Professional Documents
Culture Documents
E.A. Weiss
Agricultural Adviser
Eaglemont
Victoria
Australia
CAB INTERNATIONAL
CABI Publishing is a division of CAB International
OCAB International 1997. All rights reserved. No part of this publication may be
reproduced in any form or by any means, electronically, mechanically, by
photocopying, recording or otherwise, without the prior permission of the copyright
owners.
A catalogue record for this book is available from the British Library, London, UK
A catalogue record for this book is available from the Library of Congress, Washington
DC, USA
Introduction vii
Acknowledgements xi
2. Annonaceae 10
Cananga 10
3. Geraniaceae 24
Pelargonium 24
Geranium 50
4. Gramineae 59
Citronella 67
Lemongrass 86
Palmarosa and Gingergrass 103
Vetiver 117
5. Lamiaceae 138
Patchouli 138
6. Lauraceae 155
Camphor 156
Cassia 168
Cinnamon 180
Sassafras 195
Ocotea 198
Laurel 200
Litsea cubeba 207
7. Myristicaceae 214
Nutmeg 215
V
Vi Contents
8. Myrtaceae 235
Clove 235
Euclayptus 259
Melaleuca 302
Leptospermum 319
Pimenta 322
Myrtle 333
9. Oleaceae 342
Jasmine 342
Osmanthus 359
Glossary 579
Index 585
Introduction
vii
viii introduction
anointed with aromatic oils danced for the erotic pleasure of guests at ruler’s
feasts. Solomon, King of Israel, in his oft-quoted Song described the
exquisitely smelling ointments that stirred young maidens to love, and who
perfumed his bed with essences of myrrh, aloes and cinnamon; a common
contemporary practice by women wishing to attract a lover. It is recorded
that the great general Hannibal had good reason to regret the relationship;
after crossing the Alps and defeating the Romans he quartered his army in
Capua, capital of Campania. Defeated Romans corrupted his idle troops
and quickly made the town a trading centre for perfumes, aromatic oils and
associated sexual excess - practices which spread rapidly throughout the
Empire.
Today the association between perfumes and sexual attractiveness con-
tinues to be exploited in the electronic and print media and names of major
perfumes; while insects and some animals disperse a pheromone to attract a
mate.
The author has long been interested in aromatic plants, and wherever
possible introduced them into research programmes or integrated their
production into agricultural projects under his control. The crops selected
are those the author believes should be encouraged in countries where cash
crops are limited, or outside inputs to raise rural incomes difficult to obtain.
For this reason the important pine oils have not been included, since these
are frequently produced in more developed countries. It is emphasized that
this book is basically concerned with growing essential oil plants to obtain an
aromatic derivative, and not with essential oils per se.
Oil is contained in specialized structures in all or some plant parts: cavities
or ducts in the epidermis as in eucalyptus leaves or citrus fruit peel; glands or
hairs originating from epidermal cells as in the modified leaf hairs on
geranium. Why plants secrete oils or waxes has yet to be fully explained,
although certain activities can reasonably be attributed to their presence; to
deter browsing animals, as an insect repellent or to increase disease resist-
ance, while terpenes leached from eucalyptus leaves contribute to allelo-
pathic effects on the forest floor thus inhibiting germination and growth of
competitors. Highly scented oil contained in flowers is the exception, and
generally accepted as an aid to reproduction by attracting pollinators. Hence
the recommendation to integrate bee-keeping into commercial production
of essential oil crops whenever possible, as the author demonstrated most
profitably with certain oilseeds, and why honey production has been men-
tioned where applicable.
While the advantage to plants of their oil remains unclear, so also is the
mode of action of essential oils in phytotherapy. It would appear the
lipophilic character of compounds in oils could be responsible for the broad
spectrum of biological activities; however, to massage oils into the human
skin as in aromatherapy is of little therapeutic value, although it may induce
a feeling of sensual well-being!
lntroduction ix
The plant metabolic processes involved in oil secretion, the rate it occurs,
the biosynthetic pathways determining its composition and evolution of
various components, require further investigation. The detailed composi-
tion of many essential oils, however, has been published in the literature, and
reviews such as those of Dr Brian Lawrence are a valuable reference
source.
Oil composition, especially the terpene components, is now important in
plant classification as chemotaxonomy, since physical characters may be
virtually impossible to differentiate. The increasing precision of analytical
techniques, including enantioselective gas chromatography, allows very
accurate determination of oil composition. How this is linked to odour is
debatable, and assessing oil odour remains highly subjective. However,
determining a specific constituent or formulation is possible using an elec-
tronic nose, which is an array of chemical sensors linked to a neural network
(see Chapter 15).
Biotechnology can influence the breeding or development of essential oil
plants, and could theoretically produce plants containing oil of a required
composition or odour. It is also possible agronomically to modify a partic-
ular oil by varying the time of harvest, and thus regulating the abundance of
a specific constituent. Genetic manipulation can be commercially successful
as demonstrated on a fatty oil by scientists at CSIRO in Australia, who
successfully modified linseed to produce an edible oil, linola (Weiss, 1993).
French plant breeders have also used these techniques to alter the natural
scent of jasmine and rose flowers to produce novel odours aimed directly at
the fragrance industry.
Certain components of essential oils can be replaced by relatively cheap
synthetics, and although such compounds are useful in what are virtually
industrial products such as soft drinks, few are able fully to duplicate the
natural essence. Additionally, there is considerable worldwide pressure by
consumers on manufacturers to use what are perceived to be natural
compounds in edible and personal products. T o take advantage of this and
similar trends, growers in all countries must upgrade agricultural methods
and improve management techniques to either increase yield or raise profita-
bility. A range of suitable agricultural equipment is now available, including
a variety of small manually-operated machines, computer-controlled seed-
ers and transplanters, and integrated harvestingidistilling combinations.
Semi-automatic fruit pickers are available, and citrus-picking robots in-
dicate future trends.
Weed control should incorporate rotations and efficient use of herbicides,
which remain important and economical, while application machinery is
continually improving including weed-detecting sprayers. Integrated pest
management can reduce the amount of chemicals applied, since some
adversely affect oil characteristics or odour. Improvements in soil manage-
ment, especially irrigation techniques, could be more widely applied. Water
X Introduction
xi
World Trade
in Essential Oils 1
Total production of essential oils and aromatic plant derivatives is almost
impossible to quantify, as considerable amounts may be used nationally
without being recorded in production statistics; India and the ASEAN
countries for example. National trade statistics can also be variously in-
terpreted as they may include several oils under one designation, may be
collected by different ministries, and are frequently very inaccurate. The
following tables and comments are a general indication of world trends, and
considered sufficient to support the text. The period to publication of a book
of this type outdates annual statistics, and retrospective data (UNCTAD,
1982, 1986; Lawrence, 1984), while comprehensive, is of historical interest
only in a rapidly changing industry, and marketing executives normally
commission their own confidential surveys. An example of a significant and
fast change in the market followed establishment of the North American
Free Trade Area in 1994. Mexican citrus oils, particularly lime, thus enjoy
preferential treatment and local production rapidly expanded at the expense
of former suppliers. China’s production of rose derivatives in particular has
greatly expanded, and now rivals traditional European producers.
Major producers of aromatic materials discussed herein are, in alpha-
betical order, the ASEAN countries, Brazil, China, Egypt, India, Indonesia
and the USA, but their relative ranking in terms of volume or value varies
according to product. However, China is becoming increasingly important
as a primary source since it remains one of the few countries with a huge
reservoir of low-paid labour, a major advantage in the production of many
aromatics, especially where flowers are the raw material (Zhao, 1993).
Calculated world usage of flavour and fragrance materials of natural origin
in 1984 and 1994 is shown in Table 1.1.
The member states of the European Union (EU) and Switzerland are the
largest world market for aromatic materials; many are imported, reprocessed
and re-exported and there is a significant internal trade. There are historical
reasons why the EU is a large consumer of essential oils, France in particular
is a long-established centre of the European perfume industry, but recently
1
2 Chapter 1
Table 1.1. World usage of flavour and fragrance materials of natural origin in
1984 and 1994 (US$ millions).
Table 1.2. Average annual imports of selected essential oils 1980-1990 (in tonnes).
Table 1.3A. United States imports of specified essential oils, 1985, 1990-1991,
Japan is the second largest national market accounting for about 10% of
world demand. There is little domestic production and major oil imports are
cedar, clove, lemon, orange and peppermint. Imports for the period 1984 to
1994 nearly quadrupled in volume from 6000 to 22,000 tonnes but, due to
a steady increase in the international value of the Yen, total value increased
by only one half from Yen 12 to Yen 18 billion. A major reason for the
increase in imports was a greatly expanded use in flavourings.
Caution is necessary in interpreting world trade statistics as it may appear
from published data there is a surplus of some oils or demand for others is
low, but a new entrant to essential oil production should not be discouraged.
An excellent example is that of the Tasmanian Natural Plant Extracts
Cooperative Society founded in 1990, which achieved export sales of nearly
AU$ 7 million in 1995. Agronomic techniques, plant breeding, product
promotion and marketing can be directly controlled by producers or their
organizations, to supply manufacturers who are continually seeking cheaper
ingredients, new products, new perfumes, new taste sensations, new what-
ever to drive the market!
Introduction of a novel oil, or expansion of an existing oil or derivative
available only in small amounts (i.e. osmanthus and boronia) offer profitable
niche markets, while a Zimbabwe grower has successfully concentrated on
producing organically-grown oils. A completely new associated market can
Table 1.5. The estimated average annual use of citrus oils in beverages (in tonnes).
also promote essential oil production. The export offresh jasmine flowers by
air from India to the Gulf States and Arabia, for example, resulted in
research into prolonging flower life and increasing oil odour and persistence,
all of considerable importance to concrete producers. Oil derivatives can
frequently be processed locally into higher value-added products, of partic-
ular relevance to developing countries. Many components, some in very low
amounts, have an important role in determining the sensory character of an
oil, while others in greater abundance can be extracted to improve quality or
reduce transport costs, and are themselves saleable.
Governments in developing countries should thus support independent
rural cooperatives as the most suitable system for increasing essential oil
production. Such societies provide a focus for introducing new methods,
obtaining credit, owning small processing plants, transport and marketing
facilities. It is vital to the success of such initiatives that the maximum of
locally available inputs be used, even if this reduces potential output to some
degree. It may, for example, result in greater use of manpower rather than
mechanical equipment, since the opportunity cost of labour in many devel-
oping countries is virtually nil.
It may also be profitable to concentrate on low-input techniques to obtain
a small quantity of high quality oil, or a larger amount of crude oil for local
processing. Traditional methods of crop production in underdeveloped
regions, although frequently outdated, often have a sound basis for their
continued use. Replacement by more modern techniques may well require
so great a social restructuring of the local community as to make the change
very difficult, sometimes virtually impossible, as the author well knows.
Where the introduction of farm or ancillary equipment is possible, the
emphasis should be on reliability, with oil quality the target rather than
quantity. Solar energy converters for instance are becoming more efficient,
and the equipment is not only relatively cheap with a low maintenance
requirement, but also can be easily transported and erected in remote areas.
Heating water for stills and reducing the need for wood or similar fuels is
only one use. Hot water for personal hygiene also generally improves
community health and thus the potential for increased manual labour.
As human populations increase so will pressure on available land to grow
food, fuel, or provide housing, and an important feature of essential oil
plants is that some local cultivars or landraces have drought resistance,
tolerance of saline soils or water, and thus are able to utilize environments
unsuitable for other crops. Large commercial plantations of many tropical
essential oil crops were unusual, mainly because of the labour intensive
methods previously necessary. Machinery and equipment now available,
together with the technology involved in its use, have radically altered the
potential for large-scale essential oil or spice production. Examples mentioned
in the text are eucalyptus, ginger, tea-tree, and some citrus species whose
r------r--1
manufacturer
Production of manufacturer
essential oils
as by-products .+
FLAVOUR AND
FRAGRANCE HOUSE
Subsidiary
Flavour and
manufacturers
t c c c c
I f , I
Wholesaling networks
Consumers
Fig. 1 -1, Structure of the international flavour and fragance industry, (Hunter, 1996.) NFDA, National
Food and Drug Administration; IFEAT, International Federation of Essential Oils and Aroma Traders;
FEMA, Flavour and Extract Manufacturers Association; IFRA, International Fragance Association; RIFM,
Research Institute for Fragrance Materials; IOFI, International Organization of Flavour Industries.
8 Chapter 1
Bibliography
The following list includes references quoted in the introduction and this
chapter, important or interesting publications, or those covering aspects of
essential oils not discussed in this book. The proceedings of the various
international congresses on essential oils, fragrances and flavours also pro-
vide valuable information and references. Important periodicals directly
concerned with essential oils, associated products and their uses, are the
Journal of Essential Oil Research U.Ess. Oil R.), Perjhmer and Flavoriit
(Pe$um. Flav.) and Flavour & Fragrance Journal (Flav. Frag. J.). Access to
a computer also enables a wide range of databases to be consulted. Im-
portant industry organizations are the Research Institute for Fragrance
Materials (RIFAM), the Flavour and Extract Manufacturers Association
(FEMA), the International Fragrance Association (IFRA), the International
Organization of Flavour Industries (IOFI), and the International Federation
of Essential Oils and Aroma Trades (IFEAT).
World Trade in Essential Oils 9
Atal, A.K. and Kapur, B.M. (1982) Cultivation and Utilization of Aromatic Plants.
CSIR, Jammu Tawi, India.
Arctander, S. (1960) Perfume and Flavour Materials of Natural Origin. ,Mrs M.G.
Arctander, 6665 Valley View Blvd, Las Vegas, Nevada, USA.
Baerheim, S.A. and Scheffer, J.J.C. (1989) Essential Oils andAromatic Plants. Dr W.
Junk Publications, The Hague, The Netherlands.
Burkill, I.H. (1966) A Dictionary of the Economic Products of the Malay Peninsula, 2
vols. Government of Malaysia, Kuala Lumpur, Malaysia.
Grisanti, E.P. (1993) The challenges of a United Europe for the flavour and
fragrance industry. Perfum. Flav. 18(1), 3-7.
Guenther, E. (1 949-52) The Essential Oils (6 vols). D. van Nostrand Co., New York,
USA.
Hay, R.K.M. and Waterman, P.G. (1993) Volatile Oil Crops: their Biolosy, Bzo-
chemistry and Production. Longman, Harlow, Essex, UK.
Heath, H.B. (198 1) Source Book ofFlavours. The AV1 Pub. Co., Westport, Connecti-
cut, USA.
Husain, A. et al. (1988) Major Essential Oil-bearing Plants of India. Central. Inst.
Medic. & Aromatic Plts, Lucknow, India.
Hunter, M.V. (1996) The flavour and fragrance industry; structure and future
trends. Cosnzetzcs Aero. Toilet. 9(6), 20-31.
Lawless, J. (1992) The Encyclopaedia of Essential Oils. Element Books Ltd, Shaftes-
bury, Dorset, UK.
Lawrence, B.M. Reviews of Essential Oils. Published regularly in Perfumer and
Flavorist. Allured Pub. Corp., Wheaton, Illinois, USA.
Lawrence, B.M. (1984) A review of the world production of essential oils. Perfum.
Flav. 10(3), 21-33.
Miller, J.I. (1969) The Spice Trade of the Roman Empire. Oxford Press, Oxford,
UK.
Parry, J.W. (1953) The Story of Spices. Chemical Pub. Co., New York, USA.
Poucher, W.A. (1993) Poucher’s Perfumes, Cosmetics and Soaps. 9th edn, 3 vols.
Chapman & Hall, London, UK.
Purseglove, J.W. (ed.) (1981) Spices (2 vols). Longman Group, Harlow, Essex,
UK.
Rosengarten, F., Jr (1969) The Book of Spices. Livingstone Pub. Co., Wynnewood,
USA.
Stobart, R. (1977) Herbs, Spices and Flavourings. Penguin Books, Harmondsworth,
Middlesex, UK.
Suleiman the Merchant - 85 1 AD (1922) Memoirs. Gabriel Ferrand, Paris, France.
Tisserand, R. (1987) The Art ofAromatherapy (Rev. edn). C.W. Daniel Co., Saffron
Walden, Essex, UK.
UNCTAD/GATT (1974 and 1986) Essential Oils and Oleoresins. Geneva, Switzer-
land.
UNCTADiGATT (1982) Spices: a Survey of World Markets, 2 vols. Geneva, Swit-
zerland.
Watt, G. (1908) The Commercial Products of India. John Murray, London, UK.
Weiss, E.A. (1993) Linola - A new edible oil. Oils G. Fats 9(3), 23-24.
Zhao, Q. (1993) China’s perfumery industry picks-up. Pe$uuMz. Flav. 18(1), 47-48;
also 16(5), 51-52.
Annonaceae
2-
The Annonaceae, of the order Magnoliales, is a very large and rather ill-
defined family with 120 genera of mainly tropical species, half in Asia,
Australia and Polynesia, one-third in tropical America, the remainder in
Africa. The genus Cananga (DC) Hook. f. & Thoms. contains two species:
Cananga odorata, source of cananga and ylang-ylang oil, and C. latifolia Gin.
& Gagnep., producing an oil of no commercial importance. A liana Arta-
botrys odoratissimus (Lam.) Men-., known as the climbing ylang-ylang be-
cause its flowers resemble those of cananga, is not utilized for oil produc-
tion.
Cananga
10
Annonaceae 11
Botany
Canunga odoruta (Lam.) Hook. f. & Thoms. (syn. C. odorutum Baill.;
Canatzgiuwz odoratum Lam.). Commonly cananga or ylang-ylang in English;
India (Tamil) karumugai; Sri Lanka wanasapu; Philippines ilong-ilang;
Burma kodapnyan; Malaysia kenanga, which is believed to be the origin of
the genus and English name.
Th e basic chromosome number of the genus is x = 8, Cananga odoruta 2n
= 16, and although there are no accepted varieties two types are now
recognized which have minor botanical differences but produce two differ-
ent oils. C. odorata forma macrophylla produces cananga oil, and C. odoruta
forma genuinu ylang-ylang oil. In general, cananga oil is obtained from
flowers on uncultivated trees, while ylang-ylang oil is obtained from planta-
tions. The following description applies to both, but comments on cultiva-
tion, etc. refer to the production of ylang-ylang oil from planted trees.
The difference between cananga and ylang-ylang whole oils is believed to
be due to selection by growers, thus cultivated trees produce flowers with
required oil characteristics. Reprocessing of cananga whole oil, however, can
produce various grades of ylang-ylang oil as will be discussed later. It is also
possible that isolation on various Pacific Islands has resulted in local cananga
trees producing slightly different oils while remaining botanically almost
identical; C. odoruta trees growing in Java, Fiji and Samoa produce oils
which have quite different characteristics (Table 2.1).
Annonaceae 13
Cananga Ylang-ylang
Reunion Comoro Philippines Madagascar
Java Fiji extra extra extra extra
Specific 0.896 0.932 0.911 0.965
gravity (20") 0.942 0.421(28') 0,969(15') 0.975 0.958 0.970
Optical -27" - 7O -30" -19" -30" -25O
rotation (20') -87O -35O - 45" -45" -35s
Refractive 1.4788 1.491 1.465 1.4948 1.4747 1.500
index (20") 1.5082 1,494(25') 1.475(15") 1.4900 1 ,05(20°)
Solubility (viv 1:l-2 - 110.5 110.5 1 :0.5 1:0.5
95% alcohol)
Ester value 42-94 75-84 96-134 200 90-1 50 125-1 60
Figures in columns are ranges.
Source: Duve et al. (1975).
Fig. 2.1. Cananga odorata: Flowering shoot and fruit. (Department of Agriculture, Manila, Philippines.)
Annonaceae 15
Characteristic
(at 20'Cci A G R A G R A G R A G R
Specific 0.950 0.946 0.950 0.933 0.928 0.932 0.923 0.918 0.922 0.906 0.906 0.910
gravity 0.965 0.982 0.986 0.945 0.949 0.953 0.929 0.933 0.937 0.921 0.923 0.920
Optical -45" -40' -40" -44 -60° -60° -55" -68' -68$ - 63" -67" -67O
rotation -36" -25" -25" -28" - 4 4 -33" -40" -40" -40' -49" -3v -35"
Refractive 1.501 1.498 1.498 1.500 1.500 1.500 1.505 1.506 1.507 1.506 1.507 1.507
index 1.509 1.509 1.509 1.510 1.509 1.509 1,511 1.512 1.517 1.513 1.515 1.515
Acid value c3 c 2.8 (2.8 c3 c 2 . 8 c 2.8 c3 c2.8 c 2.8 <3 c2.8 c2.8
Ester value 125 130 132 90 89 89 65 56 56 38 34 34
160 182 162 120 130 130 80 89 89 58 54 60
Figures in columns are ranges.
Source: A, Afnor (1986);G, Grasse Essential Oil Association; R, Reunion.
16 Chapter 2
Ecology
Cananga is native to monsoon areas with a rainfall of at least 3000 mm
annually. It also grows well under more dry conditions, but trees which
suffer a moisture stress flower less profusely and flowers have a lower oil
content but not necessarily of lower quality. Under suitable conditions,
cananga quickly becomes established and self-sown naturalized trees now
occur on various Caribbean islands, Central America, in monsoon India and
on Zanzibar. Cananga is basically a forest tree and requires shade when
young, but heavy shade when mature retards growth and plantation trees
must be spaced so all receive direct sunlight. Shelter belts are advisable in
areas where there are strong winds. Cananga plantations are usually estab-
lished at low elevations, but there are no data on the effect of increasing
altitude on tree growth, flowering, or oil content of flowers. In China, where
cananga is grown in areas with low winter temperature, trees are protected
with straw or plastic sheets to minimize frost damage (Liao, 1984); however,
cananga should not be grown commercially in regions where frosts occur, as
the cost of protection is normally unprofitable.
Cultivation
T o establish a plantation, the area must be cleared and all debris removed or
burnt and the ash spread as fertilizer. Where high winds are common, it is
advisable to leave existing trees on borders or between planted sections as
wind-breaks. Plants may be grown from seed sown in nursery beds, but the
long easily damaged taproot, which is quickly produced following germina-
tion, renders transplanting difficult, and seed should preferably be sown in
Annonaceae 17
prepared pits where trees are to grow. Alternatively, pots should be large
enough to allow unrestricted root growth and made of biodegradable
material to allow planting directly into holes. Several seeds should be sown
per pot or pit, thinned to the most vigorous. Seed viability is frequently low
but as seed is normally available in large quantities, this presents few
problems. Fresh seed reportedly germinates less readily than seed 6-12
months old, and there may thus be a dormancy period. Soaking seed in hot
water for several hours is considered to hasten and assist germination on
Nossi-Be. Propagation by cuttings is also possible but seldom used.
Pits should be prepared well in advance of sowing, be at least 50 cm deep
to allow unrestricted taproot growth, and filled with topsoil or a mixture of
soil and well-rooted organic material. Seed should be placed about 5 cm
deep and seedlings kept moist and shaded. Initial spacing should be at least
6 X 6 m, with alternate trees removed to leave a triangular spacing should
this be necessary as trees mature; on sloping land or hillsides, trees should be
planted in rows following the contour. A circular area around seedlings
should initially be kept weed-free and inter-rows slashed or mown. Herbi-
cides are seldom used. Cananga can be intercropped in the first 2-3 years
with low-growing food crops such as pulses, but underplanting cananga in
plantations of other trees is seldom successful.
Trees are usually topped at about 3 m and pruned to maintain this height
(see Fig. 2.2) and may also be grafted in China (Liao, 1984). Topping
encourages growth of the naturally drooping lower branches, which are also
pulled down by ropes attached to small posts to keep flowers within reach of
pickers, since branches are brittle and climbing to pick flowers can cause
extensive damage.
Harvesting
Trees flower from the second year, when a small harvest may be taken, and
flower profusely from the fourth or fifth year. Flowers have little fragrance
initially, and green petals covered with fine white hair which disappears with
maturity. After 14-21 days the petals become pale green, then yellow and
their scent progressively stronger. When two small reddish spots appear on
base of the petals the flowers are ready for picking. This should be as early in
the morning as possible when oil content is highest, and normally ceases by
10 a.m., as in strong sunlight the oil content rapidly diminishes. Composi-
tion and characteristics of the oil changes as flowers develop, with the light
oxygenated compounds highest at maturity; thus inclusion of immature
flowers is detrimental to oil quality (Stashenko et al., 1995).
Rain has apparently little effect on oil content, but distilling wet flowers
reduces still efficiency and may also adversely affect oil quality. Maximum
flowering normally follows cessation of the rainy season, when 50-60% of
flower yield is obtained. Oil content is usually highest early in the dry season
and also of higher quality.
18 Chapter 2
Distillation
Flowers should be distilled immediately after picking to obtain highest yield
of best quality oil, but some delay is usual and flowers should be thinly
spread on some suitable surface out of the sun, as it is essential to avoid
fermentation. Wilting or delayed distillation normally reduces oil yield but
the effect on oil composition is less obvious; there is apparently little change
Annonaceae 19
Ylang-ylang
Extra 1 2 3 Cananga
Alpha-pinene 0.3 0.1 0.1 0.1
p-Cresyl methyl ether 8.4 3.1 1 .o 0.4
Benzyl alcohol 0.5 0.2 ng ng
p-Cresol 0.1 0.1 ng ng
Methyl benzoate 4.0 1 .o 0.3 0.1
Linalool 10.3 5.5 3.2 2.0
Benzyl acetate 12.6 4.2 1.2 0.5
Alpha-terpineol 0.1 ng ng ng
Methyl salicylate 0.1 0.1 ng ng
Geraniol 0.2 0.2 0.2 ng
Linalyl acetate 0.2 0.2 0.2 ng
Safrole 0.3 0.1 0.1 ng
Methyl anthranilate 0.1 ng ng ng
Eugenol 0.2 0.2 0.4 0.3
Geranyl acetate 4.0 3.0 2.2 2.0
lsoeugenol 0.2 0.4 ng ng
Beta-caryophyllene 6.8 11.5 12.8 16.3
Alpha-caryophyllene 3.1 4.2 4.1 8.8
Farnesene 18.0 16.8 17.0 21 .o
Delta-cadinene 8.9 15.1 20.4 16.3
Gamma-cadinene 2.0 5.3 7.0 7.5
Nerolidol 0.5 0.8 0.5 1.8
(Z).(E)-Farnesol 0.9 1.7 1.7 1.4
Benzyl benzoate 4.3 8.5 9.7 5.3
Benzvl salicvlate 1.9 1.2 2.7 1.7
ng, not given.
Source: F. Buccellato (personal communication).
22 Chapter 2
References
AFNOR (1986) Recueil de Nomzes Fraqaises des Huiles Essentielles. 2nd edn. Assoc.
FranCaise de Normalisation (AFNOR), Paris, France.
Anon. (1989) Bibliography of Cananga odorata. Med. Arom. Plt. Absts. 11(1), 91-93.
CSIR, New Delhi, India.
Anzaldo, F.E. (1980) Essential oil production in the Philippines. N S D B Tech. J.
5(2), 67-71.
Bacon, R.F. (1908) Philippine terpenes and essential oils. Philipp. J. Science (Sect. A)
3(1), 65-86; also (4) 127; ( 5 ) 265.
Buccellato, F. (1982) Ylang survey. Perfum. Flav. 7 ( 5 ) , 9-12.
Burkhill, L.H. (1966) The Dictionary of Economic Products of the Malay Peninsula, 2
vols. Government Printer, Kuala Lumpur, Malaysia.
Demarne, P-E. (1989) Production d'Ylang-Ylang et de Poivre Vert a la Societe Agricole
de Bevoay, Nossi-Be (Madagascar). CIRAD-IRAT, Saint-Denis, Reunion.
Ding, J.K. et al. (1988) Studies on constituents of the essential oils of different
varieties of C. odorata in different flowering periods. Acta Bot. Yunnananica
10(3), 331-334.
Duve, R.N., Vithalbhai, C.L. and Smith, R.M. (1975) Chemical composition of the
essential oil from flowers of endemic Cananga odorata from Fiji. Flavours
December: 341, 346.
Fisher, D.G. (1990) Leaf structure of C. odorata; morphology and anatomy. Can. J.
Bot. 68, 534-563.
Gaydou, E M et al. (1986) Composition of the essential oil of ylang-ylang from
iMadagascar. J. Agric. Food Chem. 34, 481-487.
Gaydou, E.M. et al. (1988) Multidimensional data analysis of essential oils: ylang-
ylang. J. Agric. Food Chem. 36, 574-579.
Guenther, E. (1952) Oil of cananga and ylang-ylang. In: The Essential Oils, vol. 5.
Van hlostrand, New York, USA, pp. 267-316.
Annonaceae 23
Pelargonium
The first pelargonium, P. cucullatum (L.) L’Herit. was collected from Table
Mountain in 1672 by Paul Herman, and South Africa is the centre of origin
of the genus. Van der Walt considers the reason so many species, about 70,
are found in the southwestern Cape region is the availability of a variety of
microhabitats favouring speciation. When Paul Herman published Horti
academici Lugduno-batavi catalogus in 1687 listing plants growing in the
Leyden Botanic Gardens, the Netherlands, ten Pelargonium species were
included. William A.van der Stel is believed to have imported P. peltatum L.
and P. zonale (L.) L’Herit. into the Netherlands in 1700, introduced to
England in 1701 and 17 10 respectively, and P. gruveolens L‘Herit. in 1794.
British control of the Cape of Good Hope greatly increased interest in
pelargoniums, with numerous papers in the Botanical Magazine, and pub-
lication of beautifully illustrated books including the five-volume Ger-
aniaceae by Sweet (1 820-26), Flore de Serres et des Jardins de 1’Europe (1 847)
and Les Geraniums (1 897).
24
Geraniaceae 25
but when the essential oil type was first introduced is unknown; recent
introductions are by private companies seeking superior strains. Two basic
oils are produced commercially named for the main producing regions of
Faiyum and the Delta (Benveniste and Azzo, 1992).
Following introduction to Reunion in 1870, geranium steadily became an
important local cash crop, and the planted area reached a record 15,000 ha
in 1952, but production has since fluctuated with the international oil price
(Sens, 1994). The first oil produced on Reunion was from a species of
unknown origin occurring on the island. It was quickly replaced by imported
Grasse material which became the distinct local cultivar Rose, later deter-
mined to be a P. capitatum X P. radens cross (Demarne and Van der Walt,
1989; Demarne, 1992). Chemotypes may also exist, as plants from Reunion
cultivated elsewhere produced an oil with a different composition to the
majority of those grown on the island (Mallavarapu et al., 1993). The main
plantations are in the west and south west on steep slopes virtually useless for
other crops, a major reason why oil production on Reunion continues. The
majority of geranium plantings are small individual plots whose owners
either combine to operate collective stills, hire a portable still or sell their
herbage to one of the larger operators with modern stills producing high-
quality oil (Demarne, 1992). Reunion lies in the Indian Ocean cyclone belt
and these terrible storms periodically play havoc with the industry.
Geranium oil has been produced in a number of East and Central African
countries, principally Kenya, but oil production has virtually ceased except
for a small quantity for domestic consumption in South Africa. It is believed
geranium was first introduced into Kenya either from South Africa by Cape
Dutch settlers, or from India by engineers engaged to build the Uganda
railway (Matheson and Bovill, 1950). These species are quite different to the
indigenous Pelargonium spp. which occur wild in East Africa, the most
common being the yellow-flowered P. fischeri. The first commercial planting
was at Njoro in 1925 using introduced species, which produced an oil
known as Mawah oil, the Ki-Swahili word for flowers generally. By 1937
there were 800 ha under geranium, but production virtually ceased during
the Second World War and, although subsequently resumed, it never
reached the original level. The author established in Western Kenya in 1964
a commercial plantation of 100 ha using plants selected from Algerian and
Reunion cultivars imported by the Department of Agriculture (see Fig. 3.1).
This operated profitably for several years, producing oil midway between
Mawah and Reunion types, but production ceased when the company was
taken over. This was the last attempt at large-scale geranium oil production
in eastern and central Africa. Geranium oil was also produced in northern
Tanganyika (Tanzania) from the late 1940s, but never expanded despite
official support. Subsequent trial plantings in several regions indicated a
substantial area was suitable for geranium as an oil crop (Weiss,
Geraniaceae 27
Botany
The genus Pelargonium L’Herit. contains about 250 species divided into 15
sections, with some 200 species occurring in South Africa, 18 in the rest of
Africa, eight in Australasia, two in Madagascar, two in the Middle East, and
one each on the islands of St. Helena and Tristan da Cunha (Van der Walt
and Ward-Hilhorst, 1977; Webb, 1984; Van der Walt, 1985). The genus
name is derived from the Greek pelargos, a stork, and the first legitimate
publication of the name pelargonium is considered to be in Aiton’s Hortus
Kewensis, 1789, which contained much material from L’Heritier’s un-
published work Compendium Generlogium. Within the genus fifteen sub-
genera are recognized, including Ciconium, Dibrachya and Pelargonium con-
sidered as having great influence on modern pelargoniums. The subgenus
(section) Pelargonium contains about 25 species (Van der Walt, 1985),
mainly shrubs or sub-shrubs with woody, branched or very branched stems,
many with variegated or fragrant leaves, and often drought-resistant. The
basic chromosome number of Pelargonium is x = 11, with the somatic num-
ber of 2n = 88 for P. graveolens and its close relative P. radens, while P.
capitatum is 2n = 66 (Albers and Van der Walt, 1984), but the important
Reunion cultivar Rose is a heptaploid, 2n = 77 and male sterile.
The four species basically concerned in the development of essential oil
cultivars but which have not been cultivated commercially to produce an oil
are Pelargonium graveolens, P. odoratissimum,P. capitatum and P. radens, while
P. roseum is considered a P.graveolens hybrid. A most important commercial
essential oil producer is the cultivar Rose, a hybrid of P. capitatum X P.
radens but often referred to as P. graveolens in the literature without qualifica-
tion, and to a lesser extent cultivars of P.capitatum X P.graveolens. Many
other Pelargonium spp. contain an essential oil described in the literature, but
none are commercially cultivated.
Pelargoniums cultivated for their essential oil were often identified in the
literature as P.graveolens cultivars regardless of botanical accuracy. Most
were selections originating in Europe which became locally acclimatized
Geraniaceae 29
when planted elsewhere, but some are deliberate hybrids. Most hybrids are
crosses between two of the four species described below.
Pelargonium graveolens L‘Heritier ex Aiton (syn. Geranium radula Roth.; P.
intermedium Knuth), commonly geranium in English and French is a small,
much-branched, perennial shrub to 1.3 m but generally less, naturally
forming a dense spreading bush. The leaves are fragrant, mint-scented, on
stems and branches, opposite, 6.5-13.5 cm by 6-12 cm, villous, highly
divided, with five to seven palmate lobes; the margins are toothed and
somewhat revolute; the petioles are long, sturdy, and subtended by stipules
(Fig. 3.2). Species colour is a dull green although grey-green, yellow-green,
dark or light green leaves occur on cultivars. An essential oil can be obtained
from leaves and fresh green shoots, but is of no commercial importance.
Herbage also contains tannins whose recovery in India is considered desir-
able (Balakrishnan et al., 1982). The inflorescence is axillary with small
umbels of 3-7 pink, rosy or pale purple flowers. The hairy calyx tube divides
in mid-length into five lobes and five entire petals, the upper two slightly
larger and longer, and pelargoniums are thus the only members of the
Geraniaceae with zygomorphic flowers. The ovary is superior, five-lobed,
each lobe containing a single pendulous ovule; the style is short and hairy.
The fruit is long and pointed, and although pelargoniums do not normally
produce seed, some cultivars seed readily. The seed is small, oblong-ovoid
and brownish, the testa is hard with attached long hairs. Viability is usually
very poor and germination erratic.
Pelargonium odoratissimum (L.) L’Herit. (syn. Geranium odoratissimum L.)
is a perennial, prostrate spreading shrub with caudex stems, normally short
but up to 30 cm, usually with many slender, weak, soft-pubescent, fre-
quently trailing branches. The leaves are fragrant, apple-scented, light
green, soft, round-cordate, very obtuse, blade 2.5-3.0 cm, similar in
breadth, obscurely three or more lobed, and the margins dentate-crenate.
The petioles are up to 10 cm long, stipule is triquetrous or broadly ovate,
usually connate; the peduncles are long and opposite leaves. The inflores-
cence bears 3-10 tiny flowers, which are white (occasionally pink), with
small red spots; the calyx is more or less pubescent, the spur 8 mm or less;
the sepals or lobes lanceolate, acute, with transparent margins; the petals, to
16 mm long, are linear-spatulate, with a rounded apex.
Pelargonium radens H.E. Moore (syn. P. radula (Cav.) L’Herit.) is an erect,
perennial densely branched shrub with a stout stem to 1.5 m. The leaves are
fragrant, mint-scented, deeply palmately-parted; the lobes are narrow linear,
pinnatified; rough-hispid above but soft-pubescent below, with the margins
prominently revolute. The peduncles are short and hispid, bearing 3-8 pale
purple flowers with darker purple streaks and margins. The flowers are
pedicillate, the calyx tube is short, and the lobes on the sepals are setose and
glandular.
30 Chapter 3
Shoot Inflorescence
Flower Petals
Leaves
this could partially account for the reported differences in general growth, oil
composition and characteristics from such local ‘Bourbon’ plants.
Cultivar Rose is an erect, perennial shrub to 1.3 m, forming compact
clumps to 1 m diameter, with an extensive, spreading root system seldom
penetrating below 30 cm. The multiple stems are soft, green to grey-green,
becoming darker and woody with age, and are more or less hairy. Junctions
of main or large secondary stems with the root at ground level often become
highly lignified and brittle with age. The number and length of stems and
degree of branching varies; an unnamed Algerian cultivar stated to have
originated on Reunion and grown by the author in East Africa, tended to be
more upright than the well-branched type directly imported from Reunion.
Stem height and branching may be modified or stimulated by growth
regulators (Bhattacharjee and Thimmappa, 199l), while selection for a
more upright plant would assist mechanical harvesting.
The leaves on stems and branches are opposite, fragrant, rose-scented,
soft, hirsute, with 5-7 palmate lobes, highly divided; the base is cordate, the
apex obtuse, with somewhat revolute margins. Size is variable but averages
50 X 70 mm, petiole 30 mm; the stipules are asymmetrical, triangular, 6 X
4 mm (Fig. 3.3). An essential oil is obtained from leaves by distillation or
solvent extraction, whose main constituents are citronellol, geraniol, iso-
menthone, citronellyl formate and geranyl formate (Table 3.1). The main
characteristics of Reunion and other oils are shown in Table 3.2.
The inflorescence is terminal and the peduncle 15-60 mm long with small
compact pseudo-umbels of 3-1 0 small, rose-violet coloured flowers. The
flowers generally are as P. cupitutum, with the important exception that male
flowers are sterile. Main flowering in Reunion is in spring and again in
August-September. Oil content of flowers has been reported at 0.6%
(Fleisher and Fleisher, 1985), and although this may be higher than leaf oil
content, flowers contribute little to total oil yield. Flower oil is also very
similar to leaf oil, and nothing is gained by separate distillation, encouraging
flowering or increasing the number of flowers per plant. Where the last two
are desired, spraying with gibberellin (GA) may produce the required effect,
and usually prolongs flowering. There is usually one main flowering period,
but plants should not normally be allowed to reach full bloom before cutting,
since oil content and composition are adversely affected.
The oil is contained in small glands distributed over green parts of the
plant, and their number and size govern the amount of oil a plant can
produce. Since oil glands are most numerous in the leaves, leaf size and
number are important factors affecting profitable oil production. Personal
observation and discussions with growers confirmed there are differences
between individual plants in any large population and, as noted previously,
this may be due to the original introductions. Considerable variation was
also noted in the rate of regrowth and development of leaves after cutting in
plants from cultivars grown by the author in western Kenya. Growth
32 Chapter 3
regulators can increase leaf size, but in Egypt GA affected the ratio between
oil components (Eid and Rofeel, 1980).
The biosynthesis and metabolism of oil in leaves has still to be fully
determined, although some pathways have been described (Chubinidze et
al., 1993). The rate of oil accumulation and its composition however, vary
with leaf age, and oil content and geraniol content of oil are highest in
youngest and lowest in oldest leaves. Oil and geraniol content of young and
old leaves in India was, in per cent, 1.56 and 34.6,0.05 and 13.9 respectively
(Rajeswara Rao et al., 1993). Oil content is usually stated as a proportion of
herbage harvested by weight (kg oil per tonne), and thus varies with the age
of plants at cutting, time of year and efficiency of still operation. The
following figures are from supervised commercial plantations with modern
stills designated C, or controlled field experiments E: India 0.04-0.13% E
(Balakrishnan et al., 1982; Kumar et al., 1985) and 0.10-0.15% C (Dor-
aswamy and Sundaram, 1982); Kenya 0.07-0.1 1% C (Weiss, 1966), and
Mawah type 0.10-0.20% C (Beckley, 1931); Reunion 1.5-3.0% C (De-
marne, 1992); Tanzania 0.05-0. l l % E (Weiss, 1955-60); Algerian 0.17%
Table 3.1, Main oil constituents from cultivar Rose from Reunion.
Ecology
Basically a warm-temperate plant, geranium will produce a profitable oil
yield in sub-tropical regions under suitable management, within the con-
straints of a frost-free growing period, sufficient rain and a mainly cold dry
season. The effect of location on oil yield and characteristics is well docu-
mented, but little data are available on how individual climatic factors other
than rainfall influence the oil or its major components. Thus it is almost
impossible to predict the type of oil which will be produced when plants are
introduced to a new area. While the oil will be basically that of the original
cultivar, its individual fingerprint will only be established by analysing oil
produce from plants in their new environment.
French oil from Grasse is considered to have the finest and most rose-like
odour. Reunion oil, rich in citronellol, may have a minty odour but is now
considered to be the standard by which other oils are judged. Moroccan oil
approaches that of France, while Russian oil is often less strongly scented.
East African oils were of good quality but in a category of their own. Spanish
oils generally contained around 45% geraniol and 25% citronellol; from
Japan not more than 22% geraniol; Israeli oils contain little geraniol but 45%
citronellol. Oil produced from P. radens in Zimbabwe was dark with a
pronounced minty note (Sholto-Douglas, 1969).
Geranium is grown commercially between 300-1200 m in the south of
France, Algeria, and Reunion where a limit of 1400 m is observed; in the
Kenya highlands between 2000-2500 m; between 1500-2000 m in the hill
country of India; from 0-500 m in the Black and Caspian Sea regions of
Russia; and at low altitudes in Australia and the USA. Plants grown at higher
elevations in sub-tropical climates frequently produce a lighter oil than oil
from the same cultivar grown at lower altitudes. In Tanzania, a Reunion
cultivar grown at 1200 m produced oil with a much softer more rosy note
than oil from plants growing at 250 m (Weiss, 195551960), while in
Zimbabwe oil from plants grown at 330 m resembled true Reunion type,
Geraniaceae 35
while oil from plants at a higher altitude resembled a lighter Algerian oil
(Sholto-Douglas, 1969). Plants of the same cultivar grown at three altitudes
in India, 542 m, 900 m and 2200 m gave highest yield at 542 m and lowest
at 2200 m, but oil from the latter plants had more menthone, citronellol and
nerol; geraniol rose progressively from 22.4% at 542 m to 30.2% at 2200 m
(Rajeswara Rao et al., 1990a).
Geranium produces maximum leaf growth with high oil content under
warm sunny conditions, and cloudy days or low temperatures adversely
affect both. A daytime temperature of 20-25°C is the optimum, but 5°C
above and below is acceptable; below 6°C normally inhibits growth, below
3°C is frequently fatal if prolonged, and frost at any stage of growth usually
kills. In the Caspian region of Russia where it is grown as an annual,
200-220 frost-free days are required and the necessity to replant each year is
very costly. It is very doubtful if geranium to produce oil can be grown on a
large scale as an annual and be commercially viable.
Annual rainfall of 1000-1500 mm evenly distributed over the main
growing period is required but regions with sharply defined wet and dry
seasons can produce good crops provided there are no extended periods of
waterlogging. Very heavy rains following onset of the monsoon, or cyclones,
can be most damaging; on Reunion for example a majority of plants may be
killed when soils remain saturated. Prolonged mist or high humidity reduces
leaf growth by encouraging fungal disease. Although geranium is drought
tolerant, long dry spells severely retard growth, reduce oil content, and
change oil characteristics. The effect in Kenya on oil and herbage yield of a
mainly dry 3 month period, compared to a mainly wet 3 month period is
shown in Table 3.3.
Thus the climatic factors producing a higher yield of herbage have the
opposite effect on oil, and vice versa; this is important to growers insofar as
a reduction in one was compensated by a rise in the other, the yield of oil per
hectare being similar. Oil produced after 3 wet months however, had a
slightly milder note than oil produced after 3 dry months (Weiss, 1967).
Plants cut in India in the spring (semi-dry period) also had a greater oil
content than those cut after monsoon rains (Mani et al., 1982; Prakasa Rao
et al., 1995), while oil from plants grown in semi-arid conditions showed
similar variation (Bhattacharya et al., 1993). Where there are no pronounced
Table 3.3. The effect of dry and wet periods on oil and herbage yield of
geranium in Kenya.
Oil
Herbage
tiha kgit kgiha
After 3 wet months 28.2 0.73 20.6
After 3 dry months 21.8 1.07 23.3
36 Chapter 3
generally increased herbage yield (Rajeswara Rao et al., 1990b, c). Very high
levels of ammonium sulphate were applied to an irrigated crop in Israel; 500
kgiha was soil incorporated prior to planting, 300 kgiha as a top dressing
before the first cut, and a similar amount after each cut. This produced the
extremely high yield of 70 tiha from three cuts in a year, and a calculated
total oil yield of 185 kgiha (Fleisher and Fleisher, 1985). There is usually
little or no difference in green material yield between the various nitrogenous
fertilizers, and the cheapest is normally applied. While high levels of nitrogen
generally promotes plant growth, related effects may modify a yield increase;
rust caused by Puccinia pelargonii-zonalis is often more severe.
Phosphate is usually necessary to ensure rapid rooting of cuttings and
subsequent herbage production, but has little effect on the oil content of
leaves or oil characteristics. Increasing seedbed superphosphate (17-18% P)
from 0 to 135 kgiha doubled herbage yield in the first and second year after
planting in Kenya, but yield rapidly declined in the third and subsequent
years, and a further application would probably have been profitable (Weiss,
1967,1969). In India, 25-40 kgiha superphosphate is common and a similar
amount is often applied in alternate years over the life of a plantation (Gulati
etal., 1982; Kumaretal., 1985). Acompoundfertilizerof50:130:lOONPK
at 600-800 kgiha increased yield on Reunion by 30% with phosphate the
most important (Gailleton, 1967), but the standard sugarcane fertilizer of
15:7:24 or 18:7:30 NPK is now more commonly applied at levels related to
the average oil yield per hectare; at 200 kg oil it is 200 kgiha, at 30-40 kg oil
it is 400-600 kgiha, and above 50 kg oil it is 700 kg of 9.5:14:30 NPK plus
600 kg ammonium nitrate or 300 kg ureaiha (Chablier, 1992).
Potassium levels, especially the ratio to nitrogen and phosphate, must be
accurately determined, since response to this nutrient is usually erratic. It is
frequently part of a compound NPK fertilizer, but may not be required or
required at a much lower level than available in the compound. Experi-
mental results in India showed potassium is frequently lacking in many
geranium-growing regions; potassium chloride is commonly recommended
at an annual rate of 35-70 kgiha (18-36 kg K).
Calcium in the seed-bed frequently increases herbage yield in the pres-
ence of adequate amounts of the major nutrients, and is often applied in
conjunction with magnesium; as dolomite limestone or ground magnesium
limestone when locally available. The amount applied is usually small, since
geranium prefers a slightly acid soil. From the small amount of data available
it appears no field planting of geranium has been adversely affected by a
minor or trace element deficiency. In pot trials magnesium frequently
increased herbage yield, and also in field trials when applied in the seed-bed.
No significant effect on herb oil content or oil composition has been
reported. Inclusion of molybdenum and copper increased herbage and oil
yield of well-fertilized plants in India.
38 Chapter 3
Cultivation
It is essential that lands should be well ploughed and all roots and other plant
debris burnt, as geranium plants are very susceptible to root infections. This
major cultivation should take place as far in advance as possible of planting-
out cuttings, since it will provide an invaluable opportunity to produce a
level seedbed and destroy all persistent weeds and grasses. The latter are
extremely difficult to remove once geranium plants are well established
without causing damage to the relatively shallow roots or extensive bran-
ches. Where geranium is grown as a smallholder crop the minimum of
cultivation is normal; on Reunion for example, trash from a preceding cane
crop is piled and burnt, and the land hoed prior to planting cuttings. On
sloping lands where terracing is not possible, permanent lines of soil-binding
plants such as Pennisetum spp. or Vetiveria spp. should be established at
regular vertical intervals, to reduce the danger of soil erosion. Geranium is
normally planted in pure stands for commercial oil production, but when
grown by smallholders may be interplanted. In general, the total return from
Geraniaceae- 39
Harvesting
The first cut should not be taken until plants are well established and bushy,
normally 6-8 months after planting out. Cutting too early after establish-
ment often kills many plants, or regrowth is very slow. When to cut may be
determined by weather or other farming operations, but to obtain maximum
oil yield plants should be routinely sampled to determine oil content;
however, cutting is usually related to plant height or flowering. Oil content
normally increases from onset of flowering, is highest at full bloom, and very
rapidly decreases thereafter (Table 3.6). In southern India, oil content was
highest in July and lowest in February, while specific gravity was highest in
January and lowest in September (Mani and Sampath, 1982). Oil composi-
tion also changes over time, which tracer studies have shown to be sub-
stantial (Banthorpe et al., 1983). These changes also correspond to the main
growth periods, and thus cutting at the optimum time for oil content also
generally produces the highest quality oil.
Cutting can be manual or mechanical at 12-20 cm above the ground, with
field trials to establish the most profitable height in relation to distillation
time and oil yield per tonne of material. The importance of cutting height
was demonstrated in India and Israel; leaf blades, petioles and flowers
contain practically all the oil, the stem little, and from 15 cm below the tip
virtually none (Vadivel and Sampath, 1982; Kumar et al., 1985; Fleisher and
Fleisher, 1985). The whole stem contains so little oil it is a waste of effort to
distil, but there is currently no practical method of harvesting leaves alone.
Smallholder plots are invariably cut by hand, with only green leafy shoots
selected to reduce volume throughput of stills, and a major pruning is
42 Chapter 3
Distillation
Once cutting has commenced on a commercial geranium plantation, stills
should be operated continuously to keep pace with the volume of material,
since it is generally unprofitable to have sufficient stills working only part of
the day to meet demand over the relatively short period of optimum herbage
oil content. Wherever possible, commercial plantations should install a
combination trailer-still system as described in the section on harvesting
eucalyptus. Harvesting can be extended provided it is accepted that oil
characteristics will probably differ during the period, and it will be necessary
separately to store the oil obtained until it can be reprocessed or blended to
the required type. The author obtained two quite different oils in the dry and
wet seasons in Kenya as noted, and cutting when plants were in full bloom
44 Chapter 3
tended to produce an oil with lower geraniol and citronellol content than
cutting prior to blooming. Figure 3.5 shows a still which was built under the
author's supervision in Kenya.
Allowing herbage to wilt before distilling depends on the facilities avail-
able; where it can be thinly spread on a clean shaded surface wilting can be
an advantage, but the duration must be accurately established (Chan-
dravadana and Kumar, 1989; Rajeswara Rao et al., 1992)). Field wilting has
a number of disadvantages; most importantly oil content falls rapidly in
strong sunlight, with the added risk of soil contamination.
Geranium herbage can be hydro- or steam-distilled, the oil obtained being
basically the same under a similar level of management. A substantial
amount of geranium oil is produced by small owner-operated unsophisti-
cated stills, whose efficiency in terms of cost per kilogram of oil is low. Such
producers should co-operate to establish a modern distillery to produce high
quality oil, while governments in some countries actively support the devel-
opment and introduction of more modern stills, Reunion for example (Vera
et al., 1995). Solar heating can substantially reduce the cost of pre-heating
water to produce steam and replace all or part of the huge amounts of wood
required; 200-400 kg wood per tonne of herbage processed is common. T h e
material used to construct still bodies may have an important effect on oil
composition, and must be considered when new distilleries are planned
(Guerere et al., 1985).
Herbage can be directly loaded or chopped and blown into the still body,
but whatever method is used, it is essential the charge should not be too
densely packed, as this will cause channelling or hot spots. Modern stills use
cohobation, since some oil remains in the distillation water. This oil, the
secondary oil, may reach 25% of potential oil yield and can be solvent
extracted. In Israel, secondary oil was higher in free alcohols but had a lower
ester content. Analysis of the two oils gave the following (secondary oil in
brackets): specific gravity (20"C), 0.899 (0.899), refractive index 1.468
(1.468), optical rotation -10.9" (-5.9"), acidvalue 2.6 ( 4 . 9 , estervalue 75
(36), ester value after acetylation 232 (250) (Fleisher and Fleisher, 1985).
When herbage is hydro-distilled the liquid remaining in the still body is
not equivalent to this secondary oil. It may have been overheated or
remained in the still for several distillations and contain compounds which
adversely effect oil quality. It should be discarded. Spent herbage should
preferably be composted for return to the plantation as mulch, or used on
other crops. On large plantations the volume is substantial and an efficient
method of removing it from the distillery for treatment or dumping is
essential.
Crude oil should be dried, filtered and stored in opaque airtight contain-
ers. Storage temperature should be below 10°C for extended periods and
not higher than 15°C for shorter periods. In general, acid number and
specific gravity increase and optical rotation and ester number decrease in
oils stored in ambient conditions. Opaque containers of glass, aluminium or
steel are preferred, as many plastic containers are unsuitable. An indication
of the physical and chemical changes which can occur in an Egyptian oil has
been published (Oda, 1982). Leaves may be solvent extracted with hexane,
but the concentrate must be again extracted with ethanol to remove the
wax.
Lawrence, 1994; Southwell and Stiff, 1995). Many Pelargonium spp. pro-
duce an essential oil and composition of the most important have been
reviewed (Lawrence, 1992). Oil from all origins is frequently adulterated,
the finer the original quality the more common the occurrence. Some
countries, including Reunion and Egypt, have a high standard of purity and
oils purchased directly from reliable suppliers are usually unadulterated.
Geranium
pilot projects have been unsuccessful. Wild plants in Bulgaria are re-
cumbent, with short stout stems arising from underground rhizomes. Leaves
tend to bunched towards the stem apex. Zdravetz is found in quantity
between 800-1700 m on rather infertile soils in Bulgaria; in Yugoslavia, soils
rich in calcium were reported to induce azulene concentration, while plants
growing in soils with a high selenium content had no azulene in the oil
(Mihajlov and Tucakov, 1975). Few pests or diseases have been accurately
recorded from wild zdravetz; recently Gnomopnia geranii-macrowhizi is de-
scribed as damaging roots in Bulgaria (Fakirova, 1995).
Harvesting normally begins in September and continues until snow falls
or the first frost. Herbage can be distilled fresh or wilted, with a yield of
Commercial State
sample' specification Guenther'
Density (40°C) 0.930-0.952 0.938-0.968 0.9515 (25°C)
Refractive index 1.5024-1.5135 1.500-1.519 1.5076 (25°C)
Optical rotation (40°C) - 4 8 to -8.6" - I 0 to -9" -4'35' (25%)
Acid number 1.2-1.87 Max 3.0 1.9
Ester number 4.97-7.91 3.5-8.0 6.5
(after acetylation) (20.75-49.78) ng (35.5)
Acetyl number ng 20.7-50.0 ng
Melting-point ("C) ng 34-50 ng
Free alcohols (%)a 5.80-1 1.88 4-1 2 ng
Total alcoholsa 7.70-14.50 5-1 6 ng
(combined alcohols) ng 1-4 ng
Gerrnacrone (%) 47.75-50.13 42-53 ng
~~~ ~~ ~
0.08-0.125%, and plants exposed to direct sunlight had a higher herbage oil
content than those growing mainly in shade. Plants growing in various
districts also had different herbage oil levels, indicating the potential for
increasing oil content by selection. Herbage is loaded into stills without
undue compaction, distilled for approximately 4 hours, and about 1 kg of oil
is obtained from 800-1 200 kg herbage. Hydro-distillation with cohobation
is normal, as this extracts almost all the oil and produces an oil closest to the
plant's natural scent.
Crude oil at ambient temperature is a bluish-green semi-viscous liquid
containing large colourless, odourless prismatic crystals. On exposure to air
the oil becomes gradually less green, finally a brownish green, the acid
number increases and the characteristic odour becomes modified by a
pronounced resinous note. This note is also very strong in cohobation water.
Since the crystals in crude oil are odourless, only the liquid (eleoptene) is of
interest to perfumers and contains 2% monoterpenes and 25% sesqui-
terpenens.
Zdravetz oil is a pale olive-green or pale yellowish-green somewhat
viscous liquid, which at room temperature is normally 30% liquid, 70%
crystals by volume; crystals begin to dissolve at 32°C and completely
dissolve between 45OC and 52OC. The main characteristics of zdravetz oil
are shown in Table 3.9, with alcohols calculated as geraniol although this is
absent from the oil; similarly germacrone has been traditionally quoted as
stearoptene. The main components of zdravetz oil are sesquiterpenes (to
90%), of which about half is germacrone (Ognyanov, 1985). Oil odour is
warm-fresh, sweetly herbaceous, rosy-woody, tenacious and used almost
54 Chapter 3
References
Albers, F. and Van der Walt, J.J.A. (1984) Untersuchungen zur Karyologie und
Mihosporogenese von Pelargonium sect. Pelargonium. Plt. Syst. Evol. 147(3),
177-188.
Arctander, S. (1960) Perjiume and Flavour Materials of Natural Origin. Mrs M.G.
Arctander, 6665 Valley View Blvd, Las Vegas, Nevada, USA.
Balakrishnan, R. et al. (1982) Potentialities of geranium in South India. In: Cult. Util.
Aromatic Plants. pp. 570-572. CSIR, Jammu-Tawi, India.
Banthorpe, D.V., Long, D.R.S. and Pink, C.R. (1983) Biosynthesis of geraniol and
related monoterpenes in P.graveolens. Phytochem. 22( 1 l), 2459-2463.
Beckley, V.A. (1931) Geranium. Bull. 19, Dept. Agric., Nairobi, Kenya.
Beneviste, B. and AZZO,N. (1992) Geranium oil. Kato Tech. Bull. &+Newsletter, 2,
1-8. Kato Worldwide Ltd., New York, USA.
Bhattacharjee, S.K. and Thimmappa, D.K. (1991) Effects of growth hormone,
length of cutting and number of leaves on the regeneration of adventious roots
in Pelargonium graveolens. Ann. Agric. Sci. Ain Shams University, Cairo 36(20),
57 3-578.
Bhattacharya, A.K. et al. (1993) Composition of the oil of rose-scented geranium
(Pelargonium spp.) grown under the semiarid tropical climate of south India. J.
Ess. Oil Res. 5(2), 229-231.
Brown, J.T. and Charlwood, B.V. (1986) The control of callus formation and
differentiation in scented pelargoniums. J. Plant. Physiol. 123(5), 409-41 7.
Chablier, P.F. (1992) La Fertilisation. In: Le Geranium Rosat a la Reunion, CIRAD,
Saint Denis, Reunion.
Chandravadana, M.V. and Kumar, T.V. (1989) Yield and oil quality of scented
geranium as affected by biomass storage, duration and different temperatures.
Ind. Perjium. 33(4), 246-248.
Charlwood, B.V. et al. (1988) Pelargoniums: flavour, fragrances and the new
technology. Plants Today (MarchiApril) pp. 42-46.
Chubinidze, V.V., Beriashvili, T.V. and Chubinidze D.V. (1993) The metabolism of
the volatile oil and citronellol in geranium (Pelargonium roseum L.) cells. Bioh-
khim 58(6), 874-878.
Cilek, J.E. and Schreiber, E.T. (1994) Failure of the mosquito plant Pelargonium X
citrosuin to repel adult Aedes albopictus and Culex quinquejasciatus in Florida. 3.
Amer. Mosquito Cont. Assoc. 10(4), 473-476.
Demarne, F.E. (1989) L’amelioration varietal du ‘Geranium rosat’ (Pelergoniumsp.)
- Contribution systemic, caryological et biochemique. DSc Thesis, Univ. de
Paris-sud, Centre D’Orsay, Paris, France.
Demarne, F.E. (1992) Histoire de la culture, botanique, et systematique. In: Le
Geranium Rosat a la Reunion. CIRAD, Saint Denis, Reunion.
Geraniaceae 55
Demarne, F.E. and Van der Walt, J.J.A. (1989) Origin of rose-scented pelargonium
cultivar grown on Reunion Island. S. AJ:Tyskr. Plantk. 55(2), 184-191.
Demarne, F.E. and Van der Walt, J.J.A. (1990) P. tonzentosum:a potential source of
peppermint-scented oil. S. AJ:J. Plant G. Soil 7(1), 36-39.
Demarne, F.E. et al. (1993) A study of the variation in the essential oil and
morphology of P. capitatum. J. Ess. Oil Res. 5(5), 493-499.
Dhakshinamoorthy, M. et al. (1979) Nutrient uptake studies on geranium. Ind.
Pefum. 23(3/4), 220-221.
Doraswamy, K. and Sundaram, M. (1982) Geranium cultivation in South India. In:
Cult. Util. Aromatic Plants. R.R.L., Jammu Tawi, India, pp. 573-577.
Eid, M.N. and Rofeel, 1,s. (1980) Effect of B-9 and CCC on the growth and
essential oil of geranium. Fac. Agric. Res. Bull. No. 1238; see also Res. Bull. N o .
1239. Ain Shams University, Cairo.
Fakirova, V.I. (1995) Gnomopnia geranii-macrorrhizi sp.nov. from Bulgaria. Myco-
taxon 54(2), 329-334.
Fedotova, Z.A. (1993) New geranium gall midge species (Dipteru, Cecidomyiidae)
from Kazakhstan. Vestnik Zoologii 5 , 19-25.
Fleisher, A. and Fleisher, Z. (1985) Yield and quality of essential oil from P.
graveolens cultivated in Israel. J. Sci. Food Agric. 36, 1047-1 150.
Fraisse, D. et al. (1983) SPECMA bank applications to the study of geranium
essential oils of various origins. In: Proc. 9th. Int. Ess. Oil. Cong. Singapore.
Fritz, J. (1976) Effect of nitrogen fertilization on rose geranium production. Agro.
Trop. 31(4), 369-374.
Gailleton, J.M. (1967) L’evolution de la production des huiles essentielles a la
Reunion. In: Proc. Ess. Oil Prod. in Develop. Countries. Trop. Prod. Inst.,
London, UK, pp. 4 1-6 1.
Guenther, E. (1950) The Essential Oils. Vol. 4. D. van Nostrand, New York, USA,
pp. 672-673.
Guenther, E. (1957) The essential oil industry of Kenya. The Guenther Report. pp. 4.
Fritzche Bros. Inc., New York, USA.
Guerere, M. et al. (1985) Etude d’huiles essentiells de geranium Bourbon obtenues
de trois facons differentes. Ann. Fuls. Exp. Chim. 78(836), 131-136.
Gulati, B.C. et al. (1982) Cultivation of P. graveolens as an annual crop. In Cult. Util.
Aromatic Plts. pp. 559-565. CSIR, Jammu-Tawi, India.
Husain, A. et al. (1 988) Major Essential Oil-bearing Plants ofIndia. Cent. Inst. Medic.
& Aromatic Plts, Lucknow, India, pp. 112-1 18.
Ivanovov, V.I. (1984) Development and production of machines for raising essential
oil geranium transplants on a commercial basis. Subtropich Kult. 2, 53-56.
Ivanovska, N. et al. (1993) Anticomplementary activity of polyphenols from Gera-
nium sanguineum and Geranium macrorrhizum root extracts. Compt. Rend. Acad.
Bulgare des Sci. 46(8), 131-133.
Kim, N.P. and Aprasidi, G.S. (1989) Osmotic indices and heat resistance in
geraniums under cultivation. Uzbek. Biol. Zhurn. 3, 24-28.
Koutchouloria, T.L. (1964) Selection of geranium. In: Proc. 3rd. Inter. Cong. Essen.
