
(K. Porter, pers. comm., 2005).
In Asian countries, excluding China, tion. The cost of labour for hand emascula-
approximately 40 percent of [maize] is tion and pollination is prohibitively high
planted with hybrids. In Asian countries in developed nations. Although genetic
including China approximately 76 per- systems such as cytoplasmic-nuclear male
cent of [maize] is planted with hybrids. sterility are present in cotton (Meredith,
(T. Kunta, pers. comm., 2005) 1999), to date none of them have been used
to make hybrid cotton on a large scale.
Soybeans (Glycine max (L.) Merr.) are
Kumar, 2004). In recent years a special not grown as hybrids, although nuclear
kind of genetic male sterility has been used and cytoplasmic-nuclear sterility systems
as well, to make two-line hybrid rice have been reported as operational in the
(Guohui and Longping, 2003; Yuan, 1998). crop (Ding et al., 2002; Smith, Horner and
Seed is produced on female inbred lines Palmer, 2001). Effective cross-pollination
that are homozygous for environmentally systems must be developed (probably using
sensitive recessive male sterility genes. insects such as honeybees (Apis mellifera L.)
Seed production fields are planted in an or leaf cutter bees (Megachile rotundata F.)
environment (e.g. long day and/or high and modified soybean flower morphology),
temperature) that enables expression of to enable adequate levels of seed yield
the male sterility gene and thus enables (Palmer et al., 2001).
successful seed production. Seed increase In theory one could produce apomictic
fields of the female lines are grown in an hybrids of important field crops such as
environment (e.g. short day and/or cooler rice or maize; seed multiplication would be
temperatures) that represses expression of much less expensive and hybrids could be
the male sterility genes, allowing the female sold at lower prices. Additionally, farmers
lines to reproduce via self-pollination. could reproduce the apomictic hybrids
Cotton hybrids are widely grown in via saved seed as they now can do with
India, and also in China and Viet Nam), but pure-line self-pollinated crops. Research
very little elsewhere (Anonymous, 2005; is underway to develop apomictic hybrids
James, 2002; Meredith, 1999; Roberts, 2005). but the genetics are complicated.
The Indian hybrids are produced by Development of apomictic hybrids will
means of hand emasculation and pollina- require extensive research and experimen-
Selection methods. Part 3: Hybrid breeding 235
(Duvick, 1997). One should note that than either of the parents (more tolerance
augmented stress tolerance as a result of to biotic stress), and the goal of broad
breeding for dependability is not unique to protection is reached in one generation
hybrid breeding programmes. Breeding for rather than the long-term back-crossing
dependable performance improves abiotic and selection needed to place all of the
stress tolerance of non-hybrid as well as of needed genes in a single, inbred cultivar.
hybrid cultivars (Rajaram, Singh and Ginkel,
1997; Smale et al., 2002), although heterosis 11.3.5 Customers can dictate traits of
per se plus ability to combine tolerance hybrid cultivars
traits from two parents into one hybrid does Farmers who plant hybrid seed can dic-
give a unique advantage to hybrids. tate the traits of the hybrids they plant, if
A caveat: Breeding for tolerance of abi- they buy the seed from commercial firms.
otic stress at one location does not neces- Because the existence of commercial seed
sarily give adequate tolerance to that stress companies depends on seed sales, they
in a markedly different location (e.g. the must develop hybrids (or produce seed
different locations may have a much greater of hybrids developed by the public sec-
intensity of the targeted abiotic stress, such tor) that the customer wants to buy. For
as drought at flowering time). example, the hybrids must be adapted to
the environment and farming practices of
11.3.3 Uniform expression of desired each adaptation region where the seed is to
traits be sold, and additionally the hybrids must
Uniform expression of certain traits is have specific traits and quality as desired by
highly desirable for some crops. Such uni- farmers, their potential customers.
formity is automatic in the case of cultivars Farmers may favour (or insist upon)
of self-pollinated crops such as tomato or maize hybrids with resistance to prevalent
wheat, but not so for cross-pollinated crops leaf diseases in their region, sorghum hybrids
such as maize or beets. Hybrids of cross- with tolerance to iron deficiency in alka-
pollinated crops are uniform for all traits line soils common to their region, or rice
if the hybrids are single-cross hybrids, hybrids with flavour and texture that suits
made by crossing two inbred lines. Thus, their taste (or the taste of the ultimate con-
hybrids in cross-pollinated crops present sumers of their product). In the United
new opportunities for producing a uni- States of America, during the past 70 years,
form product, if uniformity is desired or maize breeders continually have had to breed
required, as in many horticultural crops hybrids with ever increasing tolerance to the
(Wehner, 1999). stresses imposed by higher plant densities,
11.4.2 Potential uniform susceptibility famine and great hardships for millions in
to new pest genotypes the past). Genetic diversity can help, but it
As stated earlier, most hybrid genotypes is not a cure-all.
are highly uniform. This means that they Within any one season, farmers who
may be uniformly resistant to important plant uniform hybrids can spread their risk
diseases or insect pests, but they also may by growing several hybrids of differing
be uniformly susceptible to new and unex- parentage, more or less as when one grows
pected pests, or to an unexpected and a genetically heterogeneous non-hybrid
highly unfavourable growing season. cultivar. In addition, breeders constantly
This condition is not unique to hybrid develop new hybrids of genotypes different
cultivars; it is present and presents the same from the predecessor hybrids, and thereby
potential problems in self-pollinated crops provide farmers with genetic diversity in
such as wheat or soybeans (Simmonds, time, also called temporal genetic diversity
1993; van der Plank, 1963). And even (Duvick, 1984; Smale et al., 2002). Also,
genetically heterogeneous cropsor wild as noted above, hybrids in self-pollinated
plant speciescan be devastated by a new crops can contain a broader range of resist-
genotype of insect or disease if they happen ance genes (greater internal genetic diver-
to lack the needed resistance or tolerance sity) than is easily bred into individual
genes, despite their highly heterogeneous pure-line cultivars.
nature. Thus, Dutch elm disease (Ophiostoma Ultimately, however, one must recognize
ulmi Buisman) devastated the genetically that to grow hybrids of cross-pollinated
diverse populations of native American crops is to reduce the genetic diversity
elm (Ulmus americana L.) in the United in the field, in comparison with growing
States of America in the1960s (French et highly heterogeneous OPCs of those same
al., 1980). Chestnut blight (Cryphonectria crops.
parasitica (Murrill) Barr) overwhelmed
the entire wild population of American 11.4.3 Difficult to serve unique small
chestnut trees (Castanea dentate (Marsh.)) adaptation areas
in the mid-twentieth century (Anagnostakis, Seed companies usually cannot afford to
2005). Ergot (Claviceps purpurea (Fr.) breed and produce seed for very small, spe-
Tul.) infestations of rye (a genetically cialized markets. Farmers may therefore be
heterogeneous cross-pollinated crop) caused unable to find commercial hybrids that suit
numerous epidemics of poisoning in Europe their needs if they are in a small region with
during the Middle Ages and in more recent
centuries as well (Matossian, 1989). And
many centuries before those events, the early The challenge in developing countries is
Romans annually sacrificed a red dog to the that [some farmers] have no or very lim-
god Robigus in hopes that he, sufficiently ited choice because seed of improved [sor-
satisfied with the red dog, would not ravage ghum] varieties or hybrids is not available.
their fields of genetically heterogeneous Clearly, there are local preferences but it
wheat cultivars with epidemics of red rust may be impossible for large companies to
(Puccinia spp.) (see Large, 1982 for this and adequately address all of them.
other narratives of destructive plant disease (K. Porter, pers. comm., 2005)
epidemics that resulted in widespread
Selection methods. Part 3: Hybrid breeding 239
their current favourite cultivars and report impressions of the value of hybrid seed
their findings to the breeder (R. Heisey, and the veracity of seed companies (e.g.
pers. comm., 2005). Ilagan, 2004) [and is particularly frequent in
developing countries where Governmental
11.6 SEED PRODUCTION Seed Companies have the monopoly of
11.6.1 Technical aspects seed production and distribution. Editors].
Seed production techniques, such as detas- Efficient and accurate performance of
selling, cytoplasmic male sterility, hand these multiple tasks requires skilled workers,
emasculation and pollination, and various specialized equipment, and, above all, a well-
kinds of self-incompatibility, have been coordinated and well-directed organization,
discussed earlier. i.e. a hybrid-seed production company.
All of the methods that use pollination
by wind or insects on a field scale share a 11.6.2 Seed producers
common need: to establish sufficient isola- One should note that although public-sec-
tion from other sources of pollen to ensure tor breeders (e.g. academic institutions,
that out-crossing is held to acceptably low government agencies or international
levels. For some crops this can mean that research centres) do hybrid breeding (i.e.
seed production fields must be long dis- they may produce improved populations,
tances from other fields of that crop. inbred lines and well-tested final hybrids),
Seed producers also must ensure that the large-scale production and distribution
seed has satisfactory germination, is of hybrid seed to farmers is nearly always
produced in sufficient quantity to satisfy the work of commercial seed companies.
farmer needs, and, importantly, is delivered Public-sector breeders (via the institu-
in timely fashion, suited to the planting tions that employ them) will release their
schedules of the farmers who use the seed. plant breeding products for use by the
Truthful labelling is essential as well; this private sector, using a variety of forms
may seem to be an unnecessary statement, of release. In some cases, no restrictions
but in some parts of the world seed has are placed upon the released germplasm;
been sold as first-generation hybrid when in other cases, some sort of intellectual
it was not. This seriously damages farmers property protection is used to allow col-
lection of royalties or other forms of pay-
ment (e.g. Butler and Marion, 1985; ERS,
Testing [in sunflower] is necessary in all 2004; Heisey, Rubenstein and King, 2005;
the zones of the world, mainly because University System of Maryland, 2004).
each zone has some particular abiotic or Seed companies come in various sizes,
biotic stress that needs to be assessed. Thus from small local companies to large interna-
it is done in North America, throughout
the sunflower growing regions of Europe,
South America, India, Australia and other Because [sunflower] is a bee-pollinated
parts of Asia. Selection time from inbred crop, sizable isolation distances are required
genesis to culminating hybrid products can between varieties. Typically 11.5 miles
take a total of up to 10 years. [1.82.7 km] is utilized.
(G. Cole, pers. comm., 2005) (G. Cole, pers. comm., 2005)
244 Plant breeding and farmer participation
their extensive knowledge of pedigrees, use the small-plot information to guide them
performance of inbred lines in various in evaluating the adaptation of hybrids to
hybrid combinations, and, increasingly, the particular environments, cultural practices
information in genomics databases. and farmer preferences. Farmer participation
is passive in the sense that farmers make no
Hybrid formation, trials, evaluation decisions about which hybrids are entered
Professional breeders can best organize and in the trials or which ones are saved, and
carry out these tasks, although farmers can therefore is one type of Participatory Variety
(and should) participate in the evaluation Selection (PVS) (Chapter 3).
stages. Annually, breeders make many new Another step in collaborative evaluation
experimental hybrids, with full knowledge in the United States of America came about
that only a few of them will perform well after farmers adopted combines for harvest,
enough to save and release as acceptable starting in about the 1970s. Experimental
new hybrids. Typically, one must make hybrids in the final stage of trials are now
dozens or hundreds of experimental hybrids widely tested by farmers in strip trials.
for initial observation and yield trials. As Farmers plant field-length strips (e.g.
noted earlier, a few seasons of performance several rows) of experimental hybrids
and evaluation trials in appropriate (often furnished gratis by seed companies)
environments will enable the breeders to next to strips of their favourite commercial
reduce the number of experimental hybrids hybrid(s), and at harvest time a measured
to a much smaller total. This reduced length of each strip is combine harvested and
number of hybrids (advanced experimental the grain is weighed, tested for moisture, and
hybrids) can then be tested much more yield is calculated. Other notes on pertinent
widely, in particular in those environments traits may be taken at the same time.
and locations where they are to be grown Farmers use these strip-test data (typi-
and used by farmers. cally those from their own farm and also
Farmers and breeders can, and do, those from neighbouring farms) to help
collaborate in conducting many of decide which commercial hybrids to plant
these performance trials, especially in in their fields in the following season. Their
the more advanced stages of selection. participation is active in that they choose
Such collaboration is common today in which hybrids to compare in strip trials
industrialized countries, as well as in some and which hybrids to plant in the following
regions of developing countries. For example, season, but is still a type of PVS [Editors
from the first days of hybrid breeding, seed note].
companies in the United States of America Seed companies use the strip-test data
have grown the majority of their advanced to help them decide which experimental
small-plot hybrid yield trials on land of hybrids to save and promote to commercial
collaborating farmers. The professional status (i.e. production and sale) for the fol-
breeders provide seed, plant the trials and lowing season.
harvest them; the farmers provide land and Thousands of such trials are performed
cultural practices. Sometimes the farmers each season. Properly analysed, the data
have received some kind of payment, for can be statistically significant, and can help
example if the trials on average yielded less the breeders to characterize the specific
than the farmers commercial crop. Breeders adaptation(s) of each hybrid. They are a
Selection methods. Part 3: Hybrid breeding 249
Cooperatives can produce their own seed Farmers cooperatives in Viet Nam have
if they are willing to invest in the seed played a role in the hybrid rice seed industry
production elements that result in qual- for some time now. The seeds harvested are
ity seed. These are: (1) a mechanism to turned over to the provincial seed compa-
ensure getting pure seed of inbred parents; nies that in turn sell the seeds to the farm-
(2) training of qualified seed plant manag- ers. There are now nine community-based
ers; (3) identification of reliable farmer seed farmers cooperatives for hybrid rice seed
growers; (4) hiring people to manage the production in the Philippines. They mainly
various aspects of seed production (planting produce seed of the public-sector-developed
dates to ensure isolation, removing rogues, hybrids. The national research institute
proper harvest techniques, etc.); (5) acquir- provides technical backstopping whenever
ing facilities for drying and conditioning needed by the cooperatives. The Philippine
the seed for maximum quality; (6) evalu- Rice Research Institute, which is responsible
ation of final quality (germination, vigour for the development and dissemination of
test, purity, etc.); (7) proper packaging; and the technology, facilitated the establishment
(8) storage facilities. of these cooperatives. It was responsible for
(K. Porter, pers. comm., 2005) the development and dissemination of the
technology. In the beginning, seeds pro-
duced by the cooperatives were procured by
Growers (P.A.G.), consisting of approxi- the government and sold to the farmers at
mately two dozen independent seed pro- subsidized prices. The subsidy is now gradu-
duction enterprises like the Champaign ally being phased out and the marketing of
County Seed Company. P.A.G. eventually the hybrid seeds is transferred entirely to
reorganized as an independent, relative- the cooperatives. The sustainability of this
ly large-scale conventional seed company arrangement remains to be seen since seed
with its own breeding programme. In due marketing has been only recently trans-
course, an even larger company purchased ferred to the cooperatives
P.A.G. (Fernandez-Cornejo, 2004). (R.S. Toledo, pers. comm., 2005)
Despite the potential difficulties dis-
cussed in previous paragraphs, farmer
cooperatives have been formed in devel- by the seed companies directly. The seed
oping countries to produce hybrid seed. production is said to be economical and of
The Philippine cooperatives appear to high quality. The farmers are compensated
be transitioning into small seed companies on the basis of the amount of seed they
that depend on public-sector breeders for produce for the contracting company. This
inbreds and hybrids. enterprise is a huge source of income
A different way for farmers to participate and contribution to the local economy
in (and profit from) hybrid seed production (G. Srinivasan and D. Beck, pers. comm.,
is to produce hybrid seed on their farms, on 2005).
contract to commercial seed firms. In India,
hundreds of farmers in Andhra Pradesh Seed sales and distribution
produce hybrid maize seed in this way, As noted above, cooperatives might not only
guided and trained by a local contractor, or produce hybrid seed, they might distribute
252 Plant breeding and farmer participation
Bnziger, M., Edmeades, G.O. & Lafitte, H.R. Duvick, D.N. 1966. Influence of morphology
1999. Selection for drought tolerance increas- and sterility on breeding methodology. In
es maize yields across a range of nitrogen K.J. Frey, ed. Plant breeding, pp. 85138.
levels. Crop Science, 39: 10351040. Ames, USA, Iowa State University Press.
Bnziger, M., Setimela, P.S., Hodson, D. & Duvick, D.N. 1984. Genetic diversity in major
Vivek, B. 2004. Breeding for improved farm crops on the farm and in reserve.
abiotic stress tolerance in maize adapted Economic Botany, 38: 161178.
to southern Africa. In T. Fischer et al., Duvick, D.N. 1997. What is yield? In G.O.
eds. New directions for a diverse planet. Edmeades et al., eds. Developing drought-
Proceedings of the 4th International Crop and low N-tolerant maize, pp. 332335.
Science Congress, Brisbane, Australia, 26 Proceedings of a Symposium, CIMMYT, El
September1 October 2004. The Regional Batan, Mexico, 2529 March 1996., Mxico,
Institute Ltd., Gosford, Australia. D.F., CIMMYT.
Bi, X., de Palma, J., Oane, R., Khush, G.S. & Duvick, D.N. 2002. Theory, empiricism and
Benett, J. 2003. Molecular approaches for intuition in professional plant breeding. In
fixing the heterozygosity of hybrid rice. In D.A. Cleveland & D. Soleri, eds. Farmers,
S.S. Virmani et al., eds. Hybrid rice for food scientists and plant breeding: Integrating
security, poverty alleviation, and environ- knowledge and practice. pp. 189211.
mental protection, pp. 135150. Proceedings Wallingford, UK, CABI Publishing.
of the 4th International Symposium on Duvick, D.N. 2004. Keynote 2 The current
Hybrid Rice, Hanoi, Viet Nam, 1417 May state of plant breeding: How did we get
2002. Los Baos, Philippines, IRRI. here? In M. Sligh & L. Lauffer, eds. Summit
Brown, G.G. 1999. Unique aspects of cyto- Proceedings: Summit on Seeds and Breeds
plasmic male sterility and fertility restora- for 21st Century Agriculture, pp. 7191.
tion in Brassica napus. Journal of Heredity, Washington, DC, Pittsboro, USA, Rural
90: 351356. Advancement Foundation International-
Butler, L.J. & Marion, B.W. 1985. The impacts USA.
of patent protection on the U.S. seed indus- Duvick, D.N. 2005. The contribution of breed-
try and public plant breeding. NC Regional ing to yield advances in maize (Zea mays
Research Publication No. 304. University of L.). Advances in Agronomy, 86: 83145.
Wisconsin-Madison, Madison, USA. 128 p. Duvick, D.N., Smith, J.S.C. & Cooper, M.
Canola Advantage. 2005. Growing canola 2004. Long-term selection in a commer-
(available at http://www.canola-council. cial hybrid maize breeding program. In J.
org/hybridssyntheticcanola.aspx). Janick, ed. Plant Breeding Reviews, Part
J.G. Coors & Pandey, S., eds. 1999. Genetics 2, Long-term selection: crops, animals, and
and exploitation of heterosis in crops. bacteria, pp. 109151. Vol. 24.
Madison, USA, ASA/CSSA/SSSA. Edmeades, G.O., Bnziger, M. & Cortes-C.,
Crabb, R. 1993. The hybrid corn-makers. 2nd M. 1997. From stress-tolerant populations
ed. West Chicago, USA, West Chicago to hybrids: The role of source germplasm.
Publishing Company. 414 p. In G.O. Edmeades et al., eds. Developing
Ding, D., Gai, J., Cui, Z. & Qiu, J. 2002. drought- and low N-tolerant maize,
Development of a cytoplasmic-nuclear pp. 263272. Proceedings of a Symposium,
male-sterile line of soybean. Euphytica, CIMMYT, El Batan, Mexico, 2529 March
124: 8591. 1996. Mxico D.F., CIMMYT.
Selection methods. Part 3: Hybrid breeding 255
Ejeta, G. 1988. Development and spread Ha, D.T., Thao, T.D., Khiem, N.T., Trieu,
of Hageen Dura-1, the first commercial M.X., Gerpacio, R.V. & Pingali, P.L. 2004.
sorghum hybrid in the Sudan. Applied Maize in Vietnam: Production systems, con-
Agricultural Research, 3: 2935. straints, and research priorities. Mexico,
Ekasingh, B., Gypmantasiri, P., Thong- D.F., CIMMYT. 42 p.
ngam, K. & Grudloyma, P. 2004. Maize in Hanna, W.W. 1995. Use of apomixis in culti-
Thailand: Production systems, constraints, var development. Advances in Agronomy,
and research priorities. Mexico, D.F., 54: 333350.
CIMMYT. 36 p. Hassan, R.M., Mekuria, M. & Mwangi, W.
ERS [Economic Research Service, USDA]. 2001. Maize breeding research in eastern
2004. Agricultural biotechnology intellec- and southern Africa: Current status and
tual property (available at http://www.ers. impacts of past investments made by the
usda.gov/Data/AgBiotechIP/). public and private sectors 196697. Mexico,
Fernandez-Cornejo, J. 2004. The seed industry D.F., CIMMYT. 33 p.
in U.S. agriculture: An exploration of data Haussmann, B.I.G., Obilana, A.B., Ayiecho,
and information on crop seed markets, P.O., Blum, A., Schipprack, W. & Geiger,
regulation, industry structure, and research H.H. 2000. Yield and yield stability of
and development. Bulletin 786. USDA- four population types of grain sorghum in
ERS, Washington, DC. a semi-arid area of Kenya. Crop Science,
French, D.W., Ascerno, M.E. & Stienstra, W.C. 40: 319329.
1980. The Dutch Elm disease. AG-BU-0518. Heisey, P.W., Rubenstein, K.D. & King, J.L.
Minnesota Extension Service, University of 2005. Patterns of public and private sec-
Minnesota, St. Paul, USA. tor patenting in agricultural biotechnology.
Geiger, H.H. & Miedaner, T. 1999. Hybrid rye AgBioForum 8(2-3): 7382 (available at www.
and heterosis. In J.G. Coors & S. Pandey, agbioforum.org; accessed 31 Jan. 2009).
eds. Genetics and exploitation of heterosis Ilagan, L. 2004. Hybrid rice seed cheats
in crops, pp. 439450. Madison, USA, ASA/ revealed. The Manila Times, 17 August
CSSA/SSSA. 2004. Manila.
Grando, S., von Bothmer, R. & Ceccarelli, S. Iowa State Department of Agriculture. 1934.
2001. Genetic diversity of barley: use of Part VI. Iowa Corn and Small Grain
locally adapted germplasm to enhance yield Growers Association Thirty-fifth annual
and yield stability of barley in dry areas. In Iowa year book of agriculture, pp. 135148.
H.D. Cooper, C. Spillane & T. Hodgkin, Iowa State Department of Agriculture, Des
eds. Broadening the Genetic Base of Crop Moines, USA.
Production, pp. 351-372. New York, USA, James, C. 2002. Global review of commercial-
CABI/Rome, FAO/Rome, IPGRI. ized transgenic crops. 2001 Feature: Bt cot-
Guohui, M. & Longping, Y. 2003. Hybrid rice ton. Ithaca, USA, ISAAA.
achievements and development in China. In Janick, J. 1998. Hybrids in horticultural crops.
S.S. Virmani et al., eds. Hybrid rice for food In K.R. Lamkey & J.E. Staub, eds. Concepts
security, poverty alleviation, and environ- and breeding of heterosis in crop plants:
mental protection, pp. 247256. Proceedings Special publication number 25, pp. 4556.
of the 4th International Symposium on Madison, USA, CSSA.
Hybrid Rice, Hanoi, Viet Nam, 1417 May Jenkins, M.T. 1978. Maize breeding during
2002. Los Baos, Philippines, IRRI. the development and early years of hybrid
256 Plant breeding and farmer participation
maize. In D.B. Walden, ed. Maize breeding Matossian, M.K. 1989. Poisons of the past: Molds,
and genetics, pp. 1328. New York, USA, epidemics, and history. New Haven, USA, and
John Wiley and Sons. London, Yale University Press. 190 p.
Jones, D.F. 1918. The effects of inbreeding Melchinger, A.E. & Gumber, R.K.. 1998.
and cross-breeding upon development. Overview of heterosis and heterotic groups
Connecticut Agricultural Experiment in agronomic crops. In K.R. Lamkey & J.E.
Station, New Haven, USA. Staub, eds. Concepts and breeding of hetero-
Jones, D.F. 1922. The productiveness of sis in crop plants: Special publication number
single and double first generation corn 25, pp. 2944. Madison, USA, CSSA.
hybrids. Journal of the American Society of Menkir, A. & Akintude, A.O. 2001. Evaluation
Agronomy, 14: 241252. of the performance of maize hybrids,
Jordaan, J.P., Engelbrecht, S.A., Malan, J.H. & improved open-pollinated and farmers local
Knobel, H.A. 1999. Wheat and heterosis. In varieties under well-watered and drought
J.G. Coors & S. Pandey, eds. Genetics and stress conditions. Maydica, 46: 227238.
exploitation of heterosis in crops, pp. 411 Meredith, W.R. Jr. 1999. Cotton and het-
421. Madison, USA, ASA/CSSA/SSSA. erosis. In J.G. Coors & S. Pandey, eds.
Joshi, P.K., Singh, N.P., Singh, N.N., Genetics and exploitation of heterosis in
Gerpacio, R.V. & Pingali, P.L. 2005. Maize crops, pp. 451462. Madison, USA, ASA/
in India: Production systems, constraints, CSSA/SSSA.
and research priorities., Mexico, D.F., Miller, J.F. 1987. Sunflower. In W.R. Fehr,
CIMMYT. 42 p. ed. Principles of cultivar development.
Knudson, M.K., & Ruttan, V.W. 1988. The pp. 626668. New York, USA, Macmillan
R&D of a biological innovation: The case Publishing Company.
of hybrid wheat. Food Research Institute Mojayad, F. & Planchon, C. 1994. Stomatal
Studies, 21: 4567. and photosynthetic adjustment to water
Koemel, J.E. Jr, Guenzi, A.C., Carver, B.F., deficit as the expression of heterosis in sun-
Payton, M.E., Morgan, G.H. & Smith, E.L. flower. Crop Science, 34: 103107.
2004. Hybrid and pureline hard winter Morris, M.L. & Bellon, M.R. 2004.
wheat yield and stability. Crop Science, Participatory plant breeding research:
44: 107113. Opportunities and challenges for the
Kunz, K., ed. 2002. Vegetable breeding for international crop improvement system.
market development. Bangkok, East West Euphytica, 136: 2135.
Seed International Ltd. 150 p. NRC [National Research Council]. 1972.
Large, E.C. 1962. The advance of the fungi. Genetic vulnerability of major crops.
New York, USA, Dover Publications. 488 p. Report of the National Research Council
Lpez-Pereira, M.A. & Morris, M.L. 1990. Committee on Genetic Vulnerability of
Impacts of international maize breeding Major Crops. Washington, DC, National
research in the developing world, 1966 Academy of Sciences. 307 p.
1990. Mxico D.F., CIMMYT. 58 p. Nicolas, D. 2005. Bls hybrides: les ventes
Maredia, M. & Howard, J. 1998. Facilitating progressent doucement. Agro Performances,
seed sector transformation in Africa: Key April: 3233.
findings from the literature. USAID - Owen, F.V. 1945. Cytoplasmically inherited
Bureau for Africa, Office of Sustainable male-sterility in sugar beets. Journal of
Development, Washington, DC. Agricultural Research, 71: 423440.
Selection methods. Part 3: Hybrid breeding 257
Palmer, R.G., Gai, J., Sun, H. & Burton, Rick, C.M. & Chetelat, R.T. 1995. Utilization
J.W. 2001. Production and evaluation of of related wild species for tomato improve-
hybrid soybean. Plant Breeding Reviews, ment. Acta Horticulturae, 412: 2138.
21: 263307. Roberts, A. 2005. Hazera no longer novelty,
Paudyal, K.R., Ransom, J.K., Rajbhandari, gaining more attention (available at http://
N.P., Adhikari, K., Gerpacio, R.V. westernfarmpress.com/mag/farming_haz-
& Pingali, P.L. 2001. Maize in Nepal: era_no_longer/index.html).
Production systems, constraints, and pri- Rusike, J., Howard, J. & Maredia, M. 1997. Seed
orities for research. NARC and CIMMYT, sector evolution in Zambia and Zimbabwe;
Kathmandu. 48 p. Has farmer access improved following eco-
Perotti, E., Grimanelli, D., John, P., Hoisington, nomic reform? Policy Synthesis for USAID
D. & Leblanc, O. 2004. Why is transferring - Bureau for Africa, Office of Sustainable
apomixis to crops still a dream? In T. Fischer Development 31: 1-4.
et al., eds. New directions for a diverse plan- Saaten-Union. 2005. Le site du bl hybride
et. Proceedings of the 4th International Crop (available at http://www.blehybride.com/).
Science Congress, Brisbane, Australia, 26 Seiler, G.J. 1992. Utilization of wild sunflower
September1 October 2004. The Regional species for the improvement of cultivated
Institute Ltd., Gosford, Australia. sunflower. Field Crops Research, 30: 195
Pixley, K. & Bnziger, M. 2004. Open-pollinated 230.
maize varieties: A backward step or valu- Simmonds, N.W. 1993. Introgression and
able option for farmers? In D.K. Friesen & incorporation Strategies for the use of
A.F.E. Palmer, eds. Integrated approaches crop genetic resources. Biological Reviews
to higher maize productivity in the New of the Cambridge Philosophical Society,
Millennium, pp. 2228. Proceedings of the 68(4): 539562.
7th Eastern and Southern Africa Regional Simon, P.W. 2000. Domestication, historical
Maize Conference, Nairobi, Kenya, 5-11 development, and modern breeding of car-
February 2002. CIMMYT and KARI rot. Plant Breeding Reviews, 19: 158190.
(Kenya Agricultural Research Institute)., Smale, M., Reynolds, M.P., Warburton, M.,
Mexico, D.F., CIMMYT. Skovmand, B., Trethowan, R., Singh, R.P.,
Pons, L. 2003. Heart-friendly corn oil? New Ortiz-Monasterio, I. & Crossa, J. 2002.
high-oleic corn varieties make it possible. Dimensions of diversity in modern spring
Agricultural Research Magazine (USDA- bread wheat in developing countries from
ARS), 51: 1819. 1965. Crop Science, 42: 17661779.
Rajaram, S., Singh, R.P. & Ginkel, M.V. 1997. Smith, M.B., Horner, H.T. & Palmer, R.G.
Breeding wheat for wide adaptation, rust 2001. Temperature and photoperiod effects
resistance and drought tolerance. In Crop on sterility in a cytoplasmic male-sterile
Improvement for the 21st Century, pp. 139 soybean. Crop Science, 41: 702704.
163. Trivandrum, India, Research Signpost. Sperling, L., Ashby, J.A., Smith, M.E.,
Renard, M., Delourme, R., Valle, P. & Pierre, Weltzien, E. & McGuire, S. 2001. A frame-
J. 1998. Hybrid rapeseed breeding and pro- work for analyzing participatory plant
duction. In G. Thomas & A. Monteiro, eds. breeding approaches and results. Euphytica,
Proceedings of the International Symposium 122: 439450.
on Brassicas. Acta Horticulturae, 459: 291 Stephens, J.C. & Holland, R.F. 1954.
298. Cytoplasmic male-sterility for hybrid sor-
258 Plant breeding and farmer participation
ghum seed production. Agronomy Journal, Virmani, S.S., ed. 1994. Hybrid rice technolo-
46: 2023. gy: New developments and future prospects.
Swastika, D.K.S., Kasim, F., Sudana, W., Selected papers from the International Rice
Hendayana, R., Suhariyanto, K., Gerpacio, Research Conference. Los Baos, Laguna,
R.V. & Pingali, P.L. 2004. Maize in Indonesia: Philippines, IRRI. 296 p.
Production systems, constraints, and research Virmani, S.S. & Kumar, I. 2004. Development
priorities. Mexico, D.F., CIMMYT. 40 p. and use of hybrid rice technology to
Tanksley, S.D., Grandillo, S., Fulton, T.M., increase rice productivity in the trop-
Zamir, D., Eshed, Y., Petiard, V., Lopez, J. ics. International Rice Research Notes
& Beck-Bunn, T. 1996. Advanced back- 29.1/2004. Manila, IRRI.
cross QTL analysis in a cross between an Virmani, S.S., Mao, C.X. & Hardy, R., eds.
elite processing line of tomato and its wild 2003. Hybrid rice for food security, poverty
relative L. pimpinellifolium. Theoretical and alleviation, and environmental protection.
Applied Genetics, 92: 213224. Los Baos, Philippines, IRRI. 407 p.
Tatum, L.A. 1971. The southern corn leaf Wehner, T.C. 1999. Heterosis in vegetable
blight epidemic. Science, 171: 11131116. crops. In J.G. Coors & S. Pandey, eds.
Tollenaar, M. & Wu, J. 1999. Yield improve- Genetics and exploitation of heterosis in
ment in temperate maize is attributable crops, pp. 387397. Madison, USA, ASA/
to greater stress tolerance. Crop Science, CSSA/SSSA.
39: 15971604. Whitt, S.R., Wilson, L.M., TenaiIlon, M.I.,
Tracy, W.F. & Chandler, M.A. 2004. The his- Gaut, B.S. & Buckler IV, E.S. 2002. Genetic
torical and biological bases of the concept diversity and selection in the maize starch
of heterotic patterns in Corn Belt Dent pathway. Proceedings of the National
maize. Annual Corn & Sorghum Research Academy of Sciences of the USA, 99: 12959
Conference Proceedings, 810 December 12962.
2004. Audio CD. American Seed Trade Widick, J. 2005. Company history: PAG
Association. Alexandria, USA. hybrids, GSC hybrids (available at
University System of Maryland. 2004. Bylaw http://www.sunprairieseeds.com/his-
IV 3.20. University System of Maryland tory2.htm).
Policy on Intellectual Property (Approved Yuan, L.P. 1998. Hybrid rice development
by the Board of Regents, February 8, and use: innovative approach and chal-
2002; amended by the Chancellor, July 7, lenges (available at http://www.fao.org//
2004) (available at http://www.usmd.edu/ DOCREP/003/W8595T/w8595t02.htm).
Leadership/BoardOfRegents/Bylaws/ Yuan, L.P., Yang, Z.Y. & Yang, J.B. 1994.
SectionIV/IV320.html). Hybrid rice research in China. In S.S.
USDA-NASS. 2000. Crops Production Data, Virmani et al., eds. Hybrid rice for food
by States, 1866-1997. USDA-National security, poverty alleviation, and environ-
Agricultural Statistical Service, Washington, mental protection, pp. 143147. Proceedings
DC. of the 4th International Symposium on
Van der Plank, J.E. 1963. Plant diseases: Hybrid Rice, Hanoi, Viet Nam, 1417 May
Epidemics and control. New York, USA, 2002. Los Baos, Philippines, IRRI.
Academic Press. 349 p.
259
CHAPTER 12
Selection methods
Part 4: Developing open-pollinated
varieties using recurrent
selection methods
Fred Rattunde, Kirsten vom Brocke, Eva Weltzien and Bettina I.G. Haussmann
260 Plant breeding and farmer participation
12.1 INTRODUCTION
FIGURE 12.1
This chapter provides an overview of the Schematic presentation of an open recurrent
use of recurrent population improvement selection scheme with cyclic selection and
methods in variety development. The recombination combined with introgression
major stages of population improvement of new germplasm
Becker, 1993.
12.2 RECURRENT SELECTION: WHAT IS
IT AND HOW DOES IT CONTRIBUTE TO
VARIETY DEVELOPMENT?
FIGURE 12.2
Recurrent selection schemes involve cycles Expected performance distribution of
of testing, selection and recombination of progenies from the original and an
breeding units, with the possibility of improved population
deriving new varieties from each popula- Original Improved
tion cycle bulk or from the progenies population population
C0 C1
developed during each cycle (Figure 12.1).
Number of individuals
TABLE 12.1
Alternative recurrent selection methods, required number of generations (or years), and degree of
2
exploitation of the variance of additive ( A2 ) and dominance ( D ) effects
A2 D2
Notes: (1) Not equal to A2 unless p = q = 0.5 and dominance decreases to zero with inbreeding.
(2) Coefficient difficult to define unless p = q = 0.5.
done by applying mass selection for simply the degree to which they can exploit the
inherited traits. available additive genetic variance (A). The
The genetic diversity that is retained other two factors (selection intensity and
in the improved populations enables heritability) can be managed by breeders
continued improvement. Several cycles of to optimize genetic gains. Options for
directional selection should increase the managing these factors will be discussed
frequency of desirable alleles, resulting in in more detail in the sections on mass- and
higher probabilities of obtaining superior progeny-based selection methods.
progenies than in the original population It is crucial to consider the time to
(Figure 12.2). Furthermore, the genetic complete a cycle of selection since the
variability retained in a broad-based amount of progress is determined both
population enables selection for different by the gain per cycle and the time per
traits or adaptation to new target conditions cycle. Table 12.1 indicates the minimum
as the needs emerge. Examples of this are number of generations required for various
listed below. methods. The time required for one cycle
A broad-based population retaining can be reduced significantly if off-season
substantial diversity for maturity or other nursery facilities are available.
adaptive traits could be given to breeders
(researchers or farmers) in differing agro- 12.4 BREEDING OPEN-POLLINATED
ecological zones, to develop zone-specific VARIETIES USING RECURRENT
populations. SELECTION METHODS
A population retaining variability for Varieties of cross-pollinated species can
grain quality traits could be used to be developed by selection in a population
derive distinct populations for other bulk per se. This approach is appropri-
quality requirements. ate where intra-varietal heterogeneity is
Rare dwarf segregants in a tall-plant- desirable or necessary. Varieties can be
height population can be selected to derived from the population bulk per se by
develop a new short-plant-height popu- mass selection for specific highly heritable
lation with potential for increased grain traits that give the variety a more distinct
harvest index. character; for example, a narrower range
of flowering dates or more uniform plant
12.3 EXPECTED GENETIC GAIN OR THE height, or grain or plant colour. The highly
RESPONSE TO SELECTION successful pearl millet variety ICMV 155
The response to selection (R), using any was created by this method, with 59 S0
type of selection, is a function of the inten- plants mass selected during the random
sity of selection (i), the extent to which mating of the New Elite Composite Cycle
observed differences are determined by 4 bulk used to create a new variety (Singh
genetic causes (h = square root of herit- et al., 1994).
ability) and the extent of additive genetic Varieties of cross-pollinated species
variation (A), as indicated by the formula can also be developed from a set of
(Falconer, 1981): superior progenies identified during the
R = i* h*A selection phase of progeny-based recurrent
The different recurrent selection methods selection. For each variety, a separate set of
listed in Table 12.1 vary fundamentally for progenies would be identified based on a
Selection methods. Part 4: Developing open-pollinated varieties using recurrent selection methods 263
diversity of the population through selection the trait can be evaluated before flowering
of parents with outstanding performance and undesirable plants culled, full parental
for certain traits but less desirable for control can be imposed and the full extent
others will maximize the potential for of additive genetic variance can be exploited.
long-term genetic gains, but reduce the For traits that can only be observed after
possibility of deriving agronomically flowering, only the female parent can be
superior end products in the short term. In controlled, and thus only 50 percent of A
contrast, greater emphasis on population can be exploited (Table 12.1), unless plants
eliteness through more restrictive are self-pollinated and the selfed progenies
inclusion of parents for population creation are used for recombination.
will maximize opportunities for immediate
extraction of distinct finished varieties, but 12.6.1 For which selection objectives
limit long-term potential gains and benefits and conditions can mass selection be
from intra-varietal diversity. useful?
Mass selection is a very simple method of
12.6 MASS SELECTION selection, as selection is based on individual
Mass selection involves the selection of plants. This method thus requires
individual plants or even of individual minimal materials and organization for
grains or seeds (Allard, 1999). This type of implementation. Mass selection enables
selection is based on the phenotype only, maintaining a very large effective population
as a given genotype is neither replicated size even with high selection intensity.
nor tested in differing environments. Mass Several thousand plants can be evaluated
selection therefore always has confounding and several hundred retained to create the
of environmental conditions that can mask next cycle of the population. An additional
genotypic differences. As breeders can advantage of mass selection is that each
only marginally influence the extent to season results in the recombination among
which observed differences are determined differing gene blocks in the population.
by genetic causes (h = square root of This frequent recombination is essential
heritability), mass selection is only effective for breaking undesirable linkages and
for traits with higher heritability and little increasing the frequency of desirable trait
genotype by environment interaction. combinations. This is very important
One factor that can be better managed during the initial phases of a recurrent
to increase response to selection (R) is the selection programme, when new parental
intensity of selection (i) used in mass selec- materials are being recombined to form
tion. As the test units are single plants it is new populations, or when a new variety is
relatively easy to increase selection inten- formed from partially inbred progenies.
sity by increasing the area sown with the Mass selection will be most effective for
population bulk, to have a greater number traits that are highly heritable, with genetic
of plants from which to select the minimum differences that are observable on individual
number of desirable plants to constitute the plants. One study in pearl millet showed
next cycle. quite acceptable heritabilities for single
The extent of additive genetic variation plant expression of plant height (0.58), seed
that can be exploited by mass selection weight (0.52) and flowering date (0.45),
depends on the level of parental control. If but not for grain yield (0.29), based on
Selection methods. Part 4: Developing open-pollinated varieties using recurrent selection methods 265
BOX 12.1
Use of zone-specific sorghum populations as source material for variety
development
Couscous 8 6 3 3 3 2
Commercial grain 31 24 8 11 7 -
Grain storability 27 19 6 7 3 -
Fodder 50 30 12 15 14 3
Early maturity 55 36 6 12 7 2
Striga resistance 31 22 10 17 12 -
Stems
(construction) 39 26 10 10 4 1
Total 361 247 76 107 76 10
BOX 12.2
Origin of a flooding-tolerant sorghum population
mass selection offers a systematic approach through parental control that reduces the
by dividing the field into grids, and selecting extent to which selected plants are pollinat-
a common number of plants from within ed by unselected plants. Self pollination and
each grid. selection of selfed plants achieves maximum
The effectiveness of selection between parental control. The same result is achieved
plants can be maximized by ensuring that with populations of self-pollinating species
selection is conducted by the most skilled containing genetic male sterility, through
people. For example, the threshability (ease identification and selection of male-fer-
of separating grains and glumes) of sorghums tile plants. Note however that selection
in West Africa can be best observed by of selfed plants would require a separate
farmers who have years of experience and recombination to constitute the next cycle
a cultural heritage of selection for this trait. bulk. If introgression from neighbouring
Effectiveness of selection may be further fields is not desired, sufficient isolation
raised by identifying individuals who are distance would need to be maintained.
the most interested and locally respected Culling undesired plants prior to flowering
for their capabilities as seed experts. also provides parental control and could
The genetic gains from mass selection therefore double gains for traits that can
in out-crossing species can be increased be observed before flowering. Culling out
268 Plant breeding and farmer participation
tall plants in a dwarf population is one such a population may be grown by several
example. farmers or researchers, with selections from
differing sites being pooled by breeders to
Genetic Variance (A) capture selections that represent a wider
The choice of parents for creating the sample of conditions or selection criteria.
initial population determines the level and Selections produced by different breeders
usefulness of genetic diversity. The more on differing sites could be simply bulked,
diverse the parents chosen, the higher or they could be grown out in isolation
will be the expected genetic variance and for recombination, with eventual culling of
therefore the potential gain from selection. certain off-type progenies.
There is usually an optimal level of diversity
beyond which the mean performance of Selection Intensity (i)
the population would go down, thereby More intense selection (setting higher
reducing the usefulness of the population in thresholds for retaining plants) is expected
the long-term (Schnell and Utz, 1975). to increase the genetic gains. The selection
Maintaining sufficient population size intensity coefficient in the genetic gains
through selection and recombination of a formula corresponds to the number of
large enough number of plants will help standard deviations by which the mean of
maintain a desirable array of alleles, and the selected fraction exceeds the population
assure genetic variation exists for selec- mean (Falconer, 1981; Becker, 1993), and
tion in subsequent cycles (Witcombe and thus depends on the percentage of selected
Virk, 2001). Likewise, maintaining a suf- individuals.
ficiently large effective population size is An advantage of mass selection is the
indispensable to avoid inbreeding, which possibility of achieving very high selection
is of greatest concern in highly outcrossing intensity, by which rare plants possessing
species. It is exhibited as a loss of vigour the desired combination of several traits or
of the population and undirected separa- express rare forms of a given trait can be
tion into distinctive lines due to random identified. However, to realize this poten-
fixation of genes. Effects of inbreeding and tial it is necessary that a sufficiently large
strategies for avoidance are summarized by number of plants be available for selection.
Allard (1999) and Hallauer and Miranda For example, a population with 10 000
(1981), among others. Using a minimum plants could be subjected to selection with
of 200 plants to create the next cycle bulk a 2 percent selected fraction and still retain
will minimize loss of genetic variation and 200 plants for reconstituting the next cycle
genetic drift that would otherwise arise bulk. Observation of farmer selection in
from mating among a limited number of sorghum populations in Burkina Faso
parents and sampling. shows that they frequently retain selected
Mass selection, based on single plant fractions ranging from 0.2 to 5 percent.
selection in a given site, can produce more
site-specific responses than would be 12.7 PROGENY-BASED RECURRENT
obtained with multi-environment progeny SELECTION METHODS
testing. However, where the objective is A range of methods for population
to produce an improved population and improvement rely on testing, selecting and
eventual varieties with wider adaptation, recombining families rather than individual
Selection methods. Part 4: Developing open-pollinated varieties using recurrent selection methods 269
at least 200 progenies that will allow an higher and direct benefits to participating
appropriate intensity of selection (15 to 30 farmers less than in the case of mass selec-
percent) and still retain a sufficient number tion.
of progenies for recombination or a subse-
quent stage of selection. The initial creation 12.8 EVOLUTION OF POPULATION
of progenies (S1, full-sib (FS) or half-sib IMPROVEMENT PROGRAMMES
(HS) for example) can thus be done by Although population improvement
selection from thousands of plants. Mass programmes can follow a single selection
selection by several farmers, each sowing methodology for improvement of a given
the same bulk in their own fields, has been trait or set of traits over many cycles
useful in achieving suitably large numbers (Rattunde and Witcombe, 1993), this may not
of selected progenies. often be the case. Population improvement
Progeny evaluations conducted in suf- may begin by conducting several cycles of
ficient test environments is also important mass selection to narrow and clean up
to effectively assess genetic potential and the population to a more acceptable range
to sample the environmental diversity. For for critical adaptation or quality traits.
example, the wide range of sowing dates, Populations may reach appropriate ranges
soil and rainfall conditions for sorghum for simply inherited traits after a few cycles
production in even a single agro-ecologi- of selection and little further progress will
cal zone of West Africa requires a mini- be made by selecting for these same traits.
mum of four to six test environments to Improvement of more complexly-inherited
provide some measure of representation. traits, such as yield, would require changing
Conducting progeny trials on farmers to progeny-based selection methods.
fields, although logistically challenging, can Practical population improvement
help achieve the necessary, and appropriate, programmes can also undergo major changes
sampling of test environments. in the breeding objectives in response to
Progeny-based trials conducted with evolving needs and opportunities. For
farmers presents several challenges not example, population improvement by
encountered with mass selection. The large ICRISAT-Mali was initially conducted on
number of progenies and more complicated a sorghum population of tall plant height as
trial designs requires researcher assistance, this plant height corresponds to what most
at least during planting and harvesting, farmers grow in the target Sudanian zone of
or even researcher management of on- West Africa, and tall parental materials had
farm trials. Trial designs can be modified the required suite of adaptive and quality
to make on-farm progeny testing feasi- characteristics. However, the convergence
ble. Individual farmers could, for example, of farmers priority setting that placed
grow a single replication or even a subset highest value on increased yields, the
(incomplete block) of test entries. Modern hypothesis that reducing heights could raise
statistical procedures and computing power harvest index and thus grain yields, and the
now make analysis of the widest range of identification of novel dwarf segregants in
incomplete and unbalanced designs pos- the ongoing population improvement work,
sible. Issues of how benefits and costs are led to a major shift to dwarf population
shared also need to be considered, since and variety development, as described in
land and labour requirements may be much Box 12.3. Further, this sorghum population
Selection methods. Part 4: Developing open-pollinated varieties using recurrent selection methods 271
BOX 12.3
The evolving Guinea-race sorghum populations in Mali
BOX 12.4
Lata (Bala Berthe): A dwarf Guinea-race sorghum variety and hybrid parent
developed through population breeding
S3 progenies (n = 89) derived from the most promising S2 progenies (see Box 12.3) were further
tested in replicated, multi-environment, on-station trials for variety development. Six of the
most promising progenies were included in a 16-entry, 2-replicate, early generation variety
trial conducted by 20 farmers in 10 villages in 2005. Each entry was given a name to facilitate
discussions by farmers. The progeny Lata showed higher yield, intermediate height (2.5 m),
and was appreciated by farmers. This progeny was given to Bala Berthe, a farmer with strong
interest and expertise in selection. Bala Berthe conducted two cycles of mass selection for
panicle architecture, grain and glume characteristics and shared a portion of his seed lot with
ICRISAT. The variety Lata (Bala Berthe), following further testing in larger scale 4-entry
Variety Test Kits, was submitted for variety registration in Mali. This variety was also used
as a male parent to produce a series of experimental hybrids. Analysis of multi-environment
hybrid yield trials showed that Lata (Bala Berthe) had the highest combining ability of all male
parents in 2007.
be exploited, increasing the mean produc- Rai, K.N., Anand Kumar, K., Andrews, D.J.,
tivity as well as the genetic variation of the Rao, A.S., Raj, A.G.B. & Witcombe, J.R.
resulting inter-pool populations. 1990. Registration of ICTP 8203 Pearl
Millet. Crop Science, 30: 959.
REFERENCES Rattunde, H.F., Witcombe, J.R. & Singh, P.
Allard, R.W. 1999. Principles of Plant Breeding. 1989. Feasibility of mass selection in pearl
New York, USA, John Wiley & Sons. 254 p. millet. Crop Science, 29: 14231427.
Andrews, D.J., Gupta, S.C. & Singh, P. 1985. Rattunde, H.F.W. & Witcombe, J.R. 1993.
Registration of WC-C75 Pearl millet. Crop Recurrent selection for increased grain yield
Science, 25: 199200. and resistance to downy mildew in pearl mil-
Becker, H. 1993. Pflanzenzchtung. Stuttgart, let. Plant Breeding, 110: 63-72.
Germany, Verlag Eugen Ulmer. Rattunde, H.F.W., Weltzien-R., E., Bramel-
Falconer, D.S. 1981. Introduction to quantita- Cox, P.J., Kofoid, K., Hash, C.T., Schipprack,
tive genetics. New York, USA, Longman. W., Stenhouse, J.W. & Presterl, T. 1997.
340 p. Population improvement of pearl millet
Fehr, W.R. (editor). 1987. Principles of Cultivar and sorghum: current research, impact
Development. Vol. 1 Theory and Technique. and issues for implementation. In Genetic
USA, MacGraw-Hill Inc. Improvement of Sorghum and Pearl Millet,
Gallais, A. 1981. Amlioration des populations pp. 188212. Proceedings of an International
en vue de la cration de varits. Le slec- Conference held at Lubbock, USA, 2327
tionneur franais. pp. 523. September 1996.
Hallaurer, A.R. & Miranda-Fo, J.B. 1981. Schnell, F.W. & Utz, H.F. 1975. F1 Leistung
Quantitative Genetics in Maize Breeding. Ames, und Elternwahl in der Zchtung von
USA, Iowa State University Press. 468 p. Selbstbefruchtern. In Bericht ber die
Selection methods. Part 4: Developing open-pollinated varieties using recurrent selection methods 273
Arbeitstagung der Vereinigung ster- ulations base gntique large avec les
reichischer Pflanzenzchter, pp. 243248. agriculteurs au Burkina Faso. Agriculture,
Gumpenstein, 2527 November 1973. BAL, 17(2): 145153.
Gumpenstein, Austria. Weltzien, E. 2005. Research on priority set-
Schipprack, W. 1993. Estimation of popula- ting with farmers. In A. Christinck, E.
tion parameters and optimization of alter- Weltzien & V. Hoffmann, eds. Setting
native procedures of recurrent selection breeding objectives and developing seed
in Pearl Millet (Pennisetum glaucum (L.) systems with farmers. A handbook for prac-
R.Br.) PhD Dissertation. University of tical use in participatory plant breeding
Hohenheim, Germany. projects. Weikersheim, Germany, Margraf
Singh, P., Raj, A.G.B., Rao, M.N.V.V. & Publishers.
Witcombe, J.R. 1994. Registration of ICMV Weltzien, E., vom Brocke, K. & Rattunde, F.
155 Pearl Millet. Crop Science, 34: 1409. 2005. Planning plant breeding activities with
Strahwald, J.F. & Geiger, H.H. 1988. Theoretical farmers. In A. Christinck, E. Weltzien & V.
studies on the usefulness of doubled haploids Hoffman, eds. Setting breeding objectives
for improving the efficiency of RS in spring and developing seed systems with farmers.
barley. In Proceedings of the 7th Meeting A handbook for practical use in participa-
of Eucarpia Section on Biometrics in Plant tory plant breeding projects. Weikersheim,
Breeding. s, Norway, 25 August 1988. Germany, Margraf Publishers.
vom Brocke, K., Trouches, G., Zongo, S., Witcombe, J.R. & Virk, D.S. 2001 Number of
Bitie, A., Oualbogo, A., Barro-Kondombo, crosses and population size for participa-
C., Weltzien, E. & Chantereau, J. 2008. tory and classical plant breeding. Euphytica,
Cration et amlioration in-situ de pop- 122: 451462.
275
CHAPTER 13
Selection methods
Part 5: Breeding clonally propagated
crops
selection step are again normal asexual result was the domestication of pitiquia
reproduction (Simmonds, 1979). Hence, (Solanum stenotomum), which was most
the finally selected clone is genetically iden- probably selected from S. leptophyes or
tical with the original seed plant from S. canasense. From the view-point of the
which the selected clone is derived. In other knowledge of the twenty-first century it is
words, each seed plant is a potential variety. not surprising that suddenly potato plants
Roots and tubers, fruit and tree plant spe- with larger leaves and larger tubers were
cies have been used by human since long found. Potato spontaneously changes its
before the dawn of agriculture. They have polyploidy level by unreduced gametes and
been domesticated by IPB (Simmonds, recombination. Polyploid potatoes are more
1979) and several made a substantial yield vigorous than their diploid ancestors. The
progress by FPB in some regions of the result was the domesticated of polyploid
world. However, in other regions of the andigena (S. tuberosum subsp. andigena).
world there is not much yield progress, Andigena is the ancestor of the commercial
and in these regions there appears to be a potato in long-day temperate climatesthe
clear need of PPB for progress. We wish to so-called Irish potato (S. tuberosum subsp.
illustrate this by two examples: potato and tuberosum) (Hawkes, 1979, 1981). This IPB
sweet potato. of potato and introductions of FV of potato
An example of the needs and requirements into the Northern Hemisphere changed
of clonally propagated varieties can be the world both socio-economically and
found in potato (Solanum spp.). There are politically (Hobhouse, 1985).
about 200 wild potato species (Huamn Today, eight species of potato are still
and Ross, 1985). They usually contain cultivated in the Andes, variously diploid,
glycoalkaloids, which give tubers a bitter triploid, tetraploid and pentaploid:
taste and which are toxic when consumed (i) cultivated diploid potatoes are
in large quantities (Zitnak and Filadelfi, pitiquia (S. stenotomum), its close
1985). It is nearly certain that 100 to 130 relatives phureja (S. phureja) and
centuries ago indigenous knowledge in limea (S. goniocalyx), and ajanhuiri
the Andes and along the Pacific coast of (S. ajanhuiri), which evolved from
South America was those sites where it was interspecific recombination of diploid
possible to collect wild potato tubers where pitiquia and the diploid wild potato
species and mutants were growing that had species S. megistacrolobum;
low alkaloid content. Although these tubers (ii) cultivated triploid potatoes are
were very small, the man, or more probably chaucha (Solanum chaucha), a
a woman, made life much easier by growing hybrid between diploid pitiquia
and maintaining desirable types by cloning and tetraploid andigena, and rucki
close to their homes. This happened more (Solanum juzepczukii), a hybrid
than 8 000 years ago, and most likely between diploid pitiquia and the
independently at several places (Ugent, tetraploid wild potato species
Pozorski and Pozorski, 1982; Ugent, S. acaule;
Dillehay and Ramirez, 1987). Those types (iii) cultivated tetraploid potatoes are
were preferred that were easier to maintain, andigena and Irish potato; and finally
easier to harvest (shorter stolons) and had (iv) the cultivated pentaploid potato
larger tubers compared to other types. The is a hybrid species (Solanum
278 Plant breeding and farmer participation
the potato crop can be lost because of high DM, low-sweet or bland type, which
Colorado beetle (Leptinotaras decemlineata) has a dry mouth feeling. The first type, also
(CIP, 1977, 1980; Radcliffe, 1982; Rich, called the dessert type, has extremely high
1983). pro-vitamin A concentrations (Huang,
A simpler example for the needs and Tanudjaja and Lum, 1999) and a 50 g piece
requirements of clonally propagated varie- of fresh storage roots can meet the daily
ties can be found in sweet potato (Ipomoea requirements of a pre-schooler (Low et
batatas). Sweet potato was domesticated in al., 2007). Moreover, sweet potatoes with
the Americas more or less during the same high pro-vitamin A concentrations have
prehistoric period as the potato (OBrien, high protein and mineral concentrations
1972). The evolution of sweet potato was (Grneberg, unpublished). In the United
not as complex and diverse as that of States of America, the dessert type is
potato. There are about 500 Ipomoea spe- generally the desired sweet potato to
cies, but only the I. batatas species was meet market and consumer needs. In the
domesticated (Austin and Huamn, 1996). Caribbean, low DM orange-fleshed sweet
Again polyploidy was important. Sweet potatoes (OFSP) are consumed, but as
potato is hexaploid and its closest relative a staple, so a dryer mouth feel and less
is I. trifida (di- and tetraploid). It is certain sweet flavour is preferred. These white-
that sweet potato contains the I. trifida or yellow-fleshed varieties are known as
genome, but obviously it is not simply a bonitos or ricos (Baynes, 1972). Along the
multiple copy. Two-thirds of the sweet Pacific coast of South America we observed
potato genome corresponds to the I. trifida that sweet potatoes are mainly pale orange
genome and one-third to an ancestor very fleshed and less sweet. However, locally,
closely related to I. trifida (Shiotani and white- and purple-fleshed sweet potatoes
Kawase, 1989). Within diploid I. trifida are consumed, which clearly have different
accessions (seed families) it is also possible taste, texture and flavour compared to
to find plants that form small storage roots OFSP. In Brazil, the sweet potato storage
(Daniel Reynoso, pers. comm.). However, roots must clearly have a high DM
sweet potato has been found in the ruins concentration (28 to 30 percent DM), and
of the so-far oldest city in the Americas, usually this is a white-fleshed sweet potato;
Caral on the Pacific coast of central Peru however, locally, high DM OFSP can be
(Solis, 2004), and the crop reached Pacific found (Amauri Buso, pers. comm.).
Polynesia and parts of South-East Asia The taste preferences in sub-Saharan
(naturally or by early seafarers) before Africa are similar to those in Brazil, per-
Columbus. It was primarily the Portuguese haps because the Portuguese introduced the
that introduced it into Europe, Africa, sweet potato into Africa. All FVs are nearly
South Asia and East and South-East Asia exclusively white- or yellow-fleshed and
(Yen, 1976). have high DM concentrations; however, a
Although the taste of sweet potato in FVs few pale- to medium-orange-fleshed FVs
and MVs differ tremendously, two major can be found with high DM concentrations
types can be distinguished: (i) the orange- (Tumwegamire, unpublished). These local
fleshed, moist, low dry matter (DM) and OFSP FVs are very promising for allevia-
sweet type, which has a soft mouth feeling; tion of vitamin A deficiency in sub-Saharan
and (ii) the white- or pale-yellow-fleshed, Africa (e.g. FVs such as Ejumula, Carrot
280 Plant breeding and farmer participation
C, Carrot Dar es Salaam or Zambezi). farmers have not adopted MVs (e.g. cv.
In eastern Africa, storage root DM con- INA100, which is a high yielding OFSP
tents must be >30 percent (Mwanga et al., and fits consumer needs very well), because
2003). In southern Africa, storage DM of insufficient SPVD tolerance. Farmers
concentration between 26 and 29 percent became disappointed with new MVs and
are accepted (Laurie Sunette, pers. comm.), after a few growing seasons returned to FVs
whereas in West Africa, sweet potato must such as cv. Jonathan and cv. Huambachero.
be non-sweet, very high in DM concentra- OFSPs from the Americas with elevated DM
tions (between 30 and 35 percent DM) and (e.g. cv. Jonathan) are partially successful in
with a texture and flavour tentatively simi- southern Africa, and in south-west and
lar to yam (Dioscorea spp.) (IITA, 1981). central Asia, provided that weevil pressure is
In contrast, in India, where sweet potato not extreme. Weevil damage is associated with
consumption has been very low in the past, drought-prone regions (Central and South
today people prefer sweet potatoes with America, sub-Saharan Africa and south-
high DM, high sugar content, dark orange west and central Asia); however, weevil
flesh and a storage root shape that is cylin- species differ: Cylas formicarius in all parts
drical but tapering at both ends (Sreekanth of the tropics, C. puncticollis additionally
Attaluri, pers. comm.). in Africa, and Euscepes postfasciatus in the
In addition to regional and local West Indies. On-station and farmers field
preferences for storage-root colour, DM, experiments show that there are significant
texture and taste, the acceptability of sweet differences in weevil tolerance among sweet
potato varieties is mainly determined by pest potato genotypes (Hahn and Leuschner,
and disease pressures. However, the number 1981), but this tolerance appears to be less
of pest and diseases in sweet potato are pronounced or inexistent on-farm. At the
considerable lower than in potato. Generally, same time, farmers in drought-prone regions
sweet potato varieties must have a certain of Malawi want sweet potatoes in which
degree of tolerance to Sweet potato virus storage roots are formed deep in the soil
disease (SPVD). The disease occurs after and which are clearly tapering at the top,
infection by two viruses: the Sweet potato because they associate this with less weevil
feathery mottle virus (SPFMV) and the Sweet damage (Ibrahim Benesi, pers. comm.).
potato chlorotic stunt virus (SPCSV). The Moreover, latex in the storage root skin has
SPCSV is the more problematic component been associated with less weevil damage by
of SPVD, because yield losses due to SPFMV, farmers, and varieties like Santo Amaro from
in the absence of SPCSV co-infection, are Brazil clearly have considerably less weevil
low and SPFMV resistance in sweet potato damage than other sweet potato varieties
breaks after the plant is infected by SPCSV (Rafael Vasquez Martinez, pers. comm.).
(Gibson et al., 1998; Karyeija et al., 2000). The International Potato Center (CIP)
SPVD often causes serious yield losses in is promoting OFSPs to alleviate vitamin A
high-virus-pressure zones of sub-Saharan deficiency in the world (Low et al., 2007;
Africa, and American OFSPs have failed in Pfeiffer and McClafferty, 2007). However,
many regions of sub-Saharan Africa due to introductions from the Americas failed in
insufficient SPVD tolerance. Although the the high-SPVD-pressure zone of eastern
virus pressure of SPVD along the Pacific Africa (as did the FV Jonathan). To a
coast of South America is not extreme, certain extent this was associated with the
Selection methods. Part 5: Breeding clonally propagated crops 281
storage root flesh colour and taste. At up regional platforms for sweet potato
the same time, local African OFSP FVs, breeding in eastern, southern and West
such as Ejumula, Carrot C, Carrot Dar Africa, with a focus on dual purpose OFSP
es Salaam and Zambezi, and locally-bred (human consumption and animal feed),
OFSP MVs, such as NASPOT5 (Mwanga drought tolerant OFSP, and non-sweet high
et al., 2003), have been adopted after DM OFSP. PPB has so far mainly been a
awareness campaigns on the vitamin A research component in the organization of
deficiency problem (Regina Kapinga, pers. sweet potato breeding.
comm.). For this reason, CIP puts emphasis There are strong indications that PPB
on decentralized sweet potato breeding, in early selection steps of the sweet potato
and has recently started to recommend breeding process increases the efficiency
incorporation of at least one participatory and minimizes the risk of making wrong
selection step in the breeding process. selection decisions. In contrast to PPB, PVS
In sweet potato breeding for human is tentatively a form of on-farm evaluation
consumption, decentralization is (in the frame of a larger number of multi-
characterized by a general overall goal: that location trials) and cannot be as efficient
of developing more OFSP varieties that as PPB, because there is considerably less
meet local needs and consumer preferences, genetic variation, and, for highly heritable
to alleviate hunger and malnutrition and traits, there is nearly no genetic variation
to improve public health. The emphasis is at later breeding stages among clones. Not
on organizing OFSP breeding in eastern surprisingly, akin to the role of IPB in
and southern Africa, with national OFSP potato crop evolution, it has been shown
breeding programmes starting recently in that farmers have the ability to manage
West Africa (Ghana and Nigeria) and south- selection stages in sweet potato (Gibson
west and central Asia (India, Bangladesh and et al., 2008). This is consistent with
Sri Lanka). Breeding is almost exclusively results for potato (Gabriel and Torrez,
carried out by national agricultural 2000) and cassava (Manu-Aduening et al.,
research system (NARS) breeders and on 2006). Farmer selections are mainly made
NARS breeding stations, with currently by visual screening. This includes quality
12 NARS and two sweet potato breeders characteristics, diseases and pests, as well as
from the CGIAR system involved. NARS the growth type, which is to a certain extent
breeders are provided with funds for associated with drought adaptation in sweet
parental recombination and to consider the potato (see below on selection in early
quality trait of storage root flesh colour breeding stages). Most importantly, farmers
in the breeding process. Main emphasis in use more criteria and characters to select
breeding is given to: (1) material exchange sweet potato clones than do breeders in FPB
at the seed and clone level, (2) exchange of (Gibson et al., 2008). In such a situation,
information, knowledge and results from the risk of FPB is to ignore characters that
breeding trials by annual meetings, reports are important for good overall performance
and back-stop visits, and (3) a sweet potato of a clonal variety. However, in the study
breeding research and training build up of Gibson et al. (2008) in three provinces in
on the needs shaped among discussions Uganda, the most important characteristics
between NARS and CGIAR breeders. This for selection by farmers in early selection
has resulted in an additional aim to build stages were common to those used by
282 Plant breeding and farmer participation
(see Section 13.5) involved in appropriate are eaten fresh, or are only boiled or
choices of parents in breeding clonally roasted, and when they are processed this
propagated crops. is often carried out at the household level.
To consider the range of needs, Exceptions are sugar cane, fruit crops used
preferences and adaptation requirements for the juice industry, and to certain extent
for the large number of clonally propagated root and tuber crops (potato, cassava and
crops is out of scope in this chapter. Here sweet potato) when they are used for the
we want to give the principles of breeding starch, alcohol or biofuel industries. In
clonally propagated crops and how PPB can resource-poor environments, yields and
be carried out or linked into these breeding yield stability with low input are a priority,
programmes. The breeding objectives in addition to consumer acceptability. As
and methods will be considered for four has been mentioned above, a major factor
agricultural crops in more detail at the that determines yields, yield stability and
end of this chapter, namely: potato, sweet adaptation are pests and diseases. The most
potato, cassava, and banana or plantain. important pests and diseases of important
Table 13.1 gives the plant parts used for clonally propagated crops in agriculture
propagation, the world production and by eco-geographical region are given in
the area harvested, as well as the polyploid Table 13.2, together with the most important
level of the most important clonally quality characteristics.
propagated crops in agriculture. Obviously, Most clonally propagated crops are
quality characteristics determined by polyploid (Table 13.1). An exception
consumer preferences and market needs is cassava, which can be considered as
are key characteristics for breeding clonally a polyploid behaving like a diploid (see
propagated crops, because many of these below). Polyploidy is an important aspect
TABLE 13.1
Data on the 11 most important clonally propagated crops on a global basis
Species Planting material World production Area harvested Polyploidy
Potato Sprout tubers 315 106 t 18.8 106 ha 2x, 3x, 4x, 5x
(Solanum tuberosum)
Cassava Hardwood cuttings 226 106 t 18.6 106 ha 2x
(Manihot esculenta)
Sweet potato Sprout cuttings 124 106 t 9 106 ha 6x
(Ipomoea batatas)
Yam Root tubers 51 106 t 4.6 106 ha 3x10x
(Dioscorea spp.)
Taro Corms 12 106 t 1.8 106 ha 4x
(Colocasia esculenta)
Sugar cane Cane stalks 194 106 t 20.4 106 ha 8x
(Saccharum ofcinarum)
Banana and Plantain Corms 105 106 t 9.6 106 ha 3x
(Musa paradisiaca)
Citrus fruit Bud stick grafting on 89 106 t 5.6 106 ha 2x, 3x+1, 4x-3
(Citrus spp.) rootstocks
Grapes Hardwood cuttings 69 106 t 7.4 106 ha 6x
(Vitis vinifera)
Apple Bud stick grafting on 64 106 t 4.8 106 ha 2x, 3x
(Malus pumila) rootstocks
Strawberry Adventitious shoots 4 106 t 0.26 106 ha 8x
(Fragaria grandiora)
NOTES: FAOStat 2006 at faostat.fao.org, Sucrose production.
284 Plant breeding and farmer participation
TABLE 13.2
Quality characteristics, pests and diseases by production zone of the 11 most important clonally
propagated crops in agriculture and horticulture
Major production Quality characteristics Pest and diseases
zones
in crop evolution (as we have already seen hexaploids (six sets; 6x) (Tate, Soltis and
in potato) and has important consequences Soltis, 2005) and species with even higher
in breeding clonally propagated crops. It is polyploidy levels are known (Table 13.1).
important to note that all breeding lines The haploid level (one set; 1x) does not
or varieties of clonally propagated crops are occur as a normal stage in the life cycle of
homogenous (clone lines and varieties are a crop. However, haploid plants occur by
genetically fixed and as homogenous as non- spontaneous mutations, wide crosses and
segregating breeding lines or hybrids from anther culture (e.g. diploids are developed
breeding self-fertilized or hybrid crops). from tetraploid potatoes by pollination with
The homogenous clones are exact genetic specific clones of S. phureja and haploids by
copies of their mother plants, if mutations anther culture). Haploids are occasionally
are ignored. This is more or less obvious in used in FPB of clonally propagated crops,
potato, cassava or sweet potato field plots, especially potatoes (Hermsen and Verdenius,
or in fruit and tree plantations, provided 1973; Wenzel and Foroughi-Wehr, 1984). In
no genotype mixtures are observed. What crop evolution different polyploidy levels
is not directly obvious to an observer originated from genome mutations and by
is that each clone line or variety in the hybridization between very closely related
field or plantation is a highly heterozygous species. Autopolyploids and allopolyploids
hybrid (clone lines and varieties are highly include wheat (Triticum durum and T.
heterozygous hybrids comparable with aestivum), canola (Brassica napus) and cotton
heterozygous hybrids developed in hybrid (Gossypium spp.), and nearly all clonally
breeding). It should be noted that due propagated species are autopolyploids. The
to polyploidy, clonally propagated crops homologous chromosomes in autopolyploids
are usually more heterozygous than those are similar enough that multivalents of the
diploid crops in which hybrid breeding same homologous chromosomes are formed.
is applied. The difference between clone Doubling of chromosomes occurs if the
hybrids and seed hybrids such as maize spindle poles are not developed when the
is that the first are propagated by asexual nucleus is dividing chromosomes in mitotic
reproduction and the latter are developed and meiotic cell division. There are several
by sexual reproduction. possible outcomes of abnormal meiosis.
Natural formation of 2n gametes was
13.3 POLYPLOIDY most important in evolution of cultivated
General knowledge about polyploidy is Solanum species, and the formation is mainly
required to get an understanding of breeding determined by one recessive gene (Watanabe
clonally propagated crops. Polyploidy has and Peloquin, 1988), so this character can
a strong effect on the performance of clones be used in breeding potato. Polysomic
as well as the parentoffspring correlations. inheritance is sensitive to disorders and
A polyploid genotype contains more than therefore autopolyploids often have reduced
two homologous sets of chromosomes in fertility, and occasionally they are completely
the nucleus of somatic cells. According to infertile and propagate only asexually.
the number of chromosome sets in the Multiple chromosome sets occur
nucleus we distinguish different polyploid spontaneously in nature from 2n gametes
types: triploid (three sets; 3x), tetraploid and can be induced artificially by colchicine
(four sets; 4x), pentaploids (five sets; 5x), (an alkaloid of autumn crocus, Colchicum
Selection methods. Part 5: Breeding clonally propagated crops 287
crops is represented by vegetative plant gametes), has been proposed for breeding
parts. Most polyploids display heterosis tetraploid potatoes (Ortiz, 1998). Polyploidy,
relative to their parental species, as well as heterozygosity and heterosis make the
relative to inter-gene-pool crossings within selection of good parents in population
a species. A polyploid population contains improvement of clonally propagated crops
three, four, five, six or more alleles at each very difficult. A good parent generates
locus. Hence, considerably more effects large genetic variation around a high family
due to dominance and epistasis are possible, mean. Cross-prediction and inter-gene-pool
and the genetic variation due to dominance crosses are very important in population
and epistatic effects in polyploidy crops improvement of clonally propagated crops.
is very large compared with the genetic This aspect of clonal breeding is often
variation caused by dominance and epistatic neglected and this might be the reason for
effects in diploid crops. For this reason, the low level of breeding progress in many
the performance of clonally propagated clonally propagated crops. In contrast
crops is mainly determined by heterosis. to population improvement (selection of
Usually in breeding of clonally propagated superior parents see Section 13.7 below),
crops, an F1 clone hybrid is crossed selection within a given genetic variation
with another F1 clone hybrid, so that for variety development is relatively easy in
the offspring shows extremely extensive clonally propagated crops (discard inferior
segregation. In parentoffspring studies it material). All the genetic advantages of
is possible to determine mid-parent and clonally propagated crops can be used for
mid-offspring heterosis, as well as the best- variety development, and the genotype
parent mid-offspring heterosis (similar to finally released is in the hands of the breeder
the assessment of heterosis in a hybrid immediately after the initial crossings.
breeding programme of diploid cropssee A clonally propagated crop that has no,
Chapter 11). In polyploids, more than one or nearly no, sexual reproduction is close
allele per locus is transferred in gametes to a dead end in evolution and breeding.
to the next generation, so that, in contrast Genetic variability can only accumulate
to diploids, the genetic variation due to by mutations. However, this source of
dominance determines the response to new variation has often been used to find
selection in population improvement as enhanced types of fruits and ornamentals
long as the population is not in equilibrium (van Harten and Broertjes, 1988).
(after recombining parental material in Nevertheless, the main source of generating
controlled crossings, a population is usually new variation in clonally propagated crops
not in equilibrium). In tetraploid potato is sexual reproduction. Owing to a more
populations that are not in equilibrium, or less regular meiosis in polyploids with
one-third of the dominance variance is an even number of chromosome sets
exploitable for selection progress when (4x or 6x), sexual seed production and
selection takes place on the female and male generation of new genotypes is possible.
sides (Wricke and Weber, 1986; Gallais, Nearly all clonally propagated crops, e.g.
2004). The exploitation of the dominance potato, sweet potato and cassava, are cross-
variance in population improvement, in fertilized crops in combination with self-
combination with the selection for different incompatibility. Incompatibility alleles
levels of ploidy (using the inheritance of 2n are the reason why specifically sought
Selection methods. Part 5: Breeding clonally propagated crops 289
seed often differ considerably from plants years. It is obvious that the wide range of
raised from vegetative planting material, quality preferences and the numerous pests
(ii) the plants raised from seeds are normally and diseases in each clonally propagated
grown in pots in greenhouses, and for crop and their interaction with genotypes
most traits this is not representative of justifies decentralization and participatory
field conditions, and (iii) a single plant approaches. However, the better simulation
evaluation is usually not appropriate, with of the final target environment realized
the exceptions of susceptibility to highly with FPB justifies a stronger PPB approach.
aggressive pathogens. In field crops, an Many advocate PPB because the stress and
important factor is interplant competition. marginal field conditions of resource-poor
A genotype must be tested under conditions farmers are not adequately simulated by
that simulate the field conditions in practice. FPB (see also below). In this context the
For this reason several plants of each two clear advantages of breeding clonally
clone are tested in plots in blocks under propagated crops should be pointed out:
homogenous field conditions. The aim is an (i) no genetic changes occur in genotypes
unbiased comparison of genotypes within after seed has been produced; and (ii) the
blocks. The number of plants per plot and total genetic variation of genotypes
the plot size depends on the crop as well as (comprising the genetic variances due to
the breeding stage. Fruit trees and perennial additive, dominance and epistatic effects)
shrubs are tested in larger plots with fewer can be exploited by selection. For these
plants than potatoes, and these again are reasons, only the genotypeenvironment
tested in larger plots than cut flowers. Early (GE) interaction and the plot error must be
selection stages (A-clones and B-clones) are considered (and reduced by testing in several
tested in smaller plots than later selection environments) to identify the best clone.
stages (C-clones and D-clones). The amount
of planting material at each breeding stage is 13.4.1 Early breeding stages and PPB
determined by the propagation coefficient In the general breeding scheme
of the crop. For example potato has, (Figure 13.1) each surviving seed plant
among clonally propagated crops, a very is cloned to be raised as A-clones in
low propagation coefficient of about ten, observation plots (visual screening of
whereas sweet potato has a relatively high general clone performance), or evaluation
propagation coefficient of between 30 and plots (recording of data on specific traits
90 (depending on the field propagation of each clone). Figure 13.2 shows the
method used). This is one factor why potato planting of sweet potato A-clones. The
breeding is relatively slow (about eight to plot size of A-clones is usually a single-row
ten years from cross to variety release). plot comprising 3 to 5 plants. The trial is
Breeders do not breed for a single conducted with no replications. It is open
environment; they breed for a range to discussion whether A-clones should
of environments. Hence, the field be evaluated at two locations. Selection
evaluations must simulate the range of theory results show that it is nearly always
target environments. For this reason, and the best resource allocation to test as many
depending on the propagation coefficient, clones as possible at one location, without
the clones are tested in plots, in homogenous replications (Wricke and Weber, 1986).
blocks, at several locations and for several Many breeders use only one location at the
Selection methods. Part 5: Breeding clonally propagated crops 291
FIGURE 13.2
Planting early selection stages of sweet potato for the
accelerated breeding scheme in San Ramon (one of four locations)
early breeding stages due to the restrictions a marginal or hot-spot environment (for
of the propagation coefficient and breeding drought, salinity, biotic challenge, etc.) can
budget. However, there are several reasons be discarded, or at least considered with
to test A-clones at two locations: (i) a trial caution in good environments.
at one location can be lost (e.g. extreme A-clones are only selected for highly
weather conditions) and then a full breeding heritable traits such as general performance
step and population is lost; (ii) trials at only (growth type; tuber, root or fruit size, shape
one location are of little value (the GE and colour), resistance to pests and diseases,
interaction cannot be separated from the harvest index, dry matter and nutritional
genotypic effect); and (iii) the response to quality. Breeders nearly always conduct
selection is still very close to the optimum a visual selection at the A-clone breeding
in a wide range of scenarios, including stage. However, it can be questioned if the
the scenarios where A-clones are tested A-clones selected by the breeder match
at two locations (Grneberg et al., 2004). farmer needs and would be selected by
Moreover, information from contrasting farmers. With two locations, one location
environments can be combined if the can be easily evaluated by farmers in a
breeder tests A-clones at two locations. PPB approach, while the other location
For example, clones that clearly fail in is used by the breeder. It should be noted
292 Plant breeding and farmer participation
that visual selection of general performance nutritional quality (starch, vitamins and
can also be an efficient indirect selection micronutrients by fast through-put analysis
for yield. In sweet potato, we observed methods) (Hartmann and Buning-Pfaue,
among several thousand A-clones grown in 1998; Lu, Huang and Zhang, 2006; Zum
1 m-row plots at three locations a heritability Felde et al., 2007; Bonierbale et al., 2009).
for yield of about h2 0.4 (harvesting and In the next season, B-clonesalso
recording all A-clones for yield at all three called promising clonesare planted in
locations). It was considered as useless larger plots in 2 to 3 rows with planting
work, because a visual selection at the material obtained from selected A-clones.
first location resulted in a nearly common The B-clone trials are still conducted
set of selected clones and a heritability without replications, but generally at two
for yield of about h2 0 in the selected or more locations. The B-clone stage is
fraction. This was demonstrated for two usually the beginning of selection for low
different breeding populations grown in heritability traits such as yield, biomass
two different seasons, so that the breeding and yield stability. The determination of
scheme was changed. Only those clones stability parameters such as the slope of
that have passed the visual selection step at the regression line and deviations from
location 1 are harvested and considered for regression (Fox, Crossa and Romagosa,
storage root quality evaluations at location 2 1997) requires at least three locations.
and 3. However, relying on visual selection However, it should be noted that stability
in early breeding stages requires a person parameters from less than 6 environments
who is very experienced with sweet potato. are still of little value. As mentioned above,
We think farmer participation at the visual a strong justification for PBB is that stress
selection stage in early breeding stages and marginal field conditions of resource-
is essential to avoid genotypes entering poor farmers are not adequately simulated
later breeding stages with characteristics by FPB (Ceccarelli, 1994). Cross-over GE
(storage root size, shape, form, colour, etc.) interactions occur, and what appears to be
unacceptable to farmers. As described above, good in resource-rich environments often
farmer preferences vary substantially both does not perform well in resource-poor
within and between regions, and the visual environments. This has also been clearly
selection can be conducted by independent observed in sweet potato, and outstanding
farmer groups. The advantages of PPB are clones for resource-poor environments were
very obvious in the early selection stages discarded by FPB (e.g. the clone SR92.499-
of clonal breeding, in which large numbers 23; Grneberg et al., 2005). Usually, but not
of fixed genotypes must be screened for always, the response to selection in poor
many highly heritable traits. PPB in the production environments is smaller than in
early selection stages has been successfully good production environments. The genetic
applied in potato (Gabriel and Torrez, variance is smaller while interaction and
2000), cassava (Manu-Aduening et al., error are larger, so that the performance of
2006) and sweet potato (Gibson et al., individual clones becomes more difficult
2008), and by working with two or three to distinguish. However, outstanding
locations it can be linked into FPB, in genotypes with different growth types
which selection is conducted for traits that adapted to resource-poor environments
cannot be evaluated by farmers, such as cannot display their full potential if FPB
Selection methods. Part 5: Breeding clonally propagated crops 293
does not test in such environments. Taking 3- to 5-row plots. All important agronomic
sweet potato breeding as an example again, traits are determined, including taste and
yield stability is associated with harvest post-harvest characteristics. Furthermore,
index (Grneberg et al., 2005). Under it merits determination of the above-
drought stress, good performing sweet mentioned stability parameters: (i) slope
potato clones have a harvest index of about of the regression line, and (ii) deviations
0.5 (on the basis of fresh matter storage from regression, as well as conducting an
root yields and total fresh matter biomass Additive Main Effect and Multiplicative
yields). Vine production is of considerable Interaction (AMMI) Analysis in those
importance to farmers to obtain sufficient cases where the regression model does
planting material for the next growing not fit (Fox, Crossa and Romagosa, 1997).
season. Clones performing well in resource- Usually, a clone is considered to have stable
rich environments usually fail in drought- performance if the slope of the regression
stress environments due to insufficient vine line is close to 1, and the deviations from
production rather than to unacceptable the regression line are small. An important
storage root production. At the same question in later breeding stages is that
time, outstanding clones in drought-stress of how many locations and how many
environments show a strong increase in vine replications to use. With more locations and
production with medium storage root yields more replications, the estimation of the yield
when grown in environments with good performance of clones is more reliable. At
water supply (Andrade, unpublished). The the same time, for a given testing capacity,
selection of genotypes with desired growth increasing the number of locations and
types or desired sinksource allocations replications results in fewer clones being
in marginal environments requires that tested. Generally, the gain from increasing
breeders evaluate the breeding population the number of replications is less than
in such an environment; this characteristic that obtained by increasing the number of
cannot be determined in a resource-rich genotypes and locations. Investigations of
environment. Here we suggest linking this problem in selection theory have led
the evaluation in a marginal environment to a recommendation to conduct advanced
with the visual selection in early breeding clone trials still with no replications but in
stages. All clones that fail in the marginal the maximum number of environments that
environment (i.e. extreme reduced storage can be managed by the breeder (Utz, 1969,
root production or vine production) are cited in Wricke and Weber, 1986). However,
eliminated from all other selection steps. many scientists are still very reluctant to
conduct trials without replications. Since
13.4.2 Later breeding stages and PPB the fixed costs of experimental stations are
At the beginning of the C-clone and high, it is usually an advantage to (i) create
D-clone selection stages the breeding artificial environments on experimental
population has been reduced to between stations (by running part of a station without
30 and 300 clones. While the number of fertilizer or with less irrigation) and (ii) to
clones in later selection stages is further go on-farm to evaluate clones with farmers,
reduced, those selected clones are tested i.e. PVS. However, nearly all of the initial
in more environments and in replications. genetic variation in the breeding population
The plots for C-clones and D-clones are has been discarded at later breeding stages,
294 Plant breeding and farmer participation
1951; Wricke and Weber, 1986). In ABS, from two hybrids cannot be reproduced
independent culling is conducted: (i) in a by crossing the hybrids again) without a
poor resource environment where clones system that maintains and provides at least
undergo visual selection; (ii) only those some healthy planting material.
clones passing the first selection step are Numerous viruses, bacteria and fungi
harvested in environments 2 and 3 to are transmitted by vegetative planting
determine yield and quality of selected material. Viruses are particularly important,
good performance clones over all traits and because viral diseases cannot be controlled
environments (index values are determined chemically. Viruses are spread by vectors,
by the Pesek-Baker index (Pesek and Baker, most often aphids and whiteflies. The
1969) to assist the breeder in their selection traditional maintenance of varieties and
decisions); and (iii) only those clones that production of healthy planting material
have passed the second selection step are includes protecting the base plants of
harvested in environment 4, where clones varieties in greenhouses or under nets, and
were already planted in season 2 in a farmers to prevent the development of a vector
field under high SPVD pressure in a third population by intensive use of insecticides.
selection step to select for SPVD tolerance. The base material is also termed mother
About 300 sweet potato clones enter the later plants. However, under these conditions,
breeding stages. In two subsequent seasons only 20 to 200 plants of each variety can
and two selections steps, 4 to 5 clones are be maintained, and planting material must
finally selected for variety release (first be produced in the field. These clones in
season: 300 clones, three environments, two the field for producing healthy planting
plot replications and 5-row plots; second material are the so-called S-clones, because
season: 40 clones, 16 environments, two planting material is usually called seed in
plot replications and 5-row plots). This is clonally propagated crops. Healthy S-clone
carried out in cooperation with NARS and production is supported by (i) application
farmer groups. of insecticides against vector populations
(monitoring by yellow cards); (ii) choosing
13.6 MAINTAINING VARIETIES AND locations for S-clone production that
S-CLONE MULTIPLICATION are out of range of vector populations
As a result of clonal propagation, (i.e. locations close to the sea or in cool
maintaining varieties should not be highlands); and (iii) removing all visibly
difficult. Genetic changes in varieties do infected plants from S-clone fields.
not occur by undesired crossings nor The detection of virus infections has
by segregation, and mutations are rare. been simplified by use of the enzyme-linked
However, the opposite is the truth, and immunosorbent assay (ELISA) procedure.
maintaining clonally propagated varieties is The principle is a reaction between the
a difficult and expensive part of the breeding viruses in plants and antibodies against these
operation. The main reason is that in clonal viruses. The reaction is made visible by an
propagation through vegetative plant parts, enzymatic colour formation. In practice,
many more diseases can be transmitted some leaf sap is pressed out and the colour
compared with seed propagation. A new reaction is assessed on special test plates
variety will have no impact in practice, and coated with antibodies. In the case of sweet
even can be lost (a clone hybrid developed potato, the plants tested negative for viruses
296 Plant breeding and farmer participation
prior knowledge on the cross the breeder are rarely used in practice. One suggestion
has to make as many crosses as possible, is to determine the value of a parent on the
which is also minimizing the risk of raising basis of the offspring performance from test
genotypes with poor performance (Wricke crosses. Another suggestion is to work on
and Weber, 1986). As mentioned above, most a reduced polyploidy level, which has been
clonally propagated crops are polyploid and especially proposed for breeding tetraploid
highly heterozygous, so that dominance potatoes (Ross, 1986). However, the latter
and epistatic effects contribute considerably has been little applied in practice for
to clone performance. For this reason, it parental selection, but has often been used
should be assumed that not much is known to incorporate germplasm of wild Solanum
about the value of a cross combination species into advanced breeding populations
until it has been made and tested. This is in (Tarn et al., 1992). Parental selection on the
agreement with our observations in sweet basis of test crosses are made on a large scale
potato, where the correlation between in potato breeding programmes for long-
mid-parent and mid-offspring yields is low day, temperate climates (150 to 500 cross
(r 0.5). We currently recommend raising combinations per breeding programme,
10 to 20 genotypes per cross combination, cited by Ross, 1985). It has been observed
while increasing the number of cross that specific combining ability is nearly
combinations to the maximum possible as large as general combining ability, and
with the resources available. However, after in some cases specific combining ability
clones of these crosses have been evaluated, has been observed that is clearly larger
the good crosses should be repeated on a than general combining ability (Sanford,
large scale. An optimum for the number and 1960; Mullin and Lauer, 1966; Tai, 1976;
size of crosses can determined if estimations Killick, 1977; Veilleux and Lauer, 1981;
are available for the genotypic variance Gaur, Gopal and Rana, 1983, cited by Tarn
between crosses and within crosses, and the et al., 1992; Gopal, 1998; Kumar, 2004; De
non-genetic variance components (Wricke Galarreta et al., 2006). This is not surprising
and Weber, 1986). Breeders often generate a as long as potato breeders do not work with
large number of seeds in polycross nurseries, two clearly separate gene pools for variety
but in these only the female parent is development. In potato breeding for tropical
controlled. The correlation between parent and subtropical regions, heterosis and
and mid-offspring in breeding populations high general combining ability have been
derived from polycross nurseries is half observed between andigena and tuberosus
of mid-parentmid-offspring correlation in gene pools in tuber-propagated potatoes
controlled crosses. and in true-seed potatoes (Enrique Chujoy,
Often the parents are chosen due to their pers. comm.). However, as long as these
performance per se. For theoretical reasons, gene pools are not improved on the basis of
this cannot be very secure in clonally- general combining ability separately from
propagated crops. Clone varieties are the complementary gene pool, such effects
highly heterozygous hybrids and usually cannot be exploited in the long term.
polyploids, so that segregation in crossings In sweet potato experiments we observed
is almost unpredictable. Therefore, for a long a mid-parentmid-offspring heterosis of
time now, suggestions have been made for 84 percent among 48 cross combinations (or
better assessment of parents; however, they 184 percent if the mid-parent value is set to
298 Plant breeding and farmer participation
100 percent). This is a clear indication that colour, as well as pest and disease resistanc-
the design of breeding schemes using the es), while others are selected simultaneous-
combining ability of two gene pools merits ly by aggregating characters into an index
investigation. Two breeding gene pools are (often an intuitively formed index, such as
available for sweet potato to test heterosis: score values for overall performance).
the Jewel Gene pool, developed mainly A parent appears to have a good overall
from North American varieties, and the trait performance if no trait is below the
Zapallo-SPK Gene pool, developed mainly population average. However, only in those
from South American and African FVs. cases where trait associations are close to
The value of a parent is nearly always zero or positive can it be expected that
determined by several characteristics. In parents with good performance over all
general, parents should be recombined with traits produce offspring in which each char-
a good combining ability and good per- acter has been improved. In parental selec-
formance over all traits. The PPB study in tions, negative trait associations can be very
Uganda (Gibson et al., 2008) underlines critical. Table 13.3 gives an example for
how many characteristics are important for sweet potato, in which DM shows a strong
good performance over all traits. Moreover, negative trait association with pro-vitamin
FPB also has the aim of improving nutri- A, iron and zinc concentrations, as well as
tional quality, especially pro-vitamin A, a moderate negative trait association with
iron and zinc concentrations (Pfeiffer and storage root yield. The associations in the
McClafferty, 2006) in potato, sweet potato, example are strong enough that under vari-
cassava, plantain and other crops. With an ous scenarios of multi-trait selection the
increasing number of characters, breeders breeding population is improved for yield,
operate with larger breeding populations, pro-vitamin A, iron and zinc, whereas the
as in potato and sweet potato. Aiming at DM of the population decreases.
only 30 genotypes finally selected, and In other words, the DM is changing
assuming 10 characters each, selected in in the wrong direction even though it
sequential selection steps with a selected was selected for improvement. These sur-
fraction of ten percent (1 out of 10), then prising undesired effects in the case of
300 000 000 000 genotypes would be needed sweet potato and DM improvement in
in the original base population. Populations connection with pro-vitamin A, iron and
of this size cannot be established in prac- zinc improvement was also observed for
tice. Moreover, even if the population size the Williams selection procedure and this
is extremely large, some desired combi- index selection procedure (Williams, 1962)
nations probably do not exist, such as comes very close to intuitive selection pro-
sweet potato genotypes with high yield, cedures used by breeders in which a weight
high SPVD tolerance, high DM and high is assigned to each trait on the basis of its
pro-vitamin A, iron and zinc concentra- economic importance. The only selection
tions). Often, breeding can only approach procedure that can monitor the response
the desired genotype in several steps of to selection in each trait is the Pesek Baker
recombination and selection. In practice, index (Pesek and Baker, 1969). However,
some characters are selected sequentially this index requires estimations of genetic
(especially where there is clearly a low- variance and co-variances, but the proce-
est acceptable value (tuber size, shape and dure ensures that parents are selected that
Selection methods. Part 5: Breeding clonally propagated crops 299
TABLE 13.3
Estimations of genetic correlations for yield, dry matter, total carotenoids, iron and zinc in sweet
potato storage roots of 24 megaclones and 26 advanced breeding clones grown in at two locations
in two replications
Storage root yield Dry matter Total carotenoids Iron
develop populations in which traits are combining ability and improving gene pools
improved according to a ratio of desired on the basis of general combining ability
genetic improvements (so-called desired values (called reciprocal recurrent selection
genetic gains) given by the breeder. in maize breeding) are the most difficult
An alternative is the Elston index tasks in breeding; however, they can greatly
(Elston, 1963), in which the breeder can increase yield gains. At the same time, we
raise the threshold for the trait at risk by think that the visual selection of potential
modifying the lowest acceptable value for parents in a PPB approach should be used
each. This index can be easily applied in as additional information by the breeder. It
each replication and environment, so that should be noted that the work plan for both
index mean values for each genotype can the selection of parents for the next cycle of
be calculated together with other statistical selection and the early selection stages for
parameters (Grneberg et al., 2005). variety development are always to a certain
We are aware of only one case in which extent in common. In sweet potato breeding
PPB has been applied for the selection of at CIP we use a combination of sequential
parents in clonally propagated crops. In and simultaneous index selection in early
the Cochabamba region of Bolivia, farmers selection stages (see also above): (i) visual
selected potato parents in an andigena selection by eliminating all genotypes that
population, which had been improved do not meet the lowest acceptable values
for agronomic performance and Late for each trait (this lends itself to PPB);
blight tolerance. Selected clones in this (ii) in the remaining selected fraction (about
population were used as parents with the 2 500 clones), apply index selection for
regionally grown FV Waycha (Gabriel yield and nutritional quality traits using the
and Torrez, 2000) and the PPB approach Pesek-Baker index, with the square roots
included hand-crossing by farmers. We of variance components as desired genetic
think that the ability of farmers in the gains; and (iii) selecting for pest and disease
selection of parents is limited beyond a tolerance (mainly SPVD) in the remaining
selection of clone performance per se. Test selected fraction (about 300 clones) by
crosses, general combining ability, specific visual selection (this lends itself to PPB) and
300 Plant breeding and farmer participation
systems used in hybrid seed production. It no longer apply. Moreover, breeding TPS
is interesting that apomixis is distributed potato as a cross-fertilized crop will not lead
across many plant families. It appears not to to completely homogenous varieties. This
be controlled by a complex genetic system. is the reason why only a few TPS varieties
An example is Guinea grass (Panicum have been developed in the Northern
maximum), in which the sexual tetraploids Hemisphere. All these have been exclusively
are recessive homozygous (aaaa), whereas used for home garden production. However,
apomictic genotypes carry a dominant we think that by using hybrid selection
allele and are heterozygous (Aaaa) (Savidan, schemes and inbreeding in two separate
1983). So far, studies on apomixis have been gene pools it should be possible to develop
mainly made in tropical grasses, but more more and more homogenous and attractive
and more attention is being paid to rice and TPS varieties.
maize. There are opinions that apomixis The advantages of TPS are mainly
systems will become available to breeders, of interest in tropical and subtropical
and in this context gene isolation and an regions of the world. Under these climatic
apomixis gene have been mentioned conditions, the production, storage and
(Savidan, 2000). However, so far there is no transportation of potato planting material
such apomixis system usable in breeding is difficult. Moreover, potato yields are
programmes. The major problem is that considerable lower in tropical and
plants with the same genotype can express subtropical regions of the world than in
different degrees of apomixis. the Northern Hemisphere, so that about
20 percent of the harvest is needed as
13.9 PROPAGATION OF POTATOES BY planting material. Hence TPS varieties in
SEED the tropics can have 20 percent lower yields
Finally, the option that clonally propagated compared to clonally propagated potato
crops can be propagated by sexual seeds is varieties and remain competitive. About 20
considered. In many countries there have TPS varieties have been developed. Most
been research projects in which potatoes interesting are those varieties developed
were cultivated by seed. These are potatoes from recombination of the andigena and
that are sown instead of planted. Since the tuberosus gene pools. However, the original
planting material of clonally propagated idea of raising seedlings in nurseries and then
potatoes is often called a seed potato, the planting seedlings into the field by hand has
term true potato seed (TPS) was introduced. not been adopted. What has been adopted is
The use of TPS has two principle advantages: to raise TPS varieties in seedling nurseries to
the most important potato diseases cannot obtain healthy planting material, and then
be transmitted in true seed, and only a few to cultivate these TPS varieties for several
hundred grams of TPS are needed to cultivate growing seasons as a clonally propagated
a potato field, where usually several tonne crop, and to request true seed again after
of tubers are needed (Simmonds, 1997). yield declines are significant due to declining
This is associated with two disadvantages: health status (Fuglie, 2001). However, today,
potatoes grown from seed are weak in not more than 10 000 ha of TPS are grown,
vigour and are sensitive to many factors, mainly in Asia, which trace back to about
and the breeding method and advantages eight TPS varieties. The future of TPS is
of breeding clonally propagated crops can debatable. From the breeding perspective,
302 Plant breeding and farmer participation
the future of TPS will mainly depend on the Americas (41 million tonne) and Africa
working with two gene pools, in which a (16 million tonne) (FAO, 2006). The top 20
certain extent of inbreeding is applied, with potato producing countries are China (22
subsequent use of general combining ability percent), The Russian Federation (12 percent),
to improve these two gene pools. India (8 percent), Ukraine (6 percent), United
States of America (6 percent), Poland (3
13.10 POTATO percent), Germany (3 percent), Belarus (3
Breeding potatoes has been reviewed by percent), Canada (2 percent), France (2
Tarn et al. (1992). The andigena potato percent), United Kingdom (2 percent),
(Solanum tuberosum subsp. andigena; Turkey (2 percent), Netherlands (1 percent),
autotetraploid with 48 chromosomes) Bangladesh (1 percent), Brazil (1 percent),
originated in the highlands of South Romania (1 percent), Peru (0.8 percent),
America about 5000 BC, while today two- Spain (0.6 percent), Nepal (0.5 percent) and
thirds of world potato production is in Pakistan (0.5 percent).
temperate latitudes. Following introduction
into Europe, andigena evolved into the Irish Breeding objectives
potato (S. tuberosum subsp. tuberosum), Characteristic of potato breeding is the
which is mainly characterized by day-length large number of breeding objectives. For
neutrality, uniformity of tuber shape, shorter the Irish potato, quality traits are at least
crop duration and higher harvest index than as important as yield. Moreover, breeding
andigena. Andigena remains the predominant for resistance against numerous pest and
cultivated potato in the Andes, whereas the diseases, e.g. numerous viruses (potato leaf-
Irish potato is the potato of commerce roll virus (PLRV), Potato virus Y (PVY)
in long-day temperate climates. Potatoes and Potato virus X (PVX)), Late blight
introduced into other, tropical, regions of (Phytophthora infestans), dry rots (Fusarium
the world trace back to breeding populations spp.), soft rot and blackleg (Erwinia spp.),
derived from crossings between andigena cyst-forming nematodes (Globodera
and Irish potato. However, in the Andean rostochiensis and G. pallida) have major
region, seven other potato species are still importance in long-day temperate as well
in cultivation; most important are phureja as in tropical temperate climates. For the
(S. phureja; diploid with 24 chromosomes), andigena potato, these pests and diseases
limea, ajanhuri and rucki. In addition to are of nearly similar importance (i.e. late
these cultivated species, 160 wild potato blight can destroy the whole crop in cool,
species are known (Hawkes, 1979 and 1981; high-altitude regions, especially when the
Spooner and Hijmans, 2001), so that potato weather is wet).
might have the largest gene pool among crops. There are clear differences between
Wild and indigenous species are important tropical temperate and tropical hot
resources of pest and disease resistance for climates. At temperatures above 25C,
andigena and Irish potato. The evolution of Late blight and cyst-forming nematodes
the potato was described in the introduction decline rapidly in importance, but Early
of this chapter. Asia and Europe are the blight (Alternaria solani) and root-knot
worlds largest potato producing regions, nematodes (Meloidogyne spp.) take
with annual production of about 130 and over, and Bacterial wilt (Pseudomonas
128 million tonne, respectively, followed by solanacearum) is widespread in tropical
Selection methods. Part 5: Breeding clonally propagated crops 303
lowlands. The phureja potato (for PVY and with considerable levels of total carotenoids,
Late blight) and the wild species S. acaule including pro-vitamin A.