Oils pp. 186-190. Plovdiv, Bulgaria.
Krczal, G . et al. (1995) Transmission of Pelargonium flower break virus (PFBV) in
irrigation systems and by thrips. Plt. Disease 79(2), 163-166.
56 Chapter 3
Kumar, A., Sharma A. and Virmani, O.P. (1985) Cultivation and utilization of rose
geranium - a review. Cropmap (India) 7(3), 137-147.
Kumar, N. and Sampath, V. (1982) Study on the effect of rejuvenation pruning of
geranium. In: Proc. Nut. Semin. Med. Arom. Plts. Tamil Nadu Agric. Univ.,
Coimbatore, India, pp. 123-125.
Kustova, S. (1978) Handbook of Essential Oils. Pishevaja Promyshlennos, Moscow,
CIS, (in Russian).
Lawrence, B.M. (1994) Progress in essential oils. Perj'ium. Flav. 19(1), 40-42.
Lawrence, B.M. (1992) Progress in essential oils. Perj'ium. Flav. 17(2), 46-49 and (6)
59-60.
Leela N.K. et al. (1992) Nematicidal activity of essential oil of Pelargoniunz graveolens
against the root-knot nematode Meloidogyne incognita. Nematol. medit. 20,
57-58.
iMa, J.B. and Li, J. (1991) The chemical constituents of P. roseum oil. Acta Bot.
Yunnanica 13(1), 89-92.
Mairapetyan, S.Kh. and Aleksanyan, D. (1985) State of water and productivity of
rose geranium in outdoor hydroponics. Biolog. Zhur. Armenii 38(5), 423-427.
Mallavarapu, G.R. et al.. (1993) Chemical and agronomical investigations of a new
chemotype of geranium. J. Ess. Oil.Res. 5(4), 433-438.
iMani, A.K. et al. (1982) Effect of seasonal variation on the oil content and quality of
geranium oil. In: Proc. Nut. Semin. on Medic. Arom. Plts. Tamil Nadu Agric.
Univ., Coimbatore, India, pp. 119-120.
Matheson, J.K. and Bovill, E.W. (1950) In East African Agriculture OUP, London,
UK, pp. 999-100.
Mihajlov, M. and Tucakov, J. (1975) G. macrorrhizum essential oil from Yugoslavia.
Riv. Ital. EPPOAS 57(1 l), 635-638.
Nigam, M.C., Duhan, S.P.S. and Naqmi, A.A. (1987) G C examination of oil of
P.gruveolens raised in N. India. PufaiJ. (OctlDec), 11-13.
Nofal, M.A. (1981) New data on the composition of Egyptian geranium oil. Fac.
Agric. Res. Bull No. 1601. Ain Shams Univ., Cairo, Egypt, 14 pp.
Ntezurubanza, L. et al. (1992) Huiles essentielles de Geranium et de Lemongrass
cultivees au Rwanda. Riv. Ital. EPPOAS (Numero Speciale), 63 1-636.
Nurnazarov, M.N. and Berdyeva, N.S. (1986) Development of geranium cultivation
in Tajikstan. Mastrich Kul'tuy 5, 36-37.
Oboladze, L.U. and Kuchuloriya, L.T. (1986) The geranium variety Yubileinaya.
Muslinchnye Kul. 2, 34.
Oda, H.E-S. (1982) Studies on the production, storage and detection of adulteration
of some essential oils. PhD thesis. Fac. Agric., Cairo University, Egypt.
Ognyanov, I. (1985) Bulgarian zdravetz oil. Perj'iim & Flav. 10(6), 38-44.
Ognyanov, I. and Ivanov, D. (1958) Bulgarian zdravetz oil. Pel3r. Ess. OilRec. 49(10),
617-623.
Pellegrineschi, A. et al. (1994) Improvement of ornamental characters and fragrance
production in lemon-scented geranium through genetic transformation by
Agrobacterium rhizogenes. BiolTechnology 12(l), 64-68.
Prakasa Rao, E.V.S., Ganesha Rao, R.S. and Ramesh, S. (1995) Seasonal variation
in oil content and its composition in two chemotypes of scented geranium
(Pelargoniumspp.). J. Ess. Oil Res. 7(2), 159-163.
Geraniaceae 57
Prakasa Rao, E.V.S. et al. (1984) Intercropping studies in geranium. J. Agric. Sci.
Camb. 20,499-500.
Putievsky, E. et al. (1984) Harvesting and essential oil in P.graveolens. Hassadeh
64(8), 1552-1555.
Rajeswara Rao, B.R. and Bhattacharya, A.K. (1992) History and botanical nomen-
clature of rose-scented geranium cultivars grown in India. Znd. Perj'iurn. 36(2),
155-161.
Rajeswara Rao, B.R., Kailash Singh, and Bhattacharya, A.K. (1990a) Effect of
nitrogen and method of harvesting on herbage and essential oil yield of P.graveo-
lens. Znd. J . Agron. 35(3), 312-313.
Rajeswara Rao, B.R. et al. (1990b) Variation in yield and quality of P.graveolens
under varied climatic and fertility conditions. J.Ess. Oil Res. 23(3), 99-105.
Rajeswara Rao, B.R. et al. ( 1 9 9 0 ~ Effect
) of prilled urea and modified urea on yield
and quality of P.graveolens. Fert. Res. 23(2), 81-85.
Rajeswara Rao, B.R. et al. (1992) The essential oil profiles of rose scented geranium
(Pelargonium spp.) biomass dried prior to distillation. Znd. Perjium. 36(4),
238-240.
Rajeswara Rao, B.R. et al.. (1993) Changes in profiles of essential oils of rose-scented
geranium (Pelargonium spp.) during leaf ontogeny. J. Ess. Oil Res. 5(3),
301-304.
Satyakala, G., Muralidhar Rao, M. and Lakshmi Sita, G . (1995) In vitro micro-
propagation of scented geranium (Pelargoniumgraveolens: syn P. roseum). CUT.
Sci. 68(7), 762-765.
Sens, L. (1994) Report - Bilan Positzonnement des Planteurs de Geranium. APR-
Reunion, Saint Denis, Reunion.
Sholto-Douglas, J. (1969) Introducing essential oil crops. World Crops 21 (3),
122- 124.
Southwell, I.A. and Stiff, LA. (1995) Chemical composition of an Australian
geranium oil. J.Ess. Oil Res. 7(5), 545-547.
Stepanyan, Yu.S. and Ivanov, V.V. (1977) Geranium cultivation in Armenia. In:
Proc. 7th Znt. Cong. Ess. Oils. Kyoto, Japan, pp. 153-154.
Vadivel, B. and Sampath, V. (1982) Preliminary studies on different levels of cutting
geranium. In: Proc. Nut. Semin. Medic. Arom. Plts. Tamil Nadu Agric. Univ.,
Coimbatore, India, pp. 121-122.
Van der Walt, J.J.A. (1985) A taxonomic revision of the type section of Pelargonium
L'Herit. Bothalia 15, 345-385.
Van der Malt, J.J.A. and Ward-Hilhorst, E. (1977) Pelargoniums of South Africa, 3
vols. Pub. 1977, 1981, 1988. Nat. Bot. Gdns, Kirstenbosch, South Africa.
Vera, R.R., Chane-Ming, J. and Carbonel, M. (1995) Evolution des alambics utilises
a la Reunion pour la distillation du Pelargonium spp. Riv. Ztal. EPPOS (special
edn), 725-734.
Vernin, G. et al. (1983) Etude des huiles essentielles par CG-SM-Bongue SPECMA:
essences de geranium. Parjium. Cosmet. Arom. 52, 5 1-6 1.
Webb, W.J. (1 984) The Pelargonium Family. Croom Helm, London, UK.
Weiss, E.A. (1955-1960) In: Reports to the Board. Tanganyika Agric.Corp., Dar es
Salaam, Tanzania.
Weiss, E.A. (1961-1966) In: Reports to the Board. E.A.T.E.Corp., Eldoret, Kenya.
58 Chapter 3
Weiss, E.A. (1967) Effect of nitrogen, phosphate and time of cutting on green
material and oil yields from geranium in West Kenya. Expl. Agric. 3(2),
99-103.
Weiss, E.A. (1969) Phosphatic fertilizer usage on west Kenya crops. Agric. Digest. 18,
29-38.
Yan, Dongwei, Zhang, Zhengju and Ouyang, Ning (1994) Analysis of the chemical
compositions of essential oils from scented leaves of Pelargonium hybrids accli-
matised in Yunnan province. Fruits 49(1), 22.
Yue, D., Gosselin, A. and Desjardins, Y. (1993) Effects of forced ventilation at
different relative humidities on growth, photosynthesis and transpiration of
geranium plantlets in vitro. C a m d . J Plant Sci. 73(1), 249-259.
Zhao, Q. (1993) China’s perfumery industry picks up. Pe$ Flav. 18, 47-48.
Zukov, D.G. (1964) Essential oil production in the USSR. In: Proc. 3rd. Inter. Cong.
Essen. Oils. Plovdiv, Bulgaria, pp. 127-132.
Gramineae
4,
The Gramineae (syn. Poaceae), order Graminales, is the fourth largest
family of flowering plants with some 700 genera containing approximately
8000 species. Many are highly specialized but all produce a characteristic
fruit, an achene with a transparent pericarp, the caryopsis. A number are
cultivated on a huge scale including wheat, rice and maize, others produce a
commercially important essential oil. The Gramineae are divided into two
major sub-families, Pooideae and Panicoideae; the latter consisting of three
tribes, Paniceae, Maydeae and Andropogoneae, the last containing the
genera Cymbopogon and Vetiveria (Bor, 1960; Jagadischandra, 1975; Lazar-
ides, 1980). Classification of genera in the Panicoideae remains undefined
and also the number of species included in any specific genus, especially
Cymbopogon, as species intercross readily and may also exist in a number of
forms.
The inflorescence of Cymbopogon differs from other members of the tribe,
as the paired spikelets are carried on short racemes occurring in groups of
two, each group subtended by a moderately large semi-sheathing spathe.
The spathes with their paired racemes are borne in large branched panicles,
in some species somewhat open and spreading, in others more compact
(Fig. 4.1). The basic chromosome number is x = 10, and the somatic
number for most is 2n = 20 or 40.
The genus Cymbopogon Spreng. includes 50-60 species of tropical, per-
ennial, coarse-growing tufted grasses, many having highly aromatic foliage.
The five species described are C. nardus (Ceylon citronella) and C. winter-
ianus (Java citronella); C. citratus (West Indian lemongrass) and C. Pexuosus
(East Indian lemongrass); and C. martinii (palmarosa and gingergrass). The
genus Vetiveria Bory. currently contains ten species, including V. zizanioides
(vetiver) whose rhizome and roots are highly aromatic. Many other Cymbo-
pogon species contain a volatile oil (Singh, et al., 1982; Rizk et al., 1983;
Manzur-I-Khuda et al., 1984; Nigam, 1987; Mathela and Pant, 1988;
Mathela, 1990; Boelens, 1994).
59
60 Chapter 4
The very wide range of oil types identified also indicates the range of
material available to plant breeders, not only in respect of novel oils, but also
within existing commercially exploited Cymbopogon species. In particular,
there exists considerable potential for selection and breeding strains high in
a specific oil component. A strain of C. pendulus produces oil with a 90%
citral content; of c. wintem'anus a citronellal content of 70%; of c.jwuruncusa
piperitone to 83% (Singh and Pathak, 1994), while a strain of C.martiniivar.
sofia produces oil with 40% carvone (Sobti, 1982).
Most operations necessary to grow these grasses and obtain oil, and the
pests and diseases, are common to all and are discussed in this section.
Vetiver is discussed separately since oil is contained in rhizome and roots,
not in herbage. Smallholders contribute a substantial proportion of oils
entering international trade, but output per grower is small. Since most oil
production will probably remain in this sector, there has been considerable
official research into improving management and agronomic techniques,
especially in Sri Lanka, India, Indonesia and Malaysia.
Primary cultivation initially included clearing forest or bush, but this is
now discouraged by most governments. The greatest effort currently re-
quired is clearing land of exhausted crops and preparation for replanting.
Grass clumps are dug, piled, burnt and ash scattered. Some clumps are left
on field margins to provide splits for replanting. Experienced growers note
clumps which have outgrown or outyielded the majority, and use these for
planting material, but in general little attention is paid to selection of a
superior type. Following burning, the land is weeded and may be left fallow
until onset of the rains, then again weeded. When soil moisture is adequate,
Gramineae 61
the remaining clumps are divided and planted by opening a hole with a hoe,
the division dropped in and earth firmed with the foot. A man opening the
holes, his wife planting and children bringing further supplies of splits is
extremely effective, fast, and the proportion of rooted cuttings high if
subsequent rainfall is adequate. Little attention is paid to regular planting
intervals and, as grass clumps tend to expand to fill space available, there is
no appreciable loss of oil.
New plantings are weeded when time is available and, once established,
the grasses are well able to compete with all but the most pernicious local
weeds. The area is usually weeded after each cut and any fertilizer, ashes or
spent grass spread between clumps. Herbage is harvested by sickle and
transported to the still, or left to wilt for 1-3 days depending on local custom.
Time of cutting is important in relation to oil content of leaves but, as most
smallholders are always short of money, cutting is determined more often by
a cash need. Plants are harvested a t any convenient time, and this affects
both rate of regrowth and total yield of oil, usually to grower’s loss.
Few smallholders own stills and either sell their grass by weight to still
operators or cooperate in owning and operating a still. The latter system is
now common and actively encouraged by governments, who make low-cost
loans available to groups formed for the purpose. In India, wild plants are the
source of most palmarosa and gingergrass oil, and are cut by contract gangs
who harvest and distil1 the grass in situ in the area of their grant.
With the possible exception of vetiver, the grasses can be interplanted or
underplanted in suitable crops, which must be chosen with care, since too
much shade adversely affects leaf production and thus oil yield. Neither
method is recommended for commercial plantations unless the project is of
sufficient size to ensure the required level of management is applied to all
crops: tea and citronella for example (q.v). In smallholder plantings, the
degree of intercropping will depend almost entirely on the relative profitabil-
ity of the crops involved, and thus the respective proportion of land or time
allocated to each.
Mechanized production is basically similar to that required for small-
holder crops but land clearing will not be described. For a detailed descrip-
tion of some land-clearing methods used by the author and their cost see
Weiss (197 1). Ploughing is necessary at establishment or replanting, and the
main requirement is to leave the area clear of all major vegetation. A second
ploughing or heavy disc-harrowing at right-angles may be desirable to level
lands. Phosphatic or potassic fertilizers can be broadcast prior to this
operation, which ensures wide distribution in the root zone, particularly
important with vetiver since oil is contained in the rhizome and roots.
Leaf oil content can be affected by soil fertility, since soils which produce
luxuriant or very rapid leaf growth generally do so at the expense of oil
content. Thus, in terms of material processed by stills per kilogram of oil
obtained, soils of moderate to low fertility in which the level of plant
62 Chapter 4
containing cells easily ruptured; elevators and discharge chutes should allow
ease of movement without clogging. Forward and knife speeds on mower-
type harvesters must be so adjusted that grass is cut cleanly, without pulling
or snatching. When using forage harvesters, forward speed must be regu-
lated for similar reasons and, as the action of these harvesters varies between
models and makes, a trial run in a plantation is essential before any decision
to purchase. Harvesting on large-scale plantations should aim at cutting,
chopping and transporting grass in one continuous operation with the
minimum delay in delivery to stills. Machinery now available allows con-
siderable choice and a system suitable to local conditions easily selected (cf.
eucalyptus harvesting). At the Kimberley Experimental Station, Western
Australia, the system is based on a forage harvester which cuts, chops and
blows the grass into a trailer with high mesh sides U. Bonnardeaux, Cunnu-
nurra, personal communication).
Small farmers normally wilt herbage before distilling, and the method is
important. If herbage is left in the field, soil contamination can occur which
may taint oil or alter its composition. When left in heaps on field margins for
only a few hours, herbage ferments giving the oil a grassy note. When spread
in shade on a clean dry surface, most grasses can lose up to 50% moisture
with little loss of oil, whose value is recouped in the shorter distilling time
(Singh et al., 1994).
Hydro- and steam-distillation are most common; herbage is supported on
a grid above water which is heated to produce steam; or steam produced in
a boiler is introduced into the bottom of the still and normally produces
higher yields. Water remaining in the still following distillation should
preferably be discarded as the small amount of contained oil is seldom worth
recovering. Sri Lankan stills are traditionally made of galvanized iron,
135 cm in diameter and about 250 cm in height, tapering to a cone at the
top, with a capacity of 500 kg, and insulated to prevent heat loss by being
embedded in a stone hearth or surrounded by a layer of clay or finely
chopped grass. Openings a t the top and side facilitate charging and discharg-
ing. Such stills are called goddakans in Indonesia, and tilting types in-
troduced to facilitate discharging. Improved stills holding 500-2000 kg are
operating in most oil-producing countries.
Sri Lankan stills are located where there is sufficient water; two stills being
connected to the same condenser and operated alternately. In villages the
unit of measurement of citronella grass is the wadia, about 500 kg, approx-
imately the charge of a standard still. Older stills in India are made from mild
steel; two stills 1.8 m high X 0.9 m in diameter and one boiler which
produces about 100 kg steam per hour are recommended for 8 ha of grass
plantations (see Fig. 4.2). The key factor in operating these stills is manage-
ment; the better it is the more efficient the still and the higher the quality of
oil produced. More modern stills made of stainless steel are now operating in
most countries, and there is little difficulty in producing high-quality oil. T o
66 Chapter 4
Citronella
The citronella oil of international trade is obtained almost exclusively from
Cymbopogon nardus (L.) and Cymbopogon winterianus Jowitt. Commercial
cultivation of C. nardus is limited to Sri Lanka, but C. winterianus is
commercially cultivated in Indonesia, Guatemala, Honduras, Haiti, China,
and more recently, India and Vietnam.
Commerce generally defines the oils as Ceylon and Java types respectively
as there are characteristic differences, the most important being geraniol
content. Ceylon citronella oil generally contains 55560% geraniol and
7-15% aldehydes, Java citronella oil a minimum 35% of both geraniol and
aldehydes. Other Cymbopogon species which produce oils described as
citronella are of no commercial importance. An interspecific hybrid C.
confertiJEorus X C. jwarancusa, jamrosa, yields an oil containing geraniol
68 Chapter 4
75-80% and geranyl acetate 20-25% with a quite different odour to its
parents, described as being more terpenic and grassy (Krishnamoorthy,
1989).
Citronella oil has long been known to peoples of southern India, Sri Lanka
and Java who distilled it from the local Cymbopogon species, but its early
history is unrecorded. Poultices of leaves were also widely used in India and
Sri Lanka to treat minor cuts, abrasions and swellings, while extracts of the
leaves were used to alleviate internal disorders and as a vermifuge. Oil was
first sent from Sri Lanka to the UK by Surgeon-Major Grimm in the late
1600s, who identified the grass from which it was distilled as Arundo
odorutus. It was almost a century later that small amounts of oil as Oleum sirce
were sent to Europe, but by the end of the nineteenth century demand
exceeded Sri Lanka’s capacity to supply, and Java became an increasingly
important producer.
The first European to see Java citronella oil being distilled is believed to be
a Myn. A.J. Kaffer, at Cirebon on the north coast. He later began commer-
cial oil production on his own estate and by 1900, together with other Dutch
growers, erected modern stills. Java oil was considered to be superior to Sri
Lankan by merchants in Europe and by 19 19 Java was producing more oil
than Sri Lanka. However this rapid expansion resulted in Sri Lankan
cuttings being imported for use on lands unsuitable for Java type, and the
inferior oil produced used to adulterate Java oil. Of greater consequence was
mixing of the two grasses in plantations, which then also produced an
inferior oil. These two developments, plus the many smaller producers who
paid little attention to quality of herbage or distilling procedures, quickly
downgraded Java citronella oil on world markets. Only by introducing strict
official controls on production and marketing between the two world wars
did Java oil regain its reputation for quality and dependability. Exports of
citronella oil from Indonesia in the years 1989-1 995 ranged from 560 to 650
tonnes, roughly half total production.
Citronella was introduced to Taiwan (then Formosa) from Java by a Mr
Kiyoshi Iwamoto in 1912, who successfully developed commercial oil
production. By 1940 some 360 t was produced, almost all exported to Japan.
Production in the immediate post-war years was erratic with many citronella
plantations being replaced by officially encouraged food crops. However,
growing demand rapidly increased production to a record 3000 t in 1965,
but it subsequently declined and by 1984 the industry had virtually ceased to
exist except for a small amount of oil used domestically. Although citronella
leaves were incorporated in traditional Chinese medicines and plants grown
for this purpose, domestic demand for citronella oil was satisfied by imports.
Commercial oil production began in the early 1950% and Hainan Island
became a major producer with oil production in 1957 of almost 1000 t.
Citronella was later partially replaced by more profitable essential oil crops,
especially patchouli, and oil production declined (Chiang, 1958; Zhao,
Gramineae 69
Botany
Cymbopogon nardus (L.) Rendle (syn. Andropogon nardus Ceylon, de Jong)
lenabatu or lenabatu pengiri in Sri Lanka, and C. winterianus Jowitt (syn.
Andropogon nardus Java, de Jong) rnaha pengiri in Java, serai generally in
Indonesia, miyah serai for the oil. C. winterianus is said to be named for Mr
Winter, a pioneer of the Sri Lankan essential oil industry, and known locally
as Winter’s grass. Cymbopogonnardus var. confertiflorus (Steud.) Stapf is now
considered a separate species, C. confertiflorus, and produces an inferior
oil.
Cymbopogon nardus and C. winterianus are tufted, perennial grasses (Fig.
4.3), with long narrow leaves, and numerous stems arising from short,
rhizomatous roots (Fig. 4.4). Although C. nardus is more vigorous than C.
winterianus, the latter produces larger clumps when cultivated, to 0.5 m in
diameter. Cyrnbopogon nardus exists in a variety of forms, some wild, others
known only under cultivation, together with local cultivars in countries to
which it was introduced. By contrast, C. winterianus has a much narrower
range of variation.
Cymbopogonnardus has a more extensive and penetrating root system than
C. winterianus, which enables it to withstand periods of drought seriously
affecting growth of the latter. This root system also allows the species to
grow well on a wider range of soil types than C. winterianus, and it can be
grown profitably on soils of lower fertility. The ability of C. nardus to persist
for 10-15 years under cultivation probably owes a great deal to these wide-
spreading roots; C.winterianus seldom exceeds 6 years. The multiple stems
70 Chapter 4
Fig. 4.6. Cross-section of Java citronella leaf. A, Cross-section of leaf; B. upper epidermis; C, bundle
sheath; D, oil gland; E, cuticle; F, vascular bundle; G, lower epidermis. (Adapted from Bommegowda et
al., 1979.)
Gramineae 73
1975). The oil content of leaves differs significantly and young leaves
synthesize and accumulate most essential oil. During leaf expansion, cit-
ronellal, geraniol and citronellol content increase, but decrease with matu-
rity. Analyses of leaves from C. winterianus in India showed significant
changes in the ratio of some major oil components over leaf life (Luthra et
al., 1991).
The oil content of C. nardus in Sri Lanka averages 0.5%, of C. winterianus
grown in Indonesia and Central America 0.25-1.2%. Leaf-oil content varies
with age, time of cutting and soil fertility, but 1.33% from C. winten'unus in
Honduras is well above that obtained from commercial plantings elsewhere.
Higher figures have been reported from specially selected plants or experi-
mental plantings (Bordoloi, 1982), which indicates leaf oil content can
probably be raised fairly easily. This is also true for C. nardus, as there exists
in Sri Lanka a multiplicity of strains (some natural hybrids) with wide
variation in leaf oil content and characteristics (Iruthayathas et al., 1977,
1978).
Panicles also have a small volatile oil content, generally very similar in
geraniol and citronellal content. Comparative analyses of flower and leaf oil
in Sri Lanka showed, in per cent (with range in brackets): flower oil, geraniol
22 (12-36) and citronellal 4.75 (1-10); leaf oil, geraniol 24 (2-45) and
citronellal 4 (0.5-8.0) (Iruthayathas et al., 1977). Flower oil is of little
importance, is small in proportion to total oil yield, and would not com-
pensate for the loss of leaf oil content should citronella plants be allowed to
flower before cutting.
The major difference between citronella oil from the two species, as
noted, is in the citronellal and geraniol content; C. nardus (Ceylon type)
usually contains 55565% total geraniol and 7-1 5% citronellal; C. winter-
ianus (Java type) contains at least 35% total geraniol and a minimum of 35%
citronellal. Standard characteristics are shown in Table 4.1. Wide variation
is noticeable in published data of regional oils, and Java citronella has been
reviewed in detail (Virmani and Datta, 197 1). The main characteristics of
commercial samples from selected origins are shown in Table 4.2.
Ecology
Cymbopogon nardus and C. winterianus are tropical grasses and commercially
important between the tropics of Cancer and Capricorn, but they occupy
substantially different ecological zones when wild. Although the range of C.
nardus has been considerably extended under cultivation, C. winterianus has
a wider geographical distribution within a more restricted ecological zone.
Cymbopogon nardus can be grown successfully from sea-level in Sri Lanka to
1500 m in East Africa; C. winterianus is generally grown below 500 m and on
Taiwan oil yield fell by nearly 50% above 700 m.; high-altitude strains (to
1200 m) of C. winterianus have been selected in India. At all altitudes a well
distributed rainfall and high humidity is most important to ensure high
74 Chapter 4
herbage yield. Average yields from a series of trials are shown in Table
4.3.
Both species require a rainfall of 2000-2500 m m to produce high oil yield;
where there is a pronounced wet and dry season C. winterianus will require
irrigation if it is to persist; C. nardus is less susceptible to dry periods, but oil
production in both is reduced under water stress. Warm mists or high
/- Essential
oil
4.0
E 3.9
, I ,
0
Citronellal
v)
12
Citronellol
0 I I I I I
40
'
- A Max. temp
Min. temp.
10
0' I I I I I
1 2 3 4 5
Harvest number
Fig. 4.7. Seasonal changes in oil content and main constituents of citronella in India. (Adapted from
Prakasa Rao eta/., 1985.)
Gramineae 77
period of water stress in plants. Severe wilting will then occur, with sub-
stantial loss of oil from desiccation.
and not only provides a fertilizer, but acts as a mulch and suppresses weed
growth.
When grown commercially, C. nardus and C. winterianus generally re-
spond similarly to chemical fertilizers a t rates normally applied. Variation in
a specific response is often due to local soils or climate rather than the
fertilizer, and large-scale growers should accurately determine the most
profitable type and rate of application. Since C. winterianus is the most
important, there is more data available than on C. nardus, and this is
reflected in the text.
Nitrogen is essential for high leaf yield, but there is often considerable
variation in the response to both amounts and types; in India C. winterianus
cultivars differed by 42% herbage and 6.6% oil yields per hectare at the same
N levels (Singh et al., 1980a), while different types of urea on the same
cultivar affected both herbage yield and N recovery (Singh and Singh,
1992). Nitrogen at planting, often in an NPK compound or as diammonium
phosphate, usually promotes increased and quicker growth and thus reduces
time to the first cut. The NPK fertilizer used at planting should have a low
ratio of nitrogen to phosphate and potassium, sufficient only for initial
growth. Where there is little choice and nitrogen content high, it is probably
cheaper to choose a phosphate-potassium mixture and apply the nitrogen as
top-dressing. The main effect of top-dressing nitrogen is to increase the
number of tillers and number and size of leaves, but the response in terms of
increased herbage or oil yield may fall in succeeding years, especially to high
levels of nitrogen (Rao and Munnu Singh, 199 1).
Data are lacking on comparisons between nitrogenous fertilizers and,
although urea is considered superior, sulphate of ammonia is commonly
applied because it is easily obtainable. Methods of top-dressing are unim-
portant, provided coverage is even and timing accurate. Nitrogen normally
has little effect on oil content of leaves of either species, occasionally slightly
decreasing leaf oil content but unimportant compared to the large increase
in total herbage. In India, yield of fresh herbage rose from 12 to 18.5 tiha as
nitrogen increased from 75 to 150 kg/ha. Herbage N content rose slightly
but was not significantly different (Bommegowda et al., 1981, 1983). In
general, 75-125 kg Niha applied annually in split applications after each cut
should be profitable. In fertilizer trials, N slightly increased geraniol and
citronellal content of leaf oil, but had no significant effect on oil character-
istics or constituents at levels normally used on commercial crops.
The larger root system of C. nardus enables it to obtain greater amounts of
available soil phosphate, a major reason why the plant can continue to
produce oil for many years with little or no fertilizer. Neither species can
continue to do so indefinitely, and to maintain production both nitrogen and
phosphate are necessary; one without the other is usually unprofitable.
Where rock phosphate is available it can be broadcast and ploughed-in prior
to planting splits. A small amount of superphosphate added to planting holes
Gramineae 79
or in bands alongside and below the lines of splits will assist establishment
and promote initial growth. In most areas of Sri Lanka and India the crop
will probably benefit from a further application every 3-4 years, placed as
deeply as possible between plant rows following a cut. Reports from several
countries on both species indicated that 15-30 kg Piha is normally sufficient
for 12-1 8 months, with a further application of 15 kg Piha in the second year
following the last cut.
Phosphate appears to have little effect on leaf oil content and no sig-
nificant effect on oil characteristics or constituents at levels normally ap-
plied. Comparisons between types of phosphatic fertilizers are few but, as
with many other crops grown in the tropics, single superphosphate with its
slightly higher sulphur content often gives higher herbage yield per unit of P
than other types. Although initial P availability in rock phosphates may be
low, residual availability can be similar to other phosphatic fertilizers. Since
this residual effect is very important in terms of subsequent applications and
total cost of P per cycle, any residual effect should be accurately determined.
Both types of citronella appear to have a relatively high potassium require-
ment and official recommendations are 40-60 kg Wha placed at planting.
Major increases in herbage yield have been obtained on soils lacking po-
tassium; in Sarawak herbage yield rose from 22 to 28 tiha and oil yield from
136 to 162 kglha when 75 kg Wha was applied to a yellow podzolic soil (Ng,
1972). T o obtain these increases adequate supplies of N and P must be
available, and an N-K interaction may occur on some mineral soils.
Potassium is best applied as part of a compound fertilizer at planting, and the
recommended range is usually sufficient. Potassium applied after 2 or 3
years to both species in those areas where a substantial initial application was
profitable would probably be economic. It may also be advantageous to
apply K in several equal amounts instead of a single seedbed application
(Singh et al., 1990). There are few data comparing different potassic
fertilizers, and it is normally most profitable to use the type easily available.
Potassium has little influence on leaf-oil content but, at high levels, can have
a depressing effect. Oil composition or characteristics are generally unaf-
fected, but some experimental data indicate potassium may depress geraniol
and citronellal content, especially where nitrogen and phosphate availability
is low.
The small amounts of minor elements normally required to alleviate a
local deficiency should be included in a compound mixture applied to the
seedbed or drilled alongside plant rows after a cut. Field reports of various
chlorotic conditions or foliar discolorations following application of com-
pound NPK fertilizers may have been due to induced minor element
imbalance. An apparent magnesium deficiency on irrigated soil near Bom-
bay which caused yellowing and crinkling of leaves was eliminated by a foliar
spray of urea plus magnesium sulphate (Malwatkar et al., 1984), and
80 Chapter 4
manganese was the most effective single micronutrient enhancing oil pro-
duction at Darjeeling (Nandi and Chatterjee, 1991). Iron chlorosis of C.
winterianus had little effect on leaf oil content, but did reduce herbage yield
in India (Chattopadhyay and Subrahmanyam, 1993).
Cultivation
Land preparation being completed as described in the introduction, the next
operation is planting. Plants can be raised from seed which is normally
viable, but this is lengthy and tedious. The availability of planting stocks of
known parentage eliminates the danger of using an admixture of C. nardus
and C. winterianus seed, collected where both occur in close proximity.
Propagation is usually by dividing clumps to produce splits, a year-old
clump will provide 50-60 splits, and 1 ha of clumps is normally sufficient for
7-8 ha. A spacing of 90 X 60 cm requires about 18,500 pph, and 60 X 60 cm
27,000 pph. Treating splits with a rooting preparation or a fungicide is
usually unnecessary since replacing those which die is simple and cheap.
Infilling normally begins 3 weeks after planting when the majority of splits
will have rooted and by 4 weeks plants should begin to tiller. It is essential to
maintain the optimum pph, as this directly affects herbage and oil yield; a t
Bangalore, India, increasing the population of C. winterianus from 20,000 to
50,000 pph raised oil yield by 167 kg/ha; for every additional 5000 plants
above 20,000, herbage yield increased by 1048 kg/ha and oil 6.7 kgiha, but
after 40,000 pph the increase was not significant. Such information will
select the local optimum pph relative to the cost of producing 1 kg of oil to
achieve most efficient production. High rather than low pph also tends to
suppress weed growth. On-the-square planting at the optimum population
would probably require less labour when citronella is grown by small
producers. A spacing of 60 X 60 cm is widely used in India, Sri Lanka, Java
and South America for C. winterianus, but C. nardus is more often spaced at
90 X 90 cm. In Guatemala, three or four splits up to 60 cm in length were
planted per hole, but this is most wasteful of planting material and labour,
and probably unnecessary.
Weed control in citronella is usually manual or mechanical, hand weeding
is generally the most effective, but any cultivation must be shallow, espe-
cially in young plants as they are easily uprooted. Herbicides have been used
in trials but none officially recommended for citronella; 2,4-D applied 2
weeks before planting, and atrazine, diuron, oxyfluorfen, glyphosate and
simazine as directed sprays between cuts.
Cymbopogon nardus is not usually intercropped since the plant grows to a
substantial size and suppresses adjacent crops, but can be underplanted in
rubber, coconut, arecanut and similar tall-tree plantations; C. winterianus is
less aggressive and has been successfully intercropped with young rubber
and cocoa, pepper or vanilla vines, and can also be underplanted in similar
situations as C. nardus. Both species may have reduced herbage yield when
Gramineae 81
Harvesting
The first harvest of citronella is usually 6-8 months after planting, ideally
just prior to flowering, then 3-4 times in each following 12-months depend-
ing on rate of regrowth. The profitable life of a commercial plantation is very
variable: for C. winterianus 5-8 years in India, somewhat longer in Sri Lanka,
5-1 0 years on Java and Taiwan, 4-6 years in Guatemala and Honduras; for
C. nardus 6-10 but up to 20 years. The most productive period in terms of oil
yield per hectare is the first 3-4 years of C. winterianus, 4-6 years for C.
nardus, but better management, adequate fertilization and efficient harvest-
ing could easily prolong the useful life of most plantations and raise the
average annual oil yield. Since replanting is expensive and in practice means
the loss of a year’s oil yield, every effort should be made to keep an existing
plantation in profitable production.
The number of cuts per year and the interval between is not constant; in
Indonesia, maximum oil yield from C. nardus was by cutting four times
annually although six cuts were sometimes possible. If a nitrogen top-
dressing was delayed beyond 45 days after cutting, leaf-oil content in the
next cut was lowered (Soenardi et al. , 1980). With C. winterianus the interval
was 3.5-6.5 months in northern India and directly related to commence-
ment and duration of monsoon rains. Crops planted in April-May in Assam
are cut after 3-4 months, then every 2 months until October, but normally
only twice in the period November-February (Datta, 1982). Most growers
prefer to use height of plants as a guide to cutting and, in Puerto &CO,
82 Chapter 4
highest oil yield per hectare was obtained from plants cut eleven times at 130
cm; in Guatemala, whenever plants reached 140 cm.
Herbage yield per hectare varies very widely indeed, depending directly on
size of plantation and standard of management; C. nardus from
10,00040,000 kgha annually and from C. winten'anus 10,000-30,000 kg/ha.
The oil yield per tonne of herbage varies with the season, normally highest in
the driest and lowest in the wettest period. Oil content of C. nardus in Sri
Lanka was 1.2-1.5 and 0.25-0.50% respectively; from C. wintem'anus in
India 0.5-0.6 and 0.3-0.35%; Java averages 1.2 and 0.5%, Honduras and
Guatemala 1.8 and 0.75% respectively. This seasonal variation is well
illustrated by the extensive data obtained in Taiwan (Ping-Hsien, 1956)
which is somewhat dated but little comparable has been published.
Distillation
Cut herbage is normally wilted before distilling; in Indonesia, 3-4 hours sun-
drying is usual, longer if grass is very wet following rain; in Sri Lanka and
India it may be 1-2 days. Periods up to one month have been successful in a
number of countries, although in the Philippines lowest moisture and
highest oil yield occurred after approximately 7 days (Coronel et al., 1984).
While withering to reduce water content is normally advantageous, changes
can occur in oil composition depending on method used and period.
Withering for short periods slightly increases total oxygenated monoter-
penes, but the effect disappears as the period lengthens. Rapid withering in
hot sun also slightly increased final monoterpene hydrocarbon content over
slow withering in shade (E. Iruthayathas, Colombo, personal communica-
tion).
Distilling citronella herbage presents few technical problems (Jansz et al.,
1981; Senanayake, 198 1). Steam pressure and duration influences the yield
of oil and quality; a pressure of 100-200 kPa is recommended in Sri Lanka,
although double this gives a higher oil yield lower in geraniol and citronellal.
An initial pressure of about 650 kPa is maintained in Honduras, reduced to
250 kPa when condensation begins and raised to about 1000 kPa when
condensate flows evenly.
The usual charge in Sri Lankan stills is 500-600 kg chopped grass, firmly
pressed into the still body. Normal distillation is 2 hours, double this with a
1000-1 500 kg charge. Oil recovery in the first hour is 55-65%, in the second
to fifth hours 25, 10-11, 6-8 and 4% respectively. Distillation is seldom
continued beyond 2.5 hours, and distilling to exhaustion produces an
inferior oil. First over are the low boiling-point terpenes, followed by
citronellol, citronellal, alcohols and esters, geraniol+sesquiterpenes, then
the sesquiterpene alcohols. The distillate is separated, and the oil treated to
remove water and impurities, filtered and packed. Prolonged bulk storage
increases the oil's viscosity, it becomes dark brown, specific gravity may
increase and solubility decrease, but the refractive index is less affected. A
Gramineae 83
Sri Lankan oil had a specific gravity (20°C) 0.913, refractive index (20°C)
1.4829, acid value 0.3, ester value 18.1, after acetylation 88.2, carbonyls
38.2%. Main constituents were, in per cent, 1-carvotanacetone (1 l), ger-
aniol (20), geranyl acetate (6), perillaldehyde (12) and phellandrial (8.5).
Analyses of spent dried citronella grass gave 84% organic matter contain-
ing 2.25% N, and the ash contained, in per cent, silicon dioxide (12.4),
oxides of iron and aluminium (l.O), P (0.12), K (0.01), others (1.4);
considerable variation occurs as other analyses show silicon dioxide to 4%
and K to 7 % . Cymbopogon winterianus spents may contain 95% organic and
volatile matter, the ash containing 2% N, 1.5% K and 0.05% P. Although
fresh grass is not readily eaten by cattle, dried spent grass is an acceptable
stockfeed; it can also be used for h a f t paper pulp.
Argentina Brazil Burma Ceylon" China Guatemala India Indonesia Java Nepal Paraquav Taiwan Vietnam
Limonene 3.3 2.8 8.4 9.0 2.6 3.8 4.0 3.7 3.9 3.6 3.6 3.3 3.1
cis-Ocimene ng ng < 0.1 2.2 ng ng ng 0.5 1.3 ng ng 0.3 0.2
Citronellal 34.8 46.8 12.2 5.2 34.4 34.8 30.8 35.7 36.8 35.5 37.8 35.0 32.9
Linalool 0.9 0.6 1.5 0.8 1.o 0.5 1.o 0.9 0.8 0.7 0.5 0.7 1.o
lsopulegol (4 1.9 0.7 0.9 0.6 1.4 0.5 0.8 1.3 2.1 0.7 1.4 1.4 1.5
isomer)
Alpha- ng 0.7 2.3 1.3 < 0.1 0.5 ng 0.2 0.1 ng ng 0.5 < 0.1
bergamotene
Caryophyllene 0.1 < 0.1 0.2 1.4 0.1 < 0.1 < 0.1 0.3 0.4 0.1 < 0.1 0.1 0.3
Citronellyl-acetate 2.5 1.7 5.1 0.9 2.3 1.9 7.3 2.8 2.5 3.4 0.9 2.1 2.5
Citral b (neral) 0.5 < 0.1 0.4 0.4 0.4 0.8 0.4 0.8 0.6 0.4 0.5 0.8 1.o
Citral a (geranial) 0.7 0.5 1.9 0.7 0.6 0.7 0.5 0.8 0.8 0.5 0.4 0.7 0.8
Citronellol 14.9 13.4 20.1 6.4 13.6 17.8 23.9 15.8 15.7 14.9 14.5 14.9 15.3
Nerol 0.3 0.2 <0.1 1.o 0.2 0.4 0.2 0.3 0.3 0.2 0.2 0.3 0.3
Geraniol 22.6 21.2 22.4 20.3 24.3 29.1 19.3 23.3 21.4 24.5 24.2 21.2 23.9
Elemol 3.9 1.4 3.9 1.6 5.0 1.2 2.0 1.8 2.1 3.4 3.3 3.5 3.3
Eugenol 0.7 0.9 0.5 < 0.1 0.8 0.8 0.9 0.9 0.6 0.5 1.o 0.8 0.7
aCymbopogonnardus; ng, not given.
Source: Abstracted from published sources.
86 Chapter 4
esters, citronellol and its esters. A major reason why citronella oil is still so
used is that many perfumers favour products of natural origin over syn-
thetics; especially hydroxycitronellal with its lily of the valley fragrance
strongly preferred in high-class perfumes. Geraniol and its derivative cit-
ronellol both have a rosy odour, are often called rose alcohols and are used
extensively in perfumery. Terpeneless citronella oil is produced to order, but
as the whole oil is normally used there is no point in removing terpenes
which are of little value. Wide regional variation occurs in Java oil character-
istics, and these have been summarized (Virmani and Datta, 1971); a
variation is also found in oils produced in different districts in a particular
country, and indicates the very wide range o f material available to plant
breeders. Characteristics from selected origins are shown in Table 4.1.
Variation is not restricted to composition, to perfumers there are substantial
differences in odour, thus Guatemalan oil is classed separately.
Lemongrass
Grasses under the general name lemongrass were described early, but the
first accurate genus classification was by Dr 0. Stapf (Kew Bull. 1906,
302-364), with subsequent research to establish species relationships. The
high citral content of the oil, which emits a strong lemon scent when leaves
are crushed, gives the species their common name. Lemongrass oil is
currently obtained mainly from Cymbopogon citratus (DC.) Stapf and C.
flexuosus (Steud.) Wats. which produce West Indian and East Indian lemon-
grass oil respectively, although other species produce oil with a high citral
content.
Cymbopogon citratus is known only under cultivation, and although occur-
ring in many tropical countries it is always an introduction, i.e. the wild
lemongrass of Zaire. Cymbopogon citratus is cultivated in Argentina, Brazil,
Guatemala, Honduras, Haiti and other Caribbean islands, Java, Vietnam,
Malaysia, Sri Lanka (where it is known as sereh), Madagascar, Comoros; to
a lesser extent in the Philippines, China, India, Bangladesh, Burma, Thai-
land (called takhraz) and Africa. Grown also on a small scale in many other
tropical countries, including north-western Australia which has a climate
similar to Guatemala, and southern Russia (Bagaturiya et al., 1990).
Cymbopogonflexuosus is cultivated in southern India, mainly Kerala State,
which Indian botanists claim as the plant’s original home. It is also known
locally as Cochin or Malabar grass, and the oil similarly. Two varieties of C.
flexuosus are recognized, the white-stemmed known as wella poolu now
designated C. flexuosus var. albescens, and the red-stemmed choomanna poolu.
The latter produces genuine East Indian oil and is the type commercially
cultivated (Husain et al., 1988; Fig. 4.8). Where C. flexuosus has been
Gramineae 87
Botany
Cymbopogon citratus (DC.) Stapf. (syn. Andropogon nardus var. ceriferus
Hack.; A. citratus DC.) and Cymbopogon flexuosus (Steud.) Wats. (syn.
Andropogon nardus var. flexuosus Hack.; A.flexuosus Nees), are tufted per-
ennial grasses, with numerous stiff stems arising from a short, rhizomatous
rootstock.
Roots of C. citratus, although extensive, are not deeply penetrating and
cultivated plants seldom survive more than 4-5 years; C. flexuosus usually
survives for at least 6 years although it has a very similar root system. A C.
citratus root oil obtained in Sri Lanka had these characteristics (temperature
unspecified): specific gravity 0.94, optical rotation +3"40' to -1"40', alde-
hyde content to 82% (Nethsinga and Paskaranathan, 1976a). The stems are
stiff to 3 m but under cultivation are normally cut before maturity. Stem
colour in C. citratus is a light to medium green; C. flexuosus has two types,
one with an apple-green stem which may be almost white, the other a red
stem and both can be found growing wild in close proximity. The number of
stems (tillers) significantly affects the amount of leaves produced, and any
increase in total herbage produced by C. citratus is related to extent of
tillering.
Leaves in both species are long, narrow, 1 m X 3 cm, upright when young
but trailing at maturity. Colour is usually a light to medium green, is affected
by environment, and is an unreliable guide to identification. A detailed
description of leaf structure has been published (Metcalf, 1960). Leaves can
Gramineae 89
Ecology
Both species are tropical grasses and flourish in sunny, warm, humid
conditions, basically between the tropics of Cancer and Capricorn. Many
local cultivars, however, are adapted to areas outside these limits and both
species are capable of widening their range under cultivation. Cymbopogon
90 Chapter 4
citratus is a wholly cultivated plant, and its range is that of countries to which
it has been introduced; on the Ord Irrigation Area in Western Australia for
example it has produced very high herbage and oil yields. Cymbopogon
jlexuosus requires a more sunny climate, with a well-distributed rainfall. In
South East Asia it grows naturally on roadsides, in mixed deciduous and
dipterocarp forests, often on the margin of secondary vegetation, on coastal
rocky areas and on grassy, often limestone, slopes and ridges. It occurs
naturally in India from sea-level to 1200 m, in Thailand and neighbouring
countries to 2200 m but commercial cultivation is most profitable between
500-1000 m; in Kerala, India, it grows well between 900-1250 m on hill
slopes. The effect of seasonal change on growth and herbage yield in north-
eastern India was relatively easy to correlate with specific elements of
climate, but it was not possible to do so with oil content or characteristics
(Handique et al., 1984). Cymbopogon citratus has similar climatic require-
ments to C. jlexuosus, but flourishes naturally below 500 m. In Guatemala
and Brazil, it is grown commercially between 120-600 m; in Indonesia, from
sea-level to 500 m; in Madagascar and Comoros, 200-750 m; in Haiti and
neighbouring islands, below 450 m; in Africa, generally below 500 m.
Interestingly, the citral content of oil from lemongrass grown in the Ethio-
pian highlands was a low 13% (Abegaz et al., 1983), compared to around
75% in oil from lemongrass grown at lower altitudes in East Africa (Ekun-
dayo, 1986).
It is not the effect of altitude on general plant growth, however, that can be
most important, although intensity of sunlight and rainfall at higher altitudes
adversely affect leaf-oil content. In both species oil content decreases during
Gramineae 91
long, dry, sunny periods, which occur more often at higher than lower levels
in many tropical countries. This depressing effect can be partially overcome
by selecting hillsides which receive less sun or a prevailing wet or misty wind,
and thus are suitable for lemongrass where other cash crops are scarce.
Conversely, drainage combined with judicious site selection can reduce the
effect of excessive rainfall.
The major effect of long periods of sunshine is to desiccate plants and
cause loss of leaf oil, but other factors are involved. In Sri Lanka, the
photosynthetic rate of leaves varies with light duration and intensity, and
there is a similar variation in carbon dioxide compensation values (Herath
and Ormrod, 1979). Data indicated this variation also applied to different
lemongrass strains and could be used for selecting those suited to a partic-
ular environment. Where lack of sunlight is the local limiting factor, rows of
plants should be orientated to obtain the maximum sunshine; generally
east-west.
Both species produce highest oil yield per tonne of herbage where the
rainfall averages 2500-3000 mm annually, preferably well distributed espe-
cially during the hot season; 5000 mm will be tolerated provided soils do not
become saturated. Cyrnbopogon citrutus is considered more drought tolerant,
but this refers to vegetative persistence; oil content suffers equally in both
species in long dry spells. C. flexuosus is considered more tolerant of short
periods of waterlogging, but this should not be a consideration when
selecting lands for planting. Areas which receive only heavy monsoon-type
rains or tropical storms of great intensity but short duration are generally
unsuitable for large-scale commercial plantings without special drainage.
Assessing available data (much of it unpublished) on both species it appears
the yield of oil per tonne of herbage from cuts during. a wet season (rainy or
monsoon), when growth is rapid and lush, is 45-65% of that obtained from
the same plants cut in the dry season or period (cf. geranium). A comparison
of yield and citral content of oil from uncultivated C. citrutus in the Phil-
ippines is shown in Table 4.7, which indicates the general effect of climate
on both, and it is interesting to note the effect of altitude in ‘hilly sections’.
It was also noted that, in the Cavite region, cultivated lemongrass gave a
herbage oil yield of 0.37 and 0.64% in the cool and hot seasons respectively,
much higher than that of uncultivated lemongrass.
A daytime temperature of 25530°C is considered the optimum for max-
imum oil production, with no extremely low night temperatures. Short
periods above 30°C have little general effect on plants, but severely reduce
oil content. Frost at any stage is normally fatal. Hail will shred leaves and if
a storm should coincide with a planned cut, oil yield will be low. Winds do
little damage unless they are hot and dry, or cyclonic; one evaporates the oil,
the other can destroy a plantation!
92 Chapter 4
Table 4.7. Yield of oil citral content from Cymbopogon citratus - Philippines
season, December to February: Hot season, March to June; Rainy season, July to November.
Source: Oliveros-Belardo and Aureus (1977).
cultivar; soil of p H 7.5 gave the highest herb and oil yields per hectare but
neither was significantly greater than p H 5 . 5 ; oil content of herbage, how-
ever, fell significantly between p H 5 . 5 and 7.5 although total citral content of
oilrosefrom83.5% a t p H 4 . 8 t o 8 7 . 1 % a t p H 7 . 5 (Singhetal., 1983). Saline
soils are usually considered unsuitable for lemongrass although this may be
due more to the cultural and agronomic problems associated with growing
crops on these soils, as laboratory and greenhouse trials indicate lemongrass
will grow and produce average herbage and oil yields on highly saline soils.
In pot trials, C. JEexuosusgrown in soils with an electrical conductivity (EC)
of 11.5, 10.0 and 5.5 mmhos/cm, showed no significant reduction in herb
and oil yield (Singh and Anwar, 1985); the citral content was unaffected by
increasing salinity to 15 mmhosicm.
Any crop from which 100 t of herbage per hectare is harvested annually
also removes quantities of plant nutrients, but there are few detailed reports
of the amounts removed either per hectare or per tonne of green material. In
Western Australia, the maximum amounts removed annually per hectare
were 186 kg N, 26 kg P and 384 kg K; with an average yield of 92 tlha
Cultivation
The basic operations to establish a lemongrass plantation are described in
the introduction to this chapter. In India, Sri Lanka and countries where
there are a large number of smallholders, it may be necessary to protect
newly sown plantations against domestic stock. Although mature grass is
generally unpalatable, young growth is grazed especially when other forage is
not available.
Propagation differs between C. citratus and C. flexuosus; the former is
vegetatively propagated, the latter grown from seed. T o obtain the required
amount of seed of C. flexuosus some plants in a plantation are left uncut to
flower, or a special area for seed production established. After flowering,
November-December in southern India, plants are regularly inspected to
determine when seed is mature but not overripe, or shattering will occur.
The whole inflorescence is cut, left to dry in the shade for 1-3 days, then
threshed by any method which does not damage seeds. Seed can be sown
directly where plants are to grow and, provided soil is moist and rainfall
expected, is usually successful. Approximately 35 kgiha is required. The
great disadvantage of direct sowing is the difficulty of effectively controlling
weeds in young seedlings, since most tropical grasses are fast and expansive
growers. Chemical control of such weeds is almost impossible, and manual
weeding equally difficult.
96 Chapter 4
Seed production is prolific and viability high so there is usually little need
to conserve it; thus nursery beds should be over- rather than under-sown. A
mature plant supplies about 100-200 g seed and 10-1 2 kg of clean seed is
sufficient to produce seedlings for 1 ha. Beds should be sown in advance of
the rains, so planting-out can commence as early as possible to ensure fast
growth after establishment. Seedlings are ready to transplant in 60-80 days,
topped to 15-20 cm, and dipped in fungicide. Planting of cuttings is usually
manual, but the small mechanical planters available can be used. Mechan-
ical planting may be more expensive initially, but is offset by faster seedling
growth and lower weeding cost. Single ridges or rows are usually 30-50 but
up to 100 cm apart, or ridges can be large enough to accommodate up to
three rows of plants, becoming raised beds. In-row spacing of 12-20 cm is
common, and up to three cuttings planted together since this is cheaper than
infilling. Multiple planting usually has little effect on herbage yield but, in Sri
Lanka, two cuttings per hill at 15 X 10 cm spacing gave the highest average
yield of oil per hectare. In Kerala, 2-3 cuttings per hill at 13 X 15 cm is
favoured.
Cymbopogon citrutus is generally propagated by root division (splits), and
the usual method is to divide clumps from an existing or exhausted planta-
tion following cutting. One fully mature healthy plant will yield 50-200
splits. Size of splits is a local preference but larger splits root quicker and
grow faster. Splits should be cut to leave 10-15 cm of shoot, trimmed of
excess and old roots and preferably be dipped in fungicide. Where large
numbers are involved, this can be carried out mechanically with a rotating
drum (the author once hired a concrete mixer), and splits should be dried
before planting. Use of growth stimulants or rooting compounds is usually
unnecessary, and exposing slips to gamma radiation prior to planting had a
generally adverse effect on plant growth and oil content (Banerji and Gupta,
1982).
Cymbopogon citratus is normally sown on the flat, about 10-15 cm deep,
but ridges can be raised after emergence to provide drainage or water
retention, whichever is locally most important. On clay loams or similar
rather heavy soils, too deep planting or raising a substantial ridge when
plants are still small is inadvisable, as C. citrutus is very susceptible to root
rots at this stage; depth of planting and ridge size are less important on more
sandy soils. A spacing of 50-75 cm rows and 50 cm between plants has
consistently produced the highest oil yield per hectare irrespective of coun-
try. Row width is usually determined by the mechanical equipment used and
between-plant spacing adjusted to provide the optimum population, which
it is essential to maintain by infilling since there is convincing evidence a high
population is far more profitable. The remarks on weed control in C.
flexuosus are applicable. In general, a weeding should follow cutting and
spot-spraying with a herbicide is then possible. Diseased or dead plants can
Gramineae 97
also be easily seen, removed and burnt. Smallholders irrigate when possible,
usually after each cut. Large estates in Central America were regularly
irrigated during the dry season to maintain growth as water was freely
available, while plantations in Azerbaijan, CIS, were irrigated 15 times
between April and October (Ragimov et al., 1985). The data available
indicates a water stress prior to cutting generally reduces herbage yield but
increases oil content, an effect similar to harvesting following a dry period,
and clearly demonstrated in Western Australia on the Ord Irrigation Area
(Fig. 4.10). Soil moisture status may also affect oil components, and
requires further investigation. Interestingly a 1917 report from Florida,
USA, stated: ‘a dry sandy soil yielded oil with a 75% citral content, a moist
sandy loam 68% and a wet loam 66%’.
Cymbopogon citrutus is seldom intercropped or underplanted when grown
to produce oil since it is basically unsuitable; C. jlexuosus may be inter-
planted, but is generally the major component when grown for oil. An
interesting method of integrating C. jlexuosus into plantations of other crops
was proposed for Bangladesh, but not widely implemented (Khan, 1979);
C. citratus has been underplanted in young rubber in Malaysia and elsewhere
to help defray cost of plantation establishment.
Harvesting
The first harvest of both species is usually 6-8 months after planting, then
cut three to six times annually over life of the plantation, although in Brazil
and other South American countries plants are cut at approximately 3-4
months, as this is locally believed to induce more extensive tillering and thus
bigger, more leafy plants. The cutting interval is governed by level of
management, climate and cultivar grown, thus the more skilled the local
grower, the greater the returns possible. Sampling to determine oil content
and composition is desirable, but seldom practised. Most small growers cut
a t a locally favoured height and, in general, this results in obtaining the local
average oil yield. Too frequent cutting of C. citrutus usually increases costs
with no great addition to total oil obtained over life of the plantation. By
contrast, frequent cutting of C. jlexuosus can profitably increase the total oil
yield. It is thus essential to determine the optimum height of cut and number
of cuts by field trials; however, adequate transport, distilling and storage
facilities must also be available to benefit from the results.
T o reduce the amount of green material distilled, smallholders tend to cut
high to include mainly leafy material. After several cuts it is then necessary to
prune the plants, which is time-consuming and unpleasant work so, in Sri
Lanka and southern India, plantations are often burnt at the end of the dry
season. This is effective in reducing unnecessary herbage and kills some
plants, but also substantially reduces carry-over of insect pests, especially
stemborers.
98 Chapter 4
Cutting can begin as soon as plants are dry after the night’s dew, as wet
grass left for later collection quickly ferments. Sunny days are preferable,
since cloudy and misty conditions tend to depress leaf-oil content.
Small farmers usually harvest manually, cutting plants 7.5-10 cm above
the ground and, in Sri Lanka, one worker can harvest about 2 t of fresh grass
daily. Mechanical cutting is usual on commercial plantations with a com-
bined cutter-chopper-loader. Cutting height varies with equipment used
and must be determined by field trials; in Western Australia 20 cm was the
optimum. Cutting at 20, 30, 40 and 50 cm gave annual herbage yields of
116,97,82 and 72 tiha, and the highest yield of 134 tiha was from 20 cm and
a cutting interval of 120 days (Beech, 1977). Manual cutting at different
12 - 60
-m h
9
f e
-
E
4-
10- 50
L
.-m
$ a-
n
c
v)
Lt:
6- 30
4 I I I I 20
(b)
60
50
-2
-
v
0
.-
x
40 .=
0
30
20
Fig. 4.10. Effect of withholding irrigation water from lemongrass (Cymbopogoncitratus ex Taiwan) in
Australia. (Beech, 1977.)
Gramineae 99
Distillation
In general, C. citratus is steam distilled and methods vary considerably,
reflected in the type and quality of oil produced, while an interaction
between duration of wilting and distillation time can occur. Wilting in
Indonesia for 2 days with 1.5 hours steam distillation gave highest oil yield of
0.43% (citral 80.3%); no wilting and 45 minutes yielded 0.3% oil (86.7%
citral) (Rusli et al., 1979). Increasing distilling time from 45 to 90 to 135
minutes increased oil yield from 0.311 to 0.351 to 0.406%, but citral
content fell steadily from 86.7 to 80.3%. In Puerto Rico, a 28% solution of
sodium chloride in still water increased C. citratus oil yield by almost one-
third; the use of sea or salt water in stills has been explored elsewhere but is
nowhere common (cf. Pimenta racemosa). Local stills processing C. flexuosus
in India generally use hydro-distillation and have changed little in the last
100 Chapter 4
100 years; a unit in Kerala, India, charged with 100 kg grass and 140 1 water
required 2-2.5 hours to distil including filling and emptying the still, and
used 40 kg wood to produce 0.35 kg oil. Since distillation is probably the
most important factor influencing oil quality, there has been considerable
official research on methods of improving local stills. The CISIR, Sri Lanka,
has developed the Manakoka Still, and provides plans and advice on its
construction. The Department of Technology, Bogor, Indonesia, has devel-
oped a superior still for use by small companies or cooperatives (Rusli,
1977).
Oils correctly stored (i.e. protected from air, heat and sunlight) remain
stable for long periods, with little or no loss of citral. T o reduce the adverse
effect of local storage on citral content of C. flexuosus oil in India, several
indigenous antioxidants were tested; an extract of betel leaves (Piper betle)
proved most successful, maintaining citral content at 83% compared to 72%
in oil not treated (Kurain et al., 1984). Temperature, light and length of
storage produced some interesting results in the Philippines; rate of evapora-
tion and changes in citral content of oils from fresh, air-dried or wilted leaves
stored at different temperatures showed oil extracted from fresh leaves to be
the most stable (Brandares et al., 1987).
Spent grass can be dried and used as fuel, composted or returned directly
to the plantation (or other crops) as mulch, and also fodder for domestic
stock, either dried or as silage. Analysis of fresh and spent grass and silage
made from C. citratus in Guatemala is shown in Tables 4.8 and 4.9. When
fed to sheep, the total digestible nutrients (dry material) were 44.6%
equivalent to 1963 kcalikg digestible energy, and it is a valuable stock feed in
areas lacking fresh forage. Analysed in India for use as paper pulp, the main
constituents of C. citratus dried spent grass were (per cent), ash 6.2,
pentosans 16.6, lignin 21.3, cellulose 47 (Guha et al., 1973); in Guatemala
dry material consisted of (per cent), cell walls 67, lignocellulose 42, hemi-
cellulose 25, lignin 7.5, cellulose 30, ash 4 (Ramirez et al., 1976).
Table 4.8. Amino acid composition of fresh lemongrass and lemongrass bagassea
Over 100 compounds have been detected in the oil (Boelens, 1994), and
average levels of some major constituents are, in per cent, (C. flexuosus
outside, C. citratus within brackets); citral to 85 (80), geraniol 5 ( 3 ) , methyl
heptenone 6 (0.3), myrcene 0.8 (20), dipentene 0 (4) nerol 0 ( 2 ) , not
Table 4.9. Chemical composition and organoleptic characteristics of experimental silages containing
lemongrass bagassea.
necessarily in the same sample. Major components are shown in Table 4.10
and official specifications in Table 4.1 1. Both oils have a pronounced fresh
lemony odour; that of West Indian is less strong because of its lower citral
content while East Indian is considered more fresh, lighter, and sweeter.
Whole oils are mainly used as a source of citral, with synthetics the major
competitors, and the price of a particular batch of oil is directly related to its
citral content, minimum 75%. Citral has numerous uses, and can be further
processed to isolate the ionone group which possess a violet-like fragrance
important in Ferfumery, and vitamins, especially vitamin A.
Demand for whole oils has expanded with their use as a masking fragrance
in deodorants, waxes, polishes, detergents and insecticides, where the oil’s
generally low cost is attractive. Small amounts are also used by stockfeed
manufacturers. Little is used to flavour human foods, since fresh grass or
stalks are preferred. T h e whole oil has fungicidal properties to plant and
human pathogens (Handique and Singh, 1990) and is potentially anti-
carcinogenic (Zheng et al., 1993). A diagram of lemongrass oil use and its
major competitors is shown in Fig.4.11. Terpeneless lemongrass oil is
Palmarosa oil, commonly East Indian geranium oil, and gingergrass oil are
obtained from two varieties of Cymbopogon martinii, known in India as rosha
or russa grass. Although morphologically almost identical, the two varieties
have different growth forms, favour differing environments and yield distinct
oils. In India, which produces most commercial supplies, palmarosa oil is
obtained from C. martinii var. martinii (motia) and the less valuable ginger-
grass oil from C. martiniivar. sojia (sojia). Since the names motia and sojia are
in common use and well understood, they will be used in the text.
Cymbopogon citratus
IS EOA
Specific gravity 0.872-0.897 0.869-0.894
(20°C) (25°C)
Refractive index (at 20°C) 1.4830-1.4890 1.4830-1.4890
Optical rotation -3"tO t l o -3'tO tl"
Carbonyl compounds (as citral) Min 75% Min 75%
Residue 7% maximum
Solubility (70% alcohol) Fresh oil soluble Cloudy solutions
at 20°C. Insoluble in 70,80,90 and
in 90% 95% alcohols
at 25°C.
Cymbopogon flexuosus
IS EOA
~~ ~~~~~ ~~ ~
Lemongrass oil
and perfumery
( p - ionone,
methyl-ionone,
Fig. 4.1 1. Pattern of use of lemongrass oil. (Tropical Products Institute, London.)
Indian palmarosa and gingergrass oils are obtained almost wholly from
wild plants growing in forested regions, palmarosa mainly in the states of
Madhya Pradesh, Maharashtra, Andhra Pradesh and Assam; gingergrass in
Tamil Nadu, Andhra Pradesh and Punjab. A number of regional strains
produce oils of varying characteristics but, as herbage is generally harvested
indiscriminately, the resultant bulk oil is a blend of all types in a particular
locality. Since palmarosa and gingergrass normally occupy different hab-
itats, many bulk oils are basically true to type; it is the quality which varies.
Commercial planting of palmarosa is now being officially encouraged in
India, but there are few large plantations. In the late 1930s palmarosa was
introduced by the Dutch to Java, large commercial plantations were estab-
lished, and Java oil with its generally higher ester content was considered
superior to Indian oil. Production of gingergrass oil is restricted to India, and
there are no commercial plantings. An oil almost identical to gingergrass oil
was obtained from an Angolan variety of C. densiJClorus Stapf. but was not
grown commercially (Naves, 1960).
It is not recorded when palmarosa or gingergrass oil was first distilled in
India, but bruised leaves were used to perfume bath water, and provide
poultices to relieve the pain of bruises, aching joints and swellings. The oils
were traded between India and the Persian Gulf when the Portuguese
entered the Indian Ocean in the late sixteenth century. This route remained
in use until early in the twentieth century, when oil was shipped from
Bombay to Red Sea ports then overland to Constantinople, and thus became
erroneously known to European traders as Turkish or Indian geranium oil.
It has remained popular in the Middle East as a base for cheap scents and
soaps, and is often a mixture of palmarosa and gingergrass oils or wholly the
latter. The separation of palmarosa and gingergrass into two separate
Gramineae 105
Botany
Cymbopogon martinii (Roxb.) Wats. var. martinii (syn C. maninii Stapf var.
motia), 2n = 40, produces palmarosa oil and C. martinii Stapf var. sofia(syn.
Andropogon martinii Roxb. var. sofia), 2n = 20, produces gingergrass oil.
However, terpenoid analyses of var. sofia strains from north India suggested
a closer relationship to C. densiJ7orus than to var. martinii (Mathela et al.,
1988). Although only two varieties of C. martinii are currently recognized,
the very wide natural occurrence of both on the Indian subcontinent has
given rise to a number of locally well-adapted strains, which became named
cultivars. The close relationship between the two varieties and their prox-
imity in the wild have almost certainly resulted in widespread hybridization.
Seed should be thus collected only from areas where there are pure stands or
from plants of known parentage. Little information is available on the life
span of the two varieties in the wild, but when cultivated in India motia is said
to last 10-15 years (Singh, 1977) although many plantations become
unprofitable after 6 years (Husain et al., 1988).
The root system of sofia is more extensive than motia; the latter normally
occurs in more dry, open locations at higher levels than sofia, and its smaller
but penetrating root system reflects this environment. Sofia generally occurs
in more fertile soils at lower elevations, where an extensive root system is
advantageous. Habitat zones are not strictly delineated, and both varieties
may grow side by side at the respective upper and lower limits of their range,
or interspaced where soil or climate favours one variety. In some state forests
of Madhya Pradesh and Mysore for example, the two varieties are so
intermingled that collection of one only is virtually impossible. Both varieties
occur naturally in rounded clumps or tussocks, but while motia clumps tend
to remain separated, sojia clumps grow much closer together and can
become almost pure stands covering several hectares. Both grasses are
vigorous growers and quickly colonize suitable land. Indian Forest Depart-
ment contracts specify that grass must be cut and physically moved from the
concession to assist in its control, since it rapidly expands from forests into
grasslands where it is an unwanted component which domestic stock nor-
mally refuse to eat.
Both sojia and motia are tufted perennial grasses with numerous stiff
stems, those of motia being generally more upright and fewer per plant than
sofia; motia is also taller with stems 2.5-3.0m, Sofia 1.O-1.25 m. Stem colour
in motia is generally yellow or greenish-yellow, in sofia usually purple to dark
greenish purple, but there are many gradations between the two basic stem
106 Chapter 4
very small quantities or absent. The main components ofpalmarosa oil from
selected origins are shown in Table 4.12. T h e major constituent of ginger-
grass oil is geraniol to 65% and other tertiary alcohols, particularly perillic, in
varying proportions. Gingergrass oil differs from the other cymbopogon oils
discussed as it consists mainly of p-menthane derivatives rather than ali-
phatic monoterpene derivatives. It is frequently difficult to obtain a pure
sample of gingergrass oil as it is often mixed with poor-quality palmarosa oil.
Variation can be induced artificially: for example exposing seeds to low-
dosage gamma irradiation caused changes in oil content and oil character-
istics; 7, 10 and 15 kilorads (krad) increased oil content over control, did not
affect the refractive index, but at 10 and 15 krad increased specific gravity.
These two levels also significantly increased total geraniol content and ester
value after acetylation, but decreased acid value. All radiation levels reduced
percentage germination in direct proportion to treatment; at 15 krad germi-
nation was reduced by 60% (Gupta et al., 1981).
Both varieties produce inflorescences on long stems with extended,
usually narrow, oblong panicles of small, spicate flowers; more widely
spaced and hairy on motia, compact and stellate on sojia. Since there is wide
variation within the two varieties and considerable natural crossing, size,
shape and appearance of individual inflorescences are no reliable guide to
identification. Flowers of both produce large amounts of pollen, cross easily,
and set quantities of viable seed. Main flowering in India is September-
November, the major harvesting period.
Ecology
Palmarosa and gingergrass, although morphologically similar, are adapted
to different environments as previously noted, and that of other Cymbopogon
spp. has been described in detail (Whyte, 1968; Lazarides, 1980). T h e range
of palmarosa has also been extended by a variety of uses other than for oil. In
India and Pakistan in particular, but also in other Asian countries, the grass
is widely recommended for soil conservation purposes on erodible hillsides,
to heal gullies, on contour banks and terraces, re-afforested areas and similar
situations.
Palmarosa has a natural range in India from 12-32"N, but is grown
commercially on Java and the Seychelles almost on the Equator, and on
Madagascar around 20"s. It is usually grown at 300-1200 m but as a local
cash crop from 150-800 m depending on rainfall and soils (Husain et al.,
1988). Palmarosa was considered day-neutral, but recent studies indicate
greatest rate of oil synthesis and highest geraniol content occurs with a
daylength approaching 16 hours (Ghosh and Chatterjee, 1990). Wild pal-
marosa is often found in drier hilly areas with a rainfall below 600 mm and,
although this indicates a degree of drought resistance, a rainfall of 1500 mm
is necessary in north India for high leaf yield (Singh, 1977). However, 750
mm produces a reasonable yield and palmarosa can probably be grown as a
cash crop in many low-rainfall regions of India. The intensively managed
Java plantations with four or five annual cuttings were in areas receiving at
least 1500 mm, with supplementary irrigation. If palmarosa is to be grown as
a commercially profitable crop, it must be well fertilized and this requires
adequate levels of soil moisture; a minimum of 1500 mm, well distributed,
plus supplementary irrigation during dry periods. Palmarosa prefers warm
sunny conditions with a daytime temperature of 20-25°C; extended periods
at 25-30°C can significantly reduce oil yield and suppress flowering. Frost
causes severe damage at any growth stage, and at higher elevations plants
die-back to ground level in the cold season. A slight frost during harvesting
can halve herbage oil content. The effects of hail and high winds are similar
to citronella.
The general effect of climate on oil content and composition is mainly via
the rate of growth and physiological changes which take place in plants.
Thus a dry season will retard growth and reduce oil yield, while adequate
rains promote vegetative growth and increase oil yield. However, oil compo-
sition, especially geraniol content, is usually inversely correlated with these
effects. Annual climatic variation normally has a greater effect on oil content
and composition than any other factor, and thus adverse conditions for plant
growth can seldom be profitably offset by management, i.e. more fertilizer or
applied water. Seasonal variation is also important and harvesting at the
most suitable period is essential to obtain a high yield of good-quality oil (see
section on harvesting).
Gingergrass, although occupying the same general range in India as
palmarosa, thrives below 500 m. A rainfall of 1000 mm with no extended dry
periods produces most herbage. Heavy sustained rains as in monsoon areas
of India which saturate the soil are well tolerated, as are less well-drained
Gramineae 109
deduced from the information in Table 4.13. Since there are usually one or
two cuts annually on commercial plantings in India coinciding with flower-
ing in April and September-October, nitrogen is applied as a top-dressing in
February and August, usually ammonium sulphate or urea, since there are
no data comparing the various nitrogenous fertilizers. How the top-dressing
is applied is unimportant provided coverage is even, with foliar application
not significantly superior to the easier surface method (Singh et al., 1980b),
but foliar application increased net assimilation rate and thus rate of growth.
While the nitrogen level of 40 kg Niha is common, palmarosa also responds
to much higher rates; at Bangalore there was an almost linear rise in oil yield
by increasing urea top-dressing to 240 kg Niha; highest oil yield was also
correlated with the greatest plant population at all nitrogen levels (Prakasa
Rao et al., 1985a). A relationship may exist between high rates of nitrogen
and high geraniol content of oil, but also be correlated with a lower oil
content. This is of little practical importance compared to climatic effects
and time of harvest, which have a far greater influence on both oil yield and
geraniol content.
Data from many regions indicate a moderate phosphate requirement, and
once the minimum necessary to ensure optimum returns from nitrogen has
been determined, there is little additional benefit from larger amounts. A
general recommendation in India is 20 kg Piha at planting. The phosphate
content of maturing leaves falls less quickly than that of nitrogen or po-
tassium and a greater proportion is harvested in herbage. Since a plantation
can last for 15 years it is probable a further phosphate application every third
year or use of slow-release compounds would increase oil yield by maintain-
ing healthy plants.
There are no data on comparisons between various types of phosphatic
fertilizers, but remarks elsewhere in this chapter are relevant. Phosphate can
affect oil content of herbage and oil constituents; a deficiency reduces
response to other nutrients, suppresses volume of herbage and thus total oil
yield per hectare, but may increase geraniol content. As noted, the optimum
level of phosphate reacts positively with other nutrients to increase herbage
Gramineae 111
and oil yield per hectare, but can reduce geraniol content. For all practical
purposes the effect of phosphate on oil constituents can be ignored by
producers since the variation is generally less than that due to seasonal
influences.
Palmarosa appears to have a lower requirement of potassium than cit-
ronella or lemongrass but, as data in Table 4.13. shows, there are high levels
of potassium in leaves compared with phosphate, with a much higher level in
stems; thus substantial amounts of the nutrient are removed at harvest. The
response to potassium, however, is usually limited to alleviating a deficiency
which reduces the effect of other plant nutrients; higher levels have little or
no practical value. Potassium is normally applied pre-planting as part of a
compound fertilizer but, as with phosphate, it is possible that further
applications during the life of a plantation may be profitable. Potassium at
approximately half the phosphate level is usually recommended in regions
where there is no major deficiency. Potassium has no significant effect on
herbage oil content or oil constituents at levels normally applied.
There are few reports of minor element deficiencies, or response to their
application. Levels of calcium, chloride, iron and sodium in palmarosa are
shown in Table 4.13. and iron content is considerably higher than in many
tropical grasses. On a sandy soil of p H 8.2, a crop receiving 10 tiha FYM or
40 kg Nlha gave significant increases in herbage and oil yield when foliar
sprays of ferric sulphate and manganese sulphate were applied at tillering
and just before flowering (Pareek et al., 1984). In the presence of adequate
nitrogen and phosphate 2.5 kg/ha zinc sulphate increased herb and oil yield
in Uttar Pradesh State (Sharma et al., 1980).
Cultivation
The following remarks apply to cultivated palmarosa, which will become
more profitable as high-yielding strains are utilized. Palmarosa can be grown
from seed or splits, occasionally cuttings but, as viable seed is produced in
quantity, the first method is preferred. In Assam, seedlings outyielded splits
in both herbage and oil yield per hectare and, although there were no
significant differences in oil content of herbage, geraniol content was higher
in plants grown from seedlings, 73 compared to 66% (Kanjilal et al., 1981).
Seed must be obtained only from pure stands of palmarosa or plants
specially grown for the purpose, as seed collected from natural stands often
contains gingergrass seed, especially in India and Pakistan. A mixture of
palmarosa and gingergrass was a major factor in the decline of oil production
in the Seychelles. Seed should be sown in specially prepared seedbeds as
described for C. flexuosus, and about 2 . 5 kg fresh cleaned seed will provide
sufficient seedlings for 1 ha. Palmarosa seed is very small and usually mixed
with fine sand or a similar inert material to assist even distribution. Sound
seed germinates in 10-20 days and seedlings are ready for transplanting in
approximately 6-8 weeks when 15-20 cm high.
112 Chapter 4
Harvesting
Interesting accounts of the techniques used to harvest and distil palmarosa
in India are contained in Watt (1908) and Guenther (1950), and have
changed little in many areas of the country (Singh, 1977), reflected in the
very poor quality of much local oil (see Fig. 4.12). In plantations, harvesting
can be either manual or mechanical and equipment is now available that will
cut, chop and load herbage for direct transport to stills. This system
harvested 1 ha per hour in Madagascar; hand harvesting the same area
required 6 man-days (Salomon, 1979). The harvest period could be ex-
tended by planting early and late flowering strains, allowing more efficient
Gramineae 113
use of machinery and distilling units. Such strains have been selected from
east Indian types, with a tall selection flowering in September-October, and
a late-flowering dwarf in October-December (Verma et al., 198 1). Plants
should be cut when flowering, about 6 months after planting out, then twice
annually in subsequent years. Cutting is usually September-October in the
establishment year in India, then April-May and September-October sub-
sequently. In Africa the first cut was often in the middle of the rainy season,
December-March, with a second cut at its end.
A direct correlation exists between cutting in wet or dry seasons and
herbage yield and geraniol content of oil; in the former, herbage and total oil
yield per hectare is high but geraniol content low, the reverse occurring in
the dry season. Oil characteristics may also change, as oil from herbage
harvested in May and October in India differed markedly in specific gravity,
refractive index and optical rotation, but geraniol content was much less
affected (Gulati et al., 1966). This variation in characteristics has also been
noted in oils harvested in succeeding years when there were significant
differences in weather conditions during plant growth. Time of cutting is
important and when to cut is usually judged on the extent of flowering; in
India, oil content of herbage and geraniol content of oil are highest at full
bloom, and cutting slightly earlier than later recommended since oil and
Distillation
Wilting for a day in India reduced oil yield by half compared to distilling
fresh herbage, but careless handling or longer periods can reduce oil yield by
75%. Despite this, many contractors cutting wild stands often field wilt for
4-5 days then stack the dried grass for up to 2 1 days before distilling, partly
to reduce the volume to be manually transported. Oil yield is generally
12-24 kgiha. Palmarosa can be either hydro- or steam distilled, and remarks
in the lemongrass section are applicable. Provided stills are of the improved
designs and efficiently operated, good-quality oil should result. A major
Gramineae 115
reason why so much poor-quality oil is currently produced is that these basic
requirements are frequently ignored. Spent grass can be processed for
stockfeed and, in India, gave an average content of (per cent): dry matter 34,
crude protein 12, total carbohydrates 7 . An attempt to make spent grass
silage as food for cows and buffaloes in city dairies was unsuccessful; animals
rejected the material as unpalatable without additives, and transport costs
were too high compared with other feeds. Spent grass can be utilized in
papermaking.
Costing palmarosa oil production in India showed a plantation of less than
100 ha in 1988 was not viable, since the equipment, machinery and stills
necessary were similar for 50 to 250 ha. Calculations were based on an
annual average herbage yield of 25 t/ha, with an oil content (wet basis) of
0.1-0.15%, and a 10 year plantation life. In Uttar Pradesh, small farmer
cooperatives who obtained 25 kg/ha oil annually, had an annual expenditure
between Rs 1200 and 1500 in 1972, and received around Rs 5000 for oil
with a geraniol content of 85-90%. A lower price was obtained for oil with a
lower geraniol content. (In 1972 US$ 1.OO = Rs 15.00.) Since these figures
were published, average herbage and oil yield per hectare have nearly
doubled, but so also have costs; potential oil and herbage yield however have
increased three times. A private communication in 1994 from a leading
South Indian oil merchant stated the ratio of expenses to income was now
narrower, but an efficiently run plantation was still profitable, and profit
would rise substantially as new cultivars were utilized and the proportion of
oil exported increased.
Extending plantation life and maintaining a high annual oil yield through
efficient management are significant factors affecting long-term profitability.
If spent grass could be sold for stockfeed or papermaking, this would again
increase profitability, provided the cost of alternative fuel was less than the
price received for the grass. One large oil producer in Maharashtra State
successfully employs the nucleus plantation system, buying herbage from
small farmers to supplement output from its own plantation. Profitability of
palmarosa production in Madagascar was not only aided by the mechanical
harvesting system employed, but also by extending the use of distilling
plants. Palmarosa could be harvested in the interval between ylang-ylang
and basil, and the overall cost of operating a distillery significantly re-
duced.
Palmarosa Gingergrass
Angola India Madagascar Java India
Specific gravity (20°C) 0.8875 0.887 0.884 0.885 0.925
0.900 0.889 0.900 0.935
Refractive index (20"Cl 1.4745 1.468 ng 1.472 1.485
1.476 ng 1.477 1.489
Optical rotation (20°C) -0" -I0 ng -3" -46"
to" t3" "g t3" t40"
Ester number 62(max) 15 ng ng 8
45 ng ng 40
Geraniol (%) (total) 93 85 85 80 40
95 91 90 49
Solubility (viv alcohol 1:1.5 1:1.5-3.0 ng 1 :2.0-3.5 1:3 often
70%) insoluble
ng, not given. Figures in columns are range.
Source: Burger, 1958 (Angola); Virmani, 1968 (India); Randriaminharisoa and Gaydou, 1987
(Madagascar); Customs Department, Java.
Grarnineae 117
Palmarosa Gingergrass
A B C D D
Specific gravity a 0.882 0.882 0.855 0.878 0.900
0.896 0.894 0.897 0.890 0.929
Refractive index a 1.4715 1.472 1.473 1.471 1.476
1.4780 1.478 1.477 1.475 1.491
Optical rotation -3O -3s -2 O -2" -14"
t 3c t 3O t 3O t 3O t 54"
Ester number 10 10 ng ng ng
70 40
Geraniol (total %) 85 ng 85 88 36
94 ng 95 94 60
Solubility (viv 70% ng ng 1:2 1:1.5 Normally
alcohol) insoluble
'At 2O'Cc; ng, not given; A, Grasse Essential Oil Association; B, Afnor; C, EDN; D, Indian Standard,
Figures in columns are range.
Vetiver
Vetiver, Vetiveria zizanioides L.(Nash.), is native to India and occurs as a
flowering and non-flowering type. Both have been introduced into other
tropical countries, where they are well established or naturalized. In African
and Caribbean countries, vetiver is highly regarded as an aid in soil con-
servation, its dense root mass binding loose soil (Smyle and Magrath, 1993).
Oddly, it was in India that least use was made of the plant for this purpose,
although it is now officially recommended (Allison, 1990). Most Indian, Sri
Lankan and Malaysian oil is from wild plants although there is growing small
farmer production in the Indian states of Tamil Nadu and Andhra Pradesh.
Vetiver is grown commercially for oil in Java, the Seychelles, Reunion,
Brazil, Haiti and Japan, but in the USA, Central America and some West
African countries for other purposes. The plant and rhizome have very
ancient usage in India for their aromatic properties and are mentioned in the
Hindu Atharvaveda. As khas-khas it was sufficiently important in the twelfth
century to be taxed, but there is no accurate date for first use of the oil; in
India and Sri Lanka it is known as 'the oil of tranquillity'. Vetiver plants,
including parts of the rhizome, are used in India and other Asian countries to
make aromatic fans, mats and screens (tatties) in houses, and leaves cut prior
to harvesting are used as thatch. Rhizomes are hung in houses to scent the air
or placed in sachets to perfume clothes; a current use is to place bags of wet
rhizomes in front of electric fans to cool and scent rooms. Rhizomes used for
these purposes are often repurchased by traders, who mix them with fresh
rhizomes for sale to distillers. A drink made from fresh rhizomes is taken as
118 Chapter 4
a stimulant and tonic, but the author found the taste unpleasant. Vetiver
could be the calamus of the Christian Bible (Duke, 1983), and vetiver
subsequently appeared in European Materia Medica but was not generally
popular.
Java was a major exporter of vetiver oil prior to the Second World War, the
plants grown mainly by smallholders but distilled by Dutch companies. This
vetiver was the south Indian non-flowering type, the date of its introduction
uncertain, but it was probably brought by merchants from India. Customs
Department statistics show exports of vetiver oil from Indonesia in the years
1989-1995 ranged from 58 to 65 tonnes, with an estimated 20,000 ha under
cultivation.
Vetiver has long been cultivated in China with similar uses to those in
India, but commercial production for oil began only in the mid-l950s,
reached 10 t in the 1960s and has since declined to a few tonnes as vetiver
was replaced by more profitable essential crops. Although widely planted in
the Philippines, it is not generally cultivated to produce oil (Tung and
Balina, 1993). Whole plants are used as in India, and a decoction of the
rhizome is believed to dissolve kidney stones.
Vetiver was introduced to the island of Reunion together with other
essential oil crops and good quality oil produced by French farmers until the
mid-l950s, but is now mainly grown by small farmers (Prugnieres, 1978).
On Mauritius, a paste made from pounded fresh roots is considered an
abortifacient. Development of vetiver oil production in Central America, the
Caribbean Islands and Brazil paralleled that of lemongrass. Little is now
produced on a commercial scale, but there is considerable potential as a
small farmer crop.
Botany
Vetiveria zizanioides L. (Nash) (syn. Andropogon muricatus Retz., Anatherum
zizanioides (L.) H&C., Phalaris zizanioides L.); khas-khas generally in India
and Sri Lanka (khus-khus in Europe), akar wan@ in Indonesia; Tamil
vetivern, origin of the name vetiver (Nair et al., 1979). T e n species are
included in the genus Vetiveria and two distinct types of V. zizanioides are
recognized. Vetiver is a diploid, 2n = 20. Methods of classifying vetiver
clones have been developed to aid in assessing their value in breeding
programmes (Kresovich et al., 1994). Vetiver is a densely tufted perennial
grass to 3 m, normally developing large clumps, with a rhizome and fibrous
root system containing an essential oil.
The main rootstock is a stout, branching rhizome developing an extensive
but not deeply penetrating fibrous mat of aromatic roots. The colour varies
from whitish yellow on young roots to reddish brown on roots 18-24 months
of age. Rhizomes and roots from Haiti, Reunion and Angola may appear
black, partly due to the dark colour of the volcanic soils in which they grow.
The majority of roots occur 15-30 cm below the crown, thus root growth
Gramineae 119
and extent are directly related to soil conditions (q.v.). Since rhizome and
roots contain the essential oil, their rate of growth and extent are important
and have been the basis of selection in India, where dry root weight varied
from 32 to 58 g per plant; the weight increase due to a rise in size and number
of roots and rootlets (Selvarajam et al., 1982).
For brevity, rhizome and roots will hereafter be designated roots, except
where identification is required. Data indicates there may a relationship
between tall, compact plants and low root yield; if so, removal of low-
yielding plants and rejection as planting material should be simplified. Total
root-mass can be greatly increased by manipulating soil moisture levels; in
India, single vetiver plants were grown in earthenware pipes 500 cm long
with water applied only a t increasing depths for 12 months, when roots were
350 cm long and weighed 210 g. Solvent extracting roots in sections from
the soil surface showed the first 75 cm produced most oil, but a significant
amount was obtained in the next 125 cm. Total extractives on a dry root
basis in succeeding root sections were, in per cent: 45 cm (3.19), 30 cm
(2.46), 30 cm (1.94), 45 cm (1,61), 45 cm (1.38), 45 cm (0.81), 111 cm
(0.67). A plant growing in the field alongside yielded 1.93% extractives
(Gowda and Ramaswamy, 1960). In addition to the essential oil, roots
contain fructose, sucrose, glucose and, interestingly, free glycerol, not
usually found in plants. Mature roots have a pleasant aroma and greatest oil
content, and yield high quality oil.
Since the essential oil is contained in the roots, a most important factor
determining profitability is root oil content, and there is considerable culti-
var variation in both the weight of roots and oil content. A south Indian type,
Nilamber, at the Odakkali Research Station, gave a 3-year mean annual
yield of 3856 kg roots and 14 kg oiliha, compared with the lowest which
yielded 639 kg and 2.2 kg respectively (Nair et al., 1979). Other selections
yielded 5900 kg roots and 40 kg oil/ha (Nair et al., 1982), and up to 62 kg oil/
ha (Gupta et al., 1983), while an autotetraploid cultivar had a very high oil
content (iavania, 1991). Thus the profitability of vetiver oil production
could be greatly increased by using selected material. Root oil content
however should not be the sole factor, since total yield of roots is often more
easily increased. The average oil content of Indian material seldom exceeds
2.0% and is well below that recorded in other countries of 2.2-4.6%.
The oil content of roots collected from wild plants in India is below 1.O%,
averaging 0.15% after distilling, and only the very low cost of collection and
distilling allows this system to continue. The vetiverol content of oil also
varies substantially between selections, 67-86%, and high vetiverol content
was related to high specific gravity, esters and free alcohols (Gupta et al.,
1983; Maheshwari et al., 1986). Oils collected from wild (mainly in the
north) and cultivated vetiver (mainly in the south) of India are differentiated
by calling the former khus oil, the latter vetiver oil. Indian Standards
120 Chapter 4
India
lssa lSSb Reunion Indonesia Japan
~ ~ _ _ _~ _ ~
Institution specifications state khus oil has a high laevorotation, while vetiver
oil is dextrorotatory (see Table 4.16).
Vetiver oil is extremely complex containing mainly sesquiterpenoids, and
their biosynthesis has yet to be fully determined. The main constituent is
vetiverol, and its relative abundance normally establishes oil quality. De-
tailed analyses from the main producing countries have been published
(Lemberg and Hale, 1978). The main characteristics of oil from selected
origins and official standards are shown in Table 4.16. The range in
characteristics can be wide even in local strains; 14 Indian cultivars gave the
following at 30°C (mean in brackets): specific gravity 0.9302-0.9997
(0.9691); refractive index 1.5168-1.5205 (1.5189); acid value 19.34-50.30
(33.88); ester value 17.62-68.47 (31.89); ester value after acetylation
154.41-190.16 (167.24); free vetiverol percentage 46-78 (61); total veti-
verol percentage 67-86 (73).
The stems are glabrous, stiff, erect, usually 2-3 m, and arise from a central
branched spongy rhizome. Stem number per plant is highly variable and
apparently not directly related to root extent or weight and thus unreliable as
a factor in selecting for oil yield. The leaves are erect, mid green, to 1.5 m,
narrow 7-1 0 cm, rigid, prominently veined, keeled and glabrous, with
scabrid margins; the lower sheaths are strongly compressed and flabellate-
imbricate. Detailed descriptions have been published of leaf anatomy (Met-
calf, 1960; Rajendrudu and Das, 1981) and photosynthetic activity (Maffei
et al., 1995). The leaves contain virtually no essential oil, and prior to
harvesting are burnt off, cut and fed to stock, or used as thatch. Analyses in
India gave the following on a dry matter basis (per cent); crude protein
(6.1-6.7), ether extract (1.1-2. l), crude fibre (34.7-42.2), nitrogen-free
Gramineae 121
Ecology
Vetiver is a grass of the wet tropics, usually occurring in India below 300 m,
but altitude is apparently of less importance than climate, since it is planted
to 700 m in other countries. Vetiver grows well at lower elevations than
geranium and thus does not compete with the latter where both are grown,
as on Reunion. With good management both can be distilled in the same
plant. A moist, sunny climate with no extended hot, dry period is preferred
and soil in the root zone should not dry out as this will substantially reduce
root growth and oil content. A day temperature of 25-35OC is most suitable,
but higher temperature has little effect provided there is adequate soil
moisture or humidity; below 2OoCfor extended periods reduces growth, and
below 15°C normally inhibits development. Vetiver grows naturally in open
forest or woodland where there is sufficient moisture and intermittent shade,
and would probably tolerate this environment when cultivated; on Java
heavy shade reduced root growth and oil yield. Vetiver may be intercropped
in coconut or arecanut which have a high, relatively open canopy; it grows
well, for example, on coastal sands in southern India where there are
extensive coconut plantations.
A minimum rainfall of 1500 mm is desirable for commercial oil produc-
tion, with apparently no upper limit provided the soil is free draining; vetiver
favours moist, humid conditions but cannot withstand prolonged water-
logging. As an escape it is often found on river banks or marshy areas. In
regions with a defined rainy season, this influences time of planting as soil
moisture must be sufficient to allow splits to root quickly, since a growth
check at this stage will extend the period to harvest. The effect of time of
planting on total root and oil yield is substantial; planting monthly from May
to October reduced root yield from 4555 to 2412 kg/ha, but oil yield rose
from 7.03 to 10.38 kg/ha (Nair et al., 1979). Smallholders in Java or Reunion
harvest in the wet season and replant immediately, since the wet soil makes
digging easy and adequate soil moisture ensures quick growth of splits. This
in effect means an annual harvest, since most smallholders are unable to wait
20-24 months to sell their crop.
122 Chapter 4
Frost at any time is usually fatal. Hail causes little damage to mature
plants with well-developed roots, but if leaves of newly established plants are
shredded, growth will be retarded. Wind has little effect on established
plants, but hot dry winds may retard growth of younger ones. Vetiver is so
robust it is considered virtually cyclone-proof in Reunion, and thus an
insurance against total crop loss should one of these terrible storms occur.
selections with high root yield or root-oil content apparently give a greater
response to added nutrients.
Fertilizers should be applied to the seedbed prior to planting splits; on
Reunion a 1:2:2 NPK mixture at 80 kg/ha gave the optimum results
(Gailleton, 1977), but has since been replaced by a common cane fertilizer
of 15:12:24 NPK. Nitrogen applied alone in Georgia, CIS, had no effect on
yield of roots or oil content, but the highest yield of roots was obtained when
200 kg Nlha, 87 kg P/ha and 16 kg Wha was applied (Sichinava, 1974).
Nitrogen should not be applied unless trials establish it is either essential or
profitable, and then only at planting. The few trial results available show
phosphate promotes root growth and increases total root yield per hectare,
but there appears to be an optimum amount which must be determined
locally, as an excess may depress yield. Additionally, the often small increase
in oil yield could be worth less than the cost of fertilizer and its application.
In India, there was little increase in yield of roots and oil from two levels of
phosphate and potassium, and the 5% increase in oil between 0 and 33 kg
Plha was barely significant (Nair et al., 1979).
An increase in root and oil yield generally results from application of
potassium, but it was unclear from available data if this was due to a local soil
deficiency or a basic plant requirement. If the latter, then research into the
potassium requirement of vetiver and its use in plant metabolism would be
rewarding. There is an optimum level for potassium and higher rates
generally depress yield of roots and thus total oil yield. In India there were
significant increases in root weight and oil yield from application of 22 kg
Wha over 4 years, although there was annual variation in both when neither
was significant (Nair et al., 1979). However, as the mean oil increase was
nearly 2 kg/ha this would have been very profitable at that time. The only
record of a minor element deficiency is a severe iron-induced chlorosis on
vetiver growing in a peat soil in Malaysia (Ng, 1972).
Cultivation
The remarks on cultivation in the introduction to this chapter are applicable.
Roots are the most important plant part, thus ploughing to a depth of at least
30 cm should be the aim in more clayey soils. If vetiver is to follow vetiver, it
is especially important to remove old roots since these have a lower content
of poor-quality oil. A major objective should be maximum weed reduction,
particularly tropical grasses such as Imperata cylindrica, Sorghum halepense,
Digitaria spp., and it may be necessary manually to remove clumps or spot-
spray with a herbicide. On sloping lands, contour banks or broad-based
terraces reduce the danger of soil erosion, and vetiver is planted in many
countries to bind soil on these terraces, but not harvested.
Vetiver can be grown from seed or splits, but commercial plantations use
splits. Micropropagation has been successful in the laboratory (Mucciarelli
124 Chapter 4
et al., 1993). Methods noted in the lemongrass section are generally applica-
ble when seed is used, and production of suitable splits is similar to
citronella. Selection of parent material is important since it is very difficult to
counteract the adverse effects of planting low yielding or low oil-content
material. In India, oil yield increased from the local average of 67 to 88 kgiha
without a similar increase in volume of roots harvested when selected
material was planted, and oil cost per unit substantially reduced. At least
150,000 splits per hectare should be available, more if two or three splits are
planted per hill. Most small farmers plant manually but many horticultural
planters are suitable. A spacing of 30-45 cm rows X 20-30 cm is common
for ridge or bed planting; 40 X 40 cm for surface planting. Accurate in-row
spacing within reasonable limits is not essential, since roots in suitable soil
normally expand to fill the available space. Splits can be sown on ridges, beds
or the flat, and the system chosen depends mainly on the soil type and need
for irrigation or drainage. There is little difference in root and oil yield per
hectare between the various methods on those soils where one method is not
obligatory; over 3 years in India, planting on ridges, 60 and 70 cm wide beds,
or on the flat gave average yields in kgiha of roots (oil in brackets): 4087 (14),
41 95 (1 l), 3725 (1 1) and 3834 (1 l), with no significant differences between
them (Nair et al. , 1979). Narrow beds with two rows of splits is common in
southern India. Twin rows are preferable when mechanical diggers are used,
with row and bed spacing to suit the machinery. Planting should be into
moist soil following irrigation or after rain; in India in May but not later than
July. It may be necessary to fence newly sown vetiver plantations against
domestic stock, as young growth is palatable although not mature foliage.
Vetiver plants grow rapidly once established and only one weeding is
usually necessary, as leaves quickly produce a closed canopy. Weeding is
usually manual or mechanical as herbicides are seldom used except as
directed sprays after planting; the following have been so used: 2,4-D,
atrazine, glyphosate, linuron, MCPA, simazine and there is tolerance of
dicamba and picloram. Vetiver generally follows vetiver for several years,
unless a build-up of disease requires an interval between crops.
Harvesting
The most important field factor determining oil content is age of roots at
harvest, older roots containing more and higher quality oil. A period of
15-1 8 months is usually the minimum necessary to obtain a profitable yield
from cultivated plants, but wild plants are harvested at any age. After 30-36
months, oil content drops and roots are virtually worthless as oil producers.
However, roots harvested at the optimum time and stored in a clean dry
environment can be distilled several years after lifting, loss in oil content
being offset to a considerable extent by improved quality and higher price.
In southern India, oil content progressively increased from 0.10% at 10
months to 0.56% at 15 months, 0.79% at 17 months and 0.87% at
Gramineae 125
21 months, with the optimum 16-18 months (Sobti and Rao, 1977; Nair et
al., 1979). In some Indian selections high ester value was also correlated
with high vetiverol content (Gupta et al., 1983). Although 3 harvests at 11
month intervals may give a greater yield of oil at 40 kgiha than 30 kgiha from
two harvests at 18 months, the higher quality oil from more mature roots is
worth more than the increase in weight. It also costs less to produce in terms
of harvesting, distilling and replanting expenses. Despite this, most small
producers in India, Java, Reunion and elsewhere, desperate for money,
harvest roots annually to their long-term disadvantage.
Prior to lifting roots, leaves are either slashed, mowed, laid in rows and
burnt, or removed to provide the raw material for various products noted in
the beginning of this section (see Fig. 4.13). Burning is common on large
estates, and provided herbage is dry, a fast burn does not harm roots. The
ash remaining is a useful fertilizer and soil ameliorator. A worker in Reunion
is expected to cut leaves, dig roots, remove soil and pile for transport about
40 kg roots daily. The number of men required to keep a still running is thus
substantial but, when roots are dried for export, time is of lesser importance
and collected roots are baled when sufficient are available. A shaker-digger
has been developed that greatly reduces harvesting time, but cannot be used
on hillsides where much small farmer vetiver is grown (Ducreux, 1993). In
Zaire, 50 man days were required to harvest and replant 0.5 ha based on a
day of 5-6 hours. The high labour requirement for manual harvesting is a
factor limiting expansion of vetiver cultivation where mechanization is not
possible. Roots can be easily and cheaply lifted using a potato-type digger
which also recovers more roots than manual digging. Such diggers are now
available in sizes ranging from small manually guided, to large four- or six-
row power-driven machines. Cooperative ownership, as in Japan, provides a
useful method of reducing individual investment in this type of equipment.
Whether roots are lifted by hand or machine it is essential to remove soil,
stones and any other extraneous debris, and trim excess green material.
Root yield in India is 2000-5000 kg/ha, relatively low compared to those
obtained at local experimental stations where selections averaged
7000-9000 kg/ha. Since a higher root yield is apparently easier to attain than
increased oil content, a significant increase in oil yield per hectare is thus
possible following a relatively simple selection programme. Root yield on
Java was initially around 1000 kgiha, average oil content below 0.2%, but is
now 3000 kgiha at 0.5%. On Reunion the yield from small plots is
2000-2500 kgiha with oil yield of 0.6-1.5%, but up to 5000 kg/ha obtained
under improved management. Yield in Brazil of 700-1500 kg/ha is directly
related to soil fertility. An oil content of 1.0% in young and 1.5-2.0% in
older roots was very high for a commercial plantation, obtained on a British-
owned and very well-managed estate which operated until 1952. Haiti roots
contained 0.5-1.5% but no root yield, was noted; likewise 2.5-3.0% from
126 Chapter 4
Distillation
The extent to which roots are dried depends on whether they are to be
quickly distilled or transported and exported. Roots to be distilled on site or
within days are usually allowed to remain in the field, or stacked on field
margins. The interval prior to distillation must be established locally, since
there are conflicting reports on the degree of oil loss. Whether this is due to
climate, type of storage or is an inherent characteristic of local cultivars is
unknown. Roots to be transported to distant distilleries or exported are
usually washed, sun-dried and baled. Provided they have been correctly
stored, dried roots normally produce higher quality oil than fresh roots, and
good quality dry roots on Java yield up to 3% oil. Dry root exported to
Europe and distilled in modern plants produced oil of very high quality, but
at a substantially higher cost than locally distilled oil. Root age basically
governs oil yield but distillation and subsequent treatment of distillate
frequently determines oil quality. Roots less than 15 months old generally
have a low oil content, the oil has a harsher, grassy or earthy odour disliked
by manufacturers, and is often reprocessed and added to higher quality oil.
Roots over 24 months old have high oil content, a lower proportion of
fibrous material, yield high quality oil, but distilling time is prolonged. Most
Gramineae 127
suitable material is between 15 and 20 months, fibrous, clean and dry, has
the highest ratio of oil to weight of charge, and produces oil containing all the
desired components. Large modern distilleries are normally independently
owned and operated, buying raw material from producers. A major problem
is to ensure only suitable material is offered and thus an effective inspection
and buying procedure is essential. Inclusion of immature or aged roots,
partially distilled and redried material, or roots which, in India in particular,
have been recovered from tatties and the like must be quickly distinguished
and rejected. Unless an inspection system is rigourously applied, the au-
thor's experience is that the quality of offered material quickly deteriorates;
most small growers are always short of cash and will use any method to
increase their income!
Distillation is more complicated than for the other grasses discussed,
partly due to the material but also to the complex nature of the oil. A
substantial proportion of oil is obtained by direct-fired stills fuelled by wood
or dried foliage and operated by smallholders or cooperatives. Distilling is
thus time-consuming and labour intensive, made more so by the high
boiling-point of the oil's main constituents. One common technique in India
is to soak chopped roots in water for about 12 hours before distilling. A
charge of 40 kg is distilled for 12-24 hours; up to 125 kg for 72-96 hours in
larger plants. North Indian stills mainly use roots from wild plants, are
portable and taken from site to site between December and January. Opera-
tors buy roots from licensed collectors working in state forests, plus any
available cultivated material. Stills consist basically of a copper vessel and
bamboo condenser, probably little changed since their introduction cen-
turies before. Stills are normally heated by wood or cow-dung, and although
oil recovery and quality is poor, this is of little importance since most is sold
locally.
Direct-fired stills on Reunion are larger with a capacity of 400 kg of sun-
dried roots distilled for 36-48 hours. Charges are 150-500 kg on Java
distilling for up to 48 hours, and material must not be tightly packed as this
not only prolonged distilling time, but the resultant oil had lower acid and
ester numbers and vetiverol content. Maximum oil yield of 2% was achieved
at bulk density of 0.07 kgil compared with 1.43% at 0.1 kgil (Hardjono and
Deswert, 1973). Government sponsored stills in main producing countries
now use steam with cohobation, substantially reducing distilling time. Roots
are usually sun-dried to lower moisture content and mechanically chopped
into 2.5-5.0 cm lengths; a charge is 500-700 kg distilled over 36-48 hours.
The resinous material remaining in Indian stills, about 25% of total ex-
tractives, had the following characteristics: specific gravity (30°C) 1.182;
optical rotation +48.4"; acid value 46.9. Roots remaining after extraction
can be dried and used as fuel, burnt to provide ash or composted together
with leaves as fertilizer.
128 Chapter 4
compounds, but a high ratio of those with a high boiling-point, and con-
sidered more accurately to reflect the oil in vetiver roots. Combined with its
excellent fixative properties it is highly regarded by blenders. Isolates from
the carbonyl fraction are insect-repellent gain et al., 1982), the whole oil
proved toxic to mosquito larvae (Murty and Jamil, 1987), and also possessed
fungicidal properties (Dikshit and Husain, 1984).
Vetiver oleoresin is prepared by further extracting concrete, most recently
by liquid carbon dioxide, which gives a golden yellow, more mobile liquid
containing less non-volatile compounds. A very stable extract with a rich
odour is obtained by molecular distillation. Vetiver attah is produced in
India by condensing the distillate in an oil of choice, sandalwood, ground-
nut, sesame, etc. and sold wholly on its odour related to the product's end
use, as flavouring, unguent, or hair-dressing.
References
Abegaz, B., Yohannes, P.G. and Dieter, R.K. (1983) Constituents ofthe essential oil
of Ethiopian C. citratus. J . Nut. Prod. 46, 424-426.
Akhila, A., Tyagi, B.R. and Naqui, A. (1984) Variation in essential oil content of
palmarosa at different stages of plant growth. Indian Pe$um. 28(3/4),
126- 128.
Allison, G . (1990) Vetiver grass - a hedge against soil erosion. Shell Agric. 6,
14-15.
Ames, G.R. (1968) In: Essential Oils Production in Developing Countries. Tropical
Products Institute, London, pp. 75-78.
Anon. (1984) Essential oil of citronella - S r i Lanka. Analyst, 109, 1346-1348.
Arctander, S. (1960) Pe$ume and Flavour Materials of Natural Origin. Mrs M.G.
Arctander, 6665 Valley View Blvd, Las Vegas, Nevada, USA.
Ashour, F.M. (1980) Physiochemical properties and chemical composition of ve-
tiver oil. Ann. Ap'c. Sci. (Moshtohor) 12, 183-197.
Bagaturiya, N.S. et al. (1990) Lemongrass essential oil. Pishch Prom-st. (Moscow).
10, 48.
Banerji, B.K. and Gupta, M.N. (1982) Effects of gamma radiation on Jammu
lemongrass. Tamil Nadu Agric. Univ. J. Coimbatore, India.
Barooah, H.J., Khader, M.A. and Swaminathan K.R. (1988) Use of solar energy for
distillation of palmarosa oil. Ind. Pe$um. 32(1), 22-28).
Beech, D.F. (1977) Growth and oil production of lemongrass (C. citratus) in the Ord
Irrigation Area, Western Australia. Aust. J. Exptl. Agric. andAnim. Husb. 17(4),
301-307.
Boelens, M.H. (1994) Sensory and chemical evaluation of tropical grass oils. Pe$um.
Flav. 19(2), 29-45.
Bommegowda, A. et al. (1 979) Basic studies relevant to agronomic practices in Java
citronella (C. winterianus): Studies on regrowth, floral induction and yield
potential. Mysore J. Agric. Sci. 16, 115-1 17.
130 Chapter 4
Malwatkar, G.M. et al. (1984) Seasonal variation in aldehyde content in oil; leaf
browning and crinkling in Java citronella. Ind. Perjum. 28(1), 17-23.
Manzur-I-Khuda, M. et al. (1984) Essential oils of Cymbopogon species of Bangla-
desh.J. Bang. Acad. Sci. 8(2), 77-80.
Marchetti-Srnaja, J. (199 1) Mise au point sur la composition de l'huile essentielle de
vetyver. LeJ. Nature 3(1), 3-17.
Mathela, C.S. (1990) Himalayan Cymbopogon species. In: Proc. 11th Int. Cong. Ess.
Oils Frag. Fluv. Vol. 4. Aspect Publishing, London, UK, pp. 149-160.
Mathela, C.S. and Pant, A.K. (1988) Production of essential oil from some new
Cymbopogon species. Ind. Perjum. 32(1), 40-50
Mathela, C.S. et al. (1988) Chemosystematics of terpenoids in C. martinii. Biochenz.
System. Ecology 16(2), 167-169.
Metcalf, C.R. (1960) Anatomy of the Monocotyledons, Vol. 1. Gramineae. The Clar-
endon Press, Oxford, UK.
Mucciarelli, M. et al. (1993) Callus induction and plant regeneration in V. ziza-
nioides. Plt. Cell tiss. Organ Cult. 35, (3) 267-272.
Murti, J. and Moosad, K. (1949) Effect of soil type on oil content of vetiver roots.
Amer. Perjum. 54, 113-1 17.
Murty, U.S. and Jamil, K. (1987) Effect of south Indian vetiver oil against immatures
of Culex quinquefasciatus. Int. Pest. Cont. 29(1), 8-9.
Nair, E.V.G., Chinnamma, N.P. and Kumari, R.P. (1979) Review of work done on
vetiver at the Lemongrass Research Station Odakkali. Ind. PerjuuMz. 23(314),
199-20 1.
Nair, E.V.G., Rajan, K.C., Chinnamma, N.P. and Kurian, A. (1982) Studies on
vetiver hybrids. In Proc. Nut. Sem. Medic.and Aromatic Plts. Tamil Nadu
Agricultural University, Coimbatore, India.
Nandi, R.P. and Chatterjee, S.K. (1991) Improved cultivation and distillation
methods, followed by citronella plantations of Darjeeling hills. Ind. Perjum.
35(2), 80-85.
Nanjan, K. Sampathkumar, B. and Jagannathan, R. (1982) Performance of lemon-
grass at Shevaroy Hills. Proc. Nut. Semin. on Medic. and Aromatic Plts. Tamil
Nadu Agricultural University, Coimbatore, India, p. 155.
Naves, Y.N. (1960) The essential oil of gingergrass: C. martinii and C. densiflorus.
P e Ess.
~ Oil. Rec. 51(5), 242-245.
Nethsinga, C. and Paskaranathan, U. (1976a) Lemongrass - a Literature Survey. Pub.
No. 3. CISIR, Colombo, Sri Lanka. (Contains an extensive bibliography.)
Nethsingha, C. and Paskaranathan, U. (1976b) Citronella - A Literature Survey.
Publication on Spices and Essential Oil Plants No.5. CISIR, Colombo, Sri
Lanka. (contains an extensive bibliography.)
Ng, T.T. (1972) Growth performance and production potential of some aromatic
grasses in Sarawak. Trop. Sci. 14(1), 47-58.
Nigam, M.C. (1987) Chemistry of cymbopogons and their essential oils. PufuiJ.
Journal (ApriliJun), 13-20.
Ntezurubanza, L. et al. (1992) Huiles essentielles de geranium et de lemongrass au
Rwanda. Riv. Ital. EPPOS (Num. Spec.), 631-639.
Oliveros-Belardo, L. and Aureus, E. (1977) Essential oil from C. citratus growing
wild in the Philippines. In: Proc. VIZ Int. Cong. Ess. Oils. pp. 166-168. Aspect
Publishers, London.
134 Chapter 4
Singh, R. (1977) Rosha grass farming in Dehra Dun area. In: Cultivation and
Utilization of Aromatic Plants. CSIR, Jammu-Tawi, India.
Singh, R.P., Singh, B. and Singh, V. (1994) Effect of RSC in irrigation water on
citronella Java under different levels of fertilizer. J. Znd. Soc. Soil Sci. 42(1),
164-166.
Singh, R.S. and Pathak, M.G. (1994) Variability in herb yield and volatile con-
stituents of C. jwarancusa cultivars. Zndust. Crops and Prods. (Neds) 2(3),
197-199.
Singh, R.S., Pathak, M.G. and Singh, K.K. (1979) Dynamics and diurnal changes in
oil of Java citronella. Znd. Perjium. 23(2), 116-120.
Singh, R.S., Pathak, M.G. and Bodoli, D.N. (1980a) Response of Java citronella
cultivars to nitrogen under Jorhat conditions. Znd. Perj+ium.24(4), 192-198.
Singh, R.S., Singh, K.K. and Hazarika, J.N. (1980b) Effect of varying levels and
methods of applications of nitrogen to palmarosa oil yield. Chem. Pet. Chem. J.
(India) 10, 1-3.
Singh, R.S. et al. (1983) Citral content of lemongrass under different plant popula-
tion and soil pH. PafaiJ. Jan/March, 33-36.
Smyle, J.W. and Magrath, W.B. (1993) Vetiver grass - a hedge against erosion. In
Technologies for Sustainable Agriculture in the Tropics The World Bank, Wash-
ington DC, USA, pp. 109-122.
Sobti, S.N. (1982) Scope for development of new cultivars of cymbopogons as a
source of terpene chemicals. In: Cultivation and Utilization of Aromatic Plants.
CSIR, Jammu-Tawi, India, pp. 309-3 13.
Sobti, S.N. and Rao, B. (1977) Cultivation and Improvement in Vetiver. Reg. Res.
Lab., Jammu-Tawi, India.
Sobti, S.N. et al. (1982) Breeding a new substitute for palmarosa from Cymbopogon
species. In: Proc. Nut. Sem. Med. Arom. Plts. Tamil Nadu Agricultural Uni-
versity, Coimbatore, India, pp. 166-168.
Soenardi, Marlijunadi and Darmono (1980) Experiments on the time of fertilizer
application to citronella plants. Pemb. LPTZ (Bogor) 36, 21-28.
Srinivas, S.R. (1986) Atlas of Essential Oils. Published by the author, Bronx, New
York, USA.
Surburg, H. (1988) Minor components of palmarosa oil. Developments in Food
Science 18, 505-515.
Torres, R.C., Estrella, R.R.and Gueverra, B.Q.( 1994) Extraction and character-
ization of the essential oil of Philippine C. citratus. Phil. J. Sci. 123(l), 5 1-63;
also 122(3), 269-287.
Tung, Ly. and Balina, F.T. (1993) An account of the introduction of vetiver grass
(I/: zizanioides) to improve indigenous technology for soil and water conserva-
tion. Contour 5(1), 4-7.
Verma, S., Trivedi, K.C. and Gupta, R.S. (1981) Identification of two morpho-
logically distinct plant types of palmarosa. Znd. Perjium. 25(1), 87-91.
Vimilan, A.K., Srivastava, R. and Datta, S.G. (1990) Profiles in Indian essential oils.
In: Proc. 11th Znt. Cong. Ess. Oils Flav. & Frag. Vol 4, Aspect Publishing,
London, UK, pp. 181-188.
Virmani, O.P., Gulati, B.C. and Datta, S.C. (1967) Production of oil of palmarosa.
P e Ess.~ Oil. Rec. 5 8 ( 5 ) , 285-294.
Gramineae 137
Virmani, O.P. andDatta, S.C. (1971) Essential oil of C. winterianus. The FlavourInd.
2(10), 595-602 and 2(12), 710-712. (Contains a major bibliography.)
Virmani, O.P. et al. (1979) Oil of lemongrass: 1 -East Indian, 2 -West Indian. World
Crops 31(2), 72-74 and 31(2), 120-121. (Includes a very extensive bibliogra-
phy.)
Watt, G. (1908) The Commercial Products of India. John Murray, London, UK.
Whyte, R.O. (1968) Grasslands ofthe Monsoon. Faber and Faber, London, UK.
Wijesekera, R.O.B., Jayewardene, A.L. and Fonseka, B.D. (1973) Varietal differ-
ences in the constituents of citronella oil. Phytochem. 12, 2697-2704.
Xabregas, J. (1968) Una nova cultura planaltica do vetiver. Agron. Angolana. 23,
19-23.
Zeehuisen, J.J. (1952) The influence of oxygen and metallic oxides on citronella oil.
3’.Sci. Res. Indonesia. 1, 2 16-2 19.
Zheng, G.Q., Kenney, P.M. and Lamm, L.K.T. (1993) Potential anti-carcinogenic
natural products isolated from lemongrass oil and galanga root oil. 3. A g . Food
Chenz. 41(2), 153-156
Zhao, Q. (1993) China’s perfumery industry picks up. Perjum. Flav. 18(1), 47-48,
and 16(5), 51- 52
Zhixi, H.and Hua, L.J. (1988) Oil of C. winterianus. In: Proc. Inter. Con5 Ess. Oils
Flav. Frag. Cosmetics, Vol. 1. Beijing, China, p. 13.
Larniaceae
5-
The Lamiaceae (syn. Labiateae) to which the genus Pogostemon belongs, is
one of the ten largest families of flowering plants, comprising 9 subfamilies,
200 genera and about 3200 species. The family is readily identified, the
flowers having distinct and widely gaping lips, a four-part ovary, opposite
aromatic leaves and angular petioles; it is noteworthy for the number of
species producing an essential oil, a spice or both.
Patchouli
The genus Pogostemon contains 40 species mainly native to South East Asia,
with 20 occurring in India. The common European name patchouli is
probably derived from the Tamil paccilai and the species name from the
Philippine vernacular cablin. Pogostemon cablin is the prime source of patch-
ouli oil but several other species, P. comosum, P. hortensis, P. heyneanus and
P. plectranthoides are cultivated for their oil, also described as patchouli oil,
but of inferior and often variable quality. Essential oil from plants in other
genera, Microtaena cymosa Prain. (Chinese patchouli), Plectranthus patchouli
Clark., is also sold as patchouli oil, and in the seventeenth century Rum-
phius described a plant he named Melissa lotoria with a similar scent (Herb.
Amboin 5, p. 292). Since patchouli has been widely disseminated in Asiatic
and Far Eastern countries, its original home is still in doubt. Many of the so-
called indigenous species are probably long established and acclimatized
local cultivars of P. cablin. The South East Asia-China border region is
probably patchouli’s centre of origin, and the Philippines a modern centre of
domestication and dispersal. Regional names for patchouli are very similar:
nilam wangi (Indonesia), dilem wangi (Java), nilam (Atjeh) singalon (Batak),
dhalum wangi or tilam wangi (Malaysia).
Patchouli has a long history of use by man, and plants of the genus
Pogostemon producing oil very similar to the present patchouli oil were widely
used in Asia and the Far East for incense, body and garment perfumes,
138
Lamiaceae 139
insect repellents and sprinkled in temples (Morris, 1984). The plant was
known as huo-hsiang in ancient China, and an early Chinese reference to
propagation by suckers may indicate that true patchouli was cultivated there
some 2000 years ago (Laufer, 1919). Arabs used it to perfume carpets,
Indians textiles, and Chinese produced a perfumed ink for use on scrolls.
Patchouli was known as malabathron in classical Europe. The plant was first
described as patchouli by a Mr Wallich in 1835 who obtained it from Penang
(Burkhill, 1966), and later by a Frenchman Pelletier-Soutelet, who in 1845
named it P. patchouli. In 1896, Holmes identified the plant as P. cablin
(Benth), native to the Philippines (Pharmazie 561, 222). The scent of
patchouli had, however, been known in Europe for some time as the
splendid fabrics and garments from India were permeated with its exotic
odour. The first shipment of dried patchouli leaves arrived in London in
1844.
Growing patchouli for its essential oil began in the nineteenth century in
Malaysia using plants obtained from the Philippines; from Penang it was
introduced to Java in 1895. Patchouli cultivation quickly spread to other
Indonesian islands including Sumatra in 1910; by 1920 it was well estab-
lished in western Atjeh and northern Sidikalang highlands, but later became
more important on the island of Nias. The area under patchouli in Indonesia
has risen from 11,000 ha in 1983 to about 18,000 in 1995, with total oil
production officially estimated at nearly 1000 tonnes; exports during the
period 1985 to 1995 varied from 350 to 850 tonnes.
Patchouli was extensively grown in the wetter regions of southern China
in the eleventh century, by the sixteenth century had become centred on
Hainan Island, but when oil production commenced in China is uncertain.
In the last decade, annual oil production slowly decreased from about 10 t to
2 t in 1990, and recent reports indicate this level is unlikely to rise as the area
under cultivation has decreased in favour of other essential oil crops.
Patchouli has been successfully grown in Mauritius, some Caribbean
countries, West Africa and Vietnam, but only India, Brazil, Taiwan and
Seychelles established commercial oil production. In Russia, patchouli oil is
obtained from P. plectranthoides, but little information is available. Oil of this
species is used to provide medicinal products in India, where it is known as
P. benghalensis, and the species occurs naturally throughout the country. Use
of patchouli without qualification in the text refers to P. cablin and its
products.
Botany
The classification of P. cablin is complex; it is currently described as P. cablin
Benth. (syn. P. patchouli Pellet, P. patchouli var. sauves Hook, P. tomentosus
(Hassk.); or P. patchouli Bell. (syn. P. cablin (Blanco) Benth., P. heyneanus
Benth., P. sauves M. and Briq., and Mentha cablin Blanco). Pogostemon
140 Chapter 5
P. hortensis leaves were frequently mixed with patchouli leaves when ex-
ported and for this reason leaves of Malaysian origin were often rejected by
European buyers; dried leaves yield about 2% oil (Chan, 1969).
The main constituents of patchouli oil are the patchouli alcohols 23-55%
(averaging 33%) and minor alcohols 0.4-0.6% ( mainly norpatchoulenol).
There is little information on oil biosynthesis and the mechanisms involved
although certain aspects have been irivestigated (Croteau et al., 1987;
Weyerstahl et al., 1989). Thus vital data on which to base crop improvement
and plantation management programmes are lacking. For all practical
purposes oil is contained only in the leaves, and factors directly affecting this
plant part are of greatest interest to producers. Fertilizers affect growth,
number and also oil content of leaves.
The flowers are small, white, in dense interrupted whorls on short spikes,
terminally and in axils of the uppermost leaves. They are bisexual and
hypogynous, with more or less purple/mauve streaking. The stamens have
long, pale mauve filaments which give the genus its name. The calyx
segments are united into a tube 5-7 mm long, with four to five unequal teeth
closely appressed to the corolla tube, divided into two upper lips, each with
two lobes, the upper lip being longer. The androecium has four exerted
stamens, a long style and two stigmas arising from centre of the ovary, which
consists of two carpels deeply indented into four lobes, each lobe containing
a single ovule. The fruit contains four smooth, ovoid nutlets. Some regional
cultivars seldom or never flower, especially those in Malaysia and Indonesia,
since vegetative propagation has preserved a characteristic inherent in the
original introductions.
Pogostemon heyneanus (Benth.) is probably native to the Indo-Malaysian
area and known in India as pachaput; in Malaysia as nilam bukit, in Indonesia
as dilam kembang, and grows wild in all. Pogostemon heyneanus is also a highly
aromatic herb with stems to 1 m. The leaves are ovate, acuminate or obtuse,
crenate, usually double-toothed, thinner than patchouli, on long petioles.
The species is widely cultivated on a minor scale or in gardens for its fragrant
leaves, which are dried and stored with clothing to repel insects. The oil
content of dried leaves is 0.05-0.25% in India, but up to 0.75% has been
reported. Pogostemon heyneanus flowers and produces viable seed. The
flowers are white tinged with purple, in terminal and axillary spikes, the
whorls usually separate on panicles.
Oil distilled on Java was much inferior to patchouli oil and its production
abandoned; the oil is occasionally produced in India for local consumption,
and elsewhere is used as an adulterant or extender of patchouli oil. A sample
from Bangalore was a greenish brown somewhat viscous liquid, with the
following characteristics: specific gravity (1 5°C) 0.9365; refractive index
(20°C) 1.4986; optical rotation (20°C) -18"; acid value 4.35, acid value
after acetylation 92.80; ester value 6.27; soluble in 0.5-10 vol. 90% alco-
hol.
142 Chapter 5
Ecology
Patchouli flourishes in areas with high average temperature and high relative
humidity, but can be successfully grown wherever similar conditions can be
achieved by mulching, partial shade, or ridge-planting in marshy areas.
These techniques may affect rate of growth or leaf-oil content, but a high
standard of management can normally compensate and produce a profitable
oil yield. An annual rainfall of 2000-3000 mm well distributed throughout
the year is the optimum, 1750-2000 mm acceptable, but it is unlikely that
patchouli will be commercially profitable below 1500 mm without irriga-
tion. Although patchouli has a high water requirement, seedlings and young
plants cannot tolerate waterlogging, and 3-4 days of standing water can be
fatal. Relative humidity should be high and a minimum monthly average of
75% is considered necessary for sustained growth in Indonesia (Soepadyo
and Tong, 1968).
Bright sunny days are necessary for maximum leaf-oil content, although
plants tolerate partial or intermittent shade. The effect of shade is sub-
stantial; unshaded plants in Indonesia had an oil content of 5 6 % compared
to 4.0-4.7% in shaded plants. A comparison of leaf and oil yield from
partially shaded and non-shaded plants in Malaysia showed that shaded
plants gave a slightly higher herbage yield, but lacked vigour, regenerated
more slowly after cutting and leaves, although longer, were thinner and had
lower oil content (Chin and Ahmad, 1970). When grown under no shade,
half shade and full shade in India, leaf yield fell linearly from 345 to 254 g per
plant and oil yield from 3.0 to 2.5% (Nanjan et al., 1982). The adverse effect
of shade on oil content resulted in patchouli being recommended as an
intercrop only in young rubber or oil-palm plantations in Indonesia and
Malaysia. Shade can however be beneficial where the environment is less
suitable for patchouli cultivation, and cultivars may also react differently
lamiaceae 143
when grown outside their normal habitat (Jimenez et al., 1990; Radhakrish-
nan et al., 1991).
For maximum growth an average daily temperature of 25530°C is desir-
able, but there are no published data comparing plant growth at different
temperatures. It is widely considered that patchouli grows better at higher
altitudes, but this could be due to better drainage on slopes, since lower soils
generally have impeded drainage. Wind can cause severe damage to mature
patchouli in open fields but, as most patchouli is grown by smallholders in
small areas interspersed with taller crops, wind damage is usually minimal.
Cultivation
Land selected for patchouli cultivation is often cleared to a greater extent
than for other crops, with most large trees felled, burnt and the ashes spread
as fertilizer. When grown as an intercrop in plantations, patchouli benefits
from the generally higher standard of management, and has been cultivated
successfully between young rubber and oil palms in Indonesia and Malaysia,
young and mature citrus in Malagasy and Malaysia, coffee in Malagasy and
Brazil, pepper vines in several South East Asian countries and under
coconut palms in Seychelles and Mauritius.
Patchouli is normally propagated by cuttings, preferably from the central
section of semi-mature stems (Fig. 5.1). In Indonesia, cuttings from
9-month-old stems gave the highest strike rate, as did cuttings with a
carbon-nitrogen ratio of 50 (Soepadyo and Tong, 1968). Cuttings with a
high level of total and ortho-dihydroxy phenols rooted more quickly in India
Bhattacharjee and Thimmappa (1 991), and treating cuttings with IBA or
NAA also significantly increased rooting (Selvarajan and Rao, 1982). Cut-
tings 5-10 cm long are often planted in shaded nursery beds and are
normally ready for transplanting in 3-5 weeks. When propagating material is
limited or for multiplication of selected elite plants, 2.5-5.0 cm of stem
including a leaf node can be used provided the node is completely buried in
the rooting medium. Such cuttings require careful handling as the survival
Lamiaceae 145
Spacing 60 cm row Plant height (cm) Branches per plant Herbage (kgiha)
~ _ _ _ _
30 cm 60 8 2178
45 cm 72 8 1650
60 cm 77 5 1230
75 cm 65 6 995
Source lrulappan and Ponnuswami (1982b)
The method was effective, quick and mortality low. Distance between
furrows can be varied to suit other estate machinery.