(for PLRV, PVX and both Globodera spp.)
and S. demissum (for PLRV, PVY, and late Breeding methods
blight) are important resources in breeding Crossing is relatively easy. In nature,
for tolerance and resistance. It should be crossings occur easily by open pollination
noted that the pests and diseases presented by insects. For breeding purposes the
represent only the most important species. flower architecture of the potato allows
Ross (1985) provides a list of resistance easy emasculation and controlled hand
sources in wild potato species. pollination. A fruit with about 200 seeds
However, it is possible to find tolerance develops from each successful pollination. In
or resistance genes in cultivated and wild commercial breeding, controlled crossings
potatoes against nearly all potato diseases. are usually made (both parental genotypes
An exception is Bacterial wilt, for which so are clearly defined). Genotypes with good
far no useful tolerance or resistance have performance over all traits and with a certain
been found for breeding purposes. Today, degree of genetic distance are recombined.
all new Irish potato varieties contain one The value of a cross combination is usually
or more resistance genes from wild and determined in test-crosses with 100 to
other cultivated potato species. For the 200 seeds per combination. Occasionally
Northern Hemisphere, yield, crop duration, plant breeding text-books recommend
tuber size, shape and flesh colour, eye combining parents with complementary
depth, starch content, storability, cooking traits. However, many breeders find that
characteristics, taste and suitability for this results in potatoes breeding in a wild
mechanical harvesting, as well as processing segregation, so that finally only genotypes
characteristic for chips (crisps) and French can be selected with moderate performance
fries (chips) are the most important quality over all traits. Each year, breeders plant
breeding objectives. For tropical regions, 10 000 to 200 000 seeds, which trace back
yield, regional adaptability, crop duration, to 150 to 200 cross combinations. Crossings
storability, cooking characteristics, taste and are made with flower sprouts obtained from
nutritional quality are the most important cuttings of field plants, grown in greenhouses
quality breeding objectives. Outside of the in nutrient solutions. Frequencies of
Andes, crop duration is one of the most successful crosses differ tremendously
important traits (i.e. in south-west and between parental combinations and about
central Asia there is a requirement for one-eighth to one-quarter of all parental
potato varieties with less than 80 days combinations cannot be recombined due to
crop duration). Recently, focus has been no flowering, low pollen quality, no fruit
given to improve pro-vitamin A, iron and formation or genetic incompatibility. For an
zinc concentrations in tubers to alleviate overview of overcoming crossing barriers
micronutrient malnutrition in tropical in potatoes, the reader is referred to Jansky
regions (potatoes have comparatively high (2006).
iron and zinc concentrations). This has Pre-breeding crosses are important
resulted in a separate breeding programme when one requires to improve one or two
for phureja, which has the highest iron and traits in an enhanced breeding gene pool
zinc contents among potatoes, together (e.g. shorter crop duration). Pre-breeding
304 Plant breeding and farmer participation
sixteenth century into the Philippines, whence adapted to salinity, drought and marginal
it spread to other islands and the east Asian soil conditions (Woolfe, 1992).
mainland. Portuguese seafarers introduced The crop has recently received more
the crop into Europe, Africa and India. Today interest due to the very high levels of
it is cultivated in 117 countries in all tropical pro-vitamin A (concentrations of up to
and subtropical regions of the world. Asia is 700 ppm DM) in OFSPs, and hence as
the worlds largest sweet potato producing a vehicle to reduce vitamin A deficiency
region, with about 107 million tonne of problems in the world (Huang, Tanudjaja
annual production, followed by Africa and and Lum, 1999; Low, 2007). We observed
the Americas, with approximately 15 and up to 1 200 ppm -carotene on a DM
3 million tonne, respectively. The top 12 basis in clones with variety potential in
producing countries are China (80 percent), our breeding population Jewel II (this
Nigeria (2.8 percent), Uganda (2.2 percent), corresponds to 30 mg -carotene in 100 g
Indonesia (1.5 percent), Viet Nam (1.2 fresh sweet potato storage roots. A pre-
percent), United Republic of Tanzania (0.9 schooler needs 4.8 mg -carotene per day,
percent), Japan (0.8 percent), India (0.8 and it merits discussion as to what extent
percent), Burundi (0.7 percent), Kenya (0.6 OFSP should be recommended as baby
percent), Rwanda (0.6 percent) and United and weaning food. Moreover, storage roots
States of America (0.6 percent). Further provide medium levels of iron and zinc
important sweet potato producing countries (Woolfe, 1992). Recent finding of about
are Angola, Argentina, Bangladesh, Brazil, 50 ppm DM iron and 40 ppm DM zinc in
Cuba, Egypt, Ethiopia, Haiti, Korea deep orange fleshed sweet potato storage
(Democratic Republic of), Madagascar, Peru, roots (Burgos and zum Felde, pers. comm.)
the Philippines and Papua New Guinea, merits further investigation.
with annual production between 0.3 and The stems and leaves can have spinach-
0.5 million tonne (FAO, 2006). Nearly half like taste and some varieties are used in
of the sweet potato produced in Asia is used China specifically as a green vegetable.
for animal feed, with the remainder primarily Stems and leaves have on DM basis about
used for human consumption, either as fresh four times more protein, iron and zinc than
or processed products. In Africa, the crop storage roots. It appears that stems and
is cultivated almost exclusively for fresh leaves must be cooked to reach an acceptable
consumption. iron bioavailability, but investigations into
Sweet potato is a perennial vine, propa- iron bioavailability of sweet potato tops is
gated by cuttings, and usually cultivated as very limited.
an annual crop. The planting distances in There is new demand for purple-fleshed
fields vary. In Africa, planting distances are sweet potato due to the health-promoting
usually 1 m between rows and 30 cm within effects of anti-oxidant anthocyanin sub-
rows. In China, recommended planting stances, and cell lines for a potentially
distances are 75 cm between rows and ongoing production for the food industry
20 cm within rows. The crop duration is have been established (Konczak, 2006).
very short (4 to 6 months) and the crop is However, much more important appears
even cultivated in northern China. It pro- to be the demand for non-sweet sweet
duces more edible energy per hectare per potatoes, but few genotypes are non-sweet
day than wheat, rice or cassava, and is well (Kays, 2006). There is a very large genetic
306 Plant breeding and farmer participation
variation for DM, starch and sugars in the use of the crop as animal feed and
sweet potato, and a strong positive correla- folder. The breeding for high DM and
tion has been observed for DM and starch, extractable starch is relatively easy: the
whereas a strong negative correlation was target is a high starch yield per hectare.
found between sugars and DM and starch However, currently, in many regions of
(Grneberg et al., 2009). This is nearly the world the price of sweet potato starch
ideal for the breeding target of a non-sweet currently cannot compete with the price of
high-DM sweet potato type, and we think cassava starch. Only in large regions where
that the development of non-sweet sweet the growing period is too short for cassava
potatoes should not be too difficult. within the cropping system (e.g. China) is
there an economic demand for sweet potato
Breeding objectives varieties for starch production. Breeding
FPB started very late for sweet potato. for biofuel production is in its initial stages,
One of the first breeding programmes was and so far variety recommendations for this
established at Louisiana State University in purpose are made on the basis of screening
the 1920s. Today there are several strong existing successful varieties. The breeding
national breeding programmes (e.g. China, of purple-fleshed sweet potatoes as a
Japan, South Africa, Uganda, United separate breeding programme is a relatively
States of America and Uruguay) and one new trend, and so far only carried out on
international breeding programme, at CIP a small scale in Japan, Indonesia and Peru.
(Peru). Four major breeding objectives can Future targets are the non-sweet sweet
be clustered: (i) breeding of OFSP for potato, and quick cooking features (cv.
consumption of storage roots and leaves; Quick Sweet) (Katayama et al., 2006), as
(ii) breeding for high DM and extractable well as suitability for processing into chips,
starch; (iii) breeding for biofuel production, puree, juice, weaning and baby food, and
which has started in China (Dai Fu Ma, bread on the basis of a wheat-sweet potato
pers. comm.); and (iv) breeding of purple- flour mixture (Woolfe, 1992); these trends
fleshed sweet potatoes for consumption. appear nearly exclusively in east Asia, and
In breeding for consumption, it has to be for recent developments the reader should
considered that people in different regions consult proceedings, such as Liu (2008).
have very different taste preferences; the Major constraints on high yields are
extremes are low DM content, moist mouth pests and diseases, especially Sweet potato
feel, very sweet taste and deep orange flesh chlorotic stunt virus (SPCSV) and the sweet
colour, versus high DM, bland, dry mouth potato weevils. The prevailing diseases and
feel, low sweet taste and white, yellow or insects affecting sweet potato vary from
orange flesh colour. In breeding for human region to region. There are about 35 bac-
consumption, focus is more on high DM terial and fungal diseases, more than 20
OFSP varieties with elevated iron and zinc viruses or virus-like agents, 20 nematodes
concentration and a dry and less-sweet and 20 insect species known to affect sweet
mouth taste. This breeding is hampered potato (Martin and Jones, 1986).
by a strong negative genetic correlation Currently there are only four important
between storage root DM and storage root pest and diseases: SPVD, Alternaria, sweet
pro-vitamin A, iron and zinc contents. The potato weevils and the root-knot nematode.
breeding for human consumption includes The most important virus is whitefly-
Selection methods. Part 5: Breeding clonally propagated crops 307
transmitted SPCSV, which often occurs in this reason, a transgenic approach using
co-infection with Sweet potato feathery Bt genes has received attention. Recent
mottle virus (SPFMV aphid-transmitted). findings of compounds in the latex of the
Clear synergistic disease effects are seen storage root skin and the effect of these on
with SPFMV and SPCSV (the so-called weevils might of interest for breeding (P.C.
SPVD virus complex). Generally, all Stevenson, H. Muyinza, D. Hall and R.
varieties need a certain degree of tolerance Mwanga, unpubl.).
to SPVD, and there is genetic variation for
SPVD (Mwanga, Yencho and Moyer, 2002). Breeding methods
Very high tolerance or resistance is needed True seed set occurs easily in nature by cross
in eastern Africa. Currently, it is assumed pollination (by insects, mainly bees), and for
that SPFMV and all other sweet potato breeding purposes the flower architecture
viruses (except SPCSV) are not important, of sweet potato allows easy emasculation
because sweet potato has an effective virus and controlled hand pollination. A skilled
defence system, which is broken by SPCSV technician can make 200 crossings per day,
(I. Barker, pers. comm.). with a success rate of 25 percent. From
The major fungal disease in subtropical each successful cross, two or three true
America is Fusarium wilt (Fusarium seeds are obtained. Not all sweet potato
oxysporum f.sp. batatas) and in the African parents flower readily, but flowering can
highlands the main problem is Alternaria be easily induced by grafting on Ipomoea
storage root, leaf spot and stem blight nil (2n = 30 chromosomes). It should be
(Alternaria spp.). Although there are many noted that frequencies of successful crosses
bacterial and fungal diseases with a wide differ tremendously between parental
distribution, high levels of tolerance or combinations, and about one-third of all
resistance are frequently found. This is also parental combinations are incompatible,
true for resistance to nematodes. with no seed formation. The sweet potato
There has been recurrent success in seed has a hard coat and needs to be
breeding for root-knot resistance against scarified with concentrated sulphuric acid
new races of Meloidogyne spp. (Martin to obtain even and rapid germination. In
and Jones, 1986). However, in regions a well managed breeding nursery, after 3
with a pronounced dry season, the greatest months it is possible to obtain 40 to 60
constraints are sweet potato weevils (Cylas cuttings from a true seed plant if the plant
formicarius elegantulus in all parts of the is grown in the field, and 20 to 30 cuttings if
tropics, C. puncticollis and C. brunneus the plant is grown in a pot in a greenhouse.
in Africa, and Euscepes postfasciatus in The extreme genetic make up of the crop
the West Indies). It has been an objective (hexaploid, highly heterozygous, open-
to find weevil resistance for more than pollinated by insects with true seed set
50 years, but differences in weevil attack occurring easily), the short crop duration
probably depends on preference factors (45 months), and the rapid propagation
of the weevil. It is believed that dense (40 to 60 cuttings from one plant) permits
storage roots developed deep below the soil the design of a very efficient and rapid
surface are less susceptible than less dense, breeding system.
moist-fleshed storage roots. No effective The recombination of parents is still
weevil resistance has been found so far. For usually carried out in polycross nurseries
308 Plant breeding and farmer participation
shrivel and are sun dried until they shatter, ca. 100 clones are tested in replicated trials
releasing hybrid seeds that are ready for at three locations, and consumer acceptance
germination. Cassava seed have a very short is assessed. Final selections are multiplied
dormant period and germinate quickly. and enter dissemination in year 6.
No scarification is necessary. Few seeds In eastern and southern Africa, 10 000 to
germinate unless the mean temperature 50 000 true-seed plants are raised for the first
exceeds 24C, with a temperature exceeding selection step and screened for resistance
30C for at least part of the day; the best rates to major diseases and pests at 1, 3, 6 and 9
occur at 3035C. A dry heat treatment of months after sowing, namely East African
14 days at 60C is also beneficial for newly cassava mosaic disease (EACMD), African
harvested seeds. If temperatures permit and cassava mosaic disease (ACMD), Cassava
irrigation is available, the easiest method is brown streak disease (CBSD) and CBB.
to sow the seeds direct into the soil. This In a second selection step, 2 000 clones are
is successful at IITA because temperatures planted in single-row plots (3 to 5 plants) at
from January to March range from 30 11 m spacing. The observations made in the
to 35C. At CIAT, seeds are frequently first year are repeated again at 1, 3, 6 and 9
planted in a screen house and the emerging months after planting. Each clone is scored
seedlings held until they reach 2025 cm for yield, and agronomic characteristics
before being transplanted to well prepared assessed, such as branching height and angles,
soil with good moisture conditions. canopy and number of stems per plant. In a
Since many national programmes do third selection step, 20 to 50 clones are
not have a continuous cassava crossing grown in preliminary yield trials in single
programme, they rely on distribution of pre- rows with ten plants per clone and three
selected clones from the two international replications at one to three locations. In year
institutions, CIAT and IITA. The improved 6, the final selections are taken on-farm and
germplasm generated is distributed either into national variety release trails.
in the form of elite genotypes transferred In the Americas and Asia, cassava
in vitro, or as populations of recombinant improvement is closely linked with the insti-
seeds (full-sibs or half-sibs). Cassava tutions Embrapa (Brazil), FCRI Rayoung
breeding operates with larger populations (Thailand) and CIAT (Colombia). In con-
than potato or sweet potato. trast with Africa, there are no extremely
In West Africa, up to 100 000 true devastating diseases. CIAT established
seed plants are raised from field-sown 50 000 seedling selections for particular
seed, which are screened in a first selection climatic zones. Up to 20 parents from each
step for resistances to CMD and CBB. gene pool are disseminated for evaluations
At harvest, selection is for compact roots to national centres in similar edapho-cli-
with short necks, stems branching at about matic zones. From this programme a very
100 cm, with low HCN in the leaves. broad range of improved diversity has been
In the second selection step, about 3 000 developed and distributed worldwide.
clones are grown in small, non-replicated Generally, MVs in Asia can be traced back
plots. Further selection is made for disease to 100 crosses between Asian and American
resistances, yield potential and root DM parents (Kawano, 2003). Recent findings
content, and the HCN in the roots is assayed show that the general combining ability for
enzymatically. For the third selection step, cassava fresh root yields are clearly larger
312 Plant breeding and farmer participation
than the specific combining ability across (Simmonds, 1976; Ortiz, 1995). They are a
contrasting environments (Ceballos et al., staple in Samoa, the Philippines, south India
2004). This is a clear indication that heterotic and the West Indies. Around 87 percent of all
gene pools in cassava can be formed and bananas and plantains grown worldwide are
exploited by improving two gene pools with produced by small-scale farmers for home
a reciprocal recurrent selection scheme. consumption or for sale in local markets.
About two-thirds of world production is
13.13 BANANA OR PLANTAIN dessert bananas and one-third plantains.
Breeding bananas and plantains has The fruit export market comprises only
been reviewed by Rowe (1984), and Jain one-sixth of total world production. The
and Swennen (2001) have edited recent banana is the number one fruit crop in
proceedings on banana improvement, with the world, with about 70.5 million tonne
a main emphasis on biotechnology. Banana produced annually. The top ten producing
and plantain (Musa paradisiaca, Musaceae, countries are India (24 percent), Ecuador (9
usually triploid with 33 chromosomes) percent), Brazil (9 percent), The Philippines
originated in Southeast Asia. The term (8 percent), China (8 percent), Indonesia (5
plantain is used for those bananas that are percent), Costa Rica (3 percent), Mexico (2
palatable only when cooked. The crop was percent), Thailand (2 percent) and Colombia
introduced into Africa about 3 000 BPE. (2 percent). Plantains are grown as a staple
Introduction into the Americas came after food in 52 countries worldwide with a total
1 500 AD. Today the crop is cultivated production of 34 million tonne. The top ten
worldwide in the tropics. Bananas and plantain producing countries are Uganda
plantains evolved from two diploid (30 percent), Colombia (9 percent), Rwanda
wild species, Musa acuminata (AA) and (8 percent), Ghana (7 percent), Nigeria (6
M. balbisiana (BB) in the Eumusa series percent), Peru (5 percent), Cote dIvoire (4
(x = 11) of the genus Musa. An exception percent), Congo (4 percent) and Kenya (3
is the small group of Fehi bananas in percent) (FAO, 2006).
the Pacific, which have their origin in the Bananas and plantains are one of the
Australimus series (x = 10) of Musa. All very few crops in which breeders are still
export fruit bananas are triploids (AAA) trying to find an appropriate conventional
and originated from M. acuminata. All breeding method to develop new MVs.
plantains and several locally preferred fruit Nearly all cultivars are FVs and have been
bananas are hybrids between M. acuminata selected from genetic variation developed
(AA) and M. balbisiana (BB). The higher by natural evolution. In cases of crop failure
dry matter (about 58 percent) and higher due to new pathogens and diseases, FPB still
starch content of plantains compared to focuses on identifying alternative cultivars
pure M. acuminata cultivars is attributed within existing genetic variation (collections
to the BB genome. The AAB cultivars and large screening programmes). Hence,
have long curved fruits and appear like an important source for identifying new
an oversized export banana. They are cultivars are germplasm collections held in
important food crops in south India, eastern trust in genebanks, such as the International
and central Africa and tropical America. Musa Germplasm Collection in Leuven,
The ABB cultivars have thick straight fruits, Belgium. Spontaneous mutants in Musa have
which are much shorter than the AAB types played a very important role in banana and
Selection methods. Part 5: Breeding clonally propagated crops 313
plantain breeding, including the replacement widely distributed. However, for several
of the export banana cultivar Gros Michel of these FHIA cultivars, taste and cooking
(susceptible to Panama disease or Fusarium qualities are still problematic (Roux, pers.
wilt (Fusarium oxysporum sp. cubense) by comm.). Further breeding programmes
Cavendish banana cultivars, which are have been set up at the Empresa Brasiliera
resistant to most fusarium wilt pathogens, de Pesquisa Agropecuaria (Embrapa) in
and the replacement of the plantain cultivar Brazil, the Instituto de Investigaciones
Horn plantain (AAB) (susceptible to Black en Viandas Tropicales (INIVIT) in Cuba,
sigatoka (Mycosphaerella fijiensis)) by the the Centre Africain de Recherches sur
Laknau cultivar (AAB), which is tolerant Bananiers et Plantains (CARBAP) in
to Black sigatoka and closely resembles the Cameroon, the International Institute of
Horn plantain (Stover, 1972). However, the Tropical Agriculture (IITA) in Nigeria, and
cooking qualities of Laknau are inferior the National Research Centre on Banana
to Horn plantain. Owing to the low level (NRCB) in India.
of occurrence of spontaneous mutations,
mutagenic agents and mutation breeding Breeding objectives
have often been used to generate new In breeding, resistance against Panama
genetic variation in bananas and plantains, (Fusarium wilt) and sigatoka diseases are
followed by screening programmes for in the foreground. In the first half of the
plants with resistance or tolerance to twentieth century, Panama disease destroyed
pest and diseases, coupled with desirable approximately 40 000 ha of bananas in
agronomic qualities (e.g. tolerance to Central and South America. Fortunately,
Panama disease; tolerance to the toxin of resistant Cavendish cultivars could substitute
Mycosphaerella fijiensis; short; larger fruit for the predominantly grown Gros Michel
size; and earliness). The FPB programmes variety. However, Cavendish cultivars are
for bananas and plantains started in the not resistant to all fusarium wilt pathogens
early 1900s, to develop new AAA cultivars (i.e. race 4). It should be noted that Panama
for the export market, with resistance disease cannot be controlled chemically,
against Panama disease or Fusarium wilt so that use of resistant varieties is the only
(Fusarium oxysporum f.sp. cubense). way to maintain production in regions with
Despite continued breeding efforts, no new challenge from this disease. The leaf spot
banana and plantain cultivar acceptable diseases caused by Mycosphaerella musicola
by farmers and consumers was bred until (Yellow sigatoka) and M. fijiensis (Black
the 1980s (Roux, 2001). Nevertheless, by sigatoka) are costly pathogens and must be
the end of the twentieth century, efforts regularly controlled by fungicides. Cultivars
to improve Musa started to focus on the with an AAA genome are very susceptible
use of diploid and tetraploid gene pools to both sigatoka diseases. The Horn
to develop triploid and tetraploid bananas plantain is resistant to Yellow sigatoka, but
and plantains. To date, the first improved susceptible to Black sigatoka. The latter
cultivars (AAA, AAAA, AAB, AAAB and disease threatens continued cultivation of
AABB), developed at Fundacin Hondurea the plantain food crop. Triploid cooking
de Investigacin Agrcola (FHIA) in bananas of the ABB type, such as Chato,
Honduras through the International Pelipita and Saba, are highly tolerant
Musa Testing Program (IMTP), have been to the Black sigatoka pathogen. However,
314 Plant breeding and farmer participation
Chato is susceptible to bacterial wilt or tillering capacity (Ferwerda and Wit, 1969;
Moko disease caused by Pseudomonas Rowe and Richardson, 1975; Persley and
solanacearum and to race 2 of Fusarium De Langhe, 1987).
oxysporum f. sp. cubense, while Pelipita does
not meet flavour and fruit-shape preferences, Breeding methods
so that currently only Saba remains as a Triploid bananas and plantains are vegetatively
possible substitute for the Horn plantain. parthenocarpic, i.e. no pollination is
Moreover, nematodes, mainly the burrowing necessary for fruit development. In diploids,
nematode (Radopholus similis), are a major pollination often results in seeded fruits.
constraint to bananas in monoculture, Diploids are not suitable as varieties since
and outside of the Americas the Bunchy fruit size and plant vigour are low. However,
top virus is widely distributed, which is diploids are the basis for crop improvement.
transmitted by the banana aphid (Pentalonia In the initial stages of breeding efforts, a few
nigronervosa). Many diploid accessions of seeds per bunch in some triploid varieties
M. acuminata subsp. malaccensis and M. a. were used when these had been pollinated
subsp. burmannica are resistant to races by diploid genotypes. The reason for this
1, 2 and 4 of Panama disease. Sources of seed production and genetic variation is the
resistance to Yellow sigatoka are available in formation of unreduced triploid gametes in
several subspecies of M. acuminata, while some triploid female parents after pollination
M. a. subsp. burmannica is highly tolerant within diploid male parents, which produces
to the Black sigatoka fungus. The tolerance reduced haploid gametes. The progenies of
in M. acuminata accessions to sigatoka these crosses are tetraploid. This method
diseases is apparently controlled by several was used to generate genetic variation with
dominant genes. Resistance to the burrowing the female banana parent Gros Michel and
nematode has been found in the Pisang Jari the female plantain parent Laknau (AAB),
Buaya group of diploid accessions. The which closely resembles the Horn plantain.
resistance is controlled by one or very few Tetraploid hybrids (AAAA) from crosses
dominant genes and has been incorporated with Gros Michel were resistant to Panama
into diploid and polyploid progenies. Today, disease and closely approached commercial
several FHIA varieties are resistant to acceptability, but the inferior agronomic
burrowing nematodes (Kalorizou, Gowen characteristics of the diploid parents
and Wheeler, 2007). were also present in the hybrids. Triploid
Among agronomic qualities, dwarfness hybrids derived from crosses between these
is most important in bananas and plantains, tetraploid hybrids and diploid genotypes
because they are often grown in areas were useless. Unfortunately, the cooking
with periodic strong winds. Dwarf and qualities of hybrids derived from Laknau
semi-dwarf mutants have been found in were also inferior to those of the Horn
many diploid and triploid bananas and plantain. No seeds have been produced
plantains. Examples are Highgate (a dwarf from Cavendish clones and no other
mutant of Gros Michel) and the Cavendish suitable triploid parentsexcept Gros
cultivar Grand Nain. In dwarf diploids, Michel and Laknaufor seed production
the dwarfness character is controlled by unreduced gametes have been found.
by a single dominant gene. After this in This breeding method has not succeeded in
importance are fruit characteristics and creating acceptable new varieties. However,
Selection methods. Part 5: Breeding clonally propagated crops 315
the major finding of this work was that it Thompson and Aked, 2000; Ludger, 2005;
is necessary to improve the diploid male Kalorizou, Gowen and Wheeler, 2007).
parent gene pool to increase the chances However, to our knowledge, no PPB has
of developing either new tetraploid or new been applied in early breeding stages. Most
triploid varieties. likely the reason for this is that almost
Today, banana and plantain breeding no diploid clone in its performance per se
aims at producing tetraploid and triploid would achieve acceptability by farmers.
varieties on the basis of diploid accessions Nevertheless, the future of banana and
resistant to various diseases, and the plantain breeding, as in other clonally
continuous improvement of this diploid propagated crops, should be seen in testing
gene pool for agronomic qualities (i.e. plant the combining ability between two gene
height, fruit characteristics and tillering pools and in the improvement of two
capacity) as well as high pollen production. gene pools on the basis of the general
Crossings within the diploid gene pool are combining ability and reciprocal recurrent
complex: the diploid SH-2095, which was selection. In banana and plantain breeding,
later successfully used in tetraploid variety such a breeding system can be established
development, was derived from a four-way by a seed-fertile diploid gene pool with
cross of three diploid cultivars and one wild high pollen production and a seed-fertile
accession ((Sinwobogi Tjau Lagada) tetraploid gene pool, which is used as the
(Guyod a wild Musa acuminata subsp. male parent. In such a breeding programme,
malaccensis)). Nevertheless, the genetic basis PPB could easily be incorporated. However,
of diploid pollen parents with improved the important information provided by
agronomic performance is considerably the farmers would not be seen in the
wider than in the past. The currently evaluation of clone performance per se in
best diploids are continually crossed on the diploid and tetraploid gene pool, but in
triploid Highgate and Laknau, which the numbers of acceptable clones per cross
produces unreduced triploid gametes, as combination and family between genotypes
well as on seed-fertile tetraploids with good of the diploid and tetraploid gene pool, as
agronomic performance. The first results in described above in the section on selection
new potentially tetraploid varieties and the of parents and cross prediction.
later in new potentially triploid varieties.
The advantage of triploids in variety REFERENCES
development is that they are female-sterile Austin, D.F. & Huamn, Z. 1996. A synop-
due to the uneven number of chromosome sis of Ipomoea (Convolvulaceae) in the
sets. In contrast, the even number of the Americas. Taxon, 45(1): 338.
chromosome set in tetraploids requires an Baudouin, L., Baril, C., Clment-Demange,
additional selection step for female sterility A., Leroy, T. & Paulin, D. 1997. Recurrent
in variety development. Several improved selection of tropical tree crops. Euphytica,
FHIA cultivars (AAA, AAAA, AAB, 96: 101114.
AAAB and AABB) have been developed Baynes, R.A. 1972. Sweet potato varieties in the
by this breeding method, and farmers Eastern Caribbean. Caribbean Agriculture,
participate in the final breeding stages in 3(4): 2021.
acceptability studies of these tetraploid and Becker, H.C. 1993. Pflanzenzchtung.
triploid varieties (i.e. PVS) (Ssemwanga, Stuttgart, Germany, Eugen Ulmer.
316 Plant breeding and farmer participation
Bonierbale, M., Iglesias, F., Ariel, C. & Kazuo, De Galarreta, J., Ruiz, I., Ezpeleta, B.,
K. 1994. Genetic resources management of Pascualena, J. & Ritter, E. 2006. Combining
cassava at CIAT. In Root and tuber crops, ability and correlations for yield compo-
pp. 3952. Tsukuba, Ibaraki, Japan, Ministry nents in early generations of potato breed-
of Agriculture, Forestry and Fisheries. ing. Plant Breeding, 125(2): 183186.
Bonierbale, M., Grneberg, W., Amoros, W., De Haan, S. 2009. Potato diversity at height:
Burgos, G., Salas, E., Porras, E. & Zum multiple dimension of farmer-driven in situ
Felde, T. 2009. Total and individual carote- conservation in the Andes. PhD thesis,
noid profiles in Solanum phureja cultivated Wageningen University, The Netherlands.
potatoes: II. Development and application Dhaliwal, H.S. 1977. The Ph gene and the
of near-infrared reflectance spectroscopy origin of tetraploid wheats. Genetica,
(NIRS) calibrations for germplasm charac- 47(3): 177182.
terization. Journal of Food Composition and Elston, R.C. 1963. A weight-free index for
Analysis (in press). the purpose of ranking or selection with
Burgos, G., Salas, E., Amoros, W., Auqui, M., respect to several traits at a time. Biometrics,
Muoa, L., Kimura, M. & Bonierbale, M. 19: 669680.
2008. Total, individual carotenoid profiles FAO [Food and Agriculture Organization of
in the Phureja group of cultivated pota- the United Nations]. 2006. Estimates of
toes: I. Concentrations and relationships world production and harvested area in
as determined by spectrophotometry and 2006. Data from faostat.fao.org
HPLC. Journal of Food Composition and Ferwerda, F.P. & Wit, F. 1969. Outlines of
Analysis, in press. perennial crop breeding in the tropics.
Byrne, D. 1984. Breeding cassava. Plant Wageningen, The Netherlands, Veenman
Breeding Review, 2: 73135. and Zonen.
Ceballos, H., Iglesias, C.A., Prez, J.C. & Fox, P.N., Crossa, J. & Romagosa, I. 1997.
Dixon, A.G.O. 2004. Cassava breed- Multi-environment testing and genotype
ing: opportunities and challenges. Plant environment interaction. In R.A. Kempton
Molecular Biology, 56: 503516. & P.N. Fox, eds. Statistical methods for plant
Ceccarelli, S. 1994. Specific adaptation and variety evaluation, pp. 117138. London,
breeding for marginal conditions. Euphytica, Chapman & Hall.
77: 205219. Friis-Hansen, E. 1992. The failure of for-
CIP [International Potato Center]. 1977. mal plant breeding to meet the needs of
The Potato: Major Potato Diseases and resource-poor peasants in African arid
Nematodes. Lima, CIP. lands. African Arid Lands Working Paper
CIP. 1980. An interim approach to bacterial Series, No. 3/92. 12 p.
wilt control. CIP Circular, 8(12): 13. Fuglie, K.O., ed. 2001. Performance and pros-
CIP. 1984. Potatoes for the developing world. pects of hybrid true potato seed in south and
Lima, CIP. southeast Asia. Proceedings of the CIP-
Cochran, W.G. 1951. Improvement by means ADB symposium on Field Testing Hybrid
of selection. Proc. Second Berkeley Symp. TPS in the Lowland Tropics of Asia. Bogor,
Math. Stat. Prob., 449470. Indonesia, CIP. 250 p.
Cock, J. 1985. Cassava. New potential for a Gabriel, J. & Torrez, R. 2000. Participatory
neglected crop. Boulder, USA, Westview approaches in potato improvement:
Press. experience of PROINPA in Bolivia. In
Selection methods. Part 5: Breeding clonally propagated crops 317
C.J.M. Almekinders & W.S. de Boef, eds. C. Loechl, J. Low, J. Lynam, R. Mwanga, O.
Encouraging diversity. The conservation Ortiz, A. Oswald, & G. Thiele. Unleashing
and development of plant genetic resourc- the potential of sweetpotato in Sub-Saharan
es, pp. 194199. London, Intermediate Africa: current challenges and way forward.
Technology Publications. Nairobi, CIP-SSA.
Gallais, A. 2004. Quantitative genetics and Hahn, S.K. & Leuschner, K. 1981. Resistance
selection theory in autopolyploid plants. of sweet potato cultivars to African sweet
France, INRA. potato weevil. Crop Science, 21: 499503.
Gaur, P.C., Gopal, J. & Rana, M.S. 1983. Hartmann, R. & Buning-Pfaue, H. 1998. NIR
Combining ability for yield, its components determination of potato constituents. Potato
and tuber dry matter in potato. Indian Research, 41: 327334.
Journal of Agricultural Science, 53: 876879. Hawkes, J.G. 1979. Evolution and polyploidy
Gibson, R.W., Mpembe, I., Alicai, T., Carey, in potato species. In J. Hawkes, R.N. Lester
E.E., Mwanga, R.O.M., Seal, S.E. & Vetten, & A.D. Skelding, eds. The biology and
H.J. 1998. Symptoms, etiology and sero- taxonomy of the Solanaceae, pp. 636646.
logical analysis of sweet potato virus disease London, Academic Press.
in Uganda. Plant Pathology, 47: 95102. Hawkes, J.G. 1981. Recent concepts in the
Gibson, R.W., Byamukama, E., Mpembe, I., evolution of tuber bearing Solanums.
Kayongo, J. & Mwanga, R. 2008. Working pp. 4059, in Report of the planning con-
with farmer groups in Uganda to develop ference for the exploration, taxonomy and
new sweet potato cultivars: decentralisation maintenance of potato germplasm. CIP,
and building on traditional approaches. Lima.
Euphytica, 159: 217228. Hermsen, J.G. & Verdenius, J. 1973. Selection
Gopal, J. 1998. General combining ability from Solanum tuberosum group Phureja
and its repeatability in early generations of genotypes combining high-frequency
of potato breeding programmes. Potato haploid induction with homozygosity for
Research, 41(1): 2128. embryo-spot. Euphytica, 22: 244259.
Grneberg, W.J., Abidin, E., Ndolo, P., Pereira, Hillocks, R.J. & Wydra, K. 2002. Bacterial,
C.A. & Hermann, M. 2004. Variance com- fungal and nematode diseases. In R.J.
ponent estimations and allocation of Hillocks, J.M. Thresh & A.C. Bellotti,
resources for breeding sweet potato under eds. Cassava: Biology, Production and
East African conditions. Plant Breeding, Utilization, pp. 261280. Wallingford, UK,
123: 311315. CABI Publishing.
Grneberg, W.J., Manrique, K., Dapeng, Z. Hobhouse, H. 1985. Seeds of Change. Five
& Hermann, M. 2005. Genotype envi- Plants that Transformed Mankind. London,
ronment interactions for a diverse set of Sidwick & Jackson.
sweet potato clones evaluated across vary- Huang, A.S., Tanudjaja, L. & Lum, D. 1999.
ing ecogeographic conditions in Peru. Crop Content of alpha-, beta-carotene and die-
Science, 45: 21602171. tary fiber in 18 sweet potato varieties grown
Grneberg, W.J., Mwanga, R., Andrade, M. & in Hawaii. Journal of Food Composition and
Dapaah, H. 2009. Sweet potato Breeding. In Analysis, 12: 147151.
M. Andrade, I. Barker, D. Cole, H. Dapaah, Huamn, Z. 1983. The breeding potential of
H. Elliott, S. Fuentes, W.J. Grneberg, R. native Andean potato cultivars. pp. 9697,
Kapinga, J. Kroschel, R. Labarta, B. Lemaga, In W.J. Hooker, ed. Proceedings of the
318 Plant breeding and farmer participation
Levy, D. 1984. Cultivated Solanum tuberosum ment and storage in potato. Journal of
L. as a source for the selection of cultivars Experimental Botany, 55: 975982.
adapted to hot climates. Tropical Agriculture Mullin, R. & Lauer, F.I. 1966. Breeding
(Trinidad), 61: 167170. behavior of F1 and inbred potato clones.
Liu, Q., ed. 2008. Sustainable sweet potato pro- Proceedings of the American Society for
duction technology for food, energy health Horticultural Science, 89: 449455.
and environment. Proceedings of the 3rd Mumford, E., Boonhan, N., Tomlinson, J. &
China-Japan-Korea Workshop on Sweet Barker, I. 2006. Advances in molecular phy-
potato. China Agricultural University Press. todiagnostics: new solutions for old prob-
Low, J.W, Arimond, M., Osman, N., Cunguara, lems. European Journal of Plant Pathology,
B., Zano, F. & Tschirley, D. 2007. A food- 116: 119.
based approach introducing orange-fleshed Mwanga, R.O.M., Odongo, B., Turyamureeba,
sweet potatoes increased vitamin A intake G., Alajo, A., Yencho, G.C., Gibson, R.W.,
and serum retinol concentrations in young Smit N.E.J.M. & Carey, E.E. 2003. Release
children in rural Mozambique. Journal of of six sweet potato cultivars (NASPOT1
Nutrition, 137: 13201327. to NASPOT6) in Uganda. HortScience,
Lu, G., Huang, H. & Zhang, D. 2006. 38(3): 475476.
Application of near-infrared spectroscopy to Mwanga, R.O.M., Yencho, G.C. & Moyer, J.
predict sweet potato starch thermal proper- 2002. Diallel analysis of sweet potato for
ties and noodle quality. Journal of Zhejiang resistance to sweet potato virus disease.
University, Science B, 7(6): 475481. Euphytica, 128: 237248.
Ludger, J.S. 2005. Evaluation of banana and Nassar, N.M.A. 2000. Cytogenetics and evolu-
plantain (Musa spp.) cultivars in the south tion of cassava (Manihot esculenta Crantz).
of Haiti. Proceedings of the Florida State Genetics and Molecular Biology, 23(4):
Horticultural Society, 118: 258259. 10031014.
Mandal, R.C. 2006. Tropical Root and Tuber Nishiyami, I., Miyazaki, T. & Sakamoto, S.
Crops. 360 p Jodhpur, India, Agrobios. 1975. Evolutionary autoploidy in sweet
Manu-Aduening, J.A., Lamboll, R.I., Ampong potato (Ipomoea batatas (L.) Lam) and its
Mensah, G., Lamptey, J.N., Moses, E., progenitors. Euphytica, 24: 197208.
Dankyi, A.A. & Gibson, R.W. 2006. Nogler, G.A. 1984. Gametophytic apomixis.
Development of superior cassava cultivars pp. 475518, in B.M. Johri, ed. Embryology
in Ghana by farmers and scientists: the pro- of angiosperms. Berlin, Springer-Verlag.
cess adopted, outcomes and contributions NRC [National Research Council]. 1989. Lost
and changed roles of different stakeholders. crops of the Incas. Little-known plants of
Euphytica, 150: 4761. the Andes with promise for worldwide
Martin, F.M. & Jones, A. 1986. Breeding sweet cultivation. Washington, DC, National
potatoes. Plant Breeding Review, 4: 313 Academy Press for NRC.
345. OBrien, P.J. 1972. The sweet potato: its ori-
Miles, J.W. 2007. Apomixis for cultivar devel- gin and dispersal. American Anthropologist,
opment in tropical forage grasses. Crop 74: 343365.
Science, 47(S3): S238S249. Ochoa, C. 1990. The potatoes of South America.
Morris, W., Ducreux, D., Griffiths, D., Translated by D. Ugent. Cambridge, UK,
Stewart, D., Davies, H. & Taylor, A. 2004. Cambridge University Press. See pp. 314
Carotenogenesis during tuber develop- 337.
320 Plant breeding and farmer participation
Ortiz, R. 1995. Musa Genetics. In Gowen, ed. Roux, N.S. 2001. Mutation induction in Musa.
Bananas and Plantains, pp. 84109. UK, In S.M. Jain & R. Swennen, eds. Banana
Chapman and Hall. improvement: cellular, molecular biology,
Ortiz, R. 1997. Secondary polyploids, het- and induced mutations. Enfield, USA, and
erosis, and evolutionary crop breeding for Plymouth, UK, Science Publishers Inc.
further improvement of the plantain and Rowe, P. 1984. Breeding bananas and plan-
banana (Musa spp. L.) genome. Theoretical tains. Plant Breeding Review, 2: 135155.
and Applied Genetics, 94: 1131120. Rowe, P. & Richardson, D.L. 1975. Breeding
Ortiz, R. 1998. Potato breeding via ploidy bananas for disease resistance, fruit quality,
manipulations. Plant Breeding Review, and yield. Trop. Agric. Res. Services, La
16: 1585. Lima, Honduras.
Pques, L.E. 2004. Roles of European and Rowe, P.R. 1967. Performance and variability
Japanese larch in the genetic control of of diploid and tetraploid potato families.
growth, architecture and wood quality traits American Potato Journal, 44: 263271.
in interspecific hybrids (Larix eurolepis Salazar, L.F. 1996. Potato viruses and their
Henry). Annals of Science, 61: 2533. control. Lima, CIP. 214 p.
Persley, G.J. & De Langhe, E.A. 1987. Banana Sanford, L.L. 1960. Comparative evaluation of
and Plantain breeding Strategies. Canberra, clones as testers for yield, specific gravity
Australian Centre for International Research. and tuber appearance in the potato. Ph.D.
Pesek, J. & Baker, R.J. 1969. Desired improve- Thesis, Iowa State Univ., Ames, USA.
ment in relation to selection indices. Savidan, Y.H. 1983. Genetics and utiliza-
Canadian Journal of Plant Science, 49: 803 tion of apomixes for the improvement of
804. guinea grass (Panicum maximum Jacp.).
Pfeiffer, W.H. & McClafferty, B. 2007. pp. 182184, in J.A. Smith & V.W. Hays,
HarvestPlus: breeding crops for better eds. Proceedings of the 14th International
nutrition. Crop Science, 47: 88105. Grassland Congress, Lexington, Kentucky,
Radcliffe, E.B. 1982. Insect pests of potato. USA, 1524 June 1981. Boulder, USA,
Annual Review of Entomology, 27: 173 Westview Press.
204. Savidan, Y. 2000. Apomixis: genetics and breed-
Rehm, S. & Espig, G. 1991. The cultivat- ing. Plant Breeding Review, 18: 1086.
ed plants of the tropics and subtropics: Shiotani, I. & Kawase, T. 1989. Genomic
Cultivation, economic value, utilization. structure of the sweet potato and hexa-
Germany, Eugen Ulmer. ploids in Ipomoea trifida (HBK) DON.
Rich, A.E. 1983. Potato diseases. New York, Japanese Journal of Breeding, 39: 5766.
USA, Academic Press. Simmonds, N.W. 1962. The evolution of the
Ross, H. 1985. Kartoffel (Solanum tubero- bananas. London, Longmans, Green & Co.
sum L.). In W. Hoffman, A. Mudra & W. Simmonds, N.W. 1976. Evolution of Crop
Plarre, eds. Lehrbuch der Zchtung land- Plants. Harlow, UK, Longman.
wirtschaftlicher Kulturpflanzen. Vol. 2. 2nd Simmonds, N.W. 1979. Principles of Crop
edition. Berlin, Parey. Improvement. New York, USA, Longman.
Ross, H. 1986. Potato breeding problems and Simmonds, N.W. 1997. A review of potato
perspectives. Advances in Plant Breeding, propagation by means of seed, as distinct
Suppl. 13. Journal of Plant Breeding. Berlin, from clonal propagation by tubers. Potato
Parey. Research, 40: 99214.
Selection methods. Part 5: Breeding clonally propagated crops 321
Solis, R.S. 2004. Caral: the city of the Sacred der Pflanzenzchtung. Arbeiten Univ.
Fire., Peru, ITB. 260 p. Hohenheim, Vol. 49.
Soost, R.K. & Roose, M.L. 1996. Citrus. In J. van Harten, A.M. & Broertjes, C. 1988. Induced
Janick & J.N. Moore, eds. Fruit breeding, mutations in vegetatively propagated crops.
Vol. 1, Tree and tropical fruits, pp. 257323. Plant Breeding Review, 6: 5591.
New York, USA, Wiley. Veilleux, R.E. & Lauer, F.I. 1981. Breeding
Spooner, D.M. & Hijmans, R.J. 2001. Potato behaviour of yield components and hol-
systematics and germplasm collecting, lowheart in tetraploid-diploid vs. conven-
19892000. American Journal of Potato tional derived potato hybrids. Euphytica,
Research, 78: 237268. 30: 547561.
Ssemwanga, J.K., Thompson, A.K. & Aked, J. Watanabe, K. & Peloquin, S.J. 1988. Occurrence
2000. Quality and acceptability of the new of 2n pollen and ps gene frequencies in cul-
banana cultivar FHIA 3 compared to indig- tivated groups and their related wild species
enous Uganda cultivars for matooke prepa- in tuber-bearing Solanums. Theoretical and
ration. ActaHort (ISHS), 540: 561567. Applied Genetics, 78: 329336.
Stover, R.H. 1972. Banana, plantain and Wenzel, G. & Foroughi-Wehr, B. 1984. Anther
abaca diseases. Kew, UK, Commonwealth culture of Solanum tuberosum. In I.K. Vasil,
Mycological Institute. ed. Cell culture and somatic cell genetics of
Stover, R.H. 1980. Sigatoka leaf spots of banana plants, Vol. 1, pp. 293301. New York,
and plantains. Plant Disease, 64: 750756. USA, Academic Press.
Tai, G.C.C. 1976. Estimations of general Williams, J.S. 1962. The evaluation of a selec-
and specific combining abilities in potato. tion index. Biometrics, 18: 375393.
Canadian Journal of Genetics and Cytology, Witcombe, J.R., Joshi, A., Joshi, K.D. &
18: 463470. Sthapit, B.R. 1996. Farmer participatory
Tarn, T.R., Tai, G.C.C., De Jong, H., Murphy, crop improvement. I. Varietal selection
A.M. & Seabrook, J.E.A. 1992. Breeding and breeding methods and their impact
potatoes for long-day, temperate climates. on biodiversity. Experimental Agriculture,
Plant Breeding Reviews, 9: 217332. 32: 445460.
Tate, J.A., Soltis, D.E. & Soltis, P.S. 2005. Woolfe, J.A. 1992. Sweet potato: An untapped
Polyploidy in plants. In T.R. Gregory, ed. food resource. Cambridge, UK, Cambridge
The evolution of the Genom, pp. 371426. University Press.
San Diego, USA, Elsevier. Wricke, G. & Weber, W.E. 1986. Quantitative
Thresh, J.M. & Cooter, R.J. 2005. Strategies for Genetics and Selection in Plant Breeding.
controlling cassava mosaic virus disease in Berlin, de Gruyter.
Africa. Plant Pathology, 54(5): 587614. Yen, D.E. 1976. The sweet potato. In N.W.
Ugent, D.L., Dillehay, T. & Ramirez, C. 1987. Simmonds, ed. Evolution of crop plants,
Potato remains from a Late Pleistocene pp. 4245. New York, USA, Longman.
settlement in southcentral Chile. Economic Zitnak, A. & Filadelfi, M.A. 1985. Estimation
Botany, 41: 1727. of taste threshold of three potato glycoal-
Ugent, D.L., Pozorski, S. & Pozorski, T. 1982. kaloids. Canadian Institute of Food Science
Archaeological potato tuber remains from and Technology Journal, 18: 337339.
the Casma Valley of Peru. Economic Botany, zum Felde, T., Burgos, G., Espinoza, J.,
36: 182192. Bonierbale, M. & Grneberg, W.J. 2007.
Utz, H.F. 1969. Mehrstufenselektion in Analysis of carotenoid, iron, zinc and calci-
322 Plant breeding and farmer participation
um content of potato (Solanum phureja) and (13th ICNIRS), Ume, Sweden, & Vasa,
sweet potato (Ipomoea batatas) using near- Finland, 1521 June 2007. In press. Special
infrared reflectance spectroscopy (NIRS). issue of the Journal of Near Infrared spec-
In Proceedings of the 13th International troscopy (JNIRS). See www.impublica-
Conference on Near Infrared Spectroscopy tions.com/nir/page/NIR-2007
323
CHAPTER 14
ses are accepted through local research, new on definitions: a farmers variety of a self-
details and perspectives are sure to arise, pollinated crop (e.g. of barley or rice) may
and it is only by having an open minded, be composed of diverse genotypes that,
respectful attitude that outsiders can hope from a plant breeders perspective, may be
to learn and reap the benefits of collabora- different varieties. Therefore discriminating
tion with local farmers. Such an attitude among these genotypes would be selection
facilitates new insights and understandings from the perspective of farmers as it can
that can improve the accuracy and relevan- change the genetic make up of their variety,
cy of the scientific work. If plant breeders but would be choice from the perspective
think of their interactions with farmers as of plant breeders as it would not change
tests of how complete or accurate farmers the genetic make-up of varieties as they
knowledge is, breeders will lose a critical define them. At a more fundamental level,
opportunity for supporting collaboration, farmers choice of populations and varie-
respect and collegiality, and for improving ties determines the diversity available for
the quality of their own work. Thus, at least hybridization and subsequent selection of
initially, experiments and discussions with plants. For all of these reasons, we can say
farmers should be seen as opportunities to that selection and choice together deter-
learn, not to teach. mine the degree to which varieties stay
the same, change between generations, or
14.1.1 Choice as distinct from selection evolve over generations.
It is important to differentiate between Farmers and plant breeders make choices
choice of populations or varieties that between varieties and populations, especially
does not change the genetic make-up of in the initial stages of the selection process
these units, and the selection of plants when choosing germplasm for making
from within populations or varieties, with crosses (for plant breeders), and in the final
the potential to change the genetic make- stages when choosing among populations
up of these units, which may eventually or varieties generated from those crosses
result in new varieties (Cleveland, Soleri for further testing (Hallauer and Miranda,
and Smith, 2000). Farmer criteria for both 1988: 159), or for planting (farmers) or
choice and selection include agronomic, release (plant breeders). Farmers choices
economic, culinary and aesthetic charac- of varieties or populations when saving
teristics, as well as minimizing perceived seed for planting, in seed procurement and
risk. While the distinction is commonly in allocating different varieties to different
made in some participatory plant breeding growing environments also affects the
literature (e.g. Witcombe et al., 1996), the genetic diversity of their crop repertoires,
terms choice and selection are often not and establishes the diversity on which future
explicitly defined, and in some writing may selection will be based. (For simplicity, in
be used interchangeably. Obviously, distin- the discussion of choice we will use the
guishing between these is partly a function term variety to refer to both populations
of scale, as is most clearly seen in the case and varieties.)
of vegetatively propagated crops, in which
a single clone may be chosen to establish 14.1.2 A taxonomy of farmer selection
a new variety (e.g. with cassava; Pujol, A taxonomy of selection and its biological
David and McKey, 2005). It also depends effects can help to clarify the differences
326 Plant breeding and farmer participation
FIGURE 14.1
Phenotypic selection classified according to the agent of selection, and farmers goals
when the farmers are the agents, applied to traditional agricultural systems
and similarities between plant breeders a bridge between these is useful (Soleri
and farmers. Selection can be categorized and Cleveland, 2005). For plant breeding,
according to the agent carrying out the most fundamental model of the
phenotypic selection, and the intention of relationship among phenotype, genotype
the agent when it is a human (Figure 14.1). and environment is assumed to be a good
While all types of selection function in model of reality that is the basis for PBK; we
both farmer and professional breeding, will assume it is also the basis for FK. This
professional plant breeders see intentional model is universally accepted by biologists,
phenotypic selection for micro-evolution including plant breeders, but they disagree
over generations (E) as the primary among themselves about its interpretation at
goal, with other types of selection either higher levels of generalization, for example
eliminated (e.g. applying irrigation to whether selection in optimal or marginal
eliminate drought selection), controlled for environments leads to genotypes that are
(e.g. in experimental plot design to reduce better adapted to marginal environments
E2 ), or used to optimize selection for (Ceccarelli and Grando, 2002) (see
E (e.g. roguing off-types) (Cleveland and Chapter 2). This variation in scientists
Soleri, 2007). interpretations suggests that, if farmers
Figure 14.1 focuses on selection under do in fact think in terms of this basic
farmer conditions. Natural selection is not biological model, it would be a valuable
influenced by farmers, in contrast with comparator, facilitating understanding of
human or artificial selection. Artificial selec- variations (differences in higher levels of its
tion is both indirect, a result of the envi- interpretation) within and between FK and
ronments created by farmers and plant PBK on equal grounds.
breeders, e.g. in their fields and store rooms, We use the two parts of the model on
and direct, a result of human selection of which plant breeding is based (Cleveland,
planting material. Direct artificial selection Soleri, and Smith, 2000), as presented in
can be both unconscious or unintentional standard texts (e.g. Falconer and Mackay,
(based on implicit or correlated criteria), 1996: 189; Simmonds and Smartt,
when no conscious decision is made about 1999: 193).
the trait selected for, and conscious or inten- 1. Variation in population phenotype
tional (based on explicit criteria), the result (observable characteristics) ( P2 ) on
of decisions to select for certain traits. which choice (discrimination between
different groups of plants) and selection
14.1.3 A biological model to compare (discrimination among individual plants
farmer and plant breeder knowledge within a group) are based is deter-
and practice mined by genetic variation ( G2 ), envi-
Many plant breeders and other outsiders ronmental variation ( E2 ), and variation
who work with farmers make the mistake of in genotype (genetic constitution)-by-
assuming that western scientific knowledge environment (GE) interaction ( GE2 ),
and practice is always more accurate and thus P2 G2 E2 GE
2
.
better than that of farmers. To have a 2. Response to selection (R) for a trait is
way of comparing plant breeder knowledge the difference between the mean of the
(PBK) and farmer knowledge (FK), a whole population from which the parents
neutral comparator that can function as were selected and the mean in the next
328 Plant breeding and farmer participation
generation produced by planting those Soleri, 2002b). When FK differs from that
selected seeds under the same conditions. presumed by plant breeders interpretation
R is the product of two factors, h2 and of the model, we should try to understand
S (R = h2S), where S is the selection the difference in terms of the specific geno-
differential, the difference between the types and environments each works with,
mean of the selected parental group and as well as other factors in their experience.
the mean of the entire original population
(Allard, 1999: 101102; Falconer and 14.1.4 Methods for understanding
Mackay, 1996: 189; Simmonds and Smartt, farmers knowledge and practice
1999: 193). Narrow sense heritability The best starting place for collaboration may
(h2) (that part of P2 that can be passed be simple interviews with a representative
directly from parent to progeny, the random sample of households. Such
additive variance, A2 ) = A2 / P2 . Thus, interviews can provide insights critical for
artificial phenotypic selection per se is collaboration. There are many resources
a process of identifying the individuals available describing how to conduct such
with specific phenotypic traits within a interviews (e.g. Cleveland and Soleri, 1991)
population that will contribute genetic and analyse them (e.g. Stern et al., 2004).
material to the next generation, and is The key requirements are that: (i) the
distinct from the heritability of those sample is representative of the human
phenotypic traits (see Section 14.5). population with which you are working,
In our use of the basic biological model, possibly requiring a stratified sampling
we make several assumptions. (1) It models approach, based for example on gender of
empirically observable patterns in the real farmers, household socio-economic status,
world. (2) Among both farmers and plant or dominant soil type on farms; (ii) people
breeders and other scientists, there are conducting the interviews are consistent,
some who are particularly good observers respectful, open and primarily listen
of their environments, crops and interac- to and document farmers answers and
tions between these if they occur, while comments; and (iii) questions are relevant
others are poor observers, resulting in for understanding and collaboration.
variation within groups. (3) Variation in In addition to simple questions to elicit
knowledge within and between groups can basic descriptive data (household size,
also be caused by experiences with different number working in farming, area sown to
genotypes and environments, and by dif- each crop, sources of planting seed, yields,
ferent values and pre-existing knowledge. etc.), methods such as scenarios and ranking
(4) Differences between FK or PBK and exercises may use hypothetical varieties
the model do not mean that either form to better understand farmers theoretical
of knowledge is wrong, and differences knowledge, or actual varieties they are familiar
between FK and PBK do not mean that with for insights into specific experiences and
either is inferior to the other. observations (Crossa, Bellon and Franco,
Thus, experiences under diverse circum- 2002; Soleri and Cleveland, 2005). For
stances can result in local interpretations example, a scenario using hypothetical maize
of the model, by either farmers or scien- varieties was created to better understand
tists, which can be sources of learning for the GE interaction most valued by maize
both scientists and farmers (Cleveland and farmers in a study in Mexico, Cuba and
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 329
FIGURE 14.2
Ranking exercise as presented to farmers
Four types of maize are available
:
our farmer Modern variety (MV) our farmer variety with Modern variety with properties
variety (FV) properties from other from other organisms via
organisms via transgenesis transgenesis (TGMV)
(TGFV)
Rank these from best to worst for:
$ Planting in your elds
$ ou and your family to eat
Guatemala (Soleri et al., 2005). When asked and those same varieties as backgrounds for
to choose between two maize varieties a transgene: a transgenic farmers variety
with qualitative GE response to annual and a transgenic modern variety (Figure
precipitation, 61 percent of farmers preferred 14.2). We asked farmers to rank these first
a variety with lower yield potential, mean as maize seed for sowing in their own fields,
and yield variation (stable) to a variety and then again as maize grain for their
with higher yield potential, mean and yield family to eat. The FV and MV represented
variation (responsive). The answer varied, two seed systems (informal vs formal,
with farmers from more difficult growing respectively) and had different agronomic,
environments preferring the stable variety at storage and culinary characteristics with
a significantly greater frequency than those which farmers were already familiar.
in more favourable growing environments. Farmers had no previous experience with
A similar scenario was created to investigate transgenic crop varieties (TGVs). Providing
farmers attitudes towards some of the these four choices allowed us to distinguish
possible consequences of pesticidal farmers preferences for varieties or genetic
transgenes in their maize varieties and the backgrounds (FV vs MV) from their
evolution of resistance in the pests that it preference for a genetic technology (TGV
controlled. Some of these consequences were vs non-TGV), an important distinction that
reliance on the formal seed system, a higher is either overlooked or confounded in most
seed price and initially high but declining research with farmers. TGVs were described
yields over time as pest populations evolved neutrally to farmers and they were given a
resistance. The hypothetical transgenic positive example of TGVs with the potential
variety was not identified as being transgenic to decrease pest damage.
when the scenario was presented to farmers.
Of those interviewed (n = 334), 70 percent 14.2 THE CONTEXT: INDUSTRIAL AND
chose a lower yielding but more stable and THIRD WORLD AGRICULTURE
locally available variety (Soleri et al., 2005). Industrial and Third World agriculture are
Similarly, an exercise asked those farmers to different in important ways in terms of seed
rank four types of maize: their own FV, a and food systems, growing environments
conventional MV they were familiar with, and crop genotypes. They are also similar in
330 Plant breeding and farmer participation
FIGURE 14.3
Components of agricultural systems in traditionally-based
small-scale and industrial large-scale agriculture
Improvement
Conservation
(by formal plant
(by scientists ex
breeding)
situ in gene
banks)
Consumption Multiplication
(by consumers (by seed institutes
distant from or companies)
production)
Production
(by specialized,
large-scale
farms)
terms of the basic principles and processes holds sell some portion of their production
governing these variables and the interactions in the market, but they are incompletely
among them, including the outcome of integrated into these markets (Ellis, 1993).
choice and selection. Better understanding Farmers production knowledge combines
of these differences and similarities, and understanding based on theory and empiri-
their relationship to differences and cal observation with values about the social
similarities between farmers and breeders and cultural significance of farming, often
goals, knowledge and practices, can help to focused on FVs (Soleri et al., 2002).
further collaboration between farmers and Off-farm income is often critical for
plant breeders. households overall survival strategy, and
may reduce the importance of on-farm
14.2.1 Seed and food systems production. Migration of household mem-
In industrial agriculture, food production, bers, for example, may lead to labour short-
food consumption, crop improvement, seed age (Narayanan and Gulati, 2002) and to
multiplication and crop genetic resources reduced time and other resources devoted
conservation are specialized, physically to seed selection, conservation of crop
and structurally separated, and farming is genetic diversity or production, and even-
often considered to be primarily a busi- tually to loss of the knowledge on which
ness (Lyson, 2002) (Figure 14.3a). In SSTW they depend (for an example in central
agriculture these functions are combined Mexico, see Fitting, 2006).
within the farm household and community
(Figure 14.3b) (Soleri and Cleveland, 2004), Consumption
as described below. The differences due to Farm families rely on their own food
separation vs integration of these critical production for a significant proportion of
functions in seed and food systems have their food, and FVs are valued for traits that
important implications for decisions about contribute to storage, food preparation,
the best ways for farmers and breeders to taste, colour, texture and specific uses (e.g.
improve yields and quality traits, and to maize varieties grown for husks used in
minimize farmers risk. tamale production) (Soleri and Cleveland,
2001), or sticky rice FVs used for traditional
Production foods in southern China (Zhu et al., 2003).
SSTW agriculture is essential for feeding a These specialized uses mean some FVs have
significant proportion of the world popula- high market values.
tion now, and will probably remain so in
the future, even with production increases Improvement
in large-scale, industrial agriculture (Hazell Cultivation in new locations, farmers
et al., 2007; Heisey and Edmeades, 1999). changing selection criteria and growing
As mentioned earlier, over 2 billion people environments were responsible for the
live on almost half a billion small-scale tremendous increase in intraspecific crop
farms (<2 ha) in the Third World, including diversity via mass selection following
half of the worlds undernourished people domestication (Harlan, 1992; Matsuoka et
and the majority of those living in absolute al., 2002) (see Chapter 1, this volume), and
poverty (Nagayets, 2005). Food production all of these continue today. It has been best
relies on household labour, and most house- documented at a local level in vegetatively
332 Plant breeding and farmer participation
propagated crops (Elias et al., 2001), but using or selecting and saving FV seed each
also in predominantly self-pollinated crops year for planting (Louette and Smale, 2000;
such as rice (Dennis, 1987; Richards, 1986). Soleri, Smith and Cleveland, 2000). This
For cross-pollinating crops like maize, conservation is dynamic in that populations
farmers may not be interested in chang- are exposed to changing natural and artificial
ing quantitative phenotypic traits of their selection pressures, often creating locally
varieties through selection, but rather in distinct and adapted populations through
maintaining qualitative traits of interest indirect selection.
(Pressoir and Berthaud, 2004b), and can do Because food production and consump-
so successfully even in the presence of high tion and crop improvement, seed multi-
rates of gene flow at other loci (Louette, plication and conservation are all carried
Charrier and Berthaud, 1997; Pressoir and out within the same crop population, that
Berthaud, 2004a). Quantitative improve- population will not be optimized for any
ment in such species may more often be one function per se as it might be in indus-
sought through choosing new varieties or trialized systems. For example, the value of
populations, as discussed below. FV genetic diversity in its local conserva-
tion role may in some way be in conflict
Seed multiplication with the genetic composition optimal for
Farmers do not usually distinguish seed its role as an improved population (Soleri
multiplication from food production, and Smith, 1995). In this sense, farmers
although sometimes they plant separate crop populations are similar to semi-natu-
seed multiplication plots, as do some ral plant populations that
rice farmers in Sierra Leone (Richards, ...approach complex equilibria in
1986: 138144). Farmers often save a high which overall fitness is as high as the
proportion of their seed from their own varied demands of differing sites and
harvests, but often also obtain seed through seasons, complex genetic control and
informal seed systems (Ndjeunga, 2002), the long term demands of adaptability
and frequently experiment with new seed allow.
(Louette, Charrier and Berthaud, 1997), (Simmonds and Smartt, 1999: 91)
including planting seed obtained as grain For this reason, and because of the value
(Soleri et al., 2005). The result is extensive of this farming for production and dynamic
gene flow via seed and other propagules, conservation, both in situ conservation by
as well as by pollen flow, creating seed farmers and ex situ conservation in formal
systems that are predominantly local and gene banks are necessary and complemen-
genetically open (Berthaud et al., 2001; tary. However, for conservation to play a
Pressoir and Berthaud, 2004a; vom Brocke useful role, interaction between farmers
et al., 2003a). and scientists is required, e.g. to ensure
that the selection environments in ex situ
Conservation conservation do not result in evolution that
Farmers conserve crop genetic diversity makes the population unsuitable for farm-
of FVs in situ in their fields and storage ers in the event that they require renewal
containers (Qualset et al., 1997). Most of their seed from outside their communi-
in situ conservation is done indirectly ties (Soleri and Smith, 1995). In a similar
perhaps unintentionallyas a result of way, collaboration between farmers and
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 333
plant breeders needs to balance the goals with farmers, this environmental stress and
of the breeder, which will tend to focus on variation mean that selection for improved
improving specific traits, with the other performance in farmers environments
functions of the food and seed system. needs to take place in those environments,
and requires re-thinking some of the
14.2.2 Growing environments and assumptions of conventional plant breeding
genotypes (Ceccarelli and Grando, 2002). However,
Growing environments and crop genotypes many plant breeders, especially those with
of Third World farmers differ in impor- little experience with farmers growing
tant ways from those with which most environments, believe that as a general
plant breeders and agronomists in industrial principle selection should be done in
countries are familiar. Farms often consist optimal environments because there are
of a number of small, scattered fields with spillover effects to marginal environments
marginal growing environments, i.e. rela- (for discussion, see Atlin et al., 2000;
tively high levels of stress and of temporal Rajaram and Ceccarelli, 1998). Thus,
and spatial variability. For example, while while plant breeders agree on the basic
the average size of maize grain farms in the principles of selection, they can disagree
United States of America in 2003 was 79.2 vehemently about how those principals
ha (USDA NASS, 2004), in the southern should be applied to farmers environments
Mexican state of Oaxaca over 76 percent of (Ceccarelli and Grando, 2002; Cleveland,
maize farms were smaller than 5 ha in 1995 2001). One source of such disagreements
(INEGI, 2001), and in one of the communi- may be differing interpretations of empirical
ties in the Central Valleys of Oaxaca where observations and theories thought to
we have worked, the average farm size is 3.7 underlie them (see Section 14.2.3).
ha and the average maize field size is 0.8 ha Farmers often continue to use locally
(Soleri, 1999; Soleri, Cleveland and Aragn selected FVs, even when MVs produced
Cuevas, 2003). In that same Oaxacan com- by the formal plant improvement and
munity, coefficients of variation of maize seed multiplication systems are available,
yields calculated using triangulation of because FVs may be better adapted to
farmer estimates were very high, averaging marginal growing environments, and
44 percent (Soleri et al., n.d.). Indeed it has because MVs may be agronomically,
been estimated that maize farmers in that culinarily and economically inappropriate
area of Oaxaca experience production fail- (Ceccarelli et al., 1994; Evans, 1993;
ure one year in four due to drought (Dilley, Heisey and Edmeades, 1999). Farmers
1997). In addition to high levels of environ- value FVs for agronomic traits, such as
mental variability, other factors contribute drought resistance, pest resistance and
to high levels of yield variability and pro- photoperiod sensitivity, as well as for traits
duction risk. SSTW farmers typically use contributing to storage, food preparation,
low levels of external inputs, and have lim- taste, market value and appearance (Smale,
ited access to government programmes and 2002). FVs include landraces, traditional
markets, and limited influence on the poli- varieties selected by farmers, MVs adapted
cies affecting them (Ellis, 1993; Hardaker, to farmers environments by farmer and
Huirne and Anderson, 1997). natural selection, and progeny from crosses
For many plant breeders who work between landraces and MVs (sometimes
334 Plant breeding and farmer participation
the best response for all environments, Therefore, especially for those biophysi-
including ones with high E2 . However, cal aspects of genotypes and environments
empirical testing has shown this not to be that are less well understood in terms of
true in many cases (e.g. Ceccarelli, 1996; plant breeding theory, PBK may more
Ceccarelli et al., 1994, 2003; Comadran et likely to be based on each persons or insti-
al., 2008); two reasons are that, first, the tutions specific experiences with the par-
genes responsible for a quantitative trait ticular environments and crop genotypes
such as yield may be different in different they work with, and thus may be less gen-
environments (e.g. Atlin and Frey, 1990; eralizable, and more apt to be influenced by
Atlin, McRae and Lu, 2000; Venuprasad, pre-existing knowledge (including values)
Lafitte and Atlin, 2007), and, second, h2 of specific to the plant breeders social envi-
some important traits may not be entirely ronment. This means that disagreements
obscured by E2 (Al-Yassin et al., 2005) between farmers and plant breeders, and
(see Chapter 2, this volume). Working among plant breeders, could arise even
with farmers often requires that breeders though fundamental genetic and statistical
test the validity of those interpretations of principles remain constant across a range of
theory that form the basis of conventional contexts, because the art of plant breeding
plant breeding. This includes comparing is more tied to specific individuals or envi-
the genotypes and environments and goals ronments (Ceccarelli and Grando, 2002;
for improvement, and testing the assump- Soleri and Cleveland, 2001).
tions (biological, environmental, economic,
sociocultural) on which they are based, 14.2.4 Farmer knowledge
and adjusting interpretations of theory, A lack of empirical research and theoretical
and hence methods used (Ceccarelli and analysis has contributed to using overly sim-
Grando, 2002). plified definitions of FK (and often of PBK
For this reason, farmerbreeder col- as well), and the common failure to test the
laboration may often benefit from making many assumptions underlying these defini-
a clear distinction between (a) fundamental tions (Cleveland, 2006; Sillitoe, 1998). We
biological theory, (b) interpretations of can very roughly divide current views of FK
fundamental theory, and (c) methods and into two categories: there are those that see
practice, with c possibly very different FK as fundamentally different from PBK,
depending on whether it is based on a or and those that see it as fundamentally similar.
b, or on different versions of b. Many These views also form the basis of particu-
of the disagreements about plant breeding lar advocacy perspectives; generally neither
methods for participatory plant breeding considers the theoretical content of FK.
(PPB) may grow out of disagreements In the first category, definitions of FK
about differences in the interpretation of emphasize that it is primarily value-based,
fundamental biological theory, and disa- comprising intuition and skill, socially con-
greements about these interpretations may structed, and based on the local social
in turn be based on the belief of proponents and environmental contexts and culture.
that their interpretations of fundamental According to this perspective, farmer and
theory are not based on their unique expe- PBK are seen as fundamentally different,
riences and assumptions, but rather are part and attempts to explain FK in scientific
of fundamental theory. terms impede true appreciation of FK.
336 Plant breeding and farmer participation
who carefully selected as parents tion) and other linked loci. The presence
those individuals with the ability of this sweep distinguishes the sticky rice
to live and reproduce in the local favoured by upland northeast Asian peo-
environment, as well as with superior ples from the non-glutinous rice varieties
usefulness to local consumers. used by other Asian groups, and presum-
(Allard, 1999) ably would be among their fundamental
In contrast, Simmonds and Smartt (1999: 13) choice criteria, perhaps as an adaptation for
emphasize indirect selection: the art of eating with chopsticks (Olsen et al., 2006).
cultivation is perhaps the peasants most Increasing evidence for a number
potent contribution. of crops suggests that domestication
Similar to studies based on archaeologi- could have occurred over short periods
cal data, results of molecular analyses sup- relative to the ~12 000 years that crop
port the hypothesis that farmers selection plants have been cultivated (Gepts, 2004).
has been successful in achieving evolution- Domestication syndrome traits often
ary change for traits in the domestication appear to be determined by a small number
syndrome that might be indirectly or unin- of genes with large effects, suggesting that
tentionally favoured because of agronomic domestication could proceed relatively
superiority (see Chapter 1, this volume). rapidly. For example, Paterson et al. (1995)
There is also evidence that farmer selection found a small number of quantitative trait
has been a powerful force for evolutionary loci (QTLs) coding for the domestication
change based on other preferences as well. syndrome traits of seed size, photoperiod
For example, three major genes involved in sensitivity of flowering, and brittle rachis
starch metabolism in maize were found to in taxonomically distinct cereals with
have unusually low genetic diversity com- diverse centres of origin (sorghum, rice
pared with its closest wild relative (teos- and maize). In common bean (Phaseolus
inte, Zea mays subsp. parviglumis), which vulgaris L.), control of the domestication
is strong evidence of selection for specific syndrome involves genes that have a large
processing and culinary qualities impor- effect (>2530 percent) and account for a
tant for the primary manner in which substantial part of the phenotypic variation
maize has been consumed in its regions of observed (>4050 percent) (Koinange,
origin and diversity (Whitt et al., 2002). In Singh and Gepts, 1996). Simulations based
addition, three other loci contributing to on sequence variations at loci coding for
sweet maize grain phenotypes showed low biochemical or structural phenotypes in
diversity (resulting from strong selection) maize and its close and distant relatives
in only certain varieties in particular loca- have estimated that domestication
tions, evidence of further specialization could have taken from 10 (Eyre-Walker
in the non-agronomic selection pressures et al., 1998) to between 315 and 1 023
farmers have exerted on maize (Olsen et generations (Wang et al., 1999). In
al., 2006; Whitt et al., 2002). Similarly, it addition to selecting for characteristics of
appears that strong directional selection the domestication syndrome, especially
for sticky, glutinous grain quality resulted in cereals and small pulses (Harlan, 1992),
in a selective sweep affecting an area over domestication in sexually propagated crops
250 kb long that includes the locus cod- may have resulted in increased autogamy
ing for this quality (low amylase produc- and therefore homozygosity, expressed
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 339
World agriculture, are not well understood traditions or current needs (Cleveland and
(Ellstrand, 2003b; Heinemann, 2007; NRC, Murray, 1997).
2002; Snow et al., 2005). Much plant breeding has moved from
The spread of biotechnology has also the public to the private sector (Frey, 1996)
resulted in unintentional transgene flow, and thus selection criteria are increasingly
including into centres of diversity, e.g. maize vulnerable to being dominated by private
transgenes documented in Mexican FVs profit motives rather than public good
(Alvarez-Morales, 2002; Pineyro-Nelson et motives (Simmonds, 1990), which is espe-
al., 2009; Serratos-Hernndez et al., 2007). cially evident for TGVs. The major share
Such transgene flow can be difficult to pre- of agricultural biotechnology processes and
vent (NRC, 2004), the early stages of trans- products are controlled by private multina-
gene flow to FVs are extremely difficult to tional corporations with little incentive to
monitor (Cleveland et al., 2005), and the develop TGVs most appropriate for Third
effects may often be irreversible (Ellstrand, World farmers who cannot afford to pay
2003a). Potential effects of transgene flow the premium for TGV seed (CGIAR, 2006;
on FVs and farmers are both positive and World Bank, 2007: 178).
negative, and will require risk analysis and Similarly, there is increasing concentration
evaluation specifically adapted to each loca- in the seed sector, which potentially
tion crop combination within the Third reduces competition and limits the kinds
World (Cleveland and Soleri, 2005; Soleri, of crops and crop varieties produced and
Cleveland and Aragn Cuevas, 2006). made available. The largest seed companies
Transgenes can introduce novel forms of control an ever larger proportion of the
diversity into the crop populations being seed market; according to one estimate,
selected upon by farmers and plant breeders, between 1997 and 2004 the companies with
but there is no reason to expect that farmers the largest sales increased their market share
will be able to retain, discard or manipulate from 27 percent to 33 percent, and in 2004
them any differently from other genes. the top four companies owned 38 percent
of biotechnology patents (World Bank,
14.3.4 Privatization 2007: 135136).
In the early 1980s, some countries and The drive to globalize industrial-world
farmer support groups sought to do away IPRs in plants has been intensified as a result
with all intellectual property rights (IPRs) of pressure from agricultural biotechnology
in crops, establishing farmers rights to all corporations (Graff et al., 2003; Shorett,
crop genetic resources, but this move was Rabinow and Billings, 2003). This means
defeated by the United States of America that as patented TGV crops and their trans-
and other industrial nations (Fowler, 1994), genes move intentionally or unintentionally
and private rights in plants and other living around the world, so could the rights of the
organisms now dominate, with industrial companies who own them. Movement of
patents leading the way (Atkinson et al., transgenes into non-transgenic crop popu-
2003). Farmers were left with having to lations, whether producing a net benefit for
defend themselves from the advances of an the farmer or not, makes farmers vulnerable
IPR system in plants designed by industrial to IPR claims from the technology devel-
nations and corporations, a system that oper. In the United States of America and
generally does not recognize farmers many other industrialized countries, patent
342 Plant breeding and farmer participation
holders have rights to seek damages from 14.3.5 Sustainability and farmer-
farmers who end up with patented genes scientist collaboration
in their crops, even though farmers do not The search for sustainability provoked by
want them, and do not know they are there negative environmental impacts of agri-
(Janis and Kesan, 2002). The World Trade culture (Matson et al., 1997; NRC, 2002;
Organization (WTO) seeks worldwide Tilman et al., 2002), and its genetic vulner-
uniformity of laws for IPRs in plants and ability (NRC, 1991) has led to the incor-
plant DNA to facilitate global enforce- poration of more genetic variation within
ment, and many Third World countries and among varieties by the formal crop
have adopted the industrial world model improvement system (Cooper, Spillane and
(UPOV International Union for the Hodgkin, 2001; Ortiz et al., 2007). It has
Protection of New Varieties of Plants) also encouraged a re-evaluation of GE
while others have adapted their national in crops and how best to exploit this
laws to protect small-scale farmers (World for farmers conditions (Ceccarelli et al.,
Bank, 2007: 167). The spread of IPRs and 1994, 2001). The impact on farmer selec-
coupled economic control of agricultural tion will be in the greater intraspecific and
biotechnology means that Third World intravarietal diversity deployed in formally
farmers and the nation states they live in developed varieties, and greater interest in
will have a difficult time gaining mean- that system for breeding goals more simi-
ingful control of the means to intention- lar to those of farmers. Part of the interest
ally create TGVs more suited to their own in sustainability (environmental, economic
needs, if this is the path they choose. As and social) has led to collaboration between
a result, most organizations promoting farmers and scientists.
TGVs more suited to Third World farmers Participatory or collaborative plant
are advocating public-private partnerships breeding is attempting to reverse the
(CGIAR, 2006; FAO, 2004; World Bank, separation of farmers and scientists and
2007), yet it is not clear how farmers improve the outcomes of choice and
rights will fare in this collaboration, and selection in farmers terms (Cleveland
they are not being rigorously addressed. and Soleri, 2002a; PRGA, 2004; McGuire,
While most corporations deny they Manicad and Sperling, 2003; Weltzien et
would enforce their IPRs against Third al., 2003). To that end, the next sections
World farmers, there are no guarantees. focus on understanding farmers choice
In addition to transgenes, control of local and selection, and thereby enabling farmers
farmers crop genetic resources, and the and scientists to work more closely and
traditional names and other cultural prop- productively in improving the crops they
erty that go with them through industrial grow and depend upon.
IPRs, can legally and economically pre-
vent local people themselves from reaping 14.4 CHOICE OF GERMPLASM
potential benefits in a global marketplace It is important for plant breeders to under-
increasingly interested in traditional crops stand how and why farmers choose varie-
and foods (Soleri et al., 1994). There are ties of their crops, because farmer choice
already cases of this, some of which are will ultimately determine whether a new or
being challenged (Pallotini et al., 2004). improved variety will be useful. In this sec-
tion we consider choice based on perceived
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 343
risk and yield stability, and on other fac- farmers fields, or even within a field (Soleri
tors, including quality traits. et al., 2002). Variation in time is also large:
Just as there are factors favouring the in the semi-arid tropics, seasonal and
inclusion of more than one variety in a annual rainfall is highly variable, and even
farmers crop repertoire, there also are fac- in years with adequate total rainfall, rains
tors limiting the number of varieties chosen. may arrive late, end too early, stop for a
These include farmers resources, growing period or be too heavy during flowering
environments and crop reproductive biology, or harvesting. Therefore farmers may often
among other possible factors. Additionally, grow two or more varieties of many crops,
if crop varietal diversity is maintained at a each with distinct agronomic characteristics
community instead of household level, then presumably as a measure of insurance
farmers may not feel the need to maintain against vagaries of the weather, diseases, or
some varieties themselves each year, even pests (Doggett, 1988).
though they consider those varieties to be Understanding farmers choice can
part of their varietal repertoire and intend to provide valuable insights for scientific plant
grow them in the near future (for the case of breeders. In response to climate change
rice, see Dennis, 1987). in the form of the southern movement of
isohyets, policy-makers in Mali argue that
14.4.1 Varietal choice, yield stability improved short-cycle varieties are a critical
and risk part of stabilizing the countrys volatile
In much of the past plant breeding for cereal production (Dembl and Staatz,
SSTW farmers, it was assumed that high 2000). One result is that both sorghum
yielding varieties selected in more optimal breeders and farmers in southern and central
environments would outyield FVs in Mali look north for shorter cycle varieties.
farmers environments (Ceccarelli et al., Interviews in four villages in the Upper Niger
1994; Ceccarelli, Grando and Booth, 1996). River valley zone of Mali found the most
If farmers did not adopt these varieties it common reason for adoption of the three
was assumed that they were ignorant of most popular sorghum varieties was early
how to improve their growing environments maturity (Adesina, 1992). However, since
(Aquino, 1998), or if they could not afford in good rainfall years long-cycle varieties
to do so, it was assumed that they should generally have higher yields (Adesina, 1992)
get out of farming. Consideration of risk and are rated higher for quality (Ingram,
provides a different understanding of Roncoli and Kirshen, 2002), farmers do not
farmers varietal choices and other practices. give these up entirely. Their choices thus
In the conventional economic model, a risk- increase the number of varieties in their
neutral farmer would only grow the one repertoires, although the net impact on
variety that gives the highest profits per unit genetic diversity has not been investigated.
area (Smale, 2002). However, many small- Another study in Mali of farmers choices
scale farmers in marginal environments are among their traditional sorghum varieties
risk averse (Anderson and Dillon, 1992; Soleri in terms of one or more than one variety,
et al., n.d., 2008), and spatial environmental and short-cycle or long-cycle varieties,
variation increases the likelihood of cross- found that farmers make these choices in
overs in varietal performance (qualitative an effort to optimize outputs in the face of
GE; see Section 14.5.1, below) between variation in the growing environment and in
344 Plant breeding and farmer participation
FIGURE 14.4
Sample scenario used to elicit farmers knowledge of heritability
Typical local eld, variable and high stress Hypothetical uniform and optimal eld
For each eld we asked, What length will the ears that grow in this eld
be? The same as or different from the ears from which the seed was taken?
This scenario, used with farmers in Cuba and Mexico, is about a maize trait with low average heritability (ear length). We
asked farmers what the ear length of the next generation would be in a typical relatively variable, stressful field and a
hypothetical uniform, optimal field if seed from long ears only was planted. Similar scenarios were used for low and high
heritability traits for crops in the five sites in study (see Table 14.1).
Copyright 2008, D. Soleri, D.A. Cleveland, used with permission.
TABLE 14.1
Understanding farmers perceptions of heritability
Location, crop Null hypothesis #1: Null hypothesis #2:
For traits with relatively low or those with relatively In scenarios depicting a typical
high h2, distribution of farmers responses is the environment, distribution of farmers
same whether scenarios depict typical or optimal responses is the same for traits with
environments, i.e. farmers do not see a contribution relatively low or those with relatively
of environment (Env) to phenotype high h2, i.e. farmers do not see a
difference in h2 between traits
(a) Low h2 trait across (b) High h2 trait across Low v high h2 traits in typical, variable
typical and optimal Envs typical and optimal Envs Env
Cuba, maize Ear length* Husk colour Ear length v. husk colour*
Mexico, maize Ear length* Tassel colour Ear length v. tassel colour*
Mali, sorghum Panicle weight* Glume colour* Panicle weight v. glume colour*
Syria, barley Plant height* Seed colour Plant height v. seed colour*
Nepal, rice Plant height* Seed colour Plant height v. seed colour*
* Hypothesis rejected, Fishers exact test, P<0.05
Based on Soleri et al., n.d.