Spacing is often arbitrary, but 60-90 X 30-45 cm between plants or 60 X
60 cm on the square is most common, with 100 X 100 cm on very fertile
soils. A high plant population is generally to be preferred and where row
width is determined by existing machinery, in-row spacing must be adjusted
to compensate. Close spacing suppresses weed growth, tends to increase
branching, and as oil content is highest in upper leaves, this increases oil
yield per hectare. The effect of various in-row spacings with a standard row
width is shown in Table 5.1. As a guide, 4000 standard cuttings for planting
in a nursery weigh approximately 100 kg or a total of 1000 kg/ha at a spacing
of 30 X 60 cm. Cuttings or seedlings should be planted at the beginning of
the rainy season when there is usually cloud to minimize sun-scorch; in
Sumatra this is August-December and in Malaysia March-April. On land
susceptible to temporary flooding or waterlogging, cuttings can be sown on
large ridges; on sloping land or hillsides, rows or ridges should follow the
contour to minimize soil erosion.
Patchouli is normally grown in small plots and weed control is usually
manual, carried out as necessary until the canopy closes in about 6 months
and suppresses most weeds. Following the first and subsequent harvests,
two to three weedings are sufficient. Herbicides are seldom applied, but
diuron, linuron, and simazine have been successfully used pre-emergence in
established patchouli. When patchouli is grown as an intercrop in young
rubber or oil palms, herbicides may be essential for general weed control,
and glyphosate, oxadiazon, prometryn and terbacil have been used as
directed sprays.
Little information is available on the water requirements of patchouli, but
it is doubtful if patchouli would be profitable as a wholly irrigated crop, as
total water requirement is high and the cost of supplying water is constantly
increasing. In regions where monsoon rains may be interspersed with a dry
period, mulching can successfully keep soil moist and maintain plant
growth. Underplanting in plantations with high canopies such as coconut
palms which provide intermittent shade also assists in reducing moisture loss
(Seychelles Department of Agriculture, 1970).
Lamiaceae 147
Harvesting
Patchouli is ready for harvesting 4-7 months after establishment when
plants are usually about 1 m in height and the foliage becomes pale green to
light brown, and thereafter at 3-6 monthly intervals over its productive life.
Small farmers normally cut plants 10-20 cm above the soil, but where only
one annual harvest is possible, as in Russia, plants are cut as low as possible.
When using machinery, the optimum cutting height should be established,
since cutting too low can kill plants especially if followed by a dry period.
Harvesting should not take place following rain when leaves are saturated
nor early in the morning if wet with dew. During hot or dry periods,
however, cutting in the morning and evening is preferable.
Harvesting includes stems and leaves and may amount to 5-10 t exception-
ally 15-20 t/ha in the first year, falling to 1-5 t/ha in subsequent years with
leaves usually accounting for 50-60% of the total. Smallholder yields from new
plantings are generally 1-3 t h a of green leaves in the first year, falling to 1-2 ti
ha in subsequent years, mainly due to poor management and use of low
yielding plants. Under estate conditions in Johore, Malaysia, average yield of
air-dried leaves was 1500 kgiha from the first cut and 900 kgiha from
subsequent cuts. Using current techniques, 2000-3000 kgiha dried leaves is
acceptable, but local trials indicate twice this is possible and the full potential of
patchouli has yet to be determined. A major factor reducing average leaf yield
is the general use of unselected planting material, as there is substantial
variation in leaf yield between individual plants.
All plant parts contain oil, but as oil in leaves is the main source, ideally
only leaves should be harvested. It was initially recommended in Indonesia
that only the three to five uppermost pairs of leaves should be cut as soon as
these had fully developed. Such selective harvesting is still possible for
smallholders; stems with three to five pairs of mature leaves which have not
turned yellow or brown are cut, while those with young leaves are left to
mature. This system not only promotes rapid regrowth, it also ensures more
efficient working of the usually small local stills. Fresh herbage has a
moisture content of 80-85% and oil content 0.5-1.2%. Thus 1000 kg of
fresh material from smallholder crops in Indonesia yields 150 kg dried leaves
and 1-2 kg oil, in Malaysia the average is 200 kg of dried leaves and 2-3 kg
oil, and in India about 200 kg dried leaves yields 3-4 kg oil. Also in India
8000 kg of fresh leaves yielded 1600 kg shade dried leaves, then 40 kg oil
(Kumar et al., 1986)
Under smallholder management the normal life of a patchouli planting is
2-3 years as fertilizers are seldom used, growers relying on natural soil
fertility. Thus patchouli was seldom replanted on the same land, and the plot
abandoned in favour of a newly cleared area. With rapid destruction of
mature forest in patchouli-growing regions, particularly Indonesia, the
amount of new land available is continually being reduced. This has resulted
in a reduction in smallholder-grown patchouli, but if patchouli is properly
148 Chapter 5
Distillation
Unlike other members of the Lamiaceae, fresh patchouli leaves should not
be distilled as they yield only a small proportion of their oil unless distilled
under pressure or with superheated steam. Both methods greatly increase
distilling cost, and may alter oil odour and characteristics. Normal distilla-
tion apparently does not rupture cells within the leaf, but drying or light
fermentation changes the cell structure which becomes more permeable and
allows oil to be liberated.
Stems with leaves are dried for 2-5 days and how this is accomplished is
extremely important since it basically determines oil yield when locally
distilled, or quality of dried leaves if these are to be distilled elsewhere. Fresh
herbage must not be piled in heaps as this quickly encourages mould growth,
or uncontrolled fermentation. Cut material is usually dried by spreading on
wooden racks, clean, hard-packed earth, or specially constructed concrete
floors adjacent to a distillery, and regularly turned to ensure uniform drying.
Herbage must also be protected from rain and heavy dew, and locally woven
large grass mats are often used. Although sun-drying is normal, shade-
drying with air circulating freely over cut herbage is preferable. Sun-dried
herbage may also become over-dry, with consequent loss of oil or leaves
through shattering, while a drying temperature above 40°C in Malaysia
resulted in 80% oil loss (Chin and Ahmad, 1970).
Following drying, leaves are stripped from stems and placed in woven
baskets containing about 15 kg to allow fermentation, which a skilled grower
controls by smelling the leaves. Over-fermentation produces a mouldy note
in the oil, while under-fermentation reduces oil yield but has no effect on
quality. In Malaysia, leaves are not deliberately fermented but packed in
sacks for transport to a distillery, and this delay allows some fermentation to
take place. Dried leaves can be stored for long periods provided they are kept
free of moisture and no additional fermentation occurs. Storage facilities at
rural stills are usually inadequate, and in these circumstances it is preferable
temporarily to cease harvesting. A yearly minimum of 40-60 t of dried leaves
is required to establish a small district distillery; in Malaysia 20-25 t dried
leaves are required monthly for an estate distillery to be viable.
There was previously an extensive trade in dried leaves from Indonesia to
Singapore and Europe, where high quality oil was obtained, but this trade
has virtually ceased due to the high transport cost. Leaves for export were
dried, compressed, and baled with little delay; bales store well provided they
are kept in clean, dry buildings. French buyers required leaves to be shade
dried for about 14 days, free of fermented and mouldy material, and with a
minimum 3% oil content. Patchouli leaves are frequently adulterated with
Lamiaceae 149
Young plants and seedlings are frequently attacked by various mole and
other crickets, particularly Gryllidae, while grasshoppers are easily con-
trolled in smallholder crops by hand picking, equally effective against giant
snails, Achatinafulica, which can be a local pest. Larvae of various moths, the
leaf-rolling caterpillars, are more serious pests and difficult to control.
Pachyzanela spp. are most frequently recorded, particularly P. staltalis in
Indonesia. Unidentified stem borers attacked plants in India and Malaysia,
and a gall-producing weevil was reported from India which caused decay of
stems above the gall. Mites can damage individual plants but have not been
reported as a serious pest, although they are often present.
As with pests, there are only a few reports on patchouli diseases. That
some are of considerable economic importance, and may in fact limit
production in certain areas, can be assumed from remarks contained in
general reports on the crop. Most often mentioned are root diseases caused
by Sclerotium spp. in Indonesia and Fomes spp. in Malaysia, where F. lignosus
has been identified on patchouli, and becomes a problem when the crop is
interplanted in rubber plantations. Visual symptoms of F. lignosus infection
are change in leaf colour from green to reddish purple, followed by wilting
and finally defoliation. In East Africa, P. cablin was so severely attacked by
Pythium spp. leaf spots that the author was unable to maintain demonstra-
tion plots, whereas P. heyneanus was relatively unaffected and made good
growth. Cercospora spp. and Alternaria spp. cause extensive leaf damage, and
in India a severe attack ofA. alternata causes defoliation and death of plants,
while Cercospora leaf blight also causes severe damage unless controlled.
Symptoms attributed to a virus have been reported from several countries,
identified in India as patchouli mosaic virus (PMV) but in Brazil as tobacco
mosaic virus (TMV). Bemisia tabaci, a whitefly, is the insect vector in India,
and mass propagation of virus-free seedlings for distribution to growers the
only satisfactory method of control (Kukreja et al., 1990).
freshening preparations. Flavour and odour are quite distinctive, and patch-
ouli flavoured sweets are especially popular in Asia and Arabia. The oil is
used to flavour foods, mainly processed products such as baked goods, some
meats and sausages, but very seldom as a culinary spice although frequently
present in speciality cooking oils. The reduction in tar content of cigarettes
and tobaccos has affected their perceived flavour, and patchouli oil is one of
the additives used to offset this effect; the Chinese cigarette industry is the
major user of locally produced oil. The whole oil has antimicrobial and
insecticidal activity, but is normally too expensive for these uses to be
commercially exploited. These properties are, however, well appreciated in
Asian countries where the whole oil is widely used in local medicines and as
an insect repellent. In Japan and Malaysia it is considered (erroneously) to
be an antidote to snakebite.
Patchouli oil is extremely complex, is in demand as a whole oil, and is a
very important perfumery raw material. Some 60 constituents have been
detected (Akhila and Nigam, 1984; Weyerstahl et al., 1989; Dung et al.,
1990), including sesquiterpene hydrocarbons and patchouli alcohols
(patchoulol and norpatchoulenol) (Munck and Croteau, 1990). Azulene
has also been reported from the oil. Some important odour compounds and
related synthetics have been described (Weyerstahl et al., 1989). Among
cyclohexenones found in the oil, isophorone, trimethyl cyclohexenedione
and tetramethyl cyclohexenone all have a camphoraceous odour; a carbonyl
compound, beta-patchounone, possesses a strong woody-patchouli odour,
whilst alpha-natchoulone has a strong ambergris odour.
Significant regional variation occurs, with Indonesian (East Indian) oil
considered the best quality, then Seychelles oil, and no other origins ac-
cepted as substitutes. Vietnamese and Chinese patchouli oil differs markedly
from other oils, with patchouli alcohols in Chinese oil usually below 25%
and the oil has a less sweet, slightly harsher odour. Oil is normally sold on a
sample basis, but Chinese oil may be sold in grades based on solubility i.e.
1:7-9, 1:11-12, 1:13-14, the second being most popular. Patchouli oil can
be adulterated with oils having a similar basic odour, but most other
adulterants are detected on analysis, or by the formulator’s nose. In In-
donesia, a common adulterant is a balsam resinoid, although this is officially
banned. The main characteristics of patchouli oil are shown in Table 5.2.;
values for other oils have been published (Sugimura et al., 1990).
Patchouli concrete or resinoid is obtained by solvent extracting dried
leaves, and is a very viscous liquid whose colour depends on the solvent
used; benzene produces a dark, orangy-brown, petroleum ether pale orange
to dark amber. The odour is finer than that of distilled oils, very sweet,
aromatic, rich and spicy, and improves with age. Benzene extraction in India
gives an average resinoid yield of 5% containing 70-80% of alcohol-soluble
absolute. Concrete is mainly used as a direct substitute for patchouli oil.
152 Chapter 5
References
A bibliography of patchouli is contained in Medicinal and Aromatic Plant
Abstracts 10(6), 595-599, published by MAPIS, Technical and Information
Services, Hillside Road, New Delhi, India.
Akhila, A. and Nigam, M.C. (1984) GC-MS analysis of the essential oil of patch-
ouly. Fitoterapiu. 55(6), 363-365.
Bhattacharjee, S.K. and Thimmappa, D.K. (199 1) Studies on the growth hormone,
length of cuttings and number of leaves on root formation of P. patchouli. Ind.
Perjbn. 35(2), 71-76.
Burkhill, I.H. (1 966) A Dictionay of the Economic Products of the Malay Peninsula. 2
vols. Government Printer, Kuala Lumpur, Malaysia.
Chan, S.K. (1969) Recent investigation on short-term cash crops in Serdang. In:
Progress in Oil Palms. pp. 265-286. Incorporated Society of Planters, Kuala
Lumpur, Malaysia.
Chin, H.F. and Ahrnad, R. (1970) Crop Diversz'jication in Malaysia. Incorporated
Society of Planters, Kuala Lumpur, Malaysia, pp. 107-1 17.
lamiaceae 153
155
756 Chapter 6
tropical fuel and shade tree, while cassia oil is derived from Acacia famesi-
ana.
The botanic origin of commercial cinnamon and cassia products is as
follows: True, Ceylon and Seychelles cinnamon from Cinnamomum verum;
Chinese, Canton, Kwantung, Kwangsi, Yunnan, Honan and China Junk
cassia from C. cassia; Saigon, Vietnam, Danang, Annan and Tonkin cassia
from C. loureirii; Indonesian (or cassia vera), Padang, Batavia, Korintji,
Koringtoji, Java and Macassar cassia from C. burmannii; Indian cassia may
be derived from C. obtusifolium, C.tamala or C. sintok. Many of these
designations are (deservedly) falling into disuse, and most of little commer-
cial importance. Chinese cassia bark from main producing areas in Kwangsi
and Kwantung provinces is now officially sold as Tung Sing and Si Chang.
Philippine cinnamon is obtained from C. iners, C. mindanaense and C.
mercadoi and consumed domestically. The last named is said to produce the
highest quality bark (Baja-Lapis, 1979).
Camphor
Camphor oil is usually obtained by steam distilling leaves, wood and roots of
Cinnamomum camphora (including putative subspecies), but there are other
camphor oils not obtained from C. camphora, including Ocimum kilimand-
scharicum, Labiatae, described later. Additionally various subspecies
(chemotypes) of C. camphora produce oils which differ substantially in their
major components. The Japanese C. camphora, known as Hon-sho, is con-
sidered to be the true camphor tree and native to Japan, with C. camphora
occurring in other countries designated a subspecies. The validity of this
nomenclature remains in dispute. Cinnamomum camphora and proposed
subspecies have been further divided into varieties on the basis of main oil
constituent, i.e. cineole, safrole, linalool or terpineol. A major argument
against division into subspecies is that seed from a particular tree may
produce seedlings whose leaves yield oil in which any of the four main
constituents predominates. The proposed subspecies are botanically similar
to C. camphora, and will not be described separately except in relation to
their essential oil. Differentiation on the basis of oil composition is probably
the most valid method of describing what are chemotypes of C. camphora
and not true subspecies. In this section, use of camphor without qualifica-
tion refers to C. camphora and its products.
Europeans became aware of camphor via Arab traders and one of the
earliest descriptions is in Adams (Paulus Aegineta, iii, 427-429), who stated
that Sarapion, a Greek governor of the Tebaid area of the Red Sea coast
about 150 BC, gave an account of refining camphor he obtained from an
Arab, Ibn Amram, while Symeon Seth was the only Greek author to write of
camphor itself (caphura). The Arab merchant Sinbad in the ninth century
lauraceae 157
Botany
Cinnamomum camphora (L.) Nees & Eberm. (syn. C. oficinalis Nees; Laurus
camphora L.) is commonly known as camphor and its many similar names in
European languages are probably derived from the Javanese kapur. Many
Asian names are also very similar; in India the Sanskrit karpura and Hindi
kafur suggest a single original source of the oil when it became known
outside the Far East. In Japan and Taiwan, Hon-sho means true camphor
(oil) to distinguish it from other oil types, although Taiwanese know the C.
camphora tree as kusunoki.
Cinnamomum camphora is a medium to large evergreen tree to 30 m, with
a dense, wide spreading head, large lustrous green leaves and yellow flowers.
The leaves and twigs, when crushed, smell strongly of camphor. The roots
are usually extensive, but not generally penetrating, although the linalool
form is considered in Japan to be deep-rooted. When felled to provide
camphor, large lateral roots are also excavated for distillation. Roots gen-
erally contain the highest content of essential oil, to 10%, with a camphor
content to 27%. There is normally a single stem to 30-50 m in the wild,
often enlarged at the base, with a number of large branches, the lowest at
some height on the trunk. The bark is rough, often with small fissures from
which oil exudes and solidifies, and collecting this material probably pro-
vided the original source of camphor with deliberate slitting of bark to
increase the flow the first method of harvesting. The wood is yellowish
158 Chapter 6
brown, moderately hard, fine textured, takes a good polish, and has a
persistent odour of camphor. The timber is extensively used in cabinet work,
especially chests. In a camphor tree aged 30-40 years the trunk contains
about half the total tree weight, roots 20%, stump 14%, branches 13% and
leaves 5%. The largest camphor tree on Taiwan was 51 m tall, 5.3 m DBH,
with the first branch at 20 m. Total volume was calculated as 588 m3, to
produce 7200 kg oil and 2743 kg crude camphor. The tree was estimated to
be 1400 years old. The largest tree in Japan is said to be 1200 years old with
a trunk length of 31 m and 2.3 m DBH.
The leaves are on long petioles and are leathery, alternate, ovate-elliptic,
acuminate, with one main rib and usually two side ribs; 5.0-12.5 x 1.5-5 cm
but variable; the upper side is lustrous green, the lower lighter and glaucous.
Young leaves are pinkish to light red, the colour varying between the
different forms. Although considered an evergreen, a large proportion of
mature leaves are shed annually, usually coinciding with the flush of new
growth. Leaves contain an essential oil, 0.2-2.0%, camphor content 0-80%.
Young leaves generally have a higher camphor content than those more
mature, and shaded leaves of any age less camphor than those exposed to
direct light. In India, leaves from the Cochin region average the highest
camphor content, although air-dried leaves from Darjeeling had a camphor
content of almost 2%. There are four main forms of C. camphora as noted,
each yielding a leaf oil with a different major constituent; (linalool type) 80%
linalool and 10% monoterpenes; (cineole type) 76% cineole with alpha-
pinene and 1-alpha-terpineol 20%; (sesquiterpene type) 4 0 4 0 % nerolidol
and 20% each sesquiterpenoids and monoterpenes; (safrole type) 80%
safrole and 10% monoterpenoids (Shi et al., 1989; Hirota and Hiroi, 1967).
A quick method of establishing the oil type is to crush fresh leaves, which will
smell strongly of either camphor, turpentine or eucalyptus.
The inflorescence is a short, dense, axillary panicle on the ends of the
twigs, and seldom exceeds the leaves in length; the buds are scaly (Fig. 6.1).
The flowers are small and yellow or yellowish white; the perianth is mem-
branaceous. The main flowering period in India is May-June, and in Taiwan
and Japan, April-May. The fruit is an ovoid drupe to 9 mm diameter, dark
green becoming blackish when ripe, containing one globose seed. In India,
most fruit normally ripen in October, in Taiwan and Japan, in September.
Seed from the first crop of fruits in India is frequently infertile and, like
cinnamon, ripe seeds quickly lose their viability. Seeds contain nearly 50%
of a yellowish-white crystalline aromatic fat with a high lauric content, a
melting-point of 21-23"C, used for soap making in India. Fat characteristics
are (all at 25°C): specific gravity 0.925; refractive index 1.4442; acid value
0.6; saponified value 272.3; iodine value 4.0 and unsaponified matter 0.7%
(Shastri, 1978).
The essential oil is the most important product, thus how oil is produced
in the tree, its distribution and the physiological changes which can occur are
Lauraceae 159
Fig. 6.1.Cinnamonum camphora. A, Flowering shoot; B, flower in longitudinal section and reproductive
organs; C, fruit. (Purseglove et al., 1981.)
160 Chapter 6
Ecology
Cinnarnornurn carnphora is native to China, Japan and Taiwan, but has been
introduced into many other countries, and cultivated in some with little
commercial success. The various forms of C. carnphora previously noted can
occur in any region, but in some a particular type naturally predominates,
the Yu-sho in China, the Hon-sho in Taiwan and Japan; elsewhere the
dominant type depended on the introduced parent stock. Camphor is found
in many botanic gardens well outside its natural range, as an avenue tree and
in public gardens in many countries, the southern USA and India for
example. Analyses of wood and leaves of such trees usually show a low
essential oil content and camphor may be absent. Oil from leaves and twigs
from a specimen growing in the Entebbe (Uganda) botanic garden yielded
1.8% camphor, but on Mauritius camphor content was negligible.
Camphor flourishes in warm-temperate to semi-tropical regions, and
although mature trees can withstand frost, seedlings or young trees fre-
quently succumb. The degree of growth inhibitor activity in leaves is a factor
increasing frost resistance of Cinnarnornurn spp. in Russia (Bregvadze et al.,
1975); thus selection for a high degree of this activity to increase frost
resistance is possible. A minimum temperature of 15°C is considered
desirable, but no maximum recorded. Camphor is basically a forest or
woodland tree, but highest camphor content is found in wood and leaves
from trees growing in direct light. Shaded trees, or shaded leaves from a
specific tree, usually have a lower oil and camphor content.
A rainfall of 1000-3500 mm occurs in its natural habitat, but above 3500
mm the soil must be free-draining as camphor is relatively shallow rooted
and intolerant of waterlogging. In higher rainfall areas it should be planted
on hillsides. In areas with very high rainfall the camphor content of oil is
reduced, and in East Africa there was a reduction in both oil and camphor
content following prolonged rain. Camphor grows naturally from sea-level
to nearly 3000 m, but below 1000 m is considered the optimum, especially
in Japan where leaf oil is the most important product. In India, there was a
substantial difference in oil content in leaves and wood from trees growing at
sea-level near Calcutta, 1.24% with 56% camphor, and near Darjeeling at
2200 m with 2.9 and 80% camphor (Choudhury, 1958). The latter is a very
high yield and equalled 1.6% of the air-dried material, excluding that
remaining dissolved in the oil and recovered by fractionation.
accumulates are not suitable. Soil type can affect both essential oil and its
camphor content and, in Taiwan and Japan, trees growing on sandy soils
had a higher content of both compared to trees in the same area on soil with
a high clay content. A neutral to slightly alkaline soils is preferred for
plantations, p H 6.5-8.0, although there are no data comparing essential oil
and camphor production from soils with differing p H values.
When camphor is grown in plantations, no fertilizers are normally applied
after seedlings are planted out. When grown to provide leaves for distilling
and coppiced to promote bushy growth, fertilizers are necessary to ensure
high foliage production; however, little fertilizer is normally applied to
plantations other than plant residues. When stills are located close to
plantations, spent leaves may be returned as mulch. Seedbeds are normally
well fertilized and this is discussed in the section on cultivation.
Cultivation
Prior to the First World War, a high proportion of camphor traded interna-
tionally was obtained by felling wild trees principally on Taiwan (then
Formosa), with smaller production in Japan. Chinese production was al-
most wholly consumed domestically. Wild trees were thus a forest resource
and harvested in a similar manner to normal logging, with no special
protection or management. It was not until the decline in number of wild
trees threatened the industry’s future that plantations of camphor trees were
established on Taiwan, and after the Second World War in Japan. The
establishment of plantations required a more intensive management system,
but no special requirements other than those required in normal forestry.
Seedlings are planted out, protected and weeds controlled until the young
trees are well established at between 4-7 years. In Japan, plantations are
individually owned, and labour input depends on the inclination or skill of
the owner. This was usually reflected in a faster rate of tree growth or lower
seedling mortality. With the introduction of synthetic camphor, establish-
ment of plantations mainly for camphor production became uneconomic,
and leaf oil from other forms of C. cumphoru became more important.
Production of linalool type oil is paramount in Japan, but in India and Sri
Lanka the camphor content remains most important.
Camphor is normally grown from seed, the ripe fruits collected and seed
cleaned of pulp and dried. Seed is normally sown within a few months of
collection, and viability after 6 months storage is usually low, but in regions
of China with a severe winter, seeds are kept in cool storage and sown in
spring. Decorticating and soaking seeds prior to sowing increased and
hastened germination (Wang and Shao, 1984). Seeds are sown in prepared
seedbeds, and the seedlings are normally 12-14 months old at planting out.
When ready for transplanting into prepared holes, seedlings are cut back to
5-10 cm; in India the roots are also pruned. Seed from a particular form of
C. cumphoru tree produces seedlings whose leaf oil may be any of the major
162 Chapter 6
types. Where a specific oil type is needed, linalool in Japan for example, only
seedlings which produce this oil are required, and to determine which are to
be retained the leaves are crushed and the odour indicates the strain. When
grown for camphor the form is less important than camphor content.
Total oil yield per hectare can also be substantially increased by vegetative
propagation of high-yielding parent stock. A population of 2000-2500 per
ha is common but up to 5000 in some districts at approximately 2 x 2 m or
2 x 3 m. Plantations are weeded as necessary, and herbicides seldom used.
Trees are usually harvested at 16-20 months and normally make rapid
regrowth, since camphor coppices readily and vigorously.
Harvesting
When harvested for camphor the whole tree was felled, dismembered,
reduced to chips and distilled to obtain the crude oil, later refined to separate
the camphor. A full description of the process as carried out in Taiwan,
China and prior to the Second World War in Japan is contained in Hirota
(1 956). Where camphor trees are still felled to obtain crude camphor oil, the
operations involved are essentially those used to obtain sandalwood oil
(Chapter 13). When camphor was the main product in Taiwan, age of tree
was a major factor in selection for felling, since camphor content of oil
increased with age. Yield of camphor in a tree 120 years old was three to four
times that ofa tree aged 20 years, but oil content was 25% less. The average
oil and camphor yield from a mature wild camphor tree on Taiwan was 2.4%
and 1.1% respectively, from a plantation tree of the same age 1.5% and
0.8%. The oil content of wood from young camphor trees in Japan averaged
1.4%, the main components being camphor, 1,8-cineole and safrole in that
order (Fujita et al., 1974). A relationship with age also exists for other tree
types; for example the more mature the leaves and twigs collected from the
linalool-type tree, the higher the safrole or camphor content and the lower
the linalool content in the oil (Kingstone, 1966).
Plantations to supply leaves in Japan are normally harvested annually
between October and March, although in some districts two harvests are
possible; in Sri Lanka and India, up to four harvests per year are common.
Harvesting is generally manual, but mechanical harvesting is being in-
troduced in Japan. Initially this involved the use of hand-held mechanical
cutters and strippers, but fully mechanical harvesting of leaves is considered
practical, and may involve modifying the shape of bushes to suit equipment
as has occurred on tea estates, or harvesting smaller bushes more frequently.
This is possible since leaf growth is fast, new leaves becoming mature in 2-3
months. In these circumstances, it will be profitable to select strains with
high camphor or linalool content in young leaves.
The time of cutting is not critical but prolonged rain or very cloudy
weather reduces leaf oil content and camphor content of oil. After cutting,
lauraceae 163
leaves are carefully shade-dried for 2-4 days then distilled. There is appar-
ently no difference in the camphor content of fresh, wilted or shade-dried
leaves although, in Japan, oil content may be reduced by artificial drying. Oil
yield from commercial plantations is 50-150 kg/ha, and is highest in Japan
where selection is steadily increasing average oil yield. In India, camphor
yield is 1% by weight of leaves; in Japan, crude oil yield is 1.1-1.5%, with a
linalool content of 91 %; in Sri Lanka the yield is 130-140 kg/ha of crude oil
(probably distillate). A very high yield of 300 kgiha camphor was reportedly
obtained from a trial planting in Algeria, compared with 140-1 60 kg/ha from
pilot trials in the USA. The life of a leaf-producing plantarion depends
mainly on type of camphor tree selected. In Japan, the linalool type has an
effective life of approximately 20 years, but up to 60 years in India and Sri
Lanka where high-camphor forms are planted.
Distillation
Essential oil is obtained by steam or hydro-distilling wood or leaves, and
camphor obtained by pressing or rectifying crude oil. No special technique is
required, and oil remaining after removal of camphor varies according to
type of camphor tree (see section on products and specifications).
armillaria root rot, has a wide host range, and is most often recorded from
Africa, but also occurs in Asia. Other Armillaria spp. have been recorded
from China, Taiwan and Japan. Another root rot caused by Clitocybe
tabescens occurs mainly in more tropical regions of Asia and South East Asia.
Pink disease due to Corticium salmonicolor is widely distributed in the tropics,
and takes its common name from the mass of pink mycelium which covers
affected bark. Heavily infested branches are usually killed. Another fungus
which attacks small branches, twigs and shoots is Glomerella cingulata,
especially on Mauritius but also elsewhere in Asia. A canker caused by
Pkysalospora rhodina has been reported from a number of countries, but the
degree of damage on Cinnamomum spp. is usually slight.
Taiwan China
Hon-Sho Ho-Sho India Hon-Sho Yu-Sho
Specific gravity (25°C) 0.9543 0.9231 0.9428 o.91Ola 0.9270
0.9306 0.9436 0.9337
Refractive index (25%) 1.4806 1.4692 1.4713 1.4751 1.4704
1.4702 1.4727 1.4735
Optical rotation t 32'24' t 1714' t 26" t31'23' t 748'
t 22" -29O - 18"48'
Acid number 1,l ng 2.48-3.1 25 0.6-0.9
Ester number 4.1 ng 4.0-5.0 1.4-2.2
Acetyl value 23.3 ng 21.6 36.4-47.9
Alcohol content (%)(as 6.5 36.8-30.0 ng ng
CioHid)
Camphor content (%) 50.5 39.6-44.0 51.4 27.0-34.2
a 20"C, 30°C; ng, not given. Figures in columns are range.
litre oil and crystalline camphor (Tothill, 1952). Other species yielding an oil
containing camphor include Blumea balsamijera, Chysanthemum sinense var.
japonica and Menandra benghalensis.
Aniba rosaeodora Duckei (syn. A. duckei Kost), native to South America,
yields Bois de rose oil by steam distilling the wood. As oil is currently
obtained mainly from wild trees and few plantations are established, only
basic details are included. Aniba rosaeodora and what is claimed as the closely
related A. r. var. amazonica Duckei, are forest trees of the Amazon basin.
There are about 40 members of the genus, all native to tropical South and
Central America (Gottlieb & Kubitzki, 1981). Another member, A . cane-
lilla, has a cinnamon-scented bark which is locally used to make cinnamon
tea and to scent stored clothes and is probably the explorer Humboldt’s
‘Cinnamon of the Amazon’. Accepting the variety as botanically correct, the
following remarks apply to either.
Both are tall, to 30 m, straight, evergreen trees, trunk to 1 m DBH, with
large glossy green leaves. The timber was once extensively used for furniture.
Trees are felled, cut and transported to distilleries where the wood is
reduced to chips for distilling. The trees should have a minimum trunk
diameter of 35 cm DBH at about 10-15 years of age. Excessive exploitation
has greatly reduced easily accessible stands, and transport problems and
attendant costs have increased the expense of oil production. The Brazilian
government has introduced legislation requiring processing companies to
plant Aniba trees in proportion to the number felled, but it is generally
believed the number felled far exceeds those planted.
Oil obtained from comminuted wood averages 0.5-1.5% and there is
strong regional variation in wood-oil content. Yield from freshly cut logs is
higher than from trees felled and left for any period. Leaf and branch oil
content and composition have been shown to vary substantially between
seasons especially the monoterpene hydrocarbons, oxides of linalool and
linalool (Araujo et al., 1972).
Bois de rose Brazilian oil is colourless to pale yellow, with a sweet, slightly
woody odour; the main constituents are, in per cent: linalool 75-80, limo-
nene and eucalypt01 1-2, alpha-terpineol 4-6, 1-8-cineole + beta-
phelladrene 1-2, geraniol + geranial to 1 (Formacek and Kubeczka, 1982;
Buccellato, 1988). The major use is in soft drinks and confectionery. The
main oil characteristics are shown in Table 6.4.
Bois de rose Cayenne oil is usually colourless or faintly yellow, with a finer
scent than Brazilian oil, is normally available only in very small quantities
and its continued regular production is uncertain. The oil is slightly different
from Brazilian and main characteristics are included in Table 6.4. The
linalool content is usually higher, to 97% has been recorded, and is laevoro-
tatory due to the presence of I-linalool. The main oil use is in high quality
perfumes and toiletries. Linalool is also derived from Linaloe wood oil,
168 Chapter 6
Brazil Guiana
A B A
Specific gravity (15°C) 0.875 0.868 0.872
0.895 O.89ga 0.878
Refractive index (20") 1.462 1.4620 1.460
1.485 1.4685 1.463
Optical rotation - 4O -4" -12"
t5" t5" -17"
Ester content (%) (as linalyl acetate) ng ng 0.5-1 5
Alcohol content (??) (as linalool) 75-85 84-90 85-95
Solubility (viv 70% alcohol) 1:2 1 :2 1:3-4.5 (60)
a 20'Cc; ng, not given: A = commercial samples: B = EOA specification. Figures in columns are range.
Note: The linalool content is easily adjusted, and very high levels in Brazilian oils are suspect.
Mexican obtained from Bursera delpechiana, and Ho leaf oil from C. cam-
phora in Taiwan and Japan. All have been virtually replaced by synthetic
linalool.
Cassia
Commercial cassia bark and oil are obtained from various Cinnamomum spp.
as noted in the introduction to this chapter but only three are important
internationally: C. cassia (Chinese cassia), C. burmannii (Indonesian cassia),
and C. loureirii (Vietnamese cassia). Indian cassia is obtained from C.
tamala, and is of local importance only (Brown, 1955/56). In the text, cassia
without qualification will be used in a non-specific sense. The name cassia
should not be confused with the genus Cassia, Caesalpinioideae, of which C.
angustijolia supplies the well-known senna pods. Cassia bark and, later, oil
was originally obtained wholly from wild trees but increasing demand
encouraged cultivation, with China the major producer. Total production in
1995 was 300 tonnes, with Guangxi Province contributing 170 tonnes from
five large factories, and Guandong Province around 1 15 tonnes. Cultivation
also became more common in Indonesia and Vietnam, and it is most likely
that all future supplies to the international trade will be from plantations.
The cassia of European antiquity probably originated in the Assam area of
India and reached the Middle East via early Mesopotamian civilizations,
whose records on clay tablets indicate considerable knowledge of spices to
flavour and prepare foods as early as 2000 BC (Bottero, 1985). From
Mesopotamia to Egypt, thence to Rome and Greece is a natural progression,
and cassia was widely known in Europe by the fifth century BC. Pliny
describes both cinnamon and cassia in his Historia Naturalis (6.163-97), and
Lauraceae 169
apparently there were several kinds of bark and twigs known as cinnamon
and cassia. In fifteenth century England ‘synamome was for lordes, but
canelle (cassia) was for commyn people’. The powdered bark is current in
the British Herbal Pharmacopoeia as a specific remedy for flatulent dyspepsia
or colic with nausea.
Chinese cassia
The bark was probably first described in China in the fourth century BC
Elegies ofCh’u (Rosengarten, 1969), again mentioned in the third century as
kui, and the medicinal use of bark reported by T’ao Hun-kin (Laufer, 1919),
but its use must have predated these writers since the tree has been familiar
to peoples of China’s southern provinces before recorded history (Bretsch-
nider, 1895). A later mention of what could be cassia was in Periplus of the
Erythraean Sea, written about AD 200 (Huntingford, 1976). ‘Kasia bark’ was
listed as shipped from Somalia, but this was most likely re-export of Indian
or Indonesian material as the author found no Cinnamomum spp. native to
eastern Africa. Arabs and Persians later knew cassia bark as darchini, from
dar meaning wood or bark and chin meaning China. However in some
countries, including India, darchini came to mean true cinnamon!
Botany
Cinnamomum cassia Presl. ex-Blume, closely resembles C. obtusifoliumNees
of which it was originally considered a variety, and also incorrectly known as
C. aromaticum Nees. Chinese cassia produces dried buds, bark, and bark and
leaf oils of similar composition. Chinese cassia is known generally as kwei or
kui in China, widely as cassia lignea (true cassia) in international commerce.
Cinnamomum cassia is an evergreen tree to 20 m, naturally conical in shape
with small white flowers and coarse greyish-brown bark (Chun, 1921). The
tree is aromatic in all its parts, but only bark and leaf oil are economically
important. The root system of a wild, fully grown tree has apparently never
been described; however, under cultivation roots are deeply penetrating and
extensive, with normally a single taproot and several substantial laterals.
When grown from cuttings rooting is slow, commonly 6-8 months before
sufficient roots develop to allow planting out. This tardy root growth may
also account for the slow rate of above-ground growth, since at 5-7 years
trees may reach only 1.5-2.5 m. Roots contain an essential oil of no
commercial importance.
There is normally one central trunk on uncut trees to 20 m, 20-25 cm
DBH at 10-1 5 years of age. Cultivated trees are usually coppiced when 5-7
years at 2-3 m, and subsequent regular cutting keeps trees to 3-3.5 m. Trees
required to produce cassia buds or fruit are left uncut. The trunk is normally
well branched from low down, and the progressively acute angle of branches
170 Chapter 6
gives the tree its conical shape. Bark on mature trees rough, greyish brown;
on young shoots and stems smooth and greyish. Bark from mature wild trees
is locally considered especially valuable for medicinal use. The inner bark is
brown or brownish red, containing irregularly shaped oil cells in the phloem.
Bark from young shoots is removed, scraped, dried and becomes the cassia
bark of commerce. The inner bark contains 1.5-4.0% essential oil and
Kwangsi bark generally has a higher average oil content than Kwantung
bark; the main constituent is cinnamaldehyde a t 60-98%. The wide range of
yield and main constituents is apparently due to similar factors affecting
these characters in C. verum. For example, bark from trees growing at higher
altitudes, Kwangsi, is of better quality than trees from lower levels, whose
coarser bark is the cassia lignea of commerce. A detailed description of the
morphology and histology of various cassia barks is contained in Parry
(1962).
The leaves are alternate or nearly opposite, oblong-elliptic to lanceolate,
acuminate, basically cunate, petioles 1.O-1.5 cm. Young leaves are light
green, flaccid and soft; mature leaves stiff, thick, and leathery, with three
prominent ribs from the base, 8-20 X 4-7.5 cm, shiny dark green above,
finely hairy below. Fresh leaves may contain up to 1.6% oil although yield
after distillation seldom exceeds 0.5-0.6%; the main constituent is cinnam-
aldehyde at 80-95%.
The inflorescence is a lax, spreading, terminal or axillary silky-tomentose
panicle, 7.5-15 cm; the flowers are small, pubescent, white or whitish
yellow, in cymose arrangement of threes on short pedicels; the perianth is
pale yellow and deeply divided into six oblong, blunt lobes; main flowering
in China is June-August. The fruit is an ovalielliptic, black to blackish-
purple pulpy drupe, 1.O-1.5 cm, containing one seed borne in the shallow
cup of calyx lobes. Fruits are avidly eaten by birds which are the main seed-
dispersal agents; fruits mature February-March. The seed is oval, 1 cm,
dark brown with lighter stripes, usually 2500-2700 per kg. Immature fruits
are dried and become the misnamed cassia buds, Flores cassia, of commerce.
The morphology and histology of cassia buds have been described in detail
(Parry, 1962).
Ecology
Chinese cassia is a subtropical forest tree, between 30"N and 30"s latitudes
and within these limits has adapted to a range of environments. Cassia
occurs naturally in south-eastern China, northern Assam and Burma, but is
cultivated only in southern China (mainly Kwangtung and Kwangsi), and
on Sumatra. It was introduced to South America, the southern USA, Hawaii
and Sri Lanka, but is not commercially cultivated. Cassia flourishes in hot,
sunny and humid conditions, at temperatures of 1 5 3 0 ° C but is adaptable to
B very wide seasonal range from the high 30s to -7"C, and also to very large
differences in day-night temperatures. The main areas of cultivation in
Lauraceae 171
Cultivation
Cassia can be grown from seed, cuttings or layering, but the first is most
common in China. Ripe fruit is picked and treated as for cinnamon. Seed is
planted as soon as possible; if sown within 7 days germination is around 90%
after 20-40 days; after 20 days storage viability quickly falls. Seed can be
sown in pots or prepared beds, 1.5 cm deep and 3 cm between seeds. Most
seed-beds are located under trees and thus intermittent shade is provided
naturally until seedlings are strong enough to be planted out in 1-3 years.
Cuttings can be one, two or multiple leaf, from ends of new growth, and
allowed to wilt before planting. Rooting is slow (2-18 months) and the strike
rate is usually about 50%. The optimum time for taking cuttings is just after
the new flush of growth has matured, when flaccid young leaves become firm
and horizontal. Single leaf cuttings with 5 cm stems and the top axillary bud,
172 Chapter 6
Harvesting
Bark collection, stripping and drying is basically as for cinnamon and has
been described in detail (Purseglove et al., 1981). Cassia buds, kui tsz, are
the dried immature fruit including calyx and often pedicel. These are
harvested from trees especially left in plantations, or collected from wild
trees in October-November. Cassia buds are smooth, greyish or reddish
brown, 6-10 X 5 m m in diameter; the attached calyx is hard and wrinkled.
Bark, leaves, twigs, small stems and fruit buds can be used for oil production
since characteristics are similar.
Leaves are collected when the bark is harvested, but prunings and excess
foliage can be distilled at any time. The main harvest in China is June-July,
a lesser in January-February; the highest oil yield is from green leaves
harvested in May-July (Anon., 1976). Oil quality depends on time ofharvest
and leaf maturity, but there are no published data on seasonal or regional
influences. Large mature leaves are considered to yield best quality oil in
China, and a high admixture of twigs and small branches reduces quality.
Leaves are often wilted for 6-8 days before distilling, but the reason for this
is unclear since fresh leaves generally give a higher oil yield. Pests and
diseases are as for C. verum and C. camphora.
lauraceae 173
Distillation
Leaf oil was formerly produced in simple, locally constructed, direct-fired
stills but the establishment of rural communes in China allowed the in-
troduction of more modern stills, centrally operated and maintained. These
stills take charges to 200 kg, are coal fired, and cohobation is normal. The oil
quality, characteristics and cinnamaldehyde content are subject to official
standards, which are not very efficiently enforced. The oil content of fresh
leaves in China is 0.3-0.4%, seeds (buds?) 1.5% and small twigs and stems
1-2% (Anon., 1976), but Parry (1962) quoted the oil content of bark as
1.5%, buds 0.55%, budsticks (pedicels?) 1.64%, and a mixture of leaves,
petioles and young twigs as 0.77%.
nature, the only other major source being a by-product of cocaine extraction
from Erythroxylum coca. Methods of distinguishing C. verum bark oil from
bark oil of other Cinnamomum spp. have been published (Lawrence, 1969;
Salzer, 1975). The minor components used are ortho-methoxy cinnamalde-
hyde, eugenol and coumarin. Cassia bud oil is similar to leaf oil with a yield
of 2%, and aldehyde content of 80%. It is neither produced nor available
commercially. The leaf and bark oils from other Cinnamomum spp. occur-
ring in China have been described (Zhu et al., 1994)
Indonesian cassia
Fig. 6.2. Cinnamomumburmanii, fruiting shoot. (Department of Agriculture Library, Kuala Lumpur.)
176 Chapter 6
Vietnamese cassia
Indian cassia
Indian cassia bark and leaf oil are obtained from Cinnamomum tamala Nees
& Eberm. native to the subtropical Himalayas from Kashmir to Assam and
Bangladesh. The tree is known as tejpat or tajpat in Hindi and Bengali,
tejpatra in Sanskrit and tamalpatra in Gujarati. Bark is traded locally as
Indian cassia bark or Indian cassia lignea, and substantial amounts of bark
and dried leaves are exported to neighbouring countries. Fresh and dried
leaves, as tajpat leaves, are commonly sold as a spice in markets, and an oil is
obtained by hydro or steam distillation of fresh, wilted or dried leaves.
Two types of C. tamala are known; one whose leaves yield oil high in
eugenol, the other in cinnamaldehyde. The strains are somewhat dissimilar
in appearance and range, though the latter overlaps. The fruits are attractive
to birds, which are the main seed-dispersal agents. Both types apparently
grow easily from seed, as almost pure stands can be found in some areas,
Sylhet, Bangladesh for example, and commonly found growing naturally in
forests between 900 and 2000 m on the Himalayan foothills. Bark from both
is used to adulterate C. zlerum bark.
Cinnamomum tamala (eugenol type) is an evergreen primarily forest tree
generally below 8 m, with the central trunk to 45-50 cm DBH, and
mucilaginous bark. The leaves are large, dark green, generally lanceolate,
acute, three ribbed, on short, stout petioles. The flowers yellowish to white,
on loose terminal panicles. The fruit a small pulpy, purplish drupe contain-
ing one small brown seed. The tree is very long-lived, and when only leaves
are harvested, cultivated trees over 100 years old are known in northern
India (Gulati, 1982). The eugenol type occurs generally over the natural
178 Chapter 6
range of the species in India and Bangladesh and wild trees make a sub-
stantial contribution to the total of harvested leaves and bark. The tree is also
widely cultivated in large and small plantations to provide dried leaves as
spice, but leaf-oil production is less common; the bark is coarser than C.
verum bark, although possessing the true cinnamon odour, and is a used as
a substitute and adulterant.
Cinnamomum tamala (cinnamic type) is generally slightly smaller than the
eugenol type, to 7.5 m, with the central trunk 30-40 cm DBH when mature.
The leaves are opposite, glabrous, green to dark green, three ribbed, ovate-
oblong, acute, on short petioles; young leaves are lanceolate, acute, initially
pink. The flowers are small, whitish to yellow white, on loose terminal
panicles. The fruit is a small fleshy, purplish drupe, containing one dark
brown seed. Main flowering is in May and the fruits ripen in June-July. This
type of C. tamala is more restricted in its range, although locally common in
a specific area. It is especially plentiful in hills around Nainital, where it is
also cultivated on small farms. Wild and cultivated trees are harvested
primarily for their leaves, which are dried and sold as a spice, and only a
small amount of leaf oil is distilled.
Table 6.6. Characteristics of WOtypes of Cinnarnomum tarnala leaf oil from India
Table 6.7. Main components of Cinnamomum tarnala leaf oil from India (%),
Cinnamic type a
Wild Cultivated Eugenol type
~~ ~
Cinnamon
It was as spices that cinnamon and cassia first became commercially im-
portant and cinnamon is still considered one of the finest sweet spices, with
cassia a coarser substitute. Reports of cinnamon usage in dynastic Egypt or
ancient China are considered unsound and probably referred to cassia, if
indeed the references were to any kind of Cinnamomum bark. This is
particularly so in China, where early Chinese writers habitually referred to
several spices with the same character(s), the meaning often changing over
time according to the spice’s relative importance. The very ancient use of
these spices in their natural habitat to flavour food, in social and religious
ceremonies, and in herbal remedies, with their dissemination along tradi-
tional eastern trade routes is well documented. As this trade gradually
extended from Asia to Europe, so knowledge and use of the spices grew,
creating greater demand. This demand finally sent Europeans sailing in
search of their source and thus broke Arab domination of the oriental spice
trade - a monopoly the Arabs had maintained for centuries by inventing
fantastic stories of the origin of cinnamon and the dangers involved in its
collection, many retold in the Thousand and One Nights. It was to search for
a route to the spice islands that sent Columbus sailing west to make landfall
in the Americas, discover gold not spices, and led to the destruction of the
great Aztec and Inca empires.
An early mention of Ceylon cinnamon by the Arab writer Kasawini about
AD 1275 was followed by that of a Minorite friar, John of Montecorvino in
1293. The noted traveller Ibn Batutah in 1340 also referred to Ceylon
cinnamon in his books, and described the huge storehouses, bangasalus,
which then existed in what is now Colombo. In the eighteenth century the
Dutchman Francois Valentijn described bark harvesting in detail (Arasar-
atnam, 1978). Arabs called it kitfut-ed-darsini, corrupted later to kitfah
which survives as kelfar used to describe Malabar cinnamon bark.
When the Portuguese arrived in Ceylon in 1505, cinnamon bark was
obtained from wild trees, and the historian Ribeiro described various meth-
ods used to collect and prepare bark in the Kingdom of Kotte (Fatalida
historica da Ilha Ceilao. Lisbon 1836 trans.). Following the Dutch occupa-
tion, systematic cultivation began between 1767-1 770 by a colonist, de
Kok, encouraged by Governors Falk and Van der Gaff. Under the Portu-
guese and Dutch, cinnamon production in Ceylon was a state monopoly,
protected by the most stringent laws, but the British abolished the monopoly
in 1833. Removal of the very high export duty in 1843 encouraged exports,
and large shipments were made to Europe. Contemporary authors state the
Portuguese extracted an annual tax of 110,000 kg of bark; the Dutch
produced about 180,000 kg annually, and under the British this rose to
lauraceae 181
500,000 kg. The area under cultivated cinnamon in Sri Lanka was 16,000 ha
in 1850, falling to 14,000 ha in 1970 (Wijesekera et al., 1976).
The Dutch also introduced cinnamon into their East Indian colonies,
particularly Java, and Indonesia is now a major producer and exporter of
cinnamon products. The Central Bureau of Statistics reported that bark
exports in the period 1989-94 were, in tonnes, 3293, 2957, 3958, 7857,
3379, 3890, with the USA, Singapore and The Netherlands the main
destinations in that order. Liberalization of the industry is expected to
increase production of cinnamon products and exports. Cinnamon was
brought to the Seychelles in 177 1 by Pierre Poivre and, by 1880, successfully
established (Fock-Heng, 1965). Although cinnamon has been introduced to
many other tropical countries, significant commercial production is limited
to Sri Lanka, the Seychelles, Madagascar and India in that order. Where and
when cinnamon oil was first produced is unknown, but it was steam distilled
in Ceylon during the Dutch occupation. An early mention of cinnamon oil is
in European documents of the sixteenth century, when a German ordinance
was published controlling its price.
Botany
Cinnamomum verum Presl. (syn. C. zeylanicum Breyn; Laurus cinnamomum
L). The generic name is derived from the Arabic or Persian mama via the
Greek amomum meaning spice, and the prefix chini to its believed origin.
Cinnamon is commonly known as kurundu in Sri Lanka, dalchini and
ilayangam in India, in Sanskrit tamalpatra, in Dutch caneel, in Spanish and
Portuguese canela, and commercially as true cinnamon to distinguish it and
its products from those derived from other Cinnamomum spp. In this section
cinnamon will designate C. verum and its products unless stated other-
wise.
Cinnamomum verum is a bushy, evergreen tree to 15 m with numerous
branches, long, leathery, bright green leaves, small yellow flowers and ovoid
blackish fruits (Fig. 6.3). The bark and leaves are strongly aromatic. There
is considerable physical variation within the species, indeed between in-
dividual trees, and distinct local strains exist. In Sri Lanka for example, six
named types are differentiated by the taste and aroma of their leaves.
Chemical classification would be a more accurate method of differentiation;
however, some differences are probably due to local environment as soil and
climate have a substantial effect on tree growth, type of bark and oils
produced, as discussed later. That three quite distinct essential oils (bark,
leaf and root) are obtained from one tree is of particular phytochemical
interest, and the method of biosynthesis which can generate or store eugenol
in leaves, cinnamaldehyde in stem bark, and camphor in root bark is
remarkable; light may be a major factor. The biosynthetic pathways of
phenylpropanoids and terpenoid compounds leading to oils with varying
182 Chapter 6
colour could be used to select seedlings with a high bark oil content
(Krishnamoorthy et al., 1988). Bark on young shoots is removed, scraped,
dried and becomes the cinnamon quills of commerce. Five bark types are
recognized in Sri Lanka, but the major differences are probably more
environmental than genetic, since visual differences do not consistently
correlate with oil yield and type. Commercial cinnamon bark is a dull pale
brown, to 5 . 5 mm thick, with the inner surface somewhat darker, and finely
striated longitudinally. The finest quality has a delicate fragrance and tastes
warm, sweet and pleasant. The presence of calcium oxalate crystals has been
correlated with low quality bark (Mikage et al., 1987).
The leaves are opposite, leathery, generally ovate or elliptic, 5-18 X 3-10
cm, base rounded, tip acuminate; exstipulate; petiole 1-2 cm, upper surface
grooved. They are green to bright green above, dull grey-green below, but
young leaves may initially have a reddish tint. The leaves have three to five
conspicuous longitudinal light-coloured veins running from base to tip (Fig.
6.4). The anatomy of many Cinnamomum species has been described in
detail (Bakker et al., 1992).There is considerable local variation in leaf shape
and size, and four types are recognized in Sri Lanka based on taste: sweet,
neutral, pungent and bitter, but only the last two are distilled. By keeping
them separate it is possible to produce two types of oil, but there is no
commercial demand. The two leaf oils formerly available in Sri Lanka were
regional types; one high and one low in eugenol. Green leaves yield about
1% essential oil, cinnamon leaf oil, whose main constituent is eugenol at
70-95 %.
The inflorescence is a lax terminal or axilliary panicle, peduncle soft,
hairy, to 5-7 cm. The flowers are up to 3 mm diameter, pale yellow or cream,
and have an unpleasant fetid smell. The calyx is campanulate, softly hairy,
with six acutely pointed segments; a corolla is absent. The superior uni-
locular ovary tapers to a short style, and contains a single ovule. Main
flowering is April-May in Sri Lanka, January-February in India. The fruit a
black or bluish fleshy ovoid drupe, 1.5-2.0 cm when ripe, the enlarged calyx
persisting at its base. Fruits mature in 3-5 months, are edible and attractive
to birds which are the main seed dispersal agent. Ripe fruits have a distinct
odour when opened, and a flavour resembling juniper. Crushed ripe fruits
boiled in water yield an oil known in India as cinnamon suet, used in candle
making. The globular brown seeds contain approximately 33% fixed oil.
Ecology
Cinnamon is a tree of the wet tropics and a somewhat restricted natural
range in Sri Lanka, India and South East Asia. Introduced to other coun-
tries, it thrives and is grown commercially only where the climate is very
similar to its natural habitat. In the Seychelles, fruit pigeons and the
introduced Indian mynah so widely dispersed seeds that cinnamon became
dominant in secondary forest on the main islands of Mahe and Silhouette a
184 Chapter 6
century after its introduction, and seedlings appeared on almost any vacant
undisturbed land. Another important factor assisting establishment is that
cinnamon is not an attractive browse plant to cattle and, once past the
seedling stage, is normally ignored.
Cinnamon produces the finest bark in sunny regions with an average
temperature of 27-30°C, but will grow well in a wide range of environments
as specimens in botanic gardens in many countries demonstrate. Bark from
these trees is usually of indifferent quality; for example bark collected by the
author from trees growing in the Botanic Gardens, Entebbe, where the
climate is cooler and drier than Sri Lanka, was thin, brittle and of poor
quality; similarly in Melbourne, Australia. A rainfall of 2000-2500 mm with
no pronounced dry season is the optimum, thus productive regions near
Colombo, Sri Lanka and Victoria, Seychelles, have an almost similar annual
Fig. 6.4. Cinnamomum verum, cinnamon, A, Flowering shoot; B, flower in longitudinal section; C, fruits;
D, fruit in longitudinal section. (Purseglove et al., 1981,)
lauraceae 185
established estates in India which have their own stills return residue to the
plantation as mulch.
Cultivation
Cinnamon is propagated vegetatively or from seeds but as cinnamon is open
pollinated, vegetative propagation from selected parents is recommended
and modern micropropagation methods have been successfully used to
produce large numbers of seedlings (Rai and Chandra, 1987). The fruits are
avidly eaten by birds, and it is usually necessary to bag fruits for protection.
Ripe fruits for seed are heaped until the pulp rots and the seeds are then
washed, shade-dried, and planted as soon as possible since viability quickly
declines. In India, where freshly harvested seed was sown daily for the first
7 days and weekly for 7 weeks thereafter, the highest germination rate of
94% was obtained by sowing the third day after seed was harvested, fell to
50% after 14 days storage and was nil after 40 days. Seeds sown within
7 days germinated in 20-25 days; later sowings extended the period to
32-42 days (Kannan and Balakrishnan, 1967).
Seeds should be sown thickly in rows 20 cm apart and covered with
2-3 cm of soil. Four-month-old seedlings may be transplanted into baskets,
and planted out 4-5 months later. Seedlings may also be left in the nursery
until required for transplanting, but from six months shade should be
gradually reduced. Cinnamon can be propagated by cuttings, layering, or
division of rootstocks, and the last is probably the quickest method of
upgrading a plantation, since high-yielding stools can be selected. Plants are
cut to within 15 cm of the ground, suitable sections of rootstocks removed
with adhering soil and planted out immediately into prepared pits. The great
advantage of this method is that stems are ready for cutting in 12-1 8 months
compared to 2.5-3.0 years for seedlings. A comparison of cuttings of
different lengths and ages in the Seychelles showed young two-node cuttings
planted in polythene bags filled with a mixture of red soil and coconut coir
dust then placed under a polythene tent had the highest strike rate and were
ready for planting out in 12-18 months. Apical cuttings with three leaves
were best in Sri Lanka, and treatment of more mature hardwood cuttings
with IBA at 2500 ppm encouraged rooting.
A field spacing of 120 X 120 cm or 90 X 90 cm is recommended in Sri
Lanka for commercial plantations, and blanks should be infilled to ensure
this population is maintained (Wijayaratne and Pieris, 1981). For direct
sowing, a 2 X 2 m spacing is usual and up to 20 seeds are placed into
prepared holes and later thinned. In India, a common spacing is 3 X 3 m
with several seeds sown per site. Seedlings are planted out just prior to the
rainy season with supplementary watering, or just after the rains commence.
This is normally June-July in India, May-June in Sri Lanka, and October-
November in the Seychelles. Weeding three or four times annually is
necessary during the first two years, thereafter twice per year should be
lauraceae 187
Harvesting
The stems are cut during the rains to facilitate peeling, when the red flush of
the young leaves is beginning to turn green and the sap is flowing freely. In
Sri Lanka, harvesting is mainly in May-June and October-November,
although harvesting on a limited scale continues throughout the year. The
first harvest yields a thick bark of inferior quality, but quality improves from
successive cuts. The finest quality is obtained from shoots with uniform
brown thin bark, 1.0-1.25 m X 1.25 cm diameter, from central shoots, and
the middle portion of shoots. The leaves and twigs remaining after cutting
can be distilled. Cut shoots are bundled and removed for peeling and drying
(Fig. 6.5). Bark to be distilled for oil should not be left in wet bundles or
become damp, as this encourages mould or fermentation which directly
affects oil composition. A detailed description of the preparation of cinna-
mon quills, and the various grades and qualities obtained are contained in
Purseglove (1 98 1).
A brief description of the various grades is useful, since the names will
occur in the text. Quills are yellowish-brown, cylindrical, approximately 1 m
X 10 mm diameter, scraped, smooth, thin, rolled bark with occasional scars,
and fine, light-coloured wavy lines running lengthwise and clearly visible.
These have a pleasing, fragrant odour and a warm, sweet, aromatic taste.
Quillings are broken pieces of quills of various grades, length and diameter,
which approach but do not equal the full aroma and taste of sound cinna-
mon quills. Featherings are the shavings and small pieces of bark remaining
after processing inner bark into quills, with a taste similar to quillings. Chips
are small pieces of mature, unpeelable bark, occasionally fragments of inner
bark, greyish brown within, and deficient in both aroma and taste.
Bark yield per hectare in Sri Lanka depends mainly on the standard of
management. Smallholders average 100-125 kg/ha quills and 25 kgiha chips
annually; on well-managed larger plantations the first crop after 3-4 years
yields 56 to 67 kgiha quills subsequently increasing to 168-224 kg/ha. About
63 kg of cinnamon chips and approximately 2.5 t of undried leaves per
hectare are also obtained, but only 75% of leaves available are distilled, the
remainder left as mulch. On well-established plantations in India, the first
188 Chapter 6
Distillation
Cinnamon bark produces two oils, a superior type derived from the inner
bark and a lower quality from broken quills, chips and bark. In Sri Lanka, a
third oil, katta thel, is produced from bark and twigs for local consumption.
lauraceae 189
stills and obtained yields of 1.5-2.5%. Yields of 0.5-2.0% are obtained from
Sri Lanka chip material by using stills developed by the CISIR, although oil
quality frequently improves before an increase in yield.
Leaf oil in Sri Lanka is obtained by similar methods, frequently in the
same stills. Since leaves are bulky and difficult to transport, stills are usually
located at a convenient site, and constructed of materials locally available.
Thus leaf oils in India, Sri Lanka and elsewhere can be very variable, but
where more modern equipment has been installed, oil of excellent quality
and type is produced. Seychelles oil is produced by water-steam distillation
of leaves hand stripped from shoots cut for bark production. Leaf stripping
is unpleasant, hand-blistering work and thus unpopular with local people,
who gladly abandoned it for less onerous jobs in the booming tourist
industry. Leaves are collected during January-September, with an average
yield of 1.8-2.0 tiha, and are distilled as soon as possible after arrival at the
distillery. A charge is 300 kg fresh leaves, distilled for about 6 hours. Under
the government's 1982-1 986 Development Plan, two modern distilleries
plants were constructed, but neither achieved full commercial operation.
Assessing available information on leaf oil production showed that there
appears to be a lack of basic data on effects of the various techniques used:
for example, the oil yield and any change in oil composition resulting from
distilling fresh leaves or leaves wilted for 1-7 days. This information is
essential to ensure a distillery operates efficiently and profitably, and greatly
assists the integration of oil production with other plantation operations.
and in the field after transplanting. Larvae of various Popillia spp., which
attack roots, are particularly damaging in nurseries and on young seedlings.
Most often recorded are P. complanata and P. discalis. Pests recorded from
stored cinnamon quills in India and Sri Lanka include Lasioderma sem'corne,
Pyralis farinalis and Sitodrepa panicea.
The numerous pathogens recorded from cinnamon are mainly of little
importance, are normally found at a low level of infection on other local trees
and seldom warrant control measures. Diseases of C. verum also attack other
Cinnamomum spp. and are marked with an asterisk*. In India, Sri Lanka and
Indonesia, a most important disease is caused by Corticium salmonicolor*
(syn. C. javanicum). Symptoms are the formation of a pale, pinkish-white
brittle layer (crust) on young stems or branches and, unless controlled, the
infection spreads, destroying bark and finally killing shoots. The pathogen
also attacks mango, jackfruit, custard apple and other fruit trees often
growing in the vicinity of cinnamon plantations. Affected prunings and other
plant parts should be burnt. Also attacking shoots and young stems is
Phytophthora cinnamomi*; the most obvious symptom being vertical strips of
dead bark which are most numerous near ground level. The disease is
prevalent on badly drained soils. In Indonesia, it has also been recorded on
other Cinnamomum spp.
A stem disease caused by Exobasidium cinnamomi can become serious in
some years in Sri Lanka, and may also spread to leaves. Symptoms on leaves
are small yellowish concave spots whose underside bears greyish-white spore
bodies. A stem blight due to Diplodia spp. attacks young seedlings in the
nursery and when first planted out. Symptoms are small light brown patches
on stems, which spread and cause eventual death. Several root rots are
damaging to some extent; brown rot caused by Phellinus lamaensis, white rot
by Fomes lignosus* and two black root rots by Rosellinia spp. Symptoms are
roots covered with a network of black mycelium and small white starlike
spots under the bark. Rosellinia spp. attack a range of cultivated trees and
shrubs. Leaf blight caused by Pestalotia cinnamomi, known in Sri Lanka as
grey leaf spot or blight, can cause severe damage and defoliation. Symptoms
on leaves are small yellow spots, becoming grey with central black dots. A
closely related species attacks coconut, and cross-infection is suspected.
Glomerella cingulata* can become locally or seasonally severe in India with
similar effects on trees. Three other leaf diseases have been recorded, caused
by Aecidium cinnamomi, Leptosphaeria spp. and Gloeosporium spp. but the
extent of damage was not noted.
Table 6.8. Main constituents of cinnamon oils from Sri Lanka (%)
Cinnamon bark oil is a pale to dark yellow oily liquid with a strong, warm,
sweet, spicy, tenacious odour and a sweet, pungent but not bitter taste.
Lower-grade commercial oils are usually darker and lack the powerful odour
and tenacity. Oils distilled in the U K usually differ from those produced in
other European countries and the USA in having a lower specific gravity and
containing about 60% cinnamaldehyde and 10% eugenol. Oil from first
grade Sri Lanka quills, often known as Ceylon oil, is considered the best
quality. The main oil use is to flavour foods, beverages, pharmaceutical and
dental preparations, cosmetics and perfumes. Bark oil is considered to have
a strong germicidal activity and is a fungicide (Mangiarotti et al., 1990).
Research by the British Ministry of Agriculture produced a cinnamic acid
preparation with strong bird repelling properties which, when sprayed on
fruit trees and bushes, effectively protected fruit from their depredations.
Adulteration of bark oil is common, most frequently with cinnamon leaf oil,
cassia oil, clove leaf oil, eugenol and cinnamaldehyde, which is mass pro-
duced from coal-tar bases, especially toluene. Methods of detection have
been published.
The major oil constituent is cinnamaldehyde, but other components
impart the characteristic odour and flavour distinguishing this oil from other
Cinnamomum bark oils. The powerful characteristic note of high quality oil
is believed due to a combination of methyl-n-amylketone and other alde-
hydes and ketones, but the relationship between oil composition and orga-
noleptic properties has still to be accurately determined. Analyses of the
lauraceae 193
three oils are shown in Table 6.8.; however, the analytical method used can
produce differing results (Lawrence, 1994). The proportion of individual
components in Sri Lanka bark oils can vary; for example ester content is
partly dependent upon the hydrolysis occurring with a particular distillation
method. Aldehydes can range from 51-76% (average 5 9 , with eugenol
5 1 8 % but up to 31%. A high proportion of chips in distilled material can
increase eugenol content, and quillings and featherings can increase the
aldehyde content. Seychelles bark oil is generally considered inferior to Sri
Lanka, with a harsher odour probably due to camphene and camphor not
present in Sri Lankan oils. Efforts to improve quality produced oil with a
cinnamaldehyde content similar to Sri Lanka oil, but eugenol of only 3 4 % .
Few studies of Madagascar bark oil have been published; material distilled
in Europe gave oil yield of 0.6-1 .O%, aldehyde content of 62-65% (Paillot-
Cressole and Duquenois, 1952), but there is some doubt if this bark was
wholly from C. verum. A Malaysian bark oil had a cinnamaldehyde content
of 72-80%, eugenol 1-2%, and isoeugenol 1%.
Information on non-commercial sources of bark oil has been reported;
Ghana (Talalaj, 1967; Angmor et al., 1972); Burma (Kya and Min, 1970),
and some possessed unusual properties compared to commercial oils. The
major characteristics of bark oils from selected origins are shown in Table
6.9.
Leaf oil is a yellow to brownish-yellow, with a warm, spicy, somewhat
harsh odour, lacking the richness of bark oil and a slightly bitter, burning,
very spicy and powerful flavour. The main use is in perfumery and con-
fectionery and as a source of eugenol. The main characteristics of leaf oil
from selected origins are shown in Table 6.10 and the major components in
Table 6.8. Leaf oil is also produced in India and oil originating from
commercial estates in Kerala State is of international quality. Oil from other
Indian regions is very variable, and some published analyses may be from
material other than C. zlerum. The eugenol content is generally 70-85%, and
some characteristics are shown in Table 6.10. Seychelles oil is a valued
source of eugenol, usually above 90%, with phenols 78-95% and aldehydes
5%. Madagascar oil has a eugenol content of 70-90%, and a distinguishing
feature is the relatively high benzyl benzoate content. A GLC study of leaf oil
prepared from cultivated trees in Ghana showed 92% eugenol, 2% eugenol
acetate and 1% cinnamaldehyde (Angmor et al., 1972), but apparently no
significant amount of benzyl benzoate, although 3% was found in Sri
Lankan samples. A similar study on Malaysian oil gave a eugenol content of
68-82% and isoeugenol 5-9%, considerably lower than Sri Lankan oils.
Thus the geographical origin of leaf oil could be determined on eugenol and
certain minor component content, but this would require more data from
detailed analyses. The only common adulterant is clove leaf oil, used when
cinnamon leaf oil is higher in price; the reverse also occurs.
Root-bark oil is colourless to pale yellowish brown, similar in odour to
stem bark oil but weaker, lacking in fragrance, and camphoraceous. The
major component is camphor, to 60%, which crystallizes out on standing.
Oil yield from root bark at 1-2.8% is higher than from stem bark and leaves.
GLC studies of oil obtained from single sources in Sri Lanka (Wijesekera et
al., 1974), and Ghana (Angmor et al., 1972) showed monoterpene compo-
nents at 80%, cinnamaldehyde at 8%, and a trace of eugenol in the
Ghanaian sample. In the Sri Lankan sample, 22 components were identified
with 1,8-cineole (1 5%), eugenol (5%) and cinnamaldehyde (4%) the princi-
pal minor components, with monoterpene composition similar to stem bark
and leaf oils.
Oleoresin is prepared by extracting cinnamon bark with organic solvents;
the yield using ethanol is 10-12%, using benzene 2.5-4.3%; more recently
1,1,2-trichloro-1,2,2-trifluorethanehas been used (Bernard et al., 1989).
Oleoresin is a deep reddish or greenish brown rather viscous liquid, and
contains steam-volatile oil, fixed oil and other extractives dependent on the
solvent employed. Little information is available on the composition of
cinnamon oleoresins, but volatile oil content is 16-65%. The major use is in
flavourings, cake and similar mixes, pickles, prepared meats, convenience
foods and related products where flavour stability at high temperature is
important. Oleoresins are normally commercially available dispersed on
various bases and usually sold on a volatile oil content of 23-27 or 65%.
Oils obtained from a large number of other Cinnamomum spp. of little or
no commercial importance have been described from India (Gulati, 1982;
Khana et al., 1988): Sri Lanka (Wijesekera and Jayewardene, 1974; Sri-
tharan et al., 1994): Nepal (Sthapit and Tuladhar, 1993): the Philippines
(Concha and Cruz, 1968; Lawrence and Hogg, 1974): Taiwan and China
(Hiraizume, 1950: Hirota and Hiroi, 1967; Lin and Hua, 1980): Far East
(Fujita et al., 1974; Lin and Hua, 1987).
Sassafras
Botany
Sassafras albidum (Nutt.) Nees (syn. S. oflciizale Nees; S. variifolium Karst.;
S. sassafras Karst.; Laurtls sassafras L.) The generic and common name is
derived from the French or Spanish salsafras as noted, and the common
name is used in many languages almost without change. There are two other
members of the genus, S. randaieme Hayata, native to Taiwan and very
scarce, and S. tzumu Hemsl. which is more common in China and some-
times cultivated. All are generally similar in appearance. A number of
unrelated species are also called sassafras, including Californian sassafras
Umbellularia californica Nutt., swamp sassafras Magnolia virginiana L., and
Tasmanian sassafras Atherospemza moschatum Labill.
Sassafras is a deciduous tree to 35 m, but generally half this, and when
undisturbed has an irregular, pyramidal head. T h e bark is grey and vertically
fissured; the timber fragrant, tough and durable. Trunks were hollowed by
local people to make excellent canoes, and settlers later used the wood for
fence posts and casks. Wood and bark yield a bright yellow dye. The leaves
are alternate, 7.5-15 cm, three-veined, and vary in shape from more or less
entire when young to two- and then three-lobed when mature and all three
shapes can be found on the same branchlets. The colour is bright green
above, becoming orange and red in autumn. Sassafras is one of the few
deciduous laurels. Fresh leaves when crushed emit a strong sweet scent of
orange-lemon-vanilla, quite different from the odour of the root oil. Oil
distilled from fresh leaves contains citral and geraniol but little or no
safrole.
The flowers are yellowish-green, usually unisexual sometimes bisexual, on
5 cm racemes, appearing before the leaves in April-May; petals are absent
and there are six sepal lobes. T h e male flowers have nine stamens; females a
single ovary and style. T h e fruit is an ovoid bluish-black slightly waxy drupe
7 mm in diameter, with a fleshy, red pedicel and containing one dark brown
to blackish ovoid seed, 6 X 4.5 mm; its wall is papery and brittle. Propaga-
tion is by seed, suckers which are freely produced, or root cuttings. Trans-
planting of large seedlings or young trees is difficult due to the long tap
root.
Ecology
Sassafras is native to eastern and south-eastern North America from Maine
to Florida, and west to Texas with the main producing areas in the Appa-
lachian Mountain region. Sandy soils are favoured, but sassafras can be
found on heavy loams, less often on more clayey soils. Because of its wide-
spreading and fast-growing root system, sassafras quickly colonizes cleared
land or neglected farmland, and grows rapidly in the first five years. It can
thus suppress more locally desirable species in its native habitat, and there
was a continuous supply of material available for distillation from farmers
and landowners keeping their properties free of sassafras. Trees less than ten
Lauraceae 197
years of age are normally not harvested in commercial felling as the oil yield
is too low.
Table 6.1 1, Constituents and characteristics of sassafras oil from the USAa,
Characteristic Constituent %
Specific gravity (25’Cj 1.065-1.076 Safrole 80
Refractive index (20’Cj 1.530 Alpha-pinene plus
Optical rotation (25”Cj t2’0 to t 4 9 Phellandrene 10
Acid number to 1.0 Camphene 7.0
Ester number 0.5-5.0 Eugenol 0.5
Solubility (viv) alcohol 1 :2 (90%) Other 3.0
Solidification 4-6°C
a Average values.
198 Chapter 6
sassafras oil in 1995 was about 1000 tonnes, with half further processed
domestically to produce heliotropin. Australian sassafras oil is obtained by
steam distilling the bark and leaves of Doyphoru sassafrus Endl. of the small
Monimiaceae family, whose members are confined to the Southern Hemi-
sphere tropics.
Doyphoru sassafras,locally known as golden deal, is native to warmer,
high rainfall areas of eastern Australia and is a tall rain-forest tree to 42 m but
generally smaller, to 1.2 m DBH. The leaves are opposite, oblong-
lanceolate, glossy green above, dull green below, and fragrant when crushed.
The flowers are white, 2-3 cm diameter, on three-flowered axillary panicles.
The fruits are ovoid, 2 cm, containing two to four seeds; the carpels are very
hairy. The wood is pale yellow when freshly cut, darkening with exposure.
The leaves, bark and sapwood contain the essential oil which it is now
unprofitable to produce. A characteristic of undisturbed D. sassafrus stands
are circular clumps of trees, each having arisen from a decayed central
parent.
Ocotea
The genus Ocotea Abl. contains some 350 species mainly in the New World
tropics and sparsely represented in Southern Africa and Madagascar. Many
species are important timber trees: in South Africa 0. bullata (Birch) E.
May, stinkwood, was so extensively felled that it is now a protected species;
wood of 0. rodiaei of South America is so hard it must be drilled before nails
or screws can be used, and is important in maritime construction. Brazilian
sassafras oil or, more correctly, ocotea oil (and it will be so designated), is
obtained from wood of 0. pretiosa and is similar in its main components to
true sassafras oil, for which it is generally substituted. The oil was initially
and incorrectly described as obtained from 0. cymbarum. There are at least
two types of 0.pretiosa in Brazil; one whose wood is rich in safrole, the other
containing methyleugenol and virtually no safrole (Mors et al. , 1959). Since
the latter is confined to Minas Gerais State and the former to Santa
Caterina, there is little local confusion over oil source. Any trace of a
cinnamon scent in ocotea oil is due to adulteration.
Production of ocotea oil is a recent development, beginning in the late
1930s and quickly reaching several hundred tonnes annually. Production is
wholly from wild trees which are abundant in Brazilian forests particularly in
Santa Catarina State. Although trees occur naturally in other South and
Central American countries, only in Brazil has it been commercially ex-
ploited for its oil. Uncontrolled cutting has resulted in depletion of suitable
trees close to distilleries, and transport costs are now affecting profitability of
the industry. Similarly to sassafras, only basic details are included.
lauraceae 199
Fig. 6.6. Ocotea pretiosa, leafy shoot and fruit. (Agriculture Department, Georgetown, Guyana.)
200 Chapter 6
Laurel
Laurus nobilis L. commonly bay laurel, is one of two members of the genus
Laurus, the other being L. azorica (Seub.) Franco. (syn. L. canariensis Webb
& Berth.), native to the Canary Islands hence the common name Canary
Island laurel, although a number of other species were formerly included in
the genus. In 1737 Linnaeus adopted the Latin Laurus for this tree and the
Greek Daphne for the quite different spurge-laurel and related species.
Laurel is widely applied as a suffix to common names of many unrelated
plants; ironically, dried laurel leaf used as a spice is called sweet bay or bay
leaf, although true bay oil and leaves are obtained from Pimenta racemosa
(Mill.), a member of the Myrtaceae (Chapter 8). Cherry laurel, Prunus
laurocerasus L., has serrated leaves which contain prussic acid and taste of
bitter almond, and mountain laurel, Kalmia latifolia L., also has toxic
leaves.
Laurus nobilis was sacred to the god Apollo in classical Greece; legend has
it that when Daphne the nymph daughter of the earth goddess Gaia was
pursued by Apollo, slayer of her bridegroom, she entreated the Gods for
assistance, who changed her into a laurel tree. Apollo then crowned himself
with a circle of laurel leaves, and declared the tree sacred to his divinity. A
garland of woven laurel leaves was awarded as a symbol of honour or victory
in Rome, but it is said that Julius Caesar preferred a crown of Alexandrian
laurel (Ruscus racemosus), as its broader leaves covered more of his bald pate!
Later in the Middle Ages distinguished men were crowned with a wreath of
berried laurel, hence for example the English title of Poet Laureate. Uni-
versity graduates were known as Bachelors from the Latin baccalaureus
(bacco a berry and laureus of laurel) and they were forbidden to marry as this
would distract them from their studies; hence the general designation in
Europe of unmarried men as bachelors. Laurel was also said to confer the
gift of prophecy, and a withering laurel tree in a garden predicted a disaster.
As a spice it is considered a vital ingredient of the genuine bouquet gami.
Various parts of laurel have long been used in herbal medicine; an infusion
of fruit was used to suppress profuse menstruation, and to hasten childbirth.
Berry oil is also believed to assist in alleviating stomach disorders, and was
also used in veterinary medicine. Infusions of bark and leaves were fre-
quently prescribed to alleviate kidney disorders and respiratory problems.
Wild trees have been harvested for centuries in many Mediterranean
countries to produce dried leaves, particularly Turkey to which laurel is
indigenous. The international trade in dried laurel leaves (bay leaves)
exceeds 2000 t with Turkey normally supplying two-thirds. Laurel leaf oil is
of comparatively recent origin, although when it was first commercially
distilled is uncertain. The oil is produced in small quantities in Spain, Italy
and Turkey, and in other European countries from imported dried leaves.
Laurel oil is currently of little importance, but has a number of advantages
over dried leaves for food manufacturers. Should demand for oil increase,
production could be rapidly expanded, as laurel is an easily established and
managed plantation crop which can be fully mechanized.
Botany
Laurus nobilis is believed to have originated in Asia minor, but is now
indigenous to many Mediterranean and European countries and widely
introduced elsewhere as it is most adaptable. A variety of L. nobilis is native
to China, but the leaf oil has a high phenolic content similar to Pimenta
racemosa. Laurel is commonly planted in gardens as a hedge or screen, and
in tubs to produce fresh leaves for culinary use. There are many horticultural
varieties, and very popular is L. nobilis Aurea, golden or variegated laurel.
202 Chapter 6
Ecology
Laurel favours regions of moderate rainfall, with long sunny periods. Al-
though tolerant of lower temperatures when mature, it is not generally frost
resistant and young trees and seedlings usually succumb to a severe frost.
Laurel is intolerant of waterlogging and, should this coincide with very cold
periods, mortality is high. Laurel generally grows below 700 m, the upper
limit being determined by the degree and occurrence of frost, and flourished
in frost free areas of the Kenya highlands between 2000 and 3000 m. Low
temperature due to altitude affects general growth of trees, can also reduce
leaf size, and may also reduce oil content of leaves. In Turkey and Greece oil
content is higher in leaves from trees growing at or near sea-level than from
those growing on inland hills (0. Anac, Istanbul University, personal
communication). Considerable regional and even district variation in leaf oil
composition exists, especially in relation to phenol and phenol-ether con-
tent, partly due to ecological factors but mainly to a dominant regional strain
Lauraceae 203
Cultivation
Laurel is generally uncultivated but can easily be grown in commercial
plantations, or existing natural stands managed on a more systematic basis
when this is profitable. Propagation is by 10-12 cm cuttings, taken from
mature sections of current season’s shoots in late summer, for planting out
the following spring. Cuttings from shoots arising from apical buds grew
more vigorously than shoots from lateral buds in Russia and Romania
(Vadachkoriya, 1984: Selaru, 1992). Cuttings root readily if correctly plan-
ted, tended, and treated with a fungicide (Ravid and Putievsky, 1984).
There is significant variation in leaf oil content and oil components between
individual trees (Kutubidze and Kharebava, 1990), thus selecting a high-
yielding parent should quickly increase average oil yield per hectare, and
204 Chapter 6
reported variation in leaf yield per tree from various countries indicates that
selection could quickly increase average yield per hectare. Harvesting is
currently mainly manual but hand-held mechanical cutters could be more
frequently employed. In Russia when laurel was grown in hedges and
mechanically harvested, one annual or three cuts every two years were the
most profitable (Chkhaidze and Kechakmadze, 199 1).
Distillation
Oil yield from fresh laurel leaves is normally 0.5-3.5% and from dried leaves
1.25-2.5%. There is little difference in the oil content of leaves dried at
temperatures of 40 to 70°C under controlled conditions, with the recom-
mended temperature 60-70°C (Skrubis, 1982), but when over-dried in the
sun, oil quality can be adversely affected. When correctly dried, there is no
significant difference in constituents and characteristics between oil from
fresh or dried leaves; it is only the yield which varies. The method of distilling
has a direct effect on oil yield and composition, but the data are conflicting.
In general, it appears that water and steam distillation or a combination
produce oils of similar composition but varying yield. Oil produced by
hydro-diffusion differs significantly in composition, having much less
1,8-cineole and monoterpenes but more monoterpenoid alcohols, acetates
and benzenoids (Boelens and Sindreu, 1986).
Litsea cubeba
The genus Litsea Lam. contains about 200 species widely distributed in
tropical Asia, Australia and the Pacific, and many members produce fruit or
seeds containing a fixed or volatile oil. In Java, the hard fat from L. sebifera
Pers. seeds is used in candles and soaps, bark from L. diversifolia Hk.f.
produces a poor quality cinnamon extract, and an essential oil is obtained
from fruit of L. cubeba.
Litsea cubeba (Lour) Persoon. (syn. L. citrata, Tetranthera polyantha var.
citrata Nees) is native to east Asia, especially China where it is known as may
chang and extensively cultivated, and to a lesser extent in Taiwan and Japan.
Litsea cubeba is also known as tropical verbena, more accurately applied to
Aloysia triphylla Britt. the lemon verbena, to which it is not related.
Litsea cubeba is a small tropical tree with fragrant flowers and leaves, which
emit a lemony odour when crushed. The small fruit resemble cubeb pepper,
hence the species name. Steam-distilling fruit produces an essential oil with
a high citral content, to 85%; specific gravity (26°C) 0.866-0.893; optical
rotation about +7.0"; refractive index 1.46-1.48. The oil is a pale yellow
mobile liquid with an intense, lemony, fresh-fruity odour, sweeter than
lemongrass but not so tenacious. China is the main producer and consumes
a significant proportion domestically (Liu et al., 1990). An oil can also be
distilled from bark whose main component is citronellal, and from leaves
with the main component cineole, but both are of no commercial im-
portance.
References
Anac, 0. (1986) Essential oil contents and chemical composition of Turkish laurel
leaves. Perfum. Flav. 11(9,73-75.
Angmor, J.E., Dicks, W.C. and Santra, D. K. (1972) The essential oil components
of C. zeylanicum grown in Ghana. Planta Medica 21, 416-420
Angmor, J.E. et al. (1 975) Chemical changes in cinnamon oil during its preparation.
J. Pharm. Pharmacol. 27, 89pp.
208 Chapter 6
Anon. (1976) Chinese Herbology, Vol 1 2 . Nanking Herb Institute Kiangsu, People’s
Publisher, Beijing, China
Anon. (1976b) Major Economic Trees ofSouthern China. Agricultural Printer, Peking,
China.
Arasaratnam, S. (1978) Francois L‘alentijn’s Description of Ceylon. The Hakluyt
Society, London, UK.
Araujo de, V. et al. (1972) Oleos essencias da Amazonia contendo linalol. Ann. Acad.
Brazil. Cienc. 44, 31 7-3 19, Suppl.
Asakawa, Y. et al. (1971) Chemical components of the benzene extract of C. loureirii.
Flav. Znd. 2, 114-119.
Bagaturiya, N. Sh. et al. (1985) Changes in essential oil composition of laurel leaves
during its accumulation. Subtrop. Kul’tu y 5, 103-1 06.
Baja-Lapis, A. (1979) Production of cinnamon barks for local consumption and
export. Canopy Philippines No. 8 , 5-6.
Bakker, M.E., Gerritsen, A.F. and van der Schaaf, P J . (1992) Leaf anatomy of
Cinnamomum, with special reference to oil and mucilage cells. Blumea 37( l),
1-30.
Bernard, T. et al. (1989) Extraction of essential oils by refining of plant materials.
Flav. Frag. J. 4(2), 85-90.
Boelens, M.H. and Sindreu, R.J. (1986) Chemical composition of laurel leaf oil
obtained by steam distillation and hydrodiffusion. In: Prog. Ess. Oil Res. W. de
Gruyter, Berlin, Germany, pp. 99-1 10.
Bottero, J. (1 985) The cuisine of ancient Mesopotamia. Biblical Archaeologist (3),
36-47.
Bradu, B.L. and Sobti, S.N. (1988) C. tamala in X.W. Himalayas. Znd. Perfum.
32(4), 334-340.
Brasil, G.A. et al. (1 979) 0 oleo essencial de sassafras no Rio Grande do Sul Partel.
Trib. Farmaceut. 47(1), 3-6.
Bregvadze, M A . et al. (1975) The activity of endogenous growth regulators in
Cinnamomum spp. in relation to frost hardiness. Soob. Akad. Nauk. Gruzinskoi.
79(1), 153-156.
Bretschnider, E.(1895) Botanica Sinicum. Notes on Chinese Botany from Native and
Western Sources. Pt 3. Kraus Reprint 1967, Nendeln, Lichtenstein.
Brown, E.G. (1955-1956) Cinnamon and cassia: sources, production and trade.
Col. Plt Anim. Prod. 5, 257-280; and 6, 96-1 16.
Buccellato, F. (1988) Bois de rose oil. Perfum. Flav. 13(5), 35-36.
Buttery, R.G. et al. (1974) Californian hay oil. J. Ag. Food Chem. 22(5), 733-737.
Charles, D.J. and Simon, J.E. (1992) Essential oil constituents of 0. kilimandschar-
icum. J. Ess. Oil Res. 4(2), 125-128.
Chkhaidze, D.Kh. (1988) Leaf productivity and quality of bay laurel. Subtrop.
Kul’uy 3, 117-124.
Chkhaidze, D.Kh. (1990). Basic conditions for obtaining high yields on heavy clay
soils in the Kolkhida lowlands Subtrop. Kul’tuy 1, 33-39.
Lauraceae 209
Hirota, N. and Hiroi, M. (1 967) Later studies on the camphor tree, on the leaf oil of
each practical form, and its utilization. Pe$ Ess. Oil. Record. 58(6), 364-367.
Hokwerda, A.H. et al. (1982) Composition of essential oils of L. nobilis, L. var.
angustifolia and L. azorica. Planta Medica 44(2), 116-1 19.
Huntingford, G.W.B. (1976) The Periplus ofthe Erythrean Sea. The Hakluyt Society,
British Library, London, UK.
Isdyoso, S.H. (1977) Kayu manis-cinnamon. Pemb. Lem. Pew. Tuna. Zndustm. Bogor.
25, 77-85.
Kannan, K. and Balakrishnan, S. (1967) A note on the viability of cinnamon seeds.
Madras Agric. J. 54, 78-79.
Keimatsu, K. (1906) Uber das aetherische Oil von C.loureirii. Apoth. Ztg. 21, 306.
Kekelidze, N.A. (1987) Essential oils of L. nobilis bark and wood. Kim Pn’rod. Soed.
3, 458-459.
Kekelidze, N.A., Dzhanikashili, M.I. and Kutatelodze, V.V. (1987) Dynamics of
accumulation and formation of essential oil composition in bay laurel leaves
during ontogenesis. Fiziol. Biokh. Kul’. Rust. 19(6), 607-614.
Khana, R.K. et al., (1988) Essential oil from fruit rind of C. cecidodaphne. Znd.
Perfum. 32(4), 295-300.
Kingstone, B.H. (1966) Some Japanese essential oils. P e Ess. ~ Oil Record. 47(6),
356-358.
Krishnamoorthy, B. et al. (1988) Quality parameters of cinnamon. Znd. Cocoa Areca
Spices?. 12(2), 38.
Kutubidze, V.V. and Kharebava, L.G. (1990) Studies on the essential oil of laurel
plants growing at the Chakuskii field station. Subtrop. Kul’tury 3, 11 1-1 16.
Kya, P. and Min, N.C. (1970) Studies on some local Cinnamomum spp. J. Life
Sciences (Burma) 3, 197-204.
Laufer, B. (1 9 19) Sino-Zranica. Field Museum of Natural History, Chicago, USA.
Lawrence, B.M. (1969) Determination of the botanical origin of cinnamoms of
commerce by TL chromatography. Can. Inst. Food Technol. J. 2, 178-180.
Lawrence, B.M. (1984) Cinnamon in the Seychelles. Perfunz. Flav. 9(4), 53-56.
Lawrence, B.M. (1993) Progress in essential oils. Perj%m.Flav. 18(3), 65-68
Lawrence, B.M. (1994) Progress in essential oils. Perfum. Flav. 19(4), 33-35
Lin, Y.T. et al. (1977) The essential oil of the cineol tree. Proc. VZZth Cong. Ess. Oil.
Kyoto, Japan, pp. 180-182.
Lin, Z.K. and Hua, Y.F. (1980) Studies on the essential oil of C. paucifEorum.Acta
Bot. Sinica. 22(3), 252-256.
Lin, Z.K. and Hua, Y.F. (1987) Chemical constituents of 14 essential oils from
Lauraceae growing in Sichuan. Chem. Zndust. Forest. Prod. 7( l), 46-64.
Liu, J.Y., Hong, L. and Zang, X.J. (1990) Study of Litseu cubeba oil separation with
capillary supercritical fluid chromatography and capillary chromatography.
Sepu, 8, 237-240.
Lockwood, G.B. (1979) The major constituents of essential oils of C. cassia growing
in Nigeria. PZunta Medica 36(4), 380-381.
Lauraceae 211
Shi, W.Y. et al. (1989). Study on the chemical constituents of essential oil and
classification of types from C. camphora. Acta Bot. Sinica 31 (3), 209-214
Shintree, V.P. and Rao, B.S. (1932) Essential oil from the leaves of C. zeylanicum. J.
Indian Inst. Sci. 15A, 84-87.
Skrubis, B.G. (1982) The drying of laurel leaves P e ~ u m Flav.
. 7(6), 37-40.
Sritharan, R., Jacob, V.J. and Balasubramaniam, S. (1994) Thin layer chromato-
graphic analysis of essential oils from Cinnamomum species. J. Herbs, Spices &
Med. Plants 2(2), 49-63
Stanica, F. et al. (1992) Studies on micropropagation of laurel. Universit. Stiinte
Agron. Seria B. Hort. 35(1), 83-90.
Sthapit, V.M. and Tuladhar, P.M. (1993) Sugandha kokila, C. cecidodaphne, oil. J.
Herbs, Spices & Med. Plants 1(4), 3 1-35
Sullivan, J.H. (1980) Pesticide residues in imported spices. J. Agric. Food Chem. 28,
1031-1 034.
Talalaj, S. (1967) Essential oil of C. zeylanicum growingin Ghana. W.African Pharm.
9, 10-12.
Tothill, J.D. (1952) Agriculture in the Sudan. Oxford University Press, London,
UK.
Vadachkoriya, Ts.T. (1 984) Growth, development and productivity of young laurel
seedlings. Subtrop. Kul’tuy 4, 111-1 15.
Vadachkoriya, Ts.T. and Loladze, D.V. (1986) Progressive method of raising bay
laurel transplants. Subtrop. Kul’tuy 5, 108-10.
Villarica, A.S. (1956) A chemical study ofthe leaves of C. nobilis var. 1adu.J. Philipp.
Pharm. Ass. 43, 145 and 42,90.
Wang, C.L. and Shao, B.B. (1984) A preliminary study of seed dormancy and
germination of C. canzphora. Plant. Physiol. Comm. (China) 1, 29-30.
Weiss, E.A. (1983) Oilseeds. Longman, Harlow, Essex.
Wijayaratne, R. and Pieris, N. (1 98 1) Cultivation of Essential Oil Bearing Plants.
Booklet 6. Natural Products Section. CISIR, Colombo, Sri Lanka.
Wijesekera, R.O.B., and Jayewardene, A.L. (1974) Chemical constituents of the
volatile oil of a rare variety of cinnamon. J. hratn. Sci. Council Sri Lanka 2(2),
141-146.
Wijesekera, R.O.B. and Ratnasingham, K. (1975) Essential oils, Ir: Improved
technology in the field distillation of cinnamon leaf oil. J. Nut. Sci. Council Sri
Lanka 3, (2) 109-1 15.
Wijesekera, R.O.B., Jayewardene, A.L. and Rajapakse L.S. (1 974) Volatile con-
stituents of leaf, stem and root oils of cinnamon. J. Sci. Fd. Agric. 25,
1211-1220.
Wijesekera, R.O.B. et al. (1976) Cinnamon. Natural Products Monograph No. 1,
ISIR, Colombo, Sri Lanka.
Yaacob, K.B. (1980) The essential oil industry and its potential role in the develop-
ment of Malaysia. In: Proc. 8th Int. Cong. Ess. Oils. Grasse, France, pp.
677-680.
lauraceae 213
Zhu, L., Ding, D. and Lawrence, B.M. (1994) The Cinnamomum species in China.
Perjiuum. Flavour. 19(4), 17-22.
Myristicaceae
7-
The order Magnoliales contains several major families including the Myr-
isticaceae of 18 genera, Myristica L. being the largest with 70-75 species.
Two others, Pycnanthus Warb. and Virola Aubl., contain species of eco-
nomic importance; P. angolensis Warb. a tropical hardwood exported from
West Africa as ilomba, and V. surinamensis Warb. and related species
exported from South America as virola.
The Myristica are normally evergreen trees of tropical lowland rain forests
from India and Sri Lanka eastwards through South East Asia to Taiwan, the
Pacific islands and Australia. Papua New Guinea is considered to be centre
of origin and distribution of the genus with some 40 known species, 34 of
which are endemic. A number of Myristica species are cultivated on a minor
scale, but only M . fragrans Houtt., nutmeg, is grown commercially. Seeds of
many species are locally utilized, and it is possible some could be marketed
for their individual characteristics if regularly available in sufficient quantity,
since there is now considerable consumer demand for a greater range of
natural flavourings. Seed and arils from fruit of M . succadanea Blume.,
cultivated in hill regions of the Molucca Islands, are sold and exported as
nutmeg and mace, while seeds and arils from the Papua nutmeg M . argentea
Warb., and the Bombay or false nutmeg M . malabarica Lam., are used as
adulterants of genuine nutmeg products. The Papua nutmeg has a peculiar
odour and rank flavour, but the Bombay nutmeg has little odour or taste.
Myristica canarica Bedd., which grows wild in western India, has seed with a
very high fat content and, when crushed and inserted in hollow bamboo, is
used as candles. Seed of M . laurifolia Warb., wild in India and Sri Lanka, and
M . elliptica Wall. in Malaysia, are locally used as nutmeg substitutes.
Myristica muelleri, which is native to northern Australia, bears a nutmeg-
flavoured fruit used as a substitute by early European settlers. Myristica
castaneaefolia Grey which grows wild in Fiji and other Pacific Islands is
locally used to treat diarrhoea.
A relative in the Annonaceae, Monodora myristica Dunal., the calabash or
African nutmeg native to west and east Africa, has seeds which resemble
214
Myristicaceae 215
nutmeg in odour and flavour. The seed is collected for sale in local markets,
especially in Nigeria, and substantial amounts are involved nationally
(Onyenekwe and Ogbadu, 1993). A member of the Lauraceae, Cyptocaya
moschata Nees.& Mart. the Brazil nutmeg, is similarly used in South Amer-
ica. The so-called clove-nutmeg is obtained from Ravensara aromatica
Gmel.; the Californian nutmeg is Torreya californica Torr. Use of the term
nutmeg without qualification in the text refers to M . fragrans and its
products.
Nutmeg
dominate for nearly a century. The Portuguese were replaced in the sev-
enteenth century by the Dutch, who in turn monopolized the trade for
another 200 years.
The Dutch endeavoured to limit nutmeg production to Banda and
Amboina by forcible destruction of all other trees and by 1650 this was
generally effected. Many writers commented that the scheme was thwarted
by fruit pigeons which swallowed seeds and voided them on neighbouring
islands, but the original source of this statement was not identified and the
reports discredited. Fruit pigeons are, however a major dispersal agent for
Virola surinamensis; birds swallow whole ripe seeds, digest the aril and void
the seed at a distance from the parent tree (Howe and Kerckhove, 1980). In
addition to limiting production in the Moluccas, the Dutch East India
Company in 1735 burnt 570,000 kg of surplus nutmegs in Amsterdam to
successfully maintain a high price. Supplies of mace were also deliberately
restricted, and the 1806 London price was 85-90 shillings per pound plus
import duty of 7s Id per pound.
The first attempt to break the Dutch monopoly was by the French on
Mauritius, then Ile de France, when Pierre Poivre obtained 32 nutmeg
plants in the Far East and arrived at Port Louis in December 1753 with only
five survivors. Later, Provost in the ship 1’Etoile du Matin was able to collect
quantities of nutmeg seeds and seedlings and a smaller number of clove
seedlings on the island of Begy. He returned to Ile de France where their
arrival has been graphically described (Ly-Tio-Fane, 1958). Despite an
export ban, plants from his expedition were taken to the Seychelles, Reunion
(Ile de Bourbon) and Cayenne. A later governor, Cere, discovered nutmeg
was unisexual. According to a contemporary writer ‘The first French nut-
meg was picked in December 1778 in the presence of a distinguished
assembly and Cere afterwards gave a magnificent reception to celebrate the
event’. Nutmeg was introduced to Zanzibar in 1818 from Mauritius or
Reunion but, although trees thrived, nutmeg never became as important as
cloves.
During the British occupation of the Moluccas (1 796-1 802), the Hon-
ourable East India Company sent their botanist, Christopher Smith to
collect seedlings of nutmeg and clove to establish the spice in Penang and
other countries under British control. Smith trained at Kew and had sailed
with Bligh on the Providence. By 1798 there were 600 nutmeg and some
clove trees growing in Penang; two years later 5000 nutmeg plants and
15,000 cloves arrived, more in following years, until by 1802 Smith had sent
71,265 nutmeg and 55,265 clove plants to Penang. The first Penang tree
fruited in 1802 but, in general, trees were not well tended and in 1805 the
company sold their spice gardens, then containing 51,000 nutmeg and 1600
clove trees. Penang nutmeg and mace were considered superior to Mo-
luccan products, a reputation maintained until the industry declined. Sir
Stamford Raffles, who founded Singapore in 18 19, supported local nutmeg
Myristicaceae 217
planting but trees did not thrive. Most trees in Penang and Singapore were
destroyed by disease between 1859-1886, and although cultivation in
Penang was later revived, commercial production virtually ceased early this
century (Burkhill, 1966).
Introduced into Sri Lanka in 1804, nutmeg flourished and continues to do
so as a minor crop. Nutmeg was taken to the Caribbean island of St Vincent
in 1802, probably from Kew, and was among the first plants transferred to
Port-of-Spain Botanic Gardens, Trinidad, in 1824. Planted on Grenada in
1843, the first crop large enough to influence the world market was exported
in the mid-l860s, and Grenada continues to be a leading producer of
nutmeg and mace. Mace is more expensive than nutmeg and there exist
much-quoted anecdotes concerning British and Dutch colonial officials in
Europe, who unaware that both spices come from the same tree, requested
growers in Indonesia and Grenada to reduce nutmeg and increase mace
production!
Nutmegs have long been used in herbal remedies, in medicine, and
preserved in syrup were formerly a delicacy in Europe, but became un-
popular when their toxic effects were generally appreciated. Nutmegs have
long been used as a narcotic but the first recorded hallucinogenic effect was
by Lobelius in 1576, who in his Plantarurn seu Stiripium Historia described a
pregnant English lady who ‘became deliriously inebriated after eating 10-1 2
nutmegs’ apparently to induce an abortion. The physiologist J.E. Purkinje in
1829 ate three nutmegs and described the effects as similar to Cannabis
intoxication, including disorientation, hallucinations and later a deep sleep.
Subsequently a number of descriptions of nutmeg-induced poisoning, delir-
ium or hallucinations were published. The response to nutmeg intoxication
is extremely varied, some individuals experience a profound distortion of
time and space and have visual hallucinations, and freshly grated nutmeg
produces the most profound intoxication (Weil, 1965). When these effects
were first noted and how used is uncertain, none suggest a religious motive.
A Materia Medica published in Bombay in 1883 stated the Hindus of West
India take Myristica as an intoxicant, while an Ayurvedic name for nutmeg is
made shaundu, narcotic fruit. Nutmeg powder is mixed with betel nut and
snuff in certain parts of southern India, and in Indonesia powdered nutmeg
is also used as a snuff. On Zanzibar and Pemba the author found local
women chewed nutmegs (kungumanga) as an alternative to smoking the
local bhang (marijuana), with almost the same effects.
The aromatic ethers are the most likely source of hallucinations from
either M . fragrans or M . malabarica, but the mode of action remains obscure.
Myristicin constitutes about 4% of nutmeg oil, and 25% of this fraction is
elemicin, which degrades to two potent hallucinogens, TMA (trimethoxy
amphetamine) and MMDA (3-methoxy-4, 5-methylenodioxy ampheta-
mine) by becoming ammoniated in the body. Crude nutmeg and myristicine
(a synthetic) both produce a degree of monoamine oxidase inhibition in vivo
218 Chapter 7
Botany
Myristica fragrans Houtt. (syn. M . oficinalis L.f.; M . moschata Thunb.; M .
aromatica Swartz; M . amboinensis Gand.) Basic chromosome number of the
genus is x = 7, and nutmeg 2n = 42. There are no recognized varieties of M .
fragrans, but many local cultivars. The English nutmeg is from the Latin
muscus, via French mugue and medieval English notemuge; mace from
maccis. Spanish neuz moscada;Portuguese noz moscado, mace macca or clava;
Dutch notemuskaat, mace foelie; Malay buah pala; India jaiphal generally,
Tamil sadhi-kai; Sri Lanka sadhika; Indonesian pala, with local suffixes; bali,
mada, etc.
Nutmeg is a spreading dioecious evergreen tree to 15 m with dark green
leaves, yellow flowers without petals and large yellowish fruit (Fig. 7.1). All
parts of the tree are aromatic. The roots are generally extensive, superficial,
often partially exposed, and on light soils trees are often unstable. The trees
are normally 10-15 my but to 25 m, with one main trunk to 30 cm DBH.
The bark is greyish-black, becoming darker and fissured longitudinally with
age; when slashed it exudes a watery sap. An essential oil containing no
aldehydes can be obtained by steam distilling the bark to yield 0.15%.
Branching is extensive beginning almost at the butt; the twigs are slender,
glabrous, greyish-brown, and carry a large number of leaves which form a
dense canopy.
The leaves are alternate, glabrous, and exstipulate, on a 1 cm petiole.
They measure 5-1 5 X 2-7 cm and are elliptic or oblong-lanceolate, with the
base acute and the tip acuminate. They are coriaceous, shiny, medium to
dark green above and light green or subglaucous below. The blade has 8-1 1
pairs of slender nerves curving out to the edge with indistinct anastomosis;
reticulations are usually invisible above, but distinct beneath, forming a lax
network. The leaves are aromatic and steam distilling produces an oil similar
to weak nutmeg oil, of little commercial importance. The characteristics of
an Indonesian leaf oil were (all at 25°C); specific gravity 0.860-0.865;
optical rotation - 1' to + 4 O ; refractive index 1.471-1.473; acid number
0.29-0.80; ester number 1.1-5.1; after acetylation 16.2-30.1; solubility in
90% alcohol 1: 1; yield 1.3-1.4% (Rusli and Murdjanah, 1977). Distillation
time affects oil characteristics; for example specific gravity changed from
0.8606 to 0.8630 after 3 and 9 hours respectively. Where it is profitable to
distil leaf oil, fallen leaves are collected, as in Indonesia, where individual
Myristicaceae 219
trees shed 0.7 kg dry leaves weekly yielding 1.5-3.0% oil. Dried leaves
contained 10% myristicin and 80% alpha-pinene (Meyer, 194 1).
The tree is typically dioecious, a proportion bear male and female flowers,
more rarely hermaphrodite flowers, but mature male trees occasionally
produce female flowers and may eventually become female. There is no
satisfactory method of determining tree sex until flowering, but it has been
claimed nutmeg seedlings can be sexed with reasonable accuracy by a
colorimetric test using ammonium molybdate as a reagent (Phadnis and
Choudhari, 197 l), or by the shape of oxalate crystals in leaves (Nayar et al.,
1977).
Male and female inflorescences are similar; glabrous, axillary, and carry-
ing 1-10 flowers in umbellate cymes. The main axis measures 1.0-1.5 cm
and is not branched or rarely branched more than twice. The pedicels are
pale green, 1.O-1.5 cm, with a minute caducous bracteole a t the flower base.
The flowers are fragrant, creamy-yellow, waxy, fleshy and glabrous (Fig.
7.2). The calyx is bell-shaped, nectiferous at the base, with reflexed triangu-
lar lobes. Petals are absent. T h e male flowers are smaller than the female; the
androecium to 7 mm, glabrous, stalk 2 mm, apex acute, 8-10 stamens with
anthers adnate to a central column and attached to each other by their sides.
Female flowers are up to 1 cm, with the ovary sessile, puberulous, superior,
one-celled, to 7 mm, and surmounted by a very short, white, two-lipped
Fig. 7.1. Myristica fragrans, nutmeg. A, Flowering shoot of male tree; B, male flower with part of calyx
removed; C, female flower with part of calyx removed; D. female flower in longitudinal section; E, shoot
with dehiscing fruit; F, seed surrounded by aril; G, seed in longitudinal section. (Purseglove eta/.,
1981 .)
220 Chapter 7
stigma: anthesis occurs in the early morning from 3-5 a.m. T h e floral
biology and histology of flowers has been studied in detail (Armstrong and
Drummond, 1986). An essential oil of no commercial importance can be
extracted from flowers. Trees may flower throughout the year, but there are
normally two peaks; July and October in Kerala, India. Insects are the
primary pollinators and excluding them from a female tree prevented
pollination and fruit set on Grenada (Cruickshank, 1973); in India, small
beetles were particularly important (Armstrong and Drummond, 1986) and
in some areas a specific insect may be pre-eminent, Formicomus braminus in
Kerala.
T h e fruit is a yellow fleshy drupe resembling a large apricot, usually
pendulous, broadly pyriform, smooth, 6-9 cm, with a circumferential longi-
tudinal ridge and persistent remains of the stigma, and containing one seed
(Fig. 7.3). When ripe the succulent, aromatic pericarp, about 1.3 cm thick,
splits along the suture to expose the purplish-brown lustrous seed (nutmeg)
enveloped in a crimson, fleshy, lacinated membrane, the aril (mace) (see
Fig. 7.4). The pericarp, or pod, is used in Indonesia, Malaysia and Singa-
pore for jellies and preserves. Also in Malaysia piles of rotting pods provide
excellent material to grow the very popular mushroom kulutpala, and pods
are in great demand for this purpose. The seed is ovoid or ellipsoidal,
2.0-4.0 X 1.5-2.5 cm, with considerable variation in shape and size between
individual trees. Trees bearing different types of fruit in the Moluccas are
distinguished by a suffix; pala bali has large globoid fruit, pala tidore has pear-
shaped fruits (Nitta, 1993).
The seed surface is furrowed, longitudinally wrinkled, the raphe groove
extending from basal scar to apical depression; it is relatively soft when fresh,
becoming hard with age or drying, and is easily cut or scraped. A cut surface
shows the pale brown endosperm marked by many veins containing the
essential oil. When the seed is halved longitudinally, remains of the small
embryo are visible. A detailed description of the morphology and histology
of seed and aril has been published (Parry, 1962). Cut seed has a character-
istic aromatic scent and a warm slightly bitter taste. Dried seed is the nutmeg
of the international spice trade, dried aril is the spice mace, and a detailed
description of the preparation of both has been published (Purseglove et al.,
1981).
Nutmegs usually contain, in per cent: water 9, carbohydrate 30, protein 7 ,
fixed oil 33, essential oil 4.5. Analyses of Indian nutmegs gave average
values, in per cent: moisture 14.3, protein 7.5, ether extract 36.4, carbohy-
drate 28.5, fibre 11.6, mineral matter 1.7, calcium 0.12, phosphorus 0.24,
Fig. 7.4. Nutmeg. L to R: dried and wet nutmeg; aril; nutmeg enclosed in aril
five aromatic ethers 2-20; main constituents of mace oil are, per cent:
monoterpene hydrocarbons 75-95, oxygenated monoterpenes and sesqui-
terpenes 4-18, aromatic ethers 0-6. Oil composition is further discussed in
the section on products and specifications.
A number of commercially accepted grades of nutmeg are recognized, and
those currently exported by the two major producers, Indonesia and Gre-
nada, are summarized as they will occur in the text. In both countries sound
nutmegs are graded as 80s or 11OS, according to size in numbers per pound;
mixtures of sizes are exported as sound unsorted. Sound nutmegs are used
mainly for grinding and to a lesser extent for oleoresin extraction in import-
ing countries. Substandard nutmegs of two types are exported from In-
donesia, sound shrivelled and BWP (broken, wormy and punky). In Gre-
nada, floats from the water-flotation process plus badly bruised and broken
pieces of sound nutmegs are combined in a defectives grade, similar to
Indonesian BWP but generally of higher quality. Indonesian sound shriv-
elled nutmegs contain a higher percentage of volatile oil than mature sound
nutmegs and are used for grinding, oleoresin extraction and oil distillation in
importing countries. West Indian defectives and Indonesian BWP grades
are seldom distilled, since both normally yield below 8% oil. Indonesia
exports two distilling grades; the poorest BIA or ETEZ has a volatile oil
content of 8-10%; the BSL or AZWI grade contains less shell material with
12-13% volatile-oil content. Grenada exports a distillation grade to the
USA consisting solely of floats. Two types of poorest quality Indonesian
nutmegs are distilled locally and are rarely seen in international trade: an
assortment of broken and decayed fragments, and Padang originating
mainly from this region of Sumatra.
Indonesian dried mace is exported as either whole or broken blades;
Grenadian cured mace as whole pale mace, No. 1 broken mace, No. 2
broken mace, unassorted and pickings. The No. 2 broken grade consists of
material which does not acquire the desired pale colour and is generally
shipped in bags. Whole pale mace and No. 1 broken mace are packed in
cases for export.
Ecology
Nutmeg is native to the wet tropics and trees thrive with a high well-
distributed rainfall and moderately high temperature with little seasonal
variation. An annual rainfall of 2200-3700 mm is considered the optimum
for high fruit yield, but trees grow and fruit well at 1500-2500 mm with good
management; below 1500 mm where irrigation is available. In regions with
a pronounced dry season, as on Grenada where there can be 2-3 months
without rain, fruiting is more seasonal than in Indonesia. A temperature of
2 5 3 5 ° C is the optimum and, although mature trees are little affected by
lower and higher temperatures for short periods, flowering can be adversely
affected by temperatures above 35"C, or a hot dry wind. Frost will cause
extensive damage at any stage and normally precludes commercial produc-
tion; thus irrigated cultivation at higher altitudes in the dry tropics has not
been successful. The superficial root system makes trees very susceptible to
wind damage unless protected by shelterbelts. The devastating effect of
cyclone Janet in December 1955 on nutmeg plantations in Grenada has
been well documented; similar effects on a smaller scale have been noted
from other cyclone-prone regions.
Nutmeg thrives under an insular maritime climate and is seldom grown on
a commercial scale above 500 m with individual trees and small plots at
much higher elevations, but there is little information on yield, since most
are grown to supply domestic needs. Sunny sheltered hill valleys often
provide ideal sites. An open situation is preferred for mature trees, but
shading seedlings promotes rapid growth, prevents sun-scorch, and young
trees are often interplanted with fast growing species such as bananas. In
more exposed situations, intermittent shade and wind protection can be
provided by lines of trees such as Albizzia spp. or Eythrina spp.
nutmeg cannot tolerate waterlogging even for short periods, especially when
young, It grows well in hill valleys but seldom flourishes on river flats. A
neutral to slightly acid soil p H 6.5-7.5 is preferable, but nutmeg will tolerate
more acid soils provided these are well managed; saline and alkaline soils are
unsuitable.
T o produce high fruit yield, nutmeg trees require a high level of natural
soil fertility as occurs in Indonesia and Grenada, since most plantations have
a relatively low level of management and a systematic fertilizer programme is
the exception. Whatever organic manure or plant residue locally available is
spread around trees and mulching with grass, straw or similar material is
common when seedlings are first planted out. Well-rotted plant residues,
animal wastes or residue from fish processing are usually placed in the pits
into which seedlings are transplanted. More efficient producers also place
these materials in a trench round larger seedlings and young trees during
growth and before the spreading root system precludes its use. Once trees
have an established root system, the material should be spread within the
canopy diameter; in India annual application of 2.5 kg cattle manure or
40-50 kg plant residue per tree is recommended, and young trees so treated
grow faster, are larger and tend to flower earlier than unfertilized trees. T h e
application of chemical fertilizers on a regular basis is uncommon although
their use is generally beneficial, but as nutmeg is usually grown on fairly
fertile soil, it is essential the nutrient status is accurately determined before
fertilizers are applied.
Little information is available on the effect of individual plant nutrients on
tree growth, fruit yield, seed size or composition. Reports indicate a sub-
stantial response to manure and fertilizers, but there are few data from
formal trials. The official recommendation per tree in Kerala State is 20 g N,
18 g P, 50 g K in the planting hole, increasing gradually each year until at 15
years the rate reaches 500 g N, 200 g P and 1000 g K per tree. How many
growers actually apply these levels is unknown, but trees receiving any
fertilizer are obvious by their greater size and higher yield. A 13:8:24 N P K
mixture at 25 giseedling increasing to 2.5 kgitree at 10 years was suggested
by the Grenada Agricultural Department as a general guide to producers,
but only progressive growers use any fertilizer.
Cultivation
Nutmeg trees were initially planted in cleared areas in forests or jungles, but
the opportunity for so doing has been drastically reduced. When planted in
the virgin soil of these clearings, trees grew virtually untended except for
periodic slashing of grass and weed growth. Today most nutmeg trees are
planted to replace those which have died, and this is reflected in the
following remarks. Nutmegs are generally grown from seed sown in specially
prepared seedbeds or containers. Seed nutmegs should be sound and fresh,
with the testa intact and aril removed; those from immature fruits, or over
226 Chapter 7
3-5 days old seldom germinate. Seeds kept in wet moss or polythene bags
remain viable for 15 days in India, but storing seed for more than a day or
two is not recommended. Seed is still widely used to produce seedlings, but
is time-consuming since nutmegs are male and female and sex cannot be
determined until flowering.
Seed should be sown 2.5-5.0 cm deep, 30 cm apart in beds; germination
is slow, between 45 and 80 days and, despite widespread belief to the
contrary, there is no difference in speed of emergence between males and
females. Seedlings are ready for planting out after 6 months or at 15 cm high;
although seedlings aged 18-24 months and 30 cm high made better growth
than smaller seedlings in southern India. Care is essential when removing
seedlings to ensure the long tap-root is not damaged. Seedlings in banana
leaf or modern biodegradable pots can be sown intact into planting holes
without root disturbance; polythene tubes should be carefully removed.
Vegetative production from female trees is preferable, and parents should
be selected for high fruit yield, resistance to a local disease or a similar
desired characteristic. Approach-grafting and marcotting are well known
techniques and easily applied by growers. Clonal propagation and tissue-
culture methods have been developed which enable large numbers of
plantlets from selected trees to be produced, but these techniques are
relatively expensive and most trees are still grown from seed.
In well established nutmeg plantations, seedlings found growing near
parent trees can be transplanted; that they are growing near a female tree
does not guarantee they will be female. Seedlings grown from seed will be
half male, and as only 10% are sufficient to ensure pollination, excess males
may be removed, or their sex changed by grafting with scions from female
trees. Spacing should be at least 9 m apart on the square and closer spacing
invariably reduces yield. Vegetatively propagated plants are placed one per
hole, but three seed-grown seedlings are planted 90 cm apart on the triangle,
and later thinned to one female, with a male left for every ten female trees:
carried out when plants flower in 5-7 years. Young plants in exposed
situations require shade, preferably established the previous season; bananas
are ideal, give shade, mulch and an income to offset establishment cost. In
Indonesia, Canariurn commune L., the kanari or Java almond is deliberately
planted to provide shade and a wind break as it bears an edible fruit, the pili
nut which also produces a cooking oil.
Plantations, once established, usually require little attention from grow-
ers, who slash weeds and carefully clear an area round young trees to reduce
insect damage. The dense canopy of mature trees normally suppresses weed
growth, but under-tree vegetation should be kept low to enable fallen fruit to
be seen easily. Small livestock often graze under trees, but goats and cattle
should be excluded since both cause extensive damage. Herbicides are used
where profitable, and directed sprays of glyphosate and simazine are recom-
mended in young plantations, but applying many common herbicides under
Myristicaceae 227
Harvesting
Fruit is ripe only when it has split, and is picked or allowed to fall and
collected daily; the latter is more economical but, in the Caribbean, picking
is becoming general. However pickers paid by weight tend to be unselective
and can also damage trees by climbing among branches. In Malaysia and
Indonesia, high fruit is picked using long poles fitted with cutters and a
basket. After harvesting, the nutmeg in its shell with the surrounding mace is
separated from the pericarp. Mace is later detached and carefully flattened
by hand or between boards to avoid breakage, which downgrades quality.
Mace is usually sun-dried on large trays or mats, and 2-4 hours is normally
sufficient in Indonesia and Grenada, mace retaining its original scarlet
colour.
Indonesian mace is usually exported in the red form, changing to reddish-
orange by the time it reaches the consumer. Grenadian mace is stored in
darkness up to 4 months, gradually becoming brittle, horny and pale orange-
yellow; it sells at a premium to Indonesian mace. After removal of the mace,
nutmegs are dried in their shells on large wooden trays, and turned daily by
stirring with wooden paddles to prevent fermentation and promote even
drying. Sun-drying for about one week is usual in Indonesia, or in special
buildings heated by slow fires to prevent shells cracking and melting the fat
which forms a large part of the kernel. O n Grenada, nuts are air-dried for
about 8 weeks in special buildings, losing about 25% by weight. Natural
drying is lengthy and tedious, and artificial forced-air bed-drying reduced
the time to 23 days using cold air, and to 7 days using air heated to 37"C,
without affecting quality (McGraw and Sankat, 1984). Practically all of
Granada's nutmegs and an increasing proportion of Indonesian are now
mechanically shelled.
Fruit yield per tree is very variable as is the average size of nutmegs, while
standard of management also has a considerable effect on yield, especially
adequate pest and disease control. Thus inter-regional comparisons are of
little value, and the following yield figures are given as a general guide. A
228 Chapter 7
mature tree in full bearing in India yields 750-2000 fruits annually, but up to
10,000 from one 25 years old. T h e Sri Lankan average is higher at
2000-4000 per year, with an individual tree yield of 12,000 fruit. In
Grenada, the annual average was 1500 fruit per tree, but in recently
established groves from vegetatively grown seedlings is almost double. The
yield of mace varies similarly, but from sound healthy nutmegs averages
15-20% by weight.
Distillation
When sound nutmegs are distilled, the fixed oil tends to retain some volatile
oil reducing the yield. Defective nutmegs either shrivelled or infested with
weevil larvae, which consume most of the fixed oil, have higher yields; they
also cost less! Nutmegs should be comminuted to a coarse powder then
transferred immediately to the still since, once ground, there is a rapid and
substantial loss of volatiles. Distillation should be with low-pressure live
steam and cohobation may be necessary. About 80% of oil distils within two,
the remainder within ten hours. High-pressure or superheated steam should
not be used, as it carries over small quantities of myristic acid from the fixed
oil. High-quality oils are produced on Grenada from defective nutmegs; a
single preliminary crushing in a roller mill was followed by water distillation
preferably without cohobation; steam distilling also yielded satisfactory oils.
By contrast, water-and-steam distillation with cohobation was not as satis-
factory; yield was variable and quality below that of water-distilled oils.
Distilling for 6-12 hours with a 45 kg charge of crushed nutmegs at a rate of
9-1 0 litres of condensate per hour was the optimum (Coward et al., 1972). In
general, oil yield should be 6-12% by weight, and average at least 8%.
Nutmeg and mace oleoresins are prepared by extracting comminuted
spices with organic solvents. Commercial oleoresins vary in their essential oil
and fatty oil content depending on the solvent; benzene extraction yields
3 1-37% oleoresin containing a substantial amount of odourless and flavour-
less material, mainly trimyristin, removed by washing with cold ethanol in
which the fat is almost insoluble. Extraction of nutmeg directly with cold
ethanol provides 18-26% crude oleoresin; on chilling, filtering and evapor-
ating ethanol under slight vacuum, the yield is reduced to 10-12%. Mace
extracted with petroleum ether yields 27-32% oleoresin containing 8-22%
volatile oil. Hot-ethanol extracts yield 22-27% crude oleoresin; after pro-
cessing 10-1 3%.
Nutmeg contains 2 5 4 0 % fixed oil obtained by expressing crushed nuts
between heated plates in the presence of steam, or by solvent extraction. The
product, concrete, expressed oil or nutmeg butter is a highly aromatic,
orange-coloured fat, with the consistency of butter at ambient temperature;
consisting mainly of trimyristin and a high proportion of volatile oil difficult
to separate by steam distilling.
Myristicaceae 229
polar non-polar
Alpha-thujene a 2.2
Alpha-pinene 22.6 22.8
Camphene 0.3 0.3
Beta-pinene 15.4 15.7
Sabinene 18.5 18.6
Myrcene t delta-3-carene 3.2 a
Alpha-phellandrene 0.9 2.4
Alpha-terpinene 3.5 3.4
Limonene 3.9 3.7
Beta-phellandrene 2.4 a
Gamma-terpinene 5.1 5.1
pCymene 1.I a
Terpinolene 2.4 1.7(b)
Alpha-copaene 0.2 0.2
Linalool 0.4 a
Terpinen-4-01 8.0 7.7
Alpha-terpineol 1 .o 1.o
Safrole 2.0 2.1
4-Pentylanisole a 0.3
Eugenol a 0.2
lsoeugenol a 0.2
Alpha-bergamotene a 0.1
Myristicin 6.9 5.3
Delta-cadinene a 0.3
Total 97.8 93.3
a, not separated; b, including linalool.
Source: Based on RIC (1984).
232 Chapter 7
References
Armstrong, J.E. and Drummond, B.A. (1986) Floral biology of Myn'stica fragrans.
Biotropica 18(1), 32-38.
Burkhill, I.H. (1966) The Economic Products of the Malay Peninsula. Govt. Printer,
Kuala Lumpur, Malaysia.
Coward, L.D.G., Mathews, W.S.A. and Nabney, J. (1972) Investigation of the
production of essential oil using a transportable still unit. Proc. 5th Znt. Cong. Ess.
Oils 1971. Amos de Acad, Brazil de Ciencias 44, 126-132.
Cruickshank, A.M. (1973) Some aspects of the nutmeg investigation programme in
Grenada. In: Proc. Cong. on Spices April 1972. pp. 105-1 11. Tropical Products
Institute, London, UK.
Emboden, W. (1979) Narcotic Plants. Macmillan Pub., New York, USA.
Flasch, M. (1966) Nutmeg Cultivation and its Sex Problem. Pub. Mededelingen van de
Landbouwhoge School, Wageningen, The Netherlands.
Gopalam, A. and Zaccharia, T.J. (1989) Lipophilic products of extractable fat of
nutmeg. J. Plant Crops 16 (Suppl), 107-112.
Gopalan, C. et al. (1971) Nutn'tive Value of Indian Foods. Health Bull. 23, Nat. Inst.
Nutrit., Hyderabad, India.
Gopalkrishnan, M. Rajaraman, K. and Mathews, A.G. (1979) Identification of the
mace pigment. J. Food. Sci. Tech. 16(6), 261-262.
Howe, H.F. and Kerckhove, G.A.V. (1980) Nutmeg dispersal by tropical birds.
Science 210, 925-927.
Ipe, C.V. and Varghese, C.A. (1990) Economics of nutmeg cultivation in Kerala. 3.
Plant. Crops 18(1), 29-33.
234 Chapter 7
Ly-Tio-Fane, M. (1958) Mauritius and the Spice Trade. Esclapan Ltd., Port Louis,
Mauritius.
McGraw, D.R. and Sankat, C. (1984) Effect of airflow rate on the drying of
nutmegs. Trop.Ap'c (Trin). 61(3), 171-173.
Meyer, T.M. (1941) Essential oil from leaves of nutmeg. Ing. Nederland-Tradie 8,
35-36.
Nayar, B.K., Rai, R. and Vatsala, P. (1977) A simple morphological technique for
distinguishing sex of nutmeg seedlings. C u r Sci. 46, 156-157.
Nitta, A. (1993) Nutmeg found in Molucca Islands. J. Jap. Botany 68(1), 47-52.
Onyenekwe, P.C. and Ogbadu, J.C. (1993) Volatile constituents of the essential oil
of Monodora myristica.J . Sci. Food Agric. 6 1, 379-38 1.
Parry, J.W. (1962) Spices: Their Morphology, Histology and Chemistry. Chemical Pub.
Co., New York, USA.
Phadnis, N.A. and Choudhari, K.G. (197 1) Sex determination in the seedling stage
of nutmeg. Trop. Sci. 8, 265-267.
Prakashchandra, K.S. and Chandrasekharappa, G. (1984) Lipid profile and fatty
acid composition of fat extracted from arils (mace) of M . fragrans and A.
hirsutus.J. Food Sci. Tech. 21 (l), 40-42.
Purseglove, J.W. et al. (1981) Spices. Vol. 1. Longman, London, UK, pp.
174-228.
Purseglove, J.W. (1968) Tropical Crops - Dicotyledons. Longman, Harlow, UK.
RIC (1984) Royal Institute of Chemistry: monographs for seven essential oils.
Analyst 109, 1343-1360.
Rusli, S. and Murdjanah, N. (1977) The effect of grid and distillation time on the
yield and properties of nutmeg leaf oil. PLPT Zndustri (Bogor) 24(1), 11-20.
Sandford, K.J. and Heinz, D.E. (1 97 1) Effects of storage on the volatile composition
of nutmegs. Phytochem. 10, 1245-1250.
Sarath-Kumara, S.J., Jansz, E.R. and Dharmadasa, H.M. (1985) Some physical and
chemical characteristics of Sri Lankan nutmeg oil. J. Sci. Food Ap'c. 36,
93-100.
Weil, A.T. (1965) Nutmeg as a narcotic. Econ. Bot. 19, 194-217.
Myrtaceae
8-
T h e Myrtaceae consists of some 75 genera and nearly 3000 species of mainly
tropical evergreen trees and shrubs. T h e chief centres of distribution are the
American and Asian tropics and Australia. A major revision placed many of
the Asiatic Eugenia L. in the genus Syzygium Gaertn., which contains about
500 species mainly in the Asian tropics, and the most important spice and
essential oil source is clove, S. aromaticum L. T h e genus Eucalyptus L’Herit.
is also large, containing about 700 species almost entirely native to Australia;
many produce an essential oil but few are commercially exploited. The
genus Melaleuca L. has some 200 species also mainly native to Australia, but
only half a dozen have been exploited for their essential oil. T h e genus
Leptospermum Forst. contains about 80 species mainly native to Australia;
one is cultivated for its essential oil. The genus Pimenta Lindl. contains 18
species native to tropical America and the Caribbean region; two produce a
commercially important essential oil. The genus Myrtus L. contains about
100 species mainly native to South America, but only M . communis L.
produces a commercial oil.