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 347
levels: between locations, between fields evolution, breeders looked to farmers for
in one location, and between places in one their applied knowledge and practice that
field (Soleri et al., n.d., 2002). The results produced practical results in the form of
indicated that farmers (n = 208) perceive new varieties, as in the early commercial
inter- (57 percent) and intra- (30 percent) development of maize in the United States of
location GE for their major crop, though America (Wallace and Brown, 1988: 8790).
far fewer at the latter level. GE within With the increased importance of formal
a field (18 percent) was noted mostly, science in plant breeding compared with
though not exclusively, by those growing empirical heuristics, and later as plant
self-pollinated crops, and especially those breeding moved from the public to the
working at a small scale with intimate private sector (Kloppenburg, 1988), plant
knowledge of within-field soil and breeders began to eliminate farmers from
moisture variations (e.g. rice farmers in their work (e.g. Schneider, 2002). Plant
western Nepal). Similarly, 37 percent of breeders and farmers practices and concepts
farmers responded that a qualitative GE subsequently developed independently of
interaction could occur in their crop due each other, effectively separating the formal
to temporal environmental variation in and informal systems of crop improvement
the form of annual precipitation. In the and seed multiplication, with plant breeders
presence of qualitative GE, perceptions coming to dominate: a trend that has been
of the best genetic strategy may differ at least locally apparent for 200 years
and be informative about the needs of (Simmonds and Smartt, 1999: 13). Plant
particular groups or regions. As mentioned breeders focused on modern varieties widely
in Section 14.4.1 above, farmers tended to adapted to more optimal, more intensively
favour yield stability over high yield when managed environments, while many
choosing among varieties in the face of traditionally-based farmers in relatively
qualitative temporal GE. This choice was marginal environments continued to focus
significantly more frequent among farmers on traditional, specifically adapted varieties
in more difficult environments compared for their diverse, more marginal growing
with more favourable environments. environments (Ceccarelli and Grando, 2002;
Cleveland, 2001). When contemporary plant
14.5.2 Farmers selection goals breeders involve farmers in their work, it
If plant breeders misunderstand what has generally been limited to the stage of
farmers are and are not attempting to evaluating the plant breeders populations
accomplish with their selection practices, or varieties in the field (Duvick, 2002), i.e.
it can limit the potential for meaningful choosing among different populations or
collaboration and lead to inappropriate varieties, not selecting among different plants
investments of scarce time and resources. to genetically change existing populations
Such misunderstandings have grown out or varieties.
of the historical process of separation Today, many modern plant breeders
of farmers and plant breeders work consider themselves to be applied
(Cleveland and Soleri, 2007). evolutionists, whose goal is to develop
Just as early evolutionary biologists looked plant varieties better adapted to improved
to breeders for empirical demonstration growing environments, with adaptation
of results of selection that illuminated measured primarily as increased yield
348 Plant breeding and farmer participation
FIGURE 14.5
Phenotypic selection classified according to outcome of selection
(Allard, 1999: 49). Their emphasis in the ensuing proliferation of crop varieties.
selection is on achieving directional, multi- It also means formal plant breeders tend
generational, micro-evolutionary change. to judge the efficacy of farmer seed saving
This makes sense given the organization of in terms of applied evolution, i.e. the same
industrial agricultural systems (see Section criteria they apply to their own work, and
14.2.1 above, and Figure 14.3). It also means assume that farmers use these criteria as
that plant breeders often view farmers well.
selection of seeds or other propagules for In the following sections we describe
planting as a form of mass selection for phenotypic selection by farmers organized
heritable traits, the process that is assumed in terms of possible outcomes: longer-
to account for crop domestication and for term (multi-generational) genetic change
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 349
mass selection for low heritability traits in result in new varieties (i.e. the goal is E)
cross-pollinating species with the goal of (Evans, 1993: 313314). Like plant breed-
making directional change in their varieties ers (Cooper, Spillane and Hodgkin, 2001),
(Christinck, 2002: 126; Vom Brocke et farmers also encourage gene flow under
al., 2002). This research also documented some conditions, for example mixing seed
farmers intentional introgression of from different sources, planting different
modern with traditional varieties of pearl populations contiguously or in same plot,
millet, and subsequent selection, resulting and by making crosses, as a way of increas-
in increased genetic variation and long- ing the variation on which to select.
term directional change (E) in selected Farmers can be successful in maintaining
traits, such as growing period (Christinck, varietal ideotypes through direct, inten-
2002: 123; vom Brocke et al., 2003a). tional selection for key traits, especially for
However, although it is clear that farmers highly heritable phenotypic traits, like those
can understand the principle of phenotypic that define a variety. This type of selection
selection and use it to achieve goals of is probably most important for cross-polli-
evolutionary change with different crops, nated crops, such as pearl millet and maize,
this may not always, or even usually, be as discussed below, since it is much more
their goal, or the result. difficult to maintain populations in these
compared with clonally propagated and
14.5.4 Selection for genetic response, self-pollinated crops. In eastern Rajasthan,
but not evolution India, amplified fragment length polymor-
Farmers also select with the goal of phism (AFLP) analysis showed that farmers
eliminating changes in phenotypic traits maintained the ideotypes of distinct intro-
resulting from gene flow or natural or duced pearl millet FVs, even though they
indirect phenotypic selection, i.e. to achieve have the same name as local FVs, via inten-
R but not E. Best documented are farmers tional selection of panicles for their unique
attempts to maintain varietal ideotypes based phenotypes (vom Brocke et al., 2003b). In
on quantitative or qualitative phenotypic contrast, farmers in Jalisco, Mexico, regu-
traits over time in the face of gene flow larly mix maize varieties together by clas-
(Berthaud et al., 2001). Plant breeders can sifying seed obtained from diverse sources
control unwanted gene flow much more as the same variety based on ear or kernel
effectively in their experimental plots than morphology and colour, which, together
farmers can in their fields, and in industrial with planting patterns, leads to a 12 per-
agriculture farmers often buy new seed cent level of gene flow between maize
every year, especially for cross-pollinated varieties during one crop cycle (Louette,
crops like maize, eliminating most concerns Charrier and Berthaud, 1997). A control-
regarding gene flow. led experiment found that, compared with
This type of farmer selection to elim- random selection, farmer selection dimin-
inate changes may contrast with main- ished the impact of gene flow on one FV
tenance (stabilizing) selection by plant from contrasting FVs for key varietal traits
breeders, which usually has the goal of (kernel rows per ear, kernel width and
maintaining yield in the face of changing kernel colour), but did not have any effect
environments by incorporating new alleles on allelic frequencies at 9 polymorphic loci
or changing allele frequencies, and may coding for traits invisible or unimportant to
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 351
farmers (Louette and Smale, 2000). Farmers or other propagules for cross-pollinated
stated that they were not interested in (e.g. in maize; Soleri and Smith, 2002), self-
changing their varieties, but in maintain- pollinated (e.g. in barley; Ceccarelli et al.,
ing varietal ideotypes, and appeared to be 2000) and vegetatively propagated crops
achieving their goal. Research in Oaxaca, (e.g. in potato; Zimmerer, 1996). Selection
Mexico, using microsatellite data support- with this goal is also conducted as part of
ed this finding in terms of the results of MV seed multiplication (Simmonds and
farmer selection, although farmers goals Smartt, 1999: 215). Plant breeders may also
were not investigated. Extensive gene flow carry out this type of selection, for example
and little molecular genetic structure was by removing small seed, but they do this to
observed, but the maintenance of signifi- decrease the contribution of E2 to P2 , and
cantly different maize populations based so increase heritability with the goal of E.
on morphophenological traits of interest to Research on non-heritable phenotypic
farmers persisted (Pressoir and Berthaud, differences shows these can have important
2004b). intra-generational effects in terms of ecolo-
A study in Chiapas, Mexico, found gy and agronomy. Even in species with high
that cultural diversity, as measured by heritability for seed polymorphisms, envi-
ethnolinguistic groups, was not reflected ronment may be an important determinant
in maize diversity as measured by isozyme of seed size and shape, and seed polymor-
variation, but was reflected in some phism can be a significant determinant of
morphological traits (Perales, Benz and differential survival via influence on survi-
Brush, 2005). The differences observed may vorship and adult plant size (Baskin and
have been due to unidentified culturally- Baskin, 2001: 208214). In maize, for exam-
based networks or practices that structured ple, larger seed size was found to provide
these maize populations based on farmer significant advantages in the early stages of
selection for a few critical traits against a plant growth (from germination until stem
background of ongoing gene flow (Perales, elongation) (Bockstaller and Girardin,
Benz and Brush, 2005), as was found in 1994), and was correlated with better early
the study in the central valleys of Oaxaca, vigour, greater leaf area throughout the life
Mexico, (Pressoir and Berthaud, 2004b), cycle and more rapid development from
although neither study investigated farmer time of emergence to flowering (Pommel,
goals in detail. 1990; Revilla et al., 1999).
When the goal of selection is intra-
14.5.5 Selection for intra-generation generational phenotypic differentiation,
phenotypic difference the result may not be genetic response or
Although farmers are capable of pheno- evolution, especially for low-heritability
typic selection that is effective in achieving traits in cross-pollinated crops. This
goals of evolution and genetic response, hypothesis was supported by results of
perhaps the most common goal of farmer maize seed selection exercises with farmers
selection is not genetic, but solely pheno- in two communities in Oaxaca, Mexico.
typic, because most of the time a farmers The exercises were done with maize ears
primary goal in selecting seed is to obtain post-harvest, which is the way these farmers
good planting material. This often means and most others in Mexico select maize
selection for large, clean, disease-free seeds seed. Their selections resulted in high S
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 353
TABLE 14.2
Farmers expectations for response to selection for their primary selection criterion in the major crop
they grow
Country, crop, trait (n) Question A. Farmers responding that Question B. For farmers responding
response to intentional selection for IS10>RS10 to Question A, those stating that
10 cycles > random selection for 10 response to intentional selection for 11
cycles (IS10>RS10) cycles > random selection for 10 cycles +
intentional for 1 cycle (IS11>RS10+IS1)
n % P n % P
Mexico, maize, ear length (59) 23 39 * 0.000000 6 26 * 0.000000
Cuba, maize, ear length (29) 27 93 0.245614 12 44 * 0.000002
Syria, barley, plant height (21) 20 95 0.499999 11 55 * 0.000614
Nepal, rice, grain yield (40) 39 98 0.499999 17 44 * 0.000000
Mali, sorghum, grain yield (40) 35 88 0.057662 23 66 * 0.000078
Total (189) 144 76 * 0.000000 69 48 * 0.000000
One sided Fishers exact test , of the null hypothesis that, similar to plant breeders, farmers would see intentional selection
as achieving a greater response compared to random selection. Calculated using SISA (http://home.clara.net/sisa/). RS =
random phenotypic selection by farmer, IS = intentional phenotypic selection by farmer.
Based on Soleri et al., n.d.
Some farmers said large seed resulted in a hypothetical scenario asking them to
higher germination, larger seedlings, early compare random with intentional selection
vigour and higher yields, although most for 10 cycles in a typical field, in populations
farmers attributed their preference for large with phenotypic variation for the trait
seed to custom. they used as major selection criterion
It is still possible that simple mass (Figure 14.6) (Table 14.2, question A). The
selection for intra-generational phenotypic null hypothesis was that farmers did not
differences could result in R or E even if differ from plant breeders, i.e. that they
these are not farmer goals. As mentioned would all consider intentional selection to
above, it is not clear what importance this be more effective than random selection for
had during domestication and subsequent improving or at least maintaining this trait.
diversification of crops, versus intentional The majority of responses corresponded to
selection for short-term change or long- the null hypothesis of no difference between
term maintenance. For example, maize farmer and plant breeder expectations that
farmers in Uganda and the United Republic intentional selection was more effective
of Tanzania, like those in Mexico, were for increasing yield, 76.2 percent (144/189),
reported to select for large, clean kernels although those who disagreed with that
from large ears, apparently because they idea were sufficient to reject the hypothesis
believed that these germinated well and statistically (P = 0.00000). Disagreement
produced high-yielding plants (Gibson et was particularly frequent among maize
al., 2005). Interestingly, this appeared to farmers, probably due to recombination in
result in decreased resistance to maize streak that cross-pollinating crop.
virus, since resistant plants had smaller ears, These results indicate that farmers who
and plants with large ears appeared to be believe there is an advantage of intentional
non-resistant escapes. over random selection, see their goal for
As part of a comparative five-country phenotypic selection as either S or R or E.
study of FK and PBK (Soleri et al., To discriminate between these possibili-
2002, 2004), farmers were presented with ties, and with the same null hypothesis as
354 Plant breeding and farmer participation
outlined above, those farmers responding and the movement to make formal plant
to the first question that intentional selec- breeding more relevant to farmers through
tion resulted in greater yield, were asked to PPB. Understanding farmers choice and
compare random selection for 10 cycles fol- selection practices, their biological results,
lowed by one cycle of intentional selection, the knowledge and goals underlying them,
with 11 consecutive cycles of intentional and the similarities and differences with
selection. Results differed significantly from plant breeders provides a means for the two
the null hypothesis (Table 14.2, question groups to work together more effectively.
B). Among these farmers, only 23.2 percent This understanding and collaboration is
(20/86) saw 11 years of intentional selection critical for supporting all of the important
as superior. These results demonstrate that functions of SSTW agriculture, including
among those farmers favouring intentional long-term global food security.
selection, only a minority see it as provid- For PPB, this means that farmers goals
ing cumulative multi-generational change for varietal choice and phenotypic selec-
(E), while the primary selection goal of tion need to be understood in the context
the other farmers who saw an advantage of a system that integrates production,
to multi-generational intentional selection consumption, improvement, multiplication
for low-heritability yield-related traits is and conservation. The biological result of
either eliminating changes between genera- phenotypic selection needs to be evalu-
tions (R) or a non-genetic advantage they ated in terms of its possible ecological
believe is fully achieved within one year effects (via S), as well as in terms of R
(S). The large number of farmers who do and E. Additionally, farmers theoretical
not consciously see an advantage to multi- knowledge of choice and selection, not just
generational intentional selection, but who, their criteria, need to be understood by
like other farmers, select for large seed from plant breeders to fully realize the potential
large, clean ears, may do so because of cus- benefits of collaboration. The value of this
tom, as did the majority of farmers in the research will be judged by its effectiveness
selection experiment described earlier. in improving the efficiency and outcomes
of collaborative breeding by scientists and
14.6 CONCLUSIONS farmers, and improvement in the well-being
Many elements of crop variety choice of those farmers and their communities.
and plant selection in the Third World
contrast substantially with industrial ACKNOWLEDGMENTS
agricultural systems, including the growing We thank the many farmers and scientists
environments, genetic resources and we have worked with in Cuba, Egypt,
organization of the agricultural system. Ghana, Guatemala, Mali, Mexico, Nepal,
The urgency of understanding farmer Pakistan, the Syrian Arab Republic and
selection will increase in the future with the United States of America for sharing
global climate changes, the continuing loss their knowledge, both about crop gen-
of genetic resources, the rapid spread of otypes and growing environments, and
transgenic crop varieties, the development their ideas about improving plant breeding
of a global system of IPR in crop genetic and crop production. Thanks to Salvatore
resources, the need to make agriculture Ceccarelli and Elcio Guimares for com-
more sustainable while feeding more people, ments on drafts of this chapter. We grateful-
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 355
Atlin, G.N., Baker, R.J., McRae, K.B. & Lu, crop landraces and the challenge to con-
X. 2000. Selection response in subdivided serve them in situ on farms. In S.B. Brush,
target regions. Crop Science, 40: 713. ed. Genes in the field: on-farm conservation
Baldoni, L., Tosti, N., Ricciolini, C., Belaj, A., of crop diversity, pp. 2948. Boca Raton,
Arcioni, S., Pannelli, G., Germana, M.A., USA, Lewis Publishers; Rome, IPGRI;
Mulas, M. & Porceddu, A. 2006. Genetic Ottawa, IDRC.
structure of wild and cultivated olives in Ceccarelli, S. 1996. Adaptation to low/high
the central Mediterranean basin. Annals of input cultivation. Euphytica, 92: 203214.
Botany, 98: 935942. Ceccarelli, S. 1997. Adaptation to low/high
Bnziger, M. & de Meyer, J. 2002. Collaborative input cultivation. In P.M.A. Tigerstedt, ed.
maize variety development for stress-prone Adaptation and plant breeding, pp. 225236.
environments in southern Africa. In D.A. Dordrecht, Netherlands, Kluwer Academic
Cleveland & D. Soleri, eds. Farmers, scien- Publishers.
tists and plant breeding: integrating knowl- Ceccarelli, S. & Grando, S. 2002. Plant breeding
edge and practice, pp. 269296. Wallingford, with farmers requires testing the assump-
UK, CABI. tions of conventional plant breeding: les-
Baskin, C.C. & Baskin, J.M. 2001. Seeds: ecol- sons from the ICARDA barley program. In
ogy, biogeography, and evolution of dor- D.A. Cleveland & D. Soleri, eds. Farmers,
mancy and germination. San Diego, USA, scientists and plant breeding: integrat-
Academic Press. ing knowledge and practice, pp. 297332.
Berg, T. 2009. Landraces and folk varieties: Wallingford, UK, CABI.
a conceptual reappraisal of terminology. Ceccarelli, S. & Grando, S. 2007. Decentralized-
Euphytica, 166: 423430. participatory plant breeding: an example
Berlin, B. 1992. Ethnobiological classification: of demand driven research. Euphytica,
principles of categorization of plants and ani- 155: 349360.
mals in traditional societies. Princeton, New Ceccarelli, S., Grando, S. & Baum, M. 2007.
Jersey, USA, Princeton University Press. Participatory plant breeding in water-limit-
Berthaud, J., Clment, J.C., Emperaire, L., ed environments. Experimental Agriculture,
Louette, D., Pinton, F., Sanou, J. & Second, 43: 411435.
S. 2001. The role of local-level gene flow in Ceccarelli, S., Grando, S. & Booth, R.H. 1996.
enhancing and maintaining genetic diver- International breeding programmes and
sity. In H.D. Cooper, C. Spillane & T. resource-poor farmers: crop improvement
Hodgkin, eds. Broadening the genetic base in difficult environments. In P. Eyzaguirre
of crop production, Chapter 4. Wallingford, & M. Iwanaga, eds. Participatory plant
UK, CABI. breeding, pp. 99116. Proceedings of a
Bockstaller, C. & Girardin, P. 1994. Effects of workshop on participatory plant breeding,
seed size on maize growth from emergence Wageningen, Netherlands, 2629 July 1995.
to silking. Maydica, 39: 213218. Rome, IPGRI.
Breton, C., Medail, F., Pinatel, C. & Berville, A. Ceccarelli, S., Erskine, W., Hamblin, J. &
2006. From olive tree to oleaster: Origin and Grando, S. 1994. Genotype by environ-
domestication of Olea europaea L. in the ment interaction and international breed-
Mediterranean basin. Cahiers Agricultures, ing programmes. Experimental Agriculture,
15: 329336. 30: 177187.
Brown, A.H.D. 1999. The genetic structure of Ceccarelli, S., Grando, S., Tutwiler, R., Bahar,
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 357
Ceccarelli, S., Grando, S., Baum, M., Stanca, Dennis, J.V. Jr. 1987. Farmer management of
A.M., Pecchioni, N., Mastrangelo, A.M., rice variety diversity in northern Thailand.
Akar, T., Al-Yassin, A.A., Benbelkacem, Dissertation, Cornell University, Ithaca,
A., Choumane, W., Ouabbou, H., Dahan, USA.
R., Bort, J., Araus, J.-L., Pswarayi, A., Dilley, M. 1997. Climatic factors affect-
Romagosa, I., Hackett, C.A. & Thomas, ing annual maize yields in the valley of
W.T.B. 2008. Mapping adaptation of bar- Oaxaca, Mexico. International Journal of
ley to droughted environments. Euphytica, Climatology, 17: 15491557.
161: 3545. Doggett, H. 1988. Sorghum. (2nd ed.) Tropical
Conway, G. 2003. From the green revolution Agriculture Series. New York, USA,
to the biotechnology revolution: food for Longman Scientific and Technical.
poor people in the 21st century. Rockefeller Duvick, D.N. 1996. Plant breeding, an evolu-
Foundation, New York, USA (http://www. tionary concept. Crop Science, 36: 539548.
unc.edu/courses/2006spring/geog/021/001/ Duvick, D.N. 2002. Theory, empiricism and
Conway.pdf; accessed 11 March 2009). intuition in professional plant breeding. In
Cooper, H.D., Spillane, C. & Hodgkin, T. 2001. D.A. Cleveland & D. Soleri, eds. Farmers,
Broadening the genetic base of crops: an scientists and plant breeding: integrat-
overview. In H.D. Cooper, C. Spillane & T. ing knowledge and practice, pp. 239267.
Hodgkin, eds. Broadening the genetic base Wallingford, UK, CABI.
of crop production, pp. 123. Wallingford, Elias, M., Penet, L., Vindry, P., McKey, D.,
UK, CABI. Panaud, O. & Robert, T. 2001. Unmanaged
Cooper, M. & Hammer, G.L. 1996. Plant adap- sexual reproduction and the dynamics of
tation and crop improvement. Wallingford, genetic diversity of a vegetatively propagat-
UK, CABI in association with IRRI and ed crop plant, cassava (Manihot esculenta
ICRISAT. Crantz), in a traditional farming system.
Crossa, J., Bellon, M.R. & Franco, J. 2002. A Molecular Ecology, 10: 18951907.
quantitative method for classifying farmers Ellis, F. 1993. Peasant economics: farm house-
using socioeconomic variables. In M.R. holds and agrarian development. (2nd ed.)
Bellon & J. Reeves, eds. Quantitative anal- Cambridge, UK, Cambridge University
ysis of data from participatory methods Press.
in plant breeding, pp. 113127. Mexico, Ellstrand, N.C. 2003a. Current knowledge of
D.F., CIMMYT (available at http://www. gene flow in plants: implications for trans-
cimmyt.org/Research/Economics/map/ gene flow. Philosophical Transactions of the
research_tools/manual/Quantitative/quan- Royal Society of London B 358: 11631170.
titativeBellon.pdf; accessed 12 Jan. 2009). Ellstrand, N.C. 2003b. Dangerous liaisons?
David, S. 2004. Farmer seed enterprises: A When cultivated plants mate with their wild
sustainable approach to seed delivery? relatives. Baltimore, USA, Johns Hopkins
Agriculture and Human Values, 21: 387397. University Press.
Dembl, N.N. & Staatz, J.M. 2000. The Evans, L.T. 1993. Crop evolution, adapta-
response of cereals traders to agricultural tion and yield. Cambridge, UK, Cambridge
market reform. In R.J. Bingen, D. Robinson University Press.
& J.M. Staatz, eds. Democracy and develop- Eyre-Walker, A., Gaut, R.L., Hilton, H., Feldman,
ment in Mali, pp. 145165. East Lansing, D.L. & Gaut, B.S. 1998. Investigation of the
USA, Michigan State University Press. bottleneck leading to the domestication of
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 359
maize. Proceedings of the National Academy term selection experiment. Plant Breeding
of Sciences of the United States of America, Reviews, 24(2): 144.
95(8): 44414446. Gibson, R.W., Lyimo, N.G., Temu, A.E.M.,
Falconer, D.S. & Mackay, T.F. 1996. Stathers, T.E., Page, W.W., Nsemwa, L.T.H.,
Introduction to quantitative genetics. 4th Acola, G. & Lamboll, R.I. 2005. Maize seed
ed. Edinburgh, UK, Prentice Hall/Pearson selection by East African smallholder farm-
Education. ers and resistance to Maize streak virus.
FAO [Food and Agriculture Organization of Annals of Applied Biology, 147: 153159.
the United Nations]. 1996. Report on the Gouse, M., Pray, C., Schimmelpfennig, D. &
State of the Worlds Plant Genetic Resources. Kirsten, J. 2006. Three seasons of subsist-
International Technical Conference on Plant ence insect-resistant maize in South Africa:
Genetic Resources. Rome. have smallholders benefited? Agbioforum,
FAO. 2004. The State of Food and Agriculture. 9: 1522.
Agricultural biotechnology: meeting the Graff, G.D., Cullen, S.E., Bradford, K.J.,
needs of the poor? Rome. Zilberman, D. & Bennett, A.B. 2003. The
FAO. 2007. FAOSTAT data (http://faostat. public-private structure of intellectual
fao.org/site/336/default.aspx; accessed 30 property ownership in agricultural biotech-
Aug. 2007). nology. Nature Biotechnology, 21: 989995.
FAO. 2009. FAOSTAT data (http://faostat. Hallauer, A.R. & Miranda, J.B. 1988.
fao.org/; accessed 1 March 2009). Quantitative genetics in maize breeding.
Fitting, E. 2006. Importing corn, exporting 2nd ed. Ames, USA, Iowa State University.
labor: The neoliberal corn regime, GMOs, Hardaker, J.B., Huirne, R.B.M. & Anderson,
and the erosion of Mexican biodiversity. J.R. 1997. Coping with risk in agriculture.
Agriculture and Human Values, 23: 1526. Wallingford, UK, CABI.
Fitzgerald, D. 1990. The business of breeding: Harlan, J.R. 1992. Crops and man. 2nd ed.
hybrid corn in Illinois, 18901940. Ithaca, Madison, Wisconsin, USA, ASA/CSSA.
USA, Cornell University Press. Hazell, P., Poulton, C., Wiggins, S. & Dorward,
Fowler, C. 1994. Unnatural Selection: A. 2007. The future of small farms for pov-
Technology, politics, and plant evolution. erty reduction and growth. 2020 IFPRI
Yverdon, Switzerland, Gordon and Breach. Discussion Paper No. 42. Washington, DC,
Frey, K.J. 1996. National plant breeding study IFPRI.
- 1: human and financial resources devoted Heinemann, J.A. 2007. A typology of the
to plant breeding research and development effects of (trans)gene flow on the con-
in the United States in 1994. Iowa State servation and sustainable use of genetic
University, Iowa Agriculture and Home resources. CGRFA Background Study
Economics Experiment Station, Ames, USA. Paper, No. 35 Rev.1. 100 p. FAO, Rome
Friis-Hansen, E. & Sthapit, B. 2000. Concepts (available at ftp://ftp.fao.org/ag/cgrfa/bsp/
and rationale of participatory approaches bsp35r1e.pdf; accessed 12 Jan. 2009).
to conservation and use of plant genetic Heisey, P.W. & Edmeades, G.O. 1999. Part 1.
resources. In E. Friis-Hansen & B. Sthapit, Maize production in drought-stressed envi-
eds. Participatory approaches to the conser- ronments: technical options and research
vation and use of plant genetic resources, resource allocation. In CIMMYT, ed. World
pp. 1619. Rome, IPGRI. maize facts and trends 1997/98, pp. 136.
Gepts, P. 2004. Crop domestication as a long- Mexico, D.F., CIMMYT.
360 Plant breeding and farmer participation
Huang, J., Hu, R., Pray, C., Qiao, F. & Rozelle, -%20English.pdf; accessed 24 February
S. 2003. Biotechnology as an alternative 2009).
to chemical pesticides: a case study of Bt Janis, M.D. & Kesan, J.P. 2002. Intellectual
cotton in China. Agricultural Economics, property protection for plant innovation:
29: 5567. unresolved issues after J.E.M. v. Pioneer.
Huang, J.K., Hu, R.F., Rozelle, S. & Pray, C. Nature Biotechnology, 20: 11611164.
2005. Debate over a GM rice trial in China Jarvis, D., Meyer, L., Klemick, H., Guarino,
- Response. Science, 310: 232233. L., Smale, M., Brown, A.H.D., Sadiki, M.,
Hughes, C.E., Govindarajulu, R., Robertson, Sthapit, B. & Hodgkin, T. 2000. Social,
A., Filer, D.L., Harris, S.A. & Bailey, C.D. cultural and economic factors and crop
2007. Serendipitous backyard hybridization genetic diversity. In D. Jarvis, L. Meyer, H.
and the origin of crops. Proceedings of the Klemick, L. Guarino, M. Smale, A.H.D.
National Academy of Sciences of the United Brown, M. Sadiki, B. Sthapit & T. Hodgkin,
States of America, 104: 1438914394. eds. A training guide for in situ conserva-
INEGI. 2001. (CD edition.) Sistema para la tion. On-farm version 1, pp. 1127. Rome,
consulta del annuario estadstico. INEGI, IPGRI.
Aguascalientes, Mexico. Joshi, A. & Witcombe, J.R. 1996. Farmer partic-
Ingram, K., Roncoli, M. & Kirshen, P. 2002. ipatory crop improvement. II. Participatory
Opportunities and constraints for farmers varietal selection, a case study in India.
of west Africa to use seasonal precipitation Experimental Agriculture, 32: 461467.
forecasts with Burkina Faso as a case study. Kloppenburg, J. 1988. First the seed: the
Agricultural Systems, 74: 331349. political economy of plant biotechnology,
James, C. 2006. Global Status of 1492-2000. Cambridge, UK, Cambridge
Commercialized Biotech/GM Crops: University Press.
2006. ISAAA Brief No. 35. ISAAA, Ithaca, Koinange, E.M.K., Singh, S.P. & Gepts, P.
USA (available at http://www.isaaa.org/ 1996. Genetic control of the domestication
resources/publications/briefs/35/execu- syndrome in common bean. Crop Science,
tivesummary/pdf/Brief%2035%20-%20 36: 10371045.
Executive%20Summary%20-%20English. Lacy, S., Cleveland, D.A. & Soleri, D. 2006.
pdf; accessed 12 Jan. 2009). Farmer choice of sorghum varieties in south-
James, C. 2007. Global Status of ern Mali. Human Ecology, 34: 331353.
Commercialized Biotech/GM Crops: Louette, D., Charrier, A. & Berthaud, J. 1997.
2007. ISAAA Brief No. 37. ISAAA, Ithaca, In situ conservation of maize in Mexico:
USA (available at http://www.isaaa.org/ genetic diversity and maize seed manage-
resources/publications/briefs/37/execu- ment in a traditional community. Economic
tivesummary/pdf/Brief%2037%20-%20 Botany, 51: 2038.
Executive%20Summary%20-%20English. Louette, D. & Smale, M. 2000. Farmers seed
pdf; accessed 12 Jan. 2009). selection practices and maize variety char-
James, C. 2008. Executive Summary. Global acteristics in a traditional Mexican commu-
Status of Commercialized Biotech/GM nity. Euphytica, 113: 2541.
Crops: 2008. ISAAA Brief No. 39 (http:// Lyson, T.A. 2002. Advanced agricultural bio-
www.isaaa.org/resources/publications/ technologies and sustainable agriculture.
briefs/37/executivesummary/pdf/Brief%20 Trends in Biotechnology, 20: 193196.
37%20-%20Executive%20Summary%20 Matson, P.A., Parton, W.J., Power, A.G. &
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 361
sity in Chiapas, Mexico. Proceedings of the Lloyd & J.G. Hawkes, eds. Plant genetic
National Academy of Sciences of the United conservation: the in situ approach, pp. 160
States of America, 102: 949954. 175. Hants, UK, Chapman & Hall.
Pineyro-Nelson, A., Van Heerwaarden, J., Rajaram, S. & Ceccarelli, S. 1998. International
Perales, H.R., Serratos-Hernandez, J.A., collaboration in plant breeding. In H.J.
Rangel, A., Hufford, M.B., Gepts, P., Garay- Braun, ed. Wheat: prospects for global
Arroyo, A., Rivera-Bustamante, R. & Alvarez- improvement, pp. 533537. Dordrecht,
Buylla, E.R. 2009. Transgenes in Mexican Netherlands, Kluwer Academic.
maize: molecular evidence and methodolo- Revilla, P., Butrn, R., Malvar, R.A. & Ords,
gical considerations for GMO detection in A. 1999. Relationships among kernel weight,
landrace populations. Molecular Ecology, early vigor, and growth in maize. Crop
18(4): 750761. Science, 39: 654658.
Pommel, B. 1990. Influence du poids de la Richards, P. 1986. Coping with Hunger:
semence et de la profondeur de semis sur la Hazard and experiment in an African rice-
croissance et le dveloppement de la plan- farming system. London, Allen and Unwin.
tule de mas. Agronomie, 10: 699708. Rockefeller Foundation. 2007. Biotechnology,
Pressoir, G. & Berthaud, J. 2004a. Patterns breeding and seed systems for African
of population structure in maize landraces Crops: Research and product development
from the Central Valleys of Oaxaca in that reaches farmers. A conference joint-
Mexico. Heredity, 92: 8894. ly hosted by the Rockefeller Foundation
Pressoir, G. & Berthaud, J. 2004b. Population and the Instituto de Investigao Agrria
structure and strong divergent selection de Moambique (IIAM). Maputo,
shape phenotypic diversification in maize Mozambique, 2629 March 2007. The
landraces. Heredity, 92: 95101. Rockefeller Foundation, Nairobi, Kenya
PRGA. 2004. Participatory plant breeding. (available at http://www.africancrops.net/
Online (available at http://www.prgapro- rockefeller/conferences/programme.pdf;
gram.org/index.php?module=htmlpage accessed 12 Jan. 2009).
s&func=display&pid=9#ppb_resources; Schneider, J.R. 2002. Selecting with farmers:
accessed 12 Jan. 2009). the formative years of cereal breeding and
Provine, W.B. 1971. The origins of theoreti- public seed in Switzerland (1889-1936). In
cal population genetics. Chicago, USA, D.A. Cleveland & D. Soleri, eds. Farmers,
University of Chicago Press. scientists and plant breeding: integrat-
Pujol, B., David, P. & McKey, D. 2005. ing knowledge and practice, pp. 161187.
Microevolution in agricultural environ- Wallingford, UK, CABI.
ments: how a traditional Amerindian farming Serratos-Hernndez, J.-A., Gmez-Olivares,
practice favours heterozygosity in cassava J.-L., Salinas-Arreortua, N., Buenda-
(Manihot esculenta Crantz, Euphorbiaceae). Rodrguez, E., Islas-Gutirrez, F. & de-
Ecology Letters, 8: 138147. Ita, A. 2007. Transgenic proteins in maize
Qaim, M. & Zilberman, D. 2003. Yield effects in the Soil Conservation area of Federal
of genetically modified crops in developing District, Mexico. Frontiers in Ecology and
countries. Science, 299: 900902. the Environment, 5: 247252.
Qualset, C.O., Damania, A.B., Zanatta, A.C.A. Shorett, P., Rabinow, P. & Billings, P.R. 2003.
& Brush, S.B. 1997. Locally based crop The changing norms of the life sciences.
plant conservation. In N. Maxted, B. Ford- Nature Biotechnology, 21: 123125.
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 363
Sillitoe, P. 1998. What know natives? What Does it Mean for Biodiversity and
Local knowledge in development. Social Centers of Origin. Sponsored by Pew
Anthropology, 6: 203220. Initiative on Food and Biotechnology
Simmonds, N.W. 1990. The social context of (PIFB) and Fundacin Mxico-Estados
plant breeding. Plant Breeding Abstracts, Unidos para la Ciencia (FUMEC). Mexico
60: 337341. City. 2830 September.
Simmonds, N.W. & Smartt, J. 1999. Principles Soleri, D., Cleveland, D.A. & Aragn Cuevas, F.
of crop improvement. 2nd ed. Oxford, UK, 2006. Transgenic crop varieties and varietal
Blackwell Science. diversity in traditionally based agriculture:
Smale, M. 2002. Economics perspectives on the case of maize in Mexico. BioScience, 56:
collaborative plant breeding for conserva- 503-513.
tion of genetic diversity on farm. In D.A. Soleri, D., Cleveland, D.A. & Aragn Cuevas, F.
Cleveland & D. Soleri, eds. Farmers, scien- 2008. Food globalization and local diversity
tists and plant breeding: integrating knowl- - The case of tejate. Current Anthropology,
edge and practice, pp. 83105. Wallingford, 49(2): 280290.
UK, CABI. Soleri, D. & Smith, S.E. 1995. Morphological
Snow, A.A., Andow, D.A., Gepts, P., and phenological comparisons of two Hopi
Hallerman, E.M., Power, A., Tiedje, J.M. maize varieties conserved in situ and ex situ.
& Wolfenbarger, L.L. 2005. Genetically Economic Botany, 49: 5677.
engineered organisms and the environ- Soleri, D. & Smith, S.E. 2002. Rapid esti-
ment: Current status and recommenda- mation of broad sense heritability of
tions. Ecological Applications, 15: 377404. farmer-managed maize populations in the
Soleri, D. 1999. Developing methodologies Central Valleys of Oaxaca, Mexico, and
to understand farmer-managed maize implications for improvement. Euphytica,
folk varieties and farmer seed selection in 128: 105119.
the Central Valleys of Oaxaca, Mexico. Soleri, D., Smith, S.E. & Cleveland, D.A. 2000.
Dissertation, University of Arizona, Evaluating the potential for farmer and
Tucson, USA. plant breeder collaboration: a case study of
Soleri, D. & Cleveland, D.A. 2001. Farmers farmer maize selection in Oaxaca, Mexico.
genetic perceptions regarding their crop Euphytica, 116: 4157.
populations: an example with maize in Soleri, D., Cleveland, D.A., Castro Garca, F.H.
the Central Valleys of Oaxaca, Mexico. & Aragon, F. n.d. Farm characteristics asso-
Economic Botany, 55: 106128. ciated with maize diversity conservation in
Soleri, D. & Cleveland, D.A. 2004. Farmer Oaxaca, Mexico. Ms.
selection and conservation of crop varie- Soleri, D., Cleveland, D.A., Eriacho, D.,
ties. In R.M. Goodman, ed. Encyclopedia Bowannie, F., Jr., Laahty, A. & Zuni
of Plant & Crop Science, pp. 433438. New Community Members. 1994. Gifts from
York, USA, Marcel Dekker. the Creator: Intellectual property rights
Soleri, D. & Cleveland, D.A. 2005. Scenarios as and folk crop varieties. In T. Greaves, ed.
a tool for eliciting and understanding farm- IPR for Indigenous Peoples: A sourcebook,
ers biological knowledge. Field Methods, pp. 2140. Oklahoma City, USA, Society
17: 283301. for Applied Anthropology.
Soleri, D., Cleveland, D.A. &Aragn Cuevas, Soleri, D., Cleveland, D.A., Smith, S.E.,
F. 2003. Invited presentation at Gene Flow: Ceccarelli, S., Grando, S., Rana, R.B., Rijal,
364 Plant breeding and farmer participation
D. & Ros Labrada, H. 2002. Understanding ment. III. Participatory plant breeding,
farmers knowledge as the basis for collabo- a case for rice in Nepal. Experimental
ration with plant breeders: methodologi- Agriculture, 32: 479496.
cal development and examples from ongo- Stokstad, E. 2008. Dueling visions for a hun-
ing research in Mexico, Syria, Cuba, and gry world. Science, 319: 14741476.
Nepal. In D.A. Cleveland & D. Soleri, eds. Stone, G.D. 2007. Agricultural deskilling and
Farmers, scientists and plant breeding: inte- the spread of genetically modified cot-
grating knowledge and practice, pp. 1960. ton in Warangal. Current Anthropology,
Wallingford, UK, CABI. 48: 67103.
Soleri, D., Cleveland, D.A., Aragn Cuevas, Tilman, D., Cassman, K.G., Matson, P.A.,
F., Ceccarelli, S., Diarra, M., Grando, S., Naylor, R. & Polasky, S. 2002. Agricultural
Lacy, S., Michael, M., Ortiz, R., Rana, R.B., sustainability and intensive production
Rattunde, F., Rijal, D., Ros Labrada, H., practices. Nature, 418: 671677.
Sangar, A., Amadou, S., Smith, S.E., Issa, T. Uphoff, N. 2007. Agroecological alternatives:
& Weltzien, E.R. 2004. Farmers perceptions Capitalising on existing genetic potentials.
of crop genotype-environment interactions Journal of Development Studies, 43: 218
in five locations around the world. Project 236.
Report USDA NASS [National Agricultural Statistics
Soleri, D., Cleveland, D.A., Aragn Cuevas, Service]. 2004. 2002 Census of Agriculture
F., Ros Labrada, H., Fuentes Lopez, M.R. - United States Data. Crop Production--
& Sweeney, S.H. 2005. Understanding the Prospective Planting--Supplement (avail-
potential impact of transgenic crops in tra- able at http://www.nass.usda.gov/census/
ditional agriculture: maize farmers per- census02/volume1/us/st99_1_001_001.pdf;
spectives in Cuba, Guatemala and Mexico. accessed 12 Jan. 2009).
Environmental Biosafety Research, 4: 141 Venuprasad, R., Lafitte, H.R. & Atlin, G.N.
166. 2007. Response to direct selection for grain
Soleri, D., Cleveland, D.A., Glasgow, G.E., yield under drought stress in rice. Crop
Sweeney, S.H., Aragn Cuevas, F., Ros Science, 47: 285293.
Labrada, H. & Fuentes Lopez, M.R. 2008. Virk, D.S., Singh, D.N., Prasad, S.C., Gangwar,
Testing economic assumptions under- J.S. & Witcombe, J.R. 2003. Collaborative
lying research on transgenic food crops and consultative participatory plant breed-
for Third World farmers: Evidence from ing of rice for the rainfed uplands of eastern
Cuba, Guatemala and Mexico. Ecological India. Euphytica, 132: 95108.
Economics, 67(4): 667682. Virk, D.S. & Witcombe, J.R. 2007. Trade-offs
Srivastava, J.P. & Jaffee, S. 1993. Best practices between on-farm varietal diversity and high-
for moving seed technology: New approach- ly client-oriented breeding - a case study of
es to doing business. World Bank Technical upland rice in India. Genetic Resources and
Paper. Washington, DC, World Bank. Crop Evolution, 54: 823835.
Stern, R.D., Coe, R., Allan, E.F. & Dale, I.C., vom Brocke, K., Presterl, T., Christinck,
eds. 2004. Good statistical practice for natu- A., Weltzien R, E. & Geiger, H.H. 2002.
ral resource management. Wallingford, UK, Farmers seed management practices open
CABI. up new base populations for pearl millet
Sthapit, B.R., Joshi, K.D. & Witcombe, J.R. breeding in a semi-arid zone of India. Plant
1996. Farmer participatory crop improve- Breeding, 121: 3642.
Breeding for quantitative variables. Part 1: Farmers and scientists knowledge and practice in variety
choice and plant selection 365
vom Brocke, K., Weltzien, E., Christinck, A., 170. London, Intermediate Technology
Presterl, T. & Geiger, H.H. 2003a. Effects of Publications.
farmers seed management on performance Whitt, S.R., Wilson, L.M., Tenaillon, M.I.,
and adaptation of pearl millet in Rajasthan, Gaut, B.S. & Buckler, E.S. 2002. Genetic
India. Euphytica, 130: 267280. diversity and selection in the maize starch
vom Brocke, K., Weltzien, E., Christinck, A., pathway. Proceedings of the National
Presterl, T. & Geiger, H.H. 2003b. Farmers Academy of Sciences of the United States of
seed systems and management practices America, 99: 1295912962.
determine pearl millet genetic diversity in Wise, T.A. 2007. Policy space for Mexican maize:
semi-arid regions of India. Crop Science, protecting agro-biodiversity by promot-
43: 16801689. ing rural livelihoods. Global Development
Wallace, D.H. & Yan, W. 1998. Plant breeding and Environment Institute Working Paper
and whole-system crop physiology: improv- 07-01. Tufts University, Medford, USA.
ing crop maturity, adaptation and yield. Witcombe, J.R., Joshi, A., Joshi, K.D. & Sthapit,
Wallingford, UK, CABI. B.R. 1996. Farmer participatory crop
Wallace, H.A. & Brown, W.L. 1988. Corn and improvement. I. Varietal selection and breed-
its early fathers. Revised ed. Ames, USA, ing methods and their impact on biodiversity.
Iowa State University Press. Experimental Agriculture, 32: 445460.
Wang, R.-L., Stec, A., Hey, J., Lukens, L. Witt, H., Patel, R. & Schnurr, M. 2006. Can the
& Doebley, J. 1999. The limits of selec- poor help GM crops? Technology, repre-
tion during maize domestication. Nature, sentation & cotton in the Makhathini flats,
398: 236239. South Africa. Review of African Political
Wang, S., Just, D. & Pinstrup-Andersen, P. Economy, 33: 497513.
2006. Tarnishing silver bullets: Bt cotton World Bank. 2007. World Development
adoption and the outbreak of secondary Report 2008: Agriculture for Development.
pest infestation in China. Selected Paper Washington, DC, World Bank.
prepared for presentation at the American Zeven, A.C. 1998. Landraces: a review of
Agricultural Economics Association definitions and classifications. Euphytica,
Annual Meeting, Long Beach, USA, 2226 104: 127139.
July 2006 (available at http://www.grain. Zhu, Y., Wang, Y., Chen, H. & Lu, B.-R.
org/research_files/SWang_tarnished.pdf; 2003. Conserving traditional rice varieties
accessed 12 Jan. 2009). through management for crop diversity.
Weltzien, E., Smith, M.E., Meitzner, L.S. & BioScience, 53: 158162.
Sperling, L. 2003. Technical and institutional Zilberman, D., Ameden, H. & Qaim, M. 2007.
issues in participatory plant breeding - from The impact of agricultural biotechnology
the perspective of formal plant breeding. A on yields, risks, and biodiversity in low-
global analysis of issues, results & current income countries. Journal of Development
experience. PPB Monograph. PRGA Program Studies, 43: 6378.
Coordination Office, CIAT, Cali, Columbia. Zimmerer, K. 2002. Social and agroecologi-
Weltzien R., E., Whitaker, M.L., Rattunde, cal variability of seed production and the
H.F.W., Dhamotharan, M. & Anders, M.M. potential collaborative breeding of potatoes
1998. Participatory approaches in pearl mil- in the Andean countries. In D.A. Cleveland
let breeding. In J. Witcombe, D. Virk & J. & D. Soleri, eds. Farmers, scientists and
Farrington, eds. Seeds of choice, pp. 143 plant breeding: integrating knowledge and
366 Plant breeding and farmer participation
practice, pp. 83-105. Wallingford, UK, Zong, Y., Chen, Z., Innes, J.B., Chen, C.,
CABI Publishing. Wang, Z. & Wang, H. 2007. Fire and flood
Zimmerer, K.S. 1996. Changing fortunes: management of coastal swamp enabled first
biodiversity and peasant livelihood in the rice paddy cultivation in east China. Nature,
Peruvian Andes. Berkeley, USA, University 449: 459-U4.
of California Press.
367
CHAPTER 15
Raoul A. Robinson
368 Plant breeding and farmer participation
below), but these are too rare to be a gen- The main advantage of horizontal resist-
eral breeding tool. ance is that it is durable, and that it is
A second disadvantage of vertical possible to breed for increased levels of
resistance is that it is responsible for the many different quantitative variables simul-
vertifolia effect, which is the gradual loss of taneously. Participatory plant breeders can
horizontal resistance during breeding for accordingly aim at high levels of horizontal
vertical resistance, and which is described resistance to all locally important para-
in more detail below. A third disadvantage sites. This will achieve crop husbandry that
is that vertical resistances occur only against is effectively free of all parasite damage,
some species of parasite. Consequently, it is and one that is free of pesticides as well.
not possible to use vertical resistance for all And these freedoms will be permanent.
the locally important parasites. However, it must be emphasized that this is
In general, vertical resistance is not the objective. Such an objective may prove
recommended for participatory plant impossible to achieve in practice, at least in
breeding. This is mainly because the failure some crops, and in some areas. But, even if
of a wonderful new cultivar, which has this objective is unattainable, there will be at
taken years of devoted work to produce, least some improvement over current farm-
by both professional and amateur breeders, ing practices in terms of increased yields
is quite frankly heart-breaking. Nothing and decreased damage from parasites.
can be expected to discourage amateur
breeders more than this. An essential aspect 15.1.6 Two kinds of plant breeding
of participatory plant breeding is that we Clearly, the key difference in breeding crop
maintain the confidence of the participating plants for temporary and durable resist-
amateurs. Consequently, participating ances is the difference between breeding for
professional breeders should be very single-gene and multiple-gene characters.
cautious about recommending the use of Breeding for single-gene characters
vertical resistance, or even the combined requires both pedigree breeding and back-
use of vertical and horizontal resistance, in crossing, or the very modern techniques of
participatory plant breeding. marker assisted selection and genetic engi-
neering. Anyone using these techniques for
15.1.5 Horizontal resistance acquiring resistance should assume that the
Being a quantitative variable, horizontal resulting resistance will be unstable, and
resistance can be expressed at any level that it will have a very high probability of
between a minimum and a maximum. In the breaking down sooner or later.
absence of crop protection chemicals, the Breeding for multiple-gene characters,
minimum level of horizontal resistance usu- such as horizontal resistance, requires an
ally results in a total loss of crop from the entirely different breeding technique, called
parasites. And the maximum level of hori- recurrent mass selection, which is discussed
zontal resistance usually results in negligible below.
loss of crop. However, the maximum level of
horizontal resistance never provides as com- 15.2 WHY WAS TEMPORARY
plete a protection as vertical resistance. Even RESISTANCE SO POPULAR?
with the highest level of horizontal resist- During most of the twentieth century, ver-
ance, there is always some slight parasitism. tical resistance was consistently the resist-
Breeding for quantitative variables. Part 2: Breeding for durable resistance to crop pests 371
TABLE 15.1
Summary of misleading variables
Misleading Variable Problem Solution
Parasite interference Hides high levels of horizontal resistance Use relative measurements only
Epidemiological Resistance requirements vary between agro- Use on-site screening
competence ecosystems
Environmental erosion An apparent loss of horizontal resistance Each agro-ecosystem should have its own
due to re-location in an area of higher breeding programmes
epidemiological competence
Vertifolia effect The loss of horizontal resistance when Inactivate all vertical resistances, and avoid
breeding crops for vertical resistance, or any use of crop protection chemicals
under the protection of crop protection
chemicals
Parasite erosion An apparent loss of horizontal resistance Screen in an area of high parasite density
due to changes in the parasite population
False erosion An apparent loss of horizontal resistance Be more careful
due to sloppy or negligent assessments
Biological anarchy An increased epidemiological competence in This phenomenon will disappear after a few
the parasite due to the loss or debilitation seasons of freedom from pesticides, as the
of biological control agents biological controls are restored by the use of
horizontal resistance
Population immunity Makes field assessments preferable Avoid laboratory assessments of resistance
Chance escape Provides a false indication of resistance Inoculate the screening population, or use
grid screening
Quantitative vertical Looks like horizontal resistance but is not Use parents that lack vertical resistance
resistance genes, or use the one-pathotype technique
(see below).
Durable vertical This rare phenomenon provides a false hope Do not rely on this remote possibility
resistance for breeders of single-gene resistances
Sub-optimization Do not breed for a single resistance Use the holistic approach (see below)
mechanism, or resistance to single species of
parasite
to all the locally important parasites. This epidemiological competence is low, the
requires the one quality of good health. horizontal resistance will also be low,
The least parasitized individuals will be because there is little selection pressure for
fairly susceptible to many different species resistance. But where the epidemiological
of parasite. Their quality of good health competence is high, the level of horizontal
will be low. But it will be higher than all resistance will also be high, because there is
those other, less healthy individuals in the strong selection pressure for resistance.
screening population, many of which may When breeding for horizontal
have disappeared entirely. resistance, this variation in epidemiological
A final comment about parasite interfer- competence is important in two ways.
ence concerns the movement of parasites First, we must use on-site screening. This
from one farm, or one district, to another. means that the screening for horizontal
Many organic farmers are able to culti- resistance must be conducted: (i) in the
vate fairly susceptible cultivars successfully locality of future cultivation; (ii) in the time
because their neighbours are using crop of year of future cultivation; (iii) in the field
protection chemicals. The district interfer- (i.e. not in the laboratory or greenhouse);
ence is then minimal. But were all the farm- and (iv) according the farming system (e.g.
ers in that district to eschew crop protection organic or conventional, irrigated or rainfed)
chemicals, the parasite populations would of future cultivation. On-site screening is
be so large, and the district interference particularly well suited to participatory
so great, that the use of those cultivars for plant breeding.
organic agriculture might prove impossible. Second, each distinct agro-ecosystem
The other side of this coin is that progress will require its own horizontal resistance
in breeding for horizontal resistance will breeding programme for each of its crop
lead to reductions in district interference, species. This programme must be aimed at
which, in turn, will enhance the effects of the levels of epidemiological competence of
that horizontal resistance. the locally important parasites. In practice,
this is not difficult. Agro-ecosystems are
15.3.2 Epidemiological Competence usually quite large, and the necessary levels
Epidemiological competence refers to the of horizontal resistance will be discovered
ability of a crop parasite to cause an epi- by practical farming experience, spread
demic (or infestation). It is another bio- over time, as the breeding programmes
logical variable that can be expressed at any produce more and more new cultivars, with
level between a minimum and a maximum. higher and higher levels of comprehensive
Consider a wild ecosystem, which horizontal resistance.
might, perhaps, extend up a mountainside.
The epidemiological competence of a plant 15.3.3 Environmental erosion of
parasite might change along a gradient horizontal resistance
from low to high moisture, or temperature, If a cultivar has adequate horizontal resist-
or whatever factor is governing that ance in an agro-ecosystem in which the
epidemiological competence within the parasite has a relatively low epidemiological
ecosystem. The various host ecotypes competence, and that cultivar is taken to a
along that gradient will have corresponding different agro-ecosystem, where the parasite
levels of horizontal resistance. Where the has a high epidemiological competence, the
374 Plant breeding and farmer participation
It should be added that the best way Indeed, it must be strongly positive if it is
to restore biological controls is to use to become a serious crop pest or disease.
horizontal resistance. And the best way to Each individual parasite must spawn more
enhance the effects of horizontal resistance than one new individual before it dies. Now
is to restore biological controls. The two suppose that the combination of horizontal
effects are mutually reinforcing. The resistance and biological controls is such
practical effect of this is that a new cultivar that, on average, each parasite individual
with apparently inadequate horizontal spawns less than one new individual. The
resistance may well prove to have adequate parasite population growth is now negative,
horizontal resistance, once the biological even though the host individuals are not
controls are restored. This restoration immune. This situation is population
may require several seasons of cultivation immunity, and it is important in horizontal
without pesticides but, once complete, the resistance breeding in three quite different
effects can be dramatic. ways.
It should also be added that the agents First, population immunity means that
for biological control often depend on a we do not need to breed for the maximum
small population of the parasite in order levels of horizontal resistance. We need to
to maintain their own populations. A low breed for enough horizontal resistance to
level of parasitism is often desirable for this cause population immunity, and no more.
reason, provided that it has no deleterious This level is discovered from practical
effect on the yield or quality of the crop farming experience.
product. However, purchasers of organic Second, although vertical resistance can
food often like to see minor parasite be assessed on detached leaves in a test tube,
damage as evidence for freedom from crop or leaf disks in a Petri dish, or even entire
protection chemicals. plants in a growth chamber, horizontal
resistance should not be measured in this
15.3.8 Population immunity way. This is because these laboratory
Population immunity means that a host methods cannot possibly represent the
population is effectively immune, even effects of biological controls or population
though the individuals that make up that immunity. The levels of horizontal
population are less than immune. This resistance are best determined in the field
is because population growth, unlike an and, if at all possible, under conditions of
individuals growth, can be positive or restored biological controls. Once again,
negative. Positive population growth means the best determinations will be the result of
that there are more births than deaths, practical farming experience.
and the population is increasing. Negative Third, all measurements and descriptions
population growth means that there are of horizontal resistance must be relative.
more deaths than births, and the population There can be no absolute measurements.
is decreasing. The dividing line occurs when We can describe a new cultivar A as being
the births and deaths are equal, and the more resistant to a particular parasite than
population growth is then zero. cultivar B, but less resistant than cultivar
Now consider a crop parasite. In order C. But we cannot have an absolute scale
to cause an epidemic (or infestation), the of measurement comparable to the Celsius
parasite population growth must be positive. scale of temperatures. This is because these
Breeding for quantitative variables. Part 2: Breeding for durable resistance to crop pests 377
biological variables are too imprecise to of parasite and, with participatory plant
define accurately. However, this does not breeding, this will be the responsibility of
make horizontal resistance any less useful the professionals.
in farmers fields. If the parasite is wind-borne, such as most
fungal spores, or a flying insect, a previously
15.3.9 Chance escape prepared population of the parasite can be
The distribution of parasites in a screening released, blown or water-sprayed on to
population is often uneven. This is called a the screening population. Once again, the
patchy distribution and it is most common details of the techniques vary and will be
with soil-borne parasites and gregarious handled by the professionals.
insect pests. There may then be some host With some parasites, inoculation is not
individuals that have no parasites at all, feasible for technical reasons. An alternative
and they give the impression of being technique is then to ignore those parts of
highly resistant. They are known as chance the screening population that are free of
escapes and, obviously, they should not the parasite in question. Or any individual
be selected as parents of the next screening plant that is entirely free of the parasite in
generation because they could be very question can be ignored. This procedure
susceptible. But the problem is how do runs the risk of discarding some highly
we recognise them, and how do we avoid resistant potential parents, but this wastage
them? There are a number of techniques is preferable to the risk of selecting highly
that increase the accuracy of the screening, susceptible escapes as parents of the next
depending on the nature of the parasite. generation.
It is in this area that the professionals will Finally, there may be parasite gradients
be most useful to participatory amateur within the screening population in which
breeders. the intensity of parasitism changes
If the parasite is soil-borne, such as a gradually from low to high from one part
root nematode, or a fungal or bacterial of the host population to another. This
wilt organism, it is a good idea to pre- effect can be eliminated by dividing the
germinate the seedlings in flats or peat screening population into a grid of suitably
pots for later transplanting in the field. sized squares. Only the least parasitized
These flats or pots can then be inoculated individual is selected within each square,
with the parasites in question, and the very regardless of the level of that parasitism
process of transplanting will ensure an even when compared with other squares.
distribution of the parasite.
If the parasite is a gregarious insect, in 15.3.10 Quantitative vertical resistance
which all individuals tend to congregate Occasionally, vertical resistance can be
on one host plant, they can often be quantitative and it is then easily confused with
redistributed on a daily basis by disturbing horizontal resistance. It occurs, for example,
them. This can be particularly important with Hessian fly (Mayetiola destructor)
with virus vectors. of wheat (Dent, 1998). Fortunately, this
If the parasite is seed-borne, it is usually situation is rare and need not worry the
feasible to inoculate the seed before sowing. breeders of most crops. However, if it does
The details of the techniques for doing occur, quantitative vertical resistance must
this vary considerably with different kinds obviously be avoided or inactivated during
378 Plant breeding and farmer participation
originated somewhere else. This is analogous the lock of the host and, consequently, all
to cross-pollination, or allogamy. With auto-infection is matching infection. Even
auto-infection, the parasite is born, hatched parasites that reproduce sexually, such as
or spawned on the host that it is infecting. many insects, will soon reach homozygosity
This is analogous to self-pollination, or of the matching biotype. However, vertical
autogamy. We also recognize that, in a wild resistances against insects are rather rare,
pathosystem, vertical resistance can control and this may explain why there has been so
allo-infection only, and auto-infection can little crop breeding for resistance to insects
be controlled only by horizontal resistance pests.
(although horizontal resistance can also These functions of the two kinds of
control allo-infection). resistance are emergent properties that
The control of allo-infection by ver- were completely unknown until recently,
tical resistance apparently operates as a and, being unknown, they were inevitably
system of locking, with each host having ignored by crop scientists during the
a biochemical lock, consisting of several twentieth century.
resistance genes, and each parasite having a Another obvious example of sub-
biochemical key consisting of several para- optimization occurs when breeding for a
sitism genes. If the parasite key does not fit single resistance mechanism, such as hairy
the lock of the host it is allo-infecting, the leaves that repel an insect pest. Horizontal
infection fails, while if the key does fit the resistance to insects usually consists of
lock the infection succeeds. Such a system many obscure mechanisms, all of which may
ensures that the frequency of matching vary quantitatively, and which collectively
allo-infection is low, and this stabilizes the reduce the rate of population growth of
population explosion of the parasite. This that pest.
stabilization is an emergent property that The converse of sub-optimization is
can be seen only at the level of the system known as the holistic approach, which
of locking, the level of the pathosystem. leads to local optimization of all variables.
However, if every door in the town has When breeding for horizontal resistance,
the same lock, and every householder has therefore, we must not sub-optimize. We
the same key, which fits every lock, the must work at the systems level of the agro-
system of locking is ruined by uniformity. ecosystem. Within this agro-ecosystem, we
And this is exactly what we have done in must use population breeding to produce
agriculture, with our use of a single vertical adequate and durable resistance to all locally
resistance in a uniform pure line, clone or important species of parasite, by exerting
hybrid cultivar that might be grown over selection pressure for the one characteristic
a huge area as a homogeneous population. of good health. Susceptibility to only one
This was sub-optimization at its worst. important species of parasite will result in an
Auto-infection can be controlled only inferior cultivar, and this would constitute a
by horizontal resistance because auto- clear case of sub-optimization. In addition
infection can commence only after there has to good health, new cultivars should have
been a matching allo-infection. Many crop good levels of all the other variable attributes
parasites reproduce asexually to produce necessary to a productive agriculture. This
a clone. All the individuals within that approach does not necessitate participatory
clone have the same key which matches plant breeding, and some professional
380 Plant breeding and farmer participation
breeders may choose to work on their own This crop is called corn in North America
in a scientific institution. However, this but it is called maize in all other countries,
approach does depend on high numbers of and in all other languages. It was taken by
plants being screened, and many amateur the Spanish from the New World to the
breeders, working cooperatively with Iberian peninsula some five centuries ago
a professional, can lead to both greatly and tropical rust was either left behind or
increased numbers of plants screened, and it failed to survive outside the tropics. The
greatly increased attention applied to each Portuguese then took rust-free maize to
plant screened. Africa and all points east, where it was cul-
tivated for more than four centuries in the
15.4 BREEDING CROPS FOR DURABLE absence of tropical rust. This negative selec-
(HORIZONTAL) RESISTANCE tion pressure led to the level of horizontal
In this section there are inevitable resistance to tropical rust declining to its
generalizations that do not apply to all minimum natural level. In technical terms,
crop species. For example, comments about this is the Hardy-Weinberg equilibrium.
open-pollinated crops may not apply to self- In theory, it should be possible to breed
pollinated crops, or comments about annual experimentally for absolute susceptibility,
crops may not apply to perennial crops. but this is a somewhat academic point.
When planning a breeding programme, With the development of trans-Atlantic
therefore, readers should highlight only air transport in the 1940s, tropical rust was
those aspects of these descriptions that accidentally taken from the new world to
apply to their crop species of choice. the old. Devastating epidemics developed
in the low altitude, equatorial tropics. In
15.4.1 Maize in tropical Africa East Africa, a classic breeding programme
The best way to breed for horizontal for vertical resistance was initiated. Genes
resistance is to imitate the behaviour of for resistance had to be imported from
open-pollinated maize (Zea mays) in tropical America because none could
tropical Africa, following the introduction be found in the local maize populations
of the re-encounter disease called Tropical (Storey et al., 1958). Unfortunately, these
rust (Puccinia polysora). A re-encounter vertical resistances broke down so quickly
parasite is one in which the host was that none lasted long enough to be released
separated from its parasite and taken to farmers.
to another part of the world. At a later However, the appearance of this
date, the parasite is also taken to that re-encounter disease exerted positive
new area where it re-encounters its host, selection pressure for horizontal resistance
which has lost resistance in the meanwhile. in the farmers open-pollinated crops. After
Conversely, a new encounter parasite is about a dozen maize generations, the levels
one which evolved separately from its host, of disease had declined from total loss
on a botanical relative. Later the two are of crop to negligible loss of crop. The
brought together in a new encounter. An horizontal resistance had increased from its
old encounter parasite is one in which the minimum level to its maximum level. With
host and parasite have never been separated, two crops each year, this transformation
even though both may have been moved to occurred in about six years, and it had
new areas (Buddenhagen, 1987). happened without any help from plant
Breeding for quantitative variables. Part 2: Breeding for durable resistance to crop pests 381
It is not clear whether the maize in and with each breeding cycle the levels
tropical Africa lacked vertical resistance of horizontal resistance increase until no
genes, or that it had such genes but their further increase is either possible or neces-
resistances were matched so quickly that sary. This process of quantitative increase,
they went unobserved. In either event, the in which the progeny have a higher level of
positive selection pressure for horizontal a quantitative variable than their parents, is
resistance was not hindered by functioning known as transgressive segregation.
vertical resistance. So, when breeding for horizontal resist-
ance, there is no need to begin with a good
15.4.3 Genetic resources source of resistance, but there must be a
It is now a plant breeding shibboleth that reasonably wide genetic base to ensure that
breeding for resistance requires a good all the necessary polygenes are present.
source of resistance before the breeding In practice, it is much easier to breed
can even begin. This is true when breeding for horizontal resistance than it is to
for vertical resistance, because it is essential breed for high yield, high quality of crop
to have at least one gene for resistance. But product or high agronomic suitability.
it is not true when breeding for horizontal It is therefore best to use high-yielding,
resistance. This point is so important that it high-quality, agronomically suitable, but
merits careful explanation. susceptible, modern cultivars as the original
Consider a heterogeneous screening parents. With suitable selection procedures,
population in which every individual is it should be easy to gradually increase
different from every other individual, but the levels of horizontal resistance, while
in which all the plants are susceptible. Each retaining the other desirable qualities.
plant possesses about 10 percent of the total Conversely, it would be very difficult to use
polygenes that contribute to the horizontal highly resistant, primitive archetypes as the
resistances to each of the various locally original parents, and then try to improve
important parasites. The host population as their various agricultural attributes, while
a whole is thus very susceptible. However, retaining their resistance.
we may assume that each host individual The maize of tropical Africa illustrat-
possesses a different 10 percent of those ed this point conclusively. The horizontal
total polygenes. Provided that there is a resistance accumulated within very sus-
reasonably wide genetic base, this will ceptible host populations of highly prized
mean that all the polygenes are present in local landraces. No good source of resist-
the population, but their frequency is too ance was necessary, and no diminution of
low for much resistance to be expressed in the prized characteristics occurred. When
any of the individuals. The objective of the breeding for horizontal resistance, there-
breeding is to increase these resistance gene fore, the discernible qualities of the genetic
frequencies. resources must be those of yield, quality of
The most resistant plants are selected and crop product and agronomic suitability.
they become the parents of the next screen-
ing generation. Now the most resistant 15.4.4 Population breeding
individuals will possess perhaps 20 percent Recurrent mass selection means that a
of the total polygenes. In the next screening heterogeneous plant population is screened
generation, this percentage is even higher, for the best individuals, which then become
Breeding for quantitative variables. Part 2: Breeding for durable resistance to crop pests 383
the parents of the next generation. This plants to total plants. This ratio is called
process is repeated some 1015 times, the selection coefficient. In practice, this
by which time the upper limits of most means that the screening population should
quantitative variables will have been be as large as possible so that perhaps only
reached. In each breeding cycle (i.e. each one plant in a thousand becomes a parent
generation of recurrent mass selection), of the next generation. The possibilities
there should be at least 10 to 20 parents, depend very much on the nature of the
depending on the nature of the crop. These crop. If the plants are small, such as wheat,
parents may be randomly cross-pollinated, rice or beans, it is entirely feasible to use
or hand-pollinated in all combinations, a screening population of some 100 000
again depending on the nature of the crop. plants, but if the population is a tree crop,
Quantitative variables change as a result such as a fruit or nut species, such large
of selection pressures. The term pressure populations are not feasible. However, the
is used in the sense of bringing pressure to size of the screening population is not
bear, of coercion or persuasion, and selec- critical, and if a relatively small population
tion pressures can be positive or negative. is necessary because of land or labour
Positive selection pressures lead to restrictions, the breeding programme will
the increase of a variable, while negative require more time, but the deficiency will
selection pressures lead to its decrease. The be no worse than this.
mechanism of these changes is reproductive Should it transpire that the original
fitness. For example, if a heterogeneous genetic base was too narrow to accumulate
host population is susceptible to a parasite, adequate horizontal resistance, new genetic
the most resistant individuals will be material can be added to the screening
parasitized the least and will reproduce population. This may lead to an initial,
the most, while the most susceptible slight loss of horizontal resistance, but the
individuals will be parasitized the most ultimate potential will be improved.
and will reproduce the least. With each A special aspect of quantitative
generation the population as a whole will variables is that they must all be increased
gain resistance as a consequence of this simultaneously. There is little point in
positive selection pressure for resistance. having high levels of horizontal resistance
Conversely, if the parasite is absent from to all of the locally important parasites
the locality in question, as with the maize of except one. Even a single susceptibility
tropical Africa, or because of a functioning will spoil a cultivar, and make spraying
vertical resistance or the use of a pesticide or some other form of artificial control
(i.e. the vertifolia effect; see above), the necessary (see also sub-optimization,
selection pressure for horizontal resistance above). This is a major difference between
will be negative, and the frequency of breeding for single-gene and multiple-gene
genes controlling horizontal resistance characters. Pedigree breeding allows the
will decrease. This happens because any transfer of a single-gene character, such as
unnecessary genetic characteristics tend a resistance, from a wild plant to a cultivar
to decline to a level called the Hardy- by hybridization and back-crossing. A
Weinberg equilibrium. multiple-gene variable cannot be transferred
Positive selection pressure can be in this way because hybridization leads to
increased by increasing the ratio of selected an immediate dilution. Hence the need for
384 Plant breeding and farmer participation
are infecting all the other trees. If the farmers crop constituted a natural screen-
most susceptible trees are identified and ing population. Each farmers maize then
removed, and all other diseased branches constituted a landrace with exactly the right
are also removed, this parasite interference amount of horizontal resistance to tropical
(see above) will stop and the disease will rust for that latitude and that altitude. This
be controlled. Even if the disease is merely was an example of subconscious selection
reduced in intensity, further negative Because tropical rust is so sensitive to
screenings of the most susceptible trees altitude and latitude, the pathosystems,
will eventually control the disease. This and hence the agro-ecosystems, of tropical
procedure is often far more economical rust are small. However, this extreme of
than a positive screening for resistant trees, environmental sensitivity is unusual, and
followed by a subsequent replanting of the with most crop species the agro-ecosystems
entire crop with these selections. are quite large, and relatively few breeding
programmes are necessary.
15.4.8 On-site selection
The maize of Africa illustrated the 15.4.9 A holistic approach
importance of on-site selection (see The tropical maize in Africa also illustrates
above). Puccinia polysora has maximum the need for a holistic approach. Before the
epidemiological competence at the equator, appearance of tropical rust, the maize had
and at sea level. As latitude increases, the no important pests or diseases. In other
epidemiological competence decreases to words, it had high levels of comprehensive
nothing at sea level at the tropics of Cancer horizontal resistance. That is, it had
and Capricorn. As altitude increases, the adequate levels of horizontal resistance to
epidemiological competence decreases to all the locally important parasites. With the
nothing at the equator at elevations of introduction of this re-encounter parasite,
about 1 200 m. this pathosystem balance was immediately
Maize from the highlands of Kenya, lost, and it required about a dozen
where tropical rust lacks epidemiological generations of selection to restore it.
competence entirely, is extremely It can be argued that pathosystem balance
susceptible when planted at sea level near has been lost in virtually all of our modern
the equator. Conversely, maize in Malawi crops. The objective of participatory plant
was reported to be highly resistant to breeding should be to restore pathosystem
tropical rust but it proved to be very balance in each crop species in each agro-
susceptible when planted near the equator ecosystem. When we consider the many
at sea level. This was environmental erosion different crops and the many different agro-
that occurred because the maize had come ecosystems worldwide, this is too big a task
from an area of minimum epidemiological for professional plant breeders to undertake
competence of the pathogen, where it had on their own, and it is perhaps the best
suffered minimum selection pressure for justification of participatory plant breeding.
resistance. When planted in an area of
maximum epidemiological competence, its 15.4.10 Selection pressures for other
susceptibility was revealed. qualities
Because this maize in tropical Africa was If we were to produce new cultivars that
cultivated as an open-pollinated crop, each had high levels of horizontal resistance to all
386 Plant breeding and farmer participation
locally important parasites, but which had 15.5 EXAMPLES OF BREEDING FOR
reduced yield and quality of crop product, HORIZONTAL RESISTANCE
we would be sub-optimizing (see above). Simmonds (1991) has compiled a compre-
This is why we should use modern but sus- hensive review of the results of breeding for
ceptible cultivars as our genetic resource. It horizontal resistance. He gives examples
is clear that a good source of resistance is of durable resistance in 21 species of crop,
not necessary when breeding for horizontal functioning variously against airborne and
resistance, but that high yield, quality and soil-borne pathogensfungal, bacterial,
agronomic suitability are necessary. The viral, insect and nematode. Stoner (1992)
levels of various horizontal resistances are reviewed 705 papers on host resistance to
increased while selection pressures for yield insects and mites in vegetables, and she also
and quality are maintained to ensure that quotes reviews of this topic in grain crops,
these qualities are not reduced. alfalfa and cotton. She comments that, in
most studies, the resistance is a quantitative
15.4.11 Measurement of horizontal trait, but she adds that there has been little
resistance plant breeding for resistance to insects.
When measuring the results of breeding for
horizontal resistance, assessments can be 15.5.1 Potatoes in Kenya, Mexico,
relative only. That is, we can say that a new Scotland and the United States of
cultivar has either greater or less horizontal America
resistance to a particular parasite than John S. Niederhauser was one of the
another well known and well tried cultivar. pioneers of horizontal resistance. Indeed,
An alternative description can be given with he was the first scientist to reject the
the phrase spraying not necessary, but use of vertical resistance in favour of
even this must be qualified with the rider horizontal resistance, and he did this in
that this is only true in a normal season. Mexico with resistance to Late blight of
potato (Phytophthora infestans). His most
15.4.12 Crops that are difficult or famous cultivar was Atzimba, and Mexican
impossible to breed scientists have continued his work. Blight is
For technical reasons, some crops are so severe in Mexico that the popular cultivar
difficult or even impossible to breed, and Alpha has to be sprayed with fungicides 25
amateur breeders should not attempt to times each season. The new, horizontally-
improve them. These include banana, citrus, resistant cultivars, such as Sangema and
date palm, figs, garlic, hops, horseradish, Tollocan, need to be sprayed only once or
olives, pineapple, sisal and wine grapes. twice each season.
However, most of the main food crops Working in Scotland, Simmonds (1976)
are easy to breed, and none of them could demonstrated that the potatoes cultivated in
be described as being difficult to breed. Europe (Solanum tuberosum) were derived
Worldwide, it is clearly logical for amateur from the S. andigena of South America.
plant breeders to work with participatory With only four generations of recurrent
plant breeding of crops that are easy to mass selection he was able produce neo-
breed, while the professional plant breeders tuberosum from S. andigena, and he was
should work with crops that are difficult also able to accumulate useful levels of
to breed. horizontal resistance to Late blight.
Breeding for quantitative variables. Part 2: Breeding for durable resistance to crop pests 387
Stoner, K.A. 1992. Bibliography of plant resist- Vanderplank, J.E. 1963. Plant Diseases:
ance to arthropods in vegetables, 19771991. Epidemics and Control. New York, USA,
Phytoparasitica, 20(2): 125180. and London, Academic Press. 349 p.
Storey, H.H., Howland (Ryland), A.K., Vanderplank, J.E. 1968. Disease Resistance
Hemingway, J.S., Jameson, J.D., Baldwin, in Plants. New York, USA, and London,
B.J.T., Thorpe, H.C. & Dixon, G.E. 1958. Academic Press. 206 p.
East African work on breeding maize resist- Vanderplank, J.E. 1978. Genetic and Molecular
ant to the tropical American rust Puccinia basis of Plant Pathogenesis. Berlin, Germany,
polysora. Empire Journal of Experimental Springer Verlag. 167 p.
Agriculture, 22: 117.
391
CHAPTER 16
isolation; they often interact, both with agronomic, genetic, breeding, physiological
other abiotic stresses and with biotic stress. and molecular aspects of drought resistance,
Moreover, areas with a high probability of or as recently more often used, water
abiotic stresses generally have low-input productivity (Passioura, 2006). This is
agriculture (Cooper et al., 1987), because highlighted by the publication of several
the risk of losing the crop or of a low yield reviews (Ceccarelli et al., 2004; Reynolds,
discourages the farmers from using costly Mujeeb-Kazi and Sawkins, 2005; Parry,
inputs, particularly fertilizers. This results Flexas and Medrano, 2005).
in low outputs, poor human nutrition Drought has been always a challenge
and reduced educational and employment to plant breeders, despite many decades of
opportunities, especially for girls. The rural research (Blum, 1993). The development,
poor are particularly badly affected because through breeding, of cultivars with higher
of lack of access to alternative sources of and stable harvestable yield under drought
employment or food. conditions would be a major breakthrough
(Ceccarelli and Grando, 1996). However,
16.2.1 Drought drought resistance is a very elusive trait
Drought, defined as water availability below from a genetic point of view. This is
that required for maximum crop yield, is because the occurrence, severity, timing
one of the main factors limiting crop pro- and duration of drought vary from year
duction. Although it reaches the front pages to year, and although every year there are
of the media as drought warnings or when it winners, it is difficult to find those that
causes famine and death, drought is a perma- are consistently successful. To make matters
nent constraint to agricultural production in worse, drought seldom occurs in isolation;
many developing countries, and an occa- it often interacts with other abiotic stresses
sional cause of losses in agricultural pro- (particularly temperature extremes), and
duction in developed ones. Several drought with biotic stress. As mentioned earlier
warnings have been issued in recent years in this chapter, the risk of losing the crop
in Australia, Europe and the United States because of drought limits the use of inputs.
of America. Climate changes will increase Also the definition of dry areas seems
the frequency of droughts, particularly in to be an elusive issue. This is illustrated by
Southeast Asia and Central America, and by the distribution of crops in different agro-
2050 are expected to cause water shortages climatic environments. For example, in a
for 67 percent of the future population in country such as the Syrian Arab Republic,
the world (Ceccarelli et al., 2004). with a large spatial variability of rainfall
In areas where water availability is within short distances (van Oosterom,
limited, and irrigation is not available, the Ceccarelli and Peacock, 1993), bread wheat
choice of crops is restricted to a few, and (Triticum aestivum L.), durum wheat
often to only one, thus making farmers in (T. turgidum var. durum L.) and barley
those areas vulnerable for lack of options. among the cereals, and faba bean (Vicia
In fact, most of the rural poor live in faba L.), chickpea (Cicer arietinum L.) and
areas where crop productivity and crop lentil (Lens culinaris L.) among the food
diversification are limited by lack of water. legumes, are grown in progressively drier
Therefore it is not surprising that there is environments, with some overlapping.
an ongoing global research effort on social, Therefore, a dry area for faba bean or bread
394 Plant breeding and farmer participation
wheat is moderately favourable for durum Soil mineral stresses are increasingly
wheat and chickpea, and a dry area for becoming important limiting factors for
durum wheat and chickpea is moderately crop plants in many parts of the world.
favourable for barley and lentil. Acid soil and associated aluminum toxicity
At the drier end of the spectrum, barley affect over 2 billion hectares worldwide
and lentil are the only rainfed crops, and (Humphreys and Humphreys, 2005).
the other cereals or legumes are only grown Mineral nutrient deficiency can be caused
under supplementary or full irrigation. The by low nutrient status of the soil, low
situation described for the Syrian Arab mobility or availability of nutrients within
Republic applies to most countries of the the soil.
Mediterranean basin and West Asia, and Salinity is generally defined as the
for crops such as millet, sorghum and maize presence of excessive amount of soluble salts
to the dry areas of the tropics, and is only that hinder or affect the normal function of
altered by irrigation. plant growth (Shafiq-ur-Rehman, Harris
The complexity of breeding for dry and Ashraf, 2005). Saline soils have a
areas is not only due to the biological mixture of chloride salts, with sodium
complexity of drought resistance, but chloride being often dominant. Salinity can
also to the consequences of drought for be divided into primary sources in soils
the livelihood of people living in the dry derived from saline parent rocks (Sposito,
areas. In developed countries, farmers have 1989) and secondary salinization caused
various forms of social protection against by human intervention, such as irrigation
the devastating effects of drought, while (Sposito, 1989).
in developing countries farmers have to Salinization is one of the most common
survive on their own, usually selling their forms of soil degradation. Almost all con-
assets, most commonly livestock. In areas tinents have problems related to saline soils
affected by drought in developed countries (Pessarakli, 1999), and is particularly severe
farmers may prefer cultivars capable of in arid and semi-arid regions.
high yields in the few favourable years; It is estimated that 6 percent of the
this is very different in the dry areas of worlds land and 30 percent of the worlds
most developing countries with no or irrigated areas already suffer from salinity
little social assistance where farmers prefer problems (Unesco Water Portal, 2007).
varieties capable of some yields even in the
driest years. This is an example that the 16.2.3 Temperature stresses
same biological problem in different social Temperature extremes can be experienced
contexts requires different solutions. on both a daily or seasonal basis. Long-
term climatic changes lead to higher average
16.2.2 Soil toxicities and deficiencies temperatures and increase the frequency
Soil plays a major role in determining and severity of extreme temperature events.
the amount and availability of nutrients As with other stresses, early and late stages
and toxic minerals. Soil toxicities and of crop growth are particularly sensitive
deficiencies render more than one hundred to temperature extremes. Plants can be
million hectares of agricultural land marginal affected by exposure to prolonged periods
for agriculture, limiting production and of moderately high temperatures as well to
creating poverty for millions. short periods of extremely high tempera-
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 395
tures. Low temperatures can affect plants have become more frequent over most land
by chilling, which leads to physiological areas, the frequency of heavy precipitation
and developmental abnormalities, and by events has increased over most areas, and
freezing, which causes cell damage. About since 1975 the incidence of extreme high sea
15 percent of arable land is estimated to be level has increased worldwide.
affected by freezing stress (Dudal, 1976). In conclusion, higher temperatures are
Changes in temperature are the most part of the future climate for which breeders
certain aspect of climate changes. The should breed today.
most recent evidence from the Fourth
Assessment Report on Climate Change 16.3 CHALLENGING CONVENTIONAL
of the Intergovernmental Panel on BREEDING CONCEPTS
Climate Change (IPCC), published in Most plant breeders assume that it is
2007, indicates that the warming of the too slow and too difficult to breed for
climate system is unequivocal, as it is now environments where droughts or other
evident from observations of increases in stresses are unpredictable and variable. The
global average air and ocean temperatures, target is hard to define, and heritability, and
widespread melting of snow and ice, and hence response to selection, is too low to
rising global average sea level. This is achieve meaningful results. Therefore most
shown by (i) 11 of the last 12 years (1995 of the breeding for stress environments has
2006) rank among the twelve warmest been actually conducted using the same
years in the instrumental record of global basic approach that has been very successful
surface temperature (since 1850); (ii) the in areas where lack of water or other abiotic
temperature increase is widespread over stresses is seldom important.
the globe, and is greater at higher northern With few exceptions, most breeding
latitudes; (iii) global average sea level has programmes share the following concepts:
risen since 1961 at an average rate of selection has to be conducted under the
1.8 mm/yr, and since 1993 at 3.1 mm/ well-managed conditions of research
yr, with contributions from thermal stations. It is felt that environmental
expansion, and melting glaciers, ice caps noises can be kept under control, error
and the polar ice sheets; and (iv) observed variances are smaller and response to
decreases in snow and ice extent are also selection higher;
consistent with warming. Satellite data cultivars must be genetically homogenous
since 1978 show that annual average Arctic (pure lines, hybrids, clones) and must be
sea ice extent has shrunk by 2.7 percent per widely-adapted over large geographical
decade, with larger decreases in summer of areas;
7.4 percent per decade. Mountain glaciers locally-adapted landraces must be
and snow cover on average have declined replaced because they are low yielding
in both hemispheres. and disease susceptible;
The 2007 report indicates that it is also seed of improved cultivars must be
very likely that over the past 50 years, cold disseminated through mechanisms
days, cold nights and frosts have become and institutions such as variety release
less frequent over most land areas, and hot committees, seed certification schemes
days and hot nights have become more and governmental seed production
frequent, and it is likely that heat waves organizations; and
396 Plant breeding and farmer participation
the end users of new varieties are may be very different from those of
not involved in selection and testing; the breeder (Hardon and de Boef, 1993;
they are only involved at the end of Sperling, Loevinsohn and Ntabomvura,
the consolidated routine (breeding, 1993). Typical examples are crops used
researcher-managed trials, verification as animal feed, such as barley, where
trials), to verify if the choices made for breeders often use grain yield as the
them by others are appropriate or not. sole selection criterion, while farmers are
Breeders have very seldom questioned usually equally concerned with forage
these assumptions. When they have, it has yield and the palatability of both grain
been found that: and straw.
selection in well-managed research Although the chapter is largely based on
stations tends to produce cultivars that the strategies and methodologies developed
are superior to local landraces only under during the last 20 years in the ICARDA
improved managementnot under the barley breeding programme, we believe that
low-input conditions typical of the the main findings have general applicability.
farming systems of stress environments. They will be described to demonstrate
The result is that although many new that it is indeed possible to improve the
varieties outyield local landraces on a production of a typically low-input crop
research station and some are released, such as barley, grown in environments with
few if any are actually grown by farmers low and poorly-distributed rainfall, low
in difficult environments; temperatures in winter, high temperatures
poor farmers in stress environments tend and drought during grain filling, low soil
to maintain genetic diversity in the form fertility and poor agronomic management.
of different crops, different cultivars The data were mainly obtained from
within the same crop or heterogeneous three locations in the northern Syrian
cultivars, or combinations, to maximize Arab Republic (Tel Hadya, Breda and
adaptation over time (stability), rather than Bouider). They represent three distinct
adaptation over space (Martin and Adams, agricultural systems. Tel Hadya (348 mm
1987a). Diversity and heterogeneity serve average annual rainfall) is a favourable
to disperse or buffer the risk of total crop high-input environment that lends itself to
failure due to environmental variation. a wide choice of different crops. Bouider
This is in sharp contrast to the trend of (236 mm average annual rainfall) represents
modern breeding towards uniformity; the opposite extreme: a typical low-input,
resource-poor farmers seldom use the high-risk environment where barley is the
formal seed-supply systems. They only possible rainfed crop. Breda (273 mm
frequently rely on their own or on average annual rainfall) is intermediate
neighbours seed (Almekinders, Louwaars between the two, located on the edge of
and de Bruijn, 1994). Therefore, when the the area where Arabi Aswad becomes the
appropriate cultivar is selected, adoption dominant landrace. The three sites are
is much faster through non-market geographically close, located 35 (Tel Hadya),
methods of seed distribution (Grisley, 60 (Breda) and 80 km (Bouider) south-
1993); and east of Aleppo. The key aspects of these
when farmers are involved in the strategies and methodologies are: (i) direct
selection process, their selection criteria selection for specific adaptation in the target
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 397
TABLE 16.1
Grain yield (t/ha) of Tadmor and Zanbaka in 11 and 8 locations respectively in the northern Syrian
Arab Republic
Year Location (Province) Arabi Aswad Tadmor Zanbaka
Notes: The mean in parentheses is calculated from the locations in common with Zanbaka. The data are from trials
conducted in farmers fields, without fertilizer.
TABLE 16.2
Grain yield (t/ha) of Arta in 51 locations over seven cropping seasons
Year No. of sites cv. Arta cv. Arabi Abiad % increase
Note: The data are from trials on farmers fields in the Syrian Arab Republic (except those of 19861987, which are from
Breda research station).
FIGURE 16.1
Grain yields (t/ha) of pure lines derived from Syrian landraces and modern cultivars
at three levels of stress in the Northern Syrian Arab Republic
7
6.2
6
5 4.5
4
3.3
t /ha
3.1
3
2
1
1 0.6
Level of stress
Modern Landraces
FIGURE 16.2
Frequency of selection by farmers of four types of germplasm (modern, landraces,
landraces modern and landraces Hordeum spontaneum)
80
70
60
50
Percentage
40
30
20
10
0
Tel Brack Bylounan J. Aswad B. Sharky
Locations
TABLE 16.3
Gas exchange parameters of Hordeum spontaneum accessions (means of 12 accessions) and ratio
H. spontaneum/H. vulgare at Tel Hadya, Syrian Arab Republic
Parameter Units H. spontaneum H. spontaneum/H. vulgare
higher net photosynthesis and lower pre- only 176 mm rainfall (Grando, von Bothmer
dawn leaf water potential at this stage of and Ceccarelli, 2001).
development than did cultivated barley These lines had some photosynthetic
(Table 16.3). The ability of some accessions activity early in the morning, even though
of H. spontaneum to tolerate extreme levels six times less than in absence of stress,
of drought stress was evident during the stomata were open and the pre-dawn leaf
severe drought of 1987, when two lines of water potential was negative. At the same
H. spontaneum were the only survivors in time, the stomata of the black-seeded local
the breeding nurseries grown at Bouider landrace, Arabi Aswad, considered by
(Syrian Arab Republic), which had received farmers to be very resistant to drought,
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 401
were closed, even though the pre-dawn leaf heads are grown as head rows and tested for
water potential was slightly higher than in disease resistance or quality characteristics.
H. spontaneum. By midday, the stomatal Some bulks will lose the superiority shown
conductance of H. spontaneum decreased the year before because of genotype
and net photosynthesis became negative, environment interaction and because of
while the stomatal conductance of Arabi decreasing heterozygosity and associated
Aswad was zero. reduced heterotic effects. The corresponding
families will also be discarded.
16.5 BREEDING METHODS The families deriving from the populations
Individual plant selection (such as in the that maintained their superiority for three
widely used pedigree method) in crops cropping seasons will enter yield testing.
that are normally grown in dense stands When the programme is fully
is very effective for traits that are not implemented, the yield trials contain two
affected by competition. The best example types of materials: new bulks, and pure
is probably disease resistance. However, lines derived from the superior bulks of
many characters of interest to the breeder the previous cycle. If the requirements
are strongly affected by competition. for the genetic uniformity of the varieties
Among others, the ability to tolerate water to be released in a given country are very
stress is certainly greatly affected by the strict, only the pure lines will considered as
distance between plants competing for candidates for release.
limited available water. The result is that The method is based on the basic
isolated plants (such as those of a spaced- assumptions that (i) a superior bulk is
plant F2 used in the pedigree method) grow made by a large number of superior
much better than they would if planted at genotypes, and (ii) that if the superiority is
normal density. The argument that this does maintained for a period of three cropping
not matter as long as all plants are in the seasons in a highly variable environment,
same conditions ignores the possible effects the probability is small that the superiority
of genotype competition interaction. is associated with heterosis. The method
One breeding method that, in the case can also be used to test the importance
of self-pollinated crops, seems particularly of population buffering in relation to
suitable to breeding for resistance to abiotic stability.
stresses is the bulk-pedigree method, in The method is based on the exploitation
which, after producing the F1 and the F2 on of the genetic variance between
station, three years of multilocation yield populations (Vb) because estimates of Vb
testing and selection of the bulks are carried are comparatively easy and economical to
out in the target environment(s). Selection obtain, while estimates of within-population
is done between bulks by identifying the variance (Vw) are more expensive and
best populations for either yield or other much less precise because of interaction
characters. In parallel with the field testing and competitive effects (Simmonds, pers.
of the bulks, a within-bulks selection is comm.).
conducted only in those bulks that are This method has proved to be ideal for
selected for the next level of field testing: 10 use in participatory breeding programmes
to 50 heads are collected from the selected with self-pollinated crops (Ceccarelli and
populations. The progenies of the selected Grando, 2007).
402 Plant breeding and farmer participation
diploid crops (Allard and Bradshaw, 1964). that, during millennia of cultivation under
However, as modern varieties of cereal adverse conditions, natural and artificial
crops such as wheat and barley are mostly selection have not been able to identify
pure lines, they must rely on individual either an individual genotype possessing
buffering to be stable. Population buffering a key trait associated with its superior
is a mechanism of stability associated with performance, or an individual genotype
genetic heterogeneity. Varieties made with a specific architecture of different
up of a number of genotypes, such as traits. On the contrary, the combined
the landraces, are well buffered (stable), effects of natural and artificial selection
because each member of the population is has led to diversity in architecture of
best adapted to slightly different conditions genotypes, representing different
from other members of the population. The combinations of traits. These populations
stability of the individuals is sacrificed to can be extremely useful for understanding
maximize the stability of the population. mechanisms that enhance stability in stress
Although a direct relationship between environments, not only from the genetic
genetic heterogeneity and stability has yet structure point of view, but also for
to be demonstrated for landraces, it can understanding the adaptive role of given
be speculated that, being the product of traits. In fact, although variable, landraces
natural and artificial selection following grown in environments characterized by
domestication, the genetic structure of a high frequency of stress conditions tend
landraces must bear some advantage, or at to present a high frequency of a given
least cannot be a purely random outcome. expression of specific traits.
The genetic structure of landraces, For example, barley lines extracted from
therefore, may be considered an landraces collected from five sites in the
evolutionary approach to survival and Syrian steppe (Table 16.4) were compared
performance under arid and semi-arid with barley lines extracted from landraces
conditions (Schulze, 1988). It follows collected in Jordan and with a wide range of
TABLE 16.4
Mean of morphological and developmental traits in 1041 modern barley genotypes (unrelated to
Syrian or Jordanian landraces) compared with 322 pure lines extracted from Syrian landraces and
232 pure lines from Jordanian landraces
Traits Modern (n=1041) Landraces
Syrian Arab Republic
Jordan (n=232)
(n=322)
1. Early growth vigour 2.5 b 3.2 a 2.4 b
2. Growth habit 2.8 c 4.0 a 3.1 b
3. Cold tolerance 3.0 a 1.3 c 2.3 b
4. Days to heading 117.9 b 121.2 a 116.9 c
5. Grain filling 39.3 a 35.5 c 37.4 b
6. YP (t/ha) 4.398 a 3.293 c 3.947 b
7. YD (t/ha) 0.483 c 0.984 a 0.835 b
Notes: (i) Traits 1, 2, 4, 5 & 6 were scored or measured at Tel Hadya in 1987/88 (504.2 mm rainfall); trait 3 was scored at
Bouider in 1987/88 (385.7 mm rainfall); and trait 7 was measured at Bouider in 1988/89 (189 mm rainfall), on 521 modern
lines, 92 Syrian landraces, and 86 Jordanian landraces. Early growth vigour (1=good; 5=poor), Growth habit (1=erect;
5=prostrate), Days to heading (days from emergence to awn appearance), Grain filling duration (days between heading and
maturity), YP = Yield Potential, YD = Yield under Drought.
(ii) Means followed by the same letter are not significantly (P<0.05) different based on t-test for samples of unequal size.