The Arabs knew clove as karanful, believed to be the origin of the Greek
Kayophyllon. T h e modern English name of clove is from the French clou
meaning nail, probably derived from the Latin clavus, while the original
Chinese name translates as sweet-smelling nails. T o avoid unnecessary
wordage, clove will be used to designate clove tree, harvested buds and the
spice, the context ensuring no confusion.
The clove is indigenous to the Moluccas (the Spice Islands) now part of
the Republic of Indonesia, but apparently the buds were little used as a spice
or flavouring. The first recorded use was in the Chinese Han period BC
220-206, when courtiers sweetened their breath with clove buds in the
235
236 Chapter 8
clove trade. Based on Batavia, now Jakarta, the Dutch clove trade is
described in detail by Rumphius (HerbariumAmboinense, 1750). The whole-
sale destruction of wild and cultivated trees by the Dutch resulted in a great
loss of genetic diversity, a lack noted by Rumphius in 1741, when he
described only three types of trees differing mainly in size and colour of ripe
cloves; he also stated that the locals believed that cultivated cloves, if left
untended, reverted to the wild type. On Amboina island, in a protected
enclosure, is a tree believed to be at least 300 years old, reputedly planted by
locals in an attempt to evade the Dutch monopoly. Indonesia is once again
the world’s largest producer and consumer of cloves, mainly used in kretek
cigarettes. Clove production rose slowly but steadily in the decade
1952-1962 from 3000 t to 7000 t, but following the 1964 revolution on
Zanzibar (later noted), clove production in Indonesia rapidly increased and
by 1972 was 15,000 t, reaching 25,000 t in 1982, according to figures
released by the government Statistical Bureau (Biro Pusat Statisik, Dja-
karta), Domestic clove production was subsequently related directly to the
use of clove powder in kretek cigarettes, reaching 30,000 t in 1992; since
then there has been a substantial internal surplus of cloves. Reviews of the
trade in cloves and clove derivatives to 1970 is contained in Willems (1 97 1)
and to 1973 in Adamson and Robbins (1975).
Clove and nutmeg were introduced into Mauritius in 1770 from seeds
smuggled out of the Moluccas for Governor Pierre Poivre and the first crop
of cloves was picked in 1776. Seedlings were later distributed to other Indian
Ocean islands under French administration, including the Seychelles and
Reunion. Few original seedlings survived and it is locally recorded that one
tree on Reunion supplied seed from which are descended practically all trees
on Reunion and Madagascar. Cloves were found growing on Papua New
Guinea in the 1770s and were probably also introduced to neighbouring
islands (Sonnerat, Voyage en Nouvelle Guide, 1776).
Clove was first planted in Malaysia in 1786 but was not successful. Trees
from a second introduction in 1880 flourished and a small but thriving
population remains in the Penang region. Seeds from Reunion were planted
on Sainte Marie Island off the Madagascar coast in 1827, thrived and
produced about 15 t of cloves by 1880. Plantings expanded after 1885,
cultivation was extended to mainland Madagascar in 1890, and that country
is now a major producer of cloves and exporter of clove leaf oil.
Possibly the most important introduction of cloves to any country was to
Zanzibar early in the nineteenth century. The most reliable account tells
how a local Arab, Harameli bin Saleh, banished by the Sultan for murder,
took service with a French officer and obtained clove seeds in Reunion.
These he presented to the Sultan, Sayyid Said Bin Sultan, and obtained his
pardon. The first plants were grown near the royal palace of Mtoni, 6 k m
north of Zanzibar town. The Sultan realized the economic potential of the
238 Chapter 8
crop and forced local landowners to plant cloves under threat of confisca-
tion. Substantial areas were planted on Zanzibar and Pemba, helped by the
large number of slaves (Fig. 8.1). The 1872 cyclone caused widespread
destruction of trees on Zanzibar, but did not affect Pemba. The Sultan of the
day, Sayyid Barghash bin Said, enforced replanting on Zanzibar while high
prices encouraged additional planting on Pemba. Zanzibar (now part of
Tanzania) later became the world’s largest exporter of cloves but, following
the revolution of 1964, production slumped. Land was nationalized, planta-
tions were split into uneconomical 1.5 ha plots and the government paid
growers only 4% of the export price. This was below the cost of production,
there was no incentive to maintain healthy trees or replant, and the majority
are over 60 years old. Production fell to less than 1000 t, rising slowly to
3000 t i n 1995-1996, well below the pre-1964 average of 9000 t, and far less
than the annual 20,000 t of the decades to 1960. New plantations estab-
lished on the Tanzanian mainland near Tanga are now in bearing. Relative
newcomers to the clove trade are Sri Lanka and Brazil but, with the world
surplus of cloves likely to continue for the foreseeable future, both are
unlikely to expand production beyond domestic demand.
When clove oil was first produced is uncertain, and initial mention of an
elixir of cloves was for the treatment of toothache, probably a direct result of
sucking dried cloves which not only sweetened the breath but acted as a
palliative. An extract was used as a prophylactic against skin and respiratory
diseases in the Middle Ages. The major constituent of clove oil was first
described by Liebig in 1836, as nelkensaure or acid of cloves, subsequently
eugenic acid, and in 1875 as eugenol by Johann Karl Tiemann of Berlin,
who used it to produce synthetic vanillin U. Chem. Soc. 1900, 600).
Several other unrelated species produce either fruit or an essential oil
popularly called cloves or clove oil, and two are well-known members of the
Lauraceae. Ravensara aromatica Gmel, the Madagascar clove-nutmeg, is a
medium-sized tree to 12 m, with small leathery leaves and small round
aromatic fruit whose taste resembles a mixture of clove and nutmeg; oil
distilled from its leaves is commonly an adulterant of true clove leaf oil.
Dicypellium cayophyllaturn (Mart.) Nees, the Brazil clove, bears highly
aromatic flower buds very popular locally for flavouring, which are harvested
and dried in a similar manner to cloves. Oil is distilled from wood and bark,
known as clove-cassia or clove-bark oil and exported to Europe, especially
France.
Botany
Syzygium aromaticurn (L.) Merr. & Perry (syn. Eugenia arornatica Kuntze; E.
ca yophyllata Thunb.; E. calyophyllus (Sprengel) Bull. & Harr.; Cayophyllus
aromaticus L.). The genus Syzygium Gaertn. includes some 500 species
native to the Asiatic and African tropics and Australia. In addition to
essential oils and spices, some members of the Myrtaceae are cultivated for
Myflaceae 239
their fruit including the popular guava Psidium guajava L. The clove tree was
long known as Eugenia aromatica or E. caryophyllata, but revision of the large
Eugenia genus, as noted, resulted in clove being reclassified (Schmid,
1972).
The clove’s nearest relative is the wild clove common on the forested
lower slopes of the Moluccas and Papua New Guinea, which differs from
cultivated clove in having larger, less aromatic leaves and flower buds. Also,
the essential oil content is lower and the oil has different characteristics. A
potentially valuable feature of these wild species is resistance to Matibudjang
disease (q.v.). Hybrids between wild and cultivated cloves are fertile but
have not been studied cytologically, and it is not certain whether the wild
clove should be treated as conspecific or as a separate species, E. obtusijolia
Roxb. Thus it is interesting to note the findings of Dr Meijer in Indonesia,
who over many years collected buds from wild trees and analysed their oil
(Meijer and Schmid, 1948). These oils lacked eugenol but did contain
eugenone, eugenin, isoeugenitin and isoeugenital, with an odour quite
different to oil from cultivated cloves. They also speculated on whether
cultivated clove evolved by selection to the present type, or if there were
initially two quite different wild types. No true cultivars of clove are recog-
nized in Zanzibar and Madagascar since almost all existing trees are descen-
dants of the original limited introductions; there is thus little genetic variabil-
ity or opportunity for selection. Greater variability exists in cloves’ native
Moluccas and probably also elsewhere among wild and cultivated types.
Cultivated clove is an evergreen tree to 15 m, generally conical when
young becoming more cylindrical with age, and bearing glossy green leaves,
fragrant red flowers and purple fruits. It is long-lived and recorded as
remaining productive for 150 years. The seedling has a pronounced but
short tap-root, which is quickly replaced by two or three primary sinkers. In
the first year, a mass of fibrous roots spreads out from the tap-root to a radius
of 30-50 cm and a depth of 25 cm; during the second year these fibrous roots
thicken to become main horizontal laterals. The roots finally extend to a
radius approximately equal to tree height and, when roots of neighbouring
trees overlap, natural grafting can occur. The trunk, to 30 cm DRH, often
forks near the base into two or three erect branches, a habit sometimes
simulated by planting two or three seedlings close together. The bark is grey,
smooth to rough depending on age, and slash on a healthy tree is white to
rose-pink.
The wood is very hard, and from a tree which has died from natural causes
is ash-grey; from a tree which died from sudden death it is yellow, while
infection from die-back produces a reddish-brown discoloration. The main
branches are nearly upright with few side branches, giving a cylindrical
appearance to mature trees. Smaller branches and twigs are also ascending,
terete, very brittle and greyish-white. During harvesting in Zanzibar, pickers
frequently climb trees, damaging these brittle branches and leaving open
wounds, allowing entry to the pathogen causing ‘die-back’ (see section on
pests and diseases).
The leaves are simple, opposite, coriaceous, exstipulate, glabrous and
aromatic. The lamina are lanceolate or narrowly elliptic, sometimes nar-
rowly obovate, 7-1 3 X 3-4 cm, gland-dotted, densely and obscurely pinna-
tinerved. The apex is shortly or broadly bluntly acuminate, the base cuneate
and the margin wavy and recurved. The petiole is slender, 2-3 cm, some-
what swollen, the base pinkish and the leaf blade partly decurrent in the
Myrtaceae 24 1
upper portion. New bright pink leaves appear in flushes; the upper surface
later becomes glossy dark green, the lower dull and paler. Crushed leaves
emit the characteristic clove scent and, as the volatile oil also evaporates,
clove plantations after rain are permeated with the rather cloying odour.
The inflorescence is a terminal, corymbose, trichotomous panicle, shortly
pedunculate, branched from the base, shorter than the leaves and very
variable in number of flowers; from three on a simple three-forked peduncle
to 50 or more on multiple peduncles (Fig. 8.2.). The angled peduncles and
shorter 5 mm pedicels constitute the clove stems of commerce. The bracts
and bracteoles are narrow, acute, to 2-3 mm, and are quickly shed. The
flower is hermaphroditic with a fleshy hypanthium surmounted by sepals;
the hypanthium is 1.O-1.5 cm, cylindrical, angled, green in the young bud
but flushed pink at anthesis, becoming deep reddish after stamens fall. The
four calyx lobes are fleshy, triangular, slightly incurved, 3-4 mm, and easily
observed in the spice. The inflorescences are harvested when the buds have
reached full size but before they open, thus the petals together with the
enclosed stamens form the head of the dried clove. The four petals are
imbricate, tinged red, rounded, 6 mm diameter, falling as a hemispherical
calyptra as the flowers open. The stamens are numerous, the anthers are pale
yellow and ovate; the style is very stout to 3-4 mm, with the base swollen,
and pale green. The two-celled, multi-ovulate, inferior ovary is embedded in
the top of the hypanthium. The morphology and histology of clove flower
and bud has been described (Parry, 1962).
The fragrant flowers are attractive to and regularly visited by bees and
other insects which are probably the main pollinating agents. Controlled
cross-breeding is also possible (Nair et al., 1974). In Indonesia clove is
locally considered to be self-pollinated but bagged inflorescences in Zanzi-
bar never produced viable seed (Zanzibar Department of Agriculture,
1949-1965). Flowers covered with polythene bags in Sri Lanka set 30%
seed, a percentage not significantly increased by artificial pollination (Sri-
tharan and Bavappa, 1981). Few flowers develop fully to form fruit, the
greater number fall and can be gleaned and processed into an inferior spice.
The effect of growth regulators on clove has been little studied; a Zanzibar
unreplicated trial using NAA applied in December caused a yield increase
the following December (Zanzibar Department of Agriculture, 1949-1 965).
There are two flowering seasons on Zanzibar, July-September and
November-January; in Madagascar, the main flowering is July-August; in
India, September-October in the plains, December-January in the hills; in
Indonesia, October-February depending on the area. Bud initials appear
about 6 months before they are ready to be harvested, and six stages with
local names are recognized in Zanzibar. It is thus possible to make a
reasonable crop estimate some time before harvest. A profuse flowering
leading to a bumper crop occurs every 4 years in Zanzibar and Indonesia.
242 Chapter 8
Fig. 8.2. Syzygiurn aromaticurn, clove. A, Flowering branch: 6, bud in longitudinal section: C, flower in
longitudinal section; D, fruit; E,fruit in longitudinal section; F, dried clove. (Purseglove et al., 1981.)
Myrlaceae 243
Clove oil is normally obtained from dried buds and published analyses
relate to this oil type. Oil from fresh buds at different stages of maturity
varies considerably in physical characteristics, composition and odour. Data
from India are shown in Table 8.1; subsequent analyses confirmed that
eugenol content increased with maturity while eugenol acetate decreased;
caryophyllene is apparently absent in fresh buds (Gopalakrishnan et al.,
1982). Oil from fresh buds had a pleasant fruity odour, while oil from young
buds was more mellow than from mature buds.
Fruits, mother-of-cloves, are oblong fleshy drupes, shortly tapering at
each end, reddish purple, 2.5-3.5 cm long by 1.2-1.5 cm in diameter,
surmounted by four enlarged fleshy calyx lobes, and there are normally
350-375 ripe fruit per kilogram. The fruit contains a single seed, rarely two,
with a thin fleshy pericarp 2-3 mm thick. The 2 cm, oblong purplish seed is
rounded at both ends, with two large cotyledons and no endosperm; the
cotyledons lie side by side, the inner faces folded and interlocking, attached
near their centres to the large hypocotyl. Seed from freshly harvested fruit is
normally viable and, when correctly treated, germination is usually above
90%. Growth inhibitors and temperature can affect the viability of stored
seed (Hasanah et al., 1984).
Table 8.1. Composition and characteristics of clove buds and bud oil at different maturity stages from
India.
Ecology
Clove grows naturally in a tropical maritime climate but, although such
conditions can be found in many countries to which clove has been in-
troduced, commercially viable plantations exist in very few; thus conditions
favouring clove production may be more critical than originally believed.
Clove is seldom grown commercially above 200-300 m although to 600 m in
the Seychelles and Madagascar, and some of the highest-yielding planta-
tions in India are in the Western Ghats above 1000 m. Although commercial
plantations are seldom established above 500-600 m, the tree will make
healthy growth at much higher elevations, as in the botanic garden at
Entebbe, Uganda, altitude 1200 m. Such trees normally make slow initial
growth, the time to first harvest is considerably lengthened, but trees remain
healthy and produce cloves for many decades. A temperature of 24-33°C is
recorded in main clove-growing regions of Indonesia and Zanzibar, similar
to that of the Moluccas. Lower mean daily temperatures occur in India, but
there is no information on the effect of varying temperature on tree growth,
flowering or fruit development.
The most productive clove areas of Zanzibar and Madagascar have an
annual rainfall of 2000-3000 mm similar to the Moluccas 2200-3600 mm,
with a dry period coinciding with flowering, which also greatly assists
harvesting and drying. Although there are no data in support, it has been
suggested a dry period is necessary to initiate the floral differentiation of
buds (Waard, 1974). If this is so then variation in the dry period, its length
or indeed absence may be a factor influencing the highly variable annual
yield. High humidity at flowering reduced yields in Singapore, Penang and
certain areas of Indonesia. A well-distributed rainfall of at least 3000 mm is
desirable to ensure continuous regrowth for leaf-oil production. A very dry
year can kill many young and also more mature trees especially where soils
are shallow or very free-draining. Such trees often exhibit similar symptoms
to die-back and death may be very fast. Research in Indonesia indicates trees
do not exhibit wilting symptoms under high water stress and leaves become
highly cutinized. Trees reportedly recovered very slowly and highly stressed
trees frequently died; a similar phenomenon also occurred in Madagascar
(Maistre, 1955).
Clove is essentially a forest tree and flourishes in a plantation environment
with partial shade from neighbouring trees. Isolated trees exposed to strong
direct sunlight as in some botanic garden specimens frequently show foliar
scorch, reduced young growth and flower number. Young trees in planta-
tions should be shaded initially by interplanting with suitable local species;
on Zanzibar Gliricidia spp. and Peltophorum spp. were used but the most
effective was bananas with the added advantage of providing food or income
during plantation establishment.
Young growth and branches are quite brittle, high winds cause consider-
able damage, and cyclones are devastating. The well-documented Zanzibar
Myrtaceae 245
1872 cyclone caused havoc and made extensive replanting essential to save
the industry. Tree damage was related to the later spread of sudden-death
disease (Nutman and Roberts, 197 1). Although extensive structural damage
can be caused by high winds, trees are seldom blown over or uprooted as the
extremely tough trunks resist breakage. Damaged trees generally recover,
although regrowth is slow, and if trees are extensively damaged replanting is
probably the quickest method of re-establishing clove production. Damaged
trees also appear to be more susceptible to infection by pathogenic fungi.
When establishing new plantations in areas where there are few other trees,
windbreaks are essential in the initial years and permanent windbreaks may
be necessary.
Cultivation
Clove trees are cultivated primarily to produce the clove used as a spice, and
clove oil production is a complementary activity. Oils can be obtained from
buds andior stems and are basically by-products of the spice trade; leaf oil
requires different cultural methods as trees are not required to flower.
Cloves may be propagated from seeds, layered, grafted, or by simply pegging
young branches to the ground and allowing them to root (Watt, 1908; Wit,
1976). Large numbers of plantlets can be produced from selected parents
using in vitro methods of propagation (Babu et al., 1993). Interspecific
grafting on rootstocks of Eugenia brasiliensis, E. unifora L., E. jambolana
Lamk. and Syzygiurn cumini L. was successful.
The usual method is to sow seeds and special trees must be left in
plantations to provide seed. High-yielding, regular-bearing trees should be
selected and a single tree yields about 2000 sound fruit in an average season.
Self-sown seedlings are also transplanted from existing plantations. Clove is
normally cross-pollinated, thus seedlings from a single tree can vary widely
in their characteristics. Variation is less where most trees are descendants of
very few imports, as in Zanzibar and Madagascar.
Fruits for seed are allowed to ripen and fall onto a cleared area around the
tree, collected daily, and kept moist under wet sacks for 2-3 days to soften
and facilitate hulling, which is usually manual to prevent seed damage. Seed
selected for planting should be sound, large, olive-green, from single-seeded
Myflaceae 247
fruit, the radicle not blackened, and the standard on Zanzibar was ‘never
more than 1100 seeds per kg’. Seed is normally sown immediately into
nursery beds, but also into pots or directly into prepared pits in the field.
Seed should be placed vertically in the soil, radicle down, and approximately
two-thirds of the seed covered; a spacing of 20 X 20 cm is common, and
seeds normally germinate in 10-15 days. Germination of selected seed is
usually above 90%, and epigeal. Where a regular supply of seedlings is
necessary, a permanent nursery can be established. The Tanzanian and
Indonesian governments have established nurseries to produce healthy,
well-grown seedlings sold cheaply to local clove growers.
About 12 months after sowing when seedlings are 25-50 cm, they are
removed from the seedbed together with sufficient soil to fully protect the
roots, and the tap-root pruned to promote lateral root growth. One or two
seedlings are planted in prepared pits to which animal manure or compost is
often added, plus a phosphatic fertilizer. The last is probably the most
important, since there was no apparent difference in growth and yield of
trees from seedlings planted in pits with or without organic manures (Zanzi-
bar Department of Agriculture, 1949-1 965). Seedlings should be shaded
until established; banana or palm leaves are quite suitable and usually easily
available. A field spacing of 6-8 m on the triangle is recommended when
grown for clove production, and seedlings which die should immediately be
replaced. When grown solely to produce leaves for oil extraction, hedge
planting is preferable to close planting of individual trees and coppicing. A
3-5 m row spacing with 0.75-1.0 m between plants is recommended in
Madagascar (Dufournet and Rodriguez, 1972). Row spacing should allow
use of cultivating and harvesting machinery.
Weeding is essential for the first 2-3 years following planting, then ring
weeding around individual trees, mowing or slashing is usually sufficient.
Hedge planting requires three to four weedings annually, or the herbicides
used generally in plantations to control local weeds are suitable, e.g. glypho-
sate. Local trials are essential to determine the effect of particular chemicals
on clove, as in pot trials 2,4-D, dalapon and napropamide caused abnormal
development in seedlings (Freire et al., 1985). Intercropping has been
suggested as a method of controlling weeds, and can be successful in newly
established plantations, but once trees have reached 3 m their roots can be
severely damaged by the necessary cultivation. Cover crops can be useful as
has been noted in the section on fertilizers, but normally suppress only less
aggressive local weeds and are of little value against persistent stoloniferous
species. Cover crops may also deplete soil moisture; on Zanzibar for exam-
ple tropical kudzu (Pueraria phaseoloides) used more soil moisture to 1 m
depth than the local weeds; however, in Indonesia, a cover crop slashed
three times annually plus clean weeding around the base of 3 year old trees
increased rate of tree growth and retained soil moisture (Daswir, 1986).
248 Chapter 8
Harvesting
A description of harvesting and preparation of cloves as a spice has been fully
described (Purseglove et al., 1981). Clove begins to bear after 4-5 years,
attains full bearing at 20 years and can remain productive for a further 100
years. Although young trees may begin to bear after 4 years, heavy picking is
not recommended, since it not only damages young trees but adversely
affects future growth. O n Zanzibar the 4-year cycle usually produces one
poor, two medium and one excellent clove crop, and a similar cycle has been
noted in Indonesia.
There are normally two harvests annually on Zanzibar, the rnwaka in July-
October and the vuli in December-January, with harvesting on Pemba about
1 month later. It was estimated in the 1950s and 1960s that some 10,000
persons travelled annually to Pemba from Zanzibar to work as pickers (Fig.
8.3). In India, there is basically one harvest, January in the plains and
March-April in the hills. Since clove is now grown in many regions of
Indonesia the harvest occurs at different times. T h e main harvest is usually
very short, 7-10 days, and large numbers of pickers move from island to
island. In Madagascar, harvesting is usually October-November, the main
harvest in November. There is frequently a shortage of pickers in specific
districts and thus annual plantation yields vary widely. Clove clusters are
picked when buds are full-sized, most have developed the characteristic pink
flush, but no petals have fallen to expose the stamens. I n Indonesia and
Madagascar, an experienced picker paid by the day harvests about 25 kg wet
cloves, but Zanzibar pickers paid by weight regularly reaped 55 kg.
Harvested clusters are taken to a central store where the cloves are
separated from the peduncles and pedicels and dried. The peduncles and
produce a greater weight of foliage but the oil yield per kilogram de-
creases.
In Indonesia, the major supply is from fallen leaves which are collected by
family labour, bundled and sold to leaf-buyers. Leaf-fall occurs throughout
the year and 1-1.5 kg of naturally dry leaves per tree can be collected every
2-3 weeks. A major factor affecting regional leaf-oil quality and type in
Indonesia is the often scattered location of clove trees, which may also be
interspersed with other tree crops; in times of scarcity much non-clove
material can be included. The larger more modern distilleries established as
joint ventures with Dutch companies employ their own buyers to reduce
adulteration to the minimum, and leaves are transported in company
vehicles to maintain an even flow of material. The companies also buy and
refine local crude oil, and produce eugenol, rectified clove leaf oil, and
related products.
Distillation
Clove bud oil was formerly produced in Europe or North America from
imported cloves, but is now regularly produced in Madagascar and In-
donesia, and on a small scale in India. First-quality bud oil requires high
quality cloves and thus producers had to ensure high standards of drying and
packing. For this reason the guaranteed grades formerly supplied by the
governments of Zanzibar and Madagascar prior to these countries becoming
independent were a valuable guarantee. Oil can be obtained by hydro- or
steam distilling whole or comminuted buds, and this affects the oil obtained.
Comminuted buds should be distilled immediately to reduce oil loss by
evaporation. Some distillers use whole cloves which extends distillation time
up to 24 hours, as the oil has to diffuse through the mass and a proportion of
the eugenol acetate hydrolyses, increasing the apparent eugenol content of
the oil. Steam distillation has the same effect and oils obtained by using dry
steam have the highest proportion of free eugenol.
The finest UK oil containing 8 5 4 9 % eugenol is from hydro-distillation,
but in the USA much clove bud oil is direct steam distilled. Stems were
previously exported from Zanzibar, but as a significant proportion was used
in the adulteration of the more expensive clove bud oil, stem exports were
prohibited for some 50 years. Commercial quantities of clove stem oil were
prepared exclusively at the government distillery under the British admin-
istration, but rapidly diminished after 1964. The plant consisted of
12 stainless steel stills each holding about 680 kg stems; steam distilling for
16 hours yielding 557% of an almost water-white oil, which darkens with age
to yellow, sometimes violet tinted. An oil yield of 1 5 1 7 % can be obtained
from selected high quality dried stems.
Leaf oil is generally obtained by hydro- or steam distillation in locally
constructed stills; many in Madagascar are made of sheet iron or iron drums,
embedded in stone or mud hearths, and heated by a wood fire. Others, of the
Myrtaceae 25 1
more advanced Deroy type, are made of copper. Still capacity is 1000-1500
litres charged with 300-400 kg of bundled branchlets and covered with
water. Yield after continuous operation for 24 hours, which includes 5-7
hours to boil the water, is about 1.5-1.8 %. The oil obtained is dark due to
the presence of iron eugenates. Private distillers now have modern equip-
ment producing a high-quality light-coloured oil.
Distilling in Indonesia is also generally a small-scale operation, although
several large modern distilleries have been built. The smaller stills are
usually vertical cylinders 3 X 1.6 m diameter charged and discharged
through the lid. A long pipe connected to the lid is submerged in a ditch or
concrete channel, where running water serves as condenser and cooler.
Distillate is separated in a cascade of cisterns, and the oil collected and sold
to dealers. The usual charge is about 1000 kg dry leaves yielding some
20-25 kg crude oil; in the rainy season from wet leaves the yield is usually
15-20 kg. Steam distilling takes 4-6 hours, direct firing 6-10 hours. The
distilling method affects both oil yield and characteristics; combined water
and steam distilling produces more oil than steam alone, as does an increase
in distilling time. Using water and steam for 4 hours gave the highest oil yield
of 3.0%; steam alone produced oils lower in eugenol, with a lower specific
gravity and refractive index (Nyrdjanah et al., 199 1). Oil characteristics may
change when stored leaves are distilled; in Indonesia, characteristics of oil
obtained from leaves stored for 1 or 8 months were (8-month in brackets):
specific gravity at 25OC, 1.0425 (1.0486); refractive index at 25'C, 1.5282
(1.5304), optical rotation - 1'32'(- 1'39'); solubility in 70% alcohol 1:2-3
(1:2), eugenol content 83% (79%) (Muchlis and Rusli, 1978).
Unidentified scales have been reported as the most serious pest of clove
nurseries in India. T h e red tree-ant, Oecophylla longinoda, known in Zanzi-
bar as maji-ya-moto (‘firewater’), is a serious nuisance rather than a pest. It
makes large nests about 15 cm in diameter, and the very active, pugi-~acious
workers swarm on pickers during harvesting, inflicting extremely painful
bites.
Nematodes have been reported on clove roots in several countries, but
with no assessment of damage, including Dolichodorus spp., Helicotylenchus
dihystera, Rotylenchulus renijormis and Trichodorus spp. Some, however, are
serious pests of other tropical crops, and where these are grown in prisximity
to cloves there could be greater danger of damage. In Brazil for example,
Meloidogyne incognita, which is a serious pest of pepper, could require
control.
The most economically important disease of cloves in terms of its effect on
world production is sudden-death disease, which may have different causes
but basically the same symptoms. On Zanzibar and Pemba, over half the
mature trees have succumbed since 1930. So many trees died that clove
timber became a major source of fuel on Zanzibar, and the disease continues
to be the single most important factor affecting local clove production. T h e
characteristics of this disease and remedial measures are well documented
(Dabek and Martin, 1987). There are several causal organisms, one of the
most serious being Cyptosporella eugeniae, always associated with injury,
particularly when branches are broken during harvesting. T h e branch slowly
dies back, leaves turn brown and the fungus proceeds downwards. When a
fork is reached, the unaffected branch above it dies suddenly. A serious
cause of tree death first recognized in western Sumatra in 196 1 and named
Sumatra disease has become widespread and is caused by Pseudomonas
syzygii, whose major vector is the boring beetle Hindola striata. An attempt to
introduce a degree of immunity into clove by grafting clove buds onto 27
immune or resistant species of Myrtaceae was unsuccessful (Jarvit: et al.,
1986).
T h e die-back known as Matibudjang in Indonesia begins with a :general
decay of fine feeder roots, which soon produces secondary leaf shedding and
dieback, initially at top of the crown and subsequently extending throughout
the canopy. Trees slowly decline and usually die 2-3 years from onset of the
symptoms. T h e primary cause is considered to be unfavourable soil condi-
tions rather than the associated fungal infection. A similar disease in Mada-
gascar, apoplexia, is also restricted to areas of poor soil.
Leaf spots recorded from several countries include Cylindrocladiuin guin-
queseptatum and Gloeosporiumpiperatum which attack seedlings, Mycosphaer-
ella cayophyllata associated with Altemaria spp., and various sooty moulds
such as Capnodium brasiliense and Aschersonia spp. A thread blight caused by
Corticium spp. is also of minor importance. Wilts may cause death of
Myrtaceae 253
seedlings in the nursery and after transplanting; causal organisms are Cylin-
drocladium spp., Fusarium spp., Rhizoctonia bataticola, Phytophthora spp. and
Colletotrichum spp. The parasitic alga, Cephaleuros mycoidea, is most damag-
ing to leaves in Malaysia but is seldom serious on healthy trees; it has also
been reported from India.
A B C D
Beta-caryophyllene 4.35 5.13 9.86 12.50
Alpha-humulene 0.54 0.60 1.10 1.36
EugenoI 88.95 92.35 86.89 84.77
Eugenyl acetate 5.54 1.25 1.59 ng
A, Madagascan clove bud oil; B, Zanzibar clove stem oil; C, Madagascan clove leaf oil;
D, Indonesian clove leaf oil; ng, not given. Source: Adapted from Srinivas (1986).
254 Chapter 8
between amount and intensity of heat applied during distilling leaves and oil
yield, eugenol content and oil characteristics. The amount of material in the
still had a similar effect at any given heat level (Rusli et al., 1979).
Clove bud oil is a clear, colourless to yellow mobile liquid, becoming more
brown with age or contamination with iron or copper, with a strong charac-
teristic sweet and spicy clove odour, and a warm, almost burning and spicy
flavour. The composition and organoleptic properties of distilled bud oil
differ somewhat from the volatile oil present in spice; Hydro-distillation
reportedly provides the finest oils for perfumery and flavour use with a
eugenol content of 85589%. The more common steam distillation produces
oils with a eugenol content of 91-95% resulting from hydrolysis of eugenyl
acetate. Distilling whole spice provides oil with a high eugenol content and
a specific gravity above 1.06; comminuted spice has a lower eugenol content
and a specific gravity below 1.06. Bud oil stores well in sealed light-proof
containers, and although colour may darken, characteristics and quality are
little affected.
Bud oil consists basically of three components: eugenol70-90%, eugenyl
acetate to 17%, and caryophyllene sesquiterpenes (mainly beta-
caryophyllene) 5 1 2 % . The main constituents of bud and stem oils are
shown in Table 8.4., and the main characteristics in Table 8.5. There is also
regional variation in eugenol content, and oils from Comoro may have a very
high total phenol content. The distilling technique also affects eugenol
content, especially the time material is retained in the still. In general, a
longer period increases oil yield but lowers the eugenol content. When bud
oil is required for eugenol, fast distillation is preferable but requires a higher
steam pressure and is more expensive. Although bud oil composition has
Myrlaceae 255
Table 8.4. Main constituents of clove bud and stem oils from
Madagascar.
Table 8.5.Standard specifications for clove bud, stem and leaf oils.
70-80% volatile oil. Oleoresins extracted with alcohol are much darker than
those from hydrocarbon solvents and have a more powerful perfume and
flavour, but are considered less refined or delicate in odour. Extraction on an
industrial scale is by specialist firms in major consuming countries, which are
now using supercritical fluid techniques. Oleoresin is normally dispersed on
salt, dextrose, flour or rusk, and sold weight for weight equivalent in strength
to the dry spice. The main advantages over dry spice are little risk of bacterial
contamination, standard strength and quality. The oleoresin has gained in
market share from both dry spice and bud oil especially in the food industry,
mainly in meat and soup products, and is also used directly in perfumery.
Clove bud concrete is obtained by solvent extracting dried and commi-
nuted buds; benzene and petroleum ether are commonly used. Commercial
concrete is a very viscous liquid or semi-solid mass, pale olive-green through
yellow-brown to dark brown. Highest quality is usually a clear viscous liquid,
olive to pale brown, with a refreshing, oily-sweet, intensively rich and spicy
odour closely resembling that of dry buds (Arctander, 1960).
Bud absolute is prepared by extracting concrete using similar solvents;
petroleum ether produces the finest, benzene a richer odour. Absolute is a
viscous or oily liquid, semi-solid at low temperatures; olive-green, greenish
brown to orange-brown and soluble in alcohol in all proportions. Absolute
differs from bud oil in containing only those constituents present in the
unprocessed bud; there is, for example, no trace of caryophyllenes. The
odour is of clove flowers in full bloom; floral and refined, balsamic, sour-
sweet and immensely rich with none of the unwanted off-notes and flavour
found in bud oil (Arctander, 1960). The main use is in perfumery, to a lesser
Myrtaceae 257
Table 8.6. Main characteristics of clove leaf oil from seiected countries.
Eugenol can be isolated from a variety of starting materials, and can also be
manufactured synthetically. Eugenol is smoother in odour than clove oils if
less strong and used mainly in inexpensive perfumes, soaps and similar
toiletries, or where too definite a clove odour is undesirable.
Eucalyptus
Almost all eucalypts are indigenous to Australia where they constitute about
75% of tree flora. Outside Australia, three species occur in eastern In-
donesia, six in Papua New Guinea and one in the Philippines, but inter-
estingly none are native to New Zealand or New Caledonia. The imaginary
floral boundary north of Australia, Wallace's Line, originally considered the
limit of naturally occurring eucalypts, was later modified by Huxley (Fig.
8.4). The large Eucalyptus plantations in other countries have been estab-
lished mainly from Australian seed, and are now so extensive that eucalypts
can form a significant proportion of the forested area, as in Ethiopia and
145"
15"
Fig. 8.4. Eucalyptus: Wallace's line with later modifications. (Australian National University Library,
Canberra.)
260 Chapter 8
preferably both. Perhaps the greatest single factor must be the selection or
breeding of superior trees to produce uniform, high-yielding progeny.
Botany
The genus Eucalyptus was so named by the French botanist C.L. de Brutelle
L‘Heritier from the Greek eu ( well) and kalypto (covered), in allusion to the
operculum or lid which covers the stamens until they are fully developed
(Fig. 8.5). The first species to be identified and named was E. obliqua
(L’Heritier, 1788). Probably no genus has been more thoroughly investi-
gated since Candolle’s attempt in 1828, and there are at least 700 known
species (Brooker and Kleinig, 1983; Boland et al., 1984; Hall, 1987). Recent
research into the DNA sequences of Eucalyptus spp. however, may require
further revision and provide vital information to plant breeders. Hybrid-
ization occurs naturally between many Eucalyptus species, and an interesting
discussion on its occurrence and significance is contained in Pryor (1 976).
Also notable are chemovars (chemical varieties) within a species which
produce oils with distinct characteristics.
Most eucalypts are trees and many are very large (E. regnans can reach 100
m) but the majority are 10-25 or 30-35 m . Several, including E. polybractea,
have a distinctive form known as ‘mallee’ with multiple stems originating
from a common stock. T h e extremely penetrating and often extensive root
system of mature eucalypts and the resulting massive extraction of soil
moisture by transpiration was invaluable in preserving Australian soils from
excessive salinity. This high water-extractive ability of eucalypts has been
exploited in other countries to reduce local waterlogging or swamps, but
planting eucalypts in dry areas often exacerbates a soil-moisture shortage.
An unusual feature of many eucalypts is the presence of lignotubers, which
consist of a mass of vegetative buds, associated vascular tissue and sub-
stantial food reserves. Eucalypts are evergreens except for a very few tropical
species which defoliate in the dry season, and individual leaves are normally
shed between 12 and 18 months, but up to 4 years on some species. Bud, leaf
and shoot development of eucalypts is considered indeterminate, tends to be
continuous and to occur whenever the environment is favourable, and there
is no requirement for rest or overwintering.
Mature eucalypt leaves differ from juvenile leaves, and new leaves from
those more mature on the same tree. Adult leaves are usually lanceolate,
petiolate, apparently alternate, basically green, usually glaucous, with a
bluish or whitish appearance due to surface wax. Juvenile leaves may be
broad, sessile, remain opposite and glaucous, or narrow, petiolate, becom-
ing alternate and basically green. T o avoid unnecessary wordage, only the
mature leaf will be described, although regrowth from stumps, common in
essential oil production, is usually juvenile or intermediate.
Myiiaceae 263
Stamens
Receptacle
,Stigma \J
Anthers*l g
/+ Receptacle
Fig. 8.5. Sketch of eucalyptus flower and fruit. (CSIRO Library Melbourne.)
264 Chapter 8
Ecology
Eucalypts are very adaptable and can be grown and will flourish in a wide
range of environments quite unlike their native Australia. The generally
accepted range limit is between 45"N and 45"S, and outside these latitudes
there is usually some local climatic explanation for vigorous growth. Prob-
ably the main factor limiting growth is low temperature, since in the limited
areas of Australia where snow falls regularly it acts an insulator, and soil
temperature remains above 0°C. In countries where the ground freezes this
Myrtaceae 265
is invariably fatal to eucalypts, and a feature of Russian research has been the
production of frost-resistant hybrids. Seedling growth of numerous eucalypt
species in glasshouses is minimal below 10°C, increases rapidly at 20-25"C,
peaks between 20 and 30°C and is adversely affected above 30°C. T h e
difference between night and day temperature is apparently of little sig-
nificance.
Eucalypts are particularly well adapted to monsoon Asia and seasonal
rainfall areas of Africa and south America. Species native to winter rainfall
zones of Australia (June-August), particularly those species native to south-
ern Australia, grow poorly if planted where there is predominantly summer
rainfall. The reverse is not so; species which grow naturally in areas of
summer rainfall in Australia (December-February) will often grow well in
areas of winter rainfall; E. citriodoru is a successful ornamental tree in Perth
(Western Australia) and Adelaide (South Australia) although it is confined
to Queensland and does not occur naturally below latitude 26"s.
Although a particular Eucalyptus species in Australia may be widespread
and appear to occur in almost pure stands, this may not be so; in any specific
area one species can replace another across quite sharply defined boundaries
with soil moisture considered a most important factor. This could explain
the rather uneven growth of large single-species plantations the author has
seen in Central Africa and South America, where there are often abrupt soil
changes. T h e most suitable environment in Australia is included in species
266 Chapter 8
description as a guide, but individual factors affect not only general tree
growth, but may also alter leaf oil content and oil characteristics. Oil content
is frequently lowest in very dry or very cold periods, and highest in wet or
warm periods; however, oil content of mature leaves on uncoppiced trees
may show little seasonal variation.
Cultivation
Establishment of a commercial Eucalyptus plantation is still bedevilled by the
necessity to obtain pure seed where there is no economic method of large-
scale vegetative propagation. Much seed offered for commercial sale is
misnamed, and unless obtained from a reputable source (generally Aus-
tralia) will usually be a mixture of several species, or contain a substantial
amount of hybrid seed; some species also have a specific germination
requirement. In general, large seeds germinate faster and have a higher
germination percentage; there is no evidence, however, that seed size is
related to later growth rate. A vegetative method employed in South Africa
Myrtaceae 267
on E. radiata subsp. australiana and E. dives var.‘C’ was successful and could
be more widely applied. Shoots were taken from trees selected for high
foliage yield and high oil content of leaves when harvested 18 months after
coppicing at 1 year; criteria for selection were 4.5 kg air-dried leaf and 10%
oil content per tree, Cuttings 150 m m long with one pair of leaves, dipped in
1% IAA in talc, and planted in the autumn gave an 85% take, roots
appearing in 3-4 weeks, Cuttings from specific trees rooted poorly, which
may be a genetic factor since a root inhibitor exists in adult tissue of a
number of Eucalyptus spp. Since most commercial oil is obtained from E.
globulus or E. polybractea, methods of establishment and cultivation suitable
for these species have been described in detail. Only special requirements or
techniques necessary for other species are mentioned.
Mature eucalypts of most species are generally fire resistant but planta-
tions should be protected from fire especially after coppicing, since foliage is
highly flammable; stumps will be little affected and regrow. Mallee types are
not generally fire sensitive, but a hot burn through coppiced natural stands
may kill trees.
Manual or mechanical methods of weed control remain most common in
eucalypt plantations, but generally rising labour cost has increased interest
in herbicides. A number have been tested and the following proved useful:
ametryne, atrazine, glyphosate, oxadiazon, prometryne, profluralin, propa-
zine, simanzine, terbacil and terbutryne. Species vary in their tolerance to a
specific chemical and trials are necessary before large-scale use. For in-
stance, E. globulus can be susceptible to triazine herbicides, and application
rate of terbacil can be critical since there is a relatively small difference
between levels which give effective weed control and those causing damage
to seedlings or young trees.
In Australia, there was invariably severe sheet erosion in natural stands of
E. polybractea and E. dives on undulating land, which were coppiced and
mechanically harvested, and virtually weed-free. Intercropping commercial
eucalypt plantations is not normally viable, and only successful where
eucalypts form a small proportion of the cropped area. Smallholders use the
technique most successfully, especially in India, using Cymbopogon martinii
var. motia (Roxb) Watts, Mentha arvensis L., and a number of pulses and
spices. Intercropping with other essential oil crops can also be of value to
smallholders if it enables greater utilization of a local still.
Harvesting
Harvesting wild trees was initially the usual method wherever there were
extensive stands of the required Eucalyptus spp. In Australia, operations
were almost wholly manual and, according to a contemporary author, ‘The
work was hard, the hours long, the life lonely, the rewards meagre.’ Most
Eucalyptus oil is now produced from plantations in countries other than
Australia. Techniques used for harvesting different species of eucalypts for
268 Chapter 8
oil are basically similar and generally applicable. They are discussed in detail
in the sections on E. globulus and E. polybractea. Where there is a particular
requirement or technique this is mentioned. O n an international basis,
harvesting is currently influenced by traditional rather than technical rea-
sons, and will remain so until more information is available on which to base
systematic management.
Distillation
Oil is obtained by distilling foliage from trees felled for other purposes,
harvesting natural stands, or from specially established plantations. The first
involves no agricultural operations and leaves are distilled by any suitable
method a t a convenient site, but only on large plantations is this profitable.
The current method of distilling E. polybractea is described in the section on
that species, although a few portable stills remain in use by operators
working in rough or mountainous country (Fig. 8.6). Such a distillery
consists of water tanks usually 1.2 X 1.2 X 1.2 m or 2 X 1.5 X 5 m coupled
together and operated in pairs, holding 360-460 kg foliage. Between 3500
and 4500 litres of water are placed in each tank beneath a grid which
supports the charge. Prior to replacing the lid, a thick layer of clay is laid
along the flange, on this is placed a strip of sacking and a second layer of clay
applied. The lid is then fastened in position by cross-timbers, wooden
wedges and iron stirrups. In Brazil generally and also in Spain, foliage is
transported to a permanent distillery often owned by a company which pays
for delivered leaf. Some large Spanish plantation companies also own
distilleries and produce rectified oil to international standards. In Brazil,
where there are very large plantations of E. citm'odora to produce essential oil,
these are fully integrated and well-managed operations.
Modern distilling plants in all countries are very similar and discussed in
the chapter on distilling. Solvent extraction usually recovers more oil than
distillation, but requires more sophisticated equipment, and may also alter
oil composition (Ammon et al., 1985). Disposal of still residue can be a
major problem, but is frequently dried and used a fuel. Transforming this
residue into a saleable product can substantially increase overall profitabil-
ity. A Victorian company has produced 'Eucy-mulch' by converting this
residue into a sterilized mulch in 100 kg bales for sale to nurseries and home
gardeners (P. Abbott, Melbourne, personal communication).
Geographic Sub-genera
Parts affected Pathogen location involved Remarks
Foliage Virus (including Argentina, India, Corymbia Potentially serious effects.
tobacco mosaic) S. Europe Symphyomyrtus
Puccinia Brazil Corymbia (E. Likely to show some host
citriodora) specificity.
Oidium spp. Australia, South Symphyomyrtus Little potential importance
including 0. America, Europe under field conditions.
eucalypti
Fungi imperfecti, Widespread Coiymbia and Of little significance under
Ascomycetes including Symphyomyrtus field conditions.
Australia
Fungi imperfecti, Australia Monocalyptus Evidence of pronounced
Ascomycetes Symphyomyrtus host specificity
Stem and branch Cytospora spp. Widespread Corymbia and Cankers caused by
Endothia including Symphyomyrtus Endothia spp. and
havanensis Australia Diaporthe spp. potentially
important, particularly in
tropics and subtropics.
Diaporthe Brazil Corymbia and
cubense Symphomyrtus
Roots Pythium spp. Widespread All subgenera Of local significance only.
Phytophthora spp.
Verticillium spp. Australia Evidence of resistance
Fusarium spp. among seedlings.
Cylindrocladium
SPP,
P. cinnamomi, Widespread All subgenera This is a potentially
Armillaria spp. including important group of
Armillariella spp. Australia diseases in monospecific
Ganoderma spp. plantations.
Polyporus spp.
Butt and stem Numerous higher Widespread All four Potentially significant
Basidiomycetes including subgenera cause of disease in
Australia monospecific plantations.
2 72 Chapter 8
Eucalyptus citriodora
Botany
Eucalyptus citriodora is a medium to tall tree to 50 m normally with a single
stem to 1-1.2 m DBH. The bark is smooth throughout, deciduous, bluish
grey to apple-pink, sometimes red, and often blotched with all three colours.
The wood is pale, light to grey-brown, hard, strong, very tough, moderately
durable and easily worked; the basic density is 1000 kgim’. The major use is
in general and heavy construction, and it is not suitable for pulping. Kino
contains the antibiotic citriodoral. Growth of seedlings and young trees is
rapid; in India, stem girth increased from 3 cm at 15 weeks to 10 cm at 60
weeks and to 15 cm at 120 weeks; in Zaire, trees reached 13 m in 4 years; in
Brazil, cloned trees were 15.5 m at 4 years.
The foliage is generally sparse; mature leaves measure 10-15 X 1-2 cm,
are narrow to broad-lanceolate and are strongly lemon scented when
crushed. The oil yield from steam-distilled fresh leaves is 1.5% but up to
2.5%, and up to 5% from dry leaves. The oil is almost entirely mono-
terpenoid with citronellal usually at 8 5 9 0 % . Other significant constituents
are citronellol, cineole and pinenes. Young leaves frequently have a higher
oil content than mature, but in Kenya the author often found greater
Myflaceae 2 73
variation in oil content between leaves of similar age from different trees
than between leaves of different ages from the same tree; oil content may also
be independent of leaf maturity (Mwangi et al., 1981). Interestingly in
Australia, leaves and terminal branchlets from forest trees yielded
0.5-0.75% oil, from isolated trees 1.0-1.3% and cultivated trees to 2.0%,
probably due to a general reduction in inter-tree competition. Oil character-
istics may also change with age of seedlings or leaves (Gulati et al., 1968).
Two types of E. citriodora occur in India with pubescent and non-
pubescent leaves, and their oils have different characteristics and composi-
tion irrespective of location (Gulati et al., 1968). T h e first had a fresh leaf oil
yield of 1.9-2.9% with a citronellal content of 75-85%; the second
1.75-4.0% with a citronellal content of 50-73%. In Australia, at least four
chemotypes of E. citriodora occur with an aldehyde content of 10-85%. Leaf
oil from different countries varies in composition, there are substantial
differences in Indian regional oils, while in Zaire, oil from the Ituri region
had a higher aldehyde content than oil from the Kivu area. In Brazil, there is
considerable regional variation in citronellal content partly due to the
distilling method. Oil characteristics from selected origins are shown in
Table 8.9.
T h e inflorescence is terminal, corymbose, the umbels three to five flow-
ered. The flowers are cream coloured, with the peduncles terete, 5-7 cm; the
buds are ovoid, pedicillate, 10-12 X 7-8 mm. T h e operculum is hemi-
spherical, shorter than the calyx tube; the anthers are obovate. Main flower-
ing is June-July in Australia. Steam distilled fresh buds yield 0.5-0.85% oil
with the following main characteristics: apparent density at 30°C 0.9786,
refractive index at 25°C 1.4736, optical rotation -0.28", acid value 27.7,
ester value 86.12, citronellal 28.8%.
The fruit is ovoid or urceolate, pedicellate, 10 X 3 mm; the disc broad,
descending, with 3-4 valves deeply enclosed; the pedicels are slender, to
Table 8.9. Main characteristics of Eucalyptus citriodora oil from selected originsa,
10 mm. The seed is very small, usually 200,000 but up to 400,000 per kg;
the mean viability is about 30%.
Ecology
Eucalyptus citriodorus is naturally confined to subtropical and tropical zones
in Australia, between latitudes 17' and 26"S, at altitudes of 450-1000 m in
the hotter north, 70-400 m in the cooler south; the species has not been
successful at low altitudes in the wetter tropics. The rainfall is 650-1600
mm, seasonal, with a 4-5 month dry period but up to 7 months. Elsewhere,
reports indicate a minimum of 600 mm, but 900 mm is desirable for rapid
growth. Although relatively drought tolerant, hot arid areas are unsuitable
for oil production. The species is frost susceptible especially when im-
mature. In Zaire, commercial plantings in the Ituri district were around
1500 m with average annual rainfall of 1200 mm. It was planted in the
Kenya highlands and on a small scale in the Njombe area of Tanzania with
a similar climate; in Sri Lanka it was planted below 120 m including the drier
zones.
The substantial plantings of E. citriodora in Brazil, approximately 6 million
trees, are concentrated in Sao Paulo State between 22 and 23'S, at approx-
imately 100 myrainfall 800-1200 mm. The author visited large plantations
near Guatemala City around 1500 m with an annual rainfall above 1000
mm. In China, E. citriodora is extensively planted in Guizhou and Guangxi
Provinces, 22-28"N, but far exceeded in numbers by E. robusta; in Russia,
E. citriodora is planted specifically for essential oil production mainly around
the Black Sea and Caspian Sea between 40 and 45"N. The species is widely
planted in India at varying altitudes to ensure sufficient rainfall, but its range
is somewhat restricted as seedlings are frost sensitive. Main plantings are in
the Punjab and Uttar Pradesh in the north and Andhra Pradesh in central
India; it is also grown successfully in Assam with an average annual rainfall
of 2400 mm.
Environment influences oil production, and rainfall in particular directly
affects yield per hectare, leaf oil content and oil characteristics. The follow-
ing remarks are generally applicable. Leaf oil content increases in the wet
period following a dry season, then falls as the next dry season approaches,
being lowest during that period. Where there are pronounced seasonal
differences in mean maximum and minimum temperatures, leaf oil content
is normally highest in the warm period and lowest in the cool; citronellal
content of oil also follows a similar pattern, and in India is lowest in May-
June (the hot dry season). Climate also influences the strain of E. citriodora
selected for regional planting in India, where the pubescent-leaf type is less
drought resistant than the less valuable non-pubescent type. The influence
of climate was demonstrated on E. citriodora planted at two very different
sites: Kodaikanal, at an altitude of 2200 m with an annual rainfall of 1600
mm; and Bangalore at 920 m with 820 mm annual rainfall. Oil yield and
Myrtaceae 2 75
citronellal content from fresh leaves were (per cent): 1.23-1.7 1 and 7 1-93 at
Kodaikanal; 2.2-2.5 and 73-83 at Bangalore (Rao et al., 1984). While
climate may affect oil constituents and characteristics, these differences can
be masked or modified by other factors such as chemovars, and ignorance of
such chemical varieties could have incorrectly attributed oil differences to
climatic or environmental variation in early references.
Cultivation
Those operations necessary for E. globulus are also generally suitable for E.
citriodora but seedling roots are more sensitive to damage than other euca-
lypts and should be handled carefully when transplanting. Seedbeds should
be watered prior to removal of seedlings and as much soil as possible should
be retained on the roots, which should not be allowed to dry out. Seedlings
should be transported to the planting site with the minimum delay and kept
shaded during transport. Two seedlings should be placed at each site since
mortality is frequently high and it is cheaper to thin than replant. Direct
sowing of seed into prepared lands was successful in Africa, but the favour-
able conditions are infrequent and it is not generally recommended. Viabil-
ity of commercially available E. citriodora seed is extremely variable, and
although germination percentage of seed from reliable sources in Australia is
usually around 30-50%, it may be only 5 1 0 % from a local source. Seed-
lings should be transplanted on cloudy or rainy days, and in southern India
this operation is delayed until the soil profile is moist to at least 75 cm. When
grown solely to produce oil, the trees are regularly topped to maintain good
coppice growth, with a between-tree spacing of 1.5-3.0 m in rows to suit the
equipment employed.
276 Chapter 8
Harvesting
When E. citriodora is grown for oil production young trees are normally
coppiced and the optimum local height maintained by regular cutting. T h e
first harvest is 6-12 months after planting out, and often side branches only
are removed, the main stem remaining to promote regrowth. In the 2nd to
4th year the whole tree may be felled, the timber being used as fuel. T h e
stump remaining produces numerous shoots which are then harvested once
or twice annually. Management should thus aim to increase labour effi-
ciency and yield of leaves or oil per hectare. The number of harvests per year
in relation to optimum still usage and cost per kg of oil should also be
determined. T h e number of cuts varies but is usually once or twice annually.
Once a year or three cuts in 2 years was most profitable in India, and in
Kerala State little was gained by cutting at 8 months compared to 12
months, since the small decrease in total oil at the longer interval was more
than offset by the saving in overall cost. When cut at 8 monthly intervals over
2 years, yield per tree was 600 kg foliage and 7.40 kg oil; at 12-monthly
intervals 539 kg and 7.1 kg respectively (Kannan & Nair, 1965). When
grown by smallholders whose labour costs are virtually nil in cash terms, two
or three harvests per year may be profitable. In north-west India, a major
harvest is taken in April and November, a supplementary cut in July if
regrowth is sufficient.
Other factors may influence the number of cuts; stems of 2 years growth in
treeless regions are valuable for building or fencing, and cutting is postponed
to obtain a supply. Growing E. citriodora in large plantations in India is
uncommon and most oil is produced by smallholders harvesting trees
growing on their own lands; thus yield per tree is important but has received
little attention. A majority of smallholders cut their trees only when they
have time to spare, or a rise in the local oil price makes it attractive. In Africa
and some South American countries, trees are allowed to make normal
growth and then topped at 2.5 my branches also being lopped to provide
foliage for distillation. Trees are topped again at around 5.0 m ythe stem and
woody branches being used as poles or fuel. One or two harvests annually is
normal; February and November give the highest oil yield in Cuba (Miranda
and Zayas, 1985).
Foliage yield from annual cutting of coppiced trees gives 2-5 kg leaves
(exceptionally 8 kg), with selected high-yielding strains 10 kg; mature trees
yielded 300-500 kg of foliage when felled for timber. There may also be a
difference in oil yield and citronellal content between coppiced and pol-
larded trees (Shiva et al. , 1988). Leaf oil content can vary significantly and,
in commercial plantations in particular, it is important to determine the local
reasons for, and the occurrence of, this variation since such data will allow
more profitable plantation and distillery management. Oil content and
citronellal content of oil are highest in young leaves and fall as leaves mature;
however, the reduction in oil content is slow over the first 4-6 months of leaf
Myttaceae 277
Distillation
Distillation can be hydro or steam, and the methods used for E. globulus are
suitable for E. citriodora. In general, leaves and twiglets should be as fresh as
possible, since oil yield may be reduced and characteristics change after 24
hours, especially loss of the important aldehydes.
IS0 EOA
Relative density at 20/20"C 0.858-0.877 -
Specific gravity at 25125°C - 0.860-0.875
Refractive index at 20°C 1.4500-1.4590 1.4510-1.4640
Optical rotation at 20°C -2" to t4" -0.5"o t 2 '
Solubility in ethanol at 20°C 1:2(v/v) 80% 1 :3(viv) 70%
Total aldehvdes as citronellal min. 70% (mim) 65-85%
EOA, Essential Oil Association, USA; ISO, International Standards
Organisation; the Australian Standard is similar.
2 78 Chapter 8
Eucalyptus dives
Ecology
Eucalyptus dives commonly occurs in South East Australia between 3 1 and
38OS, at altitudes of 150-1400 my with a climate warm humid in summer,
cool in winter, with frost and snow at higher elevations. The annual rainfall
is 600-1 100 mm with few long dry periods. The species is not cultivated in
Australia, and has been planted only on an experimental or pilot scale in
other countries including Zaire and South Africa.
Type A B C
Yield 3.0 1.5 2.9 3.0
(fresh material) (%) 4.0 5.1 3.9 4.0
Specific 0.8892 0.8630 0.9062 0.917
gravity 15°C 0.9063 0.8870 0.9148 0.923
Optical -44O -44" -13.4" to +q
rotation - 78O - 66" -28.3"
Refractive 1.4784 1.4750 1.4669 1.4609
index 20°C 1.4810 1.4792 1.4711 1.4640
Cineole ("h) ng ng 25-45 65-75
Piperitone (%) 40-53 2-8 12-1 8 <6
Alpha-phellandrene (%) 20 60 ng ng
30 80 ng ng
Solubility (viv 70% 1:IO 1:IO 1 :3 1:1-5
alcohol)
~~~~~ ~
Cultivation
Foliage is obtained wholly from wild trees in Australia, which are felled,
allowed to coppice, then harvested annually, and the leaves and small twigs
transported to a still. The generally hilly areas in which the highest popula-
tions and therefore most exploitable stands occur, confines mechanization
to cutting regrowth and transport. Pilot plantings of E. dives in South Africa
were to determine the feasibility of cineole production. Seedlings grew well
and many reached 4 m one year after planting out, but E. dives was
outproduced in foliage yield per hectare and total cineole by E. radiata
(Donald, 1980).
Botany
Eucalyptus globulus is a medium-sized woodland tree to 20 my or tall forest
tree to 70 m, and 2 m DBH. Mature woodland trees usually have extensive
Myrtaceae 281
roots which are frequently deeply penetrating, but in plantations the roots
are often more shallow. There is usually a single trunk, much-branched, the
first often fairly high-up and on forest trees branches are usually retained for
less than half the total height. The lower bark is rough, greyish or brownish;
the upper bark smooth, pale, often with a bluish tinge, decorticating in long
strips. Seedling growth is rapid, and young trees may reach 10-15 m in 3-4
years but 8-12 m is more usual; clones from selected trees in Brazil had an
average daily growth of 1 cm (C. Aguair, Espirito Santo, Brazil, personal
communication). Too rapid growth can be detrimental, as above-ground
development can outpace root growth and young trees can be easily up-
rooted by high winds. Trees coppice readily, but the height of coppicing is
important, as noted later; lignotuber development is weak. The wood is light
yellow-brown, open textured, commonly with interlocking grain, strong and
moderately durable; its basic density is 900 kg/m3. The timber is used in
general construction and as poles, piles and railway sleepers. Mature timber
produces poor pulp, but young plantation-grown wood is suitable for paper
and rayon pulps.
The foliage is moderate, the mature leaves being dark glossy green,
alternate, falcate, lanceolate or narrow-lanceolate, 250 X 24 mm, moder-
ately reticulate and firm. Crushed leaves are strongly aromatic, with an
unpleasant odour due to the isovaleraldehyde content. The oil yield from
steam-distilled fresh foliage and small twigs is 1.O-1.5%; young foliage from
coppiced trees yields 3%. Average oil yield from commercial operations is
reported as (per cent): Spain and Portugal 0.7-0.9; Australia 1.2-2.5; India
0.8-1.2; Brazil 0.9-1.5; Colombia 0.7-0.9; California 1.O. A major factor
affecting average yield is age of harvested foliage. Thus trees felled for timber
whose foliage is distilled primarily as a by-product give much lower average
yields per tonne of foliage than coppiced plantations where the foliage is
generally immature with a higher oil content. The oil content of well-dried
leaves is approximately twice that of fresh foliage, and oil content of dried or
partially dried leaves does not fall significantly if leaves are temporarily
stored in cool dry conditions; there is only minor loss if stored up to 1 year.
Eucalyptusglobulus oil, as noted, is the most studied of all eucalyptus oils and
many analyses have been published (Lawrence, 1994). Some 100 con-
stituents have been identified with 1,s-cineole to 60-70%, and about 40
account for less than 1% of total oil volume. The approximate percentage by
volume is shown in brackets: 20 aliphatic compounds (I); 15 monoterpenes
(20-25); 15 monoterpenoid alcohols (3-5); 15 monoterpenoid carbonyls
(1-2); 5 monoterpenoid esters (1-2); 10 sesquiterpenes (3-5); 10 sesqui-
terpenoid alcohols ( 1-2); and 5 benzenoid compounds (0.5) (Boland et al. ,
199 1). Proportions are not constant, and there can be substantial regional
and seasonal variation. The main characteristics of selected oils are shown in
Table 8.12.
282 Chapter 8
Table 8.12. Main characteristics of Eucalyptus globulus oil from selected countries.
Commercial Spain-Portugal
sample standard Brazil Australia
Specific 0.910 0.906 0.9070 0.910
gravity (20°C) 0.926 0.925 0.9105 0.930
Refractive 1.4593 1.459 1.4612 1.460
index (20°C) 1.4710 1.467 1.4653 1.470
Optical O0 0" t0.46" t0'18'
rotation t9.20° t10.0" t0.68O t15O
Solubility (viv 70% 1:1.5-11 1:5 1 :I 1 :1.5-3
alcohol)
1,8-cineole (%) 55-85 min 70 65-70 70-75
Figures in columns are range. There can be substantial variation in oil characteristics
The inflorescence is axillary, simple, with one to three white flowers; the
peduncles are sometimes absent or very short and stout to 40 mm; the
pedicels are usually absent, or stout-angular to 8 mm. The buds are axillary,
obconical, usually solitary, extremely glaucous, and sessile on a very short
peduncle. The operculum is flattened with a very distinct central knob. The
anthers are versatile; the stamens mostly inflexed and all fertile. An essential
oil can be obtained from the flowers. The flowers are attractive to bees, but
of little interest to Australian commercial apiarists.
The fruit is sessile to shortly pedicellate, obconical 1.5-1.5 X 1.5-2.5 cm,
often waxy, the rim thick, disc broad and level to slightly convex, and
smooth. There are four or five valves slightly to distinctly exerted, with a very
distinct concave calycine ring. The fruit contains 1-2% essential oil with a
cineole content of 20-70%; some 30 other compounds have been identified
with sesquiterpenes more significant than monoterpenes (Baslas and Sax-
ena, 1984). A blackcurrant scent can often be clearly recognized from
immature fruit (Boelens, 1985). Seed averages 200,000 per kg with viability
around 33%. Fertile seeds are more plump and healthier looking and, with
experience, are easily selected. Plantation trees normally produce seed in
4-5 years.
Ecology
Eucalyptus globulus occurs naturally in Tasmania and southern Victoria, but
now grows in coastal areas between 31 and 43"S, from sea-level to 1100 m
with a rainfall of 600-1 100 mm annually, and does not occur where there is
less than 500 mm. Mature trees are generally frost resistant, but tem-
peratures below -5°C often kill seedlings, although seed from trees growing
at higher, colder elevations reportedly produced frost resistant seedlings.
Mature trees are very firm in the wind, but young fast-growing trees and
coppiced regrowth can be badly damaged. Mature trees are fire resistant, but
Myiiaceae 283
Cultivation
The locally successful methods of establishing plantations of any major
forest tree are suitable for E. globulus. Broad contour-terraces are used on
steep sites in Spain and Portugal, with seedlings planted on the outside of the
terrace to permit mechanical cultivation. In general, the higher the standard
of pre-planting cultivation the greater the response in terms of more rapid,
healthy growth of trees. Plantations on very rough ground or remote areas
can be established with the minimum of preparation; small circular areas are
cleared, a hole dug in the centre, a small amount of a compound fertilizer
mixed with the soil and a sturdy seedling planted. If possible, the cleared
area should be sprayed with a herbicide to reduce weed competition during
establishment. Protection of plantation seedlings may also be necessary as E.
globulus is browsed by rabbits, small animals and sometimes goats, but is
generally unpalatable to deer and other domestic stock.
Myrtaceae 285
Seed may be sown in trays and seedlings transplanted into tubes or pots at
the four-leaf stage. Seed can also be sown into specially prepared seedbeds,
transplanted into nursery beds at the four-leaf stage, to remain until required
for planting out. In India, a seedbed 1.2 X 12 m requires 100 g of viable
seed, sown 0.5 cm deep in 7.5 cm rows; producing approximately
9000-10,000 seedlings. Seed normally germinates in 7-10 days and all
viable seeds in 1 month. Seedlings can be planted out by hand or mechani-
cally, the inter-row wide enough to allow machinery to be used with
between-plant spacings of 3-5 m, the main limiting factors being soil fertility
and available soil moisture. Weeding is essential for the first year, preferably
two, and it is not necessary to clear-weed the inter-row; slashing growth to
ground level is sufficient once seedlings are established.
Harvesting
Unlike E. polybractea most E. globulus is grown for timber, and foliage
obtained from felled trees. The increasing number of self-sown trees,
especially in South America however, are harvested whenever proximity to a
collecting point or distillery makes the foliage saleable. Data on foliage yield
per tree at varying felling ages are scanty; in Spain trees 6-8 years of age
yielded 20-60 kg leaves; elsewhere, foliage yield is invariably given as the
average per hectare. Oil with highest cineole content is obtained from foliage
growing at the ends of branches and the top of the canopy, and selective
harvesting of these leaves produces the highest quality oil. Where timber is
the main product this system could be most profitable, since there will be
less green material, labour and transport costs, and still capacity smaller.
Since felling is frequently year-round, this ensures a continuous supply of
foliage, but oil yield and cineole content are highest in warmest and lowest in
the coldest months. When grown for paper, pulp or firewood, a coppice
rotation of 10-1 2 years is common in Portugal; in South America 5-1 2 years
dependent on growth and use, i.e. firewood, charcoal, pit props. When
grown to supply timber or telephone poles, the rotation is 15-30 years.
Distillation
Well-dried and properly stored leaves normally have approximately twice
the oil content of fresh leaves from the same batch of foliage, and similar oil
characteristics and constituents, but there are no data on how these are
affected by delays in the field. In South America, cut branches are com-
monly left in the field for 2-3 days before being stripped and transported to
the distillery; elsewhere excess leaves can be kept in covered stores for up to
a week. As crude globulus oil is normally rectified, variation in constituents
is relatively unimportant. Crude globulus oil has an unpleasant odour due to
the presence of lower aliphatic aldehydes, which can be removed by first
286 Chapter 8
drying over anhydrous sodium sulphate, then distilling over 1-2% caustic
soda, which produces a dark-brown syrup, eucalyptus tar or resin oil, used in
industrial soaps and disinfectants.
IS0 BS EOA
Apparent density 0.904 0.905
(20") 0.924 0.925 (25")
Relative density 0.906 - -
(20°C) 0.925
Refractive index 1.4590 1.4580 1.4580
(20°C) 1.4670 1.4700 1.4700
Optical rotation 0" -5" -2"
(20°C) t1O0 t10" t10"
Solubility (viv 70% ethanol.20"C) 1:5 1 :5 1 :4-5
Cineole content 70% min 70% min 70%
ISO, Indian Standards Office; BS, British Standards Institute; EOA, Essential Oils Association.
Figures in columns are range.
Eucalyptus polybractea
Botany
Eucalyptus polybractea is a multi-stemmed species to 5-10 m, which may
form almost pure stands of 5000 treesiha, but is usually associated with
other mallee species. The tree typically develops a number of similar-sized
stems from one rootstock (large lignotubers) to form the distinctive mallee
growth; the stems are much branched from low on the trunk. The lower bark
is somewhat rough, thin, fibrous, tight, and light to dark grey; the upper bark
is smooth, light grey, pinkish grey or orange tinted. The only use for the
wood is as fuel. Trees when cut coppice quickly and profusely, and are thus
most suitable for mechanized harvesting. The foliage is relatively dense on
many branches. Young leaves which develop from stumps have a bluish
tinge, hence the common name. Mature leaves are petiolate, narrow-
lanceolate, 100 X 10 mm, concolorous, densely reticulate, and a dull bluish
grey or bluish green. The leaves are highly aromatic when crushed. The oil
yield from steam-distilled fresh young foliage and twigs averages 1.5-2.5%
but 5% from managed natural stands, with a 1,8-cineole content of 70-90%.
Other significant constituents are p-cymene, terpinen-4-01, alpha-pinene,
limonene, beta-pinene, and sabinene; there are 44 compounds in amounts
greater than 0.01 %. The main characteristics are shown in Table 8.14.
288 Chapter 8
Ecology
In Australia, E. polybractea is confined to small areas between 34 and 37"S,
at altitudes of 150-350 m with an average rainfall of 470-540 mm and few
long dry periods. The mean maximum temperature is 29-32OC and mean
minimum 2-4"C, with frost occurring at higher altitudes. Rainfall or avail-
able soil moisture directly affects oil yield as the following comments from a
Victorian still operator's records lodged with the Forestry Department
indicate: 'In a wet year, trees growing in hard country increased their oil
yield per tonne of leaf by 1-2.5 kg, whilst trees growing on granitic sands
decreased by 1-2.5 kg per tonne. In a dry year trees on ridges yielded 5-7 kg
oil per tonne whilst those growing in gullies yielded up to 23 kg oil per
tonne.'
stands of blue mallee, stated that sites close to stills regularly harvested since
the end of World War One showed no obvious loss in either leaf or oil
production (Ferguson, 1985).
Cultivation
Eucalyptus polybractea oil is obtained mainly from natural stands, which are
first heavy rolled or chain-felled, debris allowed to dry and burnt; then
cleared of stones, large stumps etc. which would impede the use of a forage
harvester. When establishing new plantations E. polybractea is normally
raised from seed in special containers, since direct sowing has not been
successful. Seed is very small, viability 50-80%, and enough seed sown to
provide at least 12,000 treesiha. Cuttings can also be taken from young
suckers growing close to rootstocks and the strike rate is usually high. Such
suckers are especially useful for infilling blanks. The following information is
based on New South Wales Department of Agriculture recommendations
(Small, 1986a). Seedlings can be raised in pots (peat or paper) or as open-
root stock. Establishment losses with potted seedlings are less than 5% but
can be 30% with open-root seedlings. Coarse river sand is used for 10 cm
deep, well-drained nursery beds which should be protected until the seed
has germinated and the seedlings established. Seed is sprinkled on the
surface at 2 g/m2 (yielding about 2000 seedlings), or sown directly into
containers or seed boxes and pricked out at 4-6 weeks.
T o prepare lands for planting, all roots, stumps and similar debris should
be removed, ploughed or deep-disced, and this can be limited to row lines.
Where sheet erosion may occur, contour banks are necessary or ripping on
the contour may be sufficient, leaving small ridges raised by the tines to
direct surface flow. Immediately before transplanting, row lines should be
disced to break up a surface pan, kill weeds and provide a seedbed for
mechanical planting. Seedlings are trimmed to 15 cm before transplanting
to give 9000 treesiha in twin or single rows to suit harvesting machinery.
Weed control is important in the first year, and mowing is sufficient. Short
periods of intensive sheep grazing is an effective method of controlling
annual weed growth in established trees.
Harvesting
Hand harvesting foliage is now restricted to areas too small or too uneven to
use machinery or from trees felled for other purposes; a worker is expected to
cut and load 0.6-1.0 tonne per 8-hour day for transport to the still. A few
‘eucy factories’ remain on permanent sites in south east Australia, consisting
of two vats 4 X 3 m in diameter sunk into a pit lined with bricks. The vat lid
is made from concrete or iron, and sealed in place with mud. Steam from an
adjacent boiler at a maximum pressure 350 kPa passes up through the
charge into a condensing pipe submerged in an adjacent dam. Distilling lasts
290 Chapter 8
2-4 hours, loading and unloading the vat 1-1.5 hours. Although unsophisti-
cated, these stills are effective, cheap to operate and produce good quality
oil. The main drawback is the necessity to build a dam to fill with rainwater
from runoff, or locate a creek with a reliable water supply which could be
dammed, not easily found in outback Australia. In dry years the amount of
water restricted distilling and thus all other operations, and work usually
ceased until sufficient water became available. Such shutdown periods have
little effect on the operator, as maintenance costs for the distillery or
harvested area are negligible, only family labour employed, and it is these
special circumstances that allow such small distilleries to survive.
In a mechanized operation, trees are initially coppiced to bring older-
growth mallee into production. In plantations, trees are coppiced after
15-1 8 months to a height depending on the equipment used, and this is the
first harvest. Second and subsequent cuts are every 12-24 months, usually
the longer. For forage harvesting on level land, coppice height can be down
to 5-8 cm, but on uneven ground it may be 30 cm and tractor tyres
‘overcoated’ with old tyres to prevent spiking. Oil yield increases for several
years after initial harvest of coppiced natural stands because most foliage
harvested is high-oil-content juvenile leaves. Harvesting can continue all
year round. Partial mechanization has proved profitable and has been
adopted by several medium-sized operators. Coppiced growth approx-
imately 1-1.2 m high is harvested by chain saw, mechanically shredded and
fed into a mobile distilling vat, usually a rectangular steel vessel with wheels
and a towing attachment. A grid fitted about 30 cm above the floor supports
the charge, and steam pipes are set below the grid. A hinged rear wall swings
open to allow removal of spent leaf. Two men working for approximately 8
hours can fill one vat which is then towed to the still. Two vats are used, one
being distilled while the other is refilled. A 2700 kg charge, steam distilled for
6 hours, yields 90-100 kg oil (3-4%). This system is also most commonly
used for harvesting E. dives and is suitable for many other eucalypt species,
since it is very flexible and inexpensive. The limiting factor is a constant
supply of water in sufficient volume to allow regular operation.
The most successful and profitable operators have wholly mechanized
harvesting and distilling. Coppiced growth to 5 cm diameter is cut by a
heavy-duty forage harvester, and although this tends to bruise leaves and can
reduce oil yield by 5-10%, this is offset by the operation’s speed, ease and
cheapness. After cutting, foliage is fed directly into a 3 t capacity mobile vat
which is towed to the distillery and positioned in a bay where a suspended lid
is lowered and sealed. A flexible outlet connects the lid to a stainless steel
multitubular condenser (Fig. 8.7). Steam is passed through the leaf, requir-
ing about half an hour to heat the charge and a further hour and a half to
complete distillation. Spent leaf fires the boiler, or is spread on harvested
areas as mulch. Usually 5-6 vats are distilled per 8-hour day; a two-bay, four
vat system should easily maintain a production rate of 200 kg of oil per day.
Myrtaceae 29 1
Eucalyptus radiata
Eucalyptus radiata Sieb. ex D C . is commonly known as narrow-leaved
peppermint or black peppermint. There are two legitimate subspecies, E.
radiata subsp. radiata (syn. E. amygdalina var. nitida; E. radiata var. phellan-
dra) and E. radiata subsp. robertsonii (Blak) J&B. E. radiata subsp. aus-
traliana or E. australiana B.& S., frequently referred to as an oil producer in
many texts, is now considered a synonym for E. radiata and will not be
mentioned separately. The subspecies australiana was commonly referred to
in Australia as ERVA in non-scientific literature.
Botany
Eucalyptus radiata is a small to medium-sized woodland tree to 10-1 5 m,
(sometimes a larger forest tree to 20-30 m), with a central stem to 1 m DBH,
Vapour pipe
a n
Cold
water
Fig. 8.7. Integrated harvesting and distillation unit. (Courtesy of Dr B.E.J. Small.)
Myrtaceae 293
which often forks low down, with many large branches which themselves
branch forming a large spreading tree. The bark is generally rough on the
trunk and larger branches, persistent, finely fissured longitudinally and grey
or grey-brown; it is smoother and greyer on the upper branches (Boland et
al., 1984). The wood is pale brown darkening on exposure, the gum veins
are common and it is soft, with a density of 720 kg/m3 and generally used as
fuel.
The foliage is dense and often retained nearly to ground level on smaller
trees growing in open woodlands. The mature leaves are narrow to broadly
lanceolate, 150 X 15 mm, thin, horizontal, concolorous, and pale to me-
dium green. T h e venation is oblique or semi-longitudinal and densely
reticulate. T h e leaves smell strongly of peppermint when crushed. T h e oil
yield from steam-distilled fresh foliage and twigs is to 3.556%; yield from dry
material in natural stands is usually below 1.5% in Australia, and lowest
yield in any district is in the coldest months. In South Africa, a selected tree
provided 4.5 kg dried leaf with 10% oil when harvested 18 months after
being coppiced (Donald, 1980). The main constituents and characteristics
of the oil are shown in Table 8.15 and have been described in detail (Boland
et al., 1991).
T h e inflorescence is axillary, simple, multiflowered, white, peduncles
terete, to 8 mm, the buds pedicellate, clavate to 6 X 4 mm, with no scar. The
operculum is hemispherical to conical and slightly beaked. Main flowering is
October-January in Australia. The fruit is pedicellate, cupular to truncate-
globose, to 6 X 6 mm, the disc level and the rim thick, with three to five
valves.
Ecology
T h e species occurs naturally in eastern Australia usually below 1000 m in
areas with an annual rainfall of 650-1 100 m m with few dry periods a t higher
altitudes but to 5 months in drier and lower elevations, and average summer
temperature 24-27"C, with winter frost common at higher elevations. T h e
species has been planted in other countries, grows well and produces high
quality oil under suitable conditions, e.g. in Stellenbosch, South Africa
(Donald, 1980).
Cultivation
Only wild trees are harvested in Australia and methods used to obtain leaves
and distil oil from E. globulus or E. polybractea are suitable. In South Africa,
seedlings were raised in seedbeds, pricked out into pots at the four-leaf stage
and later planted out at 3 X 1.5 m. Seedlings were mechanically planted
through a plastic mulch which had been machine-laid along rows over a
previously installed drip-irrigation line. Trees were irrigated when weekly
rainfall fell below 25 mm, and cultivated twice annually. After 1 year many
exceeded 4 m but all were topped to 30 cm; the second harvest was 18
months later when regrowth reached 7 m. The average herbage oil content
was 9.84%. The results indicated a properly managed high-yielding clonal
plantation can yield 1000 kg of oil per ha per cut, with a cineole content
averaging 90%.
Eucalyptus smithii
Eucalyptus smithii Baker is commonly known as gully-gum or Smith’s gum
and is a medium to tall forest tree, frequently found in gullies and valleys,
hence the common name.
Botany
Trees normally have a single trunk to 45 m, 1.5 m DBH, but a mallee form
may occur on exposed sites. Branching begins high up, with a compact
crown. The bark is rough, dark-brown, persistent, tight, with narrow longi-
tudinal fissures, decorticating from the upper trunk and branches in long
ribbons hanging conspicuously from the crown. New bark is smooth, white
or creamy white. Freshly stripped bark has a strong, pleasant odour of
peppermint. The wood is pale, close-grained, hard, density not recorded,
moderately durable; the timber of limited value, and susceptible to boring by
Lyctus spp. The foliage is generally sparse although isolated trees develop
more branches and a spreading crown. Mature leaves are narrow-lanceolate,
180 X 15 mm, concolorous, dull green and moderately reticulate. The
leaves are highly aromatic with a peppermint scent when crushed. The oil
Myifaceae 295
yield from steam-distilled fresh young leaves and twigs is 2-4% in Australia,
1-1.50% in Zaire and to 1.75% in Guatemala. The major constituent is
1,g-cineole and the main oil characteristics are shown in Table 8.15.
The inflorescence is axillary, simple, usually with seven white flowers; the
peduncles are slender, angular or flattened, to 12 mm. The buds are
pedicellate, clavate, diamond or oval and 7 X 4 mm. A scar is present and the
operculuin is conical. The main flowering is January-March in Australia.
The fruit is pedicellate, ovoid to subglobular, sometimes campanulate to 7 X
7 mm, with the rim thick, the disc broad and usually with three valves.
Ecology
The species occurs naturally in south-east Australia on the seaward side of
coastal table-lands between 34 and 38"S, up to 1200 m, with a rainfall of
750-1500 mm annually and few dry periods, but frost at higher elevations.
The species has been planted in Zaire at 2000 m, annual rainfall 1500-2000
mm and made excellent growth.
Cultivation
The species is not cultivated in Australia, but oil is obtained from plantations
established in Swaziland, South Africa, and on a smaller scale in Argentina
296 Chapter 8
and Uruguay. Generally, E. smithii makes faster growth than E. globulus and
its foliage also contains more oil of a similar or higher cineole content. Given
suitable conditions, it could be more profitable for oil production under
plantation management than E. globulus. In Australia, foliage is obtained
from wild trees which are felled and allowed to coppice, then harvested
annually as for E. dives. In Zaire, Kivu Province, plantation trees were
topped at approximately 3 m then allowed to coppice, and cut twice
annually giving a total oil yield of 150-600 kgiha. The low 1.0% oil yield was
stated to be partly due to inclusion of considerable woody material in stills to
facilitate distillation using low-pressure steam. In Swaziland and South
Africa, plantations are operated using methods described for E. globulus or
E. polybractea. In Guatemala, the tree is treated similarly to E. globulus.
Eucalyptus viridis
Botany
Eucalyptus viridis is a medium-sized tree of typical mallee habit, to 3-9 m but
up to 15 m, normally with a number of stems from a common rootstock,
each much branched from low down. The bark is rough, thin, blackish, with
small longitudinal fissures, persistent on lower stems; becoming papery,
ribbony, finally smooth, grey to pinkish grey with height. The wood is
yellow, darkening with age to brown, and used mainly for fuel. Mature
leaves are a clear slightly glossy green, narrow, linear to linear-lanceolate,
partially erect, 100 X 8 mm, alternate, concolorous, densely reticulate and
with prominent oil glands. The leaves are highly aromatic with a strong
peppermint scent when crushed. The oil yield from steam-distilled fresh
leaves and twigs is 1.00-1.25%. The major constituent is 1,8-cineole to
85 %; minor constituents are alpha-pinene, cymene, sesquiterpenes, cumi-
nal, phellandral and 1-cryptone. The main characteristics are shown in Table
8.15.
Myrtaceae 297
Ecology
This species occurs in large but scattered stands in eastern Australia, is
remarkably drought resistant, and often found in arid areas bordering
deserts. It is excellent for low windbreaks or shade trees in dry regions, or
where long dry periods are fatal to other species. These advantages, how-
ever, are not widely appreciated, since there is no record of the species being
grown outside Australia.
Cultivation
As for E. polybractea, and where E. viridis co-exists the two are harvested
indiscriminately since the oils are basically similar.
together. The main characteristics are shown in Table 8.16 which is com-
piled from published and unpublished sources.
Eucalyptus elata
Eucalyptus elata Dehnh. is commonly known as river white gum and is
naturally restricted to the higher coastal areas of south-east Australia. It is a
medium to tall tree to 30 m but may reach 45 m, trunk to 0.5-1 .Om DBH.
The bark frequently strips naturally from the trees, leaving a clear, smooth
greyish-white trunk, giving the species its common name. It prefers moist
but not waterlogged conditions in valleys and alluvial flats. The oil has a high
piperitone-piperitol-phellandrene content, and is similar in composition
and characteristics to E. dives Type oil, and also described as E. numerosa
Type A.
Eucalyptus leucoxylon
Eucalyptus leucoxylon F.Muel1. is commonly known as yellow gum. There are
a number of subspecies differing little in appearance, which are not differ-
entiated by local oil producers and harvested indiscriminately where they
coexist. It is a medium-sized woodland tree to 16 m in wetter areas of
southern Australia, on granite sands and loams. It grows well in coastal areas
of south-eastern USA, but only isolated specimens have been planted. The
wood is pale brown with yellow or pink tints, and is hard, strong and durable,
with a density of about 1010 kg/m3. It is used for structural engineering,
Myrtaceae 299
fences, railway sleepers, building and poles. The species commonly occurs
in association with E. sideroxylon and E. elaeophora, when foliage is harvested
and distilled together; thus undiluted E. leucoxylon oil is s ldom available.
P
Oil (or mixture) is pale yellow with a somewhat harsh odour; the main
constituent is cineole to 75%, plus alpha-pinene and sesquiterpenes. The
honey is a pale straw, dense clear and mild, candies somewhat quickly and is
of very high quality.
Eucalyptus macarthurii
Eucalyptus macarthurii D. & M. Camden is commonly known as woolly-butt
and is a medium to tall tree to 40 m, restricted to a very small rather wettish
area of New South Wales a t an average altitude of 600 m. The tree coppices
strongly and produces substantial amounts of foliage. The wood is very pale,
moderately heavy but of low durability, and the timber is of little value. This
species has been introduced into other countries, has proved surprisingly
frost hardy, and is recommended for planting in southern Russia and South
Africa. Only in South Africa, Zaire and Brazil was commercial oil produc-
tion attempted. Analysis of a Russian oil gave (percentage in brackets):
1,8-cineole (30.2); terpene alcohols (1 5.9); terpenyl acetate (8.9); globulol
(8.6); sesquiterpene hydrocarbons (3.9); alpha-terpineol (3.2); alpha-
pinene (2.5); sesquiterpene alcohols (2.5); palustrene (1.7); camphene
(0.26); p-cymene (0.18); beta-pinene (0.07) (Antonov and Zhdanov, 1986).
This oil contained terpenes not found in Australian oil from this species, see
below.
Eucalyptus macarthurii oil is one of the few eucalypt oils used exclusively in
perfumery. The bark contains an oil similar in composition to leaf oil but
much lower in yield, usually below 0.15%. The crude oil is yellowish to pale
orange, occasionally more reddish. Rectified oil is generally pale to medium
yellow, containing 80% geranyl acetate; the odour is fresh with a sweet rosy
scent. Unrectified oil has a much coarser odour due to presence of lower
aliphatic aldehydes. Australian leaf oil usually contains 44-56% geranyl
acetate, 28-40% alpha- beta- and gamma-eudesmol collectively, plus cit-
ronellal to 2%, geraniol to 5% and isovaieraldehyde between 2-3%. The
alpha- and beta-pinenes are not found in this oil (Boland et al., 199 1). Oil
constituents can vary widely, and may be more than a regional or environ-
mental reaction, since in India three distinct chemovars occur (Madan et al.,
197 1). These contained as major components, geraniol and geranyl acetate;
eudesmol and geranyl acetate; or 1,8-cineole. Selection to produce an oil
high in geranyl acetate and free of cineole would be of considerable value to
perfumers.
Eucalyptus maidenii
Eucalyptus maidenii F.Muel1. is commonly known as maidens gum and is a
tall to very tall forest tree to 75 m restricted to a small warm to cool-humid
300 Chapter 8
area on the coastal borders of New South Wales and Victoria. The wood
light brown, hard, moderately strong and moderately durable, with a density
of about 930 kg/m3, and is used for general construction purposes. The
species produces a clear golden honey, profusely from pure stands. Euca-
lyptus maidenii has been introduced into many countries as a timber or
amenity tree, but only in South Africa and South America was oil produced.
There is little reference to E. maidenii in South America and it is usually
grouped with E. globulus in statistics, although specialist growers sell the oil
as E. nzaidenii oil. The main constituents of the oil are cineole to 55% and
alpha-pinene, and the major uses are as for E. globulus oil with which it is
frequently bulked.
Eucalyptus sideroxylon
Eucalyptus sideroxylon Cunn. ex Woolls. subsp. sideroxylon is commonly
known as red ironbark and is a medium-sized woodland tree to 25 m whose
foliage is harvested with other species where they co-exist. The species has
been planted in several countries for its very hard, durable timber used
extensively in heavy engineering and railway sleepers. The wood has a
density of about 1170 kg/m3. Oil was produced commercially in South
Africa for use as a flotation agent in the mining industry. The main
constituents are cineole to 7 5 %, alpha-pinene and sesquiterpenes. Analysis
of Indian leaf oil showed 98 constituents, with 1,8-cineole being the most
important (Gupta et al., 1990). Two types of oil are produced by E.
sideroxylon in Spain; that from red flowering trees is high in cineole, and that
from white flowering trees is high in phellandrene (Garcia-Martin and
Garcia-Vallejo, 1977). The red flowering variety may be named E. side-
roxylon var. rosa in botanic gardens and arboreta, but this is a horticultural
designation only. The species produces high quality honey, pale and of good
density.
Eucalyptus staigerana
Eucalyptus staigeruna F.Muel1, ex Bailey is commonly known as lemon-
scented ironbark and is a medium-sized tree to 20 m, whose leaves are highly
lemon-scented. The major drawbacks to its wider cultivation are slow
growth, rather straggly appearance and lack offrost resistance. It is grown for
its oil in Guatemala, the Seychelles, Brazil and Zaire. Oil has not been
produced commercially in Australia, although it has a much higher leaf oil
content than E. citriodora, to 4% from fresh foliage. The main constituents of
Australian leaf oil are (per cent): geraniol (9-18), methyl geranate (1 1-18),
geranyl acetate (4-14), limonene (1-14), beta-phellandrene (12-34) and
neral (8-12). Over 30 of the compounds identified in amounts greater than
0.01% are monoterpenes (Boland et al., 1991). Eucalyptus staiger ana was
Myrtaceae 301
Eucalyptus tereticornis
Eucalyptus tereticornis Sm., commonly known as forest red gum, is native to
eastern Australia and Papua New Guinea. It occurs on medium to heavy
soils, and is drought and frost resistant. It is a medium to tall tree to 40 m
with an open crown and is well branched, with the bark mottled white and
grey with a bluish tinge. T h e leaves are linear to 20 cm, thick, glossy green,
and conspicuously veined. The panicles have 4-8 white flowers, and the fruit
is a horned capsule. The species is widely planted in India and when
plantations are felled the foliage is distilled. T h e oil is rich in pinenes, and
total potential oil production is estimated at 4700 t to produce almost 3000 t
pinene (Shiva et al., 1989).
The following Eucalyptus spp. have also been distilled for oil, some as an
admixture with the dominant local harvested species, and a detailed descrip-
tion of many other Eucalyptus leaf oils has been published (Boland et al.,
1991). Eucalyptus cneorifolia DC. (Kangaroo Island mallee) has oil with a
cineole content to 75%, plus alpha- and beta-phellandrene and p-cymene.
An attempt to cultivate this species for oil production in Morocco was
unsuccessful. Eucalyptus elaephora F.Muel1. (now regarded as a synonym of
E. goniocalyx F. Muell ex Miq.) has an oil high in cineole, 70-85%, free of
phellandrene and other terpenic hydrocarbons, and volatile aldehydes are
very low. A hybrid (unspecified) of the species in Spain reportedly gave oil
with a cineole content of 79-93%. T h e foliage of Eucalyptus deglupta Blume.
(from Kamarere, Papua New Guinea) has an oil content between 0.4 and
0.8%, and is very high in nerolidol to 45%, plus alpha-pinene, alpha-
phellandrene, p-cymene and ocimene. There are a number of varieties of
Eucalyptus oleosa F. Muell ex Miq. (red mallee) which produce an oil high in
cineole, between 45 and 95%. The dry leaves ofEucalyptus youmanii Blak. &
McK. (Youman’s stringy-bark) have a rutin content of 20%, and several
unsuccessful attempts have been made to commercially extract the rutin
from foliage collected from natural stands.
302 Chapter 8
Melaleuca
The generic name Melaleuca L. is derived from the Greek melos (dark, black)
plus leukon (white). This apparently contradictory name was given to the
first species described, M . leucadendron (L.) L., which had white papery bark
on the higher stem and branches and a black lower trunk, a combination
which commonly occurs in Australia due to fire. The genus is predominantly
Australian, but some species occur in Asia, South East Asia and the Pacific
region. The 200 species currently described are widely distributed in Aus-
tralia (Byrnes, 1986), but a proposed revision could substantially increase
this number. The common generic name of paperbark refers to the paper-
like bark which can often be peeled from larger trees in broad strips. Leaves
of many species contain essential oil but only M . alternijolia (tea-tree oil), M .
cajupti (cajuput oil) and M . quinquenervia (niaouli oil) have been commer-
cially exploited. Cajuput oil was first to attain commercial importance for its
medicinal and pharmaceutical uses; niaouli oil has similar properties but to
a lesser extent; tea-tree oil has many medical uses and is commercially
produced in Australia.
Melaleuca spp. with leaves rich in terpinen-4-01 and 1,8-cineole were
reportedly used by Australian Aborigines to relieve colds and headaches; the
leaves were simply crushed in the hand and the volatile oil inhaled. This
usage was continued by early European settlers, who also made infusions of
crushed leaves (Lassak and McCarthy, 1983). The name tea-tree was first
used by Captain Cook in his ‘Voyage Toward The South Pole’ in 1777,
when he recorded the use of Melaleuca or Leptospermum leaves to brew a ‘tea’
Myrtaceae 303
he instructed his men to drink to prevent scurvy. Later settlers also used
Leptospermum leaves as a tea substitute during periodic shortages of true tea.
The common name of tea-tree applies to a number of species of Lep-
tospermum and Melaleuca. Ti-tree is incorrect when applied to the trees or
their oils, as this is the Maori name for the New Zealand cabbage-tree,
Cordyline australis Hook. f.
The first oil distilled from Australian Melaleuca was from M . alternifolia in
1924 (Penfold, 1925). Since then the composition, characteristics, and
actual and potential uses of oil from a number of other species have been
determined, and the results are available in the literature. Some important
characteristics and major components of oil from selected species are
included at the end of this section.
Many Melaleuca spp. have a number of chemotypes, the oils differing
substantially in their major components; they also cross-pollinate readily
and seedlings may thus have foliage of varying leaf-oil content and composi-
tion. T o obtain plants with uniform high oil content leaves, vegetative
propagation of selected parents is necessary. Cuttings may be used although
the strike rate is erratic for undetermined reasons, and tissue culture has
been successful in the laboratory, but there is currently no viable method of
large-scale commercial production of seedlings.
The Melaleuca spp. discussed have a number of common features (Fig.
8.8). The trees are normally evergreen, the foliage is sometimes sparse, and
many leaves are shed in very dry years. Most species flower profusely and
produce abundant nectar and pollen, thus they are a favourite of Australian
bee-keepers for their excellent honey. A characteristic is that the flowering
spike may continue to grow after fruiting, and become a leafy branchlet. The
bark is distinctive, and in many species consists of thin paper-like layers of
cork separated by thin fibrous layers, which may reach 5 cm in thickness
before peeling naturally. Outer layers are loose, frequently torn, ragged and
partially unrolled on the lower trunk. This papery bark had a number of
traditional Aboriginal uses and is now in demand by gardeners as a filling for
hanging baskets; the author so uses bark from trees in his Melbourne garden.
The bark is potentially a raw material for the production of fibreboard and
insulation. Timber of larger trees is moderately dense, 700-850 kg/m3, hard
and often cross-grained, but sawn timber does not readily season. It was
formerly used for a wide range of building and similar purposes as the wood
is termite and water resistant, but the main current use is for domestic
firewood.
Despite the widespread occurrence of Melaleuca spp. in Australia, little
was known regarding insect pests or the degree of damage inflicted by a
particular species. Recent research indicates there are some 400 species of
herbivorous insects associated with M . quinquenervia which also occur on
other members of the genus, and it appears that no one species causes major
damage but the combined effects of the normal level of infestation of all is
304 Chapter 8
Melaleuca alternifolia
Botany
Melaleuca alternijolia (Maid. & Bet.) Cheel. (syn. M . linariijolia var. alter-
nijolia Maid. & Bet.) is a shrub or small tree to 5-7 m, occasionally higher,
usually with a single trunk but multiple stems may originate from a common
Myfiaceae 305
rootstock; the trees coppice quickly and vigorously. Branching may be high
or low, few or many, and in dense stands there are few branches per stem
which grow almost vertically. Similarly to other trees growing in swampy
areas, the roots a t base of the stem are often partly exposed and adventitious
roots occasionally occur. The amount of foliage is basically governed by
growth of individual trees and is limited to the crown and the ends of
branches in thick stands, but on isolated trees may be a dense mass from low
down. Mild trees cut for foliage are usually ready for re-cutting in 16-24
months, half this period in plantations. The leaves are green or bluish green,
glabrous, soft, very narrow, usually less than 1.5 X 20 mm, alternate, entire,
and stoutly petiolate. The oil yield from steam-distilled fresh foliage and
twiglets is 2.0- 3.0% but from individual trees to 5%; the main constituent
is terpinen-4-01, at 25-45%, but can be higher from individual trees. The
main characteristics and major components are shown in Tables 8.17 and
8.18. One minor component is viridiflorene not previously known to occur
naturally. There is considerable variation in cineole content, apparently
related to habitat.
The flowers are generally creamy-white, sometimes yellow tinted, sessile,
in open or condensed multi-flowered terminal spikes, giving a fluffy appear-
ance to a tree in full bloom. Main flowering in Australia is October-
December. The fruit is a woody capsule, globular, usually less than 0.5 mm
diameter, occurring in elongated clusters around branchlets. The seeds are
brown, normally 10,OOOig and viability is generally poor.
A B C D E
Alpha-pinene 2.1 1.8 2.2 2.7 2.0
Beta-pinene 0.4 0.4 0.7 1 .I 0.3
Sabinene 0.2 0.1 0.1 ng ng
Myrcene 0.4 0.5 1.5 2.5 ng
Alpha-phellandrene 0.8 0.6 ng ng 0.9
Alpha-terpinene 7.1 6.8 1.7 0.3 5.7
Limonene 1.4 0.9 3.7 7.9 1.5
1,8-Cineole 3.0 7.0 30.3 64.1 3.5
Gamma-terpinene 15.7 13.3 4.3 0.7 10.9
p-Cymene 6.2 2.4 5.6 0.6 13.9
Terpinolene 3.4 2.5 0.9 0.2 2.6
Linalool 0.2 0.1 ng 0.3 0.9
Terpinen-4-01 45.4 52.9 18.0 1.7 38.2
Alpha-terpineol 5.3 5.8 6.3 11.8 6.2
A, commercial oil low in 1,8-cineole; B, from selected trees: C, from selected trees, medium 1,8-cineole;
D, from selected trees, high 1 ,8-cineole; E, commercial (adulterated) oil: ng, not given.
Source: Williams and Home (1988).
Ecology
Melaleuca alternifolia is naturally confined to the warmer, wetter, east coast
of Australia, between 32 and 27"S, seldom above 300 my and often forms
dense almost impenetrable thickets. The mean summer maximum is
27-3 1"C, and mean minimum 17-19°C; mean winter maximum is 18-2 1"C
and mean minimum 6-7 "C, and the species is frost sensitive. The average
annual rainfall is 1000-1 600 mm, falling mainly in the Australian summer
and autumn. The species has not been planted outside its natural habitat
except as a specimen tree in arboreta or botanic gardens, and such trees
often grow much taller and more upright than wild trees. Leaf oil content is
lower in cooler than warmer months, but seasonal variation is less than that
due to variation in leaf age at harvest. Data suggest rainfall in the 6 months
preceding cutting has a significant effect on total oil yield (Small, 1981,
1986b). Substantial short-term variation in oil content of foliage occurs, and
the reason for this is unclear. What is important is the effect on plantation
profitability, as the oil content of dried foliage was 2.9 and 1.8% on two
successive days and 1.8 and 3.4% on two others. Cutting on a low-oil-
content day would thus have reduced by 33-50% the amount of oil pro-
duced (Murtagh and Ethrington, 1990). Thus daily determination of oil
content during cutting should be routine.
Leaf oil composition of the species changes from north to south of its
natural range; in the north oil is high in terpinen-4-01 and low in 1,8-cineole,
southward the cineole content gradually rises until finally it is similar to that
Myrtaceae 307
Cultivation
Oil was initially obtained exclusively from wild stands which are sufficiently
extensive to supply existing demand. Branches were slashed and roughly
stripped of leaves, and the mass of material transported to stills similar to
those used for distilling eucalyptus foliage. The work was labour intensive,
unpopular and, as wages rose, oil production became unprofitable. Com-
mercial plantations have since been established, and oil output can now
satisfy growing world demand. Cultivations necessary to establish a planta-
tion of M . alternifolia are similar to those for any plantation crop. In
Australia, an annual cash crop may be sown in the first and perhaps second
year to obtain a return to offset clearing costs prior to establishment (C.
Dean, Ballina, New South Wales, personal communication),
Seed can be directly sown, but this is not recommended as the seed is very
small and thus difficult to sow evenly or at the correct spacing. Seed from
selected parent trees is also very limited, and direct sowing wasteful of this
important asset in plantation profitability. Seedbeds are a mixture of coarse
sand and loam, treated with herbicide, and preferably fumigated. Alterna-
tively, seedlings can be raised in pots from seed or cuttings, and tissue
culture of selected high-yielding parents has been successful. The choice of
seedlings from nursery beds, potted seedlings or cuttings depends almost
entirely on circumstances. Establishment losses using open-rooted seedlings
can reach 30%, below 5% from potted seedlings; however, the cost of the
latter is usually much higher than nursery-raised plants. Seed is sown in
308 Chapter 8
nursery beds at 0.5 glm', and a watering-can fitted with a fine rose is very
effective to ensure even distribution. Seed sown in seed boxes is pricked out
to individual pots when 4-6 weeks old (Fig. 8.9). Optimum times for
transplanting in Australia are December-February, and May-July. A me-
chanical transplanter is used, and seedlings prepared to suit the equipment
(Fig. 8.10). Soil should be moist to 50 cm or irrigation available. Seedlings
can also be watered-in at planting, using commercially available equip-
ment.
Row spacing should suit the equipment to be used for inter-row cultiva-
tions and harvesting, and in-row spacing adjusted to ensure optimum plant
population. For M . ulternijolia in Australia, this is at least 35,000 trees/ha.
Close in-row spacing promotes leaf growth and thus oil yieldiha, and
maintenance of an optimum population is also a major factor in plantation
profitability. Weed control is essential until seedlings are well established,
preferably until the first harvest. Mechanical and chemical methods can be
used, but once trees are well grown and being regularly cut, a mown grass
cover is sufficient and assists in controlling soil erosion.
Harvesting
The method of harvesting commercial plantations of E. polybructea is suita-
ble (Figs 8.11, 8.12). The first cut is at 15-18 months, subsequently at
12-15 month intervals. Plants should be cut when stem or shoot diameter is
less than 2 cm and 5-1 0 cm above soil level. In colder regions, cutting should
be so arranged that new shoots are not developing when frost may occur.
Foliage yield in a year of average rainfall and a population of 30,000-35,000
treesiha, should be 8-10 tiha, and oil content offresh herbage 1.O-2.0%. Oil
yield per hectare rises for the first 3 years after establishment in Australia,
then remains at this level except for annual variation caused by seasonal
differences. In the first 3 years of an irrigated planting, oil yields were 70, 100
and 150 kgiha, then levelled off at about 150 kgiha. Oil yield from non-
irrigated (rainfall only) plantings was 15, 20 and 30 kgiha, then 30 kgiha.
Irrigation sharply increases oil yield but the economics of so doing must be
accurately assessed. Increased oil yield also means increased cost of harvest-
ing more foliage, and of all other charges which flow from an increased
throughput of material. The effective age of a plantation has yet to be
determined, but under good management should be measured in decades
since wild stands have been regularly harvested for 70 years.
Distillation
The system used to distil E. polybractea is basically suitable, with modifica-
tions to suit differences in distillate behaviour (Johns and Rudolph, 199 1).
In general, the rate of recovery of terpinen-4-01 and 1,8-cineole is directly
related to steam volume, whereas other components require fine chopping
company ceased to exist in 1975 all records pertaining to the company, early
history of the tea-tree oil industry and most medical and other data were
apparently destroyed. Production of Australian oil from established planta-
tions has steadily increased and totalled 200 t in 1995, most of which was
exported.
Crude oil is dark yellowish-green but steam-distilled oil is white to pale
yellowish green; the odour is spicy aromatic, combining elements of carda-
mom, nutmeg and sweet marjoram with a camphoraceous slightly bitter,
warm and spicy taste. The major components are (per cent); terpinen-4-01
(to 45), gamma-terpinene (to 25), 1,8-cineole (3-17), limonene (to 5) not
necessarily in the same sample. Oil should preferably be stored in glass or
stainless steel containers as it is a powerful solvent and interacts with a
variety of plastics. The medical and dental uses of the oil are now well
documented, a most important attribute is the ability to penetrate unbroken
skin. Detailed reviews of the medicinal properties and uses for the oil have
been published (Williams et al., 1993; Williams and Home, 1995; AUS-
TEAM Ltd, 1995).) The oil also has definite antimicrobial activity (Carson
and Riley, 1995). Although mainly used for medicinal and veterinary
purposes, the oil could be of wider interest in perfumery, as it blends well
with other oils and introduces its own distinctive note.
Melaleuca cajuputi
Botany
Melaleuca cajuputi Powell (syn. M . minor Sm., M . leucadendron var. cajuputi
(Roxb) Nied., M . leucadendron L. var. minor Sm.) is a medium sized tree to
30 m (half this in more dry areas), with usually a single trunk to 1.5 m DBH,
but several stems may grow from the original rootstock. Branching is seldom
extensive. The trees frequently have small crowns, especially in pure stands,
with smaller branches and often slender twigs. The bark is whitish, papery,
thin and flaking. Major roots can appear above ground near the stem base,
312 Chapter 8
Ecology
Melaleuca cajuputi occurs naturally from Burma and Thailand through
South East Asia to northern Australia, and has been introduced into a
number of other countries, particularly India, where it has become natural-
ized. In many it is known by one of its synonyms, and thus the accuracy of
references is difficult to determine, since the species with which it was
confused were also introduced. Wherever the species occurs it favours moist
to swampy conditions and in suitable areas forms large almost pure stands.
Swamp forests of M . cajuputi are a feature of monsoonal areas in Indonesia,
Malaysia, and Papua New Guinea. On drier areas and in bushland subject to
fire, the very vigorous regrowth of the species allows it to recover quickly,
and over time can become dominant and form dense stands. These stands
are the basis of the local cajuput oil industry since the supply of foliage is
almost inexhaustible; fires which severely damage or kill other species
merely result in a renewed supply of leaves! It is also resistant to salt-laden
winds, and is sometimes planted on shorelines as windbreaks or amenity
shelter belts.
Cultivation
There are no economically viable plantations of M. cajuputi, and all com-
mercial oil supplies are currently obtained from wild trees. Branches are
slashed, stripped of leaves, and transported to mainly portable stills. The
vigorous regeneration of cut trees has been noted and it is possible to cut-
over an area once annually if necessary. This, combined with extensive areas
under the species, easily produces enough oil to meet world demand. A very
readable description of how the industry was organized in Malaysia is that of
Penfold (Guenther, 1950). Since then there has been little basic change, and
cajuput oil production remains a cottage industry. A similar situation exists
in Indonesia, and although there have been attempts to modernize the
industry, oil of any quality is so readily saleable locally there is no real
pressure for improvement. Crude oil is also exported from Malaysia and
Indonesia for rectification mainly in Europe. Small plantations were estab-
lished in India and treated similarly to Eucalyptus, but oil yield was too low to
be profitable and it is unlikely that commercial plantations of the species will
be viable elsewhere.
314 Chapter 8
Melaleuca quinquenervia
Melaleuca quinquenervia, commonly known as five-veined paperbark, or
broad-leaved paperbark, also occurs in Indonesia, New Caledonia and
southern Papua New Guinea. This species is source of niaouli oil produced
in New Caledonia, formerly attributed to M . viridijlora or M . leucadeizdron
(Blake, 1968). Another source of confusion is that M . quinquenervia also has
several chemovars which produce leaf oils of differing composition. Thus it
is difficult to allocate data and reports in the literature to the correct species,
particularly in respect of analyses prior to 1968.
Botany
Melaleuca quinquenervia (Cav.) S.T.Blake (syn. Metrosideros quinquenervia;
Melaleuca smithii; M . maidenii, etc.) produces a large penetrating root
system, partially exposed in damp situations. Roots grow vigorously and, if
the main trunk is severed, quickly produce new shoots. Shoots can also
appear some distance from the main trunk, and thus trees quickly spread to
colonize suitable areas. This is well demonstrated in the Florida Everglades,
to which it was imported in 1906 to turn what was then regarded as merely
useless swamp into productive timber plantations. It has since colonized
almost 160,000 ha of this environmentally important area. Aerial spraying
with aboricides has been partially successful, and methods of biological
control with insects is possible (Balciunas and Burrows, 1993; Burrows et
al., 1994).
It is a medium tree to 8-12 m (occasionally to 25 m), with usually a single
stem. On trees in open stands branching occurs fairly low down and spreads
out from the stem; in dense stands the few branches are closely appressed
and give the appearance of multiple stems. The bark is typical of the genus.
The sapwood is pinkish; the heartwood pinkish grey, moderately strong,
with a density of 725-800 kg/m3.The timber is durable in water and soil but
does not season well, contains silica which blunts saws, and has few com-
mercial uses. The leaves are usually a dull green, alternate, petiolate,
lanceolate-oblanceolate, straight to slightly oblique, 5-9 X 0.6-2.5 cm,
coriaceous, typically with five longitudinal veins, and often obscured glan-
dular dots. New leaves appear silky grey due to a dense coat of fine hair.
Leaves, when crushed, have a resinous scent and a taste similar to eucalyptus
oil.
The oil yield from steam-distilled fresh foliage and twiglets is 1.5-3.5%
since there is considerable variation between individual trees; oil yield from
field stills in New Caledonia is normally below 1.5%. Similarly to M .
alternijolia there are strains or regional types whose leaf oil contains more or
less cineole; one strain common in New Caledonia yields oil with a high
1,s-cineole content. The main characteristics of oil from Australia and New
Caledonia are shown in Table 8.20. There are three major chemovars in
Australia; one produces cineole-rich oil almost identical with that from New
Caledonia, the others produce oil rich in nerolidol or linalool. The first is
readily distinguished as crushed leaves release an odour similar to that of
medicinal eucalyptus oil; the other two have a distinct sweetish floral odour.
316 Chapter 6
Four chemovars have been identified in Madasgascar, whose oils differ from
those in Australia in their major component (Ramanoelina et al. , 1994).
The inflorescence is a terminal spike in the upper axils, one to three
together but sometimes singly, measuring 4.0-9.0 X 2.5-3.5 cm, and
densely flowered in the typical bottle-brush shape (Fig. 8.13). The flowers
are fragrant and usually white or creamy-white. The species is open-
pollinated and remarks referring to M . alternifolia apply. The main flowering
in Australia is April-July. Nectar and pollen production are profuse; honey
is light to dark amber, strongly flavoured with the distinctive tea-tree taste
and candies readily. The fruit is globoid, approximately 5 mm diameter,
rather woody and may persist on twigs for 12 months. The seeds are very
small, numerous, narrow, brown, and viability is generally low.
Ecology
Melaleuca quinquenervia is confined to the east coast of Australia between 11
and 34"S, the southern coastal areas of Papua New Guinea between 7 and
10"s; and between 22 and 23"s on New Caledonia; from sea-level to 100 m
in Australia, to 300 m in New Caledonia and Papua New Guinea. Trees
flourish in areas with a minimum average rainfall of 1000 mm, and a
temperature of 15525°C. It favours swampy sites surrounded by open forest,
but the species is resilient and will colonize relatively dry and otherwise
unfavourable areas. In Australia, it forms almost pure stands in suitable
localities, but elsewhere usually coexists with other Melaleuca spp. including
M . viridzj?ora and Eucalyptus spp. On New Caledonia it is often the locally
dominant species especially in the northern Gomen area. Oil derived from
M. quinquenervia trees on New Caledonia is the cineole-rich type, but oils
derived from the same species in Australia vary with the region in which trees
grow. Many analyses in the literature, however, refer to M . viridiflora and
thus the extent ofvariation inM. quinquenerviahas still to be determined. Oil
from trees growing in Egypt gave a cineole content of 57.2% (Aboutabl et al. ,
1991).
Cultivation
Oil is obtained wholly from wild trees and these are more than sufficient to
meet present demand. Should plantations or a more efficient method of
producing the oil become profitable, the system outlined for M . alternifolia
will be suitable. Foliage is obtained by slashing branches and stripping leaves
which are transported to stills. Modern plants in New Caledonia produce
Myrtaceae 317
good quality oil with a higher recovery rate than the normal 1.0% from
traditional stills.
Fig. 8.13. Melaleuca quinquenenlia, flowering shoot (Herbarium, Royal Botanic Gardens. Melbourne.)
318 Chapter 8
sufficient to give niaouli oil the specific taste and flavour preferred in France.
Its main use is as a direct substitute for eucalyptus and cajuput oil in medical
and pharmaceutical products and, should these three oils become com-
petitive in France, niaouli oil would probably be replaced by the cheaper
eucalyptus oil; eucalyptus oil is also the most common adulterant. Oil was
produced in Australia up to the mid-1970s but could not compete with
similar eucalyptus oils and production ceased. Australian niaouli oil is
almost identical to New Caledonian oil and a direct substitute.
Melaleuca bracteata
Melaleuca linariifolia
Melaleuca linariijolia (Sm.) is confined to the east coast of Australia mainly
south of M . alternifolia. The oil content of fresh foliage is 1.5-4.0 from
individual trees. Oil constituents vary substantially between individual trees
and also regionally. The major constituents are (range in per cent);
lJ8-cineole (4-8), alpha-terpinene (8-1 2), gamma-terpinene (18-25) and
terpinen-4-01 (32-42).
Melaleuca dissitiflora
Melaleuca dissitijlora F. Muell. is a semi-desert species of inland areas. The
oil content of dry foliage is 1.4-4.2% from individual trees. Similarly to other
species, oil constituents vary. The major constituents are (range in per cent):
alpha-terpinene (6-lO), gamma-terpinene (13-20), 1.8-cineole (3-1 5 ) , and
terpinen-4-01 (30-52) (Southwell and Stiff, 1990; Kerry et al., 1990; Wil-
liams and Lusunzi, 1994).
Leptospermum
The previously ill-defined Leptospermum Forst. genus has been divided into
two groups and several subgroups (Thompson, 1989; Bean, 1992). Eighty
species are currently included, most native to Australia, with two occurring
in South East Asia, one in Papua New Guinea and one in New Zealand.
Thompson’s group 1 consists of 38 species, group 2 of 41 species, including
the L. petersonii subgroup of 6 species. Interspecific hybrids occur naturally,
but many more are produced by horticulturalists to provide a range of
popular garden shrubs. Few of the 80 species have been cultivated for their
essential oil, although most are aromatic. The chromosome number for the
genus is stated as 2n = 22, 2n = 44.
DBH. If damaged or cut, multiple stems arise from the stump. The tree is
widely planted as an ornamental or kept clipped as a hedge. The bark is light
brown, furrowed and fibrous on the lower stem, smooth on the higher stem
and branches. The foliage is relatively dense, especially on harvested trees or
those damaged by fire and regenerated. The leaves are petiolate, smooth,
dark to medium green, with one or two sometimes prominent veins, narrow
elliptical 3-4 X 1.5-3 mm, with obtuse or acuminate tips, and flat or slightly
recurved margins. Crushed leaves have a strong lemon scent, hence the
popular name. The oil yield from steam-distilled fresh leaves and twiglets is
usually 1-1.5%, but from individual trees up to 2.5% The main constituents
are, in per cent: citral 40-50, citronellal to 35, plus eugenol, isopulegol and
hydrocarbons. The main characteristics of Type oil are: specific gravity
(15OC) 0.872-0.8856; refractive index (21OC) 1.4688-1.4757; optical rota-
tion + 3'30' to + 5'; solubility (70% alcohol) 1:l to 1:5; total aldehydes
(citral+citronellal) 75-80%. There are chemovars of L. petersonii producing
oils with different characteristics and composition to the Type oil; one has
little or no citral or citronellal but various terpenes including gamma-
terpinene; a second has to 20% citral plus geraniol.
The flowers are borne on axillary or terminal racemes and are solitary,
white to creamy-white, on short lateral branches. Main flowering in Aus-
tralia is December-January. The fruit is globoid, 6-7 mm in diameter, and
the capsule is domed. The seed is linear-cuneiform to 2 mm, and viability is
usually below 50%.
Ecology
The species is irregularly distributed in north-eastern Australia with re-
stricted local occurrence, but a very wide distribution as a popular garden
plant. This indicates adaptability to a range of climatic conditions not
normally found in its natural habitat and thus the potential of the species for
commercial exploitation. Rainfall in its natural range is 800-1500 mm, with
dry periods of 3-6 months; temperature to 35°C in the north to -5°C in the
south. The species is considered both drought and frost resistant and there
may thus be local strains adapted to a region. In Kenya it grew well in the
highlands at 1000 m with an average rainfall of 1000-1200 mm and occa-
sional frosts. In Guatemala the author saw healthy, well grown trees in
somewhat neglected plantations on the coastal plateau and inland with a
rainfall of 1000-1500 mm.
which about 100 g of 10:20:5 (NPK) fertilizer had been placed; no other
fertilizer was applied. Plants grew well and were cut annually.
Trial plantings established by the author in western Kenya received
approximately 90 g triple-superphosphate per planting hole, plus 200 kgiha
ammonium sulphate top-dressing applied annually immediately the rainy
season commenced (Weiss, 1963-1 966). In Australia, ammonium nitrate
had no effect on herbage yield or oil content up to 40 kg Niha. There was,
however, a major change in oil constituents with increasing nitrogen, as
geraniol and citronellal content peaked at 40 kg Niha; noted as the first
record of such an effect on the essential oil of an Australian native plant
(Diatloff, 1990).
Cultivation
Trials in Australia showed L. petersonii can be grown as a commercial crop,
but no plantations exist. In Kenya and Guatemala, plantations were success-
ful and oil produced and sold in commercial quantities. Seedlings are
normally grown from seed obtained from wild trees, but as there are
chemovars of L. petersonii careful selection of parent trees is necessary.
Vegetative propagation of suitable parent stock is essential to raise leaf oil
levels and increase citral and citronellal content if plantations are to be a
viable proposition. The original Kenya planrations were established from
Australian-supplied seed, sown in boxes, transferred to nursery beds at
12-15 cm, then planted into prepared holes, spaced at 124 X 124 cm; all
cultivation including weeding and harvesting was manual. The first harvest
was 3 years after planting out then annually, the cutting period being
determined mainly by work on other more important essential oil crops and
still capacity. This system worked extremely well, was cost-effective, and
produced good quality oil, but the plantation did not survive the transition
from European to Kenyan ownership at independence. Small plantations in
the Kivu region of Zaire were worked on a similar basis and suffered a similar
fate.
A pilot plantation established by the author in western Kenya was planned
to be fully mechanized following planting out of seedlings. Spacing was 120
X 40 cm between plants to produce a hedge to be cut by forage harvester.
The plants grew strongly and the first cut was at 2.5 years, coppicing was fast
and regrowth vigorous with a subsequent cut at 10 months. No further work
was possible as the estate was take over by the government for smallholdings.
Oil yield over the two cuts averaged 400 kgiha (approximately 1.5%), and
analyses showed oil was similar to Australian Type oil with 46% citral and
34% citronellal. On this basis such a plantation would be profitable. Stills for
eucalyptus oils are suitable and steam distilling preferred.
322 Chapter 8
Pimenta
The genus Pimenta Lindl. contains 18 species of aromatic trees or shrubs
native to tropical America and the Caribbean. The genus name is derived
from the Spanish pimenta (peppercorns), and was originally applied to P.
dioica because its berries resembled the then more familiar spice. The most
important commercially are P. dioica L. Merr. providing the spice pimento
(allspice), and P. racemosa (Mill.) Moore the source of oil of bay. The genus
is closely related to Myrtus L. and Eugenia L., differing mainly in the flowers
and fruits, hence the confusion in early identification. The history and
cultivation of P. dioica to produce the spice has been described in detail
(Purseglove et al., 198 1). Production of pimento berry oil, pimento leaf oil
and their derivatives, and composition and characteristics of the oils will be
discussed in this section. Oil of bay should not be confused with bay leaf, a
Myrtaceae 323
spice made from leaves of Laurus nobilis L., or with oil obtained from the
Californian bay Umbelhlaria californica Nutt.
Pimenta racemosa
This species is commonly known as bay or bay rum tree and the latter is to
be preferred, since the former may be confused with the true bay. In the
Caribbean, it is known as ausu, malagueta, wild cinnamon and bay-berry,
pimenta de Tabasco (Cuba), bois d’lndie (Haiti and other French-speaking
islands), ozua or canelillo (Dominican Republic), bay boom (Surinam).
Malagueta is incorrect and more accurately applied to the unrelated Aframo-
mum melegueta (Rose.) Schum., Zingiberaceae, source of melegueta pepper,
native to West Africa and introduced by the Portuguese to their South
American colonies.
Botany
Pimenta racemosa (syn. P. acris (SW) Kostel.; Myrcia acrzs; Myrtus acris;
Amomis cayophyllata (Jacq.) Krug & Urban). A number of local P.racemosa
varieties or races have been described, most of doubtful authenticity. One
locally known as ausu and named P. racemosa var. grisea (Kiaersk) Fosberg
has dense, finely grey or white hairs on undersides of leaves, young twigs,
flower stems and fruits (Little and Wadsworth, 1964). The leaves yield oil
high in geraniol, methyl eugenol and isoeugenol (Tucker et al., 1991).
Another designated P.racemosa var. citriodora (syn. P. acris var. citriodora)
coexists in certain areas; the leaves yield a strongly lemon scented oil
containing 33% neral and 54% geranial. A believed chemovar of P.racemosa
produces an anise-scented oil containing 32% methyl chavicol, 43% methyl
eugenol but only small amounts of eugenol and chavicol. Indiscriminate
harvesting and distillation of either with true bay rum produces inferior
quality oil.
Bay rum is a small to medium-sized forest tree to 15 m occasionally 20-25
m, with a single trunk to 25 cm DBH, often slightly angled or grooved. The
bark is greyish, sometimes scaly, and is often shed giving the trunk a mottled
appearance. The inner bark is pinkish, with a slightly bitter astringent taste.
The sapwood is light brown, the heartwood mottled brownish red or
blackish, very hard, heavy, strong, tough, durable and termite resistant.
Timber from mature trees is used for general carpentry and posts, and is an
excellent fuel. Branching is normally extensive on uncut trees, usually fairly
high on the stem and uncut trees have a dense columnar crown. The twigs
are initially green becoming brown with age, angled when young, with gland
dots, usually smooth but sometimes finely grey-hairy. The trees are ever-
green and leaves are shed at 1.5-2.5 years. The leaves are simple, opposite,
entire, elliptic to obovate with a tapering base; large, 5-13 X 3-5 cm, midrib
324 Chapter 8
sunken and margins rolled under. The leaves are leathery, glandular on
lower surface, highly aromatic when crushed, and a dark shiny green,
sometimes finely hairy beneath; the petioles are short and reddish. Although
all tree parts are generally aromatic, an oil is commercially extracted only
from fresh or dried leaves and twiglets.
The oil content of mature leaves is generally similar or lower than that of
juvenile, but the phenol content is higher. The oil content of fresh mature
leaves is 1-3%, but can be 5% from specific trees; the main constituents are
phenols to 70%, mainly eugenol and chavicol. The major characteristics and
constituents are shown in Table 8.21. Leaf oil from four chemovars growing
in the Dominican Republic differed substantially from P. racemosa and each
other (Tucker et al., 199 1). Since there is a more than ample leaf supply to
satisfy current oil demand, little selection and propagation of high-yielding
trees has occurred. Decoctions of leaves are widely used by local herbalists to
alleviate respiratory infections, and mixed with shark oil to treat pleurisy.
Inflorescences are terminal on branchlets, the small fragrant white to
greenish-white flowers on much-branched cymes or corymbs, 2.5-7.5 cm.
The light green hypanthium is usually less than 3.5 mm and extends beyond
the ovary. The flowers have five broad, spreading sepals, less than 1.5 mm,
and five white, rounded petals, widely spreading to 4 mm. The flowers differ
from those of P. dioica, which have four sepals and petals. The flowers
contain numerous white stamens to 1.5 mm with small anthers, a pistil
composed of an inferior two-celled ovary and a slender, slightly bent white
style to 5 mm. Flowering is April-August within the Caribbean. The
flowering shoots and flowers yield the highest quality oil, but shoots are
normally harvested together with leaves (Fig. 8.14). The fruits are small
fleshy ovoids, 0.5-1.5 cm, 90 mm diameter, black when ripe drying to a dark
reddish-brown and containing one to three 4-7 mm diameter dark brown
seeds, viability extremely variable. The fruits are attractive to birds which are
the major seed dispersal agent, and bay rum trees are thus widely naturalized
in the Caribbean.
A B
Specific gravity (25°C) 0.95-0.99 0.957-0.988
Refractive index (20°C) 1.51-1.52 1.5075-1.5170
Optical rotation (20°C) 0" to - 3 O -4"4' to -2'56'
Solubility in 70% alcohola 1 :I-2 1 :I
Phenol content (%) (mainly eugenol and 55-65 (max) 55-65
chavicol)
a Fresh oil only (solubility declines rapidly with age); A, EOA specification; B, Commercial samples
Myrtaceae 325
Ecology
The bay rum tree is native to the Caribbean and north-eastern South
America and has been successfully introduced into south-eastern USA, Sri
Lanka, West and East Africa and Indonesia. It is frequently planted in
botanic gardens and arboreta, indicating a much wider potential range, but
is grown commercially for oil principally in Dominica and Puerto Rico.
Trees flourish at 15-35"C, below 15°C having a greater adverse effect than
above 35OC. Clear sunny days promote foliar growth, reduce the incidence
of leaf disease, and leaf-oil content is frequently higher during warm,
relatively dry periods than the converse. The tree is frost sensitive. An annual
rainfall of 2500 mm well distributed throughout the year is the optimum,
and few months should receive less than 200 mm. Trees will grow satisfacto-
rily on 1250-1 500 mm but regrowth following cutting is slow and so extends
the period between harvests that commercial plantations are unprofitable.
An annual rainfall of 750 mm is sufficient where there are large numbers of
wild trees to provide the foliage necessary. Trees flourish from sea-level to
750 m, with the upper limit set by the minimum temperature. Leaf-oil
content may decrease with altitude, but this has not been confirmed by
accurate sampling.
Cultivation
Most plantations of bay rum trees (groves) are established on land unsuit-
able for other crops with only the minimum of clearing; usually consisting of
cutting and burning scrub, and moving large stones to field boundaries.
Seedlings are then planted into prepared holes, and subsequent cultivation
limited to slashing or mowing weeds. Seedlings can be raised in nursery beds
or pots and techniques used to produce eucalyptus seedlings are generally
suitable. Where there are large areas ofwild trees, the most common method
is to collect self-sown seedlings of the required age, and plant them im-
mediately into prepared holes. T o establish a nursery, seed should be
collected when fruits are mature, usually August-October in the Caribbean.
Seeds must be removed from fruits, washed and planted within 48 hours
since viability is frequently low and germination erratic (between 2 and 6
weeks).
Selection of high-yielding trees as parent stock is little practised, but
methods of vegetative propagation successful with pimento will probably be
equally effective on bay trees. Budding and approach-grafting are well-
known techniques among orchardists, while cross-budding between P.
dioica and P. racemosa is also usually successful. The low world price for bay
leaf oil has generally precluded such relatively expensive operations, but if
groves are to be profitable on an increasingly competitive world market,
upgrading trees is essential. That this is possible is shown by the high leaf-oil
content of 4% obtained from individual trees compared with the common
Myrtaceae 327
1.5-2.5%, and average oil yield from local distillers of below 2%. Micro-
propagation could provide large numbers of improved plants should it
become profitable to use this and similar techniques.