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 405
modern (non-landrace) barley genotypes. each of these traits. The variability around a
The Syrian lines showed a higher frequency mean expression of each characterwhich
of genotypes with prostrate or semi- already allows a high degree of adaptation
prostrate growth habit, cold tolerance might perhaps be considered as a fine-tuning
and short grain-filling period, and a lower mechanism to cope with environmental
frequency of genotypes with good growth fluctuations. Thus, 321 lines derived from
vigour and early heading. Their average Syrian landraces were classified according
grain yield in unfavourable conditions (at to the score for early growth vigour in
Bouider in 1989) was 0.984 t/ha (ranging three classes: good vigour (score <2.5);
from 0.581 to 1.394 t/ha), more than twice intermediate (score = 2.53.5); and poor
the average grain yield of modern genotypes vigour (score >3.5). Each class was then
(0.483 t/ha, ranging from crop failure to classified according to the score for growth
1.193 t/ha). The average yield in favourable habit (erect <2.5; semi-prostrate = 2.53.5;
conditions of the Syrian landraces (3.293 t/ prostrate >3.5). No genotypes were found
ha) was 75 percent of the average yield in the good vigour-erect, intermediate
in favourable conditions of the modern vigour-erect, poor vigour-erect, and poor
germplasm (4.398 t/ha). vigour-semi-prostrate classes (Table 16.5).
Although this particular set of data is The groups were compared not only for
based on one environment only, it confirms the two traits used in their classification,
the existence of the trade-off between yield but also for days to heading, cold tolerance
in unfavourable conditions and yield in and length of the grain-filling period. Lines
favourable conditions discussed earlier. with good early growth vigour tend to be
Landraces collected in Jordan, from sites less cold tolerant, earlier and with a longer
with milder winters than the Syrian steppe, grain-filling period. This small percentage
have a higher frequency of genotypes that of genotypes will presumably have a yield
have better early growth vigour, more erect advantage in years with slightly milder
habit, less cold tolerance, slightly longer winter temperatures, absence of late frosts
grain-filling period and earlier heading and less severe terminal stress. The highest
than Syrian landraces. Their average frequency of genotypes (71.3 percent)
grain yield in unfavourable conditions combines intermediate early-growth vigour
was only slightly lower (0.835 t/ha) than with semi-prostrate or prostrate growth
Syrian landraces, while their average yield habit. These genotypes are slightly more
in favourable conditions (3.947 t/ha) was cold tolerant than the first group, but
in between the Syrian landraces and the are slightly later in heading. However,
modern germplasm. Syrian landraces they are better equipped to escape terminal
therefore show a combination of escape drought because of the shorter grain-
(early maturity) and avoidance (prostrate filling period. About one-quarter of the
habit and cold tolerance result in good genotypes (22.1 percent) have poor early
ground cover) mechanisms. growth vigour but a very prostrate growth
In addition to the high frequency of habit (growth habit score = 4.2) and a high
combinations of escape and avoidance level of cold tolerance (1.3). Their slightly,
traits, landraces possess another powerful although significantly, later heading is not
mechanism. They are composed of a number necessarily a negative attribute, mostly
of genotypes with a variable expression for because it is compensated for by a very
406 Plant breeding and farmer participation
TABLE 16.5
Frequency of different combinations of early growth vigor (GV), and growth habit (GH), and mean
values of cold tolerance (CT), days to heading (DH) and length of the grain filling period (GF) in a
sample of 322 lines of barley collected in the dry areas of the Syrian Arab Republic (from same trials
as indicated in notes of Table 16.4)
Groups % GV GH CT DH GF
Notes: All collection sites are included in the Palmyra region, as defined by Weltzien (1988).
the trials are grown in farmers fields using is gained as part of the selection process.
the host farmers agronomic practices; Last, but not least, the public investment in
selection is conducted by farmers in seed production is nearly always paid off
farmers fields, so that farmers are the by farmers adoption.
key decision-makers; and The programme started in the Syrian
the traditional linear sequence of Scientist Arab Republic in 1996 and was expanded to
Extension Farmers is replaced by a Algeria, Egypt, Eritrea, Islamic Republic of
team approach, with Scientists, Extension Iran, Jordan, Morocco, Tunisia and Yemen,
Staff and Farmers participating in all using the same bulk-pedigree method
major steps of variety development (see described earlier. Four types of impact can
Figure 9.1). be observed, considered below.
In a conventional breeding programme,
the most promising lines are released as Variety development
varieties, their seed is produced under New varieties were spontaneously
controlled conditions (certified seed) and disseminated from farmer to farmer as
only then can farmers decide whether to early as three years after starting the
adopt them or not. In many developing programme. In the Syrian Arab Republic,
countries the process results in many several thousand hectares are planted with
varieties being released but only a small two varieties, and 12 varieties have been
fraction being adopted. The major adopted by farmers and are under seed
consequence of the PPB concept is that multiplication (Table 16.6). Varieties are
the process transforms the delivery phase adopted both in dry areas and in wetter areas
of a plant breeding programme from being in a much shorter time than in a conventional
supply driven to being demand driven. breeding programme. It also confirms the
Under PPB, it is the initial farmers importance of landraces (Tadmor, Arta,
preference that drives the decision of which SLB and JLB lines, Zanbaka, A. Abiad
variety to release. As a consequence, adoption and A. Aswad) as well as H. spontaneum
rates are higher, and risks are minimized, as when farmers opinion becomes part of the
intimate knowledge of varietal performance breeding process.
TABLE 16.6
Varieties adopted from the PPB programme by farmers in the Syrian Arab Republic
Pedigree Name Location Rainfall
FIGURE 16.3
Grain yield (t/ha) of two lines with improved drought resistance in comparison
with the local landrace tested in 2004, 2005, and in 2006 in rainfed location in
the Syrian Arab Republic receiving less than 200 mm of total rainfall
1.2
0.8
t/ha
0.6
0.4
0.2
0
2004-1 2004-2 2005 2006-1 2006-2 mean 2004-1 2004-2 2005 2006-1 2006-2 mean
Improved Local
Line 1=H.spont.41-1/Tadmor//SLB45-090/H.spont.41-2/3/H.spont.41-1/Tadmor//H.spont.41-1/Tadmor
Line 2=Arta/3/Arar/H.spont.19-15//Hml
410 Plant breeding and farmer participation
below average and crop yields were severely and between 0.5 and 3.0 t/ha of biomass
affected. In some areas, the rainfall was so yield (Ceccarelli et al., 2004).
low that the crop did not even germinate; in Two of these lines were tested by farmers
many others the crop failed to produce grain. on large areas (520 ha) in 2004, 2005 and
The PPB trials, planted in eight farmers 2006, which were all very dry (Figure 16.3).
fields in the Syrian Arab Republic, were In comparison with the local landrace,
affected by different intensities of drought. which itself is considered to be drought
At one extreme, the rainfall was only 50 mm resistant, the two lines showed an average
in the entire season and no germination yield advantage of 44 percent, ranging from
occurred. At the other extreme, the rainfall 9 percent to 67 percent. This includes one
was 252 mm rainfall and average grain yield case in which the local landrace failed, and
was 1.8 t/ha (ranging from 1.0 to 3.2 t/ha). one of the improved lines yielded 0.5 t/
The driest sites, where some new barley ha. Both lines are derived from crosses
entries were able to produce some grain with a pure line of H. spontaneum (the
or some biomass, received between 87 and wild progenitor of cultivated barley), an
130 mm. Average grain and biomass yield indication that some H. spontaneum lines
were very low but some lines were able to can contribute significantly to enhance the
produce between 0.3 and 0.5 t/ha of grain drought resistance of cultivated barley.
TABLE 16.7
Rainfall (mm) and average yield (in parenthesis in t/ha) in five locations in the dry areas of the
Syrian Arab Republic during the three years 20032005
Location No. of lines 2003 2004 2005 Mean annual
rainfall
20002005 (mm)
TABLE 16.8
Grain yield (as percentage of the local check) of the highest yielding lines during 2003, 2004 and
2005 in five dry locations in the northern Syrian Arab Republic receiving between 186 and 308 mm
total annual rainfall
Location Line 2003 2004 2005 Mean
The second example derives from trials first and Al Bab and J. Aswad the second).
conducted in dry locations in the northern As these lines are the product of one cycle
Syrian Arab Republic during the period of selection, further progress is expected
20032005. Two locations (Bylounan and with additional cycles of recombination
Melabya) represent some of the driest areas and selection.
in the Syrian Arab Republic, where barley
is the only possible rainfed crop; J. Aswad 16.12 CONCLUDING REMARKS
and Siebatt are in slightly wetter areas, The objective of this chapter has been to
while Al Bab is a location characterized by discuss what plant breeders can do when
colder winters than the other four. In the the target environment of their breeding
two driest locations, rainfall varied from programme is characterized by chronic
176 mm to 245.3 mm total annual rainfall low yields due to numerous factors,
and average grain yield from 0.432 to 1.496 t/ such as climatic, nutritional and abiotic
ha during the testing period. In the two stresses. The data are mostly derived from
wetter locations, rainfall varied from 215 barley and from one type of dry area (dry
to 319 mm total annual rainfall and average Mediterranean with cold winters and hot
grain yield from 0.478 to 1.406 t/ha, while summers, and crops grown on current
Al Bab was the wettest of the five locations rainfall). However, the paper illustrates
but not the highest yielding because of the some general concepts that, with some
low temperatures in winter (Table 16.7). modifications, could be useful in other
In the five locations, we tested between crops and in other types of dry area.
6 and 12 lines (including the checks) repre- The first concept is that in these
senting the result of two cycles of decentral- environments, climatic, nutritional and
ized participatory selection starting from a biotic stresses usually occur together
common set of 165 lines. The yield of the (though not necessarily all of them all of the
best lines is shown in Table 16.8, expressed time); and, so far, there is little substitute
as a percentage of the local check. for actually exposing the breeding material
At the two driest sites, Bylounan and to a real field situation. Although little
Melabya, five lines outyielded the local practiced, the idea is not new. Nearly
check on average over 3 years by between forty years ago Hurd (1971) published
3.2 percent and nearly 9 percent, but only a paper with the title: Can We Breed for
three lines were consistently superior to the Drought Resistance? The first sentence of
local check in each of the three years. In the the paper was My answer to the above
two wetter locations, five lines outyielded very pertinent question is a confident and
the local check by between 7.5 percent optimistic Yes. He concluded: One
and 14.7 percent, but only the two lines method is to grow large populations in
in J. Aswad and two of the four lines in early generations under typical dry growing
Siebatt consistently outyielded the local conditions. Twenty years later, Bramel-
check. In Al Bab, two lines consistently Cox et al., (1991) recognized that the key
outyielded the local check by slightly more to increased production with fewer external
than 35 percent; two lines (ArabiAbiad/ inputs would be through a re-evaluation of
Arar//H.spont.41 and SLB28-53/SLB21- the identification and use of selection and
81) were among the highest yielding lines in testing environments.
two locations (Bylounan and Melabya the Although this concept is obvious to
412 Plant breeding and farmer participation
many and not new, selection for stress stress environment, it is probably for that
environments is still seldom done in the reason.
target environments and it still a highly The main conclusion of this paper is
controversial issue, as it is the relationship that breeding for stress environments is
between high yield under optimum possible, provided it is conducted with
conditions and high yield under abiotic strategies and methodologies that little have
stress conditions (see, for example, Chapter in common with those used in breeding
18 in this volume). This may not always for favourable environments. Adaptation
be necessarily a deliberate choice of one over time can be improved by breeding
breeding strategy or another, but is simply for specific adaptation to a given type of
due to the distance of suitable selection sites stress environment. This can be achieved by
from main cities, with all the associated taking advantage of the temporal variability
inconveniences. We hypothesize that in of stress environments, which permits
these cases an interesting solution may exposure of the same breeding material to
be offered by farmers participation in variable combinations of stresses over a
breeding (Ceccarelli and Grando, 1997; (relatively) short period and to accumulate
Ceccarelli, Grando and Baum, 2007). favourable alleles at the several loci involved
Conducting selection in farmers fields in drought resistance through successive
has the advantage of exploiting genetic cycles of recombination and selection.
differences under farm conditions, with the We are aware that this is fundamentally
additional advantage of making use of the different from the modern trend of plant
farmers knowledge of the crop. breeding towards broad adaptation over
The second concept is the use of space. The difference represents the
germplasm usually ignored by most plant contrasting interests of farmers and seed
breeders, such as landraces and wild relatives. companies. Farmers are interested in
This approach is a direct consequence of cultivars that are consistently superior on
choosing to work in the target environment their farm, regardless of how they perform
and has led to the development of a number at other locations or in other countries.
of barley cultivars, now grown in a number Seed companies, however, want to market
of farmers fields in the central and northern as much seed of as few cultivars as possible.
Syrian Arab Republic and in environments Breeders have been breeding, perhaps
considered too difficult and therefore unconsciously, more for seed companies and
beyond the plant breeders domain. for their personal prestige than for farmers.
The third concept is that in dry areas, The two objectives coincide when selection
every effort should be made to control and target environments are similar, but
environmental variability in trial and this approach has by-passed millions of
nurseries evaluation. When working at small farmers in difficult environments.
stress sites, the breeder should forget the Recent advances in plant genomics have
typical research-station style of work. The enabled one to dissect various molecular
methodology, experiment designs and plot mechanisms (signal transduction pathways)
techniques used in the very homogeneous involved in drought, cold and salt stress
environment of the experiment station are tolerance and in identifying various
not suitable; in fact, when the conclusion is genes involved in such stress tolerance.
reached that progress cannot be made in a Information generated in genomics should
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 413
be integrated into practical plant breeding. Austin, R.B., Ford, M.A. & Morgan, C.L. 1989.
Various genes identified, in both model Genetic improvement in the yield of win-
plants and crop plants, could be used in ter wheat: a further evaluation. Journal of
future for developing stress-tolerant plants Agricultural Science, Cambridge, 112: 295
through either marker-assisted selection or 301.
direct gene transfer. Blum, A. 1988. Plant breeding for stress envi-
ronments. Boca Raton, USA, CRC Press.
ACKNOWLEDGEMENTS Blum, A. 1993. Selection for sustained pro-
This chapter is based on the work carried duction in water-deficit environments. In
out at ICARDA over more than twenty International Crop Science I, pp. 343347.
years. Many people have made this work Madison, USA, CSSA.
possible, and we would like to acknowledge Blum, A., Golan, G., Mayer, J., Sinmena, B.,
the technical assistance of Mr A. Ayyan, Mr Shpiler, L. & Burra, J. 1989. The drought
R. Azzo, Mr M. Hamzeh, Mr G. Kashour, response of landraces of wheat from the
Mr A. Khorea, Mr M. Michael and Mr northern Negev Desert in Israel. Euphytica,
H. Pashayany. The work, dedication and 43: 8796.
support of many farmers in Syria are Blum, A., Golan, G. & Mayer, J. 1991. Progress
gratefully acknowledged. achieved by breeding open-pollinated
We thank Der Bundesminister fr cultivars as compared with landraces of
Wirtschaftliche Zusammenarbeit (BMZ), sorghum. Journal of Agricultural Science,
the Italian Government, the International Cambridge, 117: 307312.
Development Research Centre (IDRC) Bramel-Cox, P.J., Barker, T., Zavala-Garcia,
and the OPEC Fund for International F. & Eastin, J.D. 1991. Selection and test-
Development for their financial support. ing environments for improved perform-
ance under reduced-input conditions. In
REFERENCES Plant breeding and sustainable agriculture:
Acevedo, E. & Ceccarelli, S. 1989. Role of phys- considerations for objectives and meth-
iologist-breeder in a breeding programme ods, pp. 2956. CSSA Special Publication,
for drought resistance conditions. In F.W.G. No. 18. Madison, USA, CSSA.
Baker, ed. Drought Resistance in Cereals, Brown, A.D.H. 1978. Isozymes, plant popula-
pp. 117140. Wallingford, UK, CABI. tion genetic structure, and genetic conser-
Allard, R.W. & Bradshaw, A.D. 1964. vation. Theoretical and Applied Genetics,
Implications of genotype environment 52: 145157.
interactions in applied plant breeding. Crop Brown, A.D.H. 1979. Enzyme polymorphism
Science, 4: 503508. in plant populations. Theoretical Population
Almekinders, C.J.M., Louwaars, N.P. & de Biology, 15: 142.
Bruijn, G.H. 1994. Local seed systems and Brush, S.B. ed. 1999. Genes in the field: On-farm
their importance for an improved seed conservation of crop diversity. Boca Raton,
supply in developing countries. Euphytica, USA, IPGRI/IDRC/Lewis Publ.
78: 207216. Ceccarelli, S. 1984. Utilization of landraces
Asfaw, Z. 1989. Variation in hordein polypep- and H. spontaneum in barley breeding for
tide pattern within Ethiopian barley, dry areas. Rachis, 3(2): 811.
Hordeum vulgare L. (Poaceae). Hereditas, Ceccarelli, S. 1996. Positive interpretation of
110: 185191. genotype by environment interactions in
414 Plant breeding and farmer participation
relation to sustainability and biodiversity. Ceccarelli, S., Grando, S. & van Leur, J.A.G.
In M. Cooper & G.L. Hammers, eds. Plant 1987. Genetic diversity in barley landraces
adaptation and crop improvement, pp. 467- from Syria and Jordan. Euphytica, 36: 389
486. Wallingford. UK, CAB International, 405.
Andra Pradesh, India, ICRISAT and Ceccarelli, S., Grando, S., Baum, M. & Udupa,
Manila, IRRI. S.M. 2004. Breeding for drought resist-
Ceccarelli, S., Acevedo, E. & Grando, S. 1991. ance in a changing climate. In S.C. Rao
Breeding for yield stability in unpredict- & J. Ryan, eds. Challenges and strategies
able environments: single traits, interaction for dryland agriculture, pp. 167190. CSSA
between traits, and architecture of geno- Special Publication No. 32. Madison, USA,
types. Euphytica, 56: 169185. ASA/CSSA.
Ceccarelli, S. & Grando, S. 1996. Drought Chapman, S.C. 2008. Use of Crop Models
as a challenge for the plant breeder. Plant to understand genotype by environment
Growth Regulation, 20: 149155. interactions for drought in real-world and
Ceccarelli, S. & Grando, S. 1997. Increasing the simulated plant breeding trials. Euphytica,
efficiency of breeding through farmer par- 161: 195208.
ticipation. In Ethics and equity in conserva- Cleveland, D.A., Soleri, D. & Smith, S.E. 2000.
tion and use of genetic resources for sustain- A biological framework for understanding
able food security, pp. 116121. Proceeding farmers plant breeding. Economic Botany,
of a Workshop to develop guidelines for 54: 377390.
the CGIAR, Foz do Iguasu, Brazil, 2125 Comadran, J., Russell, J.R., van Eeuwijk, F.,
April 1997. Ceccarelli, S., Grando, S., Stanca, A.M.,
Ceccarelli, S. & Grando, S. 1999. Barley lan- Francia, E., Pecchioni, N., Akar, T.,
draces from the Fertile Crescent: a lesson Al-Yassin, A., Benbelkacem, A., Choumane,
for plant breeders. In S.B. Brush, ed. Genes W., Ouabbou, H., Bort, J., Araus, J.-L.,
in the field: on-farm conservation of crop Pswarayi, A., Romagosa, I., Hackett, C.A.
diversity, pp. 5176. Boca Raton, USA, & Thomas, W.T.B. 2008. Mapping adapta-
IPGRI/IDRC/Lewis Publishers tion of barley to droughted environments.
Ceccarelli, S. & Grando, S. 2002. Plant breed- Euphytica, 161: 3545.
ing with farmers requires testing the Cooper, P.J.M., Gregory, P.J., Tully, D. & Harris,
assumptions of conventional plant breed- H.C. 1987. Improving water use efficiency of
ing: Lessons from the ICARDA barley annual crops in the rainfed farming systems
program. In D.A. Cleveland & D. Soleri, of West Asia and North Africa. Experimental
eds. Farmers, scientists and plant breed- Agriculture, 23: 113158.
ing: integrating knowledge and practice, Cox, T.S., Shroyer, J.P., Liu, B.H., Sears, R.G.
pp. 297332. Wallingford, UK, CABI. & Martin, T.J. 1988. Genetic improvement
Ceccarelli, S. & Grando, S. 2007. Decentralized- in agronomic traits of hard red winter
participatory plant breeding: an example wheat cultivars from 1919 to 1987. Crop
of demand driven research. Euphytica, Science, 28: 756760.
155: 349360. Damania, A.B., Jackson, M.T. & Porceddu,
Ceccarelli, S., Grando, S. & Baum, M. 2007. E. 1985. Variation in wheat and barley
Participatory plant breeding in water-limit- landraces from Nepal and the Yemen Arab
ed environments. Experimental Agriculture, Republic. Zeitschrift fr Pflanzenzuchtung,
43: 411435. 94: 1324.
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 415
Damania, A. & Porceddu, E. 1983. Variation Wellard & A. Bebbington, eds. Cultivating
in landraces of Turgidum and Bread Wheats Knowledge. Genetic diversity, farmer exper-
and sampling strategies for collecting wheat imentation and crop research, pp. 6471.
genetic resources. In Proceedings of the 6th Int. Techn. Publ. Ltd.
International Wheat Genetic Symposium, Humphreys, M.O. & Humphreys, M.W. 2005.
Kyoto, Japan, 28 November3 December Breeding for stress resistance: general prin-
1983, pp. 123136. ciples. In M. Ashraf & P.J.C. Harris, eds.
Davis, J. 1990. Breeding for intercrops-with Abiotic stresses: plant resistance through
special attention to beans for intercrop- breeding and molecular approaches,
ping with maize. In S.R. Waddington, pp. 1946. New York, USA, Food Products
A.F.E. Palmer & O.T. Edge, eds. Research Press.
Methods for Cereal/Legume Intercropping. Hurd, F.A. 1971. Can we breed for drought
Proceedings of a Workshop on Research resistance? In K.L. Larson & J.D. Eastin,
Methods for Cereal Legume Intercropping eds. Drought injury and resistance in crops,
in Eastern and Southern Africa, Lilongwe, pp. 7788. CSSA Special Publication, No. 2.
Malawi, 2327 January 1988. Madison, USA, CSSA/ASA.
Dudal, R. 1976. Inventory of the major soils of Kasuga, M., Liu, Q., Miura, S., Yamaguchi-
the world with special reference to mineral Shinozaki, K. & Shinozaki, K. 1999.
stress hazards. In M.J. Wright, ed. Plant Improving plant drought, salt, and freez-
Adaptation to Mineral Stress in Problem Soils, ing tolerance by gene transfer of a single
pp. 314. Ithaca, USA, Cornell University stress-inducible transcription factor. Nature
Agricultural Experiment Station. Biotechnology, 17: 287291.
Elings, A. & Nachit, M.M. 1991. Durum wheat Lagudah, E.S., Flood, R.G. & Halloran, G.M.
landraces from Syria. I. Agroecological and 1987. Variation in high molecular weight
morphological characterization. Euphytica, glutenin subunits in landraces of hexaploid
53: 211224. wheat from Afghanistan. Euphytica, 36: 39.
Erskine, W. & Choudhary, N.A. 1986. Variation Martin, G.B. & Adams, W.M. 1987a. Landraces
between and within lentil landraces from of Phaseolus vulgaris (Fabaceae) in Northern
Yemen Arab Republic. Euphytica, 35: 695 Malawi. I. Regional variation. Economic
700. Botany, 41: 190203.
Grando, S., von Bothmer, R. & Ceccarelli, S. Martin, G.B. & Adams, W.M. 1987b. Landraces
2001. Genetic diversity of barley: Use of of Phaseolus vulgaris (Fabaceae) in Northern
locally adapted germplasm to enhance yield Malawi. II. Generation and maintanance of
and yield stability of barley in dry areas. In variability. Economic Botany, 41: 204215.
H.D. Cooper, C. Spillane & T. Hodgkin, eds. Nakashima, K., Shinwari, Z.K., Sakuma, Y., Seki,
Broadening the genetic base of crop produc- M., Miura, S., Shinozaki, K. & Yamaguchi-
tion, pp. 351-372. Wallingford, UK, CABI. Shinozaki, K. 2000. Organization and
Grisley, W. 1993. Seed for bean production in expression of two Arabidopsis DREB2
sub-Saharan Africa: issues, problems, and genes encoding DRE-binding proteins
possible solutions. Agricultural Systems, involved in dehydration- and high-salinity-
43: 1933. responsive gene expression. Plant Molecular
Hardon, J.J. & de Boef, W.S. 1993. Linking Biology, 42: 657665.
farmers and breeders in local crop devel- Parry, M.A.J., Flexas, J. & Medrano H. 2005.
opment. In W. de Boef, K. Amanor, K. Prospects for crop production under
416 Plant breeding and farmer participation
drought: research priorities and future of water use and crop yield in temperate
directions. Annals of Applied Biology, cereals. Crop Science, 42: 111121.
147: 211226. Russell, W.A. 1984. Agronomic performance
Passioura, J.B. 1986. Resistance to drought of maize cultivars representing different
and salinity: avenues for improvements. eras of breeding. Maydica, 29: 375390.
Australian Journal of Plant Physiology, Schulze, E.D. 1988. Adaptation mechanisms
13: 191201. of non-cultivated arid-zone plants: useful
Passioura, J.B. 2002. Environmental biology lesson for agriculture? In F.R. Bidinger &
and crop improvement. Functional Plant C. Johansen, eds. Drought research pri-
Biology, 29: 537546. orities for the dryland tropics, pp. 159177.
Passioura, J.B. 2006. Increasing crop produc- Patancheru, India, ICRISAT.
tivity when water is scarcefrom breeding Seki, M., Narusaka, M., Abe, H., Kasuga,
to field management. Agricultural Water M., Yamaguchi-Shinozaki, K., Carninci,
Management, 80: 176196. P., Hayashizaki, Y. & Shinozaki, K. 2001.
Porceddu, E. & Scarascia Mugnozza, G.T.1984. Monitoring the expression pattern of 1300
Genetic variation in durum wheat. In Arabidopsis genes under drought and
Proceedings of the 6th International Wheat cold stresses by using a full-length cDNA
Genetic Symposium, Kyoto, Japan, 28 microarray. Plant Cell, 13: 6172.
November3 December 1983, pp. 253264. Serraj, R. & Sinclair, T.R. 2002. Osmolyte
Pswarayi, A., van Eeuwijk, F., Ceccarelli, S., accumulation: Can it really help increase
Grando, S., Comadran, J., Russell, JR., Francia, crop yield under drought conditions? Plant
E., Pecchioni, N., Li Destri, O., Akar, T., Cell and Environment, 25: 333341.
Al-Yassin, A., Benbelkacem, A., Choumane, Shafiq-ur-Rehman, Harris, P.J.C. & Ashraf
W., Karrou, M., Ouabbou, H., Bort, J., Araus, M. 2005 Stress environments and their
J.L., Molina-Cano, J.L., Thomas, W.T.B. & impact on crop production. In M. Ashraf
Romagosa, I. 2008. Barley adaptation and & P.J.C. Harris, eds. Abiotic stresses: plant
improvement in the Mediterranean basin. resistance through breeding and molecular
Plant Breeding, 127(6): 554560. approaches, pp. 318. New York, USA,
Ramanjulu, S. & Bartels, D. 2002. Drought- Food Products Press.
and desiccation-induced modulation of Simmonds, N.W. 1983. Plant Breeding: The
gene expression in plants. Plant Cell and state of the art. In T. Kosuge, C.P. Meredith
Environment, 25: 141151. & A. Hollaender, eds. Genetic engineer-
Rasmusson, D.C. 1987. An evaluation of ing of plants. An agricultural perspective,
ideotype breeding. Crop Science, 27: 1141 pp. 525. New York, USA, Plenum Press.
1146. Simmonds, N.W. 1984. Decentralized selec-
Reynolds, M.P., Mujeeb-Kazi, A. & Sawkins, tion. Sugar Cane, 6: 810.
M. 2005. Prospects for utilising plant-adap- Simmonds, N.W. 1991. Selection for local
tive mechanisms to improve wheat and adaptation in a plant breeding programme.
other crops in drought- and salinity-prone Theoretical and Applied Genetics, 82: 363
environments. Annals of Applied Biology, 367.
146: 239259. Sinclair, T.R. 2000. Model analysis of plant
Richards, R.A., Rebetzke, G.J., Condon, A.G. traits leading to prolonged survival dur-
& van Herwaarden, A.F. 2002. Breeding ing severe drought. Field Crops Research,
opportunities for increasing the efficiency 68: 211217.
Breeding for quantitative variables. Part 3: Breeding for resistance to abiotic stresses 417
CHAPTER 17
breeding staple crops such as rice, wheat, resource-poor nations will also be discussed.
barley, maize, cassava, potatoes and beans Sufficient genotypic variation in the trait
with the ability to fortify themselves with to be selected is necessary for conventional
micronutrients. It offers a sustainable, cost- breeding to be feasible, so we will discuss
effective alternative to other strategies such this as a first step, with reference to the five
as individual supplementation and fertiliza- key micronutrients. Breeding principles dis-
tion, which is more likely to reach those cussed in this chapter are applicable to both
most in need and has the added advantage traditional and participatory plant breeding.
of requiring no change in current consumer
behaviour to be effective (Graham, Welch 17.2 GENOTYPIC VARIATION OF
and Bouis, 2001). Once a one-off invest- MICRONUTRIENT CONCENTRATION IN
ment is made to breed bio-fortified seed, STAPLE FOOD CROPS
recurrent costs are low and germplasm can 17.2.1 Fe & Zn
be shared globally. Bio-fortification, com- Over the last decade, there have been consid-
mercial fortification and supplementation erable efforts in several international research
are complementary strategies for reaching centres such as the International Maize and
malnourished populations. Furthermore, Wheat Improvement Centre (CIMMYT) in
bio-fortification can increase farm produc- Mexico and the International Rice Research
tivity as certain micronutrients, such as Zn Institute (IRRI) in the Philippines, to iden-
and Se, that improve human nutrition can tify wheat and rice germplasm with high Fe
help plants resist diseases and other envi- and Zn concentration, which has also been
ronmental stresses (HarvestPlus, 2007). the subject of several reviews (Graham et
Breeding criteria for micronutrient- al., 1999; Rengel, Batten and Crowley, 1999;
enriched staple food crops have been Cakmak et al., 2000).
reviewed recently (Welch and Graham,
2004). These criteria include (i) maintaining Wheat
crop productivity, (ii) evidence for stability At CIMMYT, in one study, 170 wheat lines
of micronutrient enrichment traits across selected out of 550 initially screened lines
various edaphic and climatic zones, were grown in a replicated trial (Ortiz-
(iii) demonstration of significant effects of Monasterio and Graham, 2000). This
enriched micronutrients on human health, study identified three promising sources
(iv) demonstration of bio-availability of high grain Fe and Zn concentration:
of enriched micronutrients for human wild species, landraces and breeding lines.
nutrition, and (v) consumer acceptance. Fe concentration was in the range of 25
In this chapter, we will focus on genetic to 56 mg/kg dry weight (DW), while Zn
potential, genotype environment interac- concentration varied from 25 to 65 mg/kg
tions, screening protocols, breeding strat- DW. In a second study, a group of 154 lines
egies for enhancing grain micronutrient from the breeding programme were grown
accumulation. Physiological and molecular together. Lines were identified with up to
mechanisms of uptake, translocation and 73 mg Fe/kg DW and 92 mg Zn/kg DW.
deposition of micronutrients in the grains or Our group at the University of Adelaide
other edible parts of major staple food crops also observed significant variation in grain
such as wheat, rice, maize, beans and cassava, Zn and Fe concentrations in commercial
which are consumed by billions of people in cultivars and advanced breeding lines
422 Plant breeding and farmer participation
TABLE 17.1
Zn and Fe concentrations (mg/kg DW) in grains of durum and bread wheat genotypes grown in
standard potting mix with all nutrients supplied adequately in a glasshouse
No. of entries Fe Zn
Mean conc. (SD) Range Mean conc. (SD) Range
Modern bread wheat 25 36 (6) 2753 39 (8) 2553
(T. aestivum)
Synthetic hexaploid 36 41 (8) 3267 41 (9) 2866
wheat (T. aestivum)
Durum wheat 24 42 (7) 2956 51 (6) 3962
(T. dicoccon)
(T turgidum) 191 33 (8) 1762 30 (12) 1281
Note: Standard deviation, SD, is given in parentheses.
Source: Genc et al., unpublished.
TABLE 17.2
Zn and Fe concentrations (mg/kg DW) in grains of bread wheat genotypes in field trials in Australia
Location and year No of entries Fe Zn
Mean conc. (SD) Range Mean conc. (SD) Range
Birchip-2000 28 37 (3) 3141 16 (2) 1219
Birchip-1999 42 38 (2) 3242 25 (3) 2031
Birchip-1998 39 42 (4) 3655 23 (2) 1931
Horsham-1998 30 33 (3) 2740 16 (1) 1319
Bute-1997 42 32 (3) 2738 18 (2) 1521
Lameroo-1996 35 33 (3) 2739 20 (2) 1524
Note: Standard deviation, SD, is given in parentheses
Source: Graham et al., unpublished.
of wheat in glasshouse and field trials study of 825 accessions, including wild
conducted over a number of years in emmer wheat (Triticum turgidum subsp.
Australia (Tables 17.1 and 17.2). In general, dicoccoides), grain concentrations were 14
grain Zn and Fe concentrations were higher to 190 mg/kg DW and 15 to 109 mg/kg DW
in the glasshouse than in field studies, for Zn and Fe, respectively. In this study,
which can be attributed to better growing no yield data were reported, thus we do
conditions (well-watered and fertilized) in not know whether high concentrations are
the glasshouse than in the field. In the field associated with low yield. In the meantime,
studies, grain Zn concentration was in the despite lower Fe and Zn concentrations
range of 12 to 31 mg/kg DW, a narrower than in wild species, more screening is
range than that found at CIMMYT. This needed of elite germplasm (modern wheat
narrower range and also lower values (<15 genotypes and advanced breeding lines) for
mg/kg DW) in grain Zn concentration are high Fe and Zn concentration in the grain,
indicative of Zn deficiency in the field. as they have already improved agronomic
Grain Fe concentration varied from 27 performance (Graham et al., 1999).
to 55 mg/kg DW. Recently, a large-scale
screening by Cakmak et al. (2004) has Rice
identified wild wheat accessions with even There also exists considerable genotypic
higher Fe and Zn concentrations than those variation for grain Zn and Fe concentration
reported previously. In this comprehensive in rice. Researchers at IRRI and the
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 423
beta-carotene
8 alpha-carotene vitamin A deficiency-associated diseases.
There have been several reports of
6
genetic variation for carotenoids in maize.
4 One of the first was that of Brunsen and
Quackenbush (1962), who showed that the
2
total concentration of carotenoids varied
0 significantly between high- and low-carotene
6 9 13 15 18 21 23 27 30 33 36 39 42
inbred lines, with an even greater variation
Days post-anthesis between low and high for provitamin A
Source: Graham and Rosser, 2000. Reprinted with carotenoids. Blessin et al. (1963) reported a
permission from the UNU Food and Nutrition Bulletin
(2000) 21: 404409.
range of 0.9 to 4.1 mg/kg for carotenes, and
18.6 to 48 mg/kg for xanthophylls in 39 maize
inbreds. In the same year, Quackenbush
preservation of carotenoids present in et al. (1963) observed concentrations of
immature wheat. up to 7.3 mg/kg carotenes, and a range of
Selection against highly pigmented 2 to 33 mg/kg lutein in 125 inbred lines.
varieties in several staple foods has led to The value of extensive screening for high-
very little variation for this trait in modern accumulation varieties was indicated in the
cultivars. However, germplasm banks report of Egesel (1997), who found a range
where landraces and old varieties are stored of just 0.13 to 2.9 mg/kg -carotene in 200
hold the key to retrieving genetic variation. maize families. More recently, 16 yellow
These valuable resources can be used as seeded maize lines were reported to have
a source of variation for the introduction 143 to 278 mg/kg carotenoids (Maziya-
of desirable traits back into commercially Dixon et al., 2000). This study measured
profitable and locally grown varieties, for total carotenoid concentrations rather than
nutritional benefit. Reports from screening defining provitamin and non-provitamin
of genetic resources obtained from carotenoids, and although valuable for
germplasm banks indicate that many of the calculation of total carotenoid intake, gives
older varieties have substantial variation for no idea of the provitamin A potential of the
carotenoid concentration. cultivars.
Maize Wheat
White maize was previously highly desired Wheat is another staple food that has been
for reasons of cleanliness and apparent subjected to selection for colour, though
purity, while maize varieties with high with different outcomes depending on the
concentrations of pigmentation were used end use, either bread or pasta. Bread wheat
as stock feed. While people suffered from varieties (Triticum aestivum) have been
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 425
Carotenoids (g/kg)
selection against pigmentation has resulted 350 Pro VA carotenoids
300
in current cultivars containing very low 250
concentrations of carotenoids. In Australia, 200
there is an exception, a South Australian 150
100
wheat by the name of Krichauff, which has
50
relatively high concentrations of caroten- 0
oids in comparison with other current cul- BW1 BW2 BW3 BW4 BW5 DR1 DR2 DR3
provide an alternative to agronomic bio- in the trial. Other wild wheat relatives may
fortification and thus minimize the need for also be efficient accumulators of Se and
Se fertilizers (Broadley et al., 2006). other minerals (Piergiovanni et al., 1997).
Few data have been published on varietal Studies of genotypic variation of I in
differences for Se accumulation for most diverse plant species have yielded variable
crop species. However, in Lycopersicon findings. A Japanese survey found no sig-
(tomatoes and related plants), four-fold nificant difference in leaf I concentration
differences in shoot Se accumulation have for plants grown on similar soils (Yuita et
been found (Pezzarossa et al., 1999), and al., 1982), while an earlier survey of differ-
in Brassica (broccoli) a significant genotype ent pasture species grown together in the
effect for Se concentration in heads was field found a thirty-four-fold variation in
found in hybrids, but not inbreds. However, leaf I concentration, with perennial rye-
the effect of environment was around grass (Lolium perenne L.) the most efficient
ten times stronger than that for genotype I accumulator (Johnson and Butler, 1957).
(Farnham et al., 2007). Evidence is scarce for significant
genotypic variation in grain density of
Wheat I in wheat (Shinonaga et al., 2001), and
In surveys and trials conducted by our no variation was detected in our South
group, involving diverse wheat germplasm Australian trial, where varieties were grown
and a total of eleven datasets in South at three locations, with three replications. I
Australia and Mexico, grain Se concentra- concentrations were low, typically less than
tions were in the range of 5 to 720 g/kg 20 g/kg (range 1060 g/kg DW) in whole
DW, but much of this variation was associ- grain (Lyons et al., unpublished).
ated with spatial variation in soil-available
Se. South Australian soils are renowned Maize
for their microspatial variability, which Little research has been conducted on
makes detection of genotypic differences genotypic variation in Se or I concentration
in grain Se density difficult. No significant in maize kernels. Our group has conducted
genotypic variation in grain Se density a limited survey of diverse maize
among modern commercial bread or durum genotypes grown in the United States
wheat varieties was detected in this study of America (Illinois) and Nigeria. No
(Lyons et al., 2005), which agrees with significant variation was detected for either
earlier research (Noble and Barry, 1982; micronutrient; however, the soils at both
Grela, 1996; Tveitnes, Singh and Ruud, sites were low in available Se, resulting
1996). However, the ancient diploid wheat, in kernel Se levels at the Nigerian site,
Aegilops tauschii L. and rye (Secale cereale) for example, of just 5 g/kg DW (range
were found in our studies to be 42 percent 310 g/kg DW). Such low levels may have
and 35 percent higher (p < 0.001; p= 0.03), limited expression of possible varietal Se
respectively, in grain Se concentration than differences (Lyons et al., unpublished).
other cereals in separate field trials, and
in a hydroponic trial rye was 40 percent Rice
higher (P < 0.001) in foliar Se content than Rice appears to be a more promising cereal
two wheat landraces. Ae. tauschii was also for genotypic variation in Se, with differ-
higher in Zn, Fe and Mn than other wheats ences detected in other studies (Nan and
428 Plant breeding and farmer participation
components in endosperm than in bran (Kang, 1990; Hill, Becker and Tigerstedt,
(Lyons et al., 2005c). Further research with 1998; Annicchiarico, Chapter 20 this
a large number of genotypes is required to volume). From a breeding point of view, it
determine the extent of genotypic variation is important to understand the nature and
in distribution of Zn in the grain, and also extent of GE interactions for designing
to assess the potential for breeding for this breeding strategies and selection procedures
trait. (Eiseman, Cooper and Woodruff,
1990). Much effort has been directed to
17.3.2 Se & I understanding GE interactions in relation
As noted above, genotypic variation was to yield, but little attention has been given
found by our group in rice bran and, to a to grain nutrient density. Graham et al.
lesser extent, endosperm for Se, but only (1999) recognized that environment (soil
in bran for I. Se concentration in rice bran type, fertilizer management and climate)
was around three times that in endosperm, can have a strong influence on nutrient
while I concentration in bran was around density of grains; thus it is important
nine times that in endosperm (mean 730 to grow out the seed to be compared
vs 80 g/kg DW) (Lyons et al., 2005b). for at least one generation in the same
As most rice is eaten in the polished form, environment to minimize the variation in
with the bran removed, selection for a nutrient density associated with previous
higher proportion of grain Se and I stored growing conditions. Only then can valid
in endosperm may be worthwhile, as noted comparisons of genetically controlled
for Zn and Fe above. variation be made.
Fertilizer management can also influence
17.5 GENOTYPE ENVIRONMENT micronutrient density in the grain. In
INTERACTIONS AND THEIR EFFECTS ON studies with maize varieties grown at
MICRONUTRIENT TRAITS different nitrogen (N) and water levels,
GE interaction is an important issue for Feil et al. (2005) found that N fertilization
plant breeders. A significant GE interaction reduced grain concentrations of Zn and Mn,
implies that rankings of genotypes differ with which was attributed to higher grain yield
environment, indicating the need for testing as a result of N application. In contrast,
at various sites or seasons. For example, a continuous irrigation did not affect grain
genotype may exhibit high-micronutrient- nutrient density. However, the rankings
density traits in one environment, but not of varieties remained unchanged by water
in others. A significant GE interaction regime and N levels, pointing to stability of
can be classified as either (i) non-crossover, varietal differences in grain nutrient density
where the ranking of genotypes remains over a range of N and water levels. At
consistent in different environments and present, there is little information available
it is the degree of accumulation that is on the effects of fertilization on grain
affected; or (ii) crossover, where the rank micronutrient density in rice or wheat.
of individual cultivars is affected by the
environment (Baker and Kosmolak, 1977). 17.5.1 Fe & Zn
GE interactions and statistical methods Wheat
for analysis and interpretation of GE From field trials conducted by our
interactions have been reviewed elsewhere group in different locations and years
430 Plant breeding and farmer participation
TABLE 17.3
Fe and Zn concentrations (mg/kg DW) in grains of wheat cultivars grown at different locations and
years in Australia
Cultivar Zn concentration Fe concentration
Lameroo Bute Birchip Horsham Birchip Birchip Lameroo Bute Birchip Horsham Birchip Birchip
1996 1997 1998 1998 1999 2000 1996 1997 1998 1999 1999 2000
Excalibur 22 18 21 17 25 18 37 32 44 35 40 37
Krichauff 20 19 25 18 24 18 36 34 46 38 39 38
Songlen 24 20 25 19 31 16 37 35 45 38 42 35
Trident 18 17 25 19 23 17 35 32 45 35 38 38
Yallaroi 20 17 22 16 24 14 32 32 45 36 40 37
Mean 21 18 24 18 26 17 35 33 45 36 40 37
Standard 2 1 2 1 3 2 2 1 1 2 2 1
deviation
Source: Graham et al., unpublished.
TABLE 17.4
Fe and Zn concentrations (mg/kg DW) in grains of wheat cultivars grown at two levels of Zn
fertilization at Lameroo (1996) and Bute (1997) in South Australia
Cultivar Zn concentration Fe concentration
Lameroo Bute Lameroo Bute
Nil +Zn Mean* Nil +Zn *Mean Nil +Zn *Mean Nil +Zn *Mean
Barunga 14 22 18 15 20 17 38 36 37 33 31 32
Cascades 14 21 17 11 20 15 38 35 36 35 31 33
Excalibur 14 22 18 11 18 14 36 37 36 34 32 33
Frame 12 19 15 12 19 15 32 31 32 35 34 35
Halberd 13 23 18 12 17 14 35 33 34 36 35 36
Janz 12 18 15 11 16 13 29 29 29 29 27 28
Krichauff 12 20 16 12 19 15 38 36 37 37 34 35
RAC750 12 22 17 13 21 17 33 31 32 31 31 31
RAC809 11 19 15 10 16 13 36 35 35 34 30 32
RAC812 11 17 14 12 20 16 33 31 32 35 34 34
RAC820 15 22 18 10 15 13 33 33 33 34 32 33
RAC826 13 21 17 12 18 15 33 34 34 36 33 34
RAC832 13 21 17 14 24 19 30 31 31 38 35 37
RH911996 15 21 17 10 16 13 32 33 32 31 31 31
RH912025 14 20 17 11 18 14 36 31 34 33 35 34
Songlen 14 24 19 11 20 15 36 37 37 36 35 35
Tammin 14 20 17 10 17 14 37 35 36 34 31 32
Trident 12 18 15 11 17 14 35 35 35 35 32 33
WI334 11 17 14 11 16 14 33 33 33 35 33 34
WI94063 11 18 15 13 19 16 31 30 30 29 30 30
WI94091 10 14 12 12 18 15 29 27 28 32 29 31
Yallaroi 17 21 19 12 17 15 34 32 33 32 32 32
Yanac 11 17 14 11 16 13 31 30 31 35 31 33
Mean 13 20 12 18 34 33 34 32
Range 1017 1424 1015 1824 2938 2737 2938 2735
NOTES: Genotype Zn fertilization interaction for grain Zn and Fe concentrations was non-significant, while genotype
location interaction was significant (LSD0.05=3 for Zn and Fe). +Zn treatment received granular zinc (zinc oxysulphate, 32
percent Zn) at a rate of 7 kg/ha at seeding and a foliar spray (Zincsol, 16.7 percent Zn) at a rate of 2 L/ha at the early growth
stage.
Source: Graham et al., unpublished.
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 431
the detection of what may be relatively Quantitative Trait Loci, QTL) identified
small (for example, 10 percent) genotypic were found to co-segregate for grain Zn
variations in Se uptake efficiency between accumulation and vegetative Zn accumula-
wheat cultivars under these field conditions tion at anthesis, indicating little prospect
is virtually impossible. Background soil for screening for grain Zn accumulation
variation in available I has also been found even as early as anthesis. Moreover, the
to be substantial at the South Australian only QTL detected for shoot Zn concen-
sites we have used, although less so than tration or content (chromosome 4) did
for Se. This large microspatial variation in not co-segregate with Zn concentration or
soils makes it difficult, if not impossible, content at either anthesis or maturity, sug-
to accurately assess genotypic differences gesting that screening for grain Zn accumu-
across environments for Se and I. This and lation at the early stage will not be reliable
narrow genotypic variation reported so or relevant to grain Zn accumulation.
far may be the reasons why to date there Inductively Coupled Plasma Optical
have been no studies reported on GE Emission Spectrometry (ICPOES) is com-
interactions for Se and I. monly used to determine mineral nutrient
concentrations in plant tissues, and allows
17.6.SCREENING AND ANALYTICAL the determination of interactions among
METHODS FOR MICRONUTRIENTS IN the various essential nutrients, effects that
FOOD CROPS can be large and important. This method
Where should screening be carried out: is fast and reliable, but can be costly for
field or greenhouse? The principles of both breeding programmes in both developed
controlled environment and field screening and developing countries, where tens of
are reviewed elsewhere (Graham, 1984), thousands of samples are handled each
and therefore will not be dealt with in year. Are there alternative and cheaper
detail here. methods to ICPOES? The researchers at
the University of Adelaide (Australia) have
17.6.1 Fe & Zn developed a rapid, cheap and user-friend-
The data presented in Table 17.4 suggest ly assay for determination of Fe in the
that screening for grain Zn concentration grain of rice or wheat (Choi, Graham and
should be carried out in optimal growing Stangoulis, 2007). This new cost-effective
conditions, as variation in grain Zn con- assay consists of two phases. In phase one,
centration under Zn deficient conditions is the assay is used to identify high grain-Fe
rather narrow. Unlike traits such as agro- lines from thousands of samples, while in
nomic Zn efficiency, screening at the early phase two, the high-Fe lines identified in
growth stage does not appear to be suitable phase one are confirmed by ICPOES.
for detecting or identifying genotypes with
the ability to load more Zn into the grain, 17.6.2 Vitamin A
due probably to the overriding importance Several standard procedures for extraction
of re-mobilization of Zn from leaves into and identification of carotenoids from plant
grain occurring towards maturity. The evi- material have been used since the discovery
dence for this comes from a study in bar- and naming of carotene in the early 19th
ley (Lonergan, 2001) in which two of the century. Separation of carotenoids initially
four chromosomal regions (also known as involved a two-step chromatographic
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 433
Se and I traits in cereals needs to include Se-DAN complex, piazselenol (Koh and
hydroponic trials and pot trials using a Benson, 1983). Samples for I analysis are
standardized growth medium, backed up typically prepared using tetramethyl ammo-
by field studies conducted on soils that are nium hydroxide (TMAH) extraction.
relatively uniform in Se and I. Selenate is
the most mobile Se form and the dominant 17.7 INHERITANCE OF MICRONUTRIENT
available Se form in well-aerated, neutral ACCUMULATION IN FOOD CROPS
to alkaline soils (Cary and Allaway, 1969), Apart from the existence of genetic
while selenite is the major form taken up variation, breeding for enhanced grain
by rice in flooded paddy soils of lower nutrient content also requires knowledge
pH (Wang and Gao, 2001). Thus the of genetic control mechanisms.
composition of hydroponic culture media
needs to be tailored to the relevant field 17.7.1 Fe & Zn
situation. Using solution culture containing Wheat
selenite as the dominant available Se form, At present, there is little or no information
Zhang et al. (2004) have found genotypic available on genetics of Zn and Fe
variation in Se concentration in the leaves accumulation in wheat grain. The continuous
of rice seedlings, and the levels are well variation in grain Zn and Fe concentration
correlated with those in grain. and content within Recombinant Inbred
Genotypic variation in I uptake in rice Lines (RIL) (n=113) derived from Opata
may be explained by differences in the oxi- Synthetic cross (Figures 17.4 and 17.5)
dising power of the roots, which can oxidise indicates that the two traits are quantitative
the iodide ion to form molecular I, which is and controlled by several genes (Genc
then absorbed more readily. A significant et al., unpublished). It is interesting to
correlation was found between the oxidis- note that some RILs had higher Zn and
ing power of rice roots and the uptake of I Fe concentration and content than either
(Yamada et al., 2005), hence this may prove of the parents, suggesting a transgressive
to be a suitable screening method. segregation, which is probably due to
Commonly used methods of Se analysis these lines carrying favourable allele
include hydride ICPOES, ICP mass spec- combinations from both parents. This
trometry, and fluorimetry. Sample prepara- multi-gene control hypothesis is supported
tion for hydride ICPOES involves digestion by recent studies that identified several
with a nitric+perchloric acid mixture, fol- chromosomal regions associated with
lowed by hydrochloric acid digestion, then grain Zn concentration (Shi et al., 2007;
treatment with sodium borohydride (Tracy Genc et al., 2009) and content (Shi et al.,
and Moller, 1990). ICP mass spectrometry, 2007). However, these field studies were
in which plasma is used as the ionization conducted at a single location and QTLs
source, is a highly sensitive method for Se identified were mapped to either different
and I analyses (Hieftje and Vickers, 1989). chromosomes or to different regions of
Another commonly used (and time-hon- the same chromosome. Therefore, there
oured) method for both Se and I analysis is a need to test mapping populations at
is the fluorimetric method. This is based multiple sites and years to validate the
on the reaction of 2,3-diaminonaphthalene QTLs and also to determine the extent
(DAN) with Se (IV) to form a fluorescent of GE interactions on grain Zn and Fe
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 435
40
35
35
30
30
25
25
No of RILs
No of RILs
20
20
15
15
10
10
5 5
0 0
23-26
27-30
31-34
35-38
39-42
43-45
46-49
50-53
54-57
17-20
21-24
25-28
29-32
33-36
37-40
40
60
35
50 30
25
No of RILs
40
No of RILs
20
30
15
20 10
5
10
0
0.95-1.24
1.25-1.54
1.55-1.84
1.85-2.14
2.15-2.44
2.45-2.74
2.75-2.94
0
0.64-0.94
0.95-1.24
1.25-1.54
1.55-1.84
1.85-2.14
2.15-2.44
are also combined with other genes with siderophores, nicotianamine, organic acids),
smaller effects that affect accumulation. reducing agents (ferric reductase), enzymes
and transport proteins in the root cells
Wheat could enhance Fe uptake and transport
A study into the genetic origin of an increase (Grusak, Pearson and Marentes, 1999).
in carotenoid pigments in the cross between However, higher uptake and transport does
a wild barley (Hordeum chilense) and not necessarily imply higher accumulation
durum wheat located this trait to the -arm in the grain (phloem loading). For example,
of chromosome 7Hch (Alvarez, Martin and a pea mutant of cultivar Sparkle (brz)
Martin, 1998). Screening of 35 lines with accumulated 36-fold higher Fe in the leaves
various Tritordeum lines revealed that compared to Sparkle, but did not have higher
although the presence of the Hch genome Fe in the seeds (Grusak, 1994). The author
is responsible for increased carotenoid con- concluded that the rate limitation to phloem
centrations in these lines it is difficult to Fe loading was due to an unidentified
predict the effect of the interaction between ligand species that would complex with
the barley and wheat genetics. H. chilense is Fe prior to phloem loading rather than the
therefore a useful but not entirely reliable availability of Fe as substrate. This study
source of increased carotenoid concentra- was followed by the study of Marentes
tions for T. durum (Alvarez, Martin and and Grusak (1998), who demonstrated that
Martin, 1999). a second mutant of cultivar Sparkle (dgl)
had 2.5-fold higher Fe concentration in
17.8 PHYSIOLOGICAL AND MOLECULAR the embryo compared to Sparkle (163 and
MECHANISMS OF MICRONUTRIENT 65 mg/kg DW, respectively). This mutant
UPTAKE, TRANSLOCATION, RE- also had higher Fe concentration in the seed
MOBILIZATION AND ACCUMULATION coat. The authors used radiotracer 59Fe to
17.8.1 Fe & Zn determine the movement of Fe in the seed
For a breeding programme to be successful, coat, and found that Fe was symplastically
it is important to understand the processes phloem loaded. They further suggested that
leading to accumulation of nutrients Fe resided within the non-vascular seed
in the grain. Obviously an increase in coat cells, and that the cells at the inner
accumulation of Fe and Zn in the grain of surface of the seed coat may facilitate the
any plant species will require higher uptake, release of Fe to the embryo apoplast. The
translocation or re-mobilization from form of Fe in the seed coat or embryo is
source (leaves) to sink (grain). The role of still not known at present.
these processes in relation to accumulation A recent study in rice suggests that
of Fe, Zn and other micronutrients has nicotianamine (NA) and nicotianamine
been reviewed recently (Grusak, Pearson synthase (NAS) genes (OsNAS1, OsNAS2
and Marentes, 1999); thus it will not be and OsNAS3) are also involved in long-
discussed in detail here. It is interesting distance transport of Fe (Inoue et al.,
to note that Fe has been the most studied 2003), apart from their roles in the release
micronutrient in rice, while Zn has been of low-molecular weight compounds,
researched to a larger extent in wheat. phytosiderophores, from the roots
There is some suggestion that increasing of graminaceous plants. This release of
the levels of Fe-chelating agents (phyto- phytosiderophores solubilizes rhizospheric
438 Plant breeding and farmer participation
Fe(III) thus increasing plant uptake of Fe, the plant (Scholz, Seifert and Grun, 1987;
for example in rice (Takagi, 1976; Takahashi Treeby, Marschner and Romheld, 1989;
et al., 2001). Most recently, a rice metal-NA Zhang, 1991; Cakmak et al., 1996), but this
transporter, OsYLS2, has been linked to needs to be established in future studies.
phloem transport and translocation to the The positive correlations between Fe
grain of Fe (Koike et al., 2004). Increasing and Zn concentrations in cereal grains pro-
the expression of transporters such as vide evidence that both nutrients may be
OsYSL2 and OsNASs in rice and other taken up and translocated to the grain
species can enhance Fe, and to some extent through the same process. However, as the
Zn, accumulation in the grains. However, correlation coefficient is never 1, there must
further studies are required to determine be other mechanisms specific to Fe or Zn
the extent of genotypic variation in NA uptake or translocation to the grain, which
concentrations and its relative contribution warrants further investigation.
to Fe and Zn accumulation in several Having briefly discussed uptake and
varieties with low and high Fe and Zn in translocation of Fe and Zn to the grain,
the grain. two questions arise: How much of the Fe
In contrast to Fe, there is little informa- and Zn in the plant ends up in the grain?
tion available on Zn translocation in the and Where are these nutrients deposited
plant and transport to the grain. An earlier in the grain? Contrary to earlier sugges-
study (Longnecker and Robson, 1993) sug- tions of poor re-mobilization of Fe from
gested that organic complexes with citrate vegetative tissues to the grain (4 percent in
and malate may be important in re-mobili- rice, Marr et al., 1995; 20 percent in wheat,
zation of Zn in the phloem. A recent study Miller et al., 1994), a recent growth-room
indicates that this may not be the case, as study with wheat reported that 77 percent
no relationship could be found between of total shoot Fe was re-mobilized to the
the presence of complexes or ligands and grain at maturity (Garnett and Graham,
loading of Zn into the wheat grain (Pearson 2005). The lower re-mobilization of Fe in
et al., 1996). However, this result does not the earlier studies involving field-grown
rule out the possibility of other endog- plants were attributed to (i) precipitation
enous chelates that may play a role in the of Fe in the apoplasm (inactive Fe) at high
long-distance transport of micronutrients concentrations, which may result in non-
(Welch, 1995). At the same time, transport re-mobilization, (ii) saturation of either
to the grain is thought to occur predomi- the grain loading or phloem loading, and
nantly via the phloem (Pearson et al., 1995), (iii) contamination of plant tissues by soil
and is well regulated (Herren and Feller, (references in Garnett and Graham, 2005).
1997). This regulation of Zn transport has Miller et al. (1994) reported that, in wheat,
been reported by Pearson, Rengel and 70 percent of Zn in the leaves was re-mo-
Graham (1999), who suggested that low-Zn bilized to the grains, while only 42 percent
grain was not a strong sink for Zn, while of shoot Zn re-mobilized to the grain in
in the case of high-Zn grain, there may be the study of Garnett and Graham (2005).
a barrier preventing excessive accumula- The differences in the amounts of Zn re-
tion in the grain. It has also been suggested mobilized in these two studies may be due
that phytosiderophores and nicotianamine to differences in genotypes and experimen-
may facilitate Zn uptake and transport in tal conditions. It has been suggested that,
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 439
in wheat, relatively large amounts of Zn are is stored in the bran layers and lost dur-
transported into crease or inner pericarp ing milling or polishing (Muramatsu et al.,
tissues via the crease phloem, and trans- 1989). To date, little has been reported on
location to the embryo and endosperm the physiology of I in the plant system and
continues throughout grain development further studies are needed, especially on the
(Pearson et al., 1998). As Zn status of the forms in which I is transported and stored.
grain improves, more Zn is distributed
to the inner pericarp and less Zn to the 17.9 CONCLUSIONS
endosperm, outer pericarp and embryo 17.9.1 Fe & Zn
(Pearson, Rengel and Graham, 1999). There is substantial evidence for genotypic
variation to justify breeding efforts towards
17.8.2 Se & I developing high grain-Fe and -Zn varie-
In most plants, uptake, transport and ties. However, our knowledge of genet-
assimilation of selenate is the same as ics, physiological mechanisms responsible
for sulphate, and leads to synthesis of for high grain Fe and Zn trait and GE
selenocysteine and selenomethionine; interactions is very limited, and now it is
selenocysteine is then incorporated into time to focus on these areas. One impor-
proteins (Lauchli, 1993). Hence, the transfer tant point we should mention is that these
of sulphate/selenate transporter genes from proposed studies should be supplemented
a Se accumulator like Astragalus bisulcatus by bio-availability studies in animals and
may be useful for phytoremediation of high- humans. There have been some concerns
Se areas (Goodson et al., 2003). However, with respect to poor bio-availability of
this strategy may not assist with Se uptake these nutrients due to naturally occurring
on soils with low Se availability, where high phytate concentrations in the grains.
most Se is present as selenite, selenide and However, a study in rats reported that bio-
elemental Se forms (Cary and Allaway, availability of Fe and Zn remained constant
1969). Selenite absorbed by the roots in low- and high-density genotypes of cere-
undergoes a series of reduction reactions, als and beans (Welch et al., 2000). So it is a
including conversion to selenide, and finally reasonable argument that there will be an
a reaction with O-acetylserine to form increase in absorption of these nutrients as
selenocysteine (Tsang and Schiff, 1978). their concentrations increase in the grain,
Because of shared transporters, sulphate in despite their low bio-availability compared
growth media inhibits uptake of selenate to animal food sources, as observed in the
(Ferrari and Renosto, 1972), and sulphite Philippine rice study (Haas et al., 2005),
may inhibit uptake of selenite, but further though in that study, the varieties differed
studies are required to confirm this. simultaneously in Fe and Zn, giving rise to
Iodine species of lower oxidative state potential interactions in the gut. We believe
and molecular weight (iodide, -1 and 116, that breeding for these traits is a worthwhile
respectively) are absorbed more read- approach given the impact the small incre-
ily than the heavier, higher valency forms ment in absorption of these nutrients will
(iodate, +5 and 214, respectively) (Umaly have on the lives of billions of people who
and Poel, 1971). I is transported mostly in are reliant on staple food crops such as rice
the xylem, hence little is re-translocated and wheat for their dietary requirements
from the leaves into the grain, where most of Fe and Zn. Finally and importantly, if
440 Plant breeding and farmer participation
breeding for high grain-Fe and -Zn traits is at sites with different soil types and includ-
to be successful and the varieties adopted ing a wide range of germplasm grown
by farmers, the high grain-Fe and -Zn together are needed to confirm whether
traits must be linked to high yield. This has sufficient genetic variability exists to enable
been achieved in rice (Gregorio, 2002), and selection for uptake and grain loading effi-
results from wheat trials are also encourag- ciency of Se and I. Previous studies suggest
ing (R.M. Trethowan, pers. comm.). that Se and I delivered through bio-forti-
fied cereals are highly bio-available (Jiang,
17.9.2 Vitamin A Cao and Jiang, 1997; Lyons et al., 2003).
Despite extensive selection against
pigmentation in several staple foods, genetic ACKNOWLEDGEMENT
variation for carotenoid concentration Figures 17.1 and 17.2 were reprinted with
can still be revealed by screening varieties permission from the UNU Food and
available from germplasm banks, as Nutrition Bulletin (2000) 21: 404409.
illustrated in this chapter. Even within
those staple crops that have substantial REFERENCES
concentrations of carotenoids, the value AACC [American Association of Cereal
of screening ancient varieties for sources Chemists]. 1983. Approved methods of
of higher accumulation is obvious. The the AACC. Method 8-01, 8th edition.
time consuming and expensive nature of Abdel-Aal, E.-S.M., Hucl, P. & Sosulski,
carotenoid analysis still remains a significant F.W. 1995. Compositional and nutritional
restriction, though recent progress in the characteristics of spring einkorn and spelt
development of fast screening methods for wheats. Cereal Chemistry, 72: 621624.
wheat will expedite mass screenings for this Abdel-Aal, E.-S.M., Young, J.C., Wood, P.J.,
staple crop. Although much work has been Rabalski, I., Hucl, P., Falk, D. & Fregeau-
done in elucidating sources of increased Reid, J. 2002. Einkorn: a potential can-
carotenoid concentrations in staple foods didate for developing high lutein wheat.
there is still much to do before we obtain Cereal Chemistry, 79(3): 455457.
concentrations that can alleviate vitamin A Alvarez, J.B., Martin, L.M. & Martin, A.
deficiencies. In addition, it is not merely 1998. Chromosomal localisation of genes
enough to develop lines with high carotenoid for carotenoid pigments using addition
concentrations; they must be adapted to local lines of Hordeum chilense in wheat. Plant
conditions, and also be culturally acceptable Breeding, 117: 287289.
and the carotenoids bio-available. Alvarez, J.B., Martin, L.M. & Martin, A.
1999. Genetic variation for caroten-
17.9.3 Se & I oid pigment content in the amphiploid
The limited investigations carried out to Hordeum chilense Triticum turgidum
date suggest that rice may be the most conv. durum. Plant Breeding, 118: 187
promising of the major cereals for breeding 189.
to improve grain Se and I density, although Alvarez, J.B., Urbano, J.M. & Martin, L.M.
further screening of all the major cere- 1994. Effect on flour quality from inclu-
als may reveal more germplasm that can sion of the Hordeum chilense genome into
enhance these traits. For rice, in particular, the genetic background of wheat. Cereal
further pot trials and field trials conducted Chemistry, 71: 517519.
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 441
Arthur, J.R. 1999. Functional indicators of Burk, R.F. & Solomons, N.W. 1985. Trace
iodine and selenium status. Proceedings of elements and vitamins and bioavailabil-
the Nutrition Society, 58: 507512. ity as related to wheat and wheat foods.
Baker, R.J. & Kosmolak, F.G. 1977. Effects American Journal of Clinical Nutrition,
of genotype-environment interactions. 41: 10911102.
Canadian Journal of Plant Science, 68: 405 Cakmak, I., Yilmaz, A., Kalayci, M., Ekiz,
410. H., Erenoglu, B. & Braun, H. 1996. Zn
Berrow, M.L. & Ure, A.M. 1989. Geological deficiency as a critical problem in wheat
materials and soils. In M. Ihnat, ed. production in Central Anatolia. Plant and
Occurrence and distribution of selenium, Soil, 180: 165172.
pp. 226228. Boca Raton, USA, CRC Cakmak, I., Ozkan, H., Braun H.J., Welch
Press. R.M. & Romheld V. 2000. Zinc and iron
Bhaskaram, P. 2002. Micronutrient malnutri- concentrations in seeds of wild, primitive
tion, infection, and immunity: an overview. and modern wheats. Food and Nutrition
Nutrition Reviews, 60: S40S45. Bulletin, 21(4): 401403.
Blessin, C.W., Dimler, C.O., Grogan, C.O. & Cakmak, I., Torun, A., Millet, E., Feldman,
Campbell, C.W. 1963. Carotenoids of corn M., Fahima, T., Korol, A., Nevo, E.,
and sorghum. Cereal Chemistry, 40: 436 Braun, H.J. & Ozkan, H. 2004. Triticum
441. dicoccoides: An important genetic
Borghi, B., Castagna, R., Corbellini, M., Heun, resource for increasing zinc and iron
M. & Salamini, F. 1996. Breadmaking quali- concentration in modern cultivated
ty of Einkorn wheat (Triticum monococcum wheat. Soil Science and Plant Nutrition,
subsp. monococcum). Cereal Chemistry, 50: 10471054.
73(2): 208. Cary, E.E. & Allaway, W.H. 1969. The stabil-
Boyd, L.O., De Boland, A.R. & Koirtyohann, ity of different forms of selenium applied to
S.R. 1972. Distribution of phytate and low-selenium soils. Soil Science Society of
nutritionally important elements among America Proceedings, 33: 571.
the morphological components of cereal Chavez, A.L., Bedoya, J. M., Sanchez, T.,
grains. Journal of Agricultural and Food Iglesias, C., Ceballos, H. & Roca, W. 2000.
Chemistry, 20: 718721. Iron, carotene, and ascorbic acid in cas-
Brunsen, A.M. & Quackenbush, F.W. 1962. sava roots and leaves. Food and Nutrition
Breeding corn with high provitamin A in Bulletin, 21(4): 410413.
the grain. Crop Science, 2: 344347. Choi, E.Y., Graham, R. & Stangoulis, J. 2007.
Broadley, M.R., White, P.J., Bryson, R.J., Semi-quantitative analysis for selecting Fe-
Meacham, M.C., Bowen, H.C., Johnson, and Zn-dense genotypes of staple food
S.E., Hawkesford, M.J., McGrath, S.P., crops. Journal of Food Composition and
Zhao, F-J., Breward, N., Harriman, M. & Analysis, 20: 496505.
Tucker, M. 2006. Biofortification of UK Combs, G.F. 2001. Selenium in global food
food crops with selenium. Proceedings of systems. British Journal of Nutrition,
the Nutrition Society, 65: 169-181. 85: 517547.
Buckner, B., Miguel, P. S., Janik-Buckner, DEgidio, M.G., Nardi, S. & Vallega, V. 1993.
D., & Bennetzen, J. L. 1996. The y1 gene Grain, flour and dough characteristics of
of maize codes for phytoene synthesis. selected stains of diploid wheat. Cereal
Genetics, 143: 479488. Chemistry, 70(3): 298303.
442 Plant breeding and farmer participation
Eiseman, R.L., Cooper, M. & Woodruff, D.R. G.K. 2008. Quantitative trait loci analysis
1990. Beyond the analytical methodology: of zinc efficiency and grain zinc concentra-
better interpretation and exploitation of tion in wheat using whole genome aver-
genotype by environment interaction in age interval mapping. Plant and Soil, 314:
breeding. In M.S. Kang, ed. Genotype by 49-66.
environment interaction and plant breeding, Goodson, C.C., Parker, D.R., Amrhein, C. &
pp. 108117. Baton Rouge, USA, Louisiana Zhang Y. 2003. Soil selenium uptake and
State University Agriculture Center. root system development in plant taxa dif-
Egesel, C.O. 1997. Inheritance and genetic fering in Se-accumulating capability. New
variability in carotenoids, tocols and related Phytologist, 159(2): 391401.
traits in maize. MSc thesis, University of Graham, R.D. 1984. Breeding for nutritional
Illinois, Urbana-Champaign, USA. characteristics in cereals. Advances in Plant
Egesel, C.O. 2001. Genetic variation among Nutrition, 1: 57102.
maize genotypes for carotenoids and toco- Graham, R.D. & Rosser, J.M. 2000.
pherol compounds. PhD thesis, University Carotenoids in staple foods. Food and
of Illinois, Urbana-Champaign, USA. Nutrition Bulletin, 21(4): 404409.
Egesel, C.O., Wong, J.C., Lambert, R.J. & Graham, R.D., Senadhira, D. & Ortiz-
Rocheford, T.R. 2003. Gene dosage effects Monasterio, I.A. 1997. A strategy for breed-
on carotenoid concentration in maize grain. ing staple-food crops with high micro-
Maydica, 48(3): 183190. nutrient density. Soil Science and Plant
Failla, M.L. 2003. Trace elements and host defense: Nutrition, 43: 11531157.
recent advances and continuing challenges. Graham, R.D., Welch, R.M. & Bouis, H.E.
Journal of Nutrition, 133: 1448S1451S. 2001. Addressing micronutrient malnu-
Farnham, M.W., Hale, A.J., Grusak, M.A. & trition through enhancing the nutritional
Finley, J.W. 2007. Genotypic and environ- quality of staple foods: principles, perspec-
mental effects on selenium concentration tives and knowledge gaps. Advances in
of broccoli heads grown without supple- Agronomy, 70: 77142.
mental selenium fertilizer. Plant Breeding, Graham, R.D., Senadhira, D., Beebe, S.,
126(2): 195200. Iglesias, C. & Ortiz-Monasterio, I. 1999.
Feil, B., Moser, S.B., Jampatong, S. & Stamp, Breeding for micronutrient density in edi-
P. 2005. Mineral composition of the grains ble portions of staple food crops: conven-
of tropical maize varieties as affected by tional approaches. Field Crops Research,
pre-anthesis drought and rate of nitrogen 60: 5780.
fertilisation. Crop Science, 45: 516523. Gregorio, G.B. 2002. Progress in breeding for
Ferrari, G. & Renosto, F. 1972. Regulation of trace minerals in staple crops. In H.E. Bouis,
sulfate uptake by excised barley roots in ed. Plant breeding: a new tool for fighting
the presence of selenate. Plant Physiology, micronutrient malnutrition, pp. 500S502S.
49: 114116. Orlando, USA.
Garnett, T.P. & Graham, R.D. 2005. Gregorio, G.B., Senadhira, D., Htut, H. &
Distribution and remobilisation of iron Graham, R.D. 2000. Breeding for trace
and copper in wheat. Annals of Botany, mineral density in rice. Food and Nutrition
95: 817826. Bulletin, 21: 382386.
Genc, Y., Verbyla, A.P., Torun, A.A., Cakmak, Grela, E.R. 1996. Nutrient composition and
I., Willsmore, K., Wallwork, H. & McDonald, content of antinutritional factors in spelt
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 443
(T. spelta L) cultivars. Journal of the Science Hill, J., Becker, H.C. & Tigerstedt, P.M.A.
of Food and Agriculture, 71: 399404. 1998. Quantitative and ecological aspects
Grogan, C.O., Blessin, C.W., Dimler, R.J. & of plant breeding. London, Chapman and
Campbell, C.M. 1963. Parental influence Hall. 275 p.
on xanthophyll and carotenes in corn. Crop Humphries, J.M. & Khachik, F. 2003.
Science, 3: 213214. Distribution of lutein, zeaxanthin, and
Grusak, M.A. 1994. Iron transport to devel- related geometrical isomers in fruit, veg-
oping ovules of Pisum sativum L. I. Seed etables, wheat, and pasta products. Journal
import characteristics and phloem iron- of Agricultural and Food Chemistry, 15(5):
loading capacity of source regions. Plant 1322-1327.
Physiology, 104: 649655. Humphries, J.M., Mares, D.J. & Graham,
Grusak, M.A., Pearson, J.N. & Marentes, R.D. 2004. Application of reflectance
E. 1999. The physiology of micronutri- colour measurement to the estimation
ent homeostasis in field crops. Field Crops of carotene and lutein content in wheat
Research, 60: 4156. and triticale. Journal of Cereal Science,
Haas, J.D., Beard, J.L., Murray-Kolb, L.E., 40(2): 151159.
del Mundo, A., Felix, A. & Gregorio, G.B. Iglesias, C., Mayer, J., Chavez, L.S. & Calle, F.
2005. Iron-bio-fortified rice improves the 1997. Genetic potential and stability of car-
iron stores of nonanemic Filipino women. otene content in cassava roots. Euphytica,
Journal of Nutrition, 135: 28232830. 94: 367373.
HarvestPlus. Breeding Crops for Better Inoue, H., Higuchi, K., Takahashi, M.,
Nutrition 2007. URL: www.harvestplus.org Nakanishi, H., Mori, S. & Nishizawa, N.K.
Hentschel, V., Kranl, K., Hollmann, J., 2003. Three rice nicotianamine synthase
Lindhauer, M.G., Bohm, V. & Bitsch, R. genes, OsNAS1, OsNAS2 and OsNAS3,
2002. Spectrophotometric determination are expressed in cells involved in long-
of yellow pigment content and evaluation distance transport of iron and differentially
of carotenoids by high-performance liquid regulated by iron. Plant Journal, 36: 366
chromatography in durum wheat grain. 381.
Journal of Agricultural and Food Chemistry, Jiang, X.M., Cao, X.Y. & Jiang, J.Y. 1997.
50: 66636668. Dynamics of environmental supplementa-
Herren, T. & Feller, U. 1997. Influence of tion of iodine: four years experience of
increased zinc levels on phloem transport in iodination of irrigation water in Hotien,
wheat shoots. Journal of Plant Physiology, Xinjiang, China. Archives of Environmental
150: 228231. Health, 52(6): 399408.
Hershey, C.H. & Ocampo, C. 1989. Description Johnson, J.M. & Butler, G.W. 1957. Iodine
of new genetic markers in cassava. Cassava content of pasture plants. Physiologia
Newsletter, 13: 15. Plantarum, 10: 100111.
Hetzel, B.S. & Pandav, C.S. 1996. SOS for a Johnson, L.J. & Miller, E.S. 1938. Variation in
billion. The conquest of Iodine Deficiency carotenoid pigment concentration among
Disorders. Delhi, Oxford University Press. inbred and crossbred strains of corn. Cereal
Hieftje, G.M. & Vickers, G.H. 1989. Chemistry, 15: 345350.
Developments in plasma source mass Jones, G.P. 1997. Minerals. In M.L. Wahlqvist
spectrometry. Analytica Chimica Acta, (editor). Food and nutrition, pp. 249254.
216: 124. Sydney, Australia, Allen & Unwin.
444 Plant breeding and farmer participation
Jos, J.S., Nair, S.G., Moorthy, S.N. & Nair, Lyons, G.H., Stangoulis, J.C.R. & Graham,
R.B. 1990. Carotene enhancement in cas- R.D. 2003. High-selenium wheat: bio-
sava. Journal of Root Crops, 16: 511. fortification for better health. Nutrition
Kang, M.S. 1990. Understanding and utiliza- Research Reviews, 16: 4560.
tion of genotype by environment inter- Lyons, G.H., Lewis, J., Lorimer, M.F., Holloway,
action in plant breeding. In M.S. Kang, R.E., Brace, D.M., Stangoulis, J.C.R. &
ed. Genotype by environment interac- Graham, R.D. 2004. High-selenium wheat:
tion and plant breeding, pp. 5268. Baton agronomic bio-fortification strategies to
Rouge, USA, Louisiana State University improve human nutrition. Food, Agriculture
Agriculture Center. and Environment, 2(1): 171178.
Koh, T.S. & Benson, T.H. 1983. Critical reap- Lyons, G.H., Ortiz-Monasterio, I., Stangoulis,
praisal of fluorimetric method for determi- J. & Graham, R. 2005a. Selenium concentra-
nation of selenium in biological materials. tion in wheat grain: Is there sufficient geno-
Journal AOAC, 66: 918926. typic variation to use in breeding? Plant
Kohrle, J. 1999. The trace element sele- and Soil, 269: 369380.
nium and the thyroid gland. Biochimie, Lyons, G.H., Ortiz-Monasterio, I., Genc, Y.,
81(5): 527533. Stangoulis, J.C.R. & Graham, R.D. 2005b.
Koike, S., Inoue, H., Mizuno, D., Takahashi, Can cereals be bred for increased selenium
M., Nakanishi, H., Mori, S. & Nishizawa, and iodine concentration in the grain? In
N.K. 2004. OsYSL2 is a rice metal-nico- C.J. Li et al., eds. Plant nutrition for food
tianamine transporter that is regulated by security, human health and environmental
iron and expressed in the phloem. Plant protection, pp. 374375. Beijing, Tsinghua
Journal, 39: 415424. University Press.
Lacroix, L.J. & Lier, J.B. 1975. Carotenoids Lyons, G.H., Genc, Y., Stangoulis, J.C.R.,
in durum wheat: developmental patterns Palmer, L.T. & Graham, R.D. 2005c.
during two growing seasons. Canadian Distribution of selenium and other nutrients
Journal of Plant Science, 55: 679684. in wheat grain, and the effect of processing
Lauchli, A. 1993. Selenium in plants: uptake, on wheat selenium content. Biological Trace
functions, and environmental toxicity. Element Research, 103(2): 155168.
Botanica Acta, 106: 455468. Mangelsdorf, P.C. & Fraps, G.S. 1931. A direct
Lepage, M. & Sims, R.P.A. 1968. Carotenoids quantitative relationship between vitamin A
of wheat flour: their identification and in corn and the number of genes for yellow
composition. Cereal Chemistry, 45: 600 pigmentation. Science, 73: 241242.
611. Marr, K.M., Batten, G.D. & Blakeney, A.B.
Lonergan, P.F. 2001. Genetic characteriza- 1995. Relationships between minerals in
tion of QTL mapping of zinc nutrition in Australian brown rice. Journal of the Science
barley (Hordeum vulgare L.). PhD thesis, of Food and Agriculture, 68: 285291.
University of Adelaide, Australia. Marentes, E. & Grusak, M.A. 1998. Iron
Longnecker, N.E. & Robson, A.D. 1993. transport and storage within the seed coat
Distribution and transport of zinc in and embryo of developing seeds of pea
plants. InR.D. Graham, R.J. Hannam & (Pisum sativum L.). Seed Science Research,
N.C. Uren, eds. Zinc in soils and plants, 8: 367375.
pp. 113124. Dordrecht, Netherlands: Mares, D.J. & Campbell, A.W. 2001. Mapping
Kluwer Academic Publishers. components of flour and noodle colour
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 445
Piergiovanni, A.R., Rizzi, R., Pannacciulli, E. & Takagi, S. 1976. Naturally occurring iron-
Della Gatta, C. 1997. Mineral composition in chelating compounds in oat- and rice-root
hulled wheat grains: a comparison between washings. Soil Science and Plant Nutrition,
emmer (Triticum dicoccon Schrank) and 22: 423433.
spelt (T. Spelta L) accessions. International Takahashi, M., Nakanishi, H., Kawasaki, S.,
Journal of Food Science and Nutrition, Nishizawa, N.K. & Mori, S. 2001. Enhanced
48(6): 381386. tolerance of rice to low iron availabil-
Quackenbush, F.W., Firch, J.G., Brunson, A.M. ity in alkaline soils using barley nico-
& House, L.R. 1963. Carotenoid, oil, and tianamine aminotransferase genes. Nature
tocopherol content of corn inbreds. Cereal Biotechnology, 19: 466469.
Chemistry, 40: 250259. Tracy, M.I. & Moller, G. 1990. Continuous flow
Rengel, Z., Batten, G.D. & Crowley, D.E. 1999 vapour generation for inductively coupled
Agronomic approaches for improving the argon plasma spectrometric analysis. Part 1:
micronutrient density in edible portions of Selenium. Journal AOAC, 73: 404410.
field crops. Field Crops Research, 60: 2740. Treeby, M., Marschner, H. & Romheld, V.
Randolf, L.F. & Hand, D.B. 1940. Relation 1989. Mobilization of iron and other micro-
between carotenoid content and number of nutrient cations from a calcareous soil by
genes per cell in diploid and tetraploid corn. plant-borne, microbial and synthetic metal
Journal of Agriculture, 60: 5164. chelators. Plant and Soil, 114: 217226.
Scholz, G., Seifert, K. & Grun, M. 1987. The Tsang, M.L.S. & Schiff, J.A. 1978. Studies of
normalizing factor for the tomato mutant sulfate utilization by algae. Plant Science
chloronerva. 26. The effect of nicotianamine Letters, 11: 177183.
on the uptake of Mn2+, Zn2+, Cu2+, Rb+ and Tveitnes, S., Singh, B.R. & Ruud, L. 1996.
PO43- by the tomato mutant chloronerva. Selenium concentration in spring wheat
Biochemie und Physiologie der Pflanzen, as influenced by basal application and top
182: 189194. dressing of selenium-enriched fertilizers.
Shi, R., Li, H., Tong, Y., Jing, R., Zhang, F. & Fertilizer Research, 45: 163167.
Zou, C. 2007. Identification of quantitative Umaly, R.C. & Poel, L.W. 1971. Effects of
trait locus of zinc and phosphorus density iodine in various formulations on the growth
in wheat (Triticum aestivum L.) grain. Plant of barley and pea plants in nutrient solution
and Soil, 306: 95104. culture. Annals of Botany, 35: 127131.
Shinonaga, T., Gerzabek, M.H., Strebl, F. & Vasconcelos, M., Datta, K., Oliva, N.,
Muramatsu, Y. 2001. Transfer of iodine from Khalekuzzaman, M., Torrizo, L., Krishnan,
soil to cereal grains in agricultural areas of S., Oliveira, M., Goto, F. & Datta, S.K. 2003.
Austria. Science of the Total Environment, Enhanced iron and zinc accumulation in
267(1-3): 3340. transgenic rice with the ferritin gene. Plant
Stephenson, L.S., Latham, M.C. & Otteson, Science, 164: 371378.
E.A. 2000. Global malnutrition. Parasitology, Wahlqvist, M.L. 1997. Vitamins and vitamin-
121: S5-S22. like compounds. In M.L. Wahlqvist, ed.
Symcox, K. D., Shadley, J. D. & Weber, D. Food and nutrition, pp. 222248. Sydney,
F. 1987. Detection of the time of occur- Australia, Allen & Unwin.
rence of non-disjunction induced by the Wang, Z.J. & Gao, Y.X. 2001. Biogeochemical
r-x1 deficiency in Zea maize L. Genome, cycling of selenium in Chinese environments.
29: 782785. Applied Geochemistry, 16: 13451351.
Breeding for quantitative variables. Part 4: Breeding for nutritional quality traits 447
Welch, R.M. 1995. Micronutrient nutrition of genes phytoene synthase and beta-carotene
plants. Critical Reviews in Plant Science, desaturase associated with the accumulation
14: 4982. of carotenoids in maize. Theoretical and
Welch, R.M., House, W.A., Beebe, S., Applied Genetics, 108: 349359.
Senadhira, D., Gregorio, G.B. & Cheng, Z. Yamada, H., Takeda, C., Mizushima, A.,
2000. Testing iron and zinc bioavailability Yoshino, K., & Yonebayashi, K. 2005. Effect
in genetically enriched beans (Phaseolus of oxidizing power of roots on iodine
vulgaris L.) and rice (Oryza sativa L.) uptake by rice plants. Soil Science and Plant
in a rat model. Food Nutrition Bulletin, Nutrition, 51(1): 141145.
21: 428433. Yuita, K., Nobusawa, Y., Shibuya, M. & Aso,
Welch, R.M. & Graham, R.D. 2004. Breeding S. 1982. Iodine, bromine and chlorine con-
for micronutrients in staple food crops from tents in soils and plants of Japan. 1. Iodine,
a human nutrition perspective. Journal of bromine and chlorine contents in soils and
Experimental Botany, 55: 353364. plants of the basin of the Miomote River.
White, P.J., Bowen, H.C., Parmaguru, P., Fritz, Soil Science and Plant Nutrition, 28(3): 315
M., Spracklen, W.P., Spiby, R.E., Meacham, 336.
M.C., Mead, A., Harriman, M., Trueman, Zechmeister, L. & Cholnoky, L. 1940.
L.J., Smith, B.M., Thomas, B. & Broadley, Carotenoids of Hungarian wheat
M.R. 2004. Interactions between seleni- flour. Journal of Biological Chemistry,
um and sulphur nutrition in Arabidopsis 135: 3136.
thaliana. Journal of Experimental Botany, Zhang, F., Romheld, V. & Marschner, H. 1991.
55(404): 19271937. Release of zinc-mobilizing root exudates
WHO [World Health Organization]. 2003. in different plant species as affected by
Iodine Deficiency Disorders. WHO, zinc nutritional status. Journal of Plant
Geneva, Switzerland. Available at www. Nutrition, 14: 675686.
who.int/en/ Zhang, L., Wang, X., Zhou, X. & Shi, W. 2004.
Wong, J.C. 1999. Molecular marker mapping Genotypic difference in selenium accumu-
of chromosomal regions associated with lation in rice seedlings at early growth stage
levels of carotenoids and tocopherols in and the relation with selenium content in
maize kernels. MSc thesis, University of rice seeds. In Abstracts of the International
Illinois, Urbana-Champaign, USA. Symposium on Trace Elements and Health,
Wong, J.C., Lambert, R.J., Wurtzel, E.T. & Hangzhou, China, 1013 October 2004.
Rocheford, T.R. 2004. QTL and candidate pp. 2829.
449
CHAPTER 18
breakthroughs in shifting yield ceilings. wheat, barley and rice) and how a physiological
The increase in cereal production has so understanding can contribute to reach
far been obtained mostly from irrigated such a goal. The study of crop physiology
land, through the diffusion of improved can assist cereal breeding in different
varieties and agronomic practices suitable for ways: (i) improving the understanding of
specific ecosystems. Moreover, investment the factors that determine crop yield and
in irrigation has allowed the conversion adaptation through the pedagogical (for
of rainfed ecosystems when suitable. syncretic) concept of ideotype and, as a
However, the rising costs of irrigation and consequence, improving crop simulation
the problems of management, cost recovery models; (ii) defining particular secondary
and the maintenance of existing systems traits to select for (analytical breeding) when
constrain the further expansion of irrigation. choosing parents for crossing or screening
Increasing siltation of reservoirs and canals, in segregating populations; (iii) indicating
lowering of underground water levels, and the kind of genetically modified organisms
the accumulation of salt in already irrigated (GMOs) that are worth developing and how
soils are causing environmental concerns to test them; and (iv) phenotyping associated
(IRRI, 1996). Furthermore, new growth with marker-assisted selection (MAS).
sectors, such as industry and tourism, as well Special emphasis will be devoted to
as increasing population and urbanization, those aspects that may be particularly useful
are all competing for water resources. to the National Agricultural Research
Moreover, sustainability concerns constrain Systems (NARS) of developing countries,
the adoption of intensive agronomic practices where research budgets are limited and
(MacIlwain, 2004), bringing into question prioritization is necessary. As such, we
the desirability of the further expansion of will focus on alternatives for evaluating
irrigation (Araus, 2004). secondary traits in an economical way, and
Abiotic stresses frequently constrain the prospects for the new array of molecular
growth and productivity of major crop spe- techniques available.
cies such as cereals. They have been specifi-
cally covered in Chapter 16 and therefore 18.2 ANALYTICAL BREEDING
we will only marginally refer to them here. Genetic improvement may be achieved
While genetic increases in yield potential through selection either:
are best expressed in optimal environments, r EJSFDUMZ
GPSBQSJNBSZUSBJU TVDIBTHSBJO
they are also associated with better yields yield) in a target environment (Ceccarelli
under drought (Trethowan, van Ginkel and and Grando, 1996). This has been referred
Rajaram, 2002; Araus et al., 2002), nitrogen to as empirical or pragmatic breeding; or
deficiency (Ortiz-Monasterio et al., 1997) r JOEJSFDUMZ
GPSBTFDPOEBSZUSBJU
UIBUNVTU
and heat-stressed environments (Reynolds be putatively related to a higher yield
et al., 1998). However, this is disputed by potential or to an improved behaviour of
several authors, as discussed in Chapters 3, the crop when it is grown in a stressful
13, 14 and 16. environment. This is known as analytical
This chapter will discuss techniques for or physiological breeding.
improving yield potential (and eventually Traditionally, breeders have achieved
adaptiveness to unfavourable environmental yield increases by intercrossing elite lines
conditions) for small-grain cereals (such as and selecting the highest- and most stable-
452 Plant breeding and farmer participation
Time to flowering, and plant height 1996; Slafer and Whitechurch, 2001). Aside
The timing of flowering is one major trait from its adaptive advantage, the reduction
related to the adaptation of cultivars to in crop duration has permitted an increase
particular growing areas, thus determining in cropping intensity for cereals such as
crop performance under the prevalent field rice, and also allowed land to be used for
conditions (e.g. Perry and DAntuono, 1989; growing non-cereal crops in cereal-based
Passioura, 1996, 2002; Richards, 1996a; farming systems (IRRI, 1996).
Slafer and Araus, 1998). This is why time to Breeders have always selected for reduced
flowering (phenological adjustment) is one height when the initial elite material is taller
of the first attributes optimized by breeding than a threshold. For plant statures above this
programmes (Slafer, 2003). Consistent threshold (70100 cm in wheat; Richards,
changes in this trait are therefore only to be 1992; Miralles and Slafer, 1995), there is no
expected in regions in which lengthening gain in biomass while there is a proportional
or shortening the growing season may reduction in harvest index. Below this
have represented advantages for adaptation threshold, the further gain in harvest index
compared with cultivars released earlier. does not compensate for the loss in biomass,
A scenario frequently reported in the due to extremely poor radiation distribution
literature, where the manipulation of time within the canopy and consequent reductions
to heading may have a strong impact on in radiation-use efficiency (e.g. Miralles and
adaptation, is that of regions characterized Slafer, 1997). Thus, reducing height down to
by a Mediterranean climate, i.e. a dry hot an optimum range increases yield potential (a
summer and humid, temperate winter similar biomass more efficiently partitioned
(Perry and DAntuono, 1989; Loss and to grains) and simultaneously reduces the
Siddique, 1994; Acevedo et al., 1999). In risk of lodging causing yield penalties.
these environments, the crops vegetative Consequently, plant height has been reduced
and early reproductive phases occur under universally through wheat breeding during
reasonably good water availabilities. the 20th century (see a number of cases
However, as the season progresses, drought reviewed by Calderini, Reynolds and Slafer,
becomes more intense and frequent water 1999, and more recently reported studies:
stresses occur during the late reproductive Donmez et al., 2001; Brancourt-Hulmel et
phases. After anthesis, grains fill under al., 2003; Jiang et al., 2003; Ramdani et al.,
rather severe water and heat stresses. The 2003).
analysis of long-term trends in time to Reduced plant height has been one of the
flowering for cultivars released in different major causes of lack of adoption of modern
eras reveals that, in most cases, there seemed varieties in a number of crops grown by
to be little or no change in regions with farmers in developing countries as animal
climates different from the Mediterranean, feed and where the way in which total bio-
while reducing time to flowering has been mass is portioned is irrelevant (Ceccarelli et
a successful strategy when breeding for al., 2000, Annicchiarico et al., 2005).
environments characterized by terminal
stresses. This is because earliness is Dry matter production and its partitioning
probably the most effective solution for As yield has been greatly increased by
increasing yield where drought during genetic improvement during the past century
grain filling is a common event (Passioura, (Figure 18.1) and grain yield is simply a
454 Plant breeding and farmer participation
50 US ( Donmez et
et al.,
al.,2001)
Argent(Calderiniet
Argent
2001)
(Calderini etal.,
al.,1995)
1995)
in the United Kingdom. However, except
40 for that last study (Shearman et al., 2005),
30
these differences have not been associated
with increases in biomass production.