Seedlings are ready for transplanting in 18-24 months and before planting
the long tap root should be pruned to encourage lateral rooting, and
seedlings topped to 15 cm. Seedlings should be planted following rain when
soil is wet below root depth, or watered in and shaded from direct sunlight
until established. Spacing is determined mainly by topography, but on level
lands 2-5 m between trees in 5-10 m rows allows adequate space for
development and inter-row cultivations. Topped seedlings frequently pro-
duce several leading shoots and some growers leave two to produce trees
with two main stems, since experience has shown total foliage yield is higher
than from single-stemmed trees. Trees are topped at 2-3 years or 3-5 m and
maintained at this height. Misshapen, aged or diseased trees are usually cut
back almost to ground level and a new shoot allowed to grow from the
stump. Weed control normally consists of mowing grass and other herbage,
but small cleared areas may also be maintained around trees to reduce the
incidence of insect and fungal attack on stems, a technique based on the
experience of pimento growers. Herbicides are seldom used but, should they
become a viable alternative, those compounds used in pimento will probably
be suitable.
Harvesting
The harvesting interval can vary from 1-2 years and is not necessarily
dependent on regrowth; a too-long interval between cuts, 2 years or more,
can result in low foliage yield since individual leaves begin to fall after 24-36
months. Experienced producers believe that leaving mature leaves on trees
for an extended period suppresses emergence of new leaves; thus well-
managed groves are cut annually, or at least three times in 4 years. Leaves
can be harvested year-round, but usually between January and June. In
regions where there is a dry period this is preferred, since more oil is obtained
per still charge. Whether this is due to a generally higher leaf oil content or a
greater proportion of more mature leaves which have a higher oil content is
immaterial to producers, since the end result is similar. What is more
important is to ensure harvesting is coordinated with still capacity in order to
maintain regular operation. The extent and timing of harvesting, and
whether this should be seasonal or year-round, must be related to a dis-
tillery’s capacity to store and process leaves.
Leaf yield from individual trees varies greatly, with low yield and rate
of regrowth genetically determined, since vegetatively produced seedlings
exhibit parental characteristics. Yield from established groves is 9000-
35,000 kg/ha with annual cutting, a wide variation due primarily to location,
but also to the system of management; the most efficient growers obtain the
highest local average yield of foliage and oil. Where cheap or family labour is
328 Chapter 8
available, leaves and twiglets only will be plucked from trees, packed in sacks
and transported to the nearest distillery. Alternatively, leafy branches are
cut, bundled, and transported to the distillery where leaves are removed.
Harvesting wild trees is often by contract with the local distillery, and
workers engaged to deliver a regular flow of leaves. Whatever method of
harvesting is employed, cheap labour is essential and, should this situation
change, bay oil production usually becomes unprofitable.
Although groves can be laid out to suit fully mechanized operations as for
eucalyptus, this is currently too expensive to be profitable, but could become
so if groves of high yielding trees are established. The life of a grove is
indeterminate, since trees are regenerated from stumps or replaced when
they cease to bear sufficient foliage. Individual trees of 50 years of age are
known, but the effect of regular cutting and good grove management on tree
life is unknown. Since long-term high average annual oil yield per hectare is
important, the best method of harvesting to ensure this must also be
determined.
Distillation
Harvested leaves can be processed immediately, or stored for up to a week in
suitable conditions with no oil loss. The factors affecting yield and quality of
pimento leaf oil are generally accepted as applicable to bay in absence of
specific data. Leaves are normally finely chopped or comminuted, and
distillation may be by water, steam or a combination of both plus cohoba-
tion. The operation differs from eucalyptus distillation in that two types of
distillate are produced, one lighter and one heavier than water, later com-
bined to produce crude oil of acceptable phenol content. A 1:3 mixture of
sea and fresh water was used in Puerto Rico stills and produced more oil with
a higher phenol content from the same amount of leaves than fresh water
alone; a 25% sodium chloride solution was an effective substitute. Although
most stills were initially locally made of any available material, many have
been replaced by stainless steel which produces a finer, lighter coloured oil
from a distillation time using high-pressure steam of 4-6 hours. Bulk oil
from smaller producers is usually sold to the owner of a larger still, who
filters, cleans and often adjusts the phenol content to an accepted stand-
ard.
Compound Percentage
Myrcene 31.6
Limonene 1.4
1,8-Cineole 2.0
Trans-ocimene 2.1
3-Octanone 1.o
p-Cymene 0.5
3-Octanol 0.5
1-0cten-3-ol 1.3
Linalool 3.0
Terpinen-4-01 0.3
Caryophyllene 0.9
Methylchavicol 0.3
Neral 0.5
Alpha-terpineol 0.8
Geranyl acetate 0.8
Eugenol 38.6
Chavicol 11.0
Source: Lawrence (1977).
terpened oils, or by alcohol washing crude oil, and lack terpenes and
sesquiterpenes. Oleoresin has been produced in small quantities for use in
fragrances and the flavour industry.
Pimenta dioica
Pimenta dioica, commonly pimento, is cultivated in Jamaica, Mexico and
Central America, primarily to produce dried berries known as pimento or
allspice.
Botany
Pimenta dioica (L.) Merr. (syn. Myrtus pimenta L.; M . dioica L.; Eugenia
pimenta DC.; Pimenta oficinalis Lindl.) is an evergreen tree to 7-10 m
(sometimes to 15 m) with a slender trunk, branching from 1 m, secondary
branches profusely branched at extremities. The bark is smooth, shiny and
silvery pale brown and is often shed in long strips. The wood is pinkish,
close-grained, heavy, strong and durable. The leaves are simple, opposite,
entire, thinly coriaceous, punctate, with pellucid glands, borne in clusters at
end of branches and highly aromatic when crushed. The flowers are small,
white and fragrant, on many-stalked cymes in the axils of upper leaves.
The fruit is sub-globose, 4-6 mm diameter, and contains two seeds. Ripe
fruit is a deep purple to glossy black, with a mucilaginous mesocarp and is
Myrtaceae 331
sweet and spicy. On drying, the fruit is dark brown, rough, and more
aromatic. The dried berry contains an aromatic steam-volatile oil, fixed oil,
resin, protein, pentosans, starch, traces of alkaloids, pigments and minerals;
the proportions vary considerably and are most pronounced in the volatile
oil. The aroma and flavour are due to steam-volatile compounds, pungency
to non-steam-volatile constituents.
The major use of ground pimento is as a flavouring and curing agent in
processed meats and bakery products, but for domestic culinary use it is
often supplied mixed with other ground spices. Information on the chem-
istry of pimento is almost wholly from Jamaican spice and leaf oil, but there
is considerable variation in the chemical composition of both from different
areas of the Caribbean and Central America (Green and Espinosa, 1989).
The major constituent of pimento spice and leaf oil is eugenol, and thus it is
similar in general character to clove bud, clove leaf, tropical bay leaf and
cinnamon leaf oils. Although eugenol is the major component of distilled
berry and leaf oils, the two oils are organoleptically quite distinct. A detailed
description of the methods used to grow and harvest P. dioca has been
published (Purseglove et al., 1981).
Distillation
Leaves may be fresh, withered or dried and stored for 2-3 months prior to
distilling with little effect on oil yield or phenol content. Yields from dried
and fresh leaves are 0.5-3.0 and 0.3-1.25% respectively. Stills take a charge
of 1000-2200 kg, have perforated false bottoms upon which leaves are
packed to prevent channelling, and distilling is by low-pressure steam for
4-5 hours. At some distilleries the milky distillation waters are run into large
concrete tanks, the oil settling out is normal leaf oil but a dark colour, which
is collected and added to the other fractions. The combined oil is filtered and
stored for export, preferably in stainless steel drums.
Berry oil is obtained by crushing berries immediately prior to loading, and
distilling by direct steam for about 10 hours usually with cohobation; total
oil yield is 3.3-4.5%. Central American berries have a volatile-oil content of
1.5-3.0%, but commercial samples of Belize berries examined in London
yielded 3.4-4.4% (average 3.5%). Oil from Caribbean and Central Amer-
ican berries varies considerably in organoleptic properties, but there is little
reliable information on differences in composition (Green and Espinosa,
1989). Local variation is well known in Jamaica, with methyl eugenol
content from zero to greater than eugenol, and a number of chemovars
apparently exist. This is supported by oil distilled from pimento collected in
Belize, whose methyl eugenol content ranged from zero to 10%. Informa-
tion on oil composition of wild pimento from other sources in Central
America is limited, since this material is seldom distilled. Oil from Jamaican
undried green berries is similar in composition to that from dried berries, but
has a higher monoterpene content.
332 Chapter 8
Berry Leaf
BSS EOA BSS EOA
Specific 1.025 1.018 1.037 1.018
gravity 1.045 1.048 1.050 1.048
(20°C) (25°C) (20°C) (25%)
Refractive 1.526 1.527 1.531 1,532
index 20° 1.536 1.540 1.536 1.536
Optical rotation O0 0" ng t 0'30'
-5" -4" -2O
Solubility (viv 70% 1:2 1 :2 1 :2 1 :2
ethanol)
Total phenols is', vol) 65 80 80-90
BSS, British Standard Specification; EOA, Essential Oil Association, USA; ng, not given.
Figures in columns are range.
content. The main characteristics ofthe oil are shown in Table 8.23. Leaf oil
is distilled almost solely in Jamaica, and is widely used in meat products and
confectionery. It can partially replace pimento berry oil, which is generally
considered a more suitable oil for flavouring, but is much more expensive.
Myrtle
Spanish Moroccan
Specific gravity (15%) 0.911-0.930 0.900-0.912
Refractive index (20°C) 1.466-1.470 1.466-1.467
Optical rotation t2I0to $26" t23" to t24'40'
Solubility (viv 80% alcohol) 1 :2 1:>7
Acid number max 2.7 ng
Ester number 63-92 ng
Ester number (after acetylation) 80-1 17 ng
ng, not given.
Source: Mazza (1983).
334 Chapter 8
Alpha-pinene 77.1
Delta-3-carene 3.9
Limonene 18.4
Alpha-terpinene 3.6
p-Cymene 5.7
Linalool 6.0
Terpinen-Col 2.2
Alpha-terpineol 8.3
Methyl hexadecanoate 9.1
Linalyl acetate 1.9
Alpha-terpinyl acetate 2.7
Nery acetate 7.8
Methyl eugenol 2.3
1 &Cineole 69.3
a mgikg berries Source: Mazza (1983).
References
Abbott, P.S. (1977) The Eucalyptus Industry. In Proc. AFDI Con5 pp. 36-43,
Taralgon, Victoria, Australia.
Aboutabl, E.A. et al. (1991) A comparative study of the essential oil from three
Melaleuca species growing in Egypt. Flav. Frag. J. 6(2), 139-141.
Adamson, A.D. and Robbins, S.R.J. (1975) The Market f o r Cloves and Clove Products
in the United Kingdom. Tropical Products Institute, London.
Akgul, A. and Bayrak, A. (1989) The essential oil content and composition of myrtle
( M . communis) Isancs. Doga T. T.ve 0.Dergisi. 13(2), 143-147.
Ammon, D.G. et al. (1985) Rapid and accurate determination of terpenes in the
leaves of Eucalyptus species. Analyst 110, 921-924.
Anon. (1984) Note sur le Secteur des Huiles Essentielles. Centre National D'Etudes
Industrielles. Min. Econ., Tunis, Tunisia.
Antonov, A.A. and Zhdanov, A.Y. (1986) Essential oil of E. macarthun'i. Khim.
Prirod. Soedinenii 4, 5 15-5 16.
Araujo, A.C., Sacramento, C.K. and Silva, P.H. (1989) Economic evaluation of
chemical harvesting of cloves in south-east Bahia. Bol. Tech No 164-CEPLAC.
Brazil.
Arctander, S. (1960) Perfume and Flavour Materials of Natural Origin. Mrs G.
Arctander, 6665 Valley View Blvd, Las Vegas, Nevada, USA.
Ashurst, P.R. et al. (1972) The development of oils of pimento. J . Sci. Res. Council,
Jamaica. 3(1), 50; also Proc. Confer. on Spices. (1973). Tropical Products
Institute, London, UK, pp. 209-214.
AUSTEAM Ltd. (1995) Tea Tree Oil into the 21st Century. AUSTEAM Ltd,
Lismore, NSW, Australia.
Australian Essential Oil Co. Ltd (1936) Medical and Dental Data of Ti-trol and
Melasol. Australian Essential Oil Company Ltd. (rev. edn), Sydney, Australia.
Australian Standard (1985) Essential Oils - Oil of Melaleuca, terpinen-4-01 type. AS
2782-1985. Stands. Assn. Aust., Sydney, Australia.
Babu, Kn. et al. (1993) Biotechnology in spices. Ind. Hort. 38(3), 46-50.
336 Chapter 8
Baez, C.M. et al. (1992) Mineral nutrition of E. globulus ssp. globulus in relation to
yield and quality of cineol. Agricult. Techn. (Chile) 52(4), 475-479.
Baker, R.J. and Smith, H.G. (1902) A Research on the Eucalypts especially in regard to
their Essential Oils. Tech. Educ. Series 13, Tech. Museum, NSW, Sydney,
Australia.
Balciunas, J.K. and Burrows, D.W. (1993) T h e rapid suppression of the growth of
M . quinquenervia saplings in Australia by insects. 3. Aquat. Plt Management 31,
2 6 5-2 7 0.
Baslas, R.K. and Saxena, S . (1984) Chemical examination of fruits of E. globulus.
Herba Hung. 23(3), 21-3.
Bassett, I.B., Fernnowitz, D.L. and Barnetson, R. (1990) A comparative study of
tea-tree oil versus benzoyl peroxide in the treatment of acne. Med. 3. Australia
153,455-456.
Bean, A.R. (1992) T h e genus Leptospermuni in northern Australia. Austrobaileya 3,
634-65 9.
Blake, S.T. (1968) A revision of M . leucadendron and its allies. Contrib. No 1.
Queensland Herbarium, Brisbane, Australia.
Boelens, M.H. (1985) Essential oils and aroma chemicals from E. globulus. Perj%m.
Flav. 9(12), 1-14.
Boland, D.J. et al. (1979) Eucalyptus Seed. CSIRO, Canberra, Australia.
Boland, D.J. et al. (1984) Forest trees of Australia. Nelson Ltd, Melbourne, Aus-
tralia.
Boland, D.J., Brophy, J.J. and House, A.P.N. (1991) Eucalyptus leaf oils. ACZARi
CSZRO. Inkata Press, Melbourne, Australia.
Bosisto, J. (1862) Abstract of a paper on the yield and uses ofvolatile oils from nature
and imported plants in the colony of Victoria. Proc. Roy. Soc. Vic. 6, 52-6 1.
Bosisto, J. (1 878) M r Bosisto's essay on the Eucalyptus and its properties in Victoria.
Expositions Universelle de Paris 1878. Victorian State Library, Melbourne, Aus-
tralia.
Brooker, M.I.H. and Kleinig, D.A. (1983) A Field Guide to Eucalypts, Vol. 1 (Vol. 2
and rev. edn Vol. 1 pub. 1990; Vol 3, 1994). Inkata Press, Melbourne,
Australia.
Brophy, J.J. and Lassak, E.V. (1986) T h e volatile leaf oils of some central Australian
species of eucalypts. Proc. Royal Soc. NSW 119, 103-108.
Brophy, J.J. and Lassak, E.V. (1988) M . leucadendra leaf oil: two phenylpropanoid
chemotypes. Flav. Frag. 3. 3, 43-46.
Brophy, J.J., Lassak, E.V. and Boland, D.J. (1987) Volatile leaf oils of the two
subspecies of M . acaciodes. 3. Proc. R. Soc. A'SW. 120, 135-139.
Brophy, J.J., Lassak, E.V. and Boland, D.J. (1988) Volatile leaf oils of six northern
Australian broad-leaved melaleucas. J. Proc. R. Soc. NSW. 121, 29-33.
Burrows, D.W., Balciunas, J.K. and Edwards, E.D. (1994) Herbivorous insects
associated with M . quinqeunervia and its allies. Austral. Entomol. 21 (4),
137-142.
Butcher, P.A., Doran, J.C. and Slee, M.U. (1994) Intraspecific variation in leaf oils
of M . alternifolia. Biochem. System. & Ecol. 22(4), 419-430.
Butcher, P.A., Byrne, M.and Moran, G.F. (1995) Variation within and among the
chlorogiast genomes of M . alternaria and M . linarifolia. Plant System. Evol.
194( 1/2), 69-8 1.
Myrtaceae 337
Gupta, K.C. et al. (1990) Chemical constituents of essential oils of E. sideroxylon and
E. nitens. Ind. Perfum. 34(1), 1-4.
Hall, N. (1987) Botanists ofthe Eucalypts. CSIRO, Canberra, Australia.
Hasanah, K. et al. (1 984) Effects of growth inhibitors and storage temperature on
viability of clove seed. P.P. T. Industrz, Bogor 8(49), 47-54.
Heath, H.B. (1981) Source Book of Flavours. The AV1 Publishing Co., Westport,
Connecticut, USA.
Hills, W.E. (199 1) Unique qualities of eucalypt wood. In: Proc. Int. Con$ Bleached
Kraft Pulp Mills. CSIRO, Clayton, Victoria, Australia.
Iwamuro, H. et al. (1983) High-boiling constituents of clove oil. J. Agric. Biol. Chem.
47,2099-2 100.
Jahns, E. (1884) Eucalyptole. Ber. dtsch. Chem. Ges. 17, 2941-2944.
Jarvie, J.K. et al. (1986) Screening for compatibility between clove and related
myrtaceous species. P.P. T. Industri, Bogor. 11(3/4), 56-59.
Johns, M.R. and Rudolph, V. (199 1) Tea tree oil processing. Fats & Oils Internat. 5 1,
23-24.
Kannan, K. and Nair, K.P.V. (1965) Influence of interval of pruning on leaf and oil
yield of E. citriodora. Ind. Oil Soapy. 30(12), 365-367.
Kerry, A.M. Williams, L.R. and Home, V.N. (1990) Composition of oils from M.
alternijolia, M . linariijolia and M . dissitijlora. In: Proc. Con$ Modern Phytotherapy.
Macquarie University, Sydney, Australia.
Lassak, E.V. (1988) The Australian eucalyptus oil industry, past and present. Chem.
in Aust. 55, 396-406.
Lassak, E.V. and McCarthy, T. (1983) Australian Medicinal Plants. Methuen,
Melbourne, Australia.
Lawrence, B.M. (1977) Chemical evaluation of various bay oils. In: Proc. 7th Inter.
Con5 Essen. Oils pp. 172-179. Kyoto, Japan.
Lawrence, B.M. (1993) Myrtle oil. Perfum. Flav. 18(2), 52-55.
Lawrence, B.M. (1994) Progress in essential oil research. Perfum. Flav. 19(6),
60-62.
Lawrence, B.M. (1995) Progress in essential oils. Perfum. Flav. 19(5), 92, 94.
Little, E.L. and Wadsworth, F.H. (1964) Common Trees of Puerto Rico and the Virgin
Islands. Agric. Handbook No. 249. USDA, Washington DC, USA.
Madan, C.L. et al. (197 1) Introduction of E. macarthurii in India: Cultural practices
and chemical variation. Flav. Indust. 4, 246-248.
IMaistre, J. (1955) Le giroflier a Madagascar et Zanzibar. Agron. Trop. 10,
4 13-4 18.
Mazza, G. (1983) Gas chromatographic-mass spectrometric investigation of the
volatile components of myrtle berries. J. Chromatog. 264, 304-3 1 1.
Meijer, T.M. and Schmid, H. (1948) Uber die Konstitution des Eugenins. Helv.
Chem Acta 31, 1603-1607.
Miranda, M. and Zayas, J.P. (1985) Influences de la epoca de recoleccion sobre el
rendimiento de aceite escencial y la composition quimica del E. citriodora.
Rev.Cuba Farm. 19, 121-127.
Morikawa, H. et.al. (1995) Eucalyptus absorbs nitrogen dioxide. Faculty of Science,
Hiroshima University, Japan.
Motl, O., Hodacava, J. and Ubikk, V. (1990) Composition of Vietnamese cajuput
oil. Flav. Frag. J. 5(1), 39-42.
Myrtaceae 339
Muchlis, N.A. and Rusli, S. (1978) A review of several problems of essential oil
quality. P.P. T.Zndustri, Bogor 28(1), 51-58.
Murtagh, G.J. and Ethrington, R.J. (1990) Variation in oil concentration and
economic return from tea-tree oil. Aust. J. Exper. Agric. 30(5), 675-679.
Mwangi, J.W., Guantai, A X . and Muriuki, G. (1981) E. citriodora: essential oil
content and chemical varieties in Kenya. E. Afr. Agric. For. J. 46(4), 89-96.
Nabney, J. and Robinson, F.V. (1972) Constituents of pimento berry oil. Flav. Znd.
3, 50-5 1.
Nair, M.K. (1978) Clove and nutmeg. Indian Farming 7, 10-13 and 35.
Nair, R.B., Sadanandar, A.K. and Unnithan, K.R.G. (1974) Floral biology, and
breeding of clove. J. Plant. Crops I. Supp. 28-29.
Nair, E.V.G. et al. (1983) Optimum time of harvest of E. citriodora grown in plains.
Ind. Pe$um. 27(1), 54-55.
Nazeem, P.A., Wahid, P.A. and Nair, P.C.S (1993) Nutrient deficiency symptoms
in clove: visual symptoms. South India Hort. 41(6), 360-365.
Nishimura, H. and Calvin, M. (1979) Essential oil of E. globulus in California. J.
Agric. Food Chew?.27(2), 432-435.
Nizharadze, A.N. and Bagaturiya, N.S. (1974) Age induced changes in cineole
content of eucalyptus oil. Subtrop. Kultuy 3, 67-76.
Nutman, F.J. and Roberts, F.M. (1971) The clove industry and disease of the clove
tree. PANS 17(2), 147-165.
Nyrdjanah, N., Hardjo, S . and Mirdna, M. (1991) Distillation method influences the
yield and quality of clove leaf oil. Zndust. Crops Res. J. 3(2), 18-26.
Parry, J.W. (1962) Spices, their Morphology, Histology and Chemistry. Chemical Pub.
Co. Inc., New York, USA.
Penfold, A.R. (1 925) The essential oils of Melaleuca linariifolia and M . alternifolia. J .
Proc. R. Soc. NSW59, 318-325.
Penfold, A.R. and Morrison, F.R. (1951) Commercial Eucalyptus Dis. Bull. 2,
Museum Appd. Arts & Sci., Sydney, Australia.
Penfold, A.R. and Willis, J.L. (1954) The essential oil industry of Australia. Econ.
Bot. 8(4), 316- 336.
Pinto, D’a.A.J., Souza, C. and Donalisio, M. (1976) Selecting eucalypts for yield and
citral content of essential oil. Bragantia 35, CXV-CXVIII.
Pryor, L.D. (1976) Biology ofEucalypts. Edward Amold, London, UK.
Purseglove, J.W. (ed.) (1981) Spices. Vol. 1, pp. 229-330. Longman, Harlow, UK.
Ramanoelina, P.A.R. et al. (1994) Occurrence of various chemotypes in niaouli ( M .
quinquenervia) essential oils from Madagascar. J. Agric. Food Chem. 42(5),
1177-1 182.
Rao, B.R.R. et al. (1984) A note on influence of location, season and strains of E.
citriodora. Znd. Pe$um. 28(314), 153-155.
Rusli, S., Laksmanahardja, P. and Simarmata, J.P. (1979) Distillation methods
affecting yield and quality of clove leaf oil. P.P.T.Zndustri. Bogor 34(7-9),
1-12.
Schmid, R. (1972) A resolution of the Eugenia-Syzygiunz controversy (Myrtaceae).
Amer. J. Bot. 59, 423-436.
Sheffield, F.M.L. (1950) The clove tree of the Seychelles. East Af.Agric. For. J. 16,
3-8.
340 Chapter 8
Jasmine
The largest number ofJasminum species is centred on the Indian and South
East Asian region, with some 40 native to India and one apparently to South
America. Three are commercially cultivated for their essential oil, J. aur-
iculatum, J. grandiJEorumand J. sambac, but other species for their fragrant
flowers. Many Jasminum species are grown in gardens and there are a
multiplicity of so-called cultivars which are basically selected strains given
names to preserve a separate identity for their developer. Such selection is
relatively simple, since there is considerable variation within species in
respect of type and rate of growth, flower size, colour and scent. Table 9.1
lists well-known jasmine species and unrelated plants often called jasmine.
Jasmine’s fragrant flowers ha<e been used since very ancient times for
personal adornment, in religious ceremonies, strewn at feasts and in bath
water in practically all countries in which the plants occur. The combination
o f a very powerful, rich scent with a white flower has a particular aesthetic
appeal. It is possible to accurately identify the flower in paintings from
ancient Egypt and India, and Romans and Greeks used jasmine pomade.
The Roman Lucius Plotius, a triumvirate member in hiding from his
political enemies, was betrayed by the strong jasmine scent of which he was
inordinately fond. It is doubtful, however, if either the Greeks or Romans
knew how to prepare jasmine pomade and probably imported it via the
Middle East. This relationship was recently revived with the development of
342
Oleaceae 343
Table 9,l. Some Jasminum species and unrelated plants often called jasmine.
a valuable export trade in fresh jasmine flowers by air from Bombay to the
Middle East and Gulf States, which totalled 30,000 kg in 1995.
A large cheap labour force is essential to harvest jasmine flowers and this
is a major reason why Egypt has in the last decade become a significant
supplier of concentrate to world markets. European companies, especially
French, have established joint ventures to locally produce concrete for their
parent companies, and their scientists are breeding higher yielding cultivars
or flowers with novel scents.
Jasmine species were mentioned in ancient Indian literature including the
Ayurveda, lain-i-Akbari and in the Tamil Sangam period, but a variety ofJ.
sambac known as Tuscan jasmine was apparently introduced into India by
the Duke of Tuscany about 169 1. The sea trade between southern India and
the Arabian Peninsula is very ancient, with jasmine plants shipped to the
Persian Gulf and Red Sea ports and their products onward to Europe. The
movement has recently been reversed with European technology flowing to
India to improve local agronomic and processing techniques. A French joint
venture in Coimbatore reportedly processed 500,000 kg flowers in 1995,
and its specialists have bred jasmine varieties whose flowers produce an
absolute with tuberose (Polianthes tuberosa L.), and orange blossom top-
notes.
Jasmine reached China along the Silk Road and the use of its flowers to
flavour tea, mo-li-hua-cha, probably originated there. Such teas now com-
monly use J . paniculatum flowers; however, a jasmine indigenous to China
has recently been recorded (Bai, 1992). The steady decline in jasmine
concrete exports from China is notable, since China is one of the few
countries with a reservoir of low-paid labour so important in jasmine
harvesting. One reason is that domestic demand is increasing as rising
disposable incomes create a market for higher quality cosmetics.
Jasmine oil or essence has long been valued for its medicinal properties; it
is widely believed to stimulate the reproductive system not only as an
344 Chapter 9
Botany
Jasminum species are deciduous or evergreen shrubs or clambering climbers,
with opposite or alternate, pinnate leaflets, sometimes apparently simple
consisting of one leaflet. The flowers are white, pink or yellow in terminal or
axillary cymes, rarely solitary, the calyx mostly with four to nine lobes of
varying length, and the corolla salverform with slender tube and four to nine
lobes. Some species have very fragrant flowers, the principal natural source
of jasmine perfume. Growth regulators can influence flowering, and in
general chlormequat, ethephon and daminozide tend to advance flowering
and increase flower yield, while GA tends to delay flowering, although the
duration of flowering is usually unaffected. However, flower oil content and
thus concrete recovery may be depressed (Bhattacharjee and Divakar, 1989;
Gowda and Gowda, 1990). The fruit is usually a two-lobed black berry. The
ontogeny ofJasminum spp. has been described in detail (Muthuswamy and
Shanmugavelu, 1982). The basic chromosome number is x = 13, 2n = 26,
but triploid forms occur naturally.
Jasminum auriculatum Vahl. (Hindi juhi; Tamil usimalligaz) is a twining
shrub with small mostly simple, occasionally trifoliate leaves, the two lower
leaflets small, reduced to auricles or lacking. The flowers are fragrant, small,
white, in pubescent, compound, multiflowered, lax cymes, approximately
2600 flowers per kg. The flowers have five to eight elliptic corolla lobes. The
Oleaceae 345
carpels are black, solitary and globose. The flowers normally open between
6 and 7 p.m. but up to 2 hours later on hot, dry days. In southern India
where it is widely cultivated, plants flower almost continuously from April to
December, with May and September yielding the greatest proportion of the
total; 21% and 16% respectively. Selected strains have a flower yield
25-50% higher than the local average of 2000 kgiha, twelve in Tamil Nadu
gave concrete yields of 0.28-0.41%, while one selection yielded 8800 kgiha
flowers with a concrete recovery of 0.34% (Sambandamurthi and Kadar,
1982). In this region the highest flower yield is in June-July. Correlations
exist between 100 bud weight and number of petals per flower, style length
and corolla tube length (Indiresha et al., 1986).
Jasminum grandiforum L. is considered to be J. oficinale forma grand-
$orurn (L.) Kobuski, commonly referred to in the literature as J. grand-
iflorum Hort. which designation will be used in this text. It is known as
Catalonian or Royal but more often as Spanish jasmine; in Hindi chameli,
Tamil padder malligai, and in many countries there are locally named
cultivars varying substantially in growth habit, leaf, stem and flower colour.
Jasminum grandzflorum is an evergreen shrub or climber to 5 my but the
ribbed stems require support when cultivated and pruned to reduce height.
The leaves are opposite, dark green, with five to seven ovate leaflets to 2 cm;
the terminal leaflet is longer, and acuminate, the lateral leaflets sessile or
shortly petiolate, the distal pair with a broad connate base confluent with the
terminal. The flowers are very fragrant, up to 4 cm diameter, white often
purple tinged, on lax axillary or terminal cymes longer than the leaves. The
bracts are ovate to spatulate-oblong, and foliaceous (Fig. 9.1). The pedicels
of the central flowers are shorter than those of the lateral flowers; the calyx is
glabrous with teeth linear to 10 mm, and the corolla tube is up to 2 cm with
5-6 oblong lobes. The corolla tube encloses two stamens on short slender
filaments. Selected J. grandiflorum cultivars usually had higher flower yield
than other Jasminum spp. in India (Srivastava, 1991).
Since the number of flowers is often the main criterion in assessing
reaction to factors under test, picking is tedious and lengthy. The minimum
number of plants per treatment to accurately reflect the result required is
thus important; in India, six plants per plot was the optimum (Ramchander
and Bhattacharjee, 1986). Under natural conditions, seed set is generally
very low, and pollen sterility in diploids frequently above 75% and to 95% in
triploids (Chandra, 1982). Flowers normally open between 6 and 7 p.m.
and main flowering in India is May-August with July and August supplying
the major proportion of total yield. In a trial of six selections in Tamil Nadu,
one yielded 10,000 kgiha flowers with a concrete recovery of 0.29%;
concrete recovery ranged from 0.25 to 0.32% over the six.
Jasminum sambac (L.) Ait. Arabian jasmine; in Hindi mogra, Tamil kuda-
malligai, is the most common cultivated species in India and there are many
named cultivars, some yield flowers for oil, others for garlands. Leaves and
346 Chapter 9
roots are used to treat diarrhoea and fevers, and a paste of flowers to treat
conjunctivitis and dermatitis. It is basically an evergreen sub-erect shrub or
vine, with stems rather slender, naturally trailing but supported when
cultivated. The stems are woody and the branchlets pubescent. The leaves
are opposite, glabrous, usually ovate or elliptic, very variable, and with
prominent lateral nerves. There are one to three leaflets to 7.5 cm, with
undulating margins on some cultivars. The flowers are very fragrant, white,
2 . 5 cm diameter, on solitary or three-flowered terminal cymes. T h e calyx
has five to nine teeth, and is linear, subulate, to 10 mm. The corolla lobes are
narrowly oblong to orbicular, to 12.5 mm and equal to the tube. The carpels
are black when ripe. Flower yield and strength of fragrance varies sub-
stantially between cultivars, many are double flowered with rounded petals
Fig. 9.1, Jasrninurn ofkinale forma grandifloruum,flowering branch. (CSIR, New Delhi.)
Oleaceae 347
and globular buds, while single flowers have rounded multi-whorled petals
and large buds to 3 cm diameter. T h e flowers are widely used in India to
make garlands and as part of temple offerings.
The flowers open between 8 and 9 p.m. and in India main flowering is
April-August. Concrete yield seldom exceeds 0.20%, but a single flowered
cultivar yielded 4200 kgiha flowers and 4.35 kgiha concrete (Sharma et al.,
1979). An absolute prepared by enfleurage in East Africa had these charac-
teristics: specific gravity (15°C) 1.024, refractive index (20°C) 2.41; optical
rotation 1.5061’; saponification value 153.3. An Indian absolute was deep
red, with a strong powerful odour suggestive of jasmine and orange flowers.
The constituents of J. sambac absolutes from various origins including
China, Indonesia and the Philippines have been compared (Lawrence,
1994). Constituents of the flower oil have been determined (Boros and
Stermitz, 1991), and the water-soluble extract from flowers contains various
glycosides (Zhang et al., 1995).
Ecology
Jasmines are tropical or subtropical plants, native to eastern and southern
Asia, Africa, Australia and basically day-neutral. Warm sunny conditions
promote growth and flowering, and with adequate soil moisture, jasmine
can withstand short periods of very high temperature without undue stress.
Growth and flowering are depressed by shade, low daytime temperature and
cool wet conditions. In regions where there are severe frosts, pruned plants
must be protected in winter.
Jasminum auriculatum is probably native to southern India, adapted to a
region with high temperature and above-average rainfall, hence the 2000 ha
under the crop in Tamil Nadu State. Temperature has a significant effect on
the onset and duration of flowering, and a relatively low day and higher night
temperature favours bud formation; thus highest flower yield occurs in
June-July in Tamil Nadu.
Jasminum grandzflorum is native to northern Iran, Afghanistan and Kash-
mir, and has been widely introduced and commercially cultivated in many
countries, principally around the Mediterranean. Basically adapted to a
mild-temperate climate, local cultivars are often drought and frost resistant.
An above average day and below average night temperature is required for
maximum floral initiation. Budding suitable J . grandzflorum cultivars on to
the local wild J . oficinale produced frost resistant plants in Sicily and
increased disease resistance in France. Most J. grandzflorum plantations are
below 500 m, and some are at sea-level in Europe, but the effect of altitude
on a specific cultivar has not been determined. T h e cultivar ‘Siu-En’
previously considered to be a form of J. odoratissimum is cultivated on
Taiwan; it is self-supporting, with opposite, pinnate leaves of five to seven
leaflets, the terminal leaflet being sessile. The flowers are whitish yellow,
2.1-2.7 cm in diameter.
348 Chapter 9
Jasrninurn sambac has a long history of cultivation and thus widely dis-
seminated in early times, but it is probably native to southern India where
birds are the main seed dispersal agent. The many local cultivars are adapted
to a wide range of environments, although all are frost sensitive and many
have little drought resistance. In China, protection of the local cultivar is
essential if plants are to survive the winter (Deng et al., 1986). Individual
cultivars react differently to a specific environment but, as the species is
commercially cultivated only in India, growers are well aware of any local
requirements. In general, cultivars with a potentially high flower yield have a
more restricted range than those with lower, as the former require longer day
length and above average day and night temperature to produce the highest
flower yield.
Cultivation
The most important operation prior to establishing a jasmine plantation is
deep ploughing, mainly to bring roots and similar material to the surface.
These must be collected and burnt, since jasmine is very susceptible to root
diseases spread from pathogens remaining on such debris; it was believed in
France that jasmine would only succeed on land that had remained fallow
for some years. Jasmine plants can be grown from seed, but viability is
seldom above 50%, and seed less than six months old should be used. In
India, a dormancy period of one month was noted in some local cultivars
(Srivastava and Karmakar, 1986). Jasmine is normally propagated by cut-
tings, layering or budding on selected rootstock and the technique selected
is that applicable to the local cultivar, since some provide easily-rooted
cuttings, others must be layered. Cuttings of 12-20 cm should be from
terminal shoots and dipping in a rooting preparation increases the strike
rate. Irradiation generally reduces root development (Devaiah and Shriv-
astava, 1989). Mist-propagation of small cuttings or plantlets is possible, but
requires special equipment.
Layering carried out in the field is simple but lengthy, and mature one-
year-old shoots provide the best material. A slanting cut is made approx-
imately half-way through the shoot some 50 cm from the end; this portion of
350 Chapter 9
the shoot (which must remain attached to the plant) is then buried horizon-
tally about 10-1 5 cm deep and pegged down with the terminal end remain-
ing above ground. Layers must be kept moist as roots form relatively slowly,
and it may be 4-6 months before the layers can be severed from the parent
and removed for transplanting.
Cuttings are usually placed in prepared planting holes and watered. In
some soils it is preferable to open the pits at least a week before planting, and
well mix the fertilizer with the soil to avoid damaging cuttings. Since jasmine
is very susceptible to soilborne fungal diseases, dipping cuttings in a fungi-
cide or adding it to the water is recommended. A spacing of 2 X 1.5 m is
common, and will require 3500-4500 cuttingslha. Spare cuttings should be
available for infilling. Local spacing varies very widely resulting in 3000 to
30,000 pph, but a high plant population generally produces more flowers
and thus concrete per hectare in the same locality than a low population. In
India, 4500-5000 pph consistently out-yielded lower populations, but in the
highly fertile and well-watered Nile Delta of Egypt, 3000-3500 pph is
common with flower yields of 8000 kglha.
Jasmine plants require support, ranging from individual stakes and trellis
to post and wire systems as used in vineyards; the choice entirely dependent
on local preference and cost. Since jasmine plantations can remain pro-
ductive for up to 15 years, a sturdy system should be selected. Jasmine plants
make slow growth in the first 2 years after planting out, require frequent
weeding, and manual or mechanical weeding should be as shallow as
possible. Herbicides are little used and are normally also too expensive for
smallholders. In India and other Asian countries intercropping in the first 2
years is usual to defray plantation establishment cost. In southern Italy,
jasmine is intercropped in bergamot orange plantations, which may not
become productive for 10-15 years.
Irrigation is normally necessary to ensure a continuous supply of flowers,
with water applied after pruning until end of flowering; in drier areas weekly
watering may be necessary to induce continuous flowering. Surface irriga-
tion is most common since overhead irrigation during flowering reduces the
amount of b n c r e t e produced,.and may also attenuate flower scent. It is
important locally to determine the most efficient method of applying water
and the amount, since both affect plantation profitability. Monitoring of
available soil moisture (ASM) is now simple with a range of inexpensive
instruments and should be the basis of any irrigation programme. In Tamil
Nadu, irrigating at 60% ASM produced highest flower yield from J. grand-
iflorum; delaying to 20% ASM reduced yield (Thamburaj et al., 1984).
Flowering begins about 6 months after planting out although flower yield
in the first 2 years is generally low. Plants flower profusely in the third year
and may continue to do so for another 10-1 2 years. Mature plants flower for
7-9 months in warm regions, 4-6 months in more temperate areas, and
main flowering is usually July-October in Europe, January-December in
Oleaceae 35 1
India. In Bangalore the peak flowering months of July, August and Sep-
tember contributed 50% of total flower yield; October, November, De-
cember 40%. In Egypt, August contributes one-third, July and September
another third, and June, October and November the remainder. In Turkey,
August and September contributed two thirds, July and October one third.
Growth regulators can affect the onset of flowering as noted, but have little
effect on duration. Pruning at various dates also had little effect on either
onset or duration of flowering in India, which remained June-July (Mu-
thuswami and Rao, 1980).
Annual pruning is necessary to ensure flowers are within reach of pickers,
to provide new growth, and to remove dead or diseased shoots. The degree
and extent of pruning depends on the cultivar and local experience, as the
rate of regrowth varies substantially. Since pruning is so important, its
relationship to flower and concrete production has been widely studied. In
India the effect on bud initiation in local cultivars has been investigated in
detail, but may not necessarily be relevant elsewhere (Subramanian et al.,
1985; Siddagangaiah, 1990). The majority of growers cut back plants to
around half their size after flowering, usually 90 cm, and some smallholders
also strip leaves to reduce disease carry-over. Early pruning in the first week
of December in Egypt gave higher yield of flowers than late pruning in the
last week of January (Table 9.2), but February is favoured in Turkey (Anac,
1986). Similar results have been reported in other countries, although the
differences in total annual yield were sometimes greater. Prunings must be
burnt to reduce the incidence of disease in the following season, and the ash
is used as fertilizer. After pruning, a thorough weeding should remove all
persistent weeds and grasses.
Harvesting
Flowers open in the early morning and there is a rapid and significant loss of
volatiles once buds have opened; for instance the important odoriferous
constituent indole fell by 0.6-0.8 mg/100 mg flowers in the first hours after
buds opened in India. There is also a substantial difference between volatiles
contained in living and harvested flowers; benzyl acetate fell from 60% to
half-opened and fresh fully opened white flowers only must be picked, not
buds, or flowers which have turned yellow as these will affect quality of the
absolute. Rain in the flowering period must be ignored as, although wet
flowers are of little value, to let them remain tends to suppress further
flowering.
Flower yield varies from 5500-1 0,500 kgiha, average 5000-8000 kgiha,
but a modern commercial plantation should aim at 8000-10,000 kg/ha, a
not unrealistic target since selections within local cultivars at local Indian
experimental stations gave flower yields of 12,500 kgiha with a concrete
yield of 0.35%-0.40%. A breeding programme can be expected quickly and
substantially to increase potential yield in a specific region. As an approx-
imate guide, 1000 kg flowers when solvent extracted yield 2.5-3.5 kg
concrete and half this amount as absolute.
The most important factors affecting total concrete yield are number of
flowers harvested per hectare and their concrete content. Unless the cultivar
chosen has a high potential yield of both, it is almost impossible to profitably
offset the loss by other methods; agronomic, administrative or processing.
The concrete yield of nine Indian cultivars ofJ. sambac for instance, varied
from 1.2 to 12 kgiha when grown under similar conditions; with average
concrete yield from J. grandzjlorum of 0.33% (double extraction gave
Distillation
Jasmine oil can be obtained directly from flowers by steam distillation, but
the yield is so low that the process is not viable. The main constituent is
benzyl acetate to 80%. Liquid jasmine absolutes are sometimes referred to as
jasmine oil and a number of analyses have been published on this basis. An
Indian jasmine oil obtained by liquid carbon dioxide extraction was golden
yellow, free from artefacts and solvent residue, and stated to have a good
shelf life. T h e benzyl acetate cQntent averaged 16%, benzyl benzoate 19%,
and phytol + methyl linoleate 14% (Gopalakrishnan and Narayanan,
1991).
Jasmine concrete is obtained from flowers by enfleurage or solvent extrac-
tion, most recently using liquid carbon dioxide. Enfleurage entails pressing
jasmine flowers between two plates (chassis) containing a fatty base (Fig.
9.2). Each batch ofblooms is pressed for 24-48 hours, then spent flowers are
replaced by freshly picked. This operation is repeated until the fat becomes
saturated with flower oil, which takes several months. The oil is then
extracted with alcohol to yield jasmine de pomade. A yield 2.5 times that of
direct solvent extraction of flowers has been recorded. Pommade is used
exclusively as the precursor to absolute. Exhausted flowers from enfleurage
Oieaceae 355
grade raddi, and the last two are prepared from exhausted flowers used to
produce sira.
Osmanthus
The genus Osrnanthus contains mainly evergreen shrubs and generally small
trees native to South East Asia through Polynesia to North America; some
produce a locally utilized timber, Osrnanthus arnericanus in the eastern USA
and 0. sandwicensis in Hawaii, but 0. fragrans native to the Himalayas,
China and Japan is the only essential oil producer. It is frequently planted
around temples and shrines.
Osrnanthusfragrans Lour. (syn. Olea fragrans Thunb.), sweet osmanthus,
is an evergreen dioecious shrub to 5-10 m. The stem is erect and branching;
the timber white, hard, close and even grained. The leaves are glossy green,
leathery, to 6-12 cm, elliptic to oblong-lanceolate, with the tip acute or
acuminate, the base cuneate, and the margins entire or finely and sharply
toothed, sometimes undulating, reticulate beneath. The flowers are creamy
white to 6 mm, borne in clusters on slender pedicels to 8 mm in leaf axils, the
corolla divided almost to the base. Main flowering is March-April in China,
although flowering can be prolonged in warmer regions. Osrnanthus flowers
are considered to have one of the most powerful scents of any plant, and are
used in China to perfume tea. The fruit is a purple ovoid drupe to 10-12
mm.
The tree prefers an open sunny situation, and will grow on a wide range of
free-draining soils, but is intolerant of water-logging. It is considered frost
resistant but drought-tender, and flourishes under intermittent shade. Prop-
agation is by cuttings from the current year’s growth taken in late summer,
and planted in pots or seedbeds for planting out the following spring. With
proper care, the strike rate is high. Plants can be raised from seed but
germination is often slow and erratic. Harvesting and treatment of flowers as
for jasmine.
Flowers are solvent extracted to produce an absolute whose main con-
stituents are: beta-ionone, geraniol, trans-and cis-linalool oxide and linalool.
A Chinese absolute included the following main components, in per cent:
beta-ionone 34, dihydro-beta-ionone 16, gamma-declactone 13, trans-
linalool oxide 6, cis- linalool oxide 6, and linool 4 (Lawrence, 1986). There
is very wide variation in published analyses, particularly between Chinese
and Japanese absolutes, with the latter considered to be closer to the scent of
fresh flowers. Fresh absolute readily oxidizes on exposure to light and air.
Most absolute produced in China is for the domestic market, but exports
have recently increased. The main use elsewhere is in high quality fragrances
and cosmetics.
360 Chapter 9
References
Anac, 0. (1986) Gas chromatographic analysis of absolutes and volatile oil isolated
from Turkish and foreign jasmine concretes. Flav. Frag. J. 1, 15-1 19. see also
Bull. Techn. Univ. Istanbul 41, 483-486 (1988).
Anon. (1978) Annual Report on Research and Development of Rose and Jasmine
Varieties. ICAR, New Delhi, India.
Atawia, B.A., Hallabo, S.A.S. and Morsi, M.K. (1988) Effect offreezing on jasmine
flowers and their concrete and absoiute qualities. Egypt. J. Food Sci. 16(1-2),
237-247 ~
Bai, P.Y. (1992) A new species of the genus Jasminum. Acta Bot.Yunnan. 14(4),
383-384.
Bhattacharjee, S.K. and Divakar, N.G. (1989) Growth flowering and essential oil
content of different species ofJasminum as affected by ethrel. Ind. J. Hort. 46(2),
240-243.
Boros, C.A. and Stermitz, F.R. (1991) Flower oil ofJasminum sambac. J . Nut. Prod.
54, 1173.
Chandra, V. (1982) Commercial exploitation of Jasnzinum species in India. In:
Cultivation and Utilization Aromatic Plants. CSIR, Jammu-Tawi, India, pp.
525-536.
Cheng, Y.S. and Chao, Y.L. (1977) Absolute of Siu-Eng flower U.odoratissimum).
In: Proc. V1I Int.Cong.Ess. Oils. Kyoto, Japan, pp. 467-469.
Deng, G. et al. (1986) Anatomical observations on cold-injured jasmine plants, and
effectiveness of various winter-protection coverings. Nut. Sci. J. Hunan Nomzal
Univ. 9(3), 73-81.
Devaiah, K.A. and Shrivastava, M.C. (1989) Sensitiveness of jasmine cuttings to
gamma radiation. Haba Hong. 28(3), 11-14.
Gopalakrishnan, N. and Narayanan, C.S. (1991) Carbon dioxide extraction of
Indian jasmine concrete. Flav. Frag. 3. 6(2), 135-138.
Gowda, V.N. and Gowda, J.W.N. (1990) Effect of Cycocel, maleic hydrazide on the
biochemical composition of J. sambac. Ind. Perfum. 34(4), 238-242 and
243-246.
Indiresha, K.M. et al. (1 986) Path analysis for floral traits in J. auriculatum. S. Znd.
Hort. 34(5), 355-357.
Kotliarova, M.V. (1 964) Absorption method for production of essential oil from
jasmine flowers. In: Proc. 11th Znt. Cong. Ess. Oils. Plovdiv, Bulgaria, pp.
85-86.
Lawrence, B (1986) Progress in essential oils. Perfum. Flav. 11(5), 76-79.
Lawrence, B.M. (1994) Progress in essential oils. Perfum. Flav. 19(2), 64-68.
Lekskul, S. (1975) Study on flowering habit and flower yield for essential oil of Thai
jasmine (Mali-La). Miscellaneous Investigation Report 75. A S K , Bangkok,
Thailand.
Mookherjee, B.D., Trankle, R.W. and Wilson, R.A. (1990) T h e chemistry of
flowers, fruits and spices; live vs dead. Pure Appld. Chem. 62, 1357-1364.
Muthuswami, S. and Tao, V.N.M. (1980) Heat units and flowering in Jasmine. In:
Proc. Nut. Senzin. Pro. Techniquesfor Comnzer. Flower Crops, Tamil Nadu Agric.
Univ., Tamil Nadu, India.
Oleaceae 361
362
Piperaceae 363
Pepper
The pepper vine (Piper nigrum) is native to the hills of western India,
esteemed as a spice in India since time immemorial, and dispersed in trade
throughout the world (Watt, 1908). So important was pepper that the search
for the source of the spice and thus control of the trade was a significant
factor influencing world exploration and history. The long pepper, P. Zongum
L. is the pippali of Sanskrit from which the Greek, Latin and most European
names are derived. It was probably the first pepper to reach the Medi-
terranean and was originally more highly regarded in the classical world than
P. nigrum. Two peppers were recognized by the Greek Theophrastus
(372-287 BC), black pepper and long pepper. Dioscorides mentions black
and white pepper which he believed were produced on different plants. In
the Roman Empire, pepper was well established as an article of commerce,
shipped to Red Sea ports by Arabs using the Arabian Sea monsoons.
Cosmos Indicopleustes visited India and Ceylon and describes accurately
the harvesting and preparation of pepper on the Malabar coast in Topo-
graphia Christiana (circa 545 AD). By the Middle Ages pepper was of great
importance in Europe to season or preserve meats, and overcome the odours
of rancid food. Peppercorns were very expensive and accepted as currency,
which survives in the term peppercorn rent. The statutes of Ethelred in
tenth-century England required early German spice traders to pay tribute,
including 10 pounds of pepper, for the privilege of trading with London
merchants. A pepperers’ guild of wholesale merchants founded in London
in 1180 was subsequently incorporated into a spicers’ guild, succeeded in
1429 by the present Grocers’ Company. The original pepperers and spicers
were the forerunners of apothecaries, emphasizing the role spices formerly
occupied in the medical treatment of patients. Pepper was no exception, and
concurrently with its use as a spice, pepper and later pepper oil had an
important place, and most early European herbal and medical treatises
attributed a range of beneficial effects to its use as a drink or external
emollient.
Pepper probably arrived in Java with early Hindu colonists and its wide-
spread cultivation dates from at least AD 500. Marco Polo describes pepper
cultivation in Java which he visited in 1280, of Chinese sailing vessels
loading up to 6000 baskets, and the abundance of pepper on the Malabar
coast of India (Description of the World, 1298). T h e great wealth of Venice
and Genoa in the thirteenth century was due largely to the spice trade.
Portugal’s desire to obtain part of this lucrative business was mainly respon-
sible for Vasco da Gama circumnavigating Africa and arriving at Calicut on
the Malabar coast of western India on 20 May 1498. This enabled Portugal
to secure a monopoly of the spice trade, which it maintained for a century.
Lisbon became the most important European trading centre for oriental
364 Chapter 10
spices, with entrep6ts in Goa and Malacca, which also exported pepper to
China. Varthema in his Travels gave a vivid picture of the Calicut plantations
in the sixteenth century:
It is like a vine climbing on trees: from each of the branches are produced five
to eight clusters of berries, a little longer than a man’s finger; they are like
raisins but more regularly arranged, and are as green as unripe grapes; they
gather them in October and November, and lay them in the sun on mats,
when they turn black as they are seen among us, without doing anything else
to them. And you must know that these people neither prune nor hoe this tree
which produces the pepper.
Also, at the beginning of the sixteenth century, Barbosa in his Coasts of East
Africa and Malabar gives a detailed account of the plant and of the trade in
pepper from Calicut shortly after the arrival of the Portuguese. Early in the
seventeenth century the Dutch acquired pepper-producing areas near Ban-
tam in Java and Lampong, Sumatra. Although the Dutch subsequently
controlled much of the pepper trade, they never succeeded in maintaining a
complete monopoly as they did with clove and nutmeg (Boxer, 1965). The
demise of Dutch dominance coincided with the entry of the USA into the
Far Eastern spice trade, and in 1797 Captain Jonathan Carnes of Salem,
Massachusetts, sailed the schooner Rajah into New York harbour carrying
pepper worth $100,000 from Sumatra. During the next 50 years ships from
Salem and Boston played an important role in world pepper trade, and in the
first half of the nineteenth century this trade reputedly produced in Salem
some of America’s first millionaires (Putnam, 1924; Rosengarten, 1969).
Early in the nineteenth century the British organized pepper planting in
Malaysia and later Sarawak.
The main pepper producers are India at around 50,000 t annually with
exports of two thirds, Indonesia which exported an average of 35,000 t i n the
years 1993-1995; Sarawak, Malaysia, which produced an average 30,000 t
in 1990-1 994, mainly exported; and Brazil which rapidly increased produc-
tion in the last decade to 40,000 t, mainly exported. These four countries
were the founder members of the International Pepper Community, now
including other producers. Countries producing small quantities for export
are Costa Rica, Madagascar, Kampuchea, Sri Lanka and Vietnam, while
others including the Philippines consume their production domestically.
Total pepper available on the world market in the crop years 1992-1993,
1993-1994 and 1994-1995 was, in tonnes, 163,000,109,000 and 117,000,
indicating the major effect of inclement weather on the crop; Indonesia for
example lost two thirds of total production in one year due to drought.
Other Piper species and the parts used are shown in Table 10.1.
Pepper spice is the dried berries of P. nigrunz, black pepper from the whole
berry, white pepper from grinding mature berries after removal of the
mesocarp. Production and preparation of the spice has been described
Piperaceae 365
Table 10.1. Less important Piper species and the parts used,
(Purseglove et al., 1981) and of green pepper (Mathew, 1993). The uses and
derivatives of a number of species have been described: P. guineense (Ekun-
day0 et al., 1988), P. aduncum L. (Orjala et al., 1993); and Pimenta spp. from
the Caribbean in a series of papers by D r A.O. Tucker and colleagues in the
Journal of Essential Oil Research from 1991 on. All references to pepper in this
text refer to I? nigrum and its derivatives.
Botany
Piper nigrum L.: in English pepper; Dutch peper; Spanish pimienta negra;
Portuguese pimenta; Arabic jZj?l; India, Hindi kalimirch, Tamil milagu, and
many others; the Philippines paminta; Malaysia lada; Indonesia generally
lada, black pepper lada hitam.
The cytology of the genus Piper is confused; Darlington and Wylie (1955)
give the basic chromosome number of the genus Piper as x = 12, 14, 16, and
somatic number of P. nigrum as 2n = 128. Mathew (1974) gives 2n = 52 for
six cultivars and four wild types, and 2n= 104 for two other wild types.
Samuel and Bavappa (1 98 1) reported chromosome numbers from nine
Piper spp. including 2n = 26 for P. longum and 2n = 52 for P. betle. Jose and
Sharma (1 984) state three P. nigrum cultivars P. attenuatum, P. boehmer-
iaefolium and P. longum had 2n = 52, one P. nigrum cultivar and P. chaba
2n = 104, l? peepuloides 2n = 156; P. cubeba 2n = 24 with large chromosomes,
high total chromatin length, was regarded as primitive. Chemotaxonomy
has established the relationships between south Indian peppers (Ravindran
and Babu, 1994).
Piper nigrum is a perennial, glabrous woody climber to 10 m, sometimes
taller, very variable in physical characteristics, and virtually dioecious.
Cultivars are generally hermaphrodite. The root system is extensive but
shallow with a few large roots to 4 m, the major portion remaining within 1
m of the surface (often the first 40 cm), in a circle round the stem. There is
initially one main climbing stem but others are rapidly produced which
branch profusely and, under cultivation, mature vines have a bushy, colum-
nar shape (Fig. 10.1). Climbing stems are green tinged with purple when
366 Chapter 10
young, dark green with age; mature stems are 4-6 cm in diameter, woody,
with thick, flake-like bark; fruiting branches are produced from these stems.
The pungent principle ofpepper fruit, piperine, is present at 0.2% in mature
fully differentiated shoots, but a trace only in leaves and young shoots
(Semler and Gross, 1988).
Leaves are alternate, single, 8-20 X 4-12 cm depending mainly on
cultivar, petiole 2-5 cm. The lamina is ovate, entire, coriaceous, with five to
seven main veins; the base oblique, obtuse or rounded; the tip acuminate.
The leaves are usually dark shiny green above, paler and glandular below.
Since there is a direct relationship between leaves produced and fruiting
spikes, cultural techniques in commercial plantations aim for healthy, well-
foliaged vines. The leaves also contain an essential oil which differs from
berry oil, with a scent resembling lime.
The inflorescence is a pendant spike on lateral branches, 3-1 5 cm, bearing
50-1 50 very small yellowish-green to whitish-yellow flowers. The flowers
may be unisexual, but are frequently bisexual in cultivars and in general the
higher the percentage of bisexual flowers in a particular cultivar the higher
the berry yield. The flowers lack a perianth, the two to four minute stamens
being borne laterally on the ovate sessile ovary surmounted by three to five
somewhat fleshy stigmas. Cross-fertilization is usual in the wild and wind the
main agent, but self-pollination appears to be general in hermaphrodite
flowers. Main flowering is normally at the onset of the monsoon; in Kerala it
is May-August with a July peak.
The fruit is a single-seeded, sessile, indehiscent, globose drupe (the
berry), 3-6 mm in diameter, the pericarp is pulpy, with 50-60 fruits per
spike. Unripe fruit is green, becoming red when ripe and black after drying.
When dry the black wrinkled skin encloses a single whitish seed, known as a
Piperaceae 367
Ecology
Pepper is native to wet tropical forests with its centre of origin and diversity
the Western Ghats, Tamil Nadu State, India, formerly known as the
Malabar Coast, which remains India's main pepper growing region. Pepper
is also wild in Assam and Borneo, most likely an escape from early cultiva-
tion. It has since been introduced wherever there was the slightest possibility
Piperaceae 369
it would grow, so popular is the spice! A pepper vine is easily grown in a large
pot, and if trained up a suitable support will provide a plentiful supply of
fresh berries for over a decade. The adaptability of P. nigrum is demonstrated
by the large number of cultivars existing in particular countries; there are
nearly 100 in India alone. This wide range also provides plant breeders with
a choice of material when selecting a specific plant characteristic.
Pepper is basically limited to a band from 20"N to 20"s of the Equator;
commercial pepper production in Indonesia is within a few degrees of the
Equator, in Sri Lanka and Malaysia about 10"N, in southern India
12-15"N, Madagascar 15"s and Brazil to 15"s. Pepper flourishes and
produces the highest berry yield from sea-level to 500 m, but can be grown
to 1500 m provided temperature and rainfall are suitable. There is often a
decrease in berry yield or quality at the higher limit, but no comparative data
for a specific cultivar are available on the effect of altitude on growth or berry
composition.
A rainfall of 2500-3500 mm annually, well distributed over the year is the
optimum, but good berry yields can be obtained with 1500 mm provided
this falls mainly in the growing and flowering season, and soil moisture is
conserved. Warm, humid and cloudy conditions promote growth, with an
average day temperature of 2 5 3 5 ° C . Long periods of strong sunlight will
scorch unshaded plants, and the north slopes of hills are preferred for pepper
plantations in India and Sri Lanka.
Light intensity is important in determining growth and thus berry yield,
and there is not only a general reaction to this factor but also cultivar
differences (Mathai, 1986). Vines, especially seedlings, benefit from inter-
mittent or partial shade but heavy shade adversely affects flowering and fruit
370 Chapter 10
setting. Prolonged very cloudy or misty periods also have a similar but less
damaging effect.
The minimum temperature in main growing regions seldom falls below
15-1 8"C, and low temperatures depress reproductive more than vegetative
growth. Very hot dry wind inhibits growth, may kill young plants and
adversely affect pollination. Windbreaks are recommended in exposed situa-
tions, particularly when vines are not grown on living supports which
provide partial protection in addition to shade.
Cultivation
As a small farmer crop, pepper vines are normally interplanted or trained up
existing trees, and little cultivation is necessary other than maintenance of a
weed-free area round the base of the vine. In Bangladesh and the Phil-
ippines, vines are sometimes grown up existing forest trees, and the adjacent
area kept clear of undergrowth. Intercropping pepper vines in commercial
plantations of other crops is not usually successful except in specific circum-
stances, e.g. coffee in India (Korikanthimath, 1994), since there is fre-
quently a conflict between management techniques. Cocoa, for example,
requires a fairly heavy shade which is inimical to pepper, while mango is host
to insects which can devastate pepper vines (Korikanthimath, 1994).
T o establish a plantation the land must be cleared and all debris burnt and
the ash spread. Topsoil disturbance must be minimized, since the bulk of
pepper roots remain near the surface. A major decision must be whether
vines are to be grown up living or dead supports. Factors affecting this
decision are whether the plantation is on an exposed or protected site, the
extent to which shading will be required, the rainfall and the availability of
quick-growing suitable trees, or timber for supports. In Sri Lanka, supports
made from eucalyptus timber gave higher berry yield than vines trained up
living trees (Bavappa et al., 1981), but living trees as supports are becoming
more common in Indonesia (Seibert, 1988). Concrete posts or circular
frames made from stone blocks are used in addition to timber supports in
Vietnam (Pohlan and Heynoldt, 1986), and also in India (Reddy et al.,
1991). Suitable trees for live supports, in addition to existing coconut, cola
nut, kapok or rubber plantations, are Eythrina spp., Leucaena spp., Grevillea
robusta, and Gliricidia spp.; in Madagascar, vines are often trained up shade
trees in coffee plantations.
The type of support also influences fertilizer uptake, as vines on living or
dead supports differed in their uptake of phosphate at the same rate and
method of application (Geetha et al. , 1993). The height of the stakes is also
important; in Brazil 2.5 m gave a higher average berry yield than 1.5 m at
3760 and 2890 kgiha respectively (Kato and Albuquerque, 1980), but in
India and Sarawak stakes are usually much higher at 4-6 m. Spacing
between supports depends on the method of weed control, and if this is
mechanical then 3-5 m must be left between rows to allow passage of a small
tractor with implements. Spacing on non-mechanized estates is usually 2-3
m on the square, with timber supports to 4 m in Sarawak, to 9 m in India,
while living supports are pruned of branches to this height. The cost of
Piperaceae 373
node from the apex are removed, and when the terminal bud has re-
generated, the shoot is carefully cut below the seventh node. Greenhouse
trials in Indonesia indicated topping or ring-barking of parent vines before
taking cuttings increased extent and rooting ability (Yufdi, 1986).
Cuttings should preferably be rooted in containers or a nursery and
treatment with a rooting compound usually results in more rapid root
growth; pre-rooting also ensures quick growth in the field and reduces
expensive infilling. Cuttings should be planted in the nursery at least 4-5
months before planting out. Alternatively, cuttings can be planted directly
into mounds, but large numbers are required which may not be available and
losses are usually high. Cuttings are usually planted at 45" with three or four
nodes below the surface and not deeper than 10-1 5 cm. If planted direct into
mounds, leaves and branches of the top three nodes are left and trained up
a temporary stake; however in Indonesia, cuttings treated with a rooting
compound and placed horizontally in the soil produced more roots more
quickly than obliquely planted cuttings (Nuryani, 1984). Cuttings should be
planted out at the beginning of the rainy season; in India July-September; in
Sarawak, September; in Sri Lanka, October-November; and in north-east
Brazil, January-February. When plants have made sufficient growth they are
tied to supports, three climbing stems retained, and regularly pruned to
encourage development of lateral fruiting branches.
Stems are pruned to 15-22.5 cm when 8-10 nodes have developed, and
the leaves removed. When a further 9-10 nodes have developed, stems are
cut back to within 3-4 nodes of the previous cut. Vines are thus pruned
seven to eight times until they reach the top of support posts at about 3 m.
The terminal growth is pinched out periodically to prevent further growth,
and low side branches and stolons removed. The main object ofpruning is to
remove unproductive