20
Looking in closer detail at post-anthesis, it
10
seems, however, that modern cultivars may
0 be characterized by higher radiation-use
1840 1860 1880 1900 1920 1940 1960 1980 2000
Year of release
efficiency levels than their predecessors.
Source: Austin, Ford and Morgan, 1989
This was the pattern observed in the only
retrospective analysis of this characteristic
that we are aware of (Calderini, Dreccer
fraction of the dry matter accumulated by and Slafer, 1997). This increased radiation-
the crop during the growing season, one use efficiency may just be a consequence
might have expected a trend for increased of the increased demand of a larger sink
biomass as newer higher-yielding cultivars in modern cultivars (see below) compared
have been released. However, in almost all with their older counterparts. These kinds of
cases in which the physiological bases of results are in line with evidence from near-
yield improvements have been analysed, isogenic lines for semi-dwarfism (Miralles
final biomass was not associated with the and Slafer, 1997) and with the fact that the
year of release of the cultivars (Figure 18.2, number of grains growing after anthesis
top), with only a few exceptions to this positively influences photosynthetic
general trend. Thus, for example, recent efficiency (Richards, 1996b). Also, it was
genetic gains in grain yield in the United recently reported that increased post-
Kingdom have resulted from a combination anthesis biomass may be achieved by
of improved growth rate in the pre-anthesis increased sink strength through positive
period, which has driven increases in the feedback to photosynthesis (Reynolds et
number of grains per unit area, and a larger al., 2004). A higher stomatal conductance
source for grain filling through increases seems to be responsible, at least in part, for
in stem soluble carbohydrate reserves the higher photosynthetic rates (Fischer et
(Shearman et al., 2005). al., 1998; Araus et al., 2002).
Breeding for quantitative variables. Part 5: Breeding for yield potential 455
Therefore, the genetic gains in grain yield The increased number of grains per
in most countries were virtually entirely unit and per area of the modern cultivars
due to modifications in harvest index compared with their predecessors seems to
(Figure 18.2, bottom), probably related to be the consequence of a higher survival of
reductions in plant height (see above). In the floret primordia, as the number of potential
few studies in which biomass partitioning florets per spike is apparently quite similar
was analysed before maturity in cultivars (e.g. Slafer and Andrade, 1993; Miralles
released in different eras, it seemed clear that et al., 2002). Thus, the higher survival of
the partitioning effects observed at maturity, floret primordiala process taking place
as differences in harvest index, were already during the last half of the stem elongation
established by anthesis as variations in the period (Kirby, 1988)appears to be the
spike-to-stem ratio (Siddique, Kirby and most important factor leading to higher
Perry, 1989; Slafer and Andrade 1993). yield potential in modern cultivars (Slafer
A critical issue in this respect is that et al., 1994). This is in agreement with the
harvest index in most modern cultivars finding that semi-dwarfing genes, which
seems to be close to its biological maximum contribute significantly to increased yields
(ca. 60 percent; Austin, 1980). Consequently, in many breeding programmes throughout
even though breeding in the past has been the world, increase the number of grains
very successful in increasing grain yield per unit land area by increasing the survival
through reducing height and increasing of floret primordia (Miralles et al., 1998).
harvest index, it appears imperative to find Actually, the giberellic acid-insensitive
alternatives for improving biomasswhile dwarfing genes Rht-B1b and Rht-D1b, the
maintaining harvest indexif further two most important commercially, have
genetic gains in yield are to be expected. been reported to reduce plant height by
around 18 percent, simultaneously exerting
Main yield components large pleiotropic effects improving spike
The two main yield components are the fertility (Flintham et al., 1997).
number of grains per unit area and the aver-
aged individual grain weight. Most experi- Associations between attributes changed
ments analysing genetic improvement effects while selecting for higher yield
on yield have considered these components. The genetic improvement of wheat yield
The vast majority of the reported studies may be understood more mechanistically by
found that, while selecting for higher-yield- inspecting and interpreting the relationships
ing cultivars, wheat breeders have consist- between the main attributes of growth
ently increased the number of grains per and partitioning of the yield components
unit land area and produced either no trend described above. Understanding the
or even a trend to slightly reduced individ- mechanistic bases by which breeding has
ual grain weight (Calderini, Reynolds and successfully increased yield may shed light
Slafer, 1999; Donmez et al., 2001; Garca del on possible future alternatives.
Moral et al., 2002; Brancourt-Hulmel et al., Genetic gains in grain yield were almost
2003), though earlier studies have found a unequivocally due to gains in the number
slight increase in this component during the of grains per unit land area, with no gains
20th century (Cox et al., 1988) or part of it and even slight losses in the average weight
(Calderini, Dreccer and Slafer, 1995). of the grains, probably as a consequence
456 Plant breeding and farmer participation
of the lower size of the cellules induced immediately before anthesis (coinciding
by insensitivity to giberellic acid. This has with stem elongation) than to any changes
given rise to a frequently found negative in growth occurring before this time (e.g.
relationship between grain number per Kirby, 1988; Savin and Slafer, 1991; Fischer,
unit land area and the average weight of 1993; Demotes-Mainard and Jeuffroy,
those grains (Slafer, Calderini and Miralles, 2001). In all these cases, there was a strong
1996). In fact this negative relationship is relationship between grain number per unit
also frequent when yield is increased by area at maturity and spike dry matter per
management practices. Although the most unit area around anthesis (Slafer, 2003).
common interpretation in the literature Breeding seems to have increased grain
has been an increased competition among number per unit area precisely through
grains as the number of grains per unit the same mechanism: crop growth has not
are is increased; we argue that the negative clearly and systematically been affected by
relationship is not competitive in nature breeding for higher yield, but there has been
(Slafer, 2003). This is because, even though a consistent trend to reduce plant height (a
wheat breeding has been reducing the trait determined during the stem elongation
degree of post-anthesis sink limitation phase), leading to increased partitioning
to yield (e.g. Kruk, Calderini and Slafer, towards the growing spikes resulting in
1997), the photosynthetic capacity during an increased spike dry matter associated
grain filling together with the pre-anthesis with more grains per unit land area (e.g.
assimilate reserves seem to be in excess of Siddique, Kirby and Perry, 1989; Slafer and
the demands of the growing wheat grains Andrade, 1993). Studies demonstrating a
during post-anthesis (e.g. Richards, 1996b; consistent increase in yield caused by Rht
Slafer, Calderini and Miralles, 1996; Borras, genes in a wide variety of conditions also
Slafer and Otegui, 2004; Reynolds et al., point to the importance of this mechanism
2004). The main conclusion from this overall (e.g. Brooking and Kirby, 1981; Fischer and
analysis is that, with only a few exceptions, Stockman, 1986; Miralles et al., 1998).
wheat yield is limited by sink during grain In this context, one can understand the
filling and that further increases in yield positive relationship between the number
depend upon increases in sink-strength after of grains per unit area and harvest index
anthesis (to either further increase grain that can be found almost without exception
number per unit land area or to increase when comparing modern and old wheat
potential size of the individual grains). cultivars. The number of grains is simply
The number of grains per unit land a reflection of changes in partitioning
area is determined by various factors, operating before anthesis, resulting in lower
including plants per unit land area, spikes vegetative biomass.
per plant, spikelets per spike and grains
per spikelet. These factors are sensitive to 18.2.2 Yield potential versus stress
growing conditions throughout the entire adaptation: GE interaction
period from sowing to anthesis (Slafer and The limitations of empirical breeding are
Rawson, 1994). However, the number of more evident when selecting for stress (e.g.
grains per unit area seems to be far more drought) adaptation due to the existence
sensitive to changes in growth partitioning of important GE interactions, and higher
over a rather short window of time within-site variability that also diminishes
Breeding for quantitative variables. Part 5: Breeding for yield potential 457
heritability (h2) (Richards, 1996a; Araus et a high yield potential phenotype. If selec-
al., 2002) even though this can not be gen- tion can address factors of stress adaptation
eralized (Al Yassin et al., 2005; Comadran in addition to yield under stress, perhaps
et al., 2008). Thus, although selecting for higher yield potential and drought resist-
yield per se in the targeted environment ance can be recombined (Blum, 2005).
may be sensible if the stress is uniformly As mentioned earlier, this is one of the
severe (e.g. Ceccarelli and Grando, 1996), most controversial topics in plant breeding
this is not the case in many realistic situ- and the views presented in this paragraph
ations. For example, cultivars tested in a represent one of the philosophies concerning
particular set of stressful conditions may GE interaction. Different views and
not behave well in another set (Cooper interpretations are discussed elsewhere in
et al., 1997). Moreover, a crossover effect this volume (for example, Chapters 4, 14,
in the yield of genotypes of high and low 16 and 20).
yield potential when regressed against the
environmental index over a wide range of Physiological avenues for increasing yield
conditions is not often found unless the potential
severe conditions are too extreme. This As stressed above, retrospective studies
may indicate that in general, genotypes with wheat indicate that improvement in
selected under high yielding environments yield has more often been associated with
will perform better than those with lower increased partitioning of biomass to the grain
yield potential when grown in rather wide than it has with increased overall biomass
range of yielding environments (Calderini (Austin et al., 1980; Waddington et al., 1986;
and Slafer, 1999; Slafer and Araus, 2007). Sayre, Rajaram and Fischer, 1997; Calderini,
Constitutive whole-plant traits have a Reynolds and Slafer, 1999). Since harvest
major role in affecting plant water use and index is estimated to have an upper limit
plant dehydration avoidance under stress. of just over 60 percent (Austin, 1980), and
These largely determine some of the nega- since this limit is already being approached
tive relations between yield potential and (Shearman et al., 2005), it is becoming more
the ability to sustain yield under severe important than ever to understand the phys-
water shortage (Blum, 2005). Under most iological and genetic bases of radiation-use
dryland situations where crops depend on efficiency and biomass determination if yield
unpredictable seasonal rainfall, the maxi- is to go on increasing (Araus et al., 2003b).
mization of soil moisture use is a crucial Increases in biomass have started to be
component of drought resistance (avoid- reported in spring wheat (Reynolds, Sayre
ance), which is generally expressed in lower and Rajaram, 1999) and winter bread wheat
water-use efficiency (WUE) (Blum, 2005) (Shearman et al., 2005). One study has
and may explain the positive correlations revealed increases in biomass of about 10
frequently found under Mediterranean percent in spring wheat specifically associ-
conditions between carbon isotope dis- ated with the introduction of the long arm
crimination (13C, see below) and grain of chromosome 7D from a distant relative
yield (Araus et al. 1998a, 2002, 2003c). of wheat, Lophopyrum elongatum, into a
However, selection for yield under drought number of wheat backgrounds (Reynolds et
stress resulted in a dehydration-avoidant al., 2001). Detailed physiological investiga-
phenotype that is rarely compatible with tion revealed that the basis of this increase in
458 Plant breeding and farmer participation
biomass was associated with a small increase and vascular connections. Traits that have
in assimilation rate during the spike growth shown association with improved yield in
stage, and a much larger increase in photo- populations of random sister lines include
synthetic rate during grain filling, leading above-ground biomass at flowering,
to an increased number of grains per spike spike mass at flowering, and duration
(Reynolds, Pellegrineschi and Skovmand, of rapid spike growth phase (Reynolds,
2005). Further experiments, in which grain Pellegrineschi and Skovmand, 2005).
number was increased artificially in elite An alternative approach to further raise
lines with a brief light treatment during the number of grains per unit land area
the rapid spike growth stage, showed that might be to lengthen the stem elongation
these lines posses a photosynthetic capac- phase (hypothesized by Slafer, Calderini
ity in excess of that needed to fill the and Miralles, 1996; Slafer et al., 2001). The
grains they would normally set (Reynolds, hypothesis would be that a longer stem
Pellegrineschi and Skovmand, 2005). elongation phase may result in more crop
One way to exploit this excess growth during this phase, higher spike
photosynthetic capacity would be to increase dry matter at anthesis and subsequently
grain number. CIMMYT is experimenting more grains being filled. The hypothesis
with a number of approaches, one of which only makes for a viable solution if the
being to exploit the large-spike trait. Large length of the stem elongation phase can be
spikes themselves do not necessarily result manipulated and if the expectedly higher
in a higher yield should the trait not be in biomass accumulated during the phase is not
balance with other plant characteristics. For counterbalanced by reduced partitioning
example, genotypes with large spikes often to the spikes. These manipulations might
have small and shrivelled grain. Large-spike involve genes responsible for sensitivity
genotypes frequently tiller less, presumably to photoperiod or for earliness per se, as
because they carry the tiller inhibitor gene the duration of the stem elongation phase
(Richards, 1988), which results in what seems to be governed by photoperiod
is known as yield compensation. The response (e.g. Slafer and Rawson, 1997;
challenge is therefore to bring traits together Miralles and Richards, 2000; Gonzlez,
in a balanced way such that increased Slafer and Miralles, 2002) and to present
grain number is matched by an adequate genetic variation in its minimum duration,
vascular system with the ability to fill all of the intrinsic earliness (grown under long
the additional grains, and a good tillering photoperiods after removing vernalization
capacity is combined with large spikes. requirements; Slafer, 1996). Details of this
Another trait being explored is the so hypothetical alternative can be found in
called multi-ovary characteristic, which Slafer et al. (2001). Briefly, there is clear
causes a single floret to set up to four genetic variation in the duration of stem
kernels instead of just the usual one elongation, even when holding constant
(Reynolds, Pellegrineschi and Skovmand, the duration of the entire period to anthesis
2005). Currently the trait suffers from the (Slafer and Rawson, 1994; Kernich,
problem of low kernel weight, but pre- Halloran and Flood, 1997). At least part of
breeding is underway with different spike this variation may be due to sensitivity to
architectures to try to better accommodate photoperiod. Such sensitivity varies between
the large number of grains in terms of space phenological phases (Slafer and Rawson,
Breeding for quantitative variables. Part 5: Breeding for yield potential 459
1996; Gonzlez, Slafer and Miralles, 2002). requirement for their incorporation into
Exposing plots to photoperiod extensions breeding programmes. The real value of
during the stem elongation phasenatural a given trait may only be assessed by
day length was maintained before this determining the genetic gain in segregating
phaseproduced a change in the duration populations following selection. However,
of the phase, associated with changes in many traits are not available in well adapted
the number of fertile florets and grains genotypes, and their validation frequently
due to modifications in spike dry matter requires the development of appropriate
at anthesis (Gonzlez, Slafer and Miralles, breeding material, which is costly and
2003a, b). This suggests that the mechanism time consuming ( Royo et al., 2005).
by which photoperiod alters the final Moreover, the evaluation of some of the
number of grains per unit area is the same traits proposed by plant physiologists
as that determined by radiation interception can be time-consuming, and sometimes
during stem elongation (Gonzlez, Slafer even expensive, which is not practical for
and Miralles, 2005). Therefore, isolating and application to the thousands of entries
subsequently manipulating (traditionally that comprise the segregating generations
or through marker assisted selection) of breeding programmes. In addition,
the genetic bases controlling sensitivity selection in segregating populations requires
to photoperiod (as in this example, or screening at the plant level or between very
genetic bases of differences in earliness per small plots, thus hindering the use of traits
se) during stem elongation, might be an that require large field plots to be assessed.
effective avenue for increasing yield. Nevertheless, some analytical or
indirect selection criteria have been used
18.3 THE PRACTICAL USE OF for decades in breeding programmes. Plant
SECONDARY TRAITS height, days to heading or to maturity,
The putative secondary traits for a breeding photoperiod or vernalization responses,
programme assisted by analytical selection spike length, disease reaction, tillering
can be used: capacity or grain weight are examples of
r GPSUIFTFMFDUJPOPGQBSFOUTUPCFJODMVEFE traits usually evaluated in wheat and barley
in the crossing block; or breeding programmes, both conventional
r TBT EJSFDU TFMFDUJPO DSJUFSJB GPS TDSFFOJOH and participatory (Ceccarelli et al., 2000,
among a large number of genotypes (i.e. 2003) so as to provide relevant information
segregating populations) and when the about the performance of genotypes.
amount of seed available is too small to Any trait to be chosen must fulfil a set of
carry out field trials with replications (i.e. requirements related to relevance in terms
the evaluation of double-haploid lines). of crop performance, as well as how it can
Whereas intensive work is continuously be measured. These aspects are discussed
being carried out by physiologists to increase below and summarized in several diagrams
yield potential, few breeders routinely use (Figures 18.3, 18.4 and 18.5).
the latest developed physiological criteria
in their mainstream breeding programmes. 18.3.1 How to choose a trait?
One reason may be the difficulty in Two comprehensive manuals have been
evaluating the response to the selection developed recently by CGIAR Centers
of secondary traits, this being an essential for cereals such as wheat and rice. Both
460 Plant breeding and farmer participation
FIGURE 18.4
Different requirements when choosing a candidate trait
Integrative trait
Genetic correlation with yield Negative interaction GE
Heritability > yield.
Breeding for quantitative variables. Part 5: Breeding for yield potential 461
FIGURE 18.5
How a candidate trait should be measured
How to evaluate a
trait?
5. It should be possible to measure the trait into a breeding programme is necessary (i.e.
rapidly and more economically than the to fulfil point 5). This may be particularly
yield itself, and in a reliable manner. pertinent when evaluating germplasm in
6. It can be assessed in individual plants or segregating populations.
in very small plots. Secondary traits may be particularly
We can predict whether the use of suited to improving selection response
a secondary trait can enhance expected for stress conditions, if they: (i) improve
progress in selection by calculating its precision (in those case where the
genetic correlation with yield (point 1) and heritability of yield is reduced by stress);
heritability (point 2). Any trait that fulfils (ii) avoid any confounding effects of stress
the first three requirements will provide timing in yield (e.g. drought and flowering
the breeders with a useful prediction tool. dates); (iii) focus the selection on a specific
While this may be enough for a breeding type of stress; and (iv) are cheaper, easier
programme, for a direct confirmation of and faster to measure than grain yield.
the value of a trait (validation), several Moreover, when choosing traits,
approaches can be taken, including the it is necessary to keep in mind an eco-
development of lines expressing well the physiological perspective. For example, in
secondary trait and the assessment of their the case of drought, most traits proposed
performance in the target environment, as by stress physiologists appear to be
well as the identification of the co-segregation associated with stress tolerance (i.e. tolerate
of quantitative trait loci (QTLs) for the trait cell dehydration). If the target environment
as well as for yield (Lafitte, Blum and Atlin, is under severe stress-prone conditions it
2003). Once a given trait is chosen as a may be helpful to select for this kind of
candidate, a practical way to incorporate it trait. In most circumstances, however, the
462 Plant breeding and farmer participation
main effect of drought is to reduce grain a given region, selection for a flowering
yield without killing the plant. In such date that does not coincide with the period
cases, breeding for higher yield potential of water deficit (i.e. it exhibits an escape
plus traits conferring stress avoidance strategy) is a very effective way to improve
(i.e. to avoid cell dehydration) may be in drought adaptation (Araus et al., 2002). The
general a better choice (see examples in limitations to this approach are that very
Araus et al., 2002, 2003a, b) but there is early varieties may suffer yield penalty in
ample evidence that this is the exception good seasons, while late-in-season freezing
rather than the rule (Chapter 16). At the episodes may affect spike fertility.
same time, and indeed in most cases, trait
evaluation should be carried out under field Spike fertility
conditions, avoiding those experimental When stress occurs near flowering, the
situations (growth chambers, greenhouses, most sensitive growth stage, the main yield
pots) that are far from the agricultural component affected is the percentage of
growing environment. fertile spikelets. This trait, important for
any cereal, is critical for rice under water
18.3.2 Which traits to use in practice? stress (Lafitte, Blum and Atlin, 2003).
While many traits have been studied for
their use in breeding for drought resistance, Plant growth and senescence
there is a general consensus among breeders Stress may accelerate the senescence of
that only a few of them can be recommended leaves. To measure leaf desiccation, it is
for use in practical breeding programmes possible to make a visual integration of
at this time. For example, CIMMYT the symptoms, translated to a ranking.
(Reynolds, Ortiz-Monasterio and McNab, Also, to check for early senescence of
2001) and IRRI (Lafitte, Blum and Atlin, leaves, particularly the flag leaves, portable
2003) recommend the use of flowering chlorophyll meters are extensively used.
and maturity dates, spike fertility, changes At the same time, stresses such as drought
in green biomass (e.g. leaf death score) strongly affect leaf expansion (Royo et
and canopy temperature. The manuals al., 2004) and thus plant growth (Villegas
developed by both institutions include et al., 2001) and further yield (Araus
a comprehensive explanation of how to et al., 2002). Therefore the total green
measure these traits. In practical terms, biomass evaluated at a critical plant stage
these traits seem valuable when breeding (i.e. anthesis) or its change over time is
for higher yield potential and adaptation to a potentially powerful traits. A feasible
moderate degrees of stress. Development of evaluation of total biomass is only possible
modern equipment and new analytical tools in practice through indirect methods, such
should facilitate future measurements of as using spectroradiometers to measure
additional physiological traits in the field. the spectra of light reflected by the canopy
(Aparicio et al., 2000; Araus, Casadess
Phenology and Bort, 2001; Royo et al., 2003). In
This is the most widely used of secondary many cases, however, the wide range of
traits, due to its ease of measurement different spectroradiometrical indices have
(see section 18.2.1. in this chapter). If the not fulfilled expectations when evaluating
pattern of water deficit is predictable in field plots for yield and their adaptation
Breeding for quantitative variables. Part 5: Breeding for yield potential 463
with NARS, has a multi-panel leaf colour fact the most useful for routine breeding
chart developed and calibrated for use with purposes. Moreover, as the GreenSeeker
rice throughout Asia (Shukla et al., 2004; includes its own radiation source, it may
IRRI, 2005). be used independently of the atmospheric
conditions (sunny or cloudy day) and,
Biomass assessment more importantly, its cost is comparable (or
Field spectroradiometers able to measure even less) that of a SPAD (below 3 000).
the spectrum of light reflected by the Conventional digital cameras are low
canopy have been expensive devices cost devices that could be adopted for
(exceeding 12 000). However, the situation generalized use in a number of agricultural
is changing. Designed initially for nitrogen applications, including plant breeding.
management, a simple and easy-to-handle Through adequate processing of the
spectroradiometer, such as GreenSeeker, information contained in digital pictures, it
has become a potentially very useful device is possible to evaluate total green biomass
for breeding (Figure 18.6). It gives only much more cost effectively than with land-
the basic spectroradiometric indices of based portable spectroradiometers. In
green biomass, such as NDVI (as well as addition, digital pictures may also provide
the inverse of the SR). Nevertheless, as information that is not currently acquired
noted above, these single indices are in through spectral reflectance measurements,
FIGURE 18.6
An example of cheap, easy-to handle spectroradiometer: the GreenSeeker.
Evaluation of dry weight of pot plants
BIOMASS / NDVI
60,00
40,00
SHOOT ( gr. DW)
30,00
20,00
10,00
0,00
0,000 0,100 0,200 0,300 0,400 0,500 0,600 0,700 0,800 0,900
NDVI
466 Plant breeding and farmer participation
FIGURE 18.7
Some applications of digital photography: Ratio of green area to total area.
Easy-to-calculate estimator of green cover
Courtesy of Dr J. Casadess.
such as the degree of soil covered by for the reasons mentioned previously. The
the crop (Figure 18.7), the proportion of decision to advance or reject a genotype
yellow leaves, or even yield components is often complex, and in practical terms
such as the number of spikes per unit land breeders most often use a system of multiple
area (Casadess, Biel and Sav, 2005). cut offs. The usual approach is to carry out
the selection of early generations in stress-
18.3.4 When to use the traits? free environments, in order to optimize the
Grain yield is the primary trait for selection expression of desired traits in the plant and
in breeding programmes aimed at both simultaneously maximize heritability and
increasing yield potential and adaptation to response to selection. In early generations,
stress-prone environments. When breeding breeders select genotypes that presumably
for drought adaptation, the conceptual achieve the levels required for the primary
model used considers yield under drought traits evaluated in segregating populations
to be a function of: (i) yield potential; (ii) the (resistance to diseases, plant type, plant
flowering date (which indicates whether height, growth cycle, spike fertility, etc.),
the crop will avoid drought stress); and choosing only those with potential good
(iii) traits that provide drought resistance. score expression in the next generation,
Most breeders select strongly for traits when they will be assessed again. Quality
other than yield in the early segregating is frequently evaluated early, at the family
generations, and do yield testing only level, in order to detect those crosses with
at later stages, when a certain level of desirable characteristics. It may also be
homozygosity has been achieved and worth evaluating in early generations for
sufficiently large seed quantities are some additional physiological traits that
available. While acknowledging the may give an indication of yield potential or
importance of secondary traits, most crop adaptation to abiotic stresses.
breeding programmes are not able to At subsequent stages, in more advanced
integrate them into their selection schemes generations, multi-site field experiments are
Breeding for quantitative variables. Part 5: Breeding for yield potential 467
conducted in order to study the adaptation interaction with target beneficiaries (IRRI,
of lines to the target environments. The 1996).
combination of yield data with data As Bnziger et al (2000) state in a com-
regarding secondary traits in environments prehensive breeding manual for abiotic
ranging from well watered to high stress stress breeding published by CIMMYT, if
allows one to ascertain the adaptability of the variety being developed for improved
genotypes to a wide range of conditions, tolerance to any stress is unacceptable to
thus making possible more reliable farmers for other reasons and is not adopt-
decisions. At this level, selection is mostly ed, all the research work invested in that
based on the main goals of the programme, variety will be wasted. It is critically impor-
usually focusing on important commercial tant, therefore, that farmers be involved in
traits. However, data on secondary traits the selection and testing process, and that
may be decisive at this stage in interpreting researchers pay careful attention to farm-
and explaining GE interactions, mostly ers views on what constitutes an appro-
when the heritability of the secondary traits priate and attractive variety under their
is higher than that of yield, and the genetic circumstances. In such a context, farmer
correlation of these traits with yield in the participatory plant breeding represents:
target environment is high. r " EJBMPHVF CFUXFFO GBSNFST BOE TDJFO-
Examples of how to implement tists to solve agricultural problems.
indirect selection for physiological traits r "XBZUPJODSFBTFUIFJNQBDUPGBHSJDVM-
for drought resistance in different cereals tural research by developing technologies
are illustrated in Fischer et al. (2003) that are more widely adopted.
r "QBUIUPNPSFQSPEVDUJWF
TUBCMF
FRVJUB-
18.4 BEYOND BREEDING ble and sustainable agricultural systems.
18.4.1 Social context Abiotic stress and genetic response for
Halls (2001) definition of ideotype has adaptation to stress will depend upon the
a wider sense than that first proposed choice of target environment. To that end,
by Donald (1968), being a plan of the farmer participatory breeding emphasizes
phenotype of a cultivar that will perform three aspects: farmers knowledge, farmers
optimally in a specific set of climatic, ability to experiment, and farmer exchange
soil, biotic and socio-cultural conditions. of information and technologies. Thus,
Emphasis on socio-cultural aspects is now breeding programmes should include
accepted as an important part of the concept participatory on-farm trials, managed by
of ideotype and embraces the concept of farmers, as part of the testing of a new
the participatory breeding approach. cultivar. This may ensure that selection
Broad-based research, including research has been effective, and that progress made
on socio-economic aspects of cereal at the station will be transferable to the
production, is needed to characterize cereal farm level. Participatory trials should be
ecosystems in terms of people and their run concurrently with advanced multiple-
environment; to improve understanding of environment trials. Moreover, testing for
farming systems, indigenous knowledge, grain quality (and other quality attributes
and farmers practices; and to refine the of the crop), in consultation with farmers
definitions of the kinds of technologies that from the target population environments, is
should be developed. This will require close cheaper than replicated yield testing. Hence,
468 Plant breeding and farmer participation
quality screening should be carried out healthier soils. These practices are known
before multiple-environment trials, so as to as conservation (or zero) tillage, and are a
discard those varieties whose quality would potent tool for stabilizing and improving
be unacceptable to farmers (Atlin, 2003). soils (Bradford and Peterson, 2000)
A comprehensive methodological approach Although conservation tillage is not new
for farmer participatory breeding can be (Cline and Hendershot, 2002), it is an
also found in Ceccarelli et al. (2000, 2003). alternative strategy that agronomists have
been promoting recently in Asia and other
18.4.2 Crop management and parts of the world as a means of combating
sustainability of cropping systems declining crop productivity (Hobbs, Giri
Breeding is just half of the equation for and Grace, 1997). Conservation tillage has
more productive and sustainable crops, the significant agricultural and environmental
other half being agronomic management. benefits (see Bradford and Peterson, 2000).
The progress that has been achieved for Improved soil health means less erosion and
grain yield has been the result of combin- higher productivity, as well as reducing the
ing improved varieties with appropriate probability of encroachment of crops into
crop management strategies (Cooper et natural ecosystems. Less carbon emissions
al., 2004). Increased demand for staple are produced thanks to the reduced use of
crops has resulted in the intensification of fossil fuels, there is less oxidation of soil
agriculture all over the developing world, organic matter, and less burning of crop
and one of the most serious consequences residues. Other environmental impacts
of this is soil degradation. When soil is over time may include less pesticide use as a
no longer part of a natural ecosystem, if result of suppression of weeds by residues,
not properly managed, its physical and and more integrated control of insects. Dust
biological properties become degraded and levels in the atmosphere are also reduced
productivity declines. This has been docu- because crop residues protect the soil from
mented, for example, in the ricewheat sys- wind and there is less soil disturbance
tems of the Punjab in South Asia (Timsina during field operations. The adoption of
and Connor, 2001). Left unchecked, this zero-tillage on 1.3 million hectares of wheat
process eventually leads to soil loss through in South Asia has been achieved through the
erosion and problems of chemical imbal- initiatives of the RiceWheat Consortium
ance, such as salinity. Water scarcity adds for the Indo-Gangetic Plains (www.rwc.
to the problem, and is intensified in poor cgiar.org/rwc).
soils as they lose the capacity to absorb and A management approach for
retain moisture. The key to this downward increasing input use efficiency in irrigated
spiral is the loss of soil organic matter. environments is the cultivation of crops
While breeding can improve the tolerance on raised beds, such that soil movement
of cultivars to salinity and reduced mois- is reduced, resulting in greater water and
ture, it is not a sustainable solution if soils nitrogen use efficiency as well reduced
continue to degrade. Since soil tillage is the herbicide use (Sayre and Moreno-Ramos,
principal cause behind the declining levels 1997; Sayre, 2004). Another technology that
of organic matter, crop management prac- is likely to have water-saving applications
tices that minimize tillage whilst keeping and is very compatible with raised bed
crop residues on the soil surface result in cultivation is alternate-furrow irrigation.
Breeding for quantitative variables. Part 5: Breeding for yield potential 469
It is well established that when plant roots plant breeding, specifically targeted to
detect a drying of the soil profile, they send specific features of crop management. For
chemical signals to their leaves resulting in a example, it has been shown that tillage
reduction in transpiration rate mediated by practices influence the composition and
reduced stomatal conductance (Davies and intensity of mycorrhizal fungi over a range
Zhang, 1991). The result of this response of soil depths. New ideotypes may be
is a decreased growth rate, but an increase better adapted to emergence and growth
in water-use efficiency. This innate drought among crop residues, to soils with higher
response mechanism can be exploited by a levels of organic nutrient sources, and to
modification in irrigation strategy, whereby an increased level and diversity of soil
the side from which plants, growing on top fauna and micro-organisms. This potential
of ridges, are irrigated alternately (Davies, synergy between breeding and innovative
Wilkinson and Loveys, 2002). This was input-use-efficient crop management
tested in furrow-irrigated maize in China and practices has been referred to as the Doubly
proved to be highly effective in permitting Green Revolution (Conway, 1997).
substantially reduced water application
while increasing harvest index and thus not ACKNOWLEDGMENTS
significantly reducing yield (Kang et al., This study was supported in part by the
2000). This simple but effective modification European research project OPTIWHEAT
in irrigation strategy has clear potential (INCO-STRIP 015460) and by the Spanish
benefits that can probably be adapted to Ministry of Science and Technology projects
many irrigated cropping systems. AGL-2006-13541-C02-1 (for J.L. Araus),
Precision agriculture is also being applied AGL 2006-07814/AGR (for G.A. Slafer)
to increase resource use efficiency. For and AGL-2006-09226-C02-01 and RTA
example, agronomists at Oklahoma State 2005-00014-C04-01 (for C. Royo).
University have adapted spectral reflectance
sensor technology for use by farmers to REFERENCES
precisely calculate the crop requirement Acevedo, E., Silva, P., Silva, H. & Solar, B.
for in-season nitrogen application. By 1999. Wheat production in Mediterranean
measuring an index (NDVI) early in the crop environments. In E.H. Satorre & G.A.
cycle, levels of N in crop can be predicted Slafer, eds. Wheat: Ecology and Physiology
(www.nue.okstate.edu). The value of this of Yield Determination, pp. 295331. New
technology has already been demonstrated York, USA, Food Product Press.
in Mexican and Ecuadorian farmers fields Al Yassin, A., Grando, S., Kafawin, O., Tell, A.
in collaboration with CIMMYT scientists, & Ceccarelli, S. 2005. Heritability estimates
enabling them to apply precise amounts in contrasting environments as influenced
of fertilizer for yield optimization while by the adaptation level of barley germplasm.
minimizing N runoff. Annals of Applied Biology, 147: 235244.
This Tillage Revolution, as it is Annicchiarico, P., Abdellaoui, M., Kerkouli,
sometimes referred to, is gaining momentum M. & Zerargui, H. 2005. Grain yield,
worldwide (MacIlwain, 2004; Nelson, straw yield and economic value of tall
Giles and Gewin, 2004). The adoption and semi-dwarf durum wheat cultivars in
of resource-conserving crop management Algeria. Journal of Agricultural Science,
practices opens up new possibilities for 143: 5764.
470 Plant breeding and farmer participation
Aparicio, N., Villegas, D., Casadess, J., Araus, ing. In H.T. Nguyen & A. Blum, eds.
J.L. & Royo, C. 2000. Spectral reflectance Physiology and Biotechnology Integration
indices for assessing durum wheat biomass, for Plant Breeding. Part I. Physiological
green area, and yield under Mediterranean basis of yield and environmental adapta-
conditions. Agronomy Journal, 92: 8391. tion, pp. 149. New York, USA, Marcel
Araus, J.L. 2004. The problem of sustainable Dekker.
water use in the Mediterranean and research Araus, J.L., Villegas, D., Aparicio, N., Garca
requirements for agriculture. Annals of Del Moral, L.F., El Hani, S., Rharrabti, Y.,
Applied Biology, 144: 229272. Ferrio, J.P. & Royo, C. 2003c. Environmental
Araus, J.L., Casadess, J. & Bort, J. 2001. factors determining carbon isotope discrim-
Recent tools for the screening of physi- ination and yield in durum wheat under
ological traits determining yield. In M.P. Mediterranean conditions. Crop Science,
Reynolds, J.I. Ortiz-Monasterio & A. 43: 170180.
McNab, eds. Application of Physiology in Atlin, G. 2003. Improving drought tolerance
Wheat Breeding, pp. 5977. Mexico, D.F., by selecting for yield. In K.S. Fischer, R.
CIMMYT. Lafitte, S. Fukai, G. Atlin & B. Hardy,
Araus, J.L., Amaro, T., Zuhair, Y. & Nachit, eds. Breeding Rice for Drought-Prone
M.M. 1997. Effect of leaf structure and Environments, pp. 1422. Los Baos,
water status on carbon isotope discrimina- Philippines, IRRI (Available at: http://
tion in field-grown durum wheat. Plant www.knowledgebank.irri.org/drought/
Cell & Environment, 20: 14841494. drought.pdf; accessed 15 Jan. 2009).
Araus, J.L., Amaro, T., Casadess, J., Asbati, Austin, R.B. 1980. Physiological limitations
A. & Nachit, M.M. 1998a. Relationships to cereals yields and ways of reducing
between ash content, carbon isotope dis- them by breeding. In R.G. Hurd et al.,
crimination and yield in durum wheat. eds. Opportunities for increasing crop
Australian Journal of Plant Physiology, yields, pp. 319. Boston, USA, Pitman for
25: 835842. Association of Applied Biology.
Araus, J.L., Amaro, T., Voltas, J., Nakkoul, H. Austin, R.B. 1989. Genetic variation in pho-
& Nachit, M.M. 1998b. Chlorophyll fluores- tosynthesis. Journal of Agricultural Science
cence as a selection criterion for grain yield (Cambridge), 112: 287294.
in durum wheat under Mediterranean condi- Austin, R.B., Ford, M.A. & Morgan, C.L. 1989.
tions. Field Crops Research, 55: 209224. Genetic improvement in the yield of win-
Araus, J.L., Slafer, G.A., Reynolds, M.P. & ter wheat: a further evaluation. Journal of
Royo, C. 2002. Plant breeding and water Agricultural Science (Cambridge), 112: 295
relations in C3 cereals: what should we 301.
breed for? Annals of Botany, 89: 925940. Austin, R.B., Bingham, J., Blackwell, R.D.,
Araus, J.L., Bort, J., Steduto, P., Villegas, D. Evans, L.T., Ford, M.A., Morgan, C.L. &
& Royo, C. 2003a. Breeding cereals for Taylor, M. 1980. Genetic improvement in
Mediterranean conditions: ecophysiological winter wheat yield since 1900 and associated
clues for biotechnology application. Annals physiological changes. Journal of Agricultural
of Applied Biology, 142: 129141. Science (Cambridge), 94: 675689.
Araus, J.L., Slafer, G.A., Reynolds, M.P. & Baker, N.R. & Rosenqvist, E. 2004. Review
Royo, C. 2003b. Physiology of yield and article. Applications of chlorophyll flu-
adaptation in wheat and barley breed- orescence can improve crop production
Breeding for quantitative variables. Part 5: Breeding for yield potential 471
strategies: an examination of future pos- previous results and the latest trends. Plant
sibilities. Journal of Experimental Botany, Breeding, 114: 108112.
55: 16071621. Calderini, D.F., Dreccer, M.F. & Slafer, G.A.
Bnziger, M., Edmeades, G.O., Beck, D. & 1997. Consequences of breeding on bio-
Bellon, M. 2000. Breeding for Drought mass, radiation interception and radia-
and Nitrogen Stress Tolerance in Maize: tion use efficiency in wheat. Field Crops
From Theory to Practice. Mexico, D.F., Research, 52: 271281.
CIMMYT (available at: http://www.cim- Calderini, D.F., Reynolds, M.P. & Slafer, G.A.
myt.org/Research/Maize/map/guides_ 1999. Genetic gains in wheat yield and main
tools/BredDrougMaize/BredDroug.htm; physiological changes associated with them
accessed 15 Jan. 2009). during the 20th century. In E.H. Satorre
Blum, A. 2005. Drought resistance, water-use & G.A. Slafer, eds. Wheat: Ecology and
efficiency, and yield potentialare they Physiology of Yield Determination, pp. 351
compatible, dissonant, or mutually exclu- 377. New York, USA, Food Product Press.
sive? Australian Journal of Agricultural Calderini, D.F. & Slafer, G.A. 1998. Changes
Research, 56: 11591168. in yield and yield stability in wheat dur-
Borras, L., Slafer, G.A. & Otegui, M.E. 2004. ing the 20th century. Field Crops Research,
Seed dry weight response to sourcesink 57: 335347.
manipulations in wheat, maize and soy- Calderini, D.F. & Slafer, G.A. 1999. Has yield
bean: a quantitative reappraisal. Field Crops stability changed with genetic improvement
Research, 86: 131146. of wheat yield? Euphytica, 107: 5159.
Bort, J., Brown, R.H. & Araus, J.L. 1996. Casadess, J., Biel, C. & Sav, R. 2005. Turf
Refixation of respiratory CO2 in the ears of colour measurement with conventional dig-
C3 cereals. Journal of Experimental Botany, ital cameras. In J. Boaventura Cunha & R.
47: 5671575. Morais, eds. 2005 EFITA/WCCA Joint
Bradford, J.M. & Peterson, G.A. 2000. Congress on IT in Agriculture, Universidade
Conservation tillage. In M.E. Sumner, ed. de Trs-os-Montes e Alto Douro, Vila Real,
Handbook of soil science. pp. G247G269. Portugal, 2528 July 2005, pp. 804811.
Boca Raton, USA, CRC Press. Cassman, K.G., Dobermann, A., Walters, D.T.
Brancourt-Hulmel, M., Doussinault, G., & Yang, H. 2003. Meeting cereal demand
Lecomte, C., Brard, P., Le Buanec, B. & while protecting natural resources and
Trottet, M. 2003. Genetic improvement of improving environmental quality. Annual
agronomic traits of winter wheat cultivars Review of Environment and Resources,
released in France from 1946 to 1992. Crop 28: 315358.
Science, 43: 3745. Ceccarelli, S. & Grando, S. 1996. Drought
Brooking, I.R. & Kirby, E.J.M. 1981. as a challenge for the plant breeder. Plant
Interrelationship between stem and ear Growth Regulation, 20: 149155.
development in winter wheat: The effects Ceccarelli, S., Grando, S., Tutwiler, R., Baha,
of a Norin 10 dwarfing gene, Gai/Rht2. J., Martini, A.M., Salahieh, H., Goodchild,
Journal of Agricultural Science (Cambridge), A. & Michael, M. 2000. A Methodological
97: 373381. Study on Participatory Barley Breeding. I.
Calderini, D.F., Dreccer, M.F. & Slafer, G.A. Selection Phase. Euphytica, 111: 91104.
1995. Genetic improvement in wheat yield Ceccarelli, S., Grando, S., Singh, M., Michael,
and associated traits. A re-examination of M., Shikho, A., Al Issa, M., Al Saleh, A.,
472 Plant breeding and farmer participation
Kaleonjy, G., Al Ghanem, S. M., Al Hasan, and the exploitation of this mechanism to
A.L., Dalla, H., Basha, S. & Basha T. 2003. increase water use efficiency in agriculture.
A Methodological Study on Participatory New Phytologist, 153: 449460.
Barley Breeding. II. Response to Selection. Davies, W.J. & Zhang, J. 1991. Root signals and
Euphytica, 133: 185200. the regulation of growth and development
Clark, D.H., Johnson, D.A., Kephart, K.D. & of plants in drying soil. Annual Review
Jackson, N.A. 1995. Near infrared reflect- of Plant Physiology and Plant Molecular
ance spectroscopy estimation of 13C dis- Biology, 42: 5576.
crimination in forages. Journal of Range Demotes-Mainard, S. & Jeuffroy, M.H. 2001.
Management, 48: 132136. Partitioning of dry matter and nitrogen to
Cline, J. & Hendershot, R. 2002. Conservation spike throughout the spike growth period
tillage. In Encyclopaedia of Soil Science, in wheat crops subjected to nitrogen defi-
pp. 223226. ciency. Field Crops Research, 70: 153
Comadran, J., Russell, J.R., van Eeuwijk, F., 165.
Ceccarelli, S., Grando, S., Stanca, A.M., Donald, C.M. 1968. The breeding of crop
Francia, E., Pecchioni, N., Akar, T., ideotypes. Euphytica, 17: 385403.
Al-Yassin, A., Benbelkacem, A., Choumane, Donmez, E., Sears, R.G., Shroyer, J.P. &
W., Ouabbou, H., Bort, J., Araus, J.-L., Paulsen, G.M. 2001. Genetic gain in yield
Pswarayi, A., Romagosa, I., Hackett, C.A. attributes of winter wheat in the Great
& Thomas, W.T.B. 2008. Mapping adapta- Plains. Crop Science, 41: 14121419.
tion of barley to droughted environments. Febrero, A., Bort, J., Voltas, J. & Araus, J.L.
Euphytica, 161: 3545. 1994. Grain yield, carbon isotope discrimi-
Condon, A.G., Richards, R.A., Rebetzke, G.J. nation and mineral content in mature ker-
& Farquhar, G.D. 2004. Breeding for high nels of barley, under irrigated and rainfed
water-use efficiency. Journal of Experimental conditions. Agronomie, 2: 127132.
Botany, 55: 24472460. Ferrio, J.P., Bertran, E., Nachit, M.M., Royo,
Conway, G. 1997. The doubly green revolu- C. & Araus, J.L. 2001. Near infrared reflect-
tion: food for all in the twenty-first century. ance spectroscopy as a new surrogate analy-
Ithaca, USA, Cornell University Press. sis for 13C in mature kernels of durum
Conway, G. & Toenniessen, G. 1999. Feeding wheat. Australian Journal of Agricultural
the world in the twenty-first century. Research, 52: 809816.
Nature, 402: C55C58. Ferrio, J.P., Villegas, D., Zarco, J., Aparicio, N.,
Cooper, M., Stucker, R.E., DeLacy, I.H. & Araus, J.L. & Royo, C. 2005. Assessment of
Harch, B.D. 1997. Wheat breeding nurser- durum wheat yield using visible and near-
ies, target environments, and indirect selec- infrared reflectance spectra of canopies.
tion for grain yield. Crop Science, 37: 1168 Field Crops Research, 94: 126148.
1176. Fischer, R.A. 1993. Irrigated spring wheat and
Cox, T.S., Shroyer, R.J., Ben-Hui, L., Sears, timing and amount of nitrogen fertilizer. II.
R.G. & Martin, T.J. 1988. Genetic improve- Physiology of grain yield response. Field
ment in agronomic traits of hard red winter Crops Research, 33: 5780.
wheat cultivars from 1919 to 1987. Crop Fischer, R.A. & Stockman, Y.M. 1986. Increased
Science, 28: 756760. kernel number in Norin 10 derived dwarf
Davies, W.J., Wilkinson, S. & Loveys, B. 2002. wheat: Evaluation of a cause. Australian
Stomatal control by chemical signalling Journal of Plant Physiology, 13: 767784.
Breeding for quantitative variables. Part 5: Breeding for yield potential 473
Fischer, R.A., Rees, D., Sayre, K.D., Lu, Z.-M., that of radiation? Functional Plant Biology,
Condon, A.G. & Larqu-Saavedra, A. 1998. 32: 181188.
Wheat yield progress associated with higher Hall, A.E. 2001. Crop Responses to
stomatal conductance and photosynthetic Environment. Boca Raton, USA, CRC
rate, and cooler canopies. Crop Science, Press.
38: 14671475. Han, F., Romagosa, I., Ullrich, S.E., Jones
Fischer, K.S., Lafitte, R., Fukai, S., Atlin, B.L., Hayes, P.M. & Wesenberg, D.M. 1997.
G. & Hardy, B., eds. 2003. Breeding Rice Molecular marker-assisted selection for
for Drought-Prone Environments. Los malting quality traits in barley. Molecular
Baos, Philippines, IRRI. 98 p. (Available Breeding, 3: 427437.
at: http://www.knowledgebank.irri.org/ Hobbs, P.R., Giri, G.S. & Grace, P. 1997.
drought/drought.pdf; accessed 15 Jan. Reduced and zero tillage options for the
2009). establishment of wheat after rice in South
Flintham, J.E., Borner, A., Worland, A.J. & Asia. CIMMYT/Rice-Wheat Consortium
Gale, M.D. 1997. Optimising wheat grain for the Indo-Gangetic Plains. Rice-Wheat
yield: effects of Rht (gibberellin-insensi- Consortium Paper Series, No. 2. Mexico
tive) dwarfing genes. Journal of Agricultural DF, CIMMYT. 24 p.
Science (Cambridge), 128: 1125. IRRI [International Rice Research Institute].
Garca del Moral, L.F., Roca, L.F., Ramdani, 1996. IRRI toward 2020. Los Baos,
A., Rharrabti, Y., Villegas, D. & Royo, Philippines, IRRI. 43 p.
C. 2002. Genetic improvement effects on IRRI. 2005. Annual Report of the Director
durum wheat yield and yield components in General, 2004-05. Vol. 15. Los Baos,
the Mediterranean region. In F.J. Villalobos Philippines, IRRI. 135 p.
& L. Testi, eds. Proceedings VII Congress Jackson, P., Robertson, M., Cooper, M. &
of the European Society for Agronomy, Hammer, G. 1996. The role of physiologi-
pp. 9192. Crdoba, Spain. cal understanding in plant breeding; from a
Gonzlez, F.G., Slafer, G.A. & Miralles, D.F. breeding perspective. Field Crops Research,
2002. Vernalization and photoperiod 49: 1137.
responses in wheat reproductive phases. Jiang, G.M., Sun, J.Z., Liu, H.Q., Qu, C.M.,
Field Crops Research, 74: 183195. Wang, K.J., Guo, R.J., Bai, K.Z., Gao, L.M.
Gonzlez, F.G., Slafer, G.A. & Miralles, D.F. & Kuang, T.Y. 2003. Changes in the rate
2003a. Floret development and spike of photosynthesis accompanying the yield
growth as affected by photoperiod dur- increase in wheat cultivars released in the
ing stem elongation in wheat. Field Crops past 50 years. Journal Plant Research,
Research, 81: 2938. 116: 347354.
Gonzlez, F.G., Slafer, G.A. & Miralles, D.F. Kang, S., Liang, Z., Pan, Y., Shi, P. & Zhang, J.
2003b. Grain and floret number in response 2000. Alternate furrow irrigation for maize
to photoperiod during stem elongation in production in an arid area. Agricultural
fully and slightly vernalised wheats. Field Water Management, 45: 267274.
Crops Research, 81: 1727. Kernich, G.C., Halloran, G.M. & Flood, R.G.
Gonzlez, F.G., Slafer, G.A. & Miralles, D.F. 1997. Variation in duration of pre-anthesis
2005. Photoperiod during stem elonga- phases of development in barley (Hordeum
tion in wheat: is its impact on fertile floret vulgare). Australian Journal of Agricultural
and grain number determination similar to Research 48: 5966.
474 Plant breeding and farmer participation
Khush, G.S. 1999. Green revolution: preparing grass. Australian Journal of Plant Physiology,
for the 21st century. Genome, 42: 646655. 20: 361369.
Kirby, E.J.M. 1988. Analysis of leaf, stem and Miralles, D.J. & Richards, R.A. 2000. Response
ear growth in wheat from terminal spikelet of leaf and tiller emergence and primordium
stage to anthesis. Field Crops Research, initiation in wheat and barley to inter-
18: 127140. changed photoperiod. Annals of Botany,
Kruk, B., Calderini, D.F. & Slafer, G.A. 1997. 85: 655663.
Source-sink ratios in modern and old wheat Miralles, D.J. & Slafer, G.A. 1995. Yield, bio-
cultivars. Journal of Agricultural Science mass and yield components in dwarf, semi-
(Cambridge), 128: 273281. dwarf and tall isogenic lines of spring wheat
Lafitte, R., Blum, A. & Atlin, G. 2003. Using under recommended and late sowing dates.
secondary traits to help indentify drought- Plant Breeding, 114: 392396
tolerant genotypes. In K.S. Fischer, R. Miralles, D.J. & Slafer, G.A. 1997. Radiation
Lafitte, S. Fukai, G. Atlin & B. Hardy interception and radiation use efficiency
eds. Breeding Rice for Drought-Prone of near isogenic wheat lines with different
Environments, pp. 3748. Los Baos, The height. Euphytica, 97: 201208.
Philippines, IRRI (Available at: http://www. Miralles, D.F., Katz, S.D., Colloca, A. & Slafer,
knowledgebank.irri.org/drought/drought. G.A. 1998. Floret development in near iso-
pdf; accessed 15 Jan. 2009). genic wheat lines differing in plant height.
Loss, S.P. & Siddique, K.H.M. 1994. Field Crops Research, 59: 2130.
Morphological and physiological traits Miralles, D.J., Rharrabti, Y., Royo, C., Villegas,
associated with wheat yield increases in D. & Garca del Moral, L.F. 2002. Grain
Mediterranean environments. Advances in setting strategies of Mediterranean durum
Agronomy, 52: 229276. wheat cultivars released in different peri-
Mackill, D.J., Fukai, S. & Blum, A. 2003. ods (19002000). AAB (Association of
Indirect selection for physiological traits Applied Biologists) Conference Genotype
for drought tolerance. In K.S. Fischer, R. to phenotype: narrowing the gaps. Royal
Lafitte, S. Fukai, G. Atlin & B. Hardy, Agricultural College, Cirencester, UK.
eds. Breeding Rice for Drought-Prone Nelson, L., Giles, J. & Gewin, V. 2004. Organic
Environments, pp. 49-54. Los Baos, The FAQs. Nature, 428: 796798.
Philippines, IRRI (Available at: http://www. Ortiz-Monasterio, R.J.I., Sayre, K.D., Rajaram,
knowledgebank.irri.org/drought/drought. S. & McMahon, M. 1997. Genetic progress
pdf; accessed 15 Jan. 2009). in wheat yield and nitrogen use efficiency
MacIlwain, C. 2004. Organic: Is it the future under four N rates. Crop Science, 37: 898
of farming? Nature, 428: 792793. 904.
Masle, J., Farquhar, G.D. & Wong, S.C. 1992. Passioura, J.B. 1996. Drought and drought
Transpiration ratio and plant mineral con- tolerance. Plant Growth Regulation,
tent are related among genotypes of a range 20: 7983.
of species. Australian Journal of Plant Passioura, J.B. 2002. Environmental biology
Physiology, 19: 709721. and crop improvement. Functional Plant
Mayland, H.F., Johnson, D.A., Asay, K.H. Biology, 29: 537546.
& Read, J.J. 1993. Ash, carbon isotope Perry, M.W. & DAntuono, M.F. 1989. Yield
discrimination and silicon as estimators of improvement and associated characteristics
transpiration efficiency in crested wheat- of some Australian spring wheat culti-
Breeding for quantitative variables. Part 5: Breeding for yield potential 475
vars introduced between 1860 and 1982. to yield and biomass in elite spring wheat.
Australian Journal of Agricultural Research, In T. Fischer et al., eds. New directions
40: 457472. for a diverse planet. Proceedings of the
Ramdani, A., Martos, V., Villegas, D., Garca 4th International Crop Science Congress,
del Moral, L.F., Rharrabti, Y. & Royo, C. Brisbane, Australia.
2003. Changes on some agronomic and Rharrabti, Y., Villegas, D., Garca del Moral,
physiological traits of durum wheat culti- L.F., Aparicio, N., ElHani, S. & Royo, C.
vars released in Italy and Spain throughout 2001. Environmental and genetic determi-
the 20th century. Vol. 1. In N.E. Pogna, nation of protein content and grain yield in
M. Roman, E.A. Pogna & G. Galterio, durum wheat under Mediterranean condi-
eds. Proceedings of the 10th International tions. Plant Breeding, 120: 381388.
Wheat Genetics Symposium, pp. 163166. Richards, R.A. 1988. A tiller inhibitor gene
Paestum, Italy. in wheat and its effect on plant growth.
Reynolds, M.P., Ortiz-Monasterio, J.L. Australian Journal of Agricultural Research,
& McNab, A., eds. 2001. Application of 39: 749757.
Physiology in Wheat Breeding. Mexico D.F., Richards, R.A. 1992. The effect of dwarfing
CIMMYT. 240 p. (available at: http://www. genes in spring wheat in dry environments.
cimmyt.org/research/wheat/map/research_ I. Agronomic characteristics. Australian
results/wphysio/WPhysiology.pdf). Journal of Agricultural Research, 43: 517
Reynolds, M.P., Pellegrineschi, A. & Skovmand, 522.
B. 2005. Sink-limitation to yield and bio- Richards, R.A. 1996a. Defining selection cri-
mass: a summary of some investigations in teria to improve yield under drought. Plant
spring wheat. Annals of Applied Biology, Growth Regulation, 20: 157166.
146: 3949. Richards, RA. 1996b. Increasing yield poten-
Reynolds, M.P., Sayre, K.D. & Rajaram, S. tial in wheat - source and sink limita-
1999. Physiological and genetic chang- tions. In M.P. Reynolds, S. Rajaram & A.
es in irrigated wheat in the post-Green McNab, eds. Increasing Yield Potential in
Revolution period and approaches for Wheat: Breaking the Barriers, pp. 134149.
meeting projected global demand. Crop [Proceedings of a] CIMMYT International
Science, 39: 16111621. Symposium, CIANO, Ciudad, Obregon,
Reynolds, M.P., Singh, R.P., Ibrahim, A., Mexico. Mexico DF, CIMMYT.
Ageeb, O.A., Larqu-Saavedra, A. & Quick, Royo, C., Villegas, D., Garca del Moral, L.F.,
J.S. 1998. Evaluating physiological traits El Hani, S., Aparicio, N., Rharrabti, Y. &
to complement empirical selection for Araus, J.L. 2002. Comparative performance
wheat in warm environments. Euphytica, of carbon isotope discrimination and cano-
100: 8594. py temperature depression as predictors of
Reynolds, M.P., Calderini, D.F., Condon, A.G. genotype differences in durum wheat yield
& Rajaram, S. 2001. Physiological basis of in Spain. Australian Journal of Agricultural
yield gains in wheat associated with the Research, 53: 561569.
LR19 translocation from A. elongatum. Royo, C., Aparicio, N., Villegas, D., Casadess,
Euphytica, 119: 137141. J., Monneveux, P. & Araus, J.L. 2003.
Reynolds, M, Condon, A.G., Rebetzke, G.J. & Usefulness of spectral reflectance indices
Richards, R.A. 2004. Evidence for excess as durum wheat yield predictors under
photosynthetic capacity and sink-limitation contrasting Mediterranean conditions.
476 Plant breeding and farmer participation
International Journal of Remote Sensing, Siddique, K.H.M., Kirby, E.J. M. & Perry,
24: 44034419. M.W. 1989. Ear to stem ratio in old and
Royo, C., Aparicio, N., Blanco, R. & Villegas, modern wheats; relationship with improve-
D. 2004. Leaf and green area develop- ment in number of grains per ear and yield.
ment of durum wheat genotypes grown Field Crops Research, 21: 5978.
under Mediterranean conditions. European Slafer, G.A. 1996. Differences in phasic devel-
Journal of Agronomy, 20: 419430. opment rate amongst wheat cultivars inde-
Royo, C., Garca del Moral, L.F., Slafer, G., pendent of responses to photoperiod and
Nachit, M.N. & Araus, J.L. 2005. Selection vernalization. A viewpoint of the intrinsic
tools for improving yield-associated physi- earliness hypothesis. Journal of Agricultural
ological traits. In C. Royo, M.N. Nachit, N. Science (Cambridge), 126: 403419.
Di Fonzo, J.L. Araus, W.H. Pfeiffer & G.A. Slafer, G.A. 2003. Genetic basis of yield as
Slafer, eds. Durum Wheat Breeding: Current viewed from a crop physiologists perspec-
Approaches and Future Strategies, pp. 563 tive. Annals of Applied Biology, 142: 117
598. New York, USA, Haworth Press. 128.
Savin, R. & Slafer, G.A. 1991. Shading effects Slafer, G.A. & Andrade, F.H. 1993.
on the yield of an Argentinean wheat Physiological attributes related to the gen-
cultivar. Journal of Agricultural Science eration of grain yield in bread wheat culti-
(Cambridge), 116: 17. vars released at different eras. Field Crops
Sayre, K.D. 2004. Raised-bed cultivation. In Research, 31: 351367.
R. Lal, ed. Encyclopaedia of Soil Science. Slafer G.A. & Araus J.L. 1998. Improving wheat
Marcel Dekker, Book Site Online publica- responses to abiotic stresses. In A.E. Slinkard,
tion. ed. Proceedings of the 9th International
Sayre, K.D. & Moreno-Ramos, O.H. 1997. Wheat Genetics Symposium, Vol. 1, pp. 201
Applications of raised-bed planting systems 213. Saskatoon, Canada, University of
to wheat. Wheat Special Report no. 31. Saskatchewan Extension Press.
Mexico D.F., CIMMYT. 36 p. Slafer, G.A. & Araus, J.L. 2007. Physiological
Sayre, K.D., Rajaram, S. & Fischer, R.A. 1997. traits for improving wheat yield under a
Yield potential progress in short bread wide range of conditions. In J.H.J. Spiertz,
wheats in northwest Mexico. Crop Science, P.C. Struik & H.H. van Laar, eds. Scale
37: 3642. and Complexity in Plant Systems Research:
Shearman, V.J., Sylvester-Bradley, R., Scott, Gene-Plant-Crop Relations, pp. 145154.
R.K. & Foulkes, M.J. 2005. Physiological Dordrecht, Netherlands, Springer.
processes associated with wheat yield Slafer, G.A., Araus, J.L. & Richards, R.A. 1999.
progress in the UK. Crop Science, 45: 175 Physiological traits to increase the yield
185. potential of wheat. In E.H. Satorre & G.A.
Shukla, A.K, Ladha, J.K., Singh, V.K., Dwivedi, Slafer, eds. Wheat: Ecology and Physiology
B.S., Balasubramanian, V., Gupta, R.K., of Yield Determination, pp. 379415. New
Sharma, S.K., Singh, Y., Pathak, H., Pandey, York, USA, Food Product Press.
P.S., Padre, A.T. & Yadav R.L. 2004. Slafer, G.A., Calderini, D.F. & Miralles, D.J.
Calibrating the leaf color chart for nitro- 1996. Generation of yield components and
gen management in different genotypes of compensation in wheat: Opportunities
rice and wheat in a systems perspective. for further increasing yield potential. In
Agronomy Journal, 96: 16061621. M.P. Reynolds, S. Rajaram & A. McNab,
Breeding for quantitative variables. Part 5: Breeding for yield potential 477
eds. Increasing Yield Potential in Wheat: physiological traits for genetic improve-
Breaking the Barriers, pp. 101133, ment of cereal yields in Mediterranean
[Proceedings of a] CIMMYT International environments. Annals of Applied Biology,
Symposium, CIANO, Ciudad, Obregon, 146: 6170.
Mexico. CIMMYT, Mexico DF, Mexico. Tambussi, E.A., Casadesus, J., Munn-Bosch,
Slafer, G.A. & Rawson, H.M. 1994. Sensitivity S. & Araus, J.L. 2002. Photoprotection
of wheat phasic development to major in water stressed plants of durum wheat
environmental factors: a re-examination of (Triticum turgidum L. var. durum): chang-
some assumptions made by physiologists es in chlorophyll fluorescence, spectral
and modellers. Australian Journal of Plant signature and photosynthetic pigments.
Physiology, 21: 393426. Functional Plant Biology, 29: 3544.
Slafer, G.A. & Rawson, H.M. 1996. Responses Timsina, J. & Connor, D.J. 2001. Productivity
to photoperiod change with phenophase and management of rice-wheat cropping
and temperature during wheat develop- systems: Issues and challenges. Field Crops
ment. Field Crops Research, 46: 113. Research, 69: 93132.
Slafer, G.A. & Rawson, H.M. 1997. Phyllochron Trethowan, R.M., van Ginkel, M. & Rajaram,
in wheat as affected by photoperiod under S. 2002. Progress in breeding for yield and
two temperature regimes. Australian Journal adaptation in global drought affected envi-
of Plant Physiology, 24: 151158. ronments. Crop Science, 42: 14411446.
Slafer, G.A. & Whitechurch, E.M. 2001. Voltas, J., Romagosa, I., Muoz, P. & Araus,
Manipulating wheat development to improve J.L. 1998. Mineral accumulation, carbon
adaptation and to search for alternative isotope discrimination and indirect selec-
opportunities to increase yield potential. In tion for grain yield in two-rowed barley
M.P. Reynolds, J.L. Ortiz-Monasterio & A. grown under semiarid conditions. European
McNab, eds. Application of Physiology in Journal of Agronomy, 9: 147155.
Wheat Breeding, pp. 160170. Mexico D.F., Villegas, D., Aparicio, N., Nachit, M.M.,
CIMMYT. 240 p. (available at: http://www. Araus, J.L. & Royo, C. 2000. Photosynthetic
cimmyt.org/research/wheat/map/research_ and developmental traits associated with
results/wphysio/WPhysiology.pdf). genotypic differences in durum wheat yield
Slafer, G.A., Calderini, D.F., Miralles, D.J. & across the Mediterranean basin. Australian
Dreccer, M.F. 1994. Pre-anthesis shading Journal of Agricultural Research, 51: 891
effects on the number of grains of three 901.
bread wheat cultivars of different potential Villegas, D., Aparicio, N., Blanco, R. & Royo,
number of grains. Field Crops Research, C. 2001: Biomass accumulation and main
36: 3139. stem elongation of durum wheat grown
Slafer, G.A., Abeledo, L.G., Miralles, D.J., under Mediterranean conditions. Annals of
Gonzalez, F.G. & Whitechurch, E.M. 2001. Botany, 88: 617627.
Photoperiod sensitivity during stem elon- Waddington, S.R., Ranson, J.K., Osmanzai, M.
gation as an avenue to raise potential yield & Saunders, D.A. 1986. Improvement in the
in wheat. Euphytica, 119: 191197. yield potential of bread wheat adapted to
Slafer, G.A., Araus, J.L., Royo, C. & Garca northwest Mexico. Crop Science, 26: 698
Del Moral L.F. 2005. Promising eco- 703.
479
CHAPTER 19
Andrew R. Barr
480 Plant breeding and farmer participation
detection systems applied to identify single r UIF FBTF PG CJPBTTBZ UIJT NBZ SFTUSJDU
nucleotide polymorphisms (SNPs). the size of the population that can be
Marker development and implementa- assayed).
tion can be divided into the following steps: The process used to decide on the best
r JEFOUJGZ QBSFOUT EJGGFSJOH JO UIF USBJU PG approach to marker development is sum-
interest; marized in Figure 19.1. There are five fun-
r EFWFMPQ B QPQVMBUJPO PG QMBOUT TFHSF- damental questions to be asked.
gating for the trait of interest (using 1. Is a marker needed? Marker
doubled haploids, single-seed descent or development is slow and expensive, so
F2 populations); it must be clear that there will be a real
r TDSFFO UIF QPQVMBUJPO GPS UIF USBJUT PG benefit to the breeding programme
interest; through access to a molecular marker.
r DPOTUSVDU MJOLBHF NBQT PG UIF DSPTT
This will involve comparing the cost
using bulked segregant analysis or of marker development and use with
genomic fingerprinting; more traditional screening methods.
r JEFOUJGZNPMFDVMBSNBSLFSTUIBUDPTFH- 2. Is the trait simply inherited or is it
regate with the trait of interest; controlled by many genes? If the
r UFTUUIFBQQMJDBCJMJUZBOESFMJBCJMJUZPGUIF inheritance is simple, then bulked
markers in predicting the trait in related segregant analysis would usually be
families (also referred to as marker vali- the preferred method for marker
dation); identification. However, if inherit-
r QSPEVDF DMFBS BOE TJNQMF QSPUPDPMT GPS ance is controlled by several genes or
assaying the markers; is strongly influenced by the environ-
r NPEJGZ CSFFEJOH TUSBUFHZ UP PQUJNJ[F ment (low heritability), full map con-
use of MAS relative to alternative selec- struction or linkage disequilibrium
tion techniques; and mapping would be the best options.
r JNQMFNFOU JOUP UIF CSFFEJOH QSP- 3. Cost of map construction? The shift
grammes. to automation and high-throughput
marker analysis is dramatically
19.2 DEVELOPING MOLECULAR reducing the cost and time involved
MARKERS in marker development. For many
The technique used to identify a molecular species, map construction will be
marker linked to a trait of interest will sufficiently cheap to replace Bulked
depend upon the type of trait and the Segregant Analysis (BSA).
resources and marker systems available. 4. Is one parent critical or a common
There are several features of the trait that source of trait? The question relates
will be important in devising the most to the diversity of germplasm
efficient marker development strategy. available for screening. If there are
These include: multiple sources of the desirable
r UIF NPEF PG JOIFSJUBODFXIFUIFS TJN- trait, linkage disequilibrium mapping
ple or polygenic; offers the options of localizing loci
r UIFIFSJUBCJMJUZPGUIFUSBJUBOEUIFSFMJ- from multiple sources, while full map
ability of the alternative (bio-assay, bio- construction will only allow the genes
chemical analysis, etc.); and from one source to be localized.
482 Plant breeding and farmer participation
TABLE 19.1
Comparison of major marker systems
Marker Usual loci DNA Approx. time Comments
system per assay amount per assay
5. Is there access to large-scale pheno- provided the basis for most early marker
typing systems? Linkage disequilibrium development work. In this process,
mapping is dependent upon the analysis a population is produced from parents
of a large pool of germplasm and differing in the trait of interest. Molecular
will require substantially more marker markers, able to detect polymorphisms
assays than full map construction. This between the two parents, are screened
is most effectively achieved through against the population so that each line will
automation and high-throughput have been assayed for the target trait or traits
genotyping. and a large number of markers. Linkage
The marker systems that are currently between the markers is assessed based on
available, and their relative advantages the segregation pattern. Several public and
and disadvantages are summarized in commercial software packages are available
Table 19.1. to assist in the map construction. In the first
The data in Table 19.1 represent phase of map construction, the markers
averages for these marker systems. are divided into linkage groups, usually
Although Table 19.1 would suggest that based on paired comparisons. Markers that
random amplified polymorphic DNA fall into a linkage group are then ordered
(RAPD) markers are the most efficient, through two- and three-point analysis.
the unreliability of these markers and the It is often difficult to assign marker-
inability to transfer them between crosses has based linkage groups to known chromo-
greatly limited their use. The full potential somes. For most major crop species, good
of newer systems, such as Diversity Arrays quality linkage maps have been published.
Technology (DarT) (Jaccoud et al., 2001), is Markers from these maps can be used as
as yet untapped. reference points to determine chromosome
Essentially five basic systems have been designations. For some species, such as
used to develop markers linked to traits wheat, chromosome substitution, trans-
of interest. The relative merits of each and location or deletion lines can be used to
their applicability are outlined below. determine the chromosome designation for
specific probes, and hence linkage groups.
19.2.1 Fully mapped approaches With a good quality linkage map,
Complete linkage maps generated from the target trait can be mapped relative
screening the progeny of a cross have to the molecular markers by measuring
Marker-assisted selection in theory and practice 483
FIGURE 19.1
Marker development strategies
Is a marker needed?
Yes No Traditional test
e.g. bio-assay
Genetics of trait?
Simple Complex
Cost of map construction? One parent critical or common source of trait?
High Low
BSA Yes No
Parental choice Access to large-scale phenotyping system?
No Yes
Full map
Population construction
construction via Association mapping
~50, RILs or DHs
population
Population
construction ~200 Choose breeding individuals ~000
RILs or DHs
Genotype and construct
Phenotyping Genotype
map
Formation of +/- bulks Phenotype Phenotype
Association of markers with Retrospective association
QTL mapping
trait mapping
Validation in related
Validation Validation
populations
Implementation Implementation Implementation
the frequency of trait co-segregation with to give reasonable genome coverage (about
markers. 1 every 10 to 20 cM). The work involved
Full map construction is difficult and in screening this number of markers is
labour intensive, particularly in species considerable if markers such as RFLPs
with a large number of linkage groups, are used. For some cultivated species, the
where linkage maps must be constructed for germplasm base may be small and the
each chromosome. Usually ten to twenty level of polymorphism between varieties or
markers are desirable for each chromosome cultivars low. In order to reveal sufficient
484 Plant breeding and farmer participation
FIGURE 19. 2
Bulked segregant analysis in the identification of
AFLP markers linked to a disease resistance locus
Parents, differing in their resistance phenotype, are used to produce a population segregating for resistance. This
population is phenotyped for resistance and two bulks are generated: resistant and susceptible. The parental DNA
and the DNA pooled from the bulks are analysed with a range of AFLP markers to identify bands present in the
resistant parent and the resistant bulk (arrows).
polymorphisms for full genome coverage it outweigh the time and costs associated with
may be necessary to screen several thousand full map construction relative to Bulked
probes. Combining RFLP with various Segregant Analysis (BSA).
other marker systems offers an alternative Linkage maps of most major crop species
approach. AFLP markers are particularly have now been constructed, usually based
useful due to the high multiplexing ratio. around RFLP markers. These established
Even though the level of polymorphism maps are a valuable resource and a good
for AFLP bands is usually lower than for starting point in identifying the most
RFLPs, the large number of bands that appropriate markers to use to obtain good
can be scored in a single assay makes this genome coverage.
a valuable marker system for enhancing
linkage maps. 19.2.2 Bulked Segregant Analysis
It is also often important to consider BSA is a technique that allows rapid, cheap
how many traits of significance may be development of markers for simply inherited
segregating in the population used for traits (Michelmore, Paran and Kesseli, 1991).
marker development. If several important The first step involves pooling individuals
traits are segregating, the benefits derived from the two phenotypic extremes of a
from constructing a full linkage map may segregating F2, doubled haploid or similar
Marker-assisted selection in theory and practice 485
population. DNA isolated from the two screening of the pools may be greater for
pools is then screened with DNA markers, SSR markers and a large number will
usually RAPD or AFLP, and polymorphic be required, but the benefits in having
bands identified. Clear polymorphisms seen markers that are immediately applicable
between the two pools will be derived from in selection programmes may overcome
regions of the genome that are common this drawback. Good collections of SSRs
between the individuals that made up the are available for most major crop species,
pools. The remainder of the genome will be but are limited for minor species.
randomly contributed by the parents and
should show no polymorphisms between 19.2.3 Partial maps
the pools. The principle of this technique is The partial map strategy represents a
illustrated in Figure 19. 2. compromise between the limited resolution
This technique has clear advantages of a BSA and the expensive and, in some
for marker identification relative to cases unnecessarily detailed, fully mapped
construction of complete linkage maps. approach. The main advantage of the partial
Only a few weeks are required to screen map approach is obviously the cost savings
the pools with the AFLP or RAPD primers in comparison with full map construction.
and the cost will be between 20 percent and As the name suggests, partial maps
30 percent of that for full map construction. are constructed of selected regions of
For traits controlled by a single gene, such the genome. A partial mapping approach
as many disease resistances, only five to ten is normally applied when some prior
plants are required for each bulk, meaning knowledge of the genetic control of the
that only small populations are required. trait or gene locations is available, which
However, there are also disadvantages to suggests that a full map is not required for
this technique. development of the markertrait association
r 5IFNFUIPEJTOPUFGGFDUJWF BMUIPVHIOPU but where the genetic control is too complex
impossible) for complex traits controlled for BSA to be successfully used.
by several unlinked loci. The technique Other scenarios where partial maps
can be used for traits controlled by are useful are where BSA has been used
fewer than three major loci. to develop a markertrait association and
r 5IFBDUVBMMPDBUJPOPGUIFHFOFPGJOUFSFTU consensus maps used to obtain a tentative
is not revealed, as only linked markers genomic location for the gene, but where:
are detected. r UIFUSBJUBQQFBSTDMPTFMZMJOLFEUPPUIFS
r 5IFNFUIPEDBOPOMZCFFGGJDJFOUMZVTFE traits of interest, and hence linkage
with PCR-based marker systems such relationships are crucial; or
as SSR, RAPD and AFLP markers. For r UIF NBSLFS EFOTJUZ PO FYJTUJOH NBQT
AFLP and RAPD markers, it is usually is too poor for useful markers to be
necessary to convert the marker to a developed.
sequenced tagged site (STS) marker, and
this is not a simple task. This problem 19.2.4 Pedigree-based analysis
is largely overcome through the use Pedigree-based whole-genome marker
of SSR markers or where very good development (sometimes called analysis
quality maps are available of AFLP by descent, or similar) is a technique for
markers. The effort involved in the initial developing and using marker-assisted
486 Plant breeding and farmer participation
There are four potential advantages of are based on a narrow germplasm pool
this approach in mapping genes in crop and display a low level of polymorphism
species. between cultivars. This has hampered
r *U QSPWJEFT B OFX QFSTQFDUJWF GPS USBJU the identification of molecular markers
mapping, because it uses population linked to agronomically important traits,
structures (based largely around complicating the differentiation of varieties
pedigree) and phenotypic data that and the analysis of genetic variability.
differ from those used for full map
construction of BSA. Consequently, we 19.3 IMPLEMENTING MAS IN
can expect to see new marker trait BREEDING PROGRAMMES
associations and targets for more The identification of markers associated
detailed analysis. with a trait of importance represents a first
r -% NBQQJOH BMTP QSPWJEFT EFUBJMFE step in marker application. Several further
fingerprinting information on a large steps are needed before a marker can be
number of lines and varieties and this used in a practical breeding programme
information will be valuable in several of (outlined in Figure 19.3). The key element
the breeding strategies outlined below. is marker validation.
r 5IF -% NFUIPE VTFT SFBM CSFFEJOH
populations, the material is diverse and 19.3.1 Marker validation
relevant and the most important genes (for Irrespective of the technique used to
adaptation, etc.) should be segregating in identify a marker linked to a target trait, the
such populations. The breeder is also association has been found in a particular
integrally involved in the process and this set of circumstances, usually in the progeny
may lead to improved rate and efficiency of a specific cross (in fully mapped, BSA,
of validation and adoption. Plant pedigree analysis and partial mapped
breeders are often reluctant to grow and methods). The reliability of the marker can
assess a huge number of lines with little be estimated from the closeness of linkage.
or no potential for direct commercial As the next step, the ability of the marker to
outcome, as required for complete map predict the target phenotype must be tested
construction. The advantage of LD in further populations. Usually one would
mapping to the breeder is that mapping aim to keep one parent in common in the
and commercial variety development is first test. Therefore, the predictive value of
conducted simultaneously. the marker in identifying the phenotype of
r 1BUUFSO BOBMZTJT PG NBSLFS EBUB NJHIU the plant can be estimated and compared
detect complex combinations (even with the original mapping result.
epistatic interactions) between alleles The usefulness of the marker in screening
at several loci that underlie the superior and assaying germplasm in use in the
individuals in a breeding population. breeding programme is assessed as the next
This might prove difficult to isolate step. The parents likely to be used in the
and validate using the full mapping marker screening programme are screened
approach. for polymorphisms between those that have
Molecular markers offer an easily and do not have the target trait. Given
quantifiable measure of genetic variation the narrow germplasm base of many crop
within crop species. Many crop species plants, a high proportion of markers will fail
488 Plant breeding and farmer participation
FIGURE 19.3
Development, validation and implementation of markers for MAS schemes
Marker linked at X cM
Is marker able to differentiate between donor and recipient germplasm? Consult database.
Yes No
TABLE 19.2
Hypothetical parent and marker genotypes One extreme consequence is a risk that
demonstrating the importance of marker only parents and populations that suit MAS
polymorphism in MAS schemes (see text for are constructed and selected. Hence, one
explanation)
must ask Is the risk of skewing the breeding
Line Genotype Genotype Genotype at population greater than the potential gains
at Marker at Marker target gene
locus 1 locus 2 locus in efficiency by using MAS?
Variety 1 A b R Computer simulations have shown that
Variety 2 a B r MAS is more effective than phenotypic
Variety 3 A B R selection when population sizes are large
Variety 4 A b r and heritability is low (Whittaker, Haley and
Thompson, 1997; Lande and Thompson,
genotypes increases, so too does the potential 1990). Modelling also indicates that the
problem of lack of polymorphism, and for combination of genotypic selection (i.e. MAS)
many breeding programmes this problem and phenotypic selection can be especially
has severely curtailed the use of MAS by: powerful. Modification to backcrossing,
r SFEVDJOH UIF OVNCFS PG DSPTTFT XIFSF pedigree and progeny schemes for self-
MAS can be applied; and pollinated crops and hybrid breeding to better
r MJNJUJOH UIF DPNQMFYJUZ PG DSPTTJOH exploit the power of MAS are discussed in
schemes where MAS can be applied. Sections 19.4, 19.5 and 19.9, below.
Thus, 4-way crossing plans or conver-
gent schemes are often difficult for MAS 19.3.6 Intellectual property issues
application due to lack of polymorphism The use of some marker systems may be con-
among parents. strained by intellectual property protection.
As mapping, QTL analysis, genomics and Further, research licences may not enable
marker development progress, the number users to apply MAS for selection of commer-
of markers has increased, making this prob- cial varieties without the consummation of a
lem potentially less serious in the future for commercial licence. The simplest advice is to
the major crops. Similarly, diagnostic mark- ensure you have freedom to operate before
ers render this problem redundant. embarking on any MAS project.
follows to select the best line from amongst plant can be determined at very early
the selfed progeny, usually for similarity development stages and therefore plants
to the recurrent parent, expression of carrying the gene or genes of interest
the transferred trait and sometimes for can be identified prior to flowering and
improved performance. This process backcrossed once they begin flowering. Up
is sometimes called defect elimination, to three backcrosses can be made in this way
especially when an otherwise elite cultivar in one season and plants can be grown in
has an obvious flaw, which can be rectified optimum conditions for cross-pollination.
by backcrossing. Some molecular markers, such as RFLPs,
Breeders like this technique because show a co-dominant mode of inheritance
of its relatively predictable outcomes, yet and therefore heterozygous individuals
it is widely viewed as a very conservative can be identified. This is of particular
breeding strategy and hence many benefit when introgressing a recessive
breeding programmes do not allocate gene. Molecular marker-assisted selection
a large percentage of their resources to is based on genotype (not phenotype) and
backcrossing. Further, breeders criticize therefore is not subject to environmental
backcrossing for other reasons: variation and lower assay error. Molecular
r NPSF SBQJE HFOFUJD JNQSPWFNFOU JT markers can be used for selecting regions
possible by conventional pedigree or of the recurrent parent genome unlinked
progeny methods of breeding; to the introgressed region. This reduces the
r UIFSF NBZ CF OP USBJU T
PG TVGGJDJFOU number of backcross generations required
importance to warrant a dedicated to recover the recurrent parent genotype
backcross programme; and increases the probability of obtaining
r BO BCTFODF PG B QIFOPUZQJD TFMFDUJPO a suitable introgression product.
system suited to rapid identification of
single plants during the backcrossing 19.4.1 Transfer of a single dominant
phase; gene
r BO BCTFODF PG FMJUF WBSJFUJFT UIBU This is perhaps the simplest MAS
warrant defect elimination or trait application attempted by most breeders.
enhancement; Here, MAS is used to follow the gene of
r BMPOHEFWFMPQNFOUUJNFPS interest through the backcrossing phase
r UIFSF JT MJUUMF QSPTQFDU PG TJNVMUBOFPVT (Figure 19.4) and may directly substitute
improvement in more than a single trait. for a bio-assay (such as a disease resistance
However, molecular markers can benefit test) or a chemical assay (such as an
a backcross breeding strategy in many ways isozyme test). A co-dominant marker
and turn it from a conservative strategy is used to distinguish the heterozygous
improving only one trait at a time, to a more carriers (Aa genotype) from the non-
aggressive strategy where many traits are carriers (aa genotype) during backcrossing
simultaneously improved while retaining generations, to ensure that the gene of
favoured linkage blocks. The benefits of interest is not lost before the fixation
MAS in backcrossing will be examined, first (inbreeding) stage.
in general (the following paragraph) and in The beauty of the marker test here
detail in the seven sections that follow. is that the plants can be sampled at the
The marker genotype of an individual 1- to 2-leaf stage and a result is available
Marker-assisted selection in theory and practice 493
FIGURE 19.4
Application of MAS to transfer a single dominant gene (Genetic structure
of each generation shown in brackets; MAS action shown in text box)
BC 3 F 2
within 5 days, ensuring that the carriers 19.4.2 Transfer of a single recessive
are quickly identified, perhaps re-potted, gene
fertilized and transferred to ideal When backcrossing a recessive trait using a
conditions for crossing when flowering phenotypic screening system, a generation
occurs. In contrast, some bio-assays (such of selfing is required in between each back-
as nematode resistance) may be so slow cross to expose the carriers of the recessive
that the plants will have already flowered gene of interest; in contrast, where a co-
and crossing must already have taken place dominant marker is available to distinguish
before the result is available, meaning that the Aa genotype from the AA genotype
crosses made to non-carrier plants are (Figure 19.5), backcrossing can proceed at
discarded and represent a waste of effort. the same speed as for a dominant gene.
Beckmann and Soller (1986) showed Hence, it is possible to save the selfing
the frequency of the favourable allele was step required between cycles of phenotypic
substantially increased where MAS was selection in a conventional backcrossing
applied during backcrossing. Care should programme.
be taken to ensure that the number of
individuals selected for each backcross 19.4.3 Selection of several genes
generation allows for recombination simultaneously
between the marker and the gene of One of the most exciting applications of
interest, so that false positives do not MAS in backcrossing is the potential to
represent a risk to success. transfer multiple traits. It is difficult to
494 Plant breeding and farmer participation
FIGURE 19.5
Application of MAS to transfer a single recessive gene (Genetic structure
of each generation shown in brackets; MAS action shown in text box)
BC 3 F 2
conceive and manage practical backcross- although this difficulty could be reduced
ing strategies using phenotypic selection to by fixing genes in each backcross
simultaneously transfer two traits, let alone stream, by selfing or doubled-haploid
more. In contrast, the application of MAS production, prior to intercrossing lines
is only limited by handling an appropriate from each stream; and
BCnF1 population to ensure recovery of r NBOBHJOHQPQVMBUJPOTPGBOBQQSPQSJBUF
the required number of individuals het- size.
erozygous for all target loci. Hence, quite
complex and ambitious defect elimination 19.4.4 Transfer of QTL
strategies can be developed and pursued. Backcrossing has traditionally not been
This may involve introgression of several the method of choice for the introgression
traits from a single donor parent or simul- of quantitative characters into breeding
taneous defect elimination streams with dif- populations. It is often assumed that
ferent donor parents (but the same recurrent many genes with small additive effects
parent), which are subsequently merged to condition quantitative traits and therefore
achieve the transfer of all the targeted traits. the probability of recovering a satisfactory
Conceptually, this approach is quite number of these genes through backcrossing
powerful, but in practice the difficulties is very small. While there have been a
encountered with this approach have been: number of reports of success, most attempts
r BWBJMBCJMJUZ PG QPMZNPSQIJD NBSLFST BT failed to achieve full expression of the
the number of donor parents increases, donor parent trait. Success appears to be
Marker-assisted selection in theory and practice 495
largely dependent on the number of genes of interest is finely mapped. Second, unlike
controlling the trait, the heritability of the a single major gene such as resistance to
trait, the ease and timing (generation) of a nematode, where phenotypic selection
selection, and the desired level of recovery during backcrossing is an alternative
of the recurrent parent phenotype. to MAS, the only way of selecting, for
QTL marker-assisted selection has the example, a complex grain quality trait is
potential to greatly enhance the efficiency with genotypic selection using markers. It
of quantitative trait introgression, particu- is not possible to test many aspects of grain
larly those derived from exotic germplasm. quality on single plants grown in controlled
In conventional selection for a trait of conditions. Another issue important to QTL
low heritability (most quantitative traits), transfer is accuracy of selection, and recent
an individuals phenotype is more greatly studies show MAS significantly improved
influenced by non-genetic factors, such the accuracy of selection for traits of low
as environment. For a mapped QTL, the heritability. Finally, the most important
phenotypic effect is estimated from the data issue becomes validation of the QTL. This
on many individuals; thus, the influence of process involves testing the phenotypic
non-genetic factors should be reduced sub- effect of the particular chromosomal region
stantially (Paterson et al., 1991). Therefore derived from a mapping population initially
genotypic selection for quantitative traits in a range of related germplasm, and then in
should be more efficient than phenotypic progressively less-related germplasm.
selection so long as a large proportion
of the additive genetic variance is associ- 19.4.5 Selection of donor parent
ated with the marker loci. In a backcross- Many authors have now published
ing programme, the gains in efficiency are phylograms for varieties and genotypes in
not solely restricted to reduction in selec- germplasm pools for major crops in many
tion error, but relate also to logistics and different agro-ecological zones. A similarity
total turnover time. Backcrossing is often index based on marker data can then be
the most appropriate breeding strategy for used to select from among a number of
quantitative traits controlled by a relatively possible donors for a desired trait, where
small number of loci. As the number of the ideal is the minimum genetic distance
loci increases, the number of backcross from the proposed recurrent parent. This
individuals that must be grown to have a will theoretically reduce the number of
high probability of recovering each marker backcrosses required to recover the
allele at all loci increases to a point where recurrent parent phenotype.
backcrossing becomes inefficient.
In its simplest form, there is little 19.4.6 Recovery of recurrent parent
difference between the backcross strategy genotype
required to transfer a QTL and that for In a backcrossing programme, the donor
a major gene. Nevertheless, there are parent genome proportion decreases with
important differences. First, there is an increasing number of backcrosses. As
ambiguity about the location of the QTL, the proportion of donor parent genome
often making the use of flanking markers decreases, so also should the contribution of
necessary. This usually increases the size of the donor parent to phenotypic expression.
the introgressed segment, unless the region Therefore, if the proportion of donor
496 Plant breeding and farmer participation
parent genome varies amongst individuals The minimization of the length of the
in the same backcross generation, then intact chromosomal segment of donor
selection for those individuals will greatly type carried (or dragged) along with
improve the efficiency of the backcrossing the target gene (linkage drag) can be
programme. achieved by selecting for individuals that
In some species, the range in the are heterozygous at the target locus and
percentage of donor parent marker alleles in homozygous for recurrent parent alleles at
BC1 progenies was 860 percent, compared two markers flanking the target locus. The
with the expectation of 25 percent. It estimation of the population size required
would therefore be possible to save 1 or 2 to achieve this for various flanking marker
backcrosses by using MAS to select for a scenarios has been made in a computer
high frequency of recurrent parent alleles, model called Popmin (Decoux and Hospital,
speeding the recovery of the background 2002), further enhancing the sophistication
genotype. and precision of the backcross strategy.
Hospital and Charcosse (1992) and
Openshaw, Jarboe and Bevis (1994) used 19.4.7 Accuracy of MAS during
simulations to compare the recovery of backcrossing
recurrent parent genotype following various The next general consideration is the
backcrossing schemes in maize. These studies accuracy of MAS during backcrossing,
have further refined the strategies required which will be determined by:
to optimize backcrossing. Stam and Zeven r UIFHFOFUJDEJTUBODFCFUXFFOUIFNBSLFS
(1981) estimated that the typical segment locus and the functional gene; and
length of donor parent DNA retained after r UIFOVNCFSPGNBSLFSTBWBJMBCMFOFBSUIF
three backcrosses was 51 cM in a 100 cM gene of interest.
chromosome. Frisch, Bohn and Melchinger For each backcross progeny selected,
(1999) showed that heavy selection in the probability of losing the target allele by
the BC1 in a simulated maize system was recombination when selection is performed
essential to reduce the size of the donor on a linked marker locus is simply equal to
segment, along with tightly linked flanking r, the recombination frequency between the
markers. Frisch, Bohn and Melchinger and marker and target loci. If this is continued
Hospital and Charcosse both recommend for t generations of backcrossing, the
two to three markers per 100 cM, distributed probability of losing the target allele by
across the remainder of the genome, for recombination is 1 -(1- r)t. Holland (2004)
rapid recovery of the donor parent. Frisch, provides the following examples:
Bohn and Melchinger (1999) are confident If the marker locus exhibits 10 percent
that reductions of 2 to 4 backcrosses in a recombination with the target gene,
6-backcross strategy are possible. there is a 10 percent chance of losing the
AFLPs are well suited to this purpose, target allele each generation, and a 27
subject to an appropriate technology-use percent chance of losing the target allele
licence. SSR-based protocols are also ideal after three generations of backcrossing.
and, depending on source, may be less Tightly linked markers, of course,
constrained by intellectual property restric- do much better: three generations of
tions. DarT technology may also have backcrossing and selection on a marker
application in this area. locus with 1 percent recombination
Marker-assisted selection in theory and practice 497
with the target gene has only a 3 percent where a great deal of information about
chance of losing the target allele. the genetic control of characters is known,
If tightly linked markers are not available it may be more common to transfer 1 to
or you wish to make certain that the gene of 5 chromosomal segments associated with
interest in included within the introgressed the expression of a semi-quantitative trait
segment, flanking markers can be used. than may be the case in other crops (Stuber,
Again, Holland (2004) provides a worked Polacco and Senior, 1999).
example:
If marker loci A and B flank the 19.4.9 Summary of backcross breeding
target locus, one would select backcross using MAS
progeny that have both A and B The advent of molecular marker technology
alleles from the donor parent. The has dramatically increased the potential
probability of losing the target allele efficiency, precision and flexibility of
with flanking marker selection is equal backcross breeding. However, these gains are
to the probability of selecting a double yet to be realized by most practical breeding
recombinant progeny from among programmes. The first commercial varieties
the doubly heterozygous backcross selected from backcrossing enhanced by MAS
progeny. If the flanking loci have have now been commercialized, and more will
recombination frequencies rA and rB, follow shortly. The challenge is now for new
respectively, with the target locus, the crops and breeders to realize the potential of
probability of losing the target allele the increased power of backcrossing.
due to double crossovers within the
selected region (ignoring crossover 19.5 USE OF MAS IN PEDIGREE OR
interference) is: rA. rB/(1- rA - rB+2 rA rB). PROGENY BREEDING
This probability can be much lower In many breeding programmes, backcrossing
than the probability of losing the accounts for only a small percentage of
target allele based on selection for a the total germplasm. While improvements
single marker locus. For example, if the in backcross efficiency are important, the
flanking markers each have 10 percent biggest potential impact of MAS in most
recombination frequency with the programmes for self pollinated plants lies
target locus, there is only a 1.2 percent in pedigree or progeny breeding systems,
chance of losing the target allele after based on so-called forward crossing
a single generation. In any case, with strategies (Holland, 2004).
tighter linkage, the chance of losing For perennials, and especially long-lived
the target allele and the amount of perennials, which take years to reach their
linkage drag are reduced. reproductive stage, markers are especially
valuable. For instance, if marker+trait
19.4.8 Backcrossing in cross-pollinated associations for fruit quality in stone fruits
species or pulp yield in forest trees were available,
Many of the opportunities and roles of the impact of markers would be especially
MAS in backcrossing in cross-pollinated high, as only those seedlings with some
and hybrid crops are similar to those chance of success need to be progressed
covered earlier in Section 19.4. One subtle to field trials, greatly increasing the cost-
difference is that for crops such as maize, effectiveness of the breeding work.
498 Plant breeding and farmer participation
In the following sections, the important national parks has been coined to describe
issues and opportunities in the application these linkage blocks. Few breeders have
of MAS in pedigree and progeny breeding attempted to characterize linkage blocks
schemes are discussed. in the breeding material and most do not
have the tools to exploit this information.
19.5.1 Characterization of the However, such linkage blocks have now
germplasm pool been characterized using molecular markers
The first attempt of many breeding in many crops, including barley and maize.
programmes to utilize molecular markers The usual methodology is to establish a
in their wider breeding programme is often genotypic database of genotype marker
the characterization of the genetic diversity loci. This information can be used to
of their germplasm pool (e.g. Melchinger et develop graphical genotypes, a technique
al., 1994). Markers are chosen that have wide where genotypes of lines are visualized in
genomic coverage (even better if hotspots a range of software applications, which
of specific interest within the genome are make it easier to interpret complex datasets.
known), which are then assayed over key These software applications can be used to
representatives of the germplasm base of the select individuals that carry alleles closest
programme, including significant ancestral to an ideal combination, as defined by the
material, successful parents, representative linkage block analysis. Currently, the most
genotypes from other major improvement advanced software for graphic genotypes
programmes in the world, and current elite resides in the private sector, although several
varieties. programmes are currently available in the
This information can be used to: public domain, e.g. GGT, Hypergene and
r DIPPTFEPOPSQBSFOUTDMPTFMZSFMBUFEUP Geneflow.
recurrent parents for backcrossing, as
described above; 19.5.3 Key recombination events
r DIPPTF QBSFOUT XJUI XJEF HFOFUJD One of the greatest positives to come from
diversity to maximize opportunities for marker development programmes has been
transgressive segregation for traits of the increased knowledge of the genome
interest or to define possible heterotic and the physical and genetic control of
groups; or important traits. After the first decade of
r NBQUSBJUTCBTFEPOQFEJHSFFBOENBSLFS research in this area, breeders can feel much
similarity, as demonstrated by Paull et al. more confident of designing an appropriate
(1998) for Sr 22 in wheat. Forster et al. strategy to transfer a trait from one parent
(1997) demonstrated AFLP associations to another, or indeed improve a quantitative
with salt tolerance within Hordeum trait. Hence, breeding strategies can be
spontaneum germplasm. modified to:
r SFDPHOJ[FDPVQMJOHBOESFQVMTJPOMJOLBHF
19.5.2 Linkage block analysis and between key traits;
selection r FOTVSF UIBU B CSFFEJOH QPQVMBUJPO JT PG
Breeders have long suspected that certain sufficient size to capture the key QTLs
chromosomal regions carry critical clusters contributing to a quantitative trait; and
of genes (linkage blocks) that have been r FOTVSFLFZNBKPSHFOFTGPSNUIFCBTFPG
highly conserved during selection. The term each important segregating population
Marker-assisted selection in theory and practice 499
Building Make cross between different resistant parents Make cross between different resistant parents
pyramids
Select for most highly resistant phenotype Marker-assisted selection for different
contributing loci
Use ultra virulent races, when available
Applicable to major and minor genes
Very difficult for genes of small effect
Genotype of final lines firmly established, except
Genotype of final line may not be clear for possible recombination between marker use
and gene (use flanking markers)
Dissecting Retrospective analysis of developed lines required No need, as genotype known from marker assay
pyramids to prove all important alleles are included
Resistant Susceptible cross made, F1 to test
dominance relationships, F2 analysis for number
of genes
Progeny test and/or Test-cross to confirm
500 Plant breeding and farmer participation
19.5.7 Early generation selection with a thin husk. Intuitively this was reason-
The selection theory required to implement able, and subsequent genetic and biochemi-
MAS in early generations is similar to other cal analysis showed this to be true. Other
forms of selection, although MAS is closer QTLs for malt extract were co-incident with
to simultaneous rather than tandem (or QTLs controlling levels of starch-degrading
sequential) selection, which is often a feature enzymesagain, this association was subse-
of early-generation phenotypic selection. quently proven to be causal. While trait dis-
In general, breeders visually select traits section may not be as powerful an approach
of high heritability in early generations to the resolution of complex characters as
because it is not possible to effectively expression profiling or functional genomics,
select for yield (and some other complex it is certainly proving a useful, practical tool
traits) in rows or small plots. Computer in many breeding programmes.
simulations have shown that MAS is more
effective than phenotypic selection when 19.5.9 Summary of use of markers
population sizes are large and heritability in backcrossing and pedigree and
is low (Whittaker, Haley and Thompson, progeny breeding systems
1997; Lande and Thompson, 1990). This In the previous two sections, the
challenges breeders and geneticists to design applications, opportunities and limitations
and implement practical MAS strategies to of MAS in backcrossing and pedigree and
effectively select complex traits in early progeny breeding systems were discussed.
generations. In Table 19.4, these roles are summarized
MAS of early-generation fixed lines is for a typical breeding programme for a self-
therefore the ultimate goal of the many pollinated crop.
programmes. However, at present, this is
fantasy for public-sector breeders, as a 19.6 USE OF MAS IN THE EXPLOITATION
number of limitations dictate that MAS can OF PRIMITIVE GERMPLASM
only be used in early-generation screening Genetic diversity in modern cultivars of
for very important material, comprising some crop species is diminishing due to
a small fraction of the total programme. the high selection pressure for specific
The key limitations are the cost of DNA quality and performance characteristics.
extraction, availability of suitable markers, This may ultimately lead to a reduced
staff resources for sample and data handling, capacity for breeders to respond to new
and the costs (fixed and recurrent) of high- disease pressures and may restrict breeders
throughput systems. from making major improvements in yield
and quality parameters. Introgression of
19.5.8 Trait dissection germplasm from outside the current gene
The use of QTL analysis to dissect traits is pool is essential. Such introgressions have
proving a powerful tool. Trait dissection relies typically sought to introduce major genes,
on the co-incidence of QTLs derived for dif- such as those for disease resistance, via
ferent characters, which is then used to infer backcrossingthis type of project has been
associations. For instance, in barley, the QTL a long-standing and successful part of many
for a thin husk is co-incident with QTL for breeding programmes, and it is easy to
malt extract, inferring that malt extract (the envisage a role for MAS along the lines
economically important trait) is associated discussed earlier.
Marker-assisted selection in theory and practice 501
TABLE 19.4
Summary of roles for MAS in breeding programmes in self-pollinated crops
Phase Pedigree and Progeny Backcrossing Marker roles
However, MAS may be able to do much with relatively little effort currently being
more. The QTLs for agronomic and quality invested in the identification of QTLs
traits in many programmes are largely from outside this elite (and narrow) gene
from elite breeding lines and cultivars, pool. Experience from some crops (e.g.
502 Plant breeding and farmer participation
papers on wheat, barley and rice just being 19.7.1 Enrichment of germplasm pools
published. It is likely to be more difficult in The success of the participatory
those crops with more complex genomes. approach will depend on the frequency
However, this methodology may facilitate of desirable alleles in the segregating
the exploitation of relatively underutilized populations supplied to the participating
genetic variation in wild ancestors and farmers. Hence, there is a role for MAS
related landraces of many crop species. applied to pre-select TC1F1 or BC1F1
individuals carrying desired alleles, or,
19.7 USE OF MAS IN PARTICIPATORY if greater marker throughput is possible,
PLANT BREEDING STRATEGIES individuals could be selected from
Participatory plant breeding has segregating populations for distribution
been proposed as a useful method of to the participating farmers.
simultaneously achieving genetic progress
through plant breeding, as well as facilitating 19.7.2 Pre-breeding for the
seed production and variety adoption. The participating farmers
technique has special application in resource- If the participating farmers in the breeding
poor farming communities. The basic programme have a set of minimum criteria
principles of a participatory scheme are that: (such as disease resistance genes, plant
r UIF CSFFEFS HFOFSBUFT B SFMFWBOU QPPM development or stature genes, quality
of genetic variability for the target traits, etc.), which they would like in a
environment; range of adapted backgrounds from an
r UIF CSFFEFS NVMUJQMJFT TFFE PG UIFTF earlier cycle of the participatory breeding
segregating populations for distribution process, these could be provided to the
to local participators (usually farmers or breeder for enrichment before return to
locally-based agronomists); the participating farmers.
r UIF CSFFEFS GBDJMJUBUFT USJBMT PG UIFTF
populations; 19.7.3 Seed production
r UIFQBSUJDJQBUPSTJEFOUJGZUIFTJOHMFQMBOUT A successful participatory programme will
or lines they feel are relevant to their produce many lines for commercialization
agro-ecological zone and end uses; for different, small, regional areas. The
r TFFE JT NVMUJQMJFE CZ UIF CSFFEFS
XIP task of seed production and maintenance
then facilitates merit trials at local level; can therefore be complex. Marker
r UIF QBSUJDJQBUJOH GBSNFST BOE CSFFEFS fingerprinting of lines could be used
jointly select elite lines, which are further to ensure the integrity of seed stocks,
multiplied for regional testing and seed both prior to release to the participating
production; and farmers and to maintain pure seed in the
r UIF QBSUJDJQBUJOH GBSNFST TQSFBE TFFE longer term.
of new elite material for commercial
production. 19.8 USE OF MAS IN TRANSGENIC
Similar schemes are now widely used BREEDING PROGRAMMES
in the CGIAR network and their full MAS has been extremely important in the
potential has still to be unlocked. MAS selection and monitoring of transgenics
may have some useful roles to play in following the development of transgenic
participatory schemes. plants. The reluctance of the food, feed and
504 Plant breeding and farmer participation
TABLE 19.6
Marker genotype of 22 lines from the cross Gairdner/Keel//Gairdner assessed at 47 marker loci
Keel alleles are shown in blue (with designation of B), Gairdner alleles are shown in yellow (A), heterozygotes in green
(A/B) and no result is shown in white (n/r). The primers are shown in column 2 together with their position on a
consensus map in column 3. In the column headed Keel alleles are genomic regions where MAS was applied for Keel traits
and in the column headed Gairdner alleles are genomic regions where MAS applied for Gairdner alleles. The resistance to
CCN is shown on the bottom row (as determined by a bio-assay) and can be compared to marker alleles for locus BMAC222
on chromosome 5H.
use of molecular markers for the most will also be used to develop doubled
rapid and efficient development of new haploid populations by the end of 2003.
lines. In the first year of the project, this
approach has successfully developed cv. 19.9.5 Simple graphical genotypes to
Keel backcross lines (BC1 and BC2) that demonstrate ancestry
carry the key quality genes from the Example 5: Use of graphical genotypes
three malting barley parents, including presented in a spreadsheet to show the
malt extract, on chromosomes 1H, 2HS, contribution of the landrace barley CI
2HL and 5H, and the -amylase gene on 3576 to modern Australian varieties.
4H, which influences diastatic power and Atmojdo (2002), in an unpublished
fermentability. These lines were to enter Masters thesis of the University of
field trials in 2002 for selection based on Adelaide, developed simple graphical
agronomic performance, prior to yield genotypes of a number of Australian barley
trials and malt quality evaluation in 2003. varieties of descent from an important
The key malting quality genes from cvs. land race known as CI 3576. The genetic
Alexis, Haruna nijo and AC Metcalfe are composition (Figure 19.6) of the cultivars
also being combined by merging the three implied that something important from
cv. Keel backcrossing programmes. BC3 CI 3576 in the region around Bmac093 is
generations will also be developed for each retained during breeding and selection in
of the Keel backcrossing programmes. Australia. Subsequently, mapping studies
Molecular markers will be used to identify revealed that a key development locus
elite individuals from these intercrosses with an allele conferring early flowering
and the BC3F1 generations. The elite lines was contributed from CI 3576.
FIGURE 19.6
Graphical genotype of chromosome 2H of barley for 9 Australian barley cultivars
derived from the landrace CI 3576, showing the retention of key linkage blocks
associated with adaptation and disease resistance
Schooner
O'Connor
Mundah
Arapiles
Galleon
Chebec
CI3576
Barque
Picola
Sloop
Probe
BCD175 A A A? B A? A B A? A? B
WG0516 B B B C D B B B B C
HVM36 115 115 115 110 110 115 115 115 115 115
ABG453 A A A B? A? A A A B B
ABC454 A A A B A A A A B B
EBmac607 145 145 145 145 141 141 141 141 145 145
Bmac093 152 152 152 152 152 ***** 152 152 152 152 152
EBmac684 180 183 180 180 175 180 180 180 180 180
Bmac132 182 182 182 190 190 182 182 182 190 190
EBmac715 185 194 185 190 190 192 192 185 188 188
WG0996 B B B? C F B B B? C? C
ABC309 A A A B D A A A B B
Bmag378 133 133 133 136 136 133 133 133 133 133
ABG014 B E B C F B B B B B
CDO0366 B B B A A B B B B B
PSR901 C B A A A C C A A A
AWBMA21 C D A E E C C G A A
EBmatc39 - 140 125 125 125 157 157 157 125 125
EBmac415 226 226 244 244 244 226 226 226 234 234
HVM54 142 142? 160 160 160 142 142 142 146 146
XKSUF41 A? C? B B B B C B? B? B
BCD292 E C A A A B B D A A
CDO0036 B? B B B B ***** B B B B B
WG0645 C C B B B D D A B B
Cultivars are shown in columns and marker loci in rows, with alleles for various SSR and RFLP loci shown both by colour
(CI 3576 alleles in red) and fragment size and allelic designation. (Atmojdo, unpublished M.Ag.Sc. Thesis).
510 Plant breeding and farmer participation
FIGURE 19.7
Graphical genotypes prepared in Geneflow of Barque, Galleon, Triumph
and CI 3576 comparing alleles at all available SSR loci
those parents and traits for which vali- programme may lead to a focus on breeding
dated, polymorphic markers are available; strategies based on the technology rather
r TUBGG SFTPVSDFT GPS IBOEMJOH UIF QMBOUT
than on the most appropriate strategy for
DNA and marker allele data. Many a particular environment. For example,
breeding programmes have funded their backcrossing is highly amenable to MAS,
initial work in MAS from special funds. but is a conservative breeding strategy and
The time is fast approaching for public- should not become the prime focus of a
sector breeders, and did several years breeding programme. Again, this issue
ago for private breeders, where resource can be addressed if the breeders are aware
re-allocation within the breeding team is of the potential problem and include a
required to further implement MAS; diversity of strategies in their programme.
r OFX
SPCPUJD TZTUFNT NBZ CF DBQBCMF PG Another limitation is the power of
analysing 104 to 105 loci per day, and QTL discrimination (Melchinger et al.,
will challenge the information processing 1998). Many QTLs indicated in mapping
capability of all but the largest breeding population studies disappear in validation
programmes; and populations. These QTLs may have
r UIF DPTU PG SPCPUJDT
5BR QPMZNFSBTF
been cross specific, subject to genotype
non-gel-based discrimination systems, environment interaction effects or illusory.
etc., for true, high-throughput systems. Illusory QTLs may have been artefacts
While such systems show great potential of small mapping populations, error in
for handling targeted numbers, they the phenotyping experiments or reflect
appear at present to be beyond the budget fundamental limitations of QTL analysis
of most barley breeders. However, SNP methods.
markers may be analysed on chips or
use mass spectrophotometric methods, 19.10.5 Application of MAS to yield
which could greatly reduce costs. improvement
New approaches to the improvement of
19.10.4 Risks and limitations of using MAS grain yield remain the Holy Grail of plant
The risk most frequently noted is the breeding. Yield, according to conventional
temptation to use only parents for which wisdom, is conditioned by many genes
either markers or polymorphic markers of small effect, so how can MAS play an
exist, thus narrowing genetic diversity. In important role in manipulating such a
particular, this may concentrate use of a few, complex trait? Many mapping studies have
well-characterized disease-resistance genes measured grain yield and applied QTL
to the exclusion of less well documented analysis (Teulat et al., 2001; Marquez-
sources. This risk can be minimized by Cedillo et al., 2001; Baum et al., 2001).
breeder discretion allocating a proportion of Australian researchers have also measured
the programme to new or uncharacterized grain yield in 10 barley mapping populations
sources. It might be more useful to think (Higgins et al., 2003). The number of QTLs
of this problem as a challenge: How can for grain yield varied widely, from two up
marker technology be used to expand our to eight. This range will be due to:
useful gene pool? r UIF QPQVMBUJPO TJ[F PG UIF NBQQJOH
There is also concern that a strong population and quality of the maps,
emphasis on markers in a breeding which affects the resolution possible;
Marker-assisted selection in theory and practice 513
Barr, A.R., Jefferies, S.P., Warner, P., Moody, 3.16.5, in Proceedings of the 9th Australian
D.B., Chalmers, K.J. & Langridge, P. 2000. Barley Technical Symposium.
Marker-assisted selection in theory and prac- Decoux, G. & Hospital, F. 2002. Popmin: A
tice. In Proceedings of the 8th International programme for the numerical optimization
Barley Genetics Conference, Adelaide, South of population sizes in marker-assisted back-
Australia, October 2004, pp. 167178. cross programmes. The Journal of Heredity,
Barr, A.R., Eglinton, J.K., Collins, H.M., 93(5): 383384.
Vassos, E. & Roumeliotis, S. 2004. Malting Eisemann, B., Banks, P., Butler, D., Christopher,
quality in barley a 25-piece puzzle. M., DeLAcy, I., Jordan, D., Mace, E.,
pp. 5871, in J. Spunar and J. Janikopva MacGowan, P., McIntyre, L., Poulsen, D.,
(eds). Oral Presentations. Proceedings Rodgers, D. & Shepherd, J. 2004. Pedigree-
of the 9th International Barley Genetics based genome mapping for marker-assisted
Symposium, Brno, Czech Republic, 2026 selection and recurrent parent recovery in
June 2004. Available at: http://www.ibgs. wheat and barley. Poster 3.4.1 In New direc-
cz/photos/Text/Oral_Presentations.pdf. tions for a diverse planet. Proceedings of
Baum, M., Grando, S., Backes, G., Jahoor, A., the 4th International Crop Science Congress,
Sabbagh, A. & Ceccarelli, S. 2003. QTLs Queensland, Australia, 26 September1
for agronomic traits in the Mediterranean October 2004 (available at http://www.crop-
environment identified in recombinant science.org.au/icsc2004/poster/3/4/1/1422_
inbred lines of the cross Arta H. spon- eisemannb.htm; accessed 16 Jan. 2009).
taneum 41-1. Theoretical and Applied Frisch, M., Bohn, M. & Melchinger, A.E.
Genetics, 107: 12151225. 1999. Comparison of selection strategies
Beckmann, J.S. & Soller, M. 1986. Restriction for marker-assisted backcrossing of a gene.
of fragment length polymorphism in plant Crop Science, 39: 12951301.
genetic improvement. Plant Molecular Cell Forster, B.P., Russell, J.R., Ellis, R.P., Handley,
Biology, 3: 197250. L.L., Robinson, D., Hackett, C.A., Nevo,
Bernardo, R. 1999. A model for marker- E., Waugh, R., Gordon, D.C., Keith, R. &
assisted selection among single crosses with Powell, W. 1997. Locating genotypes and
multiple genetic markers. Theoretical and genes for abiotic stress tolerance in barley
Applied Genetics, 97: 473478. a strategy using maps, markers and the wild
Bernardo, R. 2001. What if we knew all the species. New Phytologist, 137(1): 141147.
genes for a quantitative trait in hybrid Han, F., Romagosa, I., Ullrich, S.E., Jones,
crops? Crop Science, 41: 14. B.L., Hayes, P.M. & Wesenberg, D.M.
Collins, H.M, Logue, S.J., Jefferies, S.P., Stuart, 1997. Molecular marker-assisted selection
I.M. & Barr, A.R. 1999. A Study of the for malting quality and traits in barley.
physical, biochemical and genetic factors Molecular Breeding, 3: 427437.
influencing malt extract. pp. 2.44.12.44.6, Hayes, P.M., Liu, B.H., Knapp, S.J., Chen,
in Proceedings of the 9th Australian Barley F., Jones, B., Blake, T., Frankowiak, J.,
Technical Symposium. Rasmusson, D., Sorells, M., Ullrich, S.E.,
Coventry, S., Barr, A.R., Jefferies, S.P., Wesenberg, D. & Kleinhofs, A. 1993.
Chalmers, K., Logue, S.J. & Langridge, P. Quantitative trait locus effects and envi-
1999. Validating Quantitative Trait Loci for ronmental interaction in a sample of North
diastatic power in populations of malting American barley germplasm. Theoretical
barley (Hordeum vulgare L.). pp. 3.16.1 and Applied Genetics, 87: 392401.
Marker-assisted selection in theory and practice 515
Hayes, P.M., Toojinda, T. & Vivar, H. 1999. 1999. Pedigree analysis provides a useful
Mapping and using Stripe Rust resistance adjunct to QTL analysis in sorghum. In
genes in a malting quality background. Proceedings of the 11th Australian Plant
pp. 6466, in Proceedings of the Australian Breeding Conference, Adelaide, Australia,
Plant Breeding Conference, Adelaide, 1923 April, pp. 222223.
Australia. Kandemir, N., Jones, B.L., Wesenberg, D.M.,
Higgins, T.J., G.R. Steed, J.C. Fegent & S. Ullrich, S.E. & Kleinhofs, A. 2000. Marker-
Banerjee. 2003. Better barley faster: the role of assisted analysis of three grain yield QTL
marker-assisted selection. Australian Journal in barley (Hordeum vulgare L.) using
of Agricultural Research, 54: 10651408. near-isogenic lines. Molecular Breeding,
Holland, J.B. 2004. Implementation of molecu- 6(2): 157167.
lar markers for quantitative traits in breeding Lande, R. & Thompson, R. 1990. Efficiency
programmes challenges and opportunities. of marker-assisted selection in the improve-
Paper 3.4, in New directions for a diverse ment of quantitative traits. Genetics,
planet. Proceedings of the 4th International 124: 743756.
Crop Science Congress, Queensland, Langridge, P., Lance, R.C.M. & Barr, A.R.
Australia, 26 September1 October 2004 1996. Practical application of marker-
(available at http://www.cropscience.org. assisted selection. In Proceedings of the 7th
au/icsc2004/symposia/3/4/203_hollandjb. International Barley Genetics Symposium,
htm; accessed 16 Jan. 2009). Saskatoon, Canada, pp. 141149.
Hospital, F. & Charcosse, A. 1992. Marker- Marquez-Cedillo, L.A., Hayes, P.M., Kleinhofs,
assisted introgression of quantitative trait A., Legge, W.G., Rossnagel, B.G., Sato, K.,
loci. Genetics, 147(3): 14691485. Ullrich, S.E. & Wesenberg, D.M. 2001. QTL
Iguarta, E., Edney, M., Rossnagel, B.G., analysis of agronomic traits in barley based
Spaner, D., Legge, W.G., Scoles, G.J., on the doubled haploid progeny of two
Eckstein, P.E., Penner, G.A., Tinker, N.A., elite North American varieties representing
Briggs, K.G., Falk, D.E. & Mather, D.E. different germplasm groups. Theoretical
2000. Marker-assisted selection of QTL and Applied Genetics, 103(4): 625637.
affecting grain and malt quality in two-row Melchinger, A.E., Graner, A., Singh, M. &
barley. Crop Science, 40(5): 14261433. Messmer, M. 1994. Relationships among
Jaccoud, D., Peng, K., Feinstein, D. & Kilian, European barley germplasm: I. Genetic
A. 2001. Diversity arrays: A solid state diversity among winter and spring cul-
for sequence information independent tivars revealed by RFLPs. Crop Science,
genotyping. Nucleic Acids Research, 29(4): 34: 11911199.
Article e25, 7 p. Melchinger, A.E., Utz, H.F. & Schon, C.C.
Jefferies, S.P., Barr, A.R., Karakousis, A., 1998. Quantitative trait locus (QTL) map-
Kretschmer, J.M., Manning, S., Chalmers, ping using different testers and independ-
K.J., Nelson, J.C., Islam, A.K.M.R. & ent population samples in maize reveals
Langridge, P. 1999. Mapping of chromosome low power of QTL detection and large
regions conferring boron toxicity tolerance bias in estimates of QTL effects. Genetics,
in barley (Hordeum vulgare). Theoretical 149: 383403.
and Applied Genetics, 98: 12931303. Michelmore, R.W, Paran, I. & Kesseli, R.V.
Jordan, D.R., Henzell, R.G., Tao, Y.Z., 1991. Identification of markers linked to
Godwin, I.D., Cooper, M. &McIntyre, C.L. disease resistance genes by bulked seg-
516 Plant breeding and farmer participation
with adaptation to Mediterranean dry- Young, N.D. & Tanksley, S.D. 1989. RFLP
land conditions in barley. Theoretical and analysis of the size of chromosomal segments
Applied Genetics, 117: 653669. retained around the tm-2 locus of tomato
Whittaker, J.C., Haley, C.S. & Thompson, during backcross breeding. Theoretical and
R. 1997. Optimal weighting of informa- Applied Genetics, 77: 353359.
tion on marker-assisted selection Genetical Zhang, W. & Smith, C. 1992. Computer simu-
Research, 69: 137144. lation of marker-assisted selection utiliz-
Young, D. 1999. A cautiously optimistic vision ing linkage disequilibrium. Theoretical and
for marker-assisted breeding. Molecular Applied Genetics, 83 813820.
Breeding, 5: 505510.
519
CHAPTER 20
Paolo Annicchiarico
520 Plant breeding and farmer participation
(e.g. pure lines) are less buffered against yield stability concept only in relation to
environmental variation than other types, genotype responses over time, using the
such as open-pollinated populations or adaptation concept in relation to responses
mixtures of pure lines, owing to their lower in space (Barah et al., 1981; Lin and Binns,
richness in adaptive genes (Becker and 1988). This view, accepted here, agrees
Lon, 1988). with the farmers view that location is a
A few widely used terms, such as constantnot variablefactor, and yield
adaptation and yield stability, need be consistency over time is the only relevant
unequivocally defined at this stage. In an component of a genotypes yield stability.
evolutionary biology context, adaptation Breeding for wide adaptation aims
is a process, adaptedness is the level of to develop a variety that performs well
adaptation of plant material to a given in nearly all the target region, whereas
environment, and adaptability is the ability breeding for specific adaptation aims to
to show good adaptedness in a wide range of produce different varieties, each of which
environments. In a plant breeding context, performs well in a definite area (subregion)
the first two terms relate to a condition within the region. Early plant breeders
rather than a process, indicating the ability advocated the usefulness of selection for
of the material to be high-yielding in a given specific adaptation (Engledow, 1925),
environment or given conditions (to which and Falconer (1952) suggested that
it is adapted) (Cooper and Byth, 1996). specific breeding may be preferable for
Genotype adaptation is usually assessed environmentally-contrasting subregions on
on the basis of yield responses (although the basis of selection theory. However,
other variables, e.g. gross benefit, may be breeding programmes have mostly
considered), and undergoes modification considered GE interactions as simply
when better performing material becomes a hindrance to crop improvement, while
available. Breeding for wide adaptation and pursuing, even in less developed countries,
for high yield stability and reliability have a wide-adaptation strategy that tended to
sometimes been considered one and the promote varieties with high yield potential
same, insofar as the latter two terms indicate alongside technical packages designed to
a consistently good yield response across significantly improve the environment
environments. However, only the adaptive (Simmonds, 1979: 356). This trend has been
responses to locations, geographical areas, favoured by the perspectives of rapid yield
farming practices or other factors that gain offered by high input levels, the greater
can be controlled or predicted prior to profitability of targeting seed markets in
sowing can be exploited by selecting and more productive areas, and the belief that
growing specifically-adapted genotypes. selection in favourable areas produces a
For example, the knowledge of specific substantial yield gain also in less favourable
adaptation to past years, as shown by areas. The difficulty in sustaining and
positive genotype year (GY) interaction expanding high-input agricultural systems,
effects, cannot be exploited in future years, and the mounting evidence of the dimension
as the climatic conditions that generate of GE effects demonstrated between
year-to-year environmental variation are favourable and stress-prone environments
not known in advance. Therefore, some (e.g. Ceccarelli, 1989; Bnzinger, Bertrn and
authors have proposed applying the Lafitte, 1997), have led to reconsideration
522 Plant breeding and farmer participation
mean response of the tested genotypes, i.e. with respect to within-location response
zero GE interaction. Lin and Binns (1988) to year mean yield of three hypothetical
proposed a second stability measure that cultivars. Growing cultivar 2 in all years,
refers to consistency of yield across years the most stable cultivar according to the
within location (rather than across generic dynamic concept, minimizes the within-site
environments). This measure relates to the GY interaction effects. Growing the most
static stability concept, but is nearer to stable material based on the static concept,
a farmer's view of stability. The effect of i.e. cultivar 3, minimizes the year-to-year
the stable or dynamic concept of stability yield variation and implies a yield penalty
is displayed in Figure 20.1 (lower part) in favourable, high-yielding years and a
FIGURE 20.1
Exploitation by specific adaptation, and minimization by wide adaptation, of spatial GE
interaction (above); and exploitation by the static yield stability concept, and minimization
by the dynamic yield stability concept, of temporal GE interaction (below)
5 5
Exploitation cv. 1 Minimization cv. 1
by specic adaptation: by wide adaptation:
4 cv. 3 (subregion A) + ++ cv. 2 4 cv. 2 over the region
cv. 2
cv. 1 (subregion B)
cv. 3 cv. 3
Cultivar yield
3 3
Cultivar yield
2 2
+
+
1 1
Subregion A Subregion B
0 0
0 1 2 3 4 5 0 1 2 3 4 5
Location mean yield Location mean yield
5 5
Exploitation cv. 1 Minimization cv. 1
by Type 1 or Type 4 by Type 2 stability:
stability: cv. 2 cv. 2
4 4 cv. 2
cv. 3
-
cv. 3 cv. 3
Cultivar yield
3 3
Cultivar yield
2 2
+
+
1 1
0 0
0 1 2 3 4 5 0 1 2 3 4 5
Year mean yield Year mean yield
524 Plant breeding and farmer participation
yield gain in unfavourable, low-yielding and enlarges the opportunities to breed for
years. Exploiting positive GY effects in specific adaptation and yield stabilit y
unfavourable years increases the security of (Section 20.3.1).
food production or agricultural income at The adoption of biotechnologies (marker-
national and household levels (Simmonds, assisted selection; genetic engineering) will
1991), making the static stability concept hardly eliminate the need to cope with
far more attractive than the dynamic one GE interactions, because genetically based
in a wide range of cases, and particularly trade-offs between yield potential and
for public institutions and less favoured tolerance to major stresses (e.g. drought;
agricultural regions. Ludlow and Muchow, 1990) and the need to
High yield stability may be associated with choose a definite level of earliness for grain
low mean yield (or low stability with high crops (Wallace, Zobel and Yourstone, 1993)
mean yield), complicating variety targeting will limit the possibility of assembling all
or selection. As an extreme example of high useful genes in a single variety. Breeding for
stability associated with low yield, consider specific adaptation may even broaden the
a genotype that yields just above zero scope for marker-assisted selection, because
(high static stability) or is consistently the a large por tion of QTLs ( Chapter 2 in
bottom-yielding (high dynamic stability) this volume) and useful markers can be
across years or environments. Obviously, a environment specific (e.g. Romagosa et al.,
less stable, higher-yielding genotype would 1999; Ribaut et al., 2007).
be preferable. The practical interest of Targeting varieties, and defining
combining high levels of mean yield with a breeding strategy in relation to GE
yield stability has led to development of the interactions, are distinct objectives and
yield reliability concept. A reliable genotype will be treated separately, as they require in
has consistently high yield across years (or part different assumptions and analytical
environments) (Section 20.2.7). techniques, even when they can be
Farmers practices frequently imply an investigated using the same set of multi-
awareness of the effects of GE interactions. environment experiment data. Information
Landraces are preferred to improved will be given on a subset of techniques
varieties in stress-prone areas, reflecting considered of primary interest on the
their specific adaptation and higher grounds of the information generated,
yield stability (Almekinders, Lowaars the ease of application (also in relation
and Bruijn, 1994). Harvest security for to software availability) and the limited
farmers producing near subsistence level is amount of input data required. Most of
associated with wider crop heterogeneity, them are discussed in greater detail in the
obtained by mixture and multiple cropping book by Annicchiarico (2002a). Valuable
of different landraces (Clawson, 1985). In information on general or specific aspects
relatively favourable areas, mixing landrace can also be found in the books of Gauch
and improved variety seed, and introgressing (1992), Basford and Tukey (2000) and Yan
varietal germplasm into landraces, improves and Kang (2003), and in those edited by
crop reliability via better response to Cooper and Hammer (1996), Kang and
favourable years (vom Brocke et al., 2003). Gauch (1996) and Kang (2002), as well
The integration of farmers into national as in papers cited hereafter for specific
or regional breeding programmes fits well issues.
Coping with and exploiting genotype-by-environment interactions 525
TABLE 20.1
Number of subregions, and predicted and observed yield gain over the pair of most-grown cultivars,
for the pair of top-yielding cultivars as predicted by different methods (observed yield data;
modelled data alone or interfaced with a Geographical Information System), and comparison of
models for predictive ability according to two criteria, for durum wheat in Algeria
Method (1) No. of Average gain (%)(3) Model comparison (4)
subregions (2)
NOTES: (1) Key to methods: JR = joint regression, AMMI = Additive Main effects and Multiplicative Interaction with one PC
axis, and FR = two-covariate factorial regression modelling; GIS = Geographical Information System. (2) Based on modelling
of 24 genotypes across 17 test sites in 1998/1999 and 1999/2000. (3) Across values of 16 individual test sites; source:
Annicchiarico, Bellah and Chiari, 2006. (4) See Section 20.2.6 for explanation of criteria; source: Annicchiarico et al., 2002.
(5) In 2000/2001.
at new sites. Modelling entry yields can realistic predictions of these gains in the
clarify the adaptive responses, facilitate the validation data set (Table 20.1). Modelled
variety targeting, and improve the predic- data can predict cultivar differences on the
tion of future responses. This is showed in site better than observed data because all
Table 20.1 with reference to the site-specific model parameters (hence, all plot values
recommendation of two top-yielding culti- used for estimating these parameters) con-
vars based on cultivar yields on the site as cur to estimate each genotype by location
observed (entry means across two years and cell mean instead of only the plot values
four replications per year) and as modelled for the specific cell mean, thereby reducing
by each of three techniques described later the amount of uncontrolled error variation
(neglecting here modelling interfaced with a (so-called noise) in the estimated GL
Geographical Information System GIS). effect (Gauch, 1992). For trials repeated
In comparison with observed data, over time, the noise relates mainly to the
modelled data implied much fewer subre- error term for GL interaction (i.e. GLY
gions (3 or 4 instead of 15, for 17 test sites), interaction, or within-site GY interac-
thereby facilitating the variety targeting, tion). Modelling can decrease the number
and provided 35 percent higher yields on of subregions because noise effects allow-
average in a validation data set. Modelling ing lower-yielding material to occasion-
was a highly cost-efficient activity here: ally appear top-yielding have been reduced
the best model allowed for nearly doubling (Gauch and Zobel, 1997). Modelling can
(+71 percent) the gain from adopting better also facilitate the extension of results to
cultivars in comparison with observed data, new sites (see Section 20.2.6).
while requiring just a modest additional cost As anticipated, analysing GL effects
in comparison with the evaluation costs. In instead of GE effects simplifies the
addition, observed data largely overestimat- modelling of adaptive responses, besides
ed the predicted gain from improved choice being conceptually sound. This is shown in
of cultivars, while modelled data provided Table 20.2 for a model described in Section
Coping with and exploiting genotype-by-environment interactions 527
TABLE 20.2
Complexity of adaptation patterns as depicted by the number of significant principal component
(PC) axes, in the analysis of genotype environment (GE) and genotype location (GL) interaction
effects in six data sets
Data set No. of sites No. of years No. of genotypes Significant PC axes (1)
GE GL
Bread wheat 31 3 18 5 2
Durum wheat 1 6 3 9 4 2
Durum wheat 2 5 2 15 4 1
Durum wheat 3 6 2 12 4 1
Maize 1 11 3 13 4 0
Maize 2 11 3 11 3 1
NOTE: (1) P < 0.01 according to FGH2 test, in an Additive Main effects and Multiplicative Interaction (AMMI) analysis.
Source: Annicchiarico (1997).
FIGURE 20.2
Flow chart of steps for targeting genotypes from analysis
of multi-location yield trials repeated over time
start
NO
cross -over
NO YES
GxL interaction
of top-ranking
entries
yield yield
other GxE YES other GxE YES
instability of any instability of any
interactions high-yielding interactions high-yielding
entry entry
NO NO YES NO NO YES
1 2 3 4
20.2.4. Responses that are remarkably The main analytical steps for targeting
complex when evaluated on a GE basis cultivars on the basis of multi-location yield
(requiring three or more dimensions for trials repeated over time are summarized
a convenient multivariate representation) in Figure 20.2. There are four possible
become relatively simple on a GL basis conclusions, which imply the targeting of
(requiring two dimensions at most). A same material over the region or distinct
simple model also facilitates the extension material to subregions and, in both cases,
of results to new sites ( Section 20.2.6). the inclusion or exclusion of yield stability
528 Plant breeding and farmer participation
effect. The error terms for GE effects are: location cell means subjected to analysis
(i) GLY interaction, for GL interaction; of adaptation as least squares means (i.e.
(ii) pooled experimental error, for GLY adjusting for the lack of orthogonality in
interaction; and (iii) pooled error (if location the data; Patterson, 1997). No observation
is fixed factor), for GY interaction. for some genotypelocation cell mean is a
This last class of models is particularly serious problem that may be dealt with by
useful when test sites differ for number specific techniques and support software
and/or timing of test years. It contemplates (e.g. Gauch, 1992: 157; 2007). A complete
the year factor nested into (instead of matrix of genotype by location cell means
crossed with) location: may also be obtained by eliminating from
the data set some genotypes or locations
Rijkr = m + Gi + Lj + Yk (Lj) + Br (Lj Yk) + with missing values.
GLij + GYik (Lj) + eijkr . Climatic, soil, biotic and crop management
data of locations and morphophysiological
Non-repeatable GL effects are here traits of genotypes (even limited to a subset
included in the GY interaction within site, of test sites) can help identify environmental
which is tested on the pooled experimental factors and adaptive traits that contribute to
error and acts as the error term for GL GL interaction. Environmental data may
interaction. also help extend the results to new sites,
The possible heterogeneity of and are needed (from all environments) for
experimental errors is a major concern only factorial regression modelling.
for trials not repeated over time (where the
pooled error acts as error term for GL 20.2.3 Joint regression model
interaction). In this case, a transformation This model was developed by Yates and
of variable may be envisaged when Cochran (1938) and proposed again, in
experimental errors vary as a function of the slightly different forms, by Finlay and
environment mean yield (Dagnelie, 1975: Wilkinson (1963), Eberhart and Russell
367; Annicchiarico, 2002a: 28). Analysing (1966) and Perkins and Jinks (1968). In
balanced data sets is preferable (estimating Perkins and Jinks model (here applied to
missing plot values according to the design). GL rather than GE interaction analysis),
The unbalance due to adoption of different the GLij effects are modelled as a function
experiment designs or to varying number of the location mean value (mj) or the
of experiment replications can be overcome location main effect (Lj), which represents
by performing the ANOVA on genotype an indicator of the ecological potential of
locationyear cell means, estimating the the site for the crop:
pooled experimental error and converting
the ANOVA sum of squares (SS) into values GLij = i Lj + dij = i + i mj + dij
relative to plot data analysis, as described in
Annicchiarico (2002a: 21). The absence of where i is the regression coefficient of the
some genotypelocationyear cell mean, or genotype i and dij is the deviation from
the variable number of test years per site, the model; in the second expression an
can be taken into account by using corrected intercept value i (equal to m i) is also
SS (usually Type III) in the combined present. The coefficients, with a mean
ANOVA, and estimating the genotype by value equal to zero, and the genotype mean
530 Plant breeding and farmer participation
yields are the relevant estimated parameters site mean yield interaction. For integration
of genotype adaptation. The expected (or with ANOVA results (relative to plot
modelled) yield response of the genotype i data), the SS for the two terms can be
at the site j is: calculated by multiplying the model R2 and
its complement to one, respectively, by the
Rij = m + Gi + Lj + i + i mj = mi + Lj + ANOVA GL interaction SS. The genotype
i + i mj . regressions MS should preferably be tested
using the deviations from regressions MS as
Finlay and Wilkinson (1963) proposed the error, whereas this latter MS is tested on
a simpler description of genotype response the appropriate error MS for the ANOVA
to site mean yield, based on the coefficient GL interaction. It is preferable to test each
bi, which is equal to (i + 1): b coefficient for difference to unity (or each
coefficient for difference to zero) using the
Rij = ai + bi mj deviation from regression of the individual
genotype as the error term (as provided
where the intercept values ai (different by a separate regression analysis for each
from previous i) are equal to (mi m genotype). Also non-significant regression
bi). Specific adaptation to high-yielding parameters should be used for modelling.
and low-yielding sites descends from bi The inference on expected yields is valid in
distinctly higher and lower than unity, the range of observed site mean yields.
respectively, in the presence of relatively Targeted genotypes depend on the site
high mi. Conversely, bi around unity mean yield. For example, the adaptive
indicates a lack of specific adaptation (and responses of top-yielding material reported
wide adaptation, if combined with high in Figure 20.3(A) suggest the presence of
mi). No definite indications on genotype three subregions: (i) high-yielding (site
adaptation can be inferred solely from b mean yield >4.5 t/ha), where Karel is top-
values. ranking but W 4267 and M 104 yield
The ANOVA GL interaction SS and almost as well; (ii) medium-yielding (mean
degrees of freedom (DF) are partitioned yield 34.5 t/ha), where Messapia is the
into two components: (i) the heterogeneity top-ranking cultivar; and (iii) low-yielding
of genotype regressions, with DF = (g 1) (mean yield <3 t/ha), where D 3415 is
for g genotypes; and (ii) the deviations preferable.
from regressions, with the residual For this and following models, a simple
GL interaction DF. The SS proportion outline of material statistically inferior to
accounted for by the former term, i.e. the the top-yielding entry at variable levels of
model R2, may be obtained: (i) by summing site mean yield may be provided by the
up the model SS across separate regression mean value of Dunnetts one-tailed critical
analyses of GL effects as a function of site difference (Annicchiarico, 2002a: 34). The
mean yield performed for each genotype, assessment is improved by computing
and calculating its proportion on the total Dunnetts critical difference for each test
SS of the regressions; or (ii) as the ratio site on the ground of its specific error term,
of model to total SS in the analysis of i.e. the GY interaction (for trials repeated
covariance of all GL effects as a function over time) or the pooled experimental error
of genotype main effect and genotype (Annicchiarico, Bellah and Chiari, 2006).
Coping with and exploiting genotype-by-environment interactions 531
FIGURE 20.3
Yields modelled by joint regression (A), and nominal yields modelled by factorial regression
on rainfall from January to June (B), for durum wheat varieties across Italian locations
(A) Nominal
(B)
Yield (t/ha) yield (t/ha)
Karel
6 W 4267 6
M 104
Messapia
5 5
Karel
W 4267
M 104
4 D 3415 4 Messapia
3 3
D 3415
2 2
2.5 3.5 4.5 5.5 200 300 400 500
Site mean yield Site rainfall (mm)
More precise and complex methods for by Kempton (1984), but became popular
cultivar comparison at specific values of after Gauchs (1992) comprehensive
site mean yield (or other environmental monograph. Genotype and location main
covariate) are available (e.g. Piepho, Denis effects are estimated by ANOVA. The GL
and van Eeuwijk, 1998). It is recommended interaction matrix is subjected to a double-
to adopt less critical Type 1 error rates, e.g. centred principal components analysis (i.e.
P < 0.20, to achieve a better balance with two simultaneous analyses: in the one, the
Type 2 error rates. Even so, the site-specific genotypes are individuals and the sites
recommendation of fairly large sets of original variables; in the other, vice versa),
statistically (P < 0.20) not different cultivars which models the GL effects according
provided markedly lower yields than that to a multiplicative term whose estimated
of two top-yielding cultivars (regardless parameters relate to statistically significant
of statistical comparisons) for durum principal components (PC) axes:
wheat in Algeria (Annicchiarico, Bellah
and Chiari, 2006), confirming the high GLij = uin vjn ln + dij = (uin ln)
Type 2 errors (mainly due to large within- (vjn ln) + dij = uin vjn + dij
site GY interaction) implied by statistical
differences. Type 2 errors may be very high where uin and vjn are eigenvectors (scaled
also when the pooled experimental error as unit vectors, i.e. ui2 = vj2 = 1) of the
acts as the error term (Kang, 1998). genotype i and the location j, respectively,
and ln is the singular value (i.e. the square
20.2.4 AMMI models root of the latent root or eigenvalue), for
The use in agricultural research of the PC axis n; and dij is the deviation
Additive Main effects and Multiplicative from the model. The further scaling of
Interaction (AMMI) models was proposed eigenvectors through multiplication by ln
532 Plant breeding and farmer participation
FIGURE 20.4
Nominal yields of top-yielding alfalfa cultivars as a function of the site score on the first
GL interaction principal component (PC) axis in a reference data set (A) or the score on
the first GE interaction PC axis of four artificial environments created by the factorial
combination of two drought stress levels (almost nil or high) by two soil types (B)
t/ha
Orchesienne
40 Europe
Prosementi
Garisenda
Robot
(A) Prospera
La Rocca
30
A B C Subregions
p l z a b s e c u o Locations
20
-3 -2 -1 0 1 2 PC 1
t/ha
19
Lodi
40 17 0
10 0
15
(B) La Rocca
16
8
Prosementi
35 Europe
3
7
30
NO STRESS NO STRESS STRESS STRESS Environments
SANDY-LOAM SILTY- CLAY SANDY-LOAM SILTY- CLAY
-4 -3 -2 -1 0 1 2 3 PC 1
in locations of northern Italy. For each Prospera and (to a lesser extent) Robot
genotype, Nij values may be calculated for the sites p and l; (ii) Prosementi and
just for the two sites with extreme PC 1 (to a lesser extent) Garisenda and Robot
scores (sites p and o; if the software for the sites z, a, b, s and e; and
outputs Rij values, Nij = Rij Lj) and the (iii) Orchesienne, Europe and (to a lesser
two values connected by a straight line. extent) Prosementi for the sites u, o and
The results suggest the presence of three c. These subregions are almost coincident
recommendation domains: (i) La Rocca, with those indicated in Figure 20.4(A),
534 Plant breeding and farmer participation
FIGURE 20.6
Expected pair of top-yielding durum wheat cultivars for each of 10 annual rainfall by 10
mean winter temperature values of Algerian locations in a factorial regression model
Rainfall
(mm)
250 16 , 24 3,6
300 13 , 16 6 , 22
350
400
450
500
3 , 21 3,5 3 , 22
550
600
650
700
4 5 6 7 8 9 10 11 12 13
Code of cultivars: Mean winter temperature (C)
GTA Dur = 3 M. Ben Bachir = 16
Chens = 5 Ofanto = 21
Sahel 77 = 6 Simeto = 22
Bidi 17 = 13 Polonicum = 24
using as error term the deviations from Its expression as nominal yield:
the model of the genotype i). Also non-
significant estimated parameters should be Nij = mi + i + in Vjn
used for modelling. The inference is valid in
the range of observed covariate values. may simplify its calculation and, for one-
Careful selection of environmental vari- covariate models, allows modelling of the
ables prior to analysis, on the basis of com- genotype responses to the environmental
mon sense and their putative importance in variable as straight lines (connecting the
GL interaction, is recommended, to limit Nij values calculated only for the extreme
the calculation process and avoid the risk of covariate levels in the graph), as shown in
multicollinearity. This risk tends to be small Figure 20.3(B) for response to site mean
(Vargas et al., 1999) but may be eliminated rainfall of the same cultivars already modelled
through a partial least squares regression by joint regression (modelling actual yields as
model (Aastveit and Martens, 1986). straight lines would require a perfectly linear
The expected yield of the genotype i response to rainfall for site mean yield).
in the location j according to the selected Specific targeting may be envisaged here for
model is: two subregions: (i) relatively low rainfall
(<400 mm), where Messapia is preferable;
Rij = m + Gi + Lj + i + in Vjn = mi + and (ii) relatively high rainfall (400 mm),
Lj + i + in Vjn . where Karel and W 4267 are preferable.
Coping with and exploiting genotype-by-environment interactions 537
For two-covariate models, nominal yield The model predictive ability depends on
responses would require a three-dimensional the accuracy (high SS) and the parsimony
graphical representation. However, it is (low DF) of its GL interaction parameters
possible to display the expected top-yielding (Gauch, 1992: 134). Uni-dimensional models
material and the subregions as a function (joint linear regression, AMMI-1 and one-
of sets of site values for the covariates, as covariate factorial regression models) can
shown in Figure 20.6 for two top-yielding be compared for predictive ability based
cultivars in each combination of 10 rainfall on the MS value of their GL interaction
by 10 winter temperature levels (a denser parameter (which takes account of both
grid of points would provide more fine- characteristics). The sum of the estimated
tuned indications). Targeting has to rely on variances of the GL interaction terms of
lists of expected genotype yields for more the model could provide a general criterion
complex models. for model comparison (Annicchiarico,
2002a: 49). The variances of GL interaction
20.2.6 MODEL COMPARISON AND PC axes (which are uncorrelated), or those
SCALING UP OF RESULTS of environmental covariates (which relate
Joint regression is a simple and popular to partial regression SS for each added
model, but cannot describe GL effects covariate), can be summed up because they
that are ecologically complex (e.g. because add independent pieces of information.
of variably occurring environmental Extending Beckers (1984) procedure
stresses) or mainly affected by a different for estimating the variance of genotype
environmental factor relative to site mean regressions to PC axes or environmental
yield (Annicchiarico, 1997). Factorial covariates, the variance of any component
regression allows for explicitly assessing of the GL interaction (sC2) can be estimated
the relationships of environmental variables from its MS (MC) and the error MS for the
with GL effects, thereby improving our ANOVA GL interaction as it follows
understanding of GL interaction (both (with respect to notations in Section 20.2.2,
in general and for single genotypes). It and Me = pooled error MS): sC2 = (MC
also simplifies the definition of subregions, Me) / r l, for trials not repeated over time;
because the genotype responses to new sites sC2 = (MC MGLY) / r y l , for location
can easily be predicted as a function of the and year crossed factors; and sC2 = (MC
site mean value for the significant covariates. MGY(L) / r y l , for the year factor nested
Its use may be limited, however, by the into location. Brancourt-Hulmel, Biarns-
unavailability of environmental data in the Dumoulin and Denis (1997) proposed a
complete set of test environments, or by second criterion equal to the following
the modest explicative value of the available ratio of SS to DF globally accounted for
covariates. With AMMI analysis, sites with by the GL interaction parameters of the
missing environmental data can be used model: % GL interaction SS / % GL
for modelling but excluded from analyses interaction DF. An empirical assessment
that assess the relationships of these data suggested the superiority of the former
with GL effects. AMMI modelling has a criterion, whose indication of the greater
broader range of application, but makes the predictive ability of the two-covariate
definition of subregions less straightforward factorial regression over joint regression
than for regression models. or AMMI-1 was confirmed by the greater
538 Plant breeding and farmer participation
ability of this model to predict the top- subregion definition relies on the supposedly
yielding material (i.e. the information of close relationship between geographical
practical interest) in a validation data set proximity and similarity for top-yielding
(Table 20.1). In contrast, the Brancourt- material of the sites (attributing new sites to
Hulmel, Biarns-Dumoulin and Denis the subregion to which the nearest test site
(1997) criterion ranked factorial regression belongs). If a test site is representative of a
as the least predictive. given area, this area can be attributed to the
The definition of subregions, initially subregion including the test site. Correlations
limited to test sites, should ideally be of environmental variables with PC axes of
scaled up to the entire target region. The sites (possibly assessed for a subset of test
possible procedures vary depending on the sites) can help characterize subregions and
modelling technique and the availability and assign new sites to the most similar subregion
the type of environmental data. Whenever according to the mean level of these variables
possible, the model established in relation on the site. Taking a step further, an equation
to genotype responses to test year values for estimating the scaled PC scores of sites
of relevant environmental variables (site as a function of environmental variables
mean yield; climatic covariates; etc.) is recorded in test years may be searched for
exploited to predict future responses on the by a stepwise multiple regression analysis
basis of long-term or mean values of these for each significant PC axis. The selected
variables on new sites or test sites. Test equation(s), if able to explain a fairly large
sites may be reassigned to other subregions, portion of variation (e.g. R2 60 percent),
if top-ranking material happens to differ can be exploited to predict the site PC score
for long-term conditions. One site may and, thereby, the expected nominal yields of
belong to more subregions depending on genotypes on the site (Annicchiarico, 2002a:
the crop management (e.g. irrigation level), 57).
if relevant. For joint regression, sites can The opportunity to interface GIS data
be assigned to subregions depending on allows for a very fine-tuned geographical
their long-term mean yield (as known from definition of subregions for scaling up
statistical records or, possibly, as predicted factorial regression or AMMI results
by mean values of relevant environmental (Annicchiarico, Bellah and Chiari, 2006).
variables). For factorial regression, nominal Alternatively, a simple Decision-aid
yields of genotypes can be estimated System could be developed that outputs
as a function of site mean values of the the expected nominal yields of the cultivars
significant covariates. For models with one (possibly together with an average value of
or two covariates, graphical expressions Dunnetts one-tailed critical difference) as a
such as Figure 20.3(B) or Figure 20.6 can function of the inputted site mean value of
greatly simplify the scaling up (reading the relevant environmental variables.
best-yielding material as a function of the Scaling up may introduce a bias due to
covariate value(s) on the site). neglecting some important environmental
Extending results is more complex for variable or inaccurately estimating its
AMMI models. If no environmental data is effects. This bias was assessed for durum
available or no relationship is found between wheat variety targeting in Algeria, where
environmental variation and ordination on genotype responses to test sites or new
significant GL interaction PC axes of sites, sites were predicted as a function of long-
Coping with and exploiting genotype-by-environment interactions 539
term values in a GIS of winter mean year within location (My(l)) in an ANOVA
temperature and rainfall over the season, limited to locationyear cell means of the
which were selected as covariates for relevant genotype that also includes the
both factorial regression and in a multiple location factor. High stability is indicated by
regression for predicting the site PC 1 score low My(l). However, the estimate provided
(Annicchiarico, Bellah and Chiari, 2006). by My(l) is inflated by the experimental
Data for test sites in a validation data set error variance. An unbiased estimate can be
revealed that GIS-based recommendations provided by the temporal stability variance
implied just a slight yield decrease relative Sy(l)2 (Annicchiarico, 2002a: 81):
to those based on conventional modelling
(Table 20.1), while greatly enlarging the Sy(l)2 = My(l) (Me / r)
scope for site-specific targeting.
where Me = pooled error in the general
20.2.7 Taking account of yield stability ANOVA (performed earlier according
Static and dynamic stability concepts were to Figure 20.2), and r = number of
introduced in Section 20.1. A few measures experiment replications. Sy(l)2 and My(l)
of static stability will be considered herein, are equivalent for ranking genotypes, but
reflecting the following advantages of static the former is recommended for testing
measures over dynamic ones: (i) some- genotype differences and for adoption in
what higher repeatability or heritability; yield reliability indexes. Well-performing
(ii) estimation independent from the set of genotypes can be compared for Sy(l)2 value
tested genotypes (which allows for broader by ordinary tests for variance comparison
generalization); (iii) less ambiguous agro- such as Fishers bilateral test (for two
nomic interpretation; and (iv) relevance entries) or Hartleys test (for more entries),
for increasing food security or agricultural preferably using less critical Type 1 error
income (Lin, Binns and Lefkovitch, 1986; rates (e.g. P < 0.05 or P < 0.10). For l test
Annicchiarico, 2002a: 81). The repeatability locations and y test years, DF = l (y 1) for
across different sets of environments may each estimated Sy(l)2 parameter.
vary, depending on the crop and the region. The environmental variance (S2) measures
It increases with the temporal scale of the the static stability across environments.
assessment, but remains distinctly lower With reference to notations in Section 20.2.2
than that of genotype mean yield across and for e = number of environments, its
environments. The assessment of yield sta- value for the genotype i is: Si2 = (Rij mi)2
bility requires numerous test environments / (e 1). Greatest stability is S2 = 0. This
(eight or more) to be reliable, given its high measure is simpler to compute than Sy(l)2
sampling error (Kang, 1998). but requires a more complex procedure for
The entry regression as a function of genotype comparison based on Ekbohms
environment or year mean yield (Figure 20.1) test (described in Annicchiarico, 2002a: 82)
may have various drawbacks as a stability or other tests.
measure (poor ability to describe GE Temporal stability, which meets farmers
effects; too few years to assess stability perception of stability, is relevant for site-
over time). Lin and Binns (1988) measure specific targeting of cultivars (path 4 in
of average within-site temporal stability Figure 20.2) and can be used also for
across years (or crop cycles) is the MS for wide targeting (path 2 in Figure 20.2). The
540 Plant breeding and farmer participation
FIGURE 20.8
A specific-adaptation strategy implying the generation of novel germplasm
in a national (or regional) research centre, the research-managed early selection
of novel germplasm and evaluation of genetic resources in distinct subregions,
and subsequent farmers selection stages in each subregion
parent material, the genetic structure, the each subregion for specific adaptation: see
ideotype and the single adaptive traits, and the Section 20.3.4).
role of participatory plant breeding. Decisions For targeting genotypes, only the GL
on most of these elements may conveniently effects that imply a change of top-ranking
be verified by further experimental work. genotype(s) across locations are relevant
In particular, wide- vs. specific-adaptation (as these effects define the subregions).
strategies may be compared on the basis For defining the adaptation strategy, all
of actual yield gains. A fair comparison of GL effects that arise from lack of genetic
adaptation strategies implies similar costs by correlation among sites for entry response
assuming the same total number of selection are relevant, because the results for a
environments. Specific adaptation may then given data set are extrapolated to produce
allow for higher or lower yield gains than information on the GL effects that are
wide adaptation, because the advantage of likely to be met in future breeding for the
exploiting the portion of GL effects that region. A candidate subregion includes the
relates to subdivision of the target region locations that are similar in terms of genetic
(i.e. the genotype subregion interaction) correlation.
may be offset by the greater error of the The main analytical steps to provisionally
estimated entry means that arises from the define the adaptation strategy and yield
lower number of selection environments stability targets from multi-location yield
within each subregion (the error depends trials repeated over time are summarized in
mainly on the size of GE interaction over Figure 20.9. GL or GE effects that arise
the region for wide adaptation and within from heterogeneity of genotypic variance
FIGURE 20.9
Flow chart of steps for defining the adaptation strategy and yield stability
targets from analysis of multi-location yield trials repeated over time
start
LOW NO
LOW
variance of variance of variance of
other GxE other GxE other GxE inte -
ractions within
interactions interactions
subregion
LOW
HIGH LOW HIGH LOW HIGH
1 3 5
selection for wide selection for wide selection for specic
adaptation (in few environ- adaptation (on contrasting adaptation (in few environ-
1 ments across the region) sites in the region) ments in the subregion)
2 4 6
selection for wide adaptation selection for wide adaptation and selection for specic adaptation
and yield stability (in several yield stability (in several environments and yield stability (in several
environments across the region)
region on contrasting sites in the region) environments in the subregion)
rather than lack of genetic correlation among to the genotypic variance component (e.g.
environments are irrelevant for breeding (as 200 percent). Decisions on the stability
they merely modify the size of the entry target are subregion-specific (depending
differences) and should be removed if they on results for the relevant subset of sites)
are too large. The relative size of the lack of and can be affected by other elements (e.g.
genetic correlation and the heterogeneity costs of additional selection environments;
of genotypic variance among environments emphasis on food security policies).
may be estimated or, more simply, the need
for transforming data could be verified 20.3.2 Types of data and estimation of
(see Section 20.3.2), before estimating the variance components
variance components relative to spatial and The current requirements for use of
temporal GE interaction. An analysis data sets are more stringent than those
of adaptation aimed to define candidate for targeting genotypes, given the larger
subregions (by classifying test sites on the inference space of the analysis. Ideally,
basis of their similarity for GL effects) may the sample of sites should represent the
be justified if the GL interaction variance different areas and cropping systems in
is significant and moderately large relative to proportion to their importance, and the
the genotypic variance, e.g. 3035 percent germplasm sample should represent the
(Atlin et al., 2000). Two (or possibly three) genetic base of local interest (by including
candidate subregions may be identified that the main cultivar types and origins, or
are large enough to be of practical interest breeding lines derived from recombination
and lend themselves to a definition based on from the major germplasm groups). The
geography, environmental factors or farming lack of random sampling of entries is not
practices. Wide- and specific-adaptation a limitation, because a set of elite varieties
scenarios can be compared in terms of yield or breeding lines may represent the genetic
gains predicted from original yield data of base better than a random sample.
the same data set. Wide adaptation may be The GE interaction variance components
preferred owing to low GL interaction relative to the lack of genetic correlation
variance or to high GL interaction variance and the heterogeneity of genotypic variance
with no clear advantage of specific breeding, among environments can be estimated as
with different implications for the choice of described by Cooper, DeLacy and Basford
selection environments (the analysis can also (1996). If the latter term has larger variance
help locate these environments). Adaptation than the former, it should be reduced by
strategies may also be compared according a suitable data transformation. This may
to predicted gains in other data sets (e.g. occur when the environment mean yields
including a few sites representative of the vary widely (Annicchiarico, 2002a: 51). In
candidate subregions), or actual gains from this situation, the regression of the within-
following selection work. site phenotypic variance of genotype yields
Yield stability may be justified as a target (averaged over test years and replications)
when the overall variance accounted for by as a function of site mean yield, with both
the relevant GE interaction components terms expressed on a logarithmic scale,
(either the GY plus GLY interaction or provides a quick option for verifying the
the within-site GY interaction, depending need for transforming data and indicates
on the ANOVA model) is large relative the proper transformation (Annicchiarico,
546 Plant breeding and farmer participation
FIGURE 20.10
Nominal yields of durum wheat genotypes as a function of the scaled GL interaction
PC 1 score of 14 Algerian locations (black triangles), and definition of two subregions
(A and B) based on cluster analysis of site PC 1 scores (above) and the estimation of the
main crossover point CP (below), for original and log-transformed data
2.50
3.30
2.00
1.50
3.10
1.00
A B A B
0.50 CP CP
A B A B
0.00 2.90
-0.75 -0.25 0.25 0.75 1.25 1.75 -0.75 -0.25 0.25 0.75
Location PC 1 score Location PC 1 score
data from the member sites). Other criteria analysis over the other methods (based on
are also available (DeLacy et al., 1996a). cluster analysis or the main crossover point),
In most cases, the maximum number of while confirming the need for modelling
subregions (two or three) is fixed a priori. transformed data when most GE effects
Singh, Ceccarelli and Grando, (1999) are due to heterogeneity of genotypic
proposed the estimation of the main variance among sites (Annicchiarico,
crossover point in the joint regression 2002b). Figure 20.10 provides some insight
model, i.e. the site mean yield where the into reasons for these results. Responses for
interaction of crossover type between original yields implied larger GL effects
genotypes reaches the highest frequency, as for higher-yielding sites (here tending to
a cut-off for dividing the test sites into two higher PC 1 score), owing to their higher
subregions (one high- and one low-yielding). genotypic variance. As a consequence, many
This approach was extended to AMMI-1 low-yielding sites were grouped together
and to one-covariate factorial regression by cluster analysis, because they tended
models by Annicchiarico (2002a). to appear less distinct than they really
The cited techniques often provide were. The data transformation (selected
different classification results for the same as outlined in Section 20.3.2) removed
data. Their empirical comparison based on the heterogeneity of genotypic variance
the assumption that a useful technique tends between low- and high-yielding sites (while
to maximize the selection gains for a specific- maintaining the lower genotype variance for
adaptation strategy suggested the superiority sites with intermediate PC 1 score that is
of pattern analysis and AMMI + cluster intrinsic to the adaptive responses). Cluster
548 Plant breeding and farmer participation
analysis, unlike the main crossover criterion, sites on the basis of their response in small
subdivided the sites on the basis of their trials (Fox and Rosielle, 1982). For example,
discontinuities for GL effects as expressed the cultivars La Rocca, Prosementi and
by the relevant characteristic, i.e. the site Orchesienne in Figure 20.4(A) could be
PC 1 score, thereby justifying its better used for assigning new sites to one of three
classification ability. subregions identified by AMMI + cluster
A two-subregion scenario is the first to analysis. Additional information may allow
compare with wide adaptation, as its lower estimating the relative size of subregions in
interest relative to wide adaptation rules out terms of crop growing area or, especially for
more complex specific-adaptation scenarios. private companies, seed market importance.
A small subregion that hardly justifies any The assessment of site similarity for
specific breeding may be merged with the GL effects is recommended to explore
most similar large subregion (unless farmers the opportunities for specific breeding. It
have very specific and different preferences). revealed, for example, that winter cold stress
An indication of the approximate proportion is more important than drought stress in
of the target region occupied by each subre- determining subregions for winter cereals
gion is useful for comparing the adaptation in Italy (Annicchiarico, 1997) and Algeria
strategies and possibly for other reasons (e.g. (Annicchiarico, Bellah and Chiari, 2005).
estimation of seed markets). A very rough Sometimes, however, candidate subregions
indication is provided by the proportion of for comparing adaptation strategies are
test sites assigned to each subregion in the defined a priori on a geographical, climatic
analysis of adaptation. The scaling-up of or crop management basis (e.g. Atlin et al.,
subregion definition over the region may 2000). The investigation of environment
be somewhat arbitrary, by assigning an area similarity for GE effects in fairly small
represented by a given test site to the sub- data sets including a few test sites repre-
region in which the test site was classified, sentative of distinct geoclimatic areas and
or characterizing the subregions according several environments per site (issued by
to their mean values for environmental vari- different test years and, possibly, different
ables correlated with GL interaction PC crop managements per year) may provide
scores of sites (e.g. Annicchiarico, 1992). useful indications for the adaptation strat-
These variables may also be exploited for a egy (e.g. Annicchiarico and Iannucci, 2008),
discriminant analysis of subregions which as specific breeding requires that location is
may provide a thorough up-scaling pro- the main determinant of environment clas-
cedure and, for the frequent case of two sification and crossover GE interactions.
subregions, can easily be performed by mul-
tiple regression (Annicchiarico, 2002a: 59). 20.3.4 Comparison of wide- vs.
The possible interfacing of this analysis specific-adaptation strategies
with GIS allows for a detailed geographi- The comparison may concern predicted
cal definition of subregions (Annicchiarico, or actual yield gains, and vary depend-
Bellah and Chiari, 2005). As an alternative, ing on the crop breeding system and the
a few well-performing reference genotypes selection procedures. It hypothesizes one
characterized by contrasting adaptation and selection cycle (over a definite time span)
a specific ranking order in each subregion on research-managed selection sites or, pos-
could be used to indirectly characterize new sibly, managed environments, but more
Coping with and exploiting genotype-by-environment interactions 549
selection cycles may be envisaged, espe- method for genotype and environment as
cially for actual yield gains. The adaptation random factors using all test environments
strategies imply similar costs by assuming (for wide adaptation) or their subsets (for
the same total number of selection envi- specific adaptation), fixing E and R as
ronments (as number of sites number of hypothesized for selection. R and i are set to
years), assigning sites to subregions roughly constant values in the assessment. Hereafter,
in proportion to their relative size. Specific EA and EB represent the number of selection
selection is based on entry mean yields over environments for two subregions (A and
selection environments of the target subre- B, respectively) in a specific-adaptation
gion. Selection for wide adaptation is based scenario, whereas EAB = EA + EB is the
on entry mean yields over all test environ- number of selection environments that are
ments hypothesized for specific breeding, used, in a wide-adaptation scenario, for
in agreement with Lin and Butlers (1988) parallel selection across the subregions. PA
suggestion to choose selection sites across and PB represent the proportion of the target
the region in a stratified manner and in pro- region occupied by subregions A and B (PA
portion to the relative size of site groups. + PB = 1), as estimated by the proportion
The following method (Annicchiarico, of test sites classified in each subregion
2002b) aims to compare adaptation or, more precisely, after scaling up the
strategies for inbred lines or clones in terms subregion definition (for private companies,
of predicted yield gains over the target they might rather estimate the relative
region from undefined selection sites, using commercial importance of the subregions).
original yields of a data set possibly used EA and EB should be roughly proportional
also for defining candidate subregions. In to PA and PB, respectively (e.g. for EAB = 6
general, the average predicted yield gain = 3 sites by 2 years, and PA = 0.64: EA = 4 =
over E environments can be estimated as 2 sites by 2 years, and EB = 2 = 1 site by 2
(DeLacy et al., 1996a): G = i h2 sp (where years). The average predicted gain (per unit
i = standardized selection differential, h2 area) by a wide-adaptation strategy is:
= estimated broad sense heritability on a
genotype mean basis, and sp = square root GW = i hAB2 sp(AB) [20.2]
of the estimated phenotypic variance across
environments). In particular: where hAB2 and sp(AB) are obtained from
equation [20.1] after estimating the
h2 = sg2 / [sg2 + (sge2 / E) + (se2 / E R)] [20.1] components of variance for the whole
set of environments in the data set, and
where sg2, sge2 and se2 are estimates of the inserting EAB and R values as appropriate.
variance components for genotype, GE The average predicted yield gain over the
interaction and pooled experimental error, region provided by breeding for specific
respectively, and E and R = numbers of adaptation (GS) arises from a weighted
selection environments and experiment mean of the gains GA and GB predicted
replications. The sp term is equal to the for the subregions A and B, respectively:
square root of the denominator in equation
[20.1]. In the formulae for prediction of GA = i hA2 sp(A) ; GB = i hB2 sp(B) ;
yield gains, h2 values are calculated from GS = [(GA PA) + (GB PB)] / (PA + PB) =
variance components estimated by a REML (GA PA) + (GB PB) [20.3]
550 Plant breeding and farmer participation
where heritability and phenotypic variance when chosen, a constant gain from intra-
values are obtained from equation [20.1] population selection (e.g. Annicchiarico
after estimating the components of and Piano, 2005). Predicted gains could be
variance for the subset of test environments computed as the difference between the
belonging to subregion A (values hA2 and top-ranking cultivar and the mean of all
sp(A)) or B (hB2 and sp(B)), and inserting EA or tested entries.
EB and R in the equations as appropriate. Reliable predicted gains for comparing
This procedure can easily be extended to adaptation strategies in open-pollinated
three (or more) subregions, computing the species may be obtained from multi-
specific-adaptation gains across more than environment progeny testing of individuals
two subregions. Another procedure for as half-sib or full-sib families or as selfed
the same context (i.e. predicted yield gains progenies. It suffices to substitute the
for inbred lines or clones from undefined appropriate narrow sense heritability term
selection sites) was proposed by Atlin et (Wricke and Weber, 1986) for the broad
al. (2000). Piepho and Mhring (2005) sense heritability in formula [20.1] when
expanded this approach by considering a using this equation for estimating GW and
more complex scenario that maximizes the GS according to formulae [20.2] and [20.3].
selection gains by using for specific selection For example, the following equation and
also the data from other subregions. These the square root of its denominator allow for
data are given a weight proportional to estimating h2 and sp, respectively, for half-
their relevance for the target subregion. sib progeny testing targeted to selection of
Another adaptation strategy may parent material as clone or selfed progeny:
contemplate the selection in one subregion
also for another subregion, so that the yield h2 = 0.5 sa2 / [0.25 sa2 + (0.25 sae2 / E) +
gains in the latter subregion derive from (se2 / E R)] [20.4]
correlated responses in an indirect selection
context. Formulae for comparing this where sa2, sae2 and se2 are estimated variance
strategy against the previous ones in terms components relative to the additive genetic
of predicted gains (taking conveniently variance, the interaction of additive genetic
into account the effect of GE interactions effects with environment and the pooled
within subregions, unlike more simplified error, respectively, and E and R = numbers
approaches) are reported elsewhere of hypothesized selection environments
(Annicchiarico, 2002a: 68; 2002b). and experiment replications. The REML
For open-pollinated species, the analysis performed on family plot values
comparison of strategies based on predicted for the relevant sets of test environments
gains from selection of populations provides the estimate of se2, and allow
(as represented by cultivars in a MET estimation of the other variance components
data set) is frequently out of context, from the variance among families (sg2)
as selection mainly concerns individual and the family environment interaction
plants. Preliminary indications may be variance (sge2) (assuming no inbreeding in
obtained by comparing the top-ranking the tested material) as:
cultivars according to each adaptation
strategy, as if each cultivar was used as the sa2 = 4 sg2; sae2 = 4 sge2.
unique genetic base and could provide,
Coping with and exploiting genotype-by-environment interactions 551
The previous procedures assume cultivars) and weighting them on the relative
undefined selection sites. Thus, they relate importance of the subregions. For example,
to the average screening ability of sites Ceccarelli, Grando and Impiglia (1998)
within each subregion (as estimated from selected barley genotypes within a large
the sample of test sites). It is also possible germplasm pool for wide and for specific
to compare the adaptation strategies for adaptation to an unfavourable (A) and a
predicted yield gains from selection in favourable (B) subregion, reporting the
managed environments or in previously- mean values of the selected groups tested in
defined, nearly-optimal selection sites the two subregions. For one set of material
(Annicchiarico, Bellah and Chiari, 2005). (Cohort 89), the yield gains estimated with
Predicted gains are correlated gains from respect to a set of high-yielding control
the defined selection environments to the cultivars were: GA = 0.03 t/ha, and GB
target environments. For inbred lines or = 0.08 t/ha, for specific adaptation; GA
clones, they are (DeLacy et al., 1996a): = 0.03 t/ha, and GB = 0.08 t/ha, for
wide adaptation. Based on these values,
GT/S = i r(S,T) sp(T) [20.5] the advantage of the former strategy is
manifest. If the subregions were of equal
where r(S,T) = phenotypic correlation for size (PA = PB = 0.50), the gain over the
entry mean yield between selection and region from specific (GS) and from wide
target environments, and sp(T) = phenotypic (GW) breeding could be estimated as:
standard deviation in the target environments.
The basic difference with previous formulae GS = (GA PA) + (GB PB) = (0.03 0.50)
for predicted gains of inbreds or clones is the + (0.08 0.50) = 0.055 t/ha per cycle
substitution of the appropriate r(S,T) term in GW = (GA PA) + (GB PB) = (0.03
place of h2. For agricultural sites, however, 0.50) + (0.08 0.50) = 0.025 t/ha per cycle
the gains relate to estimates of site screening
ability that may largely be affected by specific implying a greater efficiency of specific
conditions during the test years. breeding equal to GS/GW = 0.055/0.025 =
A promising specific-adaptation strategy 220 percent.
may be compared with wide adaptation Large data sets for inbred lines or clones
in terms of actual yield gains. Selection that include many entries and several test
is performed both independently within years may also be used for assessing actual
subregion (specific adaptation) and jointly yield gains of adaptation strategies, after
across subregions (wide adaptation) defining candidate subregions and selection
at previously-defined selection sites (or sites. For example, Annicchiarico, Bellah
managed environments), contemplating the and Chiari (2005) used two years and a total
same total number of selection environments of three sites of a three-year durum wheat
(roughly assigned in proportion to the data set for wide or subregion-specific entry
relative size of subregions). The selected selection, and the remaining environments
groups are compared across a sample of to assess actual gains in each subregion
environments, assessing the gains provided (estimated as yield difference between the
by each strategy in each subregion (e.g. mean of selected entries and the mean of all
in terms of difference between the group entries), averaging results across the three
mean and the mean of a set of high-yielding possible pairs of selection years (Table 20.3).
552 Plant breeding and farmer participation
TABLE 20.3
Mean yield of selected durum wheat entries, and average observed and predicted yield gains per
selection cycle in two subregions and over the Algerian durum wheat cropping region, for specific-
and wide-adaptation strategies
Specific adaptation Wide adaptation Specific/wide ratio (%)
a Selected fraction: 3 entries out of 24. Total selection environments: 6 (3 sites by 2 years), of which 2 assigned to
Subregion A (proportion of the region = 0.322) and 4 to Subregion B (proportion of the region = 0.678).
b Values averaged across three pairs of test years.
c Gain computed as the difference between the mean of selected entries and the mean of all entries.
d For defined selection locations (using equation [20.5]). Values averaged across three pairs of test years.
e For undefined selection locations (using equation [20.2] for wide adaptation, and equation [20.3] for specific adaptation; in
Gains over the region for each strategy were was given by Annicchiarico (2007) for
weighted means of those in each subregion phenotypic selection of alfalfa in artificial
(as shown for the barley data) and indicated environments capable of reproducing the
the greater efficiency of specific over wide adaptive responses occurring across the three
breeding (GS/GW = 107.1 percent) as a subregions shown in Figure 20.4(A). Direct
consequence of greater gain (+39.5 percent) selection for specific adaptation targeted
in the stressful subregion A. This procedure each of the contrasting subregions A and
is less reliable than comparisons for actual C, exploiting correlated selection gains for
gains on a larger and independent genotype the intermediate subregion B. To reduce the
sample. However, its results agreed closely evaluation costs, the selection environments
with those of predicted gains for undefined also acted as test environments for the
selection sites (GS/GW = 104.1 percent), selections (possibly introducing some bias
while being less consistent with those relative to agricultural sites).
of predicted gains for the same set of All cited procedures hypothesize
selection sites (GS/GW = 102.2 percent) growing the novel germplasm in all selection
(Table 20.3). environments. Their indication of some
Especially for cross-pollinated crops, the advantage for specific breeding probably
lack of sufficiently large data sets may limit implies larger gains after optimizing other
the comparisons of adaptation strategies elements of the breeding strategy by
based on predicted gains, giving impulse to considering, at least to some extent: (i) the
those based on actual gains. One example allocation of novel germplasm to only one
Coping with and exploiting genotype-by-environment interactions 553
subregion on the basis of crucial adaptive screening ability of the sites (as indicated
traits (or molecular markers) assessed by phenotypic correlations between selec-
preliminarily at the main research centre; or tion and target environments for entry
(ii) the use of a distinct genetic base for each yields) that should be maximized (e.g.
subregion. Indications for these elements Annicchiarico, Bellah and Chiari, 2005).
may be provided by the MET data set, Selection for wide adaptation in the
with or without further research. Anyway, presence of sizable GL interaction should
the comparison of adaptation strategies be performed across sites that contrast for
could not take account of some positive GL effects (as hypothesized for comparing
effects of breeding for specific adaptation adaptation strategies) and are jointly capable
(Section 20.3.1) that are difficult to quantify. of maximizing the screening ability (as
indicated by phenotypic correlations), rather
20.3.5 Definition and use of selection than across sites that maximize individually
environments the screening ability (which are implicitly
In the presence of sizable GL interaction, the similar for GL effects). Contrasting sites
main research centre may host a preliminary offer the opportunity for disclosing and
selection stage if its screening ability for the selecting material capable of assembling
target region is high. According to formula different adaptive traits of interest for the
[20.5] in Section 20.3.4, the screening ability region (Calhoun et al., 1994). Thus, optimal
of a site (or a managed environment) is selection sites may be identified for wide or
proportional to the phenotypic correlation specific adaptation by the same procedures.
between entry yields on the site and entry Phenotypic correlations between selection
mean yields over the target environments. and target environments also allow the
The phenotypic correlation takes account assessing of the lower yield gain expected
of the genetic correlation between selection from adopting suboptimal sites.
and target environments and the broad sense Managed or artificial selection environ-
heritability on the site (Cooper, DeLacy and ments that reproduce the genotype adap-
Basford, 1996). When breeding for specific tive responses and do not imply very high
adaptation, this preliminary selection stage implementation costs can partly replace
may also allow for the allocation of material agricultural selection sites to reduce costs
to a specific subregion on the basis of (especially when optimal sites belong to
adaptive traits. remote areas or have little infrastructure)
Multi-environment data can also help or to increase the selection gains (especially
locate optimal selection sites for research- when agricultural sites are subject to wide
managed selection (also usable for detecting GY interaction due to unpredictable cli-
parent germplasm of specific interest for matic conditions). For example, the artificial
subregions; see Figure 20.8). Preliminary environments in Figure 20.4(B) were estab-
indications may be obtained from site ordi- lished on the ground of the positive correla-
nation in the analysis of adaptation or site tions of soil clay content and drought stress
classification for GL effects. The optimal level with PC 1 score of alfalfa test sites in
selection site for a given subregion has the Figure 20.4(A) (Annicchiarico, 1992). They
highest screening ability for the relevant could reproduce the adaptive responses
target environments. When adopting more occurring in three subregions (as shown
sites within a subregion, it is the joint by three reference varieties: Figure 20.4),
554 Plant breeding and farmer participation
and may be used to select for wide or entry selection that is proportional to the
specific adaptation (Annicchiarico, 2007). frequency of its group.
Selection under the natural conditions of The optimal number of selection sites,
the breeding centre (located in subregion years and experiment replications for inbred
A), implying sandy-loam soil and negligi- lines of clones in a region or subregion may
ble stress, would produce varieties specifi- be investigated after estimating the genotypic
cally adapted to these conditions, such as and genotype-environmental variance
cultivar Lodi in Figure 20.4(B). components for a representative sample of
Managed environments may also be used elite material and target environments. The
to breed for wide adaptation to regions aim is maximizing, for about same costs,
featured by large within-site GY interaction the yield gain: G = i h2 sp , where h2 and
and small repeatable GL effects due to sp are computed by the following equation
wide year-to-year climatic variation. In and the square root of its denominator,
such regions, agricultural sites in individual respectively, for L locations, Y years and
years frequently misrepresent the target R replication numbers hypothesized for
environments, leading to low selection selection (Cooper et al., 1999):
gains (Cooper, DeLacy and Basford, 1996).
An optimal set of managed environments h2 = sg2 / (sg2 + sgl2 / L + sgy2 / Y + sgly2 / LY
can be identified by assessing the joint + se2 / RLY)
screening ability of these environments
(Cooper et al., 1995, 1997). Federer and Estimates may also relate to more
Scully (1993) proposed statistical designs to complex scenarios, e.g. two-stage selection
select material for wide adaptation across (Grneberg et al., 2004) or among-cross
a factorial combination of two or three (bulk) plus within-cross pure line selection
crop management or physical factors that (Cooper et al., 1999). Research-managed
reproduce the variation for environmental selection trials would usually include at
variables associated with GE effects. least two replications, while the number of
Selecting on agricultural sites for wide selection years is kept low (often no more
adaptation to climatically unpredictable than two) so as not to delay the release of
regions may increase its efficiency by a varieties. Thus, decisions mainly regard the
procedure proposed by Podlich, Cooper number of selection sites. Selecting also for
and Basford (1999). A large sample of target yield stability may lead to increases in this
environments is classified on the basis of number (Figure 20.9), if socio-economically
GE interaction effects, identifying a few convenient.
major groups whose relative frequency Predicted gains for different scenarios
is estimated and that are characterized relative to managed environments depend
either by a specific response of some on the h2 value over selection environments
probe genotypes or a definite value of (as affected by hypothesized L, Y and R
some crucial climatic variable(s). Each new values) and the genetic correlation between
selection environment is classified according selection and target environments (DeLacy
to the response of the probe genotypes et al., 1996a; Qiao et al., 2004).
(grown along with the tested material) In various contexts, only an unbalanced
or the relevant climatic variable(s), and is data set (with possibly many missing
given a weight on the future MET-based genotype-environment cell means) may be
Coping with and exploiting genotype-by-environment interactions 555
available for entry selection. Best Linear the consistency of response and the value
Unbiased Prediction (BLUP) entry means for farmers of a variety across the target
as estimated by a REML method should be region or subregion.
used in this case. BLUP entry main effects
(Pi) differ from Best Linear Unbiased 20.3.6 Identification of genetic
Estimate (BLUE) entry main effects (Gi, as resources, adaptive traits and useful
provided by least squares means) because markers
they are shrunk to a greater extent for entries The analysis of adaptation can also produce
with less observations, to take account of information on germplasm which, within
the greater uncertainty introduced by less a given adaptation strategy, is of special
MET. In particular, Pi effects are shrunk interest as parent for crosses or as popu-
in proportion to the difference to unity lation for recurrent selection in view of
of the broad sense heritability on an entry its adaptive response. In general, evidence
mean basis (DeLacy et al., 1996a), i.e. Pi points to a moderate heritability of adapta-
= (h2 Gi), where h2 = sg2 / [sg2 + (sge2 / ei) tion parameters (Becker and Lon, 1988).
+ (se2 / ei ri)], and ei and ri are numbers of For example, crosses for wide adaptation
environments and experiment replications could be envisaged between genotypes:
for the entry i (the latter as harmonic mean). (i) possessing high mean yield and the
The variance components are constant desired adaptive response (as indicated by
values estimated by a REML method b 1 in joint regression, value near zero
for the entire data set. This procedure in factorial regression, and genotype PC
is simpler and more reliable than other score near zero in AMMI analysis); or
methods that also consider the broad sense (ii) between pairs of genotypes possessing
heritability in single trials (DeLacy et al., high mean yield and specific adaptation of
1996a). The BLUP means can be obtained contrasting type, such as Orchesienne and
by a REML analysis performed on the La Rocca in Figure 20.4(A). Genotypes
genotype-environment cell means, holding may also be classified for adaptive response
genotype and GE interaction as random by pattern analysis (Cooper, DeLacy and
effects (while environment is fixed) (Hill Basford, 1996). The analysis of adaptation
and Rosenberger, 1985). Specific models of genetic resources with contrasting origin
may be applied to unreplicated trials or may highlight the relationship of the envi-
more complex data structures, e.g. sites and ronment of origin to the adaptive response
farms (or years) within site (Smith, Cullis as determined by evolutionary adaptation.
and Thomson, 2001; Coe, 2002). BLUP Finally, the comparison of different variety
entry means also provide more realistic types may contribute to decisions of breed-
predictions of yield gains from actually ing programmes on the genetic structure
selected entries than do BLUE means. of novel germplasm (Brancourt-Hulmel,
Breeding for wide or specific adaptation Biarns-Dumoulin and Denis, 1997).
can account for yield stability by selecting The analysis of adaptation may also provide
entries according to Kataokas index of preliminary indications on traits contributing
reliability (see Section 20.2.7) instead of to wide or specific adaptation, by correlations
mean yield over selection environments. of the estimated genotype adaptation
This measure is justified by the fact that all parameters with morphophysiological
GE effects (including GL ones) influence traits (possibly recorded in a subset of test
556 Plant breeding and farmer participation
subplots, surrounded by eight test entries) volume), each farm within a subregion may
while the others are randomly placed to host an incomplete block.
estimate the experimental error, allows for Experimental designs cannot control the
an accurate and flexible adjustment of test effects of environmental variation within
entry values as a function of: (i) the row complete or incomplete blocks. These
and column effects; or (ii) the covariate effects may be large in stress-prone sites
represented by the value of systematic and in trials with many entries. For plots
controls (which estimates the local yield arranged in a rectangular row-column
potential). As a development of Federer, array, the ANOVA residual Emn of the
Nair and Raghavaraos (1975) design, a plot located in row m and column n may
partially replicated design implies two be modelled spatially as a function of:
replications for a subset of test entries to (i) uni- or bi-directional fertility gradients
improve the estimation of the experimental estimated by polynomial functions of
error (see also Section 20.3.1). row (Rm) and/or column (Cn) numbers,
Resolvable incomplete block designs, e.g. a second-degree polynomial response
in which the incomplete blocks can be surface; (ii) the covariate Xmn represented
grouped to form a complete replication of by the mean value of the ANOVA residuals
the entries, are preferable to non-resolvable for the neighbouring plots (which estimates
ones because this double-blocking the local yield potential), through a nearest-
structure allows for better error control neighbour (or Papadakis) method: Emn = b
and for the possible analysis as a RCB Xmn + emn (where emn is the non-modelled
for missing data (Basford et al., 1996). To residual) (Brownie, Bowman and Burton,
partly overcome the constraints of lattice 1993). As a third, more complex approach,
designs for numbers of tested entries and of Cullis and Gleeson (1991) proposed
plots within incomplete block, Patterson, modelling the residuals as a function of
Williams and Hunter (1978) devised a the spatial autocorrelation between pairs
new class of resolvable designs termed of neighbouring plots assessed separately
alpha lattices in which, given k plots per along rows and columns. If there is spatial
incomplete block, the number of entries pattern, the variance between residuals
g may be whatever multiple of k. The of neighbouring plots will be lower than
efficiency of these designs increases when k that between residuals of plots far apart,
approaches the square root of g. leading to autocorrelation (estimated by
The layout of farmer-managed selection a moving average). This two-dimensional
trials depends on the total number of test separable autoregressive spatial model of
entries and the plots available per farm. first order (AR1 AR1) can describe many
Within these constraints, it should preferably spatial patterns and has frequently proved
allow for estimating an experimental error adequate. Therefore, Gilmour, Cullis and
on the farm, e.g. by unreplicated trials Verbyla (1997) proposed generally fitting
including a few replicated entries, or on the this model and then displaying by various
site and its surroundings, by assigning a diagnostic tools (e.g. the variogram of non-
complete block to each of few farms or an modelled residuals along rows and columns)
incomplete block to each of several farms. the possible presence of remaining pattern,
In less favourable instances or when farm to be modelled by additional covariates
size is small (see also Chapter 9 in this (e.g. a cyclic row or column effect due
558 Plant breeding and farmer participation
AMMI modelling, ANOVA, GGE biplot genetics, genomics and plant breeding, pp. 365
and pattern analysis; and (iv) ASREML 383. Wallingford, UK, CABI.
(Gilmour et al., 2006), powerful for REML Annicchiarico, P. 2007. Wide- versus specific-
and spatial analyses. adaptation strategy for lucerne breeding
GENSTAT, compared with other in northern Italy. Theoretical and Applied
powerful generic software (e.g. SAS; Genetics, 114: 647657.
S-PLUS), includes specific procedures Annicchiarico, P. & Iannucci, A. 2008.
for many analyses (e.g. joint and factorial Breeding strategy for faba bean in southern
regression; AMMI; REML; spatial trend) Europe based on cultivar responses across
and has a policy of free licensing to climatically-contrasting environments.
institutions of less developed countries. Crop Science, 48: 983991.
Sets of SAS instructions for several aspects Annicchiarico, P. & Piano, E. 2005. Use of
of GE interaction, REML and spatial artificial environments to reproduce and
analyses are reported in Kang (2003). SAS exploit genotype location interaction for
instructions for specific analyses can also be lucerne in northern Italy. Theoretical and
found in many individual papers. Applied Genetics, 110: 219227.
Annicchiarico, P., Bellah, F. & Chiari, T. 2005.
REFERENCES Defining subregions and estimating benefits
Aastveit, A.H. & Martens, H. 1986. ANOVA for a specific-adaptation strategy by breed-
interactions interpreted by partial least ing programs: a case study. Crop Science,
squares regression. Biometrics, 42: 829844. 45: 17411749.
Almekinders, C.J.M., Lowaars, N.P. & de Annicchiarico, P., Bellah, F. & Chiari, T. 2006.
Bruijn, G.H. 1994. Local seed systems and Repeatable genotype location interaction
their importance for an improved seed and its exploitation by conventional and
supply in developing countries. Euphytica, GIS-based cultivar recommendation for
78: 207216. durum wheat in Algeria. European Journal
Annicchiarico, P. 1992. Cultivar adaptation of Agronomy, 24: 7081.
and recommendation from alfalfa trials in Annicchiarico, P., Chiari, T., Bellah, F.,
northern Italy. Journal of Genetics and Doucene, S., Yallaoui-Yaci, N., Bazzani,
Breeding, 46: 269278. F., Abdellaoui, Z., Belloula, B., Bouazza, L.,
Annicchiarico, P. 1997. Joint regression vs. Bouremel, L., Hamou, M., Hazmoune, T.,
AMMI analysis of genotype-environment Kelkouli, M., Ould-Said, H. & Zerargui, H.
interactions for cereals in Italy. Euphytica, 2002. Response of durum wheat cultivars to
94: 5362. Algerian environments. I. Yield. Journal of
Annicchiarico, P. 2002a. Genotype environ- Agricultural Environment for International
ment interactions: challenges and oppor- Development, 96: 157187.
tunities for plant breeding and cultivar Annicchiarico, P., Russi, L., Piano, E. &
recommendations. FAO Plant Production Veronesi, F. 2006. Cultivar adaptation across
and Protection Paper, No. 174 (available at: Italian locations in four turfgrass species.
http://www.fao.org/docrep/005/Y4391E/ Crop Science, 46: 264272.
y4391e00.HTM; accessed 18 Jan. 2009). Atlin, G.N., Cooper, M. & Bjrnstad, 2001.
Annicchiarico, P. 2002b. Defining adaptation A comparison of formal and participatory
strategies and yield stability targets in breeding breeding approaches using selection theory.
programmes. In M.S. Kang (ed). Quantitative Euphytica, 122: 463475.
560 Plant breeding and farmer participation
Atlin, G.N., Baker, R.J., McRae, K.B. & Lu X. Ceccarelli, S. 1994. Specific adaptation and
2000. Selection response in subdivided tar- breeding for marginal conditions. Euphytica,
get regions. Crop Science, 40: 713. 77: 205219.
Bnzinger, M., Bertrn, F.J. & Lafitte, H.R. Ceccarelli, S. 1996. Positive interpretation of
1997. Efficiency of high-nitrogen selection genotype by environment interaction in
environments for improving maize for low- relation to sustainability and biodiversity.
nitrogen target environments. Crop Science, pp. 467486, in Cooper & Hammer, 1996,
37: 11031109. q.v.
Barah, B.C., Binswanger, H.P., Rana, B.S. & Ceccarelli, S., Grando, S. & Impiglia, A. 1998.
Rao, G.P. 1981. The use of risk aversion in Choice of selection strategy in breeding
plant breeding; concept and application. barley for stress environments. Euphytica,
Euphytica, 30: 451458. 103: 307318.
Basford, K.E. & Tukey J.W. 2000. Graphical Chapman, S.C., Hammer, G.L., Podlich, D.W.
Analysis of Multiresponse Data: Illustrated & Cooper, M. 2002. Linking bio-physical
with a Plant Breeding Trial. Boca Raton, and genetic models to integrate physiol-
USA, Chapman & Hall/CRC Press. ogy, molecular biology and plant breeding.
Basford, K.E., Williams, E.R., Cullis, B.R. & In M.S. Kang, ed. Quantitative genetics,
Gilmour, A. 1996. Experimental design and genomics and plant breeding, pp. 167187.
analysis of variety trials. pp. 125138, in Wallingford, UK, CABI.
Cooper & Hammer, 1996, q.v. Clawson, D.L. 1985. Harvest security and
Becker, H.C. 1984. Analysis of genotype intraspecific diversity in traditional tropical
environment interaction with partition- agriculture. Economic Botany, 39: 5667.
ing of environmental effects into effects Coe, R. 2002. Analyzing data from partici-
of locations and years. Vortrage fur patory on-farm trials. pp. 2334, in M.R.
Pflanzenzchtung, 7: 209218. Bellon & J. Reeves, eds. Quantitative anal-
Becker, H.C. & Lon, J. 1988. Stability anal- ysis of data from participatory methods
ysis in plant breeding. Plant Breeding, in plant breeding. Mexico, D.F., Mexico:
101: 123. CIMMYT.
Brancourt-Hulmel, M., Biarns-Dumoulin, V. Conway, G. 1998. The doubly green revo-
& Denis, J.B. 1997. Points de repre dans lution. Food for all in the 21th century.
lanalyse de la stabilit et de linteraction London, Penguin.
gnotype-milieu en amlioration des plants. Cooper, M. & Byth, D.E. 1996. Understanding
Agronomie, 17: 219246. plant adaptation to achieve systematic
Brownie, C., Bowman, D.T. & Burton, J.W. applied crop improvement A fundamental
1993. Estimating spatial variation in analysis challenge. pp. 523, in Cooper & Hammer,
of data from yield trials: a comparison of 1996, q.v.
methods. Agronomy Journal, 85: 12441253. Cooper, M. & Hammer, G.L. eds. 1996.
Calhoun, D.S., Gebeyehu, G., Miranda, A., Plant adaptation and crop improvement.
Rajaram, S. & van Ginkel, M. 1994. Choosing Wallingford, UK, CABI.
evaluation environments to increase wheat Cooper, M., DeLacy, I.H. & Basford, K.E.
grain yield under drought conditions. Crop 1996. Relationships among analytical meth-
Science, 34: 673678. ods used to study genotypic adaptation in
Ceccarelli, S. 1989. Wide adaptation: How multi-environment trials. pp. 193224, in
wide? Euphytica, 40: 197205. Cooper & Hammer, 1996, q.v.
Coping with and exploiting genotype-by-environment interactions 561
Cooper, M., Woodruff, D.R., Eisemann, R.L., spective. pp. 39124, in Cooper & Hammer,
Brennan, P.S. & DeLacy, I.H. 1995. A selec- 1996, q.v.
tion strategy to accommodate genotype-by- DeLacy, I.H., Basford, K.E., Cooper, M. &
environment interaction for grain yield of Fox, P.N. 1996b. Retrospective analysis of
wheat: managed-environments for selection historical data sets from multi-environment
among genotypes. Theoretical and Applied trials Theoretical development. pp. 243
Genetics, 90: 492502. 267, in Cooper & Hammer, 1996, q.v.
Cooper, M., Stucker, R.E., DeLacy, I.H. & Denis, J.B. 1988. Two-way analysis using cov-
Harch, B.D. 1997. Wheat breeding nurser- ariates. Statistics, 19: 123132.
ies, target environments, and indirect selec- Eberhart S.A. & Russell W.A. 1966. Stability
tion for grain yield. Crop Science, 37: 1168 parameters for comparing varieties. Crop
1176. Science 6: 3640.
Cooper, M., Rajatasereekul, S., Somrith, Engledow, F.L. 1925. The economic possi-
B., Sriwisut S., Immark, S., Boonwite, bilities of plant breeding. pp. 3140, in Proc.
C., Suwanwongse, A., Ruangsook, S., Imperial Botanical Conference. London,
Hanviriyapant, P., Romyen, P., Porn- F.T. Brooks.
uraisanit, P., Skulkhu, E., Fukai, S., Eskridge, K.M. 1990. Selection of stable culti-
Basnayake, J. & Podlich, D.W. 1999. Rainfed vars using a safety-first rule. Crop Science,
lowland rice breeding strategies for north- 30: 369374.
east Thailand. II. Comparison of intrasta- Evans, L.T. 1998. Feeding the ten billion:
tion and interstation selection. Field Crops plants and population growth. New York,
Research, 64: 153176. USA, Cambridge University Press.
Cornelius, P.L. & Crossa, J. 1999. Prediction Falconer, D.S. 1952. The problem of environ-
assessment of shrinkage estimators of mul- ment and selection. American Naturalist,
tiplicative models for multi-environment 86: 293298.
trials. Crop Science, 39: 9981009. Federer, W.T. & Scully, B.T. 1993. A parsimoni-
Crossa, J., Fox, P.N., Pfeiffer, W.H., Rajaram, ous statistical design and breeding procedure
S. & Gauch, H.G. 1991. AMMI adjustment for evaluating and selecting desirable charac-
for statistical analysis of an international teristics over environments. Theoretical and
wheat yield trial. Theoretical and Applied Applied Genetics, 86: 612620.
Genetics, 81: 2737. Federer, W.T., Nair, R.C. & Raghavarao,
Cullis, B.R. & Gleeson, A.C. 1991. Spatial anal- D. 1975. Some augmented row-column
ysis of field experiments an extension to designs. Biometrics 31: 361374.
two directions. Biometrics, 47: 14491460. Finlay, K.W. & Wilkinson, G.N. 1963. The anal-
Dagnelie, P. 1975. Thorie et mthodes statis- ysis of adaptation in a plant-breeding pro-
tiques. Vol. 2. 2nd ed. Gembloux, Belgium, gramme. Australian Journal of Agricultural
Les Presses Agronomiques. Research, 14: 742754.
Decoux, G. & Denis, J.B. 1991. INTERA. Fischer, G., Shah, M. & van Velthuizen, H.
Logiciels pour linterpretation statistique de 2002. Climate change and agricultural vul-
linteraction entre deux facteurs. Versailles, nerability. Vienna, IIASA.
France: Biomtrie INRA. Fox, P.N. & Rosielle, A.A. 1982. Reference
DeLacy, I.H., Basford, K.E., Cooper, M., Bull, sets of genotypes and selection for yield in
J.K. & McLaren, C.G. 1996a. Analysis of unpredictable environments. Crop Science,
multi-environment data An historical per- 22: 11711175.
562 Plant breeding and farmer participation
Gauch, H.G. 1992. Statistical analysis of Kang, M.S. ed. 2002. Quantitative genetics,
regional yield trials: AMMI analysis of facto- genomics and plant breeding. Wallingford,
rial designs. Amsterdam, The Netherlands: UK, CABI.
Elsevier. Kang, M.S. ed. 2003. Handbook of formulas
Gauch, H.G. 2007. MATMODEL Version 3.0: and software for plant geneticists and breed-
Open source software for AMMI and related ers. New York, USA, Haworth Press.
analyses. Ithaca, USA, Cornell University. Kang, M.S. & Gauch, H.G. eds. 1996.
Downloadable from: http://www.css.cor- Genotype-by-environment interaction.
nell.edu/staff/gauch. Boca Raton, USA, CRC Press.
Gauch, H.G. & Zobel, R.W. 1997. Identifying Kataoka, S. 1963. A stochastic programming
mega-environments and targeting geno- model. Econometrika, 31: 181196.
types. Crop Science, 37: 311326. Kempton, R.A. 1984. The use of biplots in
Gauch, H.G., Piepho, H.P. & Annicchiarico, interpreting variety by environment inter-
P. 2008. Statistical analysis of yield trials by actions. Journal of Agricultural Science,
AMMI and GGE: further considerations. 103: 123135.
Crop Science, 48: 866889. Lin, C.S. & Binns, M.R. 1988. A method for
Gilmour, A.R., Cullis, B.R. & Verbyla, A.P. analyzing cultivar location year experi-
1997. Accounting for natural and extrane- ments: a new stability parameter. Theoretical
ous variation in the analysis of field experi- and Applied Genetics, 76: 425430.
ments. Journal of Agricultural, Biological Lin, C.S. & Butler, G. 1988. A data-based
and Environmental Statistics, 2: 269273. approach for selecting locations for regional
Gilmour, A.R., Gogel, B.J., Cullis, B.R., trials. Canadian Journal of Plant Science,
Welham, S.J. & Thomson, R. 2006. ASReml 68: 651659.
User Guide Release 2.0. Hemel Hempstead, Lin, C.S. & Poushinsky, G. 1983. A modi-
UK, VSN International. fied augmented design for an early stage of
Grneberg, W.J., Abidin, E., Ndolo, P., Pereira, plant selection involving a large number of
C.A. & Hermann, M. 2004. Variance compo- test lines without replication. Biometrics,
nent estimation and allocation of resources 39: 553561.
for breeding sweet potato under East African Lin, C.S., Binns, M.R. & Lefkovitch, L.P. 1986.
conditions. Plant Breeding, 123: 311315. Stability analysis: where do we stand? Crop
Hardwick, R.C. & Wood, J.T. 1972. Regression Science, 26: 894900.
methods for studying genotype-environ- Ludlow, M.M. & Muchow, R.C. 1990. A
ment interaction. Heredity, 28: 209222. critical evaluation of traits for improving
Hill, R.R. & Rosenberger, J.L. 1985. Methods crop yields in water-limited environments.
for combining data from germplasm evalua- Advances in Agronomy, 43: 107153.
tion trials. Crop Science, 25: 467470. Patterson, H.D. 1997. Analysis of series of
IRRI [International Rice Research Institute]. variety trials. In R.A. Kempton & P.N.
2007. CROPSTAT Version 6.1. Los Baos, Fox, eds. Statistical methods for plant vari-
Philippines, IRRI. Downloadable from: ety evaluation, pp. 139161. London,
http://www.irri.org/science/software/crop- Chapman & Hall.
stat.asp. Patterson, H.D., Williams, E.R. & Hunter,
Kang, M.S. 1998. Using genotype-by-environ- E.A. 1978. Block designs for variety trials.
ment interaction for crop cultivar develop- Journal of Agricultural Science, 90: 395
ment. Advances in Agronomy, 62: 199252. 400.
Coping with and exploiting genotype-by-environment interactions 563
Perkins, J.M. & Jinks, J.L. 1968. Environmental Romagosa, I., Feng, H., Ullrich, S.E., Hayes,
and genotype-environmental components P.M. & Wesenberg, D.M. 1999. Verification
of variability. III. Multiple lines and crosses. of yield QTL through realized molecular
Heredity, 23: 339356. marker-assisted selection responses in a
Piepho, H.P. 1995. Robustness of statistical barley cross. Molecular Breeding, 5: 143
tests for multiplicative terms in the additive 152.
main effects and multiplicative interaction Simmonds, N.W. 1979. Principles of crop
model for cultivar trials. Theoretical and improvement. Harlow, UK, Longman.
Applied Genetics, 90: 438443. Simmonds, N.W. 1991. Selection for local
Piepho, H.P. 1998. Methods for comparing adaptation in a plant breeding programme.
the yield stability of cropping systems - A Theoretical and Applied Genetics, 82: 363
review. Journal of Agronomy and Crop 367.
Science, 180: 193213. Singh, M., Ceccarelli, S. & Grando, S. 1999.
Piepho, H.P. & Mhring, J. 2005. Best linear Genotype environment interaction of
unbiased prediction of cultivar effects for cross-over type: detecting its presence and
subdivided target regions. Crop Science, estimating the crossover point. Theoretical
45: 11511159. and Applied Genetics, 99: 988995.
Piepho, H.P., Denis, J.B. & van Eeuwijk, Singh, M., Malhotra, R.S., Ceccarelli, S.,
F.A. 1998. Predicting cultivar differences Sarker, A., Grando, S. & Erskine, W. 2003.
using covariates. Journal of Agricultural, Spatial variability models to improve dry-
Biological and Environmental Statistics, land field trials. Experimental Agriculture,
3: 151162. 39: 110.
Podlich, D.W., Cooper, M. & Basford, K.E. 1999. Singh, S.P., Gutierrez, J.A., Urrea, C.A., Molina,
Computer simulation of a selection strat- A. & Cajiao, C. 1992. Location-specific and
egy to accommodate genotype-environment across-location selections for seed yield in
interactions in a wheat recurrent selection populations of common bean, Phaseolus
programme. Plant Breeding, 118: 1728. vulgaris L. Plant Breeding, 109: 320328.
Qiao, C.G., Basford, K.E., DeLacy, I.H. & Smith, A., Cullis, B. & Thomson, R. 2001.
Cooper, M. 2004. Advantage of single- Analyzing variety by environment data
trial models for response to selection in using multiplicative mixed models
wheat breeding multi-environment tri- and adjustments for spatial field trend.
als. Theoretical and Applied Genetics, Biometrics, 57: 11381147.
108: 12561264. Smith, A., Cullis, B., Luckett, D., Hollamby, G. &
Ribaut, J.M., Fracheboud, Y., Monneveux, Thomson, R. 2002. Exploring variety-environ-
P., Banzinger, M., Vargas, M. & Jiang C. ment data using random effects AMMI models
2007. Quantitative trait loci for yield and with adjustments for spatial trends. Part 2:
correlated traits under high and low nitro- Applications. In M.S. Kang, ed. Quantitative
gen conditions in tropical maize. Molecular genetics, genomics and plant breeding, pp. 337
Breeding, 20: 1529. 351. Wallingford, UK, CABI.
Romagosa, I., Ullrich, S.E., Feng, H. & Hayes, van Oosterom, E.J., Bidinger, F.R. & Weltzien,
P.M. 1996. Use of additive main effects and E. 2003. A yield architecture framework
multiplicative interaction in QTL mapping to explain adaptation of pearl millet to
for adaptation in barley. Theoretical and environmental stress. Field Crops Research,
Applied Genetics, 93: 3037. 80: 3356.
564 Plant breeding and farmer participation
vom Brocke, K., Weltzien, E., Christinck, A., paradigms about photoperiod and tempera-
Presterl, T. & Geiger, H.H. 2003. Effects of ture control of crop yield. Theoretical and
farmers seed management on performance Applied Genetics, 86: 1726.
and adaptation of pearl millet in Rajasthan, Witcombe, J.T. 2001. The impact of decentral-
India. Euphytica 130: 267280. ized and participatory plant breeding on
Vargas, M., Crossa, J., van Eeuwijk, F. A., the genetic base of crops. In H.D. Cooper,
Ramrez, M.E. & Sayre, K. 1999. Using par- C. Spillane & T. Hodgkin, eds. Broadening
tial least squares regression, factorial regres- the genetic base of crop production, pp. 407
sion, and AMMI models for interpreting 401. Wallingford, UK, CABI.
genotype environment interaction. Crop Wrike, G. & Weber, W.E. 1986. Quantitative
Science, 39: 955967. genetics and selection in plant breeding.
Vargas, M., van Eeuwijk, F.A., Crossa, J. & Berlin: W. de Gruyter.
Ribaut, J.M. 2006. Mapping QTL and QTL Yan, W. & Kang, M.S. 2003. GGE biplot anal-
environment interaction for CIMMYT ysis: a graphical tool for breeders, geneti-
maize drought stress program using fac- cists, and agronomists. Boca Raton, USA,
torial regression and partial least squares CRC Press.
methods. Theoretical and Applied Genetics, Yates, F. & Cochran, W.G. 1938. The analysis
112: 10091023. of groust DTPps of experiments. Journal of
Wallace, D.H., Zobel, R.W. & Yourstone, K.S. Agricultural Science, 28: 556580.
1993. A whole-system reconsideration of
565
CHAPTER 21
21.1 HISTORICAL DEVELOPMENTS AND trading (informal sector). In recent times the
CONTEXTS concept of seed system has been broadened
Empirical evidence shows that for millennia to include the role of the informal sector
farmers selected plants from their local in seed provision. Van Amstel et al. (1996),
landraces and saved their own seed for apart from providing a comprehensive
planting. In the 1880s, early attempts in definition of the seed system, recognize two
scientific plant breeding began and the distinctive, but interacting seed delivery
first agricultural research stations were systems: the formal and the informal sectors.
established in some European countries A farmer may have adopted a modern
(Khre, 1990). The history of the organized variety from the formal sector, but may
seed sector is linked to the introduction decide to save seed from their own harvest
of new crops and knowledge-based or exchange through social networks for the
agriculture, including scientific plant next seasons planting: seed that is produced
breeding, mechanization, commercialization informally (Bishaw, 2004).
and diversification at various stages of
agricultural development (Tripp, 2001; Formal seed system
Thomson, 1979). Continued specialization The formal seed system is composed of several
eventually brought significant changes in interrelated components, namely: (i) variety
seed provision, giving birth to an integrated development, evaluation, registration and
and market-oriented organized seed sector release; (ii) seed production, processing and
in developed countries (Groosman, 1987). storage; (iii) seed marketing and distribution;
In many developing countries, however, and (iv) seed quality assurance. It is a highly
information on the history of agricultural interdependent chain of operations whose
research and organized seed production overall performance can be measured by
prior to 1950s is rather scanty. The the efficiency of the different linkages in
introduction of highly productive semi- the chain (Pray and Ramasawmi, 1991). In
dwarf wheat and rice cultivars in the late general it is a vertically organized (Louwaars,
1960s and 1970s, which triggered what 2002), large-scale operation, mostly with
is referred to as the Green Revolution, commercial interests.
probably served as a stimulus for introducing
agricultural research and the establishment Informal seed system
of organized seed production throughout At present, in developing countries, over 80
the developing countries, particularly for percent of crops are sown from seed stocks
economically important and strategic food selected and saved by farmers or exchanged
crops. From the outset the seed system and traded locally (Almekinders, Louwaars
was inherently service-oriented, with no or and de Bruijin, 1994; Alemkinders and
limited commercial interests. Louwaars, 1999). The informal seed system
operates at local level (Cromwell, Friis-
21.1.1 Seed system definitions Hansen and Turner, 1992), and may depend
The entire seed supply of a country comes on indigenous knowledge of plant and seed
from different sources, including off- selection, sourcing, retention, management
farm from commercial sources such as the and local diffusion mechanisms (Bishaw,
public or private sector (formal sector) or 2004). Apart from farmer or community
farm saved or through local exchange and practices it also includes various local-
Variety release and policy options 567
maintaining the identity and purity of new establish its identity (registration testing)
varieties during seed production and supply. and its commercial value for cultivation
According to Tripp (2001), listing varieties (farmers) and use (consumers) (perform-
based on morphological characteristics and ance testing) are the basis for final reelase.
performance was started as early as 1905 (in The ability to discriminate and
Germany), whereas seed certification was identify varieties of agricultural and
started in 1888 (in Sweden). horticultural crops is fundamental in the
The establishment of the International modern seed trade. Variety description
Crop Improvement Association (now the is essential for effective implementation
Association of Official Seed Certifying of: (i) variety maintenance (purification);
Agencies (AOSCA; www.aosca.org) in 1919 (ii) seed multiplication (roguing); (iii) seed
(Parsons, 1985) and the Organisation for certification (field inspection); (iv) granting
Economic Co-operation and Development intellectual property rights (plant variety
(OECD; www.oecd.org) seed schemes in protection); and (iv) protection of producers
1958 (Thomson, 1979) were some of the and consumers (seed certification).
first attempts to standardize variety release The degree to which the breeders are
and seed certification schemes. These involved and the way release procedures
organizations put in place evaluation, are organized and conducted is described
registration and release procedures for as a compulsory (e.g. European Union) or
accepting and listing eligible varieties, and a voluntary (e.g. United States of America)
strict generation control to maintain the release system.
varietal purity and identity by establishing
standardized certification schemes for Variety registration testing
commercial seed production (OECD, Variety registration (DUS) testing is a
2007). Likewise, many governments enacted descriptive assessment to establish the iden-
national variety and seed regulations to tity of the new variety using morphological
implement such types of schemes. Despite a characters, as well as its sufficient uniform-
long history of organized seed sectors, many ity and stability. DUS testing usually runs
developed countries enacted comprehensive for two independent growing seasons or
variety and seed regulations only fairly years, where the new variety is compared
recently. with a wide range of existing varieties to
establish its identity. A detailed DUS test-
21.2.4 Current procedures and ing procedures and crop specific guidelines
practices are available from the International Union
Once new and potential promising lines for the Protection of New Varieties of
are identified by agricultural research, it is Plants (UPOV; http://www.upov.int/en/
essential to commercialize and make their publications/tg_rom/tg_index.html).
seed available to farmers. Variety release Distinctness: A new variety must be
requires simultaneous testing of these different from existing varieties and must
promising lines for registration (DUS) and be recognizable to verify its identity and
performance (VCU) before approval for purity during seed production and use.
large-scale seed multiplication and market- Distinctness refers to a difference from
ing for commercial use. The distinctness any other variety whose existence is a
(uniformity and stability) of the variety to matter of common knowledge.
Variety release and policy options 569
Uniformity: It relates to the degree where the best performing varieties are
of variability within the variety. The eventually recommended for cultivation. In
variation observed must normally be of some countries (e.g. Ethiopia) the variety is
a demonstrable and repeatable order. tested in on-farm verification trials under
The variety must be sufficiently uniform farmer management practices before the
within its population so that individual final release.
plants could be identified to guarantee
constant quality. Variety release
Stability: It refers to the capacity of the Variety release is a culmination of several
variety to reproduce itself during seed interrelated activities, where a decision is
production without losing its distinctive taken to approve a new variety for com-
characters. The genetic make-up of the mercial use, based on the results of registra-
variety should remain the same during tion and performance testing. Almost all
subsequent years of seed production and countries have a variety release procedure
commercialization. in place, whether that is done by an ad hoc
Varietal Identity: The identity of a new committee (e.g. the Syrian Arab Republic)
variety is established by examining and or by a legally sanctioned authority (e.g.
describing the morphological characters Turkey). The varieties that meet the require-
of growing plants. The purpose of ments for registration and performance are
registration testing, whether backed by officially released and the owner makes
legislation or not, is the recognition of breeder seed available for commercial seed
varietal unit as a unique entity and to production and marketing. In some coun-
establish its identity. tries, however registration testing (DUS) is
not yet everywhere part of the requirement
Variety performance testing for variety release (e.g. Ethiopia).
Performance (VCU) testing, referred to
as variety trials, focuses on the merit Variety registers
of the new variety to the end users, i.e. The new variety, upon approval, will be
producers and consumers. The test ensures listed in a variety register to inform the
that only varieties that are found better stakeholders, i.e. seed producers, farmers,
than the existing varieties in one or more consumers and the industry. The list
agronomic character, such as grain yield could be informative or recommendatory.
or quality, or tolerance to biotic or abiotic The register is periodically updated by
stresses, are released for use by farmers. The removing obsolete and entering new
multi-location and multi-year variety trials varieties that are eligible for commercial
are conducted in different agro-ecological seed production at national or provincial
zones to identify better performing varieties, levels (e.g. Pakistan). Many countries
which could meet diverse agronomic or have a national variety register (e.g.
consumer requirements or socio-economic Crop Variety Register in Ethiopia), while
conditions. As a result, different agronomic OECD has a common variety catalogue
management practices are used and the (www.oecd.org), which enables the
new varieties are compared with well variety to be produced and marketed in
established commercial varieties. VCU all member countries.
tests usually run for two to three years,
570 Plant breeding and farmer participation
farmers. They directly channel Breeder seed (1995) argues that regulatory reforms must
of new varieties to public companies for be seen as a continuing process, and must
further seed production and distribution. be sufficiently flexible to respond to and
The access of the private sector to public promote the evolution and diversification
bred varieties remains a major stumbling of the national seed sector in developing
block in many developing countries. This is countries.
particularly important for domestic, small, In general, the majority of developing
private seed companies and small-scale countries lack well established variety
seed enterprises operating at local level release protocols and procedures in place.
serving farmers in less favourable areas, but The level of regulation is commonly
which rely on varieties from public-sector not in line with the level of institutional
sources. Such seed companies have neither development of the country, leading to
the resources nor the technical capacity incomplete implementation and insufficient
to engage in plant breeding programmes. transparency. This creates a serious lack
It is essential to have a transparent and of credibility and inconsistent application
equitable mechanism to guarantee access of these regulations by the authorities.
to public-bred varieties, as discussed under Tripp et al. (1997) identified four key
the variety release options. constraints that need to be addressed in
regulatory reform, namely: (i) the efficiency
21.3.2 Regulatory reforms of operation; (ii) application of objective
Variety development, evaluation and release and relevant standards; (iii) participation
are closely interconnected, and it would be of stakeholders; and (iv) transparency in
difficult to draw distinctions between the managing registration and performance
regulatory frameworks that govern these testing for variety release.
as two separate activities. Accordingly,
variety regulatory frameworks are the rules Harmonization of variety regulations
and regulations associated with variety The regulatory requirements governing
development, testing, registration and registration and performance testing
release (Tripp, 1997). In effect, it includes are critical elements in variety release
the procedures and practices that guide mechanisms. In the past, where these
the conduct of plant breeding; the rules regulations existed, they were prepared
governing the official release of new and implemented within their specific
varieties; and restrictions on the type of national context. Some countries have
varieties that may be offered for sale. comprehensive variety regulations, whereas
Tripp et al (1997) described the key others still have no or outdated legislation,
features and limitations of variety and which consequently do not meet the
seed regulations, and their introduction needs of a modern seed industry. With
to developing countries. Most of these globalization, these inflexible regulations
regulations are modelled upon and influenced are a serious impediment to movement of
by past historical relationships and source of varieties across national boundaries, thus
donor support to national seed programme severely limiting opportunities in regional
development. They are at times excessively and global seed trade. Harmonized variety
strict and inflexible, limiting the range of regulations (e.g. East African Community)
varieties and seeds available to farmers. Tripp would increase the choice and movement of
574 Plant breeding and farmer participation
varieties and seeds across national borders there is still lack of conclusive evidence
and stimulate regional seed trade. Given the on its impact on the commercial plant
diversity of national seed systems and the breeding industry (Morris and Ekasingh,
globalization of the seed trade, it would be 2002), on diversity of varieties in farmers
appropriate to develop a variety regulation fields (Fischer and Byerlee, 2002), and as
and release procedure that is both flexible a precondition for the development of the
nationally and harmonized regionally. private sector (Louwaars et al., 2005).
as cooking quality, taste, marketability have appeared to be a bottleneck for the flow
and storability under traditional farming of modern crop varieties from agricultural
systems. Moreover, the strategies tend to research to farmers. The major criticism of
favour wide adaptability and selection of the variety release system is the stringent
material under favourable crop management, requirement and application of detailed
where both the environment and the trial DUS and narrowly defined VCU criteria
management are unrepresentative of actual implemented by public-sector agencies.
farmer circumstances. In the end, only a For varieties produced by conventional
few average best varieties (Alemkinders plant breeding, it is important to develop
and Louwaars, 1999) become available from clear, simple and flexible registration and
public plant breeding, which too often will performance testing systems based on
subsequently be produced and distributed by criteria developed by all stakeholders. It is
inefficient public seed companies, limiting important that for variety registration some
the choice of varieties and availability of key descriptors are identified and used to
seed to farmers. help seed producers and certification agencies
In contrast, farmers in diverse, complex, recognize varieties, instead of detailed
dry and risky environments are interested examination and recording using a large
in varieties with a broad genetic base with number of morphological characteristics.
the capacity for individual and population Although conventional regulation might
buffering in stress environments, such as intend to abandon variety registration, the
heterogeneous populations. Consequently, introduction of variety protection laws does
several varieties with specific adaptation are require a detailed registration of the protected
preferred over a few varieties with wide adap- entity. Tripp et al (1997) suggested that
tation. Although selection is more effective requirements for conventional registration
in the target environment, marginal environ- and for granting PVP be treated differently
ments are inadequately represented in nation- and handled by separate agencies to minimize
al breeding programmes, or even ignored. complications. Similarly, they suggested that
In many developing countries, the the performance testing should reflect the
responsibility for variety development circumstances and preferences of farmers,
rests with public NARS. Therefore, the where suitability instead of superior yield is
NARS plant breeding strategy, protocols used as the criterion for variety release.
and procedures have greater influence on
the type and number of varieties available Prolonged testing in variety release
for release. There are serious concerns Developing a new crop variety to enter
regarding efficiency and effectiveness in the testing for release may require more than 10
variety development process, the criteria years. Thereafter, the variety must also pass
used in evaluating breeding materials, the through series of preliminary yield trials
degree of stakeholder involvement and the and meet the requirements for registration
transparency of the system (Witcombe and and performance testing for official release.
Virk, 1997). It may take another two to six years before
the variety is finally approved for release.
Criteria for variety release testing Seed production can only be initiated
In many developing countries, registration following the official release, meaning
and performance testing for varietal release that it might require an additional five
576 Plant breeding and farmer participation
or more years for a sufficient quantity of flexible and harmonized variety release sys-
commercial seed to reach the farmers. In tems in regional or international contexts.
general, there is considerable delay and cost Harmonization of release procedures
in the process from variety development are naturally an extension of harmonized
through to its release and availability of variety regulation. As discussed earlier
seed for commercial use, which is a very under regulatory reforms, collaboration
lengthy process. The variety development among countries and sharing of data could
and release process may take up to 15 years enable much quicker decisions on variety
(e.g. Uzbekistan) from the initial crossing release. Using data from other countries
nursery to the end of official state variety to reduce the number of years or to waive
testing for variety release. requirements based on data submitted from
In the public sector, variety development, tests carried out elsewhere is critical. In
variety release and seed production and recent years, for example, efforts have been
marketing are conducted by different underway to harmonize variety release
institutions, which are not properly linked procedures to integrate national seed systems
and this may exacerbate the problem and to attract foreign investment and promote
prolong the period compared with the regional trade as part of harmonized variety
private sector, where these activities are and seed regulation (e.g. Community of
integrated. Seed production and marketing Andean Nations). Regional harmonization
start only after the official release, and there of technical aspects of variety release systems
is no inbuilt mechanism for pre-release seed may reduce cost, save time, encourage private
multiplication of public-bred varieties to investment, increase choice of varieties and
expedite the availability of seeds, with a benefit farmers.
few exceptions (e.g. Ethiopia, Uzbekistan
and Zambia). Managing registration and performance
testing
Harmonizing testing and release In principle, testing for variety registration
procedures and performance should be managed by
The UPOV protocols are becoming inter- an independent and impartial agency
nationally accepted standards in DUS test- established for the purpose. In reality, such
ing for variety description, registration and agencies vary from country to country, and
protection. Under the UPOV convention, the responsibility may be vested in a single
to maximize use of available information agency or two different institutions. In some
and minimize the time for examination, cases it is possible to make decisions based
there is cooperation among countries or on tests carried out by the breeder, but at
authorities, where some institutions have the discretion of the agency. For example,
been identified and specialized in DUS in Ethiopia, Algeria and Jordan, for release
testing for specific crops. The EU is a their National Variety Release Committees
good example of a regionally harmonized depend on performance testing data supplied
release system, where varieties released in by the breeder and on-farm verification trials
one country are acceptable in all member conducted by the breeder, but reviewed by a
countries. Such an approach provides great special technical committee.
opportunity for developing testing proto- In many developing countries, how-
cols and sharing data, as well as establishing ever, there is lack of an impartial authority
Variety release and policy options 577
FIGURE 21.1
Linkages between participatory plant breeding and seed supply
Marketing
Production
Farmers and
certication
Local
seed
Local
production
germplasm
pool
(in-situ)
Variety
maintenance
Participatory
plant
breeding
Genebanks
(ex-situ)
Variety
Formal
release
plant
Variety
breeding
evaluation
increasing diversity at farm level, shortening need for maintaining identity and integrity
the breeding cycle, identifying well-adapted of participatory varieties, applying flexible
and acceptable varieties, quicker availability variety release procedures, and establishing
of varieties and seeds, empowering the alternative seed delivery systems. Some of
farmers and lowering overall breeding these options are presented and discussed
programme costs have been discussed by in more depth below.
several authors (e.g. Staphit et al., 1996;
Witcombe et al., 1996, 1999; Mangione, Institutionalizing participatory plant
Ceccarelli and Grando, 2006; Ceccarelli and breeding
Grando, 2007). Bishaw and Turner (2008) Participatory approaches are evolving
discussed the potential linkages between and still lack clearly defined procedures
PPB and seed supply systems to exploit compared with long-established formal
farmers knowledge in crop improvement breeding programmes. At least two major
and ensure rapid adoption and diffusion forms of participatory approaches have
of varieties (Figure 21.1). They advocated been recognized: PVS and PPB, the latter
national policies that recognize the role with many variant forms (breeder/farmer-
of PPB and support strategies to release, led PPB, decentralized PPB, highly-client-
produce and market varieties developed oriented plant breeding, etc.) and some
through these approaches. They also noted differences in methodological approaches,
critical issues to be addressed for the PPB type of breeding materials, and stage
approach to function properly such as the and degree of involvement by farmers.
580 Plant breeding and farmer participation
Some successes have been reported with breeder, or more likely, in farmer groups
participatory approaches, including PVS established to produce and market the seed.
and PPB in recent years. However, PVS This provides a basis for some continuity of
appears less problematic as it deals with pure seed supply and to maintain the iden-
already released or nearly finished varieties tity of the material. In the absence of such
to derive farmers varietal choices. At the arrangement, the purity and identity of the
same time, an attempt to institutionalize variety may dilute and diffuse over years.
PPB in NARS breeding programmes in its
own right is still under debate and its future Flexibility in varietal release
remains uncertain. Were PPB officially The disadvantages of formal variety
recognized and institutionalized, the issue release procedures are discussed elsewhere.
of variety release and its commercialization However, in PVS, a limited choice of
would have long been resolved at the finished or nearly finished varieties
national level. Therefore, outcomes from bred through conventional or other
PPB need to be documented and its impacts means are exposed to numerous groups of
demonstrated to influence national polices farmers across villages in widely dispersed
to overcome the hurdle. geographical areas for farmers to select
according to their preferences. In reality,
Maintaining varietal identity and PVS is closer to conventional breeding as
integrity it involves farmers only towards the end of
There are two key factors for adoption the selection programme. Therefore, PVS
and diffusion of a variety: (i) genetic integ- presents less challenge compared with PPB
rity (the inherent capacity of the variety in variety release systems, particularly if the
to reproduce itself during seed multiplica- varieties used are from conventional plant
tion); and (ii) varietal identity (its unique breeding programmes.
distinguishing characteristics established In PPB, a few representative farmers
during its development). CPB generates are involved in selecting varieties from
defined outputs (cultivated varieties) with large segregating populations. It is believed
the responsibility for maintaining the vari- that the PPB approach gives greater
ety (identity and integrity) vested in the opportunity because of wide dispersion of
breeder, or a designated agent, and ensuring sites that reflect the actual environments
a continuing source of pure material as long of crop production. Consequently, PPB
as it remains in commercial seed produc- should encourage the use of more adapted
tion. This system of variety management is material, and development of many varieties
absolutely critical in formal systems, since with specific adaptation, particularly to less
it provides a secure point of reference and, favourable environments. This may increase
by limiting the number of generations, it farmer choice, but it may create challenges
also reduces the risk of contamination. for the formal variety release system, and
Similarly, it is therefore highly desirable ultimately for seed production as well.
that the identity and integrity of PPB varie- Varieties developed through PPB do not
ties are systematically maintained and made always meet the stringent DUS and VCU
available to more farmers. To achieve this, criteria because they may lack sufficient
responsibility should be vested either in uniformity and might not always perform
the formal sector, in an individual farmer- well across the majority of test sites compared
Variety release and policy options 581
with varieties from conventional plant et al., 1996). Ceccarelli and Grando (2007)
breeding. Applying identical testing criteria outlined the PPB model for barley, where
would be unrealistic, and alternative variety early generation materials (farmers involved
release procedures must be considered. The from F3 bulk) go through four cycles of
criteria for registration and performance selection, when farmers are involved in
should be flexible and accommodate less selecting and testing the materials during
uniformity within a variety and allow a wide the subsequent years. Farmer-selected
range of varieties with specific adaptation varieties in large-scale trials (fourth cycle)
to increase the choice of niche varieties are considered adopted and should be
available to farmers. Possible scenarios for released. Alternatively, they suggested that
release of materials from PPB are considerd testing pure line (pedigree) selection from
below (Bishaw and Turner, 2008). selected bulks can be conducted on-station
and released in situations where there are
Linking to formal plant breeding stringent variety release requirements.
Conventional plant breeding exploits For PPB varieties, any detailed examina-
indigenous knowledge by involving farmers tions for VCU appeared to be redundant
at different stages in the selection process. since farmers are already part of the selection
The materials identified or selected by process and identified those meeting their
farmers can be further refined and the preferences. Some countries also release
varieties ultimately evaluated and released varieties purely based on performance test-
through the official process and the seed ing where DUS requirement would not be
become available through the formal problematic if farmers criteria are accepted
sector. Sthapit, Joshi and Witcombe (1996) (e.g. Ethiopia). The alternative approach is
described where PPB products entered to release PPB products through a separate
national trials using scientist-led breeding registration system established to accom-
schemes run in parallel with those of the modate these varieties, or even make an
farmers, with the main purpose being to outright decision to release them without
purify the variety and select for uniformity testing (e.g. Jordan). However, even if the
to meet criteria for formal release. Belay DUS criteria are relaxed, some level of
et al. (2008) demonstrated where both description is essential to identify the vari-
conventional and participatory approaches ety for purposes of seed production and
were used in a complementary mode for marketing through formal channels.
official release of a variety in Ethiopia.
Linking to formal seed supply
Linking to formal variety release PPB varieties could be exempted from
PPB products identified by farmers can release systems and directly enter seed pro-
directly enter official variety release and duction. The formal sector may take direct
registration trials, but they may encounter responsibility for large-scale seed multipli-
difficulties in terms of either uniformity or cation and marketing of PPB varieties iden-
performance for reasons already explained. tified by farmers. Given the fact that PPB
It is suggested that release committees varieties may have a larger recommendation
accept PPB data on farmer perceptions domain beyond the initial testing sites, it is
and demand for seed rather than yield data suggested that large-scale seed production
from scientist-managed trials (Witcombe and distribution and external intervention
582 Plant breeding and farmer participation
be used as a strategy to accelerate diffusion agencies, and linking them to the formal
(Joshi, Sthapit and Witcombe, 2001). sector. For local initiatives to succeed they
Alternatively, the formal sector could must address the key issues of financial
limit itself to variety maintenance and viability and sustainability without external
Breeder seed production, in order to ensure support (Bishaw and van Gastel, 2008).
small but secure supplies to local seed
producers. Virk et al. (2003) emphasized Protecting PPB varieties
the importance of formal sector (research, In the last decade or so, access to genetic
universities, department of agriculture), resources and protection of plant varieties
NGOs and self-help groups, supported by has become an important part of an
government, to ensure seed production and increasingly intense debate in formulating
dissemination once farmer-preferred varie- policy and regulatory framework at national
ties have been identified through a par- and international levels (see also Chapters
ticipatory approach. Ceccarelli and Grando 23 and 24 in this volume). Chapter 23
(2007) advocate the need for linking PPB argues for Farmers Rights (FRs) under
varieties to formal and informal channels to International Treaty, whereas Chapter 24
ensure adoption and realize impact. proposes Plant Breeders Rights (PBRs)
under national IPR laws. In reality there is
Linking to local seed supply no contradiction between the two as they
At present there is limited information address two separate issues. However, quite
on scaling up seed production to diffuse often there is confusion, mixing farmers
PPB varieties. Despite apparent strengths, rights with breeders rights. In Ethiopia,
local seed systems may not adequately the government has enacted two separate
meet requirements for wider distribution regulations for plant breeders rights and
of PPB varieties unless they are properly community farmers rights,
strengthened and linked. Almekinders, There are many forms of IPR protection
Thiele and Danial (2007) consider that for plant varieties through biological (e.g.
there is a tendency to overestimate the role hybrids) or legal control including trade
of informal farmer-to-farmer seed exchange secrets, contracts and licenses, patents,
as a diffusion mechanism for PPB varieties. and PVP laws. Among these patents for
In India, a follow-up study for a rice asexually propagated materials (since
variety identified by PVS showed seed the 1930s) and latterly for genes, gene
diffusion from farms to relatives or friends combinations and biotechnology products,
in adjacent or nearby villages typically while PVP laws for plant varieties have
over distances of less than 10 km, despite been long established as IPR protection
project intervention in providing seed systems in the field of agriculture.
through village seed pools, seed merchants First, in conventional plant breeding,
and NGOs (Witcombe et al., 1999). Some describing a variety using morphological
institutionalization of local seed systems characters and establishing its identity is
is, however, necessary by involving, for a prerequisite both for release and for
example, existing community groups, protection. For example, under the UPOV
farmers cooperatives or associations, convention, granting PVP is based on DUS
local traders and entrepreneurs, NGOs, and novelty of the variety. Theoretically,
extension services or rural development if PPB varieties meet these criteria it is
Variety release and policy options 583
assumed they could be granted immediate the criteria of protecting PPB varieties?
protection under national PVP law. At the Fourth, the purpose of PPB is to circum-
same time, it is argued that PPB varieties vent formal sector constraints in develop-
may not meet the stringent requirements of ing and making available farmer-preferred
the formal seed system and should be hence varieties, and their wider adoption and dif-
given special treatment. Could protection fusion. It should avoid as much as possible
rights therefore be given for a variety the legalistic and bureaucratic ramifications
whose identity is not clearly established? Is that might undermine its novel approaches.
it possible to enforce protection in case of What is the purpose of protecting PPB vari-
infringements of rights? eties? Should PPB varieties be considered a
Second, the fundamental purpose of public good for the entire farming commu-
PVP is to enforce PBRs, which is a private nity? Is the protection meant to address the
and exclusive ownership right over new public good nature of these varieties?
varieties, and enforced by breeders to Does simply invoking FRs as a means to
recuperate their investments. Technically, protect PPB varieties necessarily serve the
PPB varieties are products of collaboration interest of farmers? It is therefore advisable
among different stakeholders, including to further elaborate the many uncertainties
farmers, communities and breeders. and unanswered questions surrounding the
Who is the owner of PPB varieties: the protection of PPB varieties and develop
individual farmers, their communities, a working mechanism acceptable to all
the collaborating breeders or a collective parties. This will do justice in rationalizing
ownership? Who are the ultimate users or an already burgeoning and complicated legal
beneficiaries of PPB varieties? Does the arena in agriculture. Ultimately, one must
legal protection promote or hinder wider acknowledge that countries have the right
use of PPB varieties? to design IPR regimes that are compatible
Third, at present, neither FRs nor PBRs with their own agricultural development
provide sufficient regulatory framework to and serve the interests of their farmers.
protect PPB varieties, because of complex
legal and technical issues. According to 21.5 CONCLUSIONS
the International Treaty on Plant Genetic According to Srivastava (1997), there are
Resources for Food and Agriculture profound structural changes and emerging
(ITPGRFA; FAO, 2009), FRs are clearly trends in the seed industry, including
defined irrevocable rights arising from globalization of agricultural research,
the past, present and future contributions shifting to private-sector plant breeding,
of farmers in conserving, improving and increased investment in biotechnology,
making available plant genetic resources liberalization of seed trade, emergence
and the opportunity for access and benefit of private seed companies, entry of
sharing from their use by a third party. multinational seed companies, greater
At the same time, PBRs are about private attention to the informal sector, and debate
rights given on a product whose identity is of regulatory and trade agreements on
clearly established and for specific period IPR and biodiversity. These changes call
of time. Who is the source of breeding for establishing an effective, efficient and
materials for PPB varieties? Does the scope transparent variety release system to serve
of FRs under the International Treaty meet the needs of diverse economies.
584 Plant breeding and farmer participation
for Seed Regulatory Reform: An Analysis of ment. IV. The spread and impact of a
Variety Testing, Variety Regulation and Seed rice variety identified by participatory vari-
Quality Control. Agricultural Research and etal selection. Experimental Agriculture,
Extension Network. Network Paper No. 35: 471487.
69. ODI, London. 25 p. Witcombe, J.R., Joshi, A., Joshi, K.D. &
Tripp, R. & Louwaars, N. 1997. The conduct Sthapit, B.R. 1996. Farmer participatory
and reform of crop variety regulation. In crop improvement. I. Varietal selection
R. Tripp, ed. New Seed and Old Laws. and breeding methods and their impact
Regulatory Reform and Diversification of on biodiversity. Experimental Agriculture,
National Seed Systems, pp. 88120. London, 32: 445460.
ODI. 259 p. Worede, M. 1992. Ethiopia: a genebank
Tripp, R. 1995. Seed regulatory frame- working with farmers. In D. Cooper, V.R.
works and resource poor farmers: a lit- Vellv & H. Hobbelink, eds. Growing
erature review. Agricultural Administration diversity: genetic resources and local food
(Research and Extension) Network Paper security, pp. 7894. London, Intermediate
No. 51. ODI, London. 53 p. Technology Publications.
UPOV [International Union for the Protection van Amstel, H., Bottema, J.W.T., Sidik, M. &
of New Varieties of Plants]. http://www. van Santen, C.E., eds. 1996. Integrating Seed
upov.int/en/publications/tg_rom/tg_index. Systems for Annual Food Crops. Proceedings
html (accessed 8 March 2009). of a Workshop, Malang, Indonesia, 2427
Witcombe, J.R. &Virk, D.S. 1997. New direc- October 1995. CGPRT Center, Bogor,
tions for public sector variety testing. In Indonesia. 311 p.
R. Tripp, ed. New Seed and Old Laws. Virk, D.S., Singh, D.N., Prasad, S.C., Gangwar,
Regulatory Reform and Diversification of J.S. & Witcombe, J.R. 2003. Collaborative
National Seed Systems, pp. 5987. London, and consultative participatory plant breed-
ODI. ing of rice for the rainfed uplands of eastern
Witcombe, J.R., Petre, R., Jones, S. & Joshi, A. India. Euphytica, 132: 95108.
1999. Farmer participatory crop improve-
589
CHAPTER 22
in Cuba during the 1980s, versus the 22.3 THE DIVERSITY SEED FAIR
decentralized, participatory plant breeding The first diversity seed fair was held at
model are shown in Figures 22.1 and 22.2, the National Institute for Agricultural
respectively. Science (INCA) in 1999, as an approach for
In contrast to the centralized model, PSD disseminating maize seeds suitable for low-
is based on the individual farmer, through input agriculture (Ros and Wright, 1999).
FIGURE 22.2
Participatory Seed Diffusion promoting a Diversity Chain Reaction
TABLE 22.1
There, professional breeders provided Selection criteria for maize varieties, accepted as
farmers with access to diversity from the important by farmer participants
formal and informal seed systems, and the Criterion Percent of farmer
seeds were sown under relatively low input acceptance
mixed-variety rather than single-variety one private farmer; all ten had attended
planting, which led researchers to conclude the maize seed fair. Three of these farmers
that they would have to overcome were unable to maintain their seeds because
contradictions in the practice of maintaining they lacked the conditions required for
varieties through strict isolation, as conservation from season to season,
demanded by the formal system. having relied for more than fifteen years
It became clear that farmers looked not on the formal seed sector, which supplied
only at yield, but also valued aspects such improved seeds every season.
as plant height, stalk size, number of cobs, The gene pool of the maize population
and number and position of leaves. This of one Havana farmer who selected from
is an indication of the need for alternative the seed fair was found to be composed
breeding objectives. of different seed origins: one commercial
Selection criteria chosen for maize variety from the formal seed sector, five
varieties indicated that farmers, in general, half-sib families of a landrace from La
did not practice seed saving. In fact, during Palma (a neighbouring province), and four
the discussion period, several of them asked half-sib families of a landrace from Catalina
how to save seeds. de Guines (a neighbouring municipality
The general reception given to this new of the same province) (Figure 22.3). Later
participatory approach was positive, given the same farmer bulked all materials and
that farmers were historically accustomed to selected in the field the best 1 5002 000
a more top-down management procedure. plants according to cob size, plant cob
Farmers had rapidly and easily selected height and husk covering, during three
between the 70 lines on show, and a very cycles. Afterwards, at a seed fair prepared
large range of new seed lines had been by his cooperative, this bulked population
extended to them. The plant breeders was sown along with 38 landraces conserved
who started to work in PSD felt that this by the Fundamental Research Institute
diversity indicated the need to refocus (INIFAT) gene bank, 56 half sib families
seed management so that yields and cob of landraces maintained by INCA, four
quality could be improved under low input commercial varieties and a male parent of a
conditions (Ceccarelli and Grando, 2002; popular hybrid (Ortiz et al., 2006, 2007).
Ros, 2003; Acosta et al., 2003; Martnez and Subsequently, the bulked population was
Ros, 2003b). Stimulating the flow of genetic named Felo (the nickname of the local
resource variability has shown the potential farmer breeder) and two mass selection
available for increasing yield performance cycles were done. Gradually, this new seed
on trial plots for farmer acceptance. pool, under farmer management, increased
maize production and diffusion amongst
22.4 FARMER ACCESS TO GENETIC cooperatives, and the area intercropped with
DIVERSITY maize increased over the years (Table 22.2).
22.4.1 Cross-pollinated crops: the Maize rose from being one of the most
example of maize neglected crops in the cooperative to the
Three months after the Diversity Seed Fair, third important profitable crop (Ortiz et al.,
the farmers capacity to develop maize 2003a). Currently, this population, cv. Felo,
populations was assessed among nine is under seed multiplication and continued
farmers working on three cooperatives and selection, having gained recognition from all
Participatory seed diffusion: experiences from the field 595
FIGURE 22.3
Maize selection scheme used by Felo a farmer in Cuba
1 '!(&&!*$%.*$!,0
1 $.!#%"+$"&$%$!+("%' *!"*(&
%&
1 (-*#%"+$"&$%$!+("%' *!
"*(& ,%$' !-$'!+ CYCLE
(/'$'*(/+
+++!%!,$(',/(0%!+
1 %' *!+
('+!*.! 0
1 #%"+$"&$%$!+
("%' *!+
&$',$'! 0
1 (&&! *$%
.*$!,$!+
/(&+++!%!,$('0%!+
1 '!&%!)*!',("
)()-%*#0*$
!%(.*$!,0
the municipality stakeholders, and has been ing programme, and once breeders identify
registered as an official variety in Cuba. a certain population with desired charac-
Usually, the conventional model of teristics, this population is maintained in
breeding cross-pollinated crops entails isolation (Ros, 2003). The interesting fact
recombining in the first stage of the breed- learned through the Felo experience was
596 Plant breeding and farmer participation
TABLE 22.2
Maize production in Cooperative Gilberto Leon, Havana, Cuba
Year 1999 2000 2001 2002 2003 2004
TABLE 22.3
Origins of bean varieties grown at seed diversity fair
Commercial varieties Genetic diversity Accessions collected Total
conserved in gene in the participant
bank communities
(Landraces)
Black beans 17 30 16 63
Red beans 16 15 8 39
White beans 4 14 4 22
Total 37 59 28 124
the possibility of improving yields by dis- mainly with released varieties and lan-
seminating seed diversity. Each genetic pool draces, using a non-segregating population.
built up by farmers could probably be con- Farmers could access up to 124 varieties of
tinuously recombined, choosing for yield bean from different sources (Table 22.3)
improvement as well as other important grown under low-input conditions at the
traits holding cultural or market values. INCA Experimental Station. Each variety
According to the first results of PSD in was sown in a small plot, where partici-
Cuba, seed diversity fairs should become pants could select up to five varieties to be
a recombination process whereby farmers taken home and tested on their farms under
can have access to genetic diversity at their prevailing production circumstances.
community level. In this sense, farmer After more than half of the varieties had
experimentation can play two roles, first in reached the stage of physiological ripening,
continuously providing the best progeny to a meeting was held with the farmers.
the diversity gene pool at community level. In the case of bean, farmer participants
and second in providing farmers with the came from different biophysical and
opportunity to select the best recombined socio-economic contexts. Both marginal
family in a certain cycle in the field. Thus, and industrial farming systems were
PSD in a cross-pollinated crop such as represented by 42 farmers, as well as some
maize seemed to be a simple method where NARS scientists, members of NGOs,
the recurrent selection principle can be functionaries and technicians of the
applied (Maldonado et al., 2006). Ministry of Agriculture.
The bean seed diversity fair was
22.4.2 Self-pollinated crops: the attended by male and female farmers. It
example of beans was planned to carry out varietal selection
In the case of common bean, a self-polli- for women and men separately (Verde et
nated crop, PSD in Cuba has been working al., 2003). A questionnaire was used in
Participatory seed diffusion: experiences from the field 597
t
t
ut
ed nt
Ru in c e
e
Po od
e u
Fi s re a n
od an
an
Ba st r olo
ti t
Se pla
s/p
t p sist
t
ru i s
G in
d
/
Vi re s
ds
ra
ra
ia
confronted with bean diversity was (INIFAT), and rice germplasm was donated
overwhelming, and nobody expected by the Rice Research Institute (IIR), in
genetic diversity to be of such importance addition to collected material.
to farmers. For maize, most of the diversity
In fact, the main interest of farmers collected in Mexico came from local
in maize and bean was to be able to seed systems, with 8 lines provided by
select amongst the wide range of varieties CIMMYT. In Cuba, most collected maize
according their own criteria. Numerous came from local seed systems, with only
varieties conserved in the gene bank showed four commercial varieties coming from
good performance even though some had professional breeders. In every case, each
been lost off the official varieties list. The maize, bean and rice accession collected per
spirit of experimentation, the opportunity family farm was considered as a variety. In
for more such productive options, and comparison with maize and bean, only very
the gender differences detected in the first narrow rice diversity was found in the field
participatory seed selection exercises in (Moreno et al., 2003).
Cuba, inspired farmers, scientists and others In Cuba, several public organizations
stakeholders to further explore PSD in were very open to providing materials for
Cuba and abroad. Consequently, a Mexican seed diversity fairs, and these have been
and Cuban team started to collect seeds considered an important support to the PSD
from different sources, promote diversity process. The main problem in Cuba was the
seed fairs and farmer experimentation in resistance of conventional plant breeders to
their local context. facilitate segregating populations.
In Mexico, it was extremely difficult
22.5 COLLECTION OF SEED DIVERSITY to break the barriers for access to public
A collecting mission was carried out germplasm for developing seed diversity
as a multidisciplinary effort. Teams fairs at community level. At the same time,
composed of scientists from INCA and the reaction of some public plant breeders
local stakeholders, in Cuba and Mexico, was conservative.
collected beans, maize and rice landraces
in different provinces and municipalities 22.6 FARMERS ACCESS TO GENETIC
(Table 22.5). VARIABILITY
In terms of the results of these diversity The genetic diversity conserved in
collection missions (Ros et al., 2006), the conventional gene banks, accessions
teams in Cuba, La Cuenca del Papaloapan collected during the collecting mission
and Chiapas reported potential interesting undertaken by the project, and commercial
material for certain breeding programmes. varieties donated by breeders of bean,
In general terms, the farmers donated their maize and rice, were sown in 2001 in Cuba
seeds freely. In the case of Mexico, the at farm level. In La Cuenca del Papaloapan
phenotypic diversity of collected seeds of (a catchment covering the tropical area of
maize was enough to organize different Oaxaca and Veracruz states), Mexico, two
plots in both Chiapas and La Cuenca de seed diversity fairs were held for maize
Papaloapan. In Cuba, an important bean and bean, and rice plots were attempted
collection was donated by the Fundamental but it was not possible to obtain a harvest
Research Institute in Tropical Agriculture (Table 22.6).
Participatory seed diffusion: experiences from the field 599
TABLE 22.5
Characteristics of collection missions
Crop Region Number of Number of Number of Number of
accessions farmer donors municipalities communities
involved involved
TABLE 22.6
Location and number of varieties grown in seed diversity fairs in the 20022003 period in Mexico
and Cuba
Diversity plot location Crops and Farmers selecting Altitude (masl) Experimental field
no. of varieties plot topography
varieties per
location
All cultivation of the diversity plots was After farmers took seeds to be grown
undertaken according to the traditional on their farms, different workshops were
practices of the participant communities, conducted to discuss selection methods at
except in Chiapas lowlands and Cuenca del community level and experimental design
Papaloapan, where a half-technical package principles. In La Cuenca del Papaloapan,
was applied. Each accession collected was the follow-up process in maize was focused
considered a variety. In all diversity plots, in two communities: Doroteo Arango and
farmers were allowed to choose five or six Vega del Sol.
varieties to take home. In Doroteo Arango, after one selection
cycle working with professional breeders,
22.7 PARTICIPATORY PLANT BREEDING the farmers had to move off their land
AND SEED PRODUCTION because of conflicts of land tenure, and
In both Mexico and Cuba, the facilitation so their maize breeding programme was
of farmers genetic diversity through seed completely halted as all the farmers efforts
diversity fairs increased the early reaction had to be oriented toward land recovery.
obtained from the first two seed diversity In the other community, Vega del Sol,
fairs carried out in Cuban communities. In germination of distributed seeds was
Chiapas highlands, only one seed diversity poor with farmers losing all the varieties
fair was held, the other three did not reach selected at the fair, so then the farmers
harvest due to drought or flood. and professional breeders decided to
In every place where seed diversity fairs start a new collection mission in their
were held, farmers showed great interest in communities. They collected 91 accessions
introducing greater genetic diversity into in neighbourhood communities, setting up
their own farm system (Table 22.7). four experimental plots, one per colour.
In Mexico, participants appreciated that After three years of mass selection,
some traditional varieties were grown in farmer participants had sown 17 ha of land
seed diversity fairs. In this way, traditional with four maize gene pools: white, yellow,
varieties which had almost become red and black, choosing the best cob each
extinct were chosen and multiplied by cycle. Farmers from the community started
participants. to make some negotiations with tortilla
TABLE 22.7
Genetic diversity chosen by farmer participants in the seed diversity fairs
Place Crop No. of No. of varieties Chosen diversity Percent effective
participants grown in seed (a) diversity
diversity fair (b) (a/b 100)
new, more civil approach and orientation Chiapas Highland, Mexico, and Cuba
never before to rescue maize and bean the next season (A. Alda, pers. comm.;
landraces in Chiapas; the other where Mohamed, pers. comm.). Through PSD,
economic support was requested to grow farmers reinforce seed production to be
experimental plots. The second reaction exchanged for experimenting next crop
appeared to be conditioned by other rural season or simply for culinary testing, and
programmes, which supported subsidies they use seeds for promotion or in barter
for food production in the region. Some for other products. In some cases, farmers
farmer leaders in favour of the second who never grew seeds are selling seeds
reaction decided to pull out of PSD. to farmers or to the state seed company.
In Cuba and Mexico, according to the Unfortunately, the team has no details of
perceptions of the participants, yields the volume of seeds sold through PSD.
have improved in crops under the farmer Actually the official scheme of releasing
experimentation process, and farmers were certified seeds to be adopted by farmers
able to diversify and disseminate varieties has partially broken down. In PSD, as
to the rest of the communities after three in other participatory plant breeding
years of testing (Lamin, 2005). methods, farmers adopted varieties by
In general terms, the amount of seed experimentation, and released their best
produced by farmers increased exponen- options once disseminated varieties were
tially in the participating communities. certified (Ceccarelli, 2005, pers. comm.).
In this sense, the seed production process
22.8 DECENTRALIZED SEED in centralized plant breeding, with no
PRODUCTION SYSTEM participatory element, officially starts
After four working years, the research team when improved varieties are multiplied and
noted some differences in seed production certified for dissemination. In PSD, because
concepts between PSD and conventional farmers are participating in the process of
plant breeding. In PSD, a defining charac- selection from the beginning and they are
teristic is the integration within the house- continuously accessing genetic diversity,
hold or community of genetic resource con- seed production is an integral element of the
servation, plant breeding, seed production, process through which farmers decide the
crop production and food consumption. varieties or crops that have to be multiplied
In contrast, in conventional plant breeding, and disseminated.
these functions are institutionalized, special- Currently, four agrobiodiversity centres
ized and separated (Ros, 2003; Cleveland et have been built by collaborative efforts
al., 2005). Therefore, most of the farmers between farmers and professionals scientists
working with PSD test genetic diversity and in Cuba, to promote diversity through
subsequently multiply their best options diversity seed fairs, farmer experimentation
to fill different demands from the family, and seed production by farmer decision.
neighbourhood and local market. Primary diversity centres are farms with
In marginal and industrial environments, capacity to introduce, test and disseminate
the tendency was to retain as much diversity genetic diversity.
as possible. The reaction of some farmers The speed at which PSD has spread in
from marginal environments in keeping Cuba and Mexico has caused an interesting
diversity was: We need to keep various conflict: on the one hand, the legislation
options because who knows how hard is does not allow free national seed flow
Participatory seed diffusion: experiences from the field 603
because seeds are not certified, and on In the case of PSD, such estimation
the other hand, national food security has been applied to each grower who has
depends on informal seed production in selected varieties during diversity fairs
both countries. Therefore discussions to (Figure 22.6).
reconcile the differences are taking place in Indeed, the differential selection reached
both Cuba and Mexico. by farmers gave evidence of their capacity
for obtaining superior materials amongst
22.9 FARMERS GENETIC GAINS certain populations. The results strongly
As yields were increasing in the imply that farmers participating in plant
communities, a discussion emerged in selection and seed diffusion could collaborate
different communities implementing PSD in simultaneously increasing yields and
about the real influences of farmer selection diversity. In practice, access to diversity in
on yield response. In fact, the team and the form of released varieties and segregating
scientific community looked for hard populations could provide an interesting
evidence on farmer selection efficiency. fit at local level (see Rosas, Gallardo and
In conventional breeding programmes, Jimenez (2006) for segregating populations).
one of the common indicators for Other interesting evidence is the case of
determining the impact of selection pumpkin breeding (Table 22.8). The farmers
consists of estimating genetic advance who choose gene pools on farm, according
through selection (Falconer, 1960), which to their criteria, had more efficient use
is described as follows: of energy for producing food and more
S = h2 DS profitable crops
where: S = selection advance; h2 = herit- Conventional pumpkin breeding
ability and DS = selection differential, as in Cuba provides an example of the
discussed in detail in Chapter 2. possible negative economic effects when
FIGURE 22.6
Average selection differential attained by farmers in 18 bean,
10 maize and 6 rice seed diversity fairs in Cuba
t/ha
1.2
0.8
0.6
0.4
0.2
0
2001 2002 2003 2004
Year
TABLE 22.8
Economic impact of pumpkin breeding under low input conditions
Indicators (calculated as averages) Varieties bred under high Varieties bred and
input conditions sown in sown under low input
low input conditions conditions
Cost per ha under low input conditions (Cuban pesos) 702.3 708.3
Fruit yield (t/ha) 1.5 6.7
Total income (@ 0.16 Cuban pesos per kg) 240 1080
Net income per ha (Cuban pesos) -462(1) 372
Benefit:cost ratio 0.34:1 1.5:1
TABLE 22.9
Socio-economic and biophysical contexts of scaled-out Participatory Seed Diffusion
Indigenous culture Farmer literacy Research- Production
development policy potential
priority
varieties are selected in an environment was eager to know how PSDs, emerging
not representative of the target area. from the western part of Cuba, could be
The occurrence of a cross-over response practically adapted to other Cuban zones
(Ceccarelli et al., 1994; Ceccarelli and and abroad, with different biophysical and
Grando, 2002) suggests the importance of socio-economic contexts (Table 22.9).
having a realistic view about who will be What did we scale out? Chiefly we
using the products of plant breeding. scaled out:
The experience described in this chapter r 5IF EJBHOPTUJD QIBTF
MPPLJOH GPS MPDBM
attempts to maximize the role of local multi- genetic diversity, intervention entry points
sectoral efforts, including international, and enabling institutional environments,
national and local stakeholders, through for a change of paradigm.
promoting the generation of benefits at r 4FFE EJWFSTJUZ GBJST JO NBJ[F
CFBO BOE
local level by using PSD. rice, to stimulate varietal demand and
enhance farmer participation in generating
22.10 SCALING UP PARTICIPATORY benefits.
APPROACHES It was very effective to discuss the idea
As a result of the outcome of the two of PSD with a wide range of stakeholder
breeding cycles in Cuba, the team and participants; in fact, a constructive reaction
other partners decided to expand the pilot was received from government, civil
experience from the western part of Cuba society and farmers. They built up the
in the form of a PSD programme for the different teams and planned the activities,
central and western parts of Cuba, and to and immediately started to work. Local
the Highland of Chiapas and La Cuenca organizations were extremely cooperative
del Papaloapan, Mexico. The working team in supporting the process.
Participatory seed diffusion: experiences from the field 605
The teams main work objectives were important element for successful family
to understand the seed flows, leadership business under restricted financial condi-
relations and reaction of local policy-makers tions (Ros, Soleri and Cleveland, 2002).
in terms of supporting the idea. In parallel, In maize, a cross-pollinated crop, there
and as a key activity, teams collected genetic were significant agromorphological differ-
diversity from the formal and informal seed ences between farmer-collected accessions,
sector, mainly of maize and beans. even though the local maize population had
In addition, Cuba had a Popular Rice the same name: criollo, pintico, amarillo,
Movement which was highly suited to negrito, blanco, etc. One hypothesis is that
the application of PSD. The Popular Rice such diversity made it possible to improve
Movement is a peoples movement to certain complex characteristics, such as
produce rice under low input conditions yield, through farmer participation (Acosta
for self-consumption and markets within et al., 2003; Martinez, 2005). In the case
Cuba. This movement aiming at producing of beans, a self-pollinated crop, few bean
the main staple food emerged in the 1990s types existed in industrial farming systems,
in response to the collapse of conventional and in certain lowlands of Chiapas farmers
rice production handled by the large state decided to stop growing beans due to dis-
farms. Farmers were then allowed to plant ease attacks, whereas in the upland it was
rice everywhere, and the government made possible to collect different types of beans
the land available for this (Moreno et al., to be intercropped with maize.
2005). In general terms, with beans, farmers
In terms of farming approaches, in Cuba, perceived increased disease susceptibility
farmers were experiencing a special period and loss of genetic diversity over the
due to the collapse of the Socialist Block in previous decade. Limited access to new
the late 1980s (Ros, 2003), which in general genetic diversity from either the formal or
terms meant that they had very limited informal seed sectors was evident. Some
access to agrochemicals and improved seeds morphological differences were found to
of basic grains. In Chiapas, in contrast, be limiting genetic diversity within grain
upland farmers had no choice but to grow colour of farmers beans prior to the PSD
their crops in a marginal environment. In intervention (Miranda, 2005).
comparison, La Cuenca del Papaloapan Finally, the teams work showed that the
was a high-potential environment and had situation for Cuba and Mexico was common
received enormous agricultural investment in terms of limited access to financial
in the 1980s for maximizing yields according resources to buy seeds and agrochemicals
to Green Revolution philosophy. In 2001, for the production of basic grains. In the
however, farmers in this region had, for particular case of Mexico, stakeholders felt
various reasons, lost a major part of the threatened by the USA policy of selling
official financial support. cereals at very low prices. In fact, the
According to the diagnosis phase carried limited economic situation faced by Cuba,
out before the PSD intervention, farmers in relation to Green Revolution concepts,
who have more diversity and dynamic seed was not exclusive; other regions were
exchanges in maize had more profits, in suffering from similar problems and local
both Cuba and Mexico. The experimenta- innovation was emerging as a response for
tion capacity of farmers seemed to be an overcoming obstacles to producing food.
606 Plant breeding and farmer participation
FIGURE 22.7
Participation versus external cost of participatory diffusion
of bean and maize seeds in Cuba and Mexico
+
3ch
22c 18c 19c
27c
1ch
COSTS
PARTICIPATION
- 28c
6ch
20c 5ch
2ch
24c
26c
+
10ch
16ch 12ch 9ch 13ch
25c 14ch 8ch
7ch 15ch 29ch
30ch
-
Participatory seed diffusion: experiences from the field 607
and varieties brought by professional The PSD in Chiapas was largely focused
researchers, farmers, NGOs, private on the highlands, with farming systems
companies, etc. Farmers were able to on sloping areas, and with farmers having
involve communities in undertaking very low literacy levels. However, most of
participatory approaches. the characteristics represented by the high
r Medium: Farmers organized seed diversity participation and low external support in
fairs on public property, and seeds and Figure 22.7 belonged to the seed diversity
technologies were supplied by farmers fairs developed in that region.
and professional researchers; farmers were The results confirmed the hypothesis
partially able to call on participants for that local innovations are not strictly related
undertaking participatory approaches. to literacy levels. Even though farmers had a
r Low: Public or private institutions high literacy level in Cuba, the relationship
organized seed diversity fairs on between professional scientists and farmers
experimental stations, and researchers, was weak before the collapse of the socialist
extension agents, public or private countries, and it was currently taking some
functionaries took decisions. Farmers time to establish a new relationship. It has
could not involve other farmers in been a difficult process to convince the
undertaking participatory approaches. professional scientists to consider farmer
The y axis was represented by three participation as a scientific element of their
categories of external costs as follows: profession.
r High: The expenses for food, participant In general terms, the agricultural
transportation and implementation education systems did not consider farmers
of diversity plots was covered by the as collaborators or partners of research
project. work, scientific services or policy-making,
r Medium: The food expenses and partici- and decisions in agriculture had a very strong
pant transportation was paid for by the top-down character. However, research
project. The expenses of implementing institutes and development organizations
diversity plot was covered by communi- have worked directly in different ways to
ties. quickly adopt participatory plant breeding
r Low: The implementation of experimen- methodology, even though the concept was
tal plots, food and transportation was not well documented. Personal influences
covered by the communities. of researchers have played a critical role in
Figure 22.7 shows how the external cost scaling-out PSD (Chaveco et al., 2006).
decreases with an increase in participation
over the four years of project implementation. 22.12 CONCLUDING REMARKS
The results show that external costs could be Usually, the route of plant genetic resources
reduced gradually when local stakeholders collected in communities ends at research
adopt participatory methodologies, and institution gene banks, to be used in
the recognition of farmer knowledge as conventional plant breeding programmes
well as the economic benefits of farmer (Almekinders et al., 2000). The experiences
experimentation seems to be an important discussed in this chapter provide evidence
incentive for developing PSD. Farmers of how material from collecting missions
decided to incorporate trials as organic could be tested, multiplied, improved
components of their farming systems. and disseminated by farmers and local
608 Plant breeding and farmer participation
Currently, the public research institutions in Cuba. At the same time, the organizations
are suffering from severe financial leading PSD in Mexico are focusing on
restrictions; they are strongly influenced more entrepreneurial tendencies to show
by external budget changes, which are how people marginalized by top-down
very vulnerable to socio-economic or approaches can be recognized as innovators
political changes. The field experience and potential local managers of plant genetic
described in this chapter provides a clue resources. In practice, both country cases are
that genetic diversity could lead to a viable, dealing with their own contexts. However,
small, economic initiative for many local both countries are enhancing diversity,
stakeholders. farmer participation and new technological
New institutional arrangements for and institutional arrangements towards
enhancing collaborative efforts between more integrated food production.
scientists and farmers seem to be an
important issue in reaching a better ACKNOWLEDGEMENTS
understanding of local seed systems and The author is grateful to Ronnie from
agrobiodiversity incentives (Vernooy, 2003) the International Development Research
as development cells for national and Centre (IRDC); Lisse and Richard from
international development. the Canadian International Development
It is quite clear that the experience accu- Agency (CIDA); Olivier and Rodolfo from
mulated from PSD in Cuba and Mexico Swiss Development Cooperation for their
shows that innovation in agriculture is advice and funding. Special gratitude to Jose
not exclusively a business for profession- Roberto from INCA for patience, enthusi-
al scientists, but that by involving local asm and friendly discussions of the idea
stakeholders and farmers the impact of and implementation of the project. Thanks
plant breeding in different contexts might to Juan and Reynel from SEPI (Indigenous
increase. PSD has been able to revive the Peoples Secretary of Chiapas), Victor and
professional plant breeding role and farmer Margarito from UGOCP (General Farmers
knowledge in a current context. Perhaps and Workers Union), as well as Arturo
the results obtained by the collaboration from SDR (Development Secretary of
of farmers and scientists, and the difficult Chiapas Highlands), who believed in PSD
economic situation faced by national and as a tool to induce some changes in Mexico.
international public plant breeding, could Deep thanks to all my researchers, techni-
facilitate new approaches towards more cians and farmers from Cuba and Mexico:
diverse, productive, socially and economi- actually, they were the real concept makers
cally fair plant breeding in future years. of Participatory Seed Diffusion. Thanks to
The economic and energy efficiency of Nathaniel and Julia for their draft discus-
selecting varieties under real environmental sion and spelling support.
conditions, and farmers attitudes to
experimentation, become important REFERENCES
arguments to convince policy-makers Acosta, R., Ros, H., Verde, G. & Pomagualli,
to apply PSD as a transformative tool D. 2003. Evaluacin morfoagronmica de la
in agriculture. Officially, PSD has been diversidad gentica de variedades locales de
focused as a method to encourage public maz (Zea Mays L.) en La Palma, Pinar del
welfare and re-evaluate public institutions Ro. Cultivos Tropicales, 24(23): 6167.
610 Plant breeding and farmer participation
Almekinders, C. 2000. The importance of Enrquez, J.L. 2000. Cubas new agricultural
informal seed sector and its relation with revolution. The transformation of food crop
the legislative framework. Paper presented production in contemporary Cuba. Food
at GTZ-Eschborn, 45 July 2000. First-Institute for Food and Development
Begemman, F., Oetmann, A. & Esquivel, M. Policy, Department of Sociology, University
2000. Linking conservation and utilization of California. Development Report, No. 14.
of plant genetic resources in Germany and 19 p.
Cuba. In C. Almekinders & W. de Boef, Esquivel, M. & Hammer, K. 1992. Contemporary
eds. Encouraging Diversity, pp. 103112,. traditional agriculture-structure and diver-
London, Intermediate Technology. sity of the Conuco. In K. Hammer, M.
Castieiras, L. 1992. Germoplasma de Esquivel & H. Knpffer, eds. ...y tienen faxo-
Phaseolus vulgaris L. en Cuba: colecta, nes y fabas muy diversos de los nuestros.
caracterizacin y evaluacin. Ph.D. Thesis. Origin, evolution and diversity of Cuban
Universidad Agraria de La Habana, Cuba. plant genetic resources. Vol 1. Gatersleben,
Ceccarelli, S. & Grando, S. 2002. Plant breeding Germany, Institut fr Pflanzengenetik und
with farmers requires testing the assump- Kulturpflanzenforschung.
tions of conventional plant breeding: les- Falconer, D.S. 1960. Introduction to
sons from the ICARDA barley program. In Quantitative genetics. Edinburgh, UK,
D.A. Cleveland & D. Soleri, eds. Farmers, Oliver and Boyd.
scientists and plant breeding: integrat- Funes, F. 1997. Experiencias cubanas en
ing knowledge and practice, pp. 297332. agroecologia. Agricultura Orgnica.
Wallingford, UK, CABI. Agosto-Diciembre, 37.
Ceccarelli, S., Erskine, W., Hamblin, J. & Funes, F. 2002. The organic farming move-
Grando, S. 1994. Genotype by environ- ment in Cuba. In F. Funes, L. Garcia,
ment interaction and international breed- M. Bourque, N. Perez & P. Rosset, eds.
ing programmes. Experimental Agriculture, Sustainable Agriculture and Resistance;
30: 177187. Transforming Food Production in Cuba,
Cleveland, D., Soleri, D., Aragn, F., Fuentes, pp. 126. California, USA, Food First
M., Ros, H. & Stuart, H. 2005. Genes in the Books.
landscape: Genotype-by-environment inter- Lamin, N. 2005. Evaluacin del impacto del
action, farmer knowledge, and the political fitomejoramiento participativo sobre los
economy of transgenic crop varieties in the sistemas locales del frjol comn (Phaseolus
TW. In L. Cliggett & C. Pool, eds. Economies vulgaris L.) en La Palma, Pinar del Ro,
and the Transformation of Landscapes. Cuba. M.Sc. Thesis, Universidad Agraria de
Walnut Creek, USA, Society for Economic la Habana, Cuba.
Anthropology and AltaMira Press. Lopez, M., Padron, R., Moreno, I., Lopez,
Chaveco, O., Miranda, S., Ros, H. & Garca, F., Tapia, R., Acosta, R., Guzmn, G.E. &
E. 2006. La seleccin participativa de varie- Ros, H. 2005. Seleccin participativa de
dades como alternativa para revalorizar los variedades de arroz y criterios de seleccin
intereses locales. El frijol caup en Holgun. de los productores de arroz popular en
In H. Ros Labrada, ed. Fitomejoramiento Camaguey. In Program and Abstracts of
Participativo. Agricultores Mejoran 3rd Rice International Meeting, pp. 1516.
Cultivos. Havana, Instituto Nacional de Palacio de las Convenciones, La Habana,
Ciencias Agricolas. Cuba, 810 June 2005.
Participatory seed diffusion: experiences from the field 611
Maldonado, J., Pinto, R.M., Tardn, J.M. & cin agropecuarias de La Habana. Cultivos
Jantara, E. 2006. Fitomejoramiento par- Tropicales, 24(4): 115122.
ticipativo de variedades de maz criollo: Ortiz, R., Ros, H., Ponce, M., Acosta,
una experiencia en la Regin Centro Sur R., Miranda, S., Martin, L., Moreno, I.,
de Paran, Brasil. In H. Ros Labrada, Martinez, M. & De la Fe, C. 2003b. Impacto
ed. Fitomejoramiento Participativo. del fitomejoramiento participativo cuba-
Agricultores Mejoran Cultivos. Havana, no en el desarrollo sostenible local. In
Instituto Nacional de Ciencias Agricolas. III Conferencia Internacional. Desarrollo
Martnez, M. 2005. Caracterizacin y eva- Agrario y Sostenibilidad. Facultad de
luacin participativa de maz colectado en Ciencias Agropecuarias, Universidad
Catalina de Guines, La Habana y Ventas de Central Marta Abreu de las Villas. Santa
Casanova, Santiago de Cuba. M.Sc. Thesis, Clara, Villa Clara, Cuba, 1416 June 2003.
Universidad Agraria de la Habana, Cuba. Ortiz, R., Ponce, M., Ros, H., Verde, G.,
Martnez, M., Ortiz, R. & Ros, H. 2003. Acosta, R., Miranda, S., Martn, L., Moreno,
Caracterizacin y evaluacin participati- I., Martnez, M., De la F, C. & Varela,
va de maz colectado en la localidad de M. 2005. Efectividad de la experimenta-
Catalina de Gines, La Habana. Cultivos cin campesina en la microlocalizacin de
Tropicales, 24(4): 6975. variedades de frijol y la evaluacin de la
Miranda, S. 2005. La seleccin participativa de interaccin genotipo ambiente. Cultivos
variedades como mtodo para incrementar Tropicales, 24(4): 107113.
la adopcin y la diversidad e frijol comn Ortiz, R., Chvez, F., Ros, H., Caballero, R.
(Phaseolus vulgaris L.) en comunidades & Vials, M.E. 2006. La primera variedad
rurales. M.Sc. Thesis, Universidad de La de maiz obtenida por las tcnicas de fitome-
Habana, Cuba. joramiento participativo en Cuba. Cultivos
Moreno, I., Ros, H., Miranda, S., Acosta, Tropicales, 27(1): 4950.
R., Alanis, V. & Fernndez, L. 2003. Ortiz, R., Ros, H., Ponce, M., Gil, V., Chaveco,
Caracterizacin de la diversidad en dos eji- O. & Valds, R.J. 2007. Impactos del fitome-
dos arroceros de La Cuenca del Papaloapan, joramiento participativo cubano. Cultivos
Mexico. Cultivos Tropicales, 24(4): 5559. Tropicales, 28(2): 511.
Moreno, I., Medero, B., Pulido, M., Torres, S., Ros, H. 1999. Pumpkin breeding under low
Padrn, R., Torres, S., Tapia, R., Lpez, M., input conditions in Cuba. PhD Thesis.
Lopez, F., Guzmn, L., Acosta, R., Ros, H., Havana Agricultural University, La Havana,
Puldon, V., Martinez, M., Cruz, A. & de Pasos, Cuba.
R. 2005. Seleccin participativa de variedades Ros, H. & Wright, J. 1999. Early attempts
como va de incremento de la diversidad en el at stimulating seed flows in Cuba. ILEA
arroz popular. pp. 113114, in Program and Newsletter, No. 15: 3839.
Abstracts of 3rd Rice International Meeting. Ros, H., Soleri, D. & Cleveland D. 2002.
Palacio de las Convenciones. La Habana, Conceptual changes in Cuban plant breed-
Cuba, 810 June 2005. ing in response to a national socioeco-
Ortiz, R., Ros, H., Ponce, M., Verde, G., nomic crisis: the example of pumpkins. In
Acosta, R., Miranda S., Martin L., Moreno, D.A. Cleveland & D. Soleri, eds. Farmers,
I., Martnez, M., De la Fe, C. & Valera, Scientists and Plant Breeding: Integrating
M. 2003a. Impactos de la experimenta- Knowledge and Practice. Wallingford, UK,
cin campesina en cooperativa de produc- CABI Publishing.
612 Plant breeding and farmer participation
CHAPTER 23
Ronnie Vernooy
with Pratap Shrestha, Salvatore Ceccarelli, Humberto Ros Labrada,
Yiching Song and Sally Humphries
614 Plant breeding and farmer participation
Working in situ, and decentralization comes in many forms and degrees. The
The Nepal case study points to a number of International Centre for Agricultural
important features. Perhaps the first to note Research in the Dry Areas (ICARDA)
is that the LI-BIRD research team worked participatory plant breeding efforts in the
in situon farms and in communitieswith Middle East and North Africa are based
farmers as research colleagues, each com- on it. One of the advantages it offers
plementing as much as possible the others in terms of efficiency is that selection in
knowledge, skills and experience. In this farmers fields avoids the risk of useful lines
case, the research project was and remained being discarded because of their relatively
strongly LI-BIRD directed, as the team poor performance at experimental stations,
decided where to work and also maintained where conditions are almost certainly
a generally strong hand in directing the more favourable. Decentralization as an
research process, i.e. selection of varieties organizational practice could be looked at
to be tested, seed quantity, and number of with the same perspective as participation.
farmers invited to grow the new varieties.
These decisions clearly affected the results. 23.6 CASE STUDY 2: ICARDA
Although farmers benefited from intro- This study is adapted from Vernooy (2003),
duced varieties, it is likely that their rela- and based on various ICARDA research
tively limited decision-making restricted results and publications.
the potential for a more transformative In the late 1990s, a team of researchers
change. (This is an observation about the at the ICARDA pioneered a new way to
relationship between intent and result, and work with farmers in marginal rainfall
should not be seen as a critique.) environments of Morocco, the Syrian Arab
Another feature that emerges from the Republic and Tunisia. They set out to work
case study is that decentralization replaces together with farmers and aimed to fulfil
centralization as the main organizing the needs of people living and working in
principle in order to address specific local the harsh conditions of the region. In Syria,
contexts, i.e. GE interactions, and socio- for example, researchers worked with host
economic variables including age, class or farmers. In the context of Syrian farming,
caste, gender and ethnicity. Although the these were men who accepted the invitation
research described took place at only one made by the researchers to partake in the
site, as a means to validate the approach, research in nine communities (identified
LI-BIRD subsequently concluded that this by the researchers) and with two regional
principle of decentralization could be used research stations. These host farmers and
on a wider scale, and probably countrywide. their neighbours, varying from a few to
Again, here we are dealing with a researcher- a dozen or more, took care of the trials,
directed intervention, but one that could which involved experimental lines from
have potentially a much broader impact as the research station and the farmers own
it concerns an organizational principle at varieties. Farmers and breeders assessed the
the programme, and even national research results independently in successive trials
policy, level. from 1997 to 1999. Several promising new
Decentralization (see also Chapter 9) varieties were identified from these trials.
has been at the heart of many alternative It quickly became apparent that the
approaches, but, as with participation, it farmers selection criteria, largely based on
Towards new roles, responsibilities and rules: the case of participatory plant breeding 619
environmental factors, were quite different welcomed the ability to select among a
from those used by the national breeding large number of lines; some farmers have
programmes. To the surprise of many, the started seed increase of selected varieties.
selections made by the farmers were at This has opened the window to a more
least as effective as those made by the dynamic process, with new materials being
breeders. The newly introduced materials introduced at any time.
gave good yields, and this in areas where The researchers also noted that womens
plant breeding had not previously been selection criteria often differed from the
successful. Farmers also gained access to mens, highlighting the importance of
varieties that responded to preferred traits ascertaining when and why they differ.
such as tall plants, large kernels, good early They also noted that farmers became
growth vigour, high tillering and lodging empowered by their involvement in the
resistance. Seeing these promising results, research process, gaining the confidence
breeders quickly adopted the new ideas to take decisions on crosses as well as on
and attitudes, becoming supporters of the factors such as plot size and the number
participatory approach and expanding it of locations. Perhaps of equal importance
to other areas and to other crops. The to the researchers themselves, the project
team learned that earlier plant breeding revealed the need for specific training in
programmes were ineffective on marginal areas such as experimental design and
lands because they seldom included among data analysis suitable for situations where
their selection criteria those traits that are the environment (a farmers field under
important to farmers. farmer management) cannot be under the
In addition, it became clear that scientists control as it is in the research
decentralized selection in farmers fields stations. ICARDA and national partners
avoids the risk of useful lines being discarded have continued to expand their efforts by
because of their relatively poor performance scaling-up the approach in the national
at experimental stations, where conditions systems in the region and by trying out the
are almost certainly more favourable, methodology on other crops.
through fertilization or irrigation, for
example. Decentralized selection combined Research management
with farmer participation from the initial What becomes apparent from the above
stages of the breeding process is a powerful discussion and case studies is that such
methodology to fit crops to specific new approaches require a different way of
biophysical and socio-economic contexts, organizing time, labour and the research
and to respond to farmers needs and process, i.e. the roles and responsibilities
knowledge. previously described. The emphasis is on
The researchers learned a number of other step-wise producing or co-producing as
critical lessons from the project. Among effectively and efficiently as feasible a
them is the fact that farmers can handle a project through face-to-face interactions,
large number of lines or populations, or especially in the field. Bringing different
both. Most notably, in Syria in phase 2 disciplines to the table and field is one
of the work, the number of lines assessed important element. Research management
in some villages increased from around requires flexibility. It is not about
200 up to 400! In fact, farmers warmly implementing blueprints. This new method
620 Plant breeding and farmer participation
of organizing time and labour will therefore organizations. The project is already making
benefit from adaptive process management an important contribution to improving
knowledge and skills. Farmers usually Cubas food security options.
already have a significant amount of this The key element in the project has
capacity, and it is useful to build on their been to involve the farmers, and this has
expertise, and perhaps, where useful, explore been achieved through farmer research
ways to strengthen it. Researchers may in experimental groups. The project
need to be trained to acquire this capacity. team believed that strengthening the
Insights from learning theory can be of organization of farmers increases their
much value, as well as from participatory capacity to experiment and innovate and
monitoring and evaluation approaches. to make stronger demands on the formal
Start-up periods of collaborative research agricultural research system. One method
are usually very labour intensive, requiring a the researchers used to introduce farmers to
good deal of time and effort to lay a founda- new or unknown varieties or lines was the
tion of trust and to build working relation- seed fair. Initially, this took considerable
ships, both within the research team, and planning and facilitation efforts as fairs
between the core team and others involved were organized by the INCA team and at
in the research. Longer-term commitments the INCA station. Farmers were wary of
are important, to be able to create meaning- this new approach (none had ever visited
ful and effective collaboration and to cope the INCA station), but many attended out
with unavoidable setbacks, such as a crop of curiosity. What they saw overcame their
failure due to drought. Experiments, par- reservations. The researchers managed to
ticularly in plant breeding, usually require collect genetic materials for many maize and
various cycles of selection to produce use- bean varieties (later, fairs were organized
ful results, and thus time horizons should for other crops), including commercial and
not be too restricted. Organizing regular local varieties, as well as promising new
feedback opportunities and using the results lines. The farmers were impressed.
promptly to adapt or change directions is The fairs demonstrated to farmers
another important element. the diversity of their staple crops. The
researchers subsequently allowed the
23.7 CASE STUDY 3: CUBA farmers (men and women) to select materials
The study is adapted from Vernooy for testing in their own fields, under local
(2003), and draws on National Institute conditions. This proved very popular and
for Agricultural Sciences (INCA), Cuba, successful. It proved that farmers are able
research results and publications. to assess and select from a large number of
In 2000 an interdisciplinary group of options alongside breeders. Ultimately, the
dynamic researchers at INCA took on the fairs have proved to be hugely popular, so
challenge of reshaping agriculture on the much so that farmers quite spontaneously
island. They began a project designed to started to organize similar fairs in their
improve the yield and quality of the maize own communities. Initially, the researchers
and bean crops in both unfavourable and guided and supported the farmers in doing
more favourable production areas, through this, but subsequently farmers organized
a combined effort of increased varietal fairs mostly or all by themselves. Farmers,
diversity and strengthened local farmer breeders and extension agents now continue
Towards new roles, responsibilities and rules: the case of participatory plant breeding 621
to rub shoulders at fairs, assessing varieties, Understanding natural resource and crop
and selecting the ones they like best. dynamics requires taking into account both
Breeders continue to assist farmers with the biophysical and the social dimensions
experimental design on-farm, but all trials of the processes involved in managing and
are adapted to the local context. maintaining productivity and agrobiodi-
Farmers say that in addition to introduc- versity. Plant breeders have much to gain
ing new and higher yielding maize and bean from working with social scientists in an
seeds (e.g. bean yields in the Havana experi- interdisciplinary research mode to docu-
mental site have gone up on average by 15 ment and analyse the social nature of farm-
percent and in the La Palma site by an average ing, plant breeding, and of doing research.
of about 35 percent), some of which are also Social scientists have the opportunity to
more resistant to diseases, the fairs provide ground their work in real-life situations.
new knowledge about how to handle and Facilitation and convening are new and
conserve seeds. By developing closer links important additional roles for traditional
between farmers and researchers from the plant breeders. Additional training is an
formal system, the fairs have also increased important investment if these skills are
the farmers capacity for experimentation. lacking (among researchers or farmers, or
And last, but by no means least, the fairs both). Working with a diverse group of
have become social and cultural events that people including scientists in various fields,
bring rural people together, young and old, women and men farmers, and extension
and give them an opportunity to share their workers means balancing a variety of ideas,
knowledge and experiences. interests, skills and personalities. Managing
The project team also organizes regular the process of participatory planning,
field days as another way to learn more about implementation, monitoring and evaluation
farmers preferences. Here the farmers, means paying significant attention to
both men and women, are interviewed interactions and communication, as well
about their preferences. The information as ensuring openness and fairness. Building
gathered is crucial for the INCA plant and strengthening the participatory process
breeders in identifying parental materials becomes a central part of the agenda. The
and selection criteria. To date, the project following case is a good example.
has been successful at both broadening the
genetic base and improving the quality of 23.8 CASE STUDY 4: CHINA
varieties. INCA is currently extending the This study is adapted from Vernooy (2003),
methodology and results to other provinces Vernooy and Song (2004) and various
through collaboration with other Cuban Center for Chinese Agricultural Policy
agricultural research entities. Envisioned (CCAP) documents.
is the creation of a national network to In China, new plant breeding approaches
exchange experiences, new ideas and seeds, have been pioneered by CCAP, a leading
and to provide inputs into the policy- agricultural policy research institution that
making process. is part of the Chinese Academy of Sciences
(CAS), and by the Guangxi Maize Research
Interdisciplinarity and facilitation Institute (GMRI), part of the Chinese
The case studies presented so far indi- Academy of Agricultural Sciences (CAAS).
cate that interdisciplinarity is desirable. The CCAP/GMRI research aims to
622 Plant breeding and farmer participation
identify technical and institutional options two formal breeding institutes and CCAP
for developing more effective linkages and have been directly involved in project
mutually beneficial partnerships between design and implementation. The team is
the formal and farmers seed systems. The engaged in an ongoing dialogue in order to
main hypothesis is that only such new integrate the very many contributions from
institutional development can enhance the very broad expertise base. This is not
sustainable crop development. and in always easy, but so far efforts have been
situ and on-farm management of genetic very productive.
resources. It also aims to strengthen women The research uses a participatory plant
and men farmers research and management breeding methodology adapted to the
capacities to maintain agrobiodiversity in local context. Trials in six villages and
the specific Chinese context. on-station have included both participatory
A major PPB project was implemented plant breeding and participatory variety
in Guangxi province in south-west China selection experiments. The trials allow for
following an impact study carried out from comparison in terms of locality, approach,
1994 to1998 by the International Maize and objectives and the types of varieties tested.
Wheat Improvement Centre (CIMMYT) Varieties include landraces, open-pollinated
to assess the impact of CIMMYTs maize varieties, so-called waxy maize varieties,
germplasm on poor farmers in south-west and varieties introduced by CIMMYT.
China. That study critically analysed the Some of the CIMMYT varieties have been
processes of technology development and locally improved through crossings and
diffusion. One of the key findings of the selections. Agronomic traits, yields, taste
impact study was the systematic division and palatability of these improved varieties
between the formal and the farmer seed are satisfactory. They are showing better
systems. This resulted in inadequate variety adaptation to the local environments.
development, poor adoption of formally Varietal diversity is increasing.
bred modern varieties, an increasingly The projects PPB field experiments,
narrow genetic base for breeding, and a both in farmers fields and on station,
decrease in genetic biodiversity in farmers have been functioning successfully as a
fields. platform to involve the main stakeholders
The project team supported farmers from both formal and farmer systems.
groups through training, linkages and They have facilitated effective interaction,
network building, and market involvement communication and collaboration among
among farmers and with the formal system them. Through this platform, the approach
actors. Policy-changes aim to bring about and results have reached high-level decision-
conceptual change among formal research makers (at the provincial and national
and seed system actors so that they better levels), and some inroads have been made
understand farmer roles and enable farmer into the policy process. Farmers, women in
participation. The project is implemented particular, are now speaking up in meetings
by a team of women and men from various and expressing their ideas, needs and interests.
institutions and groups, from different In a still strongly top-down research and
disciplinary backgrounds and operating at policy environment, this represents a major
different levels. Five women farmer groups, change. PPB has also strengthened the local-
six villages, six township extension stations, level organizational and decision-making
Towards new roles, responsibilities and rules: the case of participatory plant breeding 623
varietal and seed quality; variety release characteristics, but they all have one thing
systems regulating the spread of varieties in common: they provide a direct link
of proven quality to farmers; phytosanitary between locally organized farmers and the
law; seed certification schemes that aim to formal agricultural research systems.
control varietal identity and purity; and
seed quality control mechanisms that check Honduras
viability, purity and health. This theme is In Honduras the number of CIALs has
discussed in more detail in Chapter 21. grown rapidly and there are now 82,
These so-called regulatory framework comprising around 900 farmers in different
components are embedded in broader soci- regions of the country, most of them in
etal institutions, including policies affecting remote mountainous areas where they are
rural development and agricultural research frequently excluded from conventional
more broadly, e.g. land tenure, taxation, agricultural research and extension services.
marketing, financing of public research, pro- The CIALs are organized into five regional
vision of credit, and provision of extension associations of a national CIAL federation,
services. Depending on context, research the Honduran Association of CIALs.
into these broader institutional questions Fifty-five of the farmer research teams
may be highly relevant. The current trend are supported by a Honduran NGO,
of shrinking budgets around the world for La Fundacin para La Investigacin
public national agricultural research seems Participativa con Agricultores de
to make this area particularly relevant. Honduras (FIPAH), which began as a
Looking for opportunities to build on project entitled Investigacin Participativa
local change processes already in motion en Centroamrica, which was supported
or to explore spaces for change becomes an initially by CIAT and then by the
important skill. The following case studies International Development Research
are examples of how spaces for change were Centre (IDRC) between 19952000; since
found or created. 2000 it has been supported by a Canadian
NGO, USC-Canada, under its Seeds of
23.9 CASE STUDY 5: HONDURAS Survival (SoS) Program, with financial
AND NICARAGUA: CREATING SPACE backing from the Canadian International
FOR EXPERIMENTATION, ENHANCING Development Agency (CIDA).
LOCAL ORGANIZATIONAL CAPACITY With a team comprising four local
The Honduras case builds on Humphries agronomists with the collaboration of a
et al. (2005), and the Nicaragua case on Canadian rural sociologist, FIPAHs
Vernooy et al. (2000) and Vernooy (2003). agronomists have successfully bridged the
Local agricultural research committees, divide between plant breeders at the regions
or CIALs to use their Spanish acronym, have largest agricultural research institution, La
sprung up all over Latin America. CIALs Escuela Agrcola Panamericana, Zamorano,
bring farmers and researchers together in and poor hillside farmers (Humphries et
a process of joint experimentation and al., 2005).
learning. The concept was developed at Achieving organizational integration
the International Centre for Tropical between farmers and scientists in Honduras
Agriculture (CIAT) in Colombia, and it is quite remarkable. In the countryside there
quickly caught on. They vary in size and are few strong community organizations
Towards new roles, responsibilities and rules: the case of participatory plant breeding 625
of the organizational context in this region At first, two CIALs were formed;
(the Calico river catchment) revealed on the over the years the number grew rapidly.
one hand that very little formal agricultural Several committees have since moved
research was carried out in the area or that on to experimenting on a larger scale,
the results of research carried out elsewhere addressing new aspects of problems in their
(by the NARS) reached the area; on the communities, such as soil fertility. A number
other hand, it was learned that farmers of new farmer-leaders have emerged,
themselves were not known to experiment including several women. Where possible,
very widely. CIALs are linking to each other to exchange
The CIAT team hypothesized ideas and results within the catchment and
therefore that there would be space for the beyond, through participation in the annual
implementation of the CIAL methodology CIAL meetings in Honduras, for example.
in terms of providing a tool for farmer They also are building bridges to formal
experimentation and strengthening of the research and technology organizations in
organizational processes in the area. CIALs the country.
could create new groups or could build upon The Honduras and Nicaragua
existing groups, and introduce new roles experience suggest that positive change is
in the community, such as by providing possible despite very difficult institutional
a service through doing research for and contexts. Through sustained efforts,
with the community, opening the door new organizational forms can emerge, a
to participatory decision-making, problem demand-driven research process can be
diagnosis and experimental design, and set in motion, and useful linkages can be
establishing new communication patterns developed with and between local, farmer-
among farmers and between farmers and led initiatives and national or international
external agencies, such as through CIAL- units, expertise and resources. These
led presentations, field-days, and direct changes do not come about easily, and
demand for support directed to outside set-backs have been numerous. However,
agencies from the NARS. the CIALs are contributing to revitalizing
The CIAT team also thought that CIALs rural innovation and to defining many new
could be players in changing the very much rules for the research and development
supply-driven mode of operations of most game.
NGOs in the area into a more demand-
driven one, as well as getting government 23.10 SYNTHESIS
agencies and universities interested in the This chapter has addressed three interrelated
area and problems and needs of farmers. elements of the division of labour in
The core idea behind the CIAT team efforts participatory plant breeding: the roles of the
to initiate a process of CIAL formation was people involved, the nature of the research
to provide local communities with a (new) management process, and a number of
way to carry out research collectively, institutional issues that influence the space
focusing on and solving a locally felt for doing things differently. Underlying
natural resource management problem these three elements is the need to pay
(to be identified through participatory attention to:
problem analysis), and thus further Getting to know the various people
enhance local organizational capacity. involved and building trust.
Towards new roles, responsibilities and rules: the case of participatory plant breeding 627
Getting to understand and respect differ- policy-makers. These efforts tell us that
ent perspectives, interests and expertise. organizational and institutional questions,
Bridging these perspectives, interests and such as those addressed here, are central to
expertise through an interdisciplinary, (participatory) plant breeding, deserving as
iterative, learning-by-doing approach. much attention as more technical issues.
Acknowledging the very real personal,
social and institutional constraints to col- ACKNOWLEDGEMENTS
laboration, and actively finding ways to We acknowledge the contributions to the
overcome them. research and insights presented here made
Communicating clearly and in a timely by the many men and women farmers with
manner. whom we have or have had the opportunity
Finding common ground through to work in countries around the world, as
deliberate planning, monitoring and well as the support received from numerous
evaluation efforts. colleagues.
Defining tasks to be accomplished,
jointly and up front, and agreeing on REFERENCES
who will do what and when, e.g. setting Horton, D., Alexaki, A., Bennett-Lartey, S.,
objectives; selecting germplasm to be Brice, K.N., Campilan, D., Carden, F., de
used; choosing breeding, propagation Souza Silva, J., Le Duong, T., Khadar, I.,
and selection methods; selecting sites Boza, A.B., Muniruzzaman, I.K., Perez, J.,
where the research will be carried out; Somarriba Chang, M., Vernooy, R. & Watts,
identifying the type of end-product to be J. 2003. Evaluating capacity development:
produced; and the means by which the experiences from research and develop-
product(s) will be distributed. ment organizations around the world. The
Recognizing the time and effort that Hague, ISNAR; Wageningen, Netherlands,
any change process requires, including CTA; Ottawa, IDRC.
the often very slow pace of change in Humphries, S., Gallardo, O., Jimnez, J.,
everyday life. Sierra, F., with members of the Association
In other words, recognizing the need to of CIALS of Yorito, Sulaco and Victoria.
explore the practical meaning of partici- 2005. Linking small farmers to the formal
pation, its potential and its limitations. research sector: lessons from a participatory
PPB experiences to date, including those bean breeding programme in Honduras.
documented in this chapter, suggest that AgREN Network Paper No. 142. London,
significant progress has been made in terms ODI.
of the development of an alternative and Joshi, K.D, with Rana, R.B. & Subedi, A. 2001.
complementary approach. This has not Farmer and researcher contributions to the
been without difficulties, constraints and selection of landraces of Ghaiya (upland
setbacks. New challenges, such as scaling rice) for Tar areas of Nepal. Pokhara, Nepal,
up (e.g. institutionalization) and scaling out LI-BIRD.
(e.g. application and adaptation to more Rling, N. 2000. Gateway to the global garden:
favourable production environments), have Beta/Gamma science for dealing with eco-
emerged and are now being researched logical rationality. Eighth Annual Hopper
in a number of countries, involving Lecture, University of Guelph, Guelph,
farmers, researchers, extensionists and Ontario, Canada, 24 October (available
628 Plant breeding and farmer participation
CHAPTER 24
In this chapter we shall first introduce A unique feature of the protection system
the different means to protect intellectual in the United States of America is that it
property. We shall then discuss international provides two forms of sui generis protec-
treaties and conventions providing the tion (PVP protection and Plant Patent pro-
frames for intellectual property legislation of tection). The Plant Patent Act was enacted
the member countries. We shall also shortly in 1930. A plant patent may be granted to
discuss the alternative ways of protecting new and distinct plant varieties that are
intellectual property by contracts, material invented or discovered, although excluding
transfer agreements (MTAs) and physical tuber-propagated plants and plants found
means to prevent unauthorized use of in uncultivated areas. Plant patents are
improved germplasm. issued for 20 years from the date of filing.
remarkable advantage of a trademark is that the state of Idaho or Roquefort for the
it is valid as long as it is in use in commerce. specific kind of French cheese.
When the limited protection time of a patent The Agreement on Trade-Related
or plant variety protection expires, a trade- Aspects of Intellectual Property Rights
mark can still be used to inform the custom- (TRIPS Agreement) provides a high level
er of the specific qualities of the product. of protection of geographical indications
Outside of breeding industry, Kodak is for wines and spirits. Currently, extension
again a well known example: the patent right of this high level of protection to other
of the regular black and white film of Kodak products, such as agricultural products of
expired about a hundred years ago, but still developing countries, is under discussion in
everybody knows exactly what they buy the World Trade Organization (WTO).
based on the strength of the trademark. Note that geographical indications,
akin to trademarks, do not protect the
24.2.5 Geographical indications information embodied in the goods nor
Geographical indications are a kind of IP any method of producing or processing
that has already been in use for rather a long the goods. Rather geographical indications
time, but has been widely recognized as a are rewarding groups of people that have
means of protection only recently. A geo- developed a product, often over centuries
graphical indication is a sign used on goods of collective knowledge. Accordingly,
that have a specific geographical origin geographical indications are considered as
and which possess qualities or a reputation a part of wider policy to award protection
that are linked to the place of the products for indigenous knowledge.
origin. Geographical indications serve as
assurance of source or quality, and they are 24.2.6 Trade secrets
important in sense similar to a trademark. Trade secrets are probably the oldest and
In various countries, protection for the cheapest way to protect ones IP: having
geographical indications is provided under a trade secret simply requires as the term
different concepts: geographical indications indicates, that the IP is kept secret. A trade
may be protected under laws against unfair secret could for example be a composition of
competition, consumer protection laws, a culture medium or a method to transform
or laws for the protection of certification a plant species. A typical trade secret in
marks. In some countries there are special the context of plant breeding is having the
laws for the protection of geographical parent lines of a hybrid variety kept secret.
indications. In the United States of America, The positive aspect in trade secrets, in
geographical indications are treated as addition to its low cost, is that there is no
trademarks. expiration date. However, the negative side
Most commonly, a geographical is that once the secret is out, the protection
indication consists of the name of the is gone and anyone is free to use the know-
place of origin of the goods. Agricultural how.
products typically have qualities that derive
from their place of production and are 24.3 PLANT BREEDERS RIGHTS UNDER
influenced by specific local factors, such as THE UPOV CONVENTION
climate or soil. Examples of geographical As the UPOV Convention provides the
indications are Idaho for potatoes from framework for the most common and well
634 Plant breeding and farmer participation
known of the sui generis systems for Plant Even though the list covers the most
Breeders Rights we shall discuss the con- important crop species, a large number of
vention in more detail here. species cannot be protected in China.
The International Union for the PBR under the UPOV Convention
Protection of New Varieties of Plants provide the breeder of a distinct, uniform
(UPOV) was established in 1961, and since and stable variety with an exclusive right for
then the provisions have been revised in a limited period for multiplication, offering
1972, 1978 and 1991. UPOV is a separate for sale, selling, exporting, importing and
intergovernmental organization and is stocking for these purposes. These breeders
partially monitored by the World Intellectual rights do however not extend to acts done
Property Organization (WIPO). Currently for non-commercial or experimental
UPOV has 58 Member countries, 25 of purposes, nor for purposes of breeding of
which are bound by the UPOV Convention new varieties. In other words, the UPOV
of 1978, 31 by the Convention of 1991 and Convention provides protection to distinct,
2 by the Convention of 1961/1972. All the uniform and stable varieties but also leaves
important agricultural producer countries certain exemptions for further breeding and
are members of UPOV. More than half non-commercial purposes.
of the member countries are developing
countries. 24.3.1 Comparison of the UPOV
The goal of the convention is to Conventions of 1978 and 1991
provide an incentive to breeders to develop Because most of the member countries are
new varieties for the benefit of society bound by the UPOV Conventions of 1978
by granting a limited monopoly to the or 1991, we briefly compare the minimum
breeders to commercialize new varieties. requirements set forth in these two
The Convention requires granting of conventions and discuss their implications.
protection for the varieties of all plant Both of the Conventions require the
genera and species. New member countries variety to be distinct, uniform and stable
of the Convention of 1991 must provide (DUS) before protection can be granted.
protection to at least 15 plant genera, and DUS-testing is mainly based on growing
within ten years from joining UPOV they tests carried out by the competent
have to provide protection to all genera. The authority of the member country where
fact that a number of important countries, protection is sought. The Convention of
such as former Soviet Union countries, have 1991 additionally requires that the variety
joined UPOV only during the last years be novel, meaning that it has not been
means that there is currently a situation sold or commercially exploited for more
where not necessarily all plant genera and than a year in member countries where an
species can be protected in these countries. application has been filed, and not been
Contrary to the requirement of all species sold in a non member country where an
being protectable, the Convention of 1978 application for variety protection is filed
requires protection of at least 24 species. An for more than 4 years (for 6 years in case
example of a 1978 Convention member is of trees and vines) before the application in
China, which became a member in 1999 and the member country.
has currently a national list of protectable The Convention of 1978 protects
species containing 41 agricultural species. commercial use of reproductive material of
Breeders rights and IPR issues 635
the protected variety, while the Convention An essential and much discussed issue in
of 1991 protects varieties and products, UPOV is the concept of Farmers Rights or
including those that are essentially derived. Farmers Privilege. Traditionally, farmers
The essential derivation provision is a similar were free to save, re-use and sell harvested
concept to doctrine of equivalence in the seeds. UPOV has brought some limitations
patent laws, aiming to prevent plagiarism. to these rights. The Convention of 1978
Essential derivation is a concept that has did not include any specific requirements
recently created a lot of discussion and for Farmers Privilege. This means that the
therefore we shall return to it later in this farmers were left with the right to save and
chapter. re-use harvested seeds of a protected variety.
Of note is that UPOV 1978 restricts The United States of America implemented
the countries where both patent and sui the Farmers Rights of UPOV 1978, so that
generis protection are available to grant the farmer was allowed not only to save
only one type of protection for one and but also to sell the saved seeds, as long as
same botanical genus or species. The the income from the saved seed was less
Convention of 1991, however, does not than 50 percent of the total income of the
include this restriction. Accordingly, in farm. Now, as the United States of America
the United States of America, which is is a member of 1991 UPOV, the farmer
a member of 1991 UPOV, the Supreme may no longer sell the saved seed, but
Court ruled in 2001 in J.E.M. AG Supply v. does not need to pay royalties on re-used
Pioneer Hi-Bred International that a plant seeds. The European Union implements the
breeder can obtain multiple protection for 1991 Convention by allowing small-scale
newly developed plant varieties; having a farmers to save and re-use seed without
sui generis based variety protection does royalty payments, while re-use of seed of
not exclude issuance of utility patent for large-scale farmers is subject to reasonable
the same if requirements for novelty, non- royalties, which usually is 50 percent of
obviousness and usefulness are fulfilled as regular royalty rate. Colombia, which is
required for patents in the United States a member of Convention of 1978, but has
(J.E.M. AG Supply v. Pioneer Hi-Bred rules mostly according to Convention of
International, 122 S.Ct. 593, 2001). 1999, allows farmers having less than 5 ha
The Convention of 1978 gives a 15-year to save seed (Louwaars et al., 2004).
protection from filing date for crops and 18
years for trees and vines. The most recent 24.4 INTERNATIONAL GOVERNANCE
Convention, of 1991, grants 20 years of OF INTELLECTUAL PROPERTY
protection from filing date for crops and 25 IPRs are based on national legislation and
years for trees and vines. therefore the rights are usually territorial,
The Convention of 1978 grants so-called so a patent is valid only in the country of
Breeders Exemption, which means that the jurisdiction that granted it. However,
breeders are allowed to use the protected during this era of globalization, there are
material, without a licence, to breed new several international treaties and conventions
varieties. In the 1991 Convention, Breeders that are setting global frames for the IP
Exemption is optional and it is up to the legislation of the member countries. Below
national government to implement legisla- we review the treaties most relevant to
tion that respects Breeders Exemption. plant breeding.
636 Plant breeding and farmer participation
the member countries do not issue patents leaves us with the translation from Latin
for plants. However, even if plant varieties being specific or of its own kind. By
may be excluded from patentability, in giving no definition to this essential term
practise there still may be a way to get patent means that member countries are left with
protection for plants: the European Patent free hands to fashion their own protection
Convention for example does not allow system. The UPOV convention is one
individual plant varieties to be patented, interpretation of what a sui generis system
but the Board of Appeals of the European can be.
Patent Office ruled in 2000 in Novartis Another essential term not defined
v. Plant Genetic Systems that genetically in the TRIPS Agreements provision for
modified plants may be protected if the protection of plants is the term effective.
invention is not limited to a single variety How effective does an effective sui generis
(Novartis v. Plant Genetic Systems, G1/98 system need to be? To clarify the meaning
Transgenic Plant/Novartis II OJ EPO 2000). of the clause several countries are calling
Here, clearly, the interpretation of the law for further discussion on Article 27(3) of
provides patent protection to genetically the TRIPS Agreement. Among others it
modified plants. Canadian patent law has been proposed that the interpretation
excludes plants, as higher life forms, from should extend the protection to traditional
patentability. However, in a recent case, the and indigenous knowledge. The discussion
highest court in Canada found that growing on Article 27(3) of the TRIPS Agreement is
transgenic plants containing a patented gene connected to the relationship of the TRIPS
infringed a patent that claims the chimeric and the Convention on Biological Diversity
herbicide-resistance inducing gene and cells (CBD), and is covered in Section 24.4.3.
containing that gene (Monsanto Canada WTO member countries have to have
Inc. v. Schmeiser 2004 SCC 34). Therefore, their IP laws in line with the TRIPS
even if Canadian law does not allow requirements. When the TRIPS Agreement
patenting of higher life forms, this decision came into effect in January 1995 it set out
implies that patent protection in Canada is transitional periods for implementation for
extended to plants if a gene present in the developed, developing and least-developed
plants genome is claimed in the patent. The countries. Developed countries had to
rational behind this is that by growing the comply with TRIPS provisions by 1996,
plant that expresses a patented gene, one is while the least-developed countries had
using the patented invention. until the beginning of 2006. Developing
countries generally had until 2000 for the
Plant breeders rights under the TRIPS implementation, but the deadline was later
Agreement postponed until 2005.
The TRIPS Agreement provides that: India is an example of a developing
members shall provide for the country that established its PVP legislation
protection of plant varieties either by in order to comply with the TRIPS
patents or by an effective sui generis requirement. India enacted its plant variety
system or by any combination thereof. protection laws in 2001. India has chosen sui
(Article 27.3 (b)). generis legislation deviating from the norms
The TRIPS agreement does not give set by UPOV (see Sahai 2003; Brahmi,
any definitions for the term sui generis but Saxena and Dhillon, 2004). The effects of
638 Plant breeding and farmer participation
the legislation remain to be seen after it is ing party providing the recourses. These
effectively implemented. requirements have induced vast discussion
and still unsolved questions of the compat-
24.4.2 The Convention on Biological ibility of TRIPS and CBD.
Diversity
The Convention on Biological Diversity 24.4.3 Relationship of the TRIPS
(CBD) was established in 1992 as an outcome Agreement and CBD
of the United Nations Conference on CBD and the TRIPS Agreement approach
Environment and Development (UNCED). the subject of IP protection from different
Currently the CBD has 168 signatories. Of perspectives: CBD has a focus on sustainable
note is that the United States of America management of biodiversity, while TRIPS
has signed but not ratified the Convention. aims to provide a framework for adequate
The main objectives of the CBD are protection for IPR to reduce distortion
to ensure conservation of biological of international trade. The relationship of
diversity, to ensure the sustainable use of the TRIPS Agreement and CBD has been
biological diversity, and to promote a fair widely debated in the TRIPS Council. A
and equitable sharing of the benefits arising concern has been raised that implementation
from the utilization of genetic resources of the TRIPS Agreement may affect the
amongst member countries. ability of the WTO member countries to
The CBD does not as such elaborate fulfill their CBD commitments.
on IPRs, but it makes a clear statement on Some developing countries have been
technology transfer as an important means to arguing against granting patent rights
reach the goals of the Convention. Because for genetic material as is possible under
much of the agricultural technology in the TRIPS, because that might limit access to
developed countries is protected by IPRs, the resource and equitable benefit sharing,
the statement of technology transfer being as required by CBD. Some countries have
an essential means to reach the goals of the required that patent applications should
Convention means that IPRs become an be accompanied by disclosures regarding
issue as well. Article 16.2 of the Convention source of origin, any related traditional
states that knowledge and evidence of equitable benefit
In the case of technology subject to sharing. Counter arguments have included
patents and other intellectual property notation that such requirements would limit
rights, such access and transfer shall availability of protection and this again
be provided on terms which recognize would violate the principles of the TRIPS
and are consistent with the adequate agreement. Furthermore, additional require-
and effective protection of intellectual ments probably would make the system
property rights. expensive and complicated to implement.
Thereby, CBD clearly recognizes IPRs. Due to these concerns, several countries
The Convention requires equitable shar- are calling for amendments to be made to
ing of benefits arising from commercial use the TRIPS Agreement to bring it into the
of the biological resources and local knowl- line with CBD. At the same time, both
edge of communities. The Convention also the TRIPS Agreement and CBD are rather
requires that access to genetic resources is flexible in their language, and therefore the
subject to prior consent of the contract- member countries have a lot of freedom
Breeders rights and IPR issues 639
to find ways to implement both without The language of the article has been
conflict. regarded as ambiguous as it is not clear
whether, for example, isolated and purified
24.4.4 The International Treaty on compounds or gene sequences are patentable
Plant Genetic Resources for Food and under this provision or not (Lettington,
Agriculture 2004). Currently parties to the treaty can
The International Treaty on Plant Genetic interpret this provision rather freely, which
Resources for Food and Agriculture means that the MTA may have different
(ITPGRFA) was agreed in June 2004. It meanings in different countries, depending
provides for a multilateral approach to on the national legislation.
access and to benefit-sharing of a selected The ITPGRFA recognizes Farmers
list of plant genetic resources for food and Rights to freely access genetic resources,
agriculture. The list includes 35 crop genera and to use and save seed. However, the
and 29 forage species. Ex situ collections of implementation of Farmers Rights is
these crops are held by the International left fully to national governments. An
Agricultural Research Centers (IACRs). implication of this is that the member
The species in the list, even if not so many, countries of the treaty have to consider
provide about 80 percent of the worlds the relationship of Farmers Rights to
food calories from plants. the already existing IP laws. The member
The goals of the Treaty are very similar countries might, for example, already have
to the CBD, but ITPGRFA specifically provisions for Farmers Rights in their
addresses access to and fair sharing of the plant variety legislation. At the same time,
benefits generated from the commercial member countries may end up protecting
utilization of the genetic resources of the some aspects of Farmers Rights through
listed species in the food and agriculture other legislation, such as laws regulating
industries. It thereby leaves utilization of commerce in seeds.
the genetic resources in the pharmaceutical
industry out of its scope, while CBD 24.5 CURRENT ISSUES IN IPRS AND
encompasses use of genetic resources in any PLANT BREEDING
field of technology. The central mechanism 24.5.1 Access to germplasm
to implement the provisions for access Plant and animal breeding is different from
and benefit sharing is a standard material any other field of technology in the sense
transfer agreement (MTA). that it is impossible to make progress in
The draft MTA attached to the terms of inventions without having access to
ITPGRFA contains the language of Article prior art. A mechanic can invent something
12.3(d) of the treaty, which has raised a lot that provides a huge technical step forward
of discussion. Article 12.3(d) states that without having the slightest idea of what is
Recipients shall not claim any IP or already out there. Opposite to this, a plant
other rights that limit the facilitated breeder can breed a better variety only by
access to the plant genetic resources having access to germplasm. Despite this
for food and agriculture, or their essential characteristic of the art of breeding,
genetic parts or components, in the inventions related to plant breeding may
form received from the Multilateral still be protected by various forms of IPRs
System. in a way similar to inventions in mechanics.
640 Plant breeding and farmer participation
This is an issue that is raised time after time, whole community of breeders. This also
because of concern that IPRs might prevent helps to keep the genetic basis for plant
free access to germplasm and thereby affect breeding as broad as possible and minimize
the capacity to breed, research and provide the threshold for access to germplasm.
better varieties for food and feed. The language in the 1978 UPOV
International treaties have provisions Convention has been interpreted to allow
that are aimed to ease access to germplasm. cosmetic modifications in breeding new
As discussed above, ITPGRFA provides varieties, such as inducement of mutations in
for ex situ collections of most important ornamental plants. Development of methods
food and feed plants. The International for genetic engineering has brought further
Agricultural Research Centers (IARCs) of prospects of rapid modification of existing
the Consultative Group on International varieties. In order to prevent protection of
Agricultural Research (CGIAR) hold over new varieties with only minimal changes
600 000 accessions of crop, forage and compared with the original variety without
agroforestry genetic resources. ITPGRFA recognition of the breeder of the initial
requires a standardized MTA to guarantee variety, the 1991 UPOV Act amended the
that no IPRs shall be claimed for material concept of essential derivation.
received from the system. The goal of the The core of the essential derivation concept
treaty is to provide fair exchange of germ- is that the scope of the Breeders Rights
plasm of the species included in the list. is extended to varieties that are essentially
Wild germplasm, in contrast, is an derived from the original breed. Essentially
important part of the art of plant breeding, derived varieties may be obtained in various
and wild germplasm might not be represented ways. The UPOV 1991 Convention gives
in genebanks alongside cultivation-based a list of methods, including selection of
germplasm (Gepts, 2004). This argument natural or induced mutants, selection of
would inevitably lead to a very broad and a somaclonal variant, selection of variant
still unsolved issue of compatibility of individual plants in the initial variety,
existing plant IP system with the rights of backcrossing and genetic engineering.
indigenous peoples traditional knowledge Through this concept, if a breeder derives a
which issue has been recently discussed in variety essentially from another variety, such
Fingers and Shuler (2004). as inserting one new gene into the initial
variety, the new variety can be protected if
24.5.2 Breeders exemption in PBRs, it is new, distinct, uniform and stable; but
and essential derivation for as long as the initial variety is protected,
The 1978 UPOV Convention provides that the essentially derived variety cannot be
a protected variety can be freely used as an exploited without authorization from the
initial source of variation for the purpose of owner of the initial variety. In practice, this
creating other varieties, and that the breeder means that the breeder of an essentially
shall not be required to obtain authorization derived variety would need a licence from
for marketing such varieties. This provision the breeder of the original variety. If the
is known as Breeders Exemption and it is essentially derived variety is derived from a
a fundamentally important part of PVP. public-sector-bred variety, there is naturally
Breeders Exemption guarantees that the no need for a licence as the original variety
germplasm sources remain accessible to the was not protected.
Breeders rights and IPR issues 641
GoldenRice, the vitamin Arich transgenic the frame very loosely, it remains for the
rice, have been patented; a freedom to countries to decide how the protection
operate study showed that there are more is to be implemented. Some developing
than 70 patents related to the technology countries have chosen to adhere to the
(Kryder, Kowalski and Krattiger, 2000). UPOV Convention; others, such as India,
However, in most of the countries where are going to implement more liberal PVP.
rice is an important commodity, none or The Farmers Right provision of Indias
only a few of these patents were in force. Plant Variety Protection and Farmers
In such countries, using or developing the Rights Act of 2001 has created a lot of
technology further is legally completely discussion, because it seems to differ from
correct. Issues may arise only when there is anything that has been created under the
trade in the technology to countries where UPOV Conventions. The Indian law
patents or other forms of protection are in allows farmers to save, use, sow, re-sow,
force. exchange, share or sell the seed, providing
According to Pardey et al. (2003) there that the farmer shall not sell the saved seeds
is still a substantial freedom to operate in any packages or containers labelled in
for most crops of major significance for a manner indicating that the seeds are
food security in poor countries. Pardey protected (Sahai, 2003). It has been argued
argues that concern of freedom to conduct that Farmers Rights provisions as liberal
research by or on behalf of developing as Indias does protects only the brand of
countries is seen as a way to draw attention the breeder. In the Indian Act, there are
away from real constraints. Real constraints also provisions for acknowledging the role
according to the same authors are lack of of rural communities as contributors of
investment in developing country research landraces and farmers varieties. A breeder
and lack of scientific skill to access modern wanting to breed an essentially derived
technology, whether protected or not. variety needs to have permission of the
Koo, Nottenburg and Pardey (2004) communities (Sahai, 2003). The Act adopts
show that from 2000 to 2002, 54 percent all the suggestions of the UPOV 1991
of the variety protection applications filed Convention as to the methods that may be
worldwide were filed in Europe or in the used to breed essentially derived varieties,
United States of America. The principal in effect almost all the modern means of
reason for the lack of filing activities in the plant breeding. This leaves the Breeders
developing countries is a lack of established Exemption extremely narrow.
protection means. At the same time, this The International Association of Plant
data also indicates that a claim that IPRs Breeders for the Protection of Plant
are limiting the freedom to exploit plant- Varieties (ASSINSEL) suggests that any
science-related inventions in the developing national legislation authorizing farm-saved
countries is an overstatement. seed without reasonable limit and without
safeguarding the legitimate interest of the
24.5.5 Farmers Rights breeders is not in conformity with the
Developing countries are required to 1991 UPOV Convention. Additionally,
introduce some form of plant variety ASSINSEL argues that any such legislation
protection under the TRIPS Agreement. would be contrary to the meaning of the
However, as the TRIPS Agreement sets TRIPS Agreement, i.e. such a system would
Breeders rights and IPR issues 643
not provide effective sui generis protection enforced; simply offsetting the value of
(ASSINSEL, no date). losses incurred by the innovator is not
The consequences of farm-saved seed severe enough a punishment.
to the breeder depend also on the contract
that the breeder makes with the farmer. 24.5.6 Other methods to protect
If the farmer pays royalties based on the unauthorized use of seeds
amount of seed originally purchased, then Contracts and MTAs
farm-saved seed naturally reduces the A breeder can control their rights over the
earnings of the breeder. Another option for material they own by contractual agreements,
the breeder is to collect royalties as end- including MTAs, which are binding legal
point royalties when the harvested crop is contracts between the technology provider
sold. By collecting end-point royalties, the and the receiver, and the most common
breeder would benefit from the farm-saved legal documents controlling use of research
seed provision provided that the farmer material. The terms of MTAs can go far
declares that the seeds they sell is of the beyond the rights provided by a patent or
protected variety. Contracts may also be other IP legislation. The MTA may include
used to oblige the farmer to keep a record so-called reach-through clauses, whereby
of their practices. Such contracts would the technology provider may get rights
help the breeder to monitor the practice of to new varieties or other inventions and
the farmer in end-point royalty cases and improvements that have been created by
would ease collecting royalties. using the material provided through the
However, not only the implementation MTA. The receiving party has to be clear as
of the law in a country is important but to what the implications are of signing an
the enforcement is as important, if not MTA before signing.
even more important. This of course means When selling seeds of a protected or a
not only enforcement of Farmers Rights, non-protected variety, a breeder may con-
but also every aspect of the IP-laws of trol the use by various kinds of contract. We
the country. Lack of enforcement of IP discussed earlier how the breeder may con-
laws may lead to a situation such that of trol income by choosing the royalty basis
Argentina, where 2550 percent of Roundup defined in the contract with the farmer.
Ready soybean seeds grown are from Two specific types of contract of
black market sources or saved by farmers significance in the seed industry are the so
from the previous years crop (Robertson, called shrink-wrap and brown-bag licences.
2000). Similarly Kowalski (2003) is worried Typically, a breeder includes in the seed
about the future of agribiotech in China package contractual language limiting the
due to weak enforcement of the IP laws. rights of the buyer. The seed bag may for
Lack of enforcement leads to lower prices example specify that the material inside
and eventually leads to unwillingness of the bag may be not be used for further
companies to invest in countries having breeding. By opening the package or by
weak enforcement of IP laws (Giannakas, planting the seeds the user agrees with the
2003). This author suggests that penalties contractual language on the package.
determined under the TRIPS Agreement By these contractual means, the breeder
have to go beyond the norms of GATT, can regulate the use of the material, even
otherwise IPRs remain inefficiently in countries without no IP legislation.
644 Plant breeding and farmer participation
However, as both MTAs and brown bag be a method to protect varieties in countries
licences are interpreted under the contract where only weak IP protection is available
laws of the country, the enforceability of or where IP law enforcement is weak. The
such means differs between countries. proponents also argue that transaction costs
would be lower if there is no need for IP
Biological methods to control re-use of protection and therefore the benefit would
seeds come to consumers and farmers through
The modern technologies developed in the reduced seed prices. Opponents of GURTs
plant sciences provide certain methods to argue that GURTs will harm the farmers by
protect varieties from unauthorized use. taking away the ability to save and re-use
These methods include hybridization and seeds and will have adverse effects on food
technologies usually called Genetic Use security and biodiversity.
Restriction Technology (GURT). The GURT technologies are still under
Hybrid technologies were developed development and in many countries
in 1930s and today hybrid varieties have genetically modified crops are in any case
been developed for most of the important not yet in cultivation. Therefore there is
cross-pollinated food and feed species. no data that could prove either the fears
When hybrid seed is used for a second of the opponents or the hopes of the
generation, part of the hybrid vigour is lost proponents to be true. However, there are
and therefore saving seed for re-use is not already indications that countries may not
an optimal solution for a farmer. Rather the allow GURTs to be used in protectable
farmers are each year dependent on seed varieties: the Indian Plant Variety and
producers new seed. Farmers Rights of 2001 does not allow
GURT is a biotechnology application the protectable varieties to include GURT
of a system providing the breeder control technologies (Sahai, 2003).
over re-use of the seed. GURTs are not
specifically developed for the purpose of 24.6. PLANT BREEDERS RIGHTS AND
enabling plant breeders to prevent re-use PARTICIPATORY BREEDING
of the seed; rather the goal in developing Farmers have worked as plant breeders for
the techniques have been for purposes centuries, but have in most cases not sought
such as preventing transgene escape into to protect their new materials through
the environment. In any case, GURTs may any form of statutory protection. Where a
also be a strong tool for preventing re-use farmer can be identified as a breeder there
of seed. would be no inherent problem in seeking
Currently it seems that there is quite a protection, insofar as what is required for
lot of discussion of the possible effects of PBR is the identification of a breeder, and
this new technology in relation to farmers, in the case of a patent an inventor.
research and the environment (e.g. Budd, There are, however, a number of other
2004). Proponents of GURTS argue that elements to the current proprietary IPR
GURTs provide seed companies and plant protocols, in that farmers and communities
breeders with stronger control over plant are unable to meet some of the other
varieties. This would enable greater cost statutory requirements as they relate to
recovery and provide more incentive for novelty, the time of the invention, prior
the plant breeding industry. GURTs would sales, etc. There are also problems that in
Breeders rights and IPR issues 645
many locations are associated with the concept, such as the time the product
inability to identify a breeder or inventor or invention has been in the market or
per se since the breeding activity is carried whether there is prior publication of the
out at the community levelthere is no information on the invention or material.
longer a definable breeder or inventor. Clearly, in community-based schemes this
A final issue that has been of concern approach is hampered by the apparent
in the literature has been the fact that informality of the system. This situation
indigenous materials are often not is clearly crucial to the area of landraces,
characterized or published. This may lead where over many generations improvement
to perchance to another person breeding have been made, but such improvements
material with the same description, who have become public goods basically because
would be able to gain protection on such of time or sale of seed (even if informal).
material on the basis of claiming novelty
for the material in the absence of published 24.6.3 Cost of filing
information to the contrary. Clearly, in some cases, there are severe
Let us therefore address each of these constraints on small communities because
issues and indicate arguments that are of the cost of filing applications. This,
associated with each side of the issues. however, is no different from the problem
that faces individual inventors who want to
24.6.1 Determination of the breeder patent an invention. The crucial question to
or inventor ask here is why the person or community
In modern crop improvement we are wants to seek protection. If the goal is
accustomed to several people being to licence the materials for income, the
involved and collaborating on the inventive answer is to spend the money and seek
or breeding step. There would therefore be the protection. If the goal is to prevent
only limited problems with identification appropriation by others of the intellectual
of the inventive steps involved and the knowledge, then simply publish the data
persons involved, thus allowing for clear in written form, take your credit, and
identification of breeders or inventors. In while others can use the invention, they
the same way that breeders then assign their cannot gain any exclusivity or proprietary
material or invention to their institution, it protection for the material.
is clearly feasible for a community or tribe
to become the title holder to the invention 24.6.4 Decisions are needed
or material bred or invented within that From the above it follows that what is
community. What cannot happen is that the really needed are educational steps and
community or tribe is deemed as inventor clear understanding as to the goals and aims
or breeder since they do not fulfil statutory of those communities that are improving
requirements. germplasm. In short, what is it that the
communities are seeking? If the key is
24.6.2 Novelty recognition of input and prevention of
In order for PBR or a patent to be granted proprietary exploitation by others, then
there is a requirement of novelty (or publication of the data serves a vital
distinctiveness). The challenge there is purpose. There are no doubt international
that the statutes tie caveats to the novelty agencies that are willing to provide funding,
646 Plant breeding and farmer participation
either to communities to protect their new Trade and Domestic Production. Ravelo,
assets, or to assist with documentation and Italy, 811 July 2004.
publication. Fingers, J.M. & Schuler, P. 2004. Poor peoples
knowledge. Protecting intellectual property
24.7 CONCLUSIONS in developing countries. Co-publication of
The plant breeding industry has encountered World Bank and Oxford University Press.
changes during the last decades. New Gepts, P. 2004. Who owns Biodiversity, and
breeding methods have raised issues of IP how should the owners be compensated?
protection. The menu of the means that a Plant Physiology, 134: 12951307.
breeder may use to protect their invention Giannakas, K. 2003. Infringement of
is not limited to PVP but includes several Intellectual Property Rights: developing
other means as well. In addition to various countries, agricultural biotechnology, and
forms of IP, breeders may also use other the TRIPS Agreement. Biotechnology and
legal forms to control use of varieties. Genetic Resource Policies Briefs, No. 5.
There is an increasing amount of Washington, DC, IFPRI (available at www.
legislation and international treaties IFPRI.org/pubs/rag/br1001/biotechbr5.
regulating issues related to IP in plant pdf; accessed 25 Jan. 2009).
breeding. Nevertheless, a lot of the Goddar, H. 2001. The Experimental Use
decision-making is still left to national Exemption: a European Perspective. In
governments. K.M. Hill, T. Takenaka & K. Takeuchi,
There is certainly no one correct and eds. Reconciling Competing Interests in
acceptable system to be implemented in each Intellectual Property, Experimental Use
and every country to provide reasonable Exception, Validity of Patents, Attracting
rights for farmers, breeders and industry. Investment, and Collaboration Among
The international treaties set frames for Patent Offices, pp. 1013, Proceedings of the
minimal protection. The individual 2001 High Technology Summit Conference,
countries are left with rather a free hand in University of Washington, Seattle, USA.
tailoring their national IP laws. CASRIP Publication Series No. 7.
Koo, B., Nottenburg, C. & Pardey, P.G. 2004
REFERENCES Plants and intellectual property: an interna-
ASSINSEL [International Association of tional appraisal. Science, 306: 12951297.
Plant Breeders for the Protection of Plant Kowalski, S. 2003. Agricultural biotechnology
Varieties]. No date. Position Paper on Farm in China. An unreachable goal? The Journal
Saved Seed (available at www.Worldseed. of World Intellectual Property, 4: 655663.
org/Position_papers/FSSe.htm). Kryder, R.D., Kowalski, S.P. & Krattiger, A.
Brahmi, P., Saxena, S. & Dhillon, B.S. 2004. 2000. The intellectual and technical prop-
The Protection of Plant Varieties and erty components of pro-vitamin A rice
Farmers Rights Act of India. Current (Golden Rice): A preliminary freedom-
Science, 86(3): 292298. to-operate review. ISASAA Briefs, No. 20.
Budd, G. 2004. Are GURTs needed, to rem- Ithaca, USA.
edy intellectual property failures and envi- Lesser, W. & Murchler, M.A. 2004. Balancing
ronmental problems with GM crops? In investment incentives and social benefits
8th ICABR International Conference on when protecting plant varieties. Crop
Agricultural Biotechnology: International Science, 44: 11131120.
Breeders rights and IPR issues 647
CHAPTER 25
Jacqueline A. Ashby
650 Plant breeding and farmer participation
Key impacts of PPB are to produce on occasions proved a vital resource for
plant varieties that are well tailored to poor developing new crop varieties (e.g. Walker,
producers needs, to shorten the amount 2006). Nonetheless, in emerging markets,
of time plant breeding programmes need such as organic agriculture, PPB can
to get appropriate materials into farmers have advantages. For example, in France,
fields and so accelerate adoption and seed formally including producers in PPB for
dissemination. This is an impact on research organic agriculture has proved useful
efficiency related to improving the rate of for determining breeding objectives and
innovation overall. In some situations, PPB methodology (Chiffoleau and Desclaux,
helps to maintain or increase plant genetic 2006). This experience illustrates how PPB
diversity in farmers fields and improves may prove useful in the debate about the
agricultural sustainability. PPB carried welfare impacts of plant breeding in view
out with farmer groups improves farmers of consumer scepticism about genetically-
organizational and social capital, as well as modified (GM) crops, and concern about
individual farmers knowledge and skills how plant breeding affects food safety.
and capacity to learn and experiment: all PPB can promote informed participation
contribute to more resilient and sustainable and trust in research among consumers and
farming systems. In addition, PPB is expected producers.
to have welfare impacts by increasing poor In summary, impacts of PPB in
farmers access to improved varieties, their international crop improvement research
productivity, nutrition, marketing and are associated with improving research
incomes. Given the important role played relevance and efficiency via feedback
by women in managing plant genetic from farmers, traders and consumers, and
resources in many farming communities, the welfare impacts of a faster and more
PPB can affect gender equity. relevant supply of new plant varieties
PPB has evolved mainly to address the to small-scale producers. There is, in
difficulties of poor farmers in developing addition, the issue of the impact of PPB
countries. Widely seen as having advantages on the costs of research and innovation.
for use in low yield potential, high stress This is a complex issue, which is still
environments, PPB is most often applied relatively under-researched, but for which
when specific adaptation is sought. For there is some evidence, discussed later.
this reason, a review of plant breeding PPB may increase research costs compared
methodologies in the CGIAR recommended to experiment-station-centred breeding
in 2001 that it should form an organic part because it is typically decentralized and
of each Centers breeding program(TAC, requires work at multiple sites. At the same
2001: 24). However, some practitioners have time, after 23 years of cooperation with
results showing that both specific and wide a PPB programme, farmers increase their
adaptation are possible (see for example, capacity to manage varietal evaluations and
Joshi, Staphit and Witcombe, 2001). trial plots independently, and may assume
In industrialized agriculture, where some of the costs of adaptive research.
wide adaptation is prized and markets drive In this situation, a criticism of PPB is
demand for research, PPB maybe less useful that overall breeding programme costs are
from a research efficiency perspective, reduced but farmers costs go up. However,
although farmers local knowledge has the benefits of PPB to farmers include a
652 Plant breeding and farmer participation
reduction in the risk of productivity and preferred cultivars. PPB can then use as
income losses from planting ill-adapted, parents cultivars identified by PVS (see, for
poorly performing varieties in their fields example, Witcombe, Joshi and Staphit, 1996;
(a common experience for poor farmers Witcombe et al., 2005). In practice, PPB is
receiving experiment-station-centred a continuum of practices and differences
recommendations). Moreover, if PPB places from PVS are not rigid (Morris and Bellon,
improved varieties and seed in farmers 2004). Several programmes use a mixture of
fields more quickly than other approaches, approaches, often because practice evolves
then farmers income stream from new over time as breeders and the participating
varieties will increase sooner. These gains farmers learn how PPB works.
must be factored into the overall cost Discussion of the impacts of PPB
benefit assessment of PPB. requires a clear definition of what is meant
Before we can draw conclusions about by participation, because this term is
the impact of PPB it is essential to make loosely applied to a diversity of approaches
some important distinctions among the dif- for involving farmers in plant breeding,
ferent types of participation used in plant and various types of participation have
breeding, and that is the topic of the next different impacts. Participation refers to
section. the relationship between producers and
breeders, well recognized as a critical factor
25.2 WHAT DOES PARTICIPATION in many successful breeding programmes
MEAN IN PARTICIPATORY PLANT (Walker, 2006). One dimension of this
BREEDING? relationship can be defined by the use of a
The term participatory plant breeding functional or an empowering approach to
(PPB) is used in this discussion to refer participation. Functional and empowering
to the entire process of setting breeding approaches to PPB can be thought of as
objectives, making crosses, developing opposite ends of a continuum in the degree
finished varieties and their release up to of participant empowerment. Functional
and including the supply of basic seed participation in plant breeding improves
to growers. For the purposes of impact research efficiency by involving prospective
assessment, PPB refers to the full spectrum users of the results (farmers, intermediaries,
of breeding activities, including participatory traders, industries and consumers) in
varietal selection (PVS), much in the same prioritizing and evaluating traits important
way that trials evaluating finished varieties to them, such as plant architecture, market
are generally understood to form part of a appeal, storage and cooking quality.
breeding programme. Some PPB specialists Functional approaches tend to leave the
distinguish PPB as a breeding programme balance of power in decision-making in the
that includes farmers making crosses, as breeding process essentially unchanged, i.e.
distinct from one in which breeders use plant breeders (and their employers) make
farmers suggestions or preferred local most of the critically important decisions.
varieties to make their own crosses. They Empowering participation changes the
use the term PVS to refer exclusively to the balance of power in decision-making in
participation of farmers in the evaluation of the breeding programme, usually in favour
finished varieties and have demonstrated that of giving non-research interest groups a
PVS is a rapid way of identifying farmers more important role in key decisions about
The impact of participatory plant breeding 653
the end product, as well as in how the education in particular uses discovery
research is carried out. An empowering learning because adults learn better when
approach to farmer participation frequently they uncover concepts and facts themselves
alters breeding objectives and procedures, than when they are told about them.
including the environments and cultural Especially in agriculture, discovery learning
practices used to screen varieties. This involves farmers in running on-farm
leads to different results, as discussed in experiments very similar to the varietal trials
more detail below (Okali, Sumberg and used in participatory breeding. The key
Farrington, 1994; Ashby, 1996). difference is that the participatory learning
A similar distinction is made between facilitator knows ahead of time what the
PPB carried out in formal plant breeding experiments will show and, indeed, has
research programmes versus farmer-led designed the experiments to demonstrate
programmes. Farmer-led PPB is typically a known practice or principle. Because
run by NGOs, and in several cases involves participatory learning for agriculture uses
extensive research, but has different experiments, it is easily confused with
objectives from public sector crop participatory research. PPB demonstration
improvement research. These may include trials use participatory learning to share
community empowerment, biodiversity existing knowledge about varieties. PPB
conservation, disaster relief or skills research experiments combine farmers and
development (McGuire, Manicad and breeders ideas to test jointly conceived
Louise, 2003). By definition, farmer-led hypotheses and co-produce knowledge
programmes aim to empower farmers, but in that is new to all concerned. Indeed, the
practice can employ as varied or narrow a performance of varieties grown on small
mix of types of empowering and functional farms in marginal environments is often
approaches to participation as formal so unpredictable that programmes starting
breeding programmes. out with a participatory learning focus
A more important distinction is between find themselves drawn inexorably into
participatory research and participatory participatory research, because their varietal
learning. Participatory research in agriculture demonstration trials did not produce the
is conducted to investigate questions for results expected.
which neither scientists nor producers have Participation in plant breeding research
an agreed explanation. Like all research, (and in research generally) is based on the
it involves risk and uncertainty about principle that participation of end users in
the outcomes of experimental treatments the co-production of knowledge generates
and it combines use of scientific method a higher level of understanding, ownership
with native empiricism. The result is new and trust in the information, and increases
knowledge, usually a blend of scientific and their capacity and willingness to make use
indigenous. The impact of PPB on this co- of it. All actors involved in PPB research,
production of new knowledge may increase including the scientists, have hypotheses
as the level of farmer-scientist cooperation but no a priori certainty of what results
and farmer empowerment increases. will be obtained. The experimental process
In contrast, participatory learning uses is undertaken in conditions of mutual
principles of discovery learning to promote uncertainty and shared risk. PPB research
sharing of established knowledge. Adult typically involves cooperation between
654 Plant breeding and farmer participation
farmers and scientists in one or all of the independently of farmers, then farmers
following: establishing breeding objectives; are limited to a participatory learning
identifying desirable traits so as to design role. The important question is: does this
plant ideotypes; selection of parents; difference affect the recommendations and
selection in early generations; and screening the eventual impact of PPB? The chapter
of advanced lines. Scientists and farmers will return to this question when evidence
bring very different kinds of complementary of PPB impact is analysed.
knowledge and expertise to PPB research,
but they have a common goal of testing Different types of participation
hypotheses to answer questions to which PPB impacts are likely to vary depending on
neither know the definitive answer. Plant the type of participation used and whether
breeding programmes also use participatory or not the primary objective of participation
learning to demonstrate finished varieties. is the co-production of new knowledge. The
A different use of participatory learning is objectives are what differentiate approaches,
when PPB programmes seek to improve not the methodologies or tools they use
farmers capacity to participate in research for facilitating participation, whether these
in an informed manner, as when farmers involve participatory rapid appraisal (PRA),
are taught basic principles of heritability, Mother-Baby trial, farmer field schools
techniques for making crosses or how to (FFS), farmer research committees (CIALs),
keep trial records. participatory technology development
In practice, PPB programmes often use a (PTD) or others (Johnson, Lilja and Ashby,
combination of both participatory research 2003). The key distinctions are:
and participatory learning at different whether participation promotes or
stages in the plant breeding process. For excludes the co-production of new
example, in the PPB methodology called knowledge between farmers and scien-
Mother-Baby Trial the Mother varietal tists ; and
trial is a researcher-designed and researcher- the timing of farmer participation:
managed experiment. This trial is a platform specifically, how early does participation
for demonstration and participatory occur in the breeding cycle?
learning by farmers about the varieties Lilja and Ashby (1999) constructed
that breeders are testing. Farmers then a typology for empirical analysis of
select those varieties they want to try out participation based on the principle that the
on their own farms. The farmer-designed way decisions are shared at different stages
and farmer-managed Baby varietal trials of a plant breeding process will structure
are a platform for participatory learning the opportunities for co-production of
by breeders about farmers criteria for new knowledge. The typology defines two
varietal selection and management. Joint groups of decision-makers: scientists who
farmer-breeder participation in research include research programmes and extension
and the co-production of new knowledge agencies; and farmers who include all
occurs when the combination of results and the intended users of the PPB varieties
recommendations from Mother and Baby such as consumers, traders and processors.
trials are made jointly. However, if breeders These are ideal types of participation
interpret data, draw conclusions and make along a continuum in which farmers are
recommendations from Mother-Baby trials progressively more empowered, from
The impact of participatory plant breeding 655
different from those developed with farmers process into three stages: design, testing
involved at the design stage. Collaborative and diffusion. In PPB (Weltzein et al.,
participation at the late, diffusion stage 2003) these stages roughly correspond to:
of PPB means farmers can only influence Design. Setting breeding goals and gen-
when, where and with whom varieties are erating variability. Decisions are made
demonstrated and multiplied for seed, but about basic parameters of variety type(s),
not the kinds of varieties available to them. preferences, and user needs. In most pro-
Early participation that enables farmers grammes, this stage involves designing
to help set breeding goals has the collateral and making crosses between diverse par-
effect of encouraging farmers to engage ents with complementary trait combina-
more actively with the breeding programme tions. It may involve building base popu-
and adopt more rapidly. For example, in a lations for cross-pollinating crops or the
community meeting, Nepali women farmers generation of new progenies for testing.
asked for quality improvements in a cold- Testing. In plant breeding, decisions are
tolerant rice variety. Farmers and breeders made about how to narrow down the
managed and screened jointly from F5 bulk new variability achieved in the design
families and the resultant were superior stage from several thousand to a few
to the best entries from the conventional hundred progenies or clones (in the
breeding programme. Released by the case of vegetatively propagated crops),
national programme, the new variety spread and includes selection in segregating
rapidly to over 30 percent of rice area generations in self-pollinated crops. In
in the participating villages (Staphit and population improvement schemes this
Subedi, 2000). Witcombe and Virk (2001) is the progeny testing stage. In plant
argue, based on a number of studies, that breeding this stage includes the testing
when a breeding programme based on a of experimental materials on-station and,
few crosses, the choice of parents is crucial increasingly, on-farm. This testing looks
and that farmer participation is highly for desired productivity traits, adaptation
effective in narrowing this choice made at and acceptability, usually in replicated
the early stage in the breeding process. A plots over a range of locations with
methodological study of PPB using fixed increasing plot sizes. Testing continues
lines and segregating populations found that until varieties are proposed for release.
farmers used a higher selection pressure than Diffusion. This includes varietal release,
breeders, selecting about half the number demonstration under farmer management
of lines on station and about one tenth on farms, and the identification of a
of the number of lines on farm compared seed production and distribution system.
to breeders. Entries selected by farmers Although this stage goes beyond the
yielded as much as those selected by breeders purely technical breeding process, the
(Ceccarelli et al., 2000). This substantial seed system may present a bottleneck
body of research demonstrates the value to eventual impact, especially in poor
of integrating farmers intimate knowledge countries, that needs to be taken into
of their production environments into key consideration early in the design stage.
breeding decisions. Lilja and Ashby (1999, 2007) constructed
In the typology described above, Lilja a matrix in which any one of the five types
and Ashby (1999) divide the innovation of participation described earlier can be used
The impact of participatory plant breeding 657
in one or more of the three stages of PPB. whereby a given intervention leads to a set
With data obtained by using this matrix for of impacts, either expected or observed
interviews with 49 PPB programmes and (Douthwaite et al., 2003). This section uses
projects about 32 key decisions in the design the impact pathway framework to compare
and testing stages of PPB, multiple corre- impact pathways for PPB and PVS, their
spondence analysis (MCA) was applied to use of various types of participation
identify types of participation used in PPB. and differences in their impacts. Impact
MCA identifies relatively homogenous pathways can be diagrammed as flow charts,
groups of cases based on selected charac- showing products of varietal selection
teristics, in this case patterns among PPB (Figure 25.1, PVS) or making crosses
programmes in the way they use different (Figure 25.2, PPB) leading to outputs and
types of participation at different stages of finally to impacts. The impact pathway is a
the PPB process. The results showed that tool that a breeding programme can use to
these PPB programmes fall into different clarify its expected or actual outcomes and
clusters based on their participation prac- impacts: for those shown in Figures 25.1
tice. The cluster with the largest number of and 25.2 the topmost pathways are generic
PPB projects (61 percent) adheres mainly to but they can be changed to reflect specific
collaborative participation. programme goals.
Products in an impact pathway refer to
25.3 IMPACT PATHWAYS FOR results over which programmes have a high
DIFFERENT TYPES OF PARTICIPATION degree of control and a high probability of
Impact pathways provide a framework for achieving. Outcomes in an impact pathway
systematically mapping the cause-effect refer to the effects of using the products in
relationships (in the form of a flow chart), the short term (usually about 2 or 3 years).
FIGURE 25.1
Impact pathways for PVS
Lower
Collegial Information about probability Improved
performance of new of releasing research
varieties un der unacceptable efciency
farmer management varieties
Farmers may
prefer and
Farmer Independen t farmer
disseminate
experiment- selections among
unreleased
ation new varieties
materials
658 Plant breeding and farmer participation
FIGURE 25.2
Impact pathways for PPB
Impacts refer to effects that take more time to traders and consumers). The outcome
to achieve. In the category of PVS products, is that more farmers adopt PVS varieties
Figure 25.1 shows that consultative over wider areas, leading to increased food
participation produces information about and income benefits. Another impact is
farmers varietal preferences, which breeders increased research efficiency due to more
then use to identify existing varieties or relevant and desirable research products.
cultivars that are new and more desirable The impact pathway for PPB in
to the target population of farmers. If Figure 25.2 illustrates how impacts change
existing varieties are not suitable, breeders as participation occurs at earlier stages in
use information about farmer preferences the breeding process. As in PVS, a product
to produce new varieties that are typically of PPB is information about farmers
evaluated with farmers. This may involve varietal preferences. However, in PPB,
collaborative participation, in which this exchange takes place early enough for
farmers are involved in decisions about breeding objectives to be defined jointly.
which varieties are advanced or released. In some PPB programmes, parents are also
Typically, farmers engage in some of their identified and crosses jointly planned. Thus,
own experimentation, either guided by the PPB involves reciprocal learning by farmers
breeding programme (as in the Mother-baby of key information about breeding strategies
trial approach, for example) or independently and some basic procedures. Farmers help
with escapes from formal trials. The end manage early selection in the breeding
product of PVS is varieties that are more programme and this activity harnesses a lot
desirable to producers (and usually also of the energy and resources that farmers
The impact of participatory plant breeding 659
other PPB products noted in Figure 25.1, Quiros and Rivers, 1989; Ashby, 1986;
may be under-reported in the literature, so Kornegay, Beltrn and Ashby, 1996), in
that we tend to have more evidence about United Republic of Tanzania (Butler et al.,
success than about failure. 1995), in Ethiopia (Mekbib, 1997), and in
Some PPB impacts are relatively easy to Rwanda (Sperling and Scheidegger, 1996;
measure using established impact-assessment Sperling, Loevinsohn and Ntabomvura,
methodology. Agronomic and economic 1993); tree species in Burundi (Franzel,
outcomes can be assessed at the farm level Hitimana and Ekow, 1995); potatoes in
by measuring yield changes, net income over Rwanda (Haugerud and Collinson, 1990),
time and externalities such as changes in pest in Bolivia (Thiele et al., 1997; Gabriel et
pressure or soil loss. Increases or decreases al., 2006), in Peru (Ortiz et al., 2004), and
in costs are also straightforward. However, in Ecuador (Andrade and Cuesta, 1997);
when empowering participation is used, rainfed rice in India (Maurya, Bottrall and
part of the effects of PPB is on productivity Farrington, 1988); rice in Bangladesh, India
and in particular on accelerating innovation and Nepal (Joshi and Witcombe, 2003; Joshi
in varietal improvement These impacts that et al., 2008), and in East India (Cortois
are external to the technology are often et al., 2001); maize in India (Virk et al.,
referred to as disembodied effects, and pose 2003, 2005), in Ethiopia (Negasa, 1991), in
a greater challenge for impact assessment as Honduras (Gmez and Smith, 1996), and in
they are more difficult to quantify (Lilja and Brazil (Machado and Fernandes, 2001); and
Dixon, 2008). barley in the Syrian Arab Republic, Morocco
and Tunisia (Ceccarelli et al., 2001a, 2003).
Impact pathway: PPB and PVS produce A careful study in Mexico (Bellon et al.,
more desirable varieties leading to higher 2000) was designed to select a subset of 17
rates of adoption populations for PPB from a set of 152 maize
Numerous studies conclude that PPB and PVS landraces. The suggestions of men and
improve the acceptability of bred varieties women in farm communities, professional
to poor farmers in difficult environments plant breeders, gene bank managers and
by including their preferences in criteria social scientists were obtained. The results
for developing, testing and release. Small- showed that when germplasm choice did not
scale farmers often rank varieties in order of include farmers ideas, traits and materials
preference differently from breeders. Many important to farm households were often
examples are available that show how PPB overlooked. The involvement of women
clarifies where there is agreement between farmers in the participatory development
breeders and farmers on desirable traits of maize seed systems in China resulted
and where they disagree: cassava in Brazil in a broadened national maize genetic base
and Colombia (Iglesias, Hernndez-R and and improved maize yields (Song, 1998).
Lpez, 1990); Hernndez, 1993; Fukuda This experience, by now so diverse with
and Saad, 2001); pearl millet in Namibia respect to crops, cultures and production
(Ipinge, Lechner and Monyo, 1996; Monyo environments, demonstrates the efficacy
et al., 1997a,b) and in India (Weltzein, of participatory selection in producing
2000); maize in Mali (Kamara, Defoer and varieties for poor farmers who are otherwise
De Groote, 1996; Defoer, Kamara and De excluded by conventional crop improvement
Groote, 1997); beans in Colombia (Ashby, programmes.
The impact of participatory plant breeding 661
A rigorous study conducted by the several clones were released which were
ICARDA barley breeding programme highly acceptable to farmers (Fukuda and
compared the number of high-yielding Saad, 2001). Weltzein (2000) explains how
varieties obtained (termed selection learning about farmers preferences and
efficiency) using different approaches. The selection criteria reoriented an international
breeder was more successful than farmers pearl millet breeding programme to identify
in selecting on station under high rainfall components for the mixtures of plant
conditions, but farmers were more successful types farmers customarily grow. The new
under stress conditions. A t-test of significant materials were well-accepted by farmers
difference showed that farmers selections who were not adapting modern varieties. A
were as high yielding as breeders selections study conducted in Syria provides evidence
(Ceccarelli et al., 2001a). Subsequently, the of higher rates of adoption of PPB barley
same programme conducted an important varieties. Farmers were planting 69 percent
set of experiments on farmer participation more area to PPB than to conventionally
in barley breeding in Morocco, the Syrian bred varieties and were willing to pay
Arab Republic and Tunisia, where barley is more for seed of PPB varieties (Lilja and
the main crop for poor farmers in marginal, Aw-Hassan, 2002). On average, farmers
rainfed areas. Breeders trials were planted reported PPB varieties had a 26 percent
both on research stations and in farmers yield advantage over conventionally bred
fields. Selection was done independently varieties. In Ghana, maize breeders released
by professional breeders and farmers several modern varieties, which had poor
and data were gathered on their selection acceptance and were not widely adopted.
criteria and selection efficiency. Farmers Subsequently, new material was tested in
used selection criteria not normally used researcher-managed trials and in farmer-
by breeders because of the importance of managed trials, and the outstanding modern
the crop as source of animal feed. Disease, a varieties were jointly selected. Overall
major selection criterion used by breeders, adoption of modern varieties expanded to
was almost entirely neglected by farmers. over two-thirds of Ghanas maize farmers
Farmers successfully selected some of the (Morris, Tripp and Dankyi, 1999). Another
highest yielding lines in their own fields and study compared matched communities with
also on station. (Ceccarelli et al., 2001b). and without PVS conducted by farmer
By successfully understanding and research committees (CIALs). Communities
incorporating farmers criteria for doing PVS had a much higher rate of
acceptability, PPB consistently enables adoption than non-PVS communities, who
breeding programmes to break through relied on other channels for seed (IPRA,
adoption bottlenecks. In Ethiopia, for 2008). The WARDA PVS programme
example, over 122 varieties of cereals, conducted in 17 West African countries
legumes, crops and vegetables were released, since 1996 used consultative participation
but only 12 varieties had been adopted to understand better what farmers need, and
by farmers, prompting a start with PPB to feed back insights to formal research for
(Mekbib, 1997). In Brazil, the national improving future on-farm productivity: 69
agricultural research institute, EMBRAPA, percent of national programme researchers
confronted years of non-adoption of new considered that by consulting women and
cassava clones. Once PPB was implemented involving them in varietal evaluation, the
662 Plant breeding and farmer participation
programme had included varietal traits that corresponding to the ideotype into farmer
women know about, and especially gender- trials, and because millet is cross-pollinated,
related varietal preferences, leading to higher the frequency of the desired traits increased
adoption of the varieties. in local germplasm through introgression.
Farmers began selecting outcrosses to
Impact pathway: PPB leads to faster provide seed for the following season, and
varietal release after 4 years, breeders selected plants from a
A study that examines this impact farmers field. These plants were intercrossed
pathway in depth was conducted by the with 30 varieties selected on-station by
ICARDA Barley Breeding programme in farmers from specially designed elite and
Syria (Ceccarelli et al., 2001a). Using the morphologically diverse nurseries, to
same breeding population, varieties were create a PPB composite population named
developed using participatory and non- MKC. MKC was far superior to the local
participatory breeding. The study found germplasm and to another population, NC
that by introducing farmer participation 90, developed by conventional breeding
at an early stage of the breeding process (Monyo et al., 1997a, b).
(in Year 3), a three-year reduction was PPB carried out in Bolivia addressed the
achieved in the time taken from initial need to develop potato varieties for specific
crosses to release. PPB made certified ecological and market niches that need to be
varieties available by Year 6 compared similar to those already valued by farmers
to Year 9 in the conventional breeding and consumers, but more productive and
programme (Lilja and Aw-Hassan, 2002) more resistant to biotic factors such as
PPB in rice and maize in India and Late blight disease (Phytophthora infestans)
Nepal found that farmers were well able and False root-knot nematode (Nacobbus
to select from large numbers of segregating aberrans). Men and women farmers were
materials and their most preferred materials trained in potato breeding techniques and,
were rapidly adopted (Staphit, Joshi and jointly with the plant breeders, generated
Witcombe, 1996). Based on experience with 12 varieties similar to the most widely
different crops, the breeders concluded consumed cultivar in Bolivia, but resistant
that PPB reduced by 3 to 4 years the to late blight, with superior yield (1025 t/
time between making a cross and farmers ha, compared with 5 t/ha from the main
receiving materials for testing. This contrasts farmer variety) and possessing agronomic
with the conventional time of 10 years (Virk traits and qualities desired by farmers.
et al., 2005, 2003) Three of the varieties showed novel
In another case, farmer participation in potential for the potato chip industry. The
screening the entire pearl millet germplasm breeders concluded that time was gained
accessions from Namibia (numbering about and adoption accelerated when farmers
1 000) proved very efficient in generating engaged early (Gabriel et al., 2006).
some basic information, such as when
farmers recognized three major classes of Impact pathway: PPBs faster varietal
materials with different clusters of desirable release leading to earlier adoption
traits, and assisted breeders to come up increases the stream of benefits to farmers
with the desired pearl millet ideotype for An economic analysis of PPB barley
Namibia. Breeders introduced material breeding in Syria provides evidence of earlier
The impact of participatory plant breeding 663
adoption impact. Over 23 PPB varieties One of the main concerns related to
are grown on several thousand hectares. research efficiency of conventional breeding
Total estimated discounted research programmes is that PPB looks very time
induced benefits to Syrian agriculture were intensive, and therefore costly. Many aspects
estimated, comparing conventional and of PPB seem likely to increase costs: on-farm
three different PPB approaches, based on a testing begins earlier, more seed is needed of
rigorous comparison using experimentally- experimental varieties, trials are dispersed
generated data on yields. Benefits from outside the experiment stations, and different
conventional breeding were estimated at kinds of personnel may be needed to interact
US$ 21.9 million. Benefits estimated for effectively with farmers. Farmers need to
the three PPB approaches ranged from be transported to experiment stations or
US$ 42.7 million to US$ 113.9 million. regional trials, and a good deal of time is
Most difference is attributed to the way spent interacting with them to involve them
PPB reduced the amount of time it took for in the early stages of the breeding process.
improved varieties to get into farmers fields In the case of a high altitude rice in Nepal,
(Lilja and Aw-Hassan, 2002). Staphit and Subedi (1996) considered their
combined PVS and PPB approach cost-
Impact pathway: more desirable varieties effective because the parents and segregating
and higher adoption rates improve products were piggybacked off the ongoing
research efficiency formal breeding process. Farmers were
New Rice for Africa (NERICA) given still segregating (F5) bulk families
implemented by WARDA, the African harvested from the most promising F4 rows,
Rice Centre, used PVS to evaluate new for evaluation in their fields. There were
varieties with men and women farmers, and important differences in the ways farmers
helped to identify cost-saving production, and breeders tested the materials. The
grain processing and consumption traits, preferred cultivars subsequently developed
in addition to yield-related characteristics, with farmers were widely adopted within
valued by men and women. Results from Cte three years. In ICARDAs study that
dIvoire show that failing to include gender- compared PPB and conventional approaches
differentiated production and consumption the operational costs of the programme
traits and focusing on the wrong attributes increased due to PPB, which included costs
leads to biased and inappropriate varietal of work off station in Syria and in several
promotions. Evaluating new varieties other countries. However, operational costs
only on yield-related characteristics (often are only 23 percent of the total budget.
gender-neutral) will cause 19 percent of Overall, the total annual budget went up by
all varieties to be wrongly categorized as 3 percent, approximately US$ 26 000. (Lilja
superior, whereas incorporating gender- and Aw-Hassan, 2002). This cost has to be
differentiated traits (labour-related, seen against the savings incurred by getting
consumption, post-harvest) reduces the varieties out to farmers three years earlier
probability of promoting varieties with using PPB. Clearly more analyses of the way
poor acceptability and instead increases PPB affects costs would help to clarify this
adoption potential (Dalton and Guei, 2003; debate, but at present we cannot conclude
Dalton, 2004; Lilja and Erenstein, 2002; that PPB automatically represents a major
Lilja and Dalton, 1998). increase in cost for a breeding programme.
664 Plant breeding and farmer participation
Impact pathway: PPB fosters new skills, opportunity for further research that could
new knowledge and social capital that contribute to its wider recognition.
speed up innovation Although not assessments of PPB
PPB involves moving off conveniently- impact, several studies find that use of
located, well-endowed experiment participatory research and learning
stations to a more decentralized breeding approaches improve skills and knowledge
programme that relies heavily on farmer- important for innovation (see, for example,
managed selection. Numerous studies Dalton et al., 2005; Classen et al., 2008).
conclude that selection by farmers offers For example, social and human capital
the greatest yield benefit over experiment benefits have been studied for members
station selection in low-yield-potential, of farmer research committees (CIALs)
marginal environments that differ in Latin America (Classen et al., 2008).
dramatically from experiment station CIAL members indicated that they had
conditions (Weltzein et al., 2001; Smith, gained more knowledge about agriculture
Castillo and Gmez, 2001; Cecarelli, 2000, and were seen as agricultural experts
2001, 2003; TAC, 2001; Virk et al., 2003; and advisors in the community. They
Morris and Bellon, 2004; Walker, 2006). had also improved their communication
Decentralization places heavy demands and leadership skills, and had increased
on breeders time and other resources, relationships with neighbours and
unless a significant degree of delegation with other outside institutions. CIAL
to farmers takes place. In this situation, members experimented more with new
farmers need to develop the skills and crops, had learned other new skills, and
knowledge required to maintain research had higher levels of commitment to
quality with minimal supervision. Low their communities, thereby leading to a
rates of varietal turnover or replacement higher level of community participation.
have been proposed as an indicator that In communities where the CIAL had
farmers are not able to access varieties identified new technology and converted
appropriate to their needs and constraints, into commercial seed producers, the
signalling that opportunities exist to start communities benefited by having easy
PPB (Walker, 2006; Brennan and Byerlee, access to new technology (e.g. new varieties,
1991). Breeders also acquire new skills such as early maturing maize varieties and
and knowledge through PPB that improve new bean varieties). One study specifically
their capacity to sustain the varietal change examines the argument that improvements
needed to increase productivity. However, in social capital from using a participatory
whether PPB fosters capacity to sustain approach reduces transaction costs, leading
this innovation remains a research gap to better cooperation and coordination
in assessment of the impacts of PPB. and improving the innovation process
Whether PPB will gain enough traction (Gandarillas Morales, 2007). Thus impact
to achieve institutionalization in breeding pathway remains poorly documented for
programmes is still an open question: PPB, reflecting the focus of most studies
without institutionalization, PPBs wider on the breeding products and outcomes
impact on innovation systems may remain rather than the social impacts of PPB. This
hypothetical. Social analysis of innovation therefore remains an area ripe for further
processes involving PPB are therefore an social analysis.
The impact of participatory plant breeding 665
Impact pathway: PPB increases inclusion in Mexico concluded that farmers as a group
of the poor and disadvantaged, especially earned a high return with a benefit:cost ratio
women, in R&D, leading to more of 3.8-1 although for the private investor
equitable distribution of benefits the returns were low. This underscores
This impact pathway remains a significant the importance working with groups of
research gap in the assessment of PPB. Only farmers to build participation. Participants
if programmes make a specific strategy of from richer households captured a larger
including the poor or other disadvantaged proportion than they invested so that
groups in the process, will PPB or PVS lead there was a transfer of wealth to the richer
to more equitable distribution of benefits. households from the intermediate investors,
It is often wrongly assumed that use of also the largest sub-group of investors.
participatory approaches will guarantee This reinforced the gender bias in the
inclusion of the poor or women and lead to distribution of benefits (Smale et al., 2003)
more equity. While these approaches make The lesson here is that participation can
it easier to engage with the poor, unless lead to the exclusion of important groups
participant selection targets a particular of beneficiaries, such as women, depending
social group, a participatory approach does on prevailing customs and norms, especially
not automatically lead to benefits for them if participation is based on self-selection.
(Kumar and Corbridge, 2002; Cleaver, 1999). Inclusion in PPB may be determined a
The implications for PPB are illustrated by priori by the choice of crop, suggesting that
the findings of a study assessing impacts if equity is an impact goal, then the decision
of participation in potato Integrated Pest with whom to do PPB should precede
Management (IPM) and PPB in Peru (Ortiz ones about where and which crop to work
et al., 2001). Women did not participate in with.
potato IPM because pest management is
not part of their traditional responsibilities 25.4 CONCLUSIONS
in the potato crop, and they said they PPB produces more acceptable varieties,
could participate more in other traditional increasing adoption. This is its most
womens crops. However, participatory extensively documented outcome and
selection of potato clones was identified as probably the most compelling incentive for
an activity in which women had an essential plant breeders to use this approach. While
contribution, because they are responsible major PPB initiatives, such as ICARDAs,
for seed management, and so they were do document yield improvements of 3050
50 percent of the participants in varietal percent with PPB, there remains a need for
selection. In contrast, a study carried out in more comparative data on yield or other
Ecuador with two indigenous communities, advantages of PPB versus other breeding
to determine farmers preferences for quinoa approaches, data that could be used to
cultivars and to improve PPB processes, assess final impacts. Another consideration
found more women than men participated is research efficiency: PPB makes research
because quinoa, a primarily subsistence crop, more relevant and the changes in cost it
is mainly managed by women (McElhinny involves do not appear to lower breeding
et al., 2007). An analysis of economic costs programme costbenefit ratios, and might
and benefits of participation by farmers in even improve these. PPB also affects the
conserving and improving maize landraces speed at which new varieties are developed
666 Plant breeding and farmer participation
and get into farmers fields. The important 26 September1 October 2004, Brisbane,
determinants of this impact are the use and Australia. Published on CDROM. Web site
timing of collaborative participation: early www.cropscience.org.au
involvement of farmers in setting breeding Ashby, J.A., Quiros, C. & Rivers, Y. 1989.
goals encourages the co-production of new Farmer participation in technology devel-
knowledge, gives farmers confidence in opment: work with crop varieties. pp. 115
and ownership of the new varieties, and 122, in Farmer First. London, Intermediate
stimulates their rapid dissemination. To Technology Publications.
realize its full potential on a large scale, Bellon, M.R., Smale, M., Aguirre, A., Taba,
PPB requires organizational, policy and S., Aragn, F., Daz, J. & Castro, H. 2000.
legal changes in both international and Identifying Appropriate Germplasm for
national plant breeding which, with few Participatory Breeding: An Example from
exceptions, represent tenacious obstacles the Central Valleys of Oaxaca, Mexico.
to the institutionalization of PPB because CIMMYT Economics Working Paper
science bureaucracies and the political elites 00-03. CIMMYT, Mexico D.F., Mexico.
that fund them, resist being accountable to Brennan, J.P. & Byerlee, D. 1991 The rate of
poor farmers as clients. In the long term, crop varietal replacement on farms: meas-
one of the most important impacts of PPB ures and empirical results for wheat. Plant
may be its effect on the relevance of these Varieties and Seeds 4: 99-106
agricultural innovation systems to the poor. Butler, L.M., Myers, J., Nchimbi-Msolla, S.,
Massangye, E., Mduruma, Z., Mollel, N. &
REFERENCES Dimosa, P. 1995. Farmer evaluation for early
Andrade, H.J. & Cuesta, X. 1997. El rol del generation bean lines preferences. Southern
usuario en la seleccin y liberacin de Africa Development Community (SADC)
variedades de papa en el Ecuador. Instituto Regional Bean Research Workshop, 24
Nacional de Investigaciones Agropecuarias October 1995. Oil and Protein Seed Centre,
(INIAP), Ecuador. Potchefstroom, South Africa. 20 p.
Ashby, J.A.1986. Methodology for the par- Ceccarelli, S., Grando, S., Singh, M.D., Shikho,
ticipation of small farmers in the design of A., Al Issam, M., Al Saleh, A., Kaleonjy, G.,
on-farm trials. Agricultural Administration Al Ghanem, S.M., Al Asma, A.L., Dalla, H.,
and Extension, 25: 161176. Basha S. & Basha, T. 2000. A methodologi-
Ashby, J.A. 1996. What do we mean by par- cal study on participatory plant breeding. I.
ticipatory research in agriculture? In New Selection Phase. Euphytica, 111: 91104.
Frontiers in Participatory Research and Ceccarelli, S., Kafawin, S.O., Saoub, H., Grando,
Gender Analysis, pp. 1522. Proceedings of S., Halila, H., Ababneh, M., Shakatreh, Y. &
the international seminar on participatory Bailey, E. 2001a. Increasing the relevance
research and gender analysis (PRGA), Cali, of breeding to small farmers: farmer par-
Colombia, 914 September 1996. CIAT ticipation and local knowledge in breeding
Publication, No.294. barley for specific adaptation to dry areas of
Ashby, J.A. & Lilja, N. 2004. Participatory Jordan. In An Exchange of Experiences from
research: Does it work? Evidence from South and South East Asia, pp. 381390.
participatory breeding. In New directions Proceedings of the international symposium
for a diverse planet. Proceedings of the on Participatory plant breeding and partici-
4th International Crop Science Congress, patory plant genetic resource enhancement,
The impact of participatory plant breeding 667
Pokhara, Nepal, 15 May 2000. Participatory Dalton, T.J. 2004. A household hedonic model
Research and Gender Analysis Program of rice traits: economic values from farm-
(PRGA), and CIAT, Cali, Colombia. 459 p ers in West Africa. Agricultural Economics,
(available at http://www.prgaprogram.org/ 31(2-3): 149159.
modules.php?op=modload&name=Down Dalton, T.J. & Guei, R.G. 2003. Productivity
loadsPlus&file=index&req=getit&lid=329; gains from rice genetic enhancements in
accessed 26 Jan. 2009). West Africa: countries and ecologies. World
Ceccarelli, S., Grando, S., Bailey, E., Amri, Development, 31: 359374.
A., El-Felah, M., Nassif, F., Rezgui, S. & Dalton, T.J., Lilja, N., Johnson, N. & Howeler,
Yahyaoui, A. 2001b. Farmer participation R. 2005. Human capital accumulation
in barley breeding in Syria, Morocco, and and productivity improvements in Asian
Tunisia. Euphytica, 122: 21536. cassava systems: Are participatory research
Ceccarelli, S., Grando, S., Tutwiler, R., Baha, approaches beneficial? Paper presented
J., Martn, A.M., Salahieh, H., Goodchild, at the American Agricultural Economics
A. & Michael, M. 2003. A methodologi- Association meeting, Providence, Rhode
cal study on participatory plant breed- Island, USA, July 2427 2005.
ing. II. Response to selection. Euphytica, Defoer, T., Kamara, A. & De Groote, H. 1997.
133: 185200. Gender and variety selection: farmers
Chiffoleau, Y. & Desclaux, D. 2006. assessment of local maize varieties in
Participatory plant breeding: the best southern Mali. African Crop Science
way to breed for sustainable agriculture? Journal, 5: 6576.
International Journal of Agricultural Douthwaite, B., Kuby, T., van de Fliert,
Sustainability, 4(2): 119130. E. & Schulz, S. 2003. Impact pathway
Classen, L., Humphries, S., FitzSimons, J., evaluation: an approach for achieving and
Kaaria, S., Jimnez, J., Sierra, F. & Gallardo, attributing impact in complex systems.
O. 2008. Opening participatory spaces for Agricultural Systems, 78: 243265.
the most marginal: learning from collective Franzel, S., Hitimana, L. & Ekow, A. 1995.
action in the Honduran hillsides. World Farmer participation in on-station tree
Development, 36(11): 24022420. species selection agroforestry: A case study
Cleaver, F. 1999. Paradoxes of participation: from Burundi. Experimental Agriculture,
Questioning participatory approaches to 31: 2738.
development. Journal of International Fukuda, W. & Saad, N. 2001. Participatory
Development, 11: 597612. research in cassava breeding with farmers
Courtois, B., Bartholome, B., Chaudhary, D., in northeastern Brazil. Empresa Brasileira
McLaren, G., Misra, C.H., Mandal, N.P., de Pesquisa Agropecuria (EMBRAPA),
Pandey, S., Paris, T., Piggin, C., Prasad, K., Embrapa Mandioca e Fruticultura Cruz
Roy, A.T., Sahu, R.K., Sahu, V.N., Sarkarung, das Almas, Brazil. 44 p.
S., Sharma, S., Sigh, A., Singh, H.N., Sinh, Gabriel, J.L., Herbas, J., Salazar, M., Ruiz,
O.N., Singh, N.K., Singh, R.K., Singh, S., J., Lpez, J., Villarroel, J. & Cossio,
Sinha, P.K., Sisodia, B.V.S. & Takhur, R. D. 2006. Participatory plant breeding:
2001. Comparing farmers and breeders A new challenge in the generation and
rankings in varietal selection for low-input appropriation of potato varieties by
environments: a case study of rainfed rice in farmers in Bolivia. PRGA Consultative
eastern India. Euphytica, 122: 537550. Group on International Agricultural
668 Plant breeding and farmer participation
Research (CGIAR); Fundacin PROINPA patory research: the case of corn in South
(Working Document no. 22). Mali. Tropicultura, 16: 100105.
Gandarillas Morales, E.G. 2007. Institutional Kumar, S. & Corbridge, S. 2002. Programmed
arrangements to improve the responsiveness to fail? Development projects and the poli-
of agricultural innovations systems to the tics of participation. Journal of Development
needs of the poor: An inverstigation of Studies, 39(2): 73103.
participatory monitoring and evaluation Johnson, N.L., Lilja, N. & Ashby, J.A. 2003.
in Bolivia. Unpublished working paper, Measuring the impact of user participation
PROINPA Foundation, Bolivia. in agricultural and natural resource man-
Gmez, F. & Smith, M. 1996. Conservation agement research. Agricultural Systems,
and enhancement of maize with small farm- 78: 287306.
ers in Honduras. A collaborative project Joshi, K.D., Staphit, B.R. & Witcombe, J.R.
between Escuela Agrcola Panamericana 2001. How narrowly adapted are the
Zamorano and Cornell International products of decentralized breeding? The
Institute for Food, Agriculture and spread of rice varieties from a participatory
Development (CIIFAD). Project report. breeding programme in Nepal. Euphytica,
15 p. (Mimeo). 122(3): 589597.
Haugerud, A. & Collinson, M.P. 1990. Plans, Joshi, K.D. & Witcombe, J.R. 2003. The
genes and people: Improving the relevance impact of participatory plant breeding
of plant breeding in Africa. Experimental (PPB) on landrace diversity: a case study
Agriculture, 26: 341362. for high altitude rice in Nepal. Euphytica
Hernndez-R, L.A. 1993. Evaluacin de nue- 134(1): 117125.
vas variedades de yuca con la participacin Joshi, K.D., Devkota, K.P., Harris, D., Rawal,
de agricultures. Working document no.130, K.B., Pandit, D.B., Tripathi, M., Musa,
CIAT, Cali, Colombia. 85 p. A.M., Johansen, C. & Witcombe, J.R. 2008.
Iglesias, C.A., Hernndez-R, L.A. & Lpez, A. Innovative ways of assessing outcomes and
1990. Participatory selection with farmers impacts of participatory research on rice
in northern Colombia. Internal mss. CIAT, in Bangladesh, India and Nepal. Paper
Cali, Colombia. prepared for Workshop on Rethinking
Ipinge, S.A., Lechner, W.R. & Monyo, E.S. Impact: Capturing the Complexity of
1996. Farmer participation in the evalu- Poverty and Change. PRGA Program,
ation of priority plant and grain traits CIAT, Cali, Colombia. March 2008
on station. The case of pearl millet in Kornegay, J., Beltrn, J.A. & Ashby, J.A. 1996.
Namibia. In K. Leuschner & C.S. Manthe, Farmers selections within segregating popu-
eds. Drought-tolerant crops of Southern lations of common bean in Colombia. In P.
Africa. Proceedings of the SADC/ Eyzaguirre & M. Iwanaga, eds. Participatory
ICRISAT Regional Sorghum and Pearl Plant Breeding, pp. 151159. Proceedings
Millet Workshop, Gaborone, Botswana, of a Workshop on Participatory Plant
2529 July 1994. ICRISAT, India. Breeding, Wageningen, Netherlands, 2629
IPRA, 2008. The CIALs at a Glance: Adoption July 1995. Rome, IPGRI.
of a Bean Variety (Available at http://www. Lilja, N. & Dalton, T.J. 1998. Developing
ciat.cgiar.org/ipra/ing/glance.htm). African public goods: rice varietal selec-
Kamara, A., Defoer, T. & De Groote, H. 1996. tion in Cte dIvoire. A discussion
Selection of new varieties through partici- paper, Organized Symposium, American
The impact of participatory plant breeding 669
Negasa, A., Tolessa, B., Franzel, S., Gedeno, Sperling, L. & Scheidegger, U. 1996. Results,
G. & Dadi, L. 1991. The introduction of an methods, and issues in participatory
early maturing maize (Zea mays) variety to selection: The case of beans in Rwanda.
a mid-altitude farming system in Ethiopia. International Institute for Environment
Experimental Agriculture, 27: 375383. and Development (IIED). Gatekeeper
Okali, C., Sumberg, J. & Farrington, J. 1994. Series no.51. UK.
Farmer Participatory Research: Rhetoric and Sthapit, B., Joshi, K.D. & Witcombe, J.R.
Reality. London, Intermediate Technology 1996. Farmer participatory crop improve-
Publications. ment. III. Participatory plant breeding, a
Ortiz, O., Nelson, R., Orrego, R. & Leon, case-study for rice in Nepal. Experimental
V. 2001. Impact evaluation of participa- Agriculture, 32: 479496.
tory development of integrated insect and Staphit, B. & Subedi, A. 2000. Participatory
disease management (IPM) for the potato approaches in plant breeding. In C.
crop in San Miguel, Peru. Report by CIP to Almekinders & W. de Boeuf, eds.
the CGIAR PRGA Program (available at Encouraging Diversity: the conservation
http://190.144.167.140/prga and search for and development of plant genetic resourc-
Ortiz; accessed 11 April 2009). es, pp. 183188. Exeter, UK, Intermediate
Ortiz, O. Garrett, K.A., Heath, J.J., Orrego, R. Technology Publications.
& Nelson, R.J. 2004. Management of potato TAC [Technical Advisory Committee of
late blight in the Peruvian highlands: evalu- CGIAR]. 2001. Systemwide Review of
ating the benefits of farmer field schools Breeding Methodologies in the CGIAR.
and farmer participatory research. Plant Rome, FAO.
Disease, 88(5): 565571. Thiele, G., Gardner, G., Torrez, R. & Gabriel, J.
Smale, M., Bellon, M.R., Aguirre, J.A., Manuel 1997. Farmer involvement in selecting new
Rosas, I., Mendoza, J., Solano, A.M., varieties: Potatoes in Bolivia. Experimental
Martnez, R., Ramrez, A. & Berthaud, J. Agriculture, 33: 275290.
2003. The economic costs and benefits of Vernooy, R. 2003. Seeds that give: Participatory
a participatory project to conserve maize Plant Breeding. Ottawa, International
landraces on farms in Oaxaca, Mexico. Development Research Centre. 83 p.
Agricultural Economics, 29(3): 265275. Virk, D.S., Singh, D.N., Prasad, S.., Gangwar,
Smith, M.E., Castillo, F. & Gmez, F. 2001. J.S. & Witcombe, J.R. 2003. Collaborative
Participatory plant breeding with maize and consultative participatory plant breed-
in Mexico and Honduras. Euphytica, ing of rice for the rainfed uplands of India.
122: 551565. Euphytica, 132: 95108.
Song, Y. 1998. New seed in old China: impact Virk, D.S., Chakraborty, M., Ghosh, J., Prasad,
of CIMMYTs collaborative programme on S.C. & Witcombe, J.R. 2005. Increasing
breeding in south-western China. PhD the- the client orientation of maize breeding
sis. Wageningen Agricultural University, using farmer participation in East India.
Netherlands. Experimental Agriculture, 41: 413426.
Sperling, L., Loevinsohn, M. & Ntabomvura, Walker, T.S. 2006. Participatory varietal selec-
B. 1993. Rethinking the farmers role in tion, participatory plant breeding, and
plant breeding: Local bean experts and on- varietal change. Background paper for
situation selection in Rwanda. Experimental the World Development Report 2008.
Agriculture, 29: 509515. Washington, DC, World Bank.
The impact of participatory plant breeding 671
Weltzein, E. 2000. Supporting farmers genetic ment. I. Varietal selection and breeding
resource management: experiences with methods and their impact on biodiversity.
pearl millet in India. In C. Almekinders & Experimental Agriculture, 32: 445460.
W. de Boeuf, eds. Encouraging Diversity: Witcombe, J.R. & Virk, D.S. 2001. Number
the conservation and development of plant of crosses and population size for par-
genetic resources, pp. 189193. Exeter, UK, ticipatory and classical plant breeding.
Intermediate Technology Publications. Euphytica 122:451-462
Weltzein, E., Smith, M.E., Meitzner, L.S. & Witcombe, J.R., Joshi, K.D., Gayawali, S.,
Sperling, L. 2003. Technical and institu- Musa, A.M., Johansen, C., Virk, D.S. &
tional issues in participatory plant breed- Staphit, B.R. 2005. Participatory plant
ing From the perspective of formal plant breeding is better described as highly
breeding. A global analysis of issues, results client-oriented plant breeding I. Four
and current experience. PPB Monograph, indicators of client orientation in plant
No 1. Cali, Colombia, PRGA. breeding. Experimental Agriculture,
Witcombe, J.R., Joshi, K.D. & Staphit, B.R. 41: 299319.
1996. Farmer participatory crop improve-
This book provides a comprehensive description and
assessment of the use of participatory plant breeding in
developing countries. It is aimed at plant breeders, social
scientists, students and practitioners interested in
learning more about its use with the hope that they all
will nd a common ground to discuss ways in which plant
breeding can be benecial to all and can contribute to
alleviate poverty.
ISBN 978-92-5-106382-8
9 789251 063828
I1070E/1/09.09/2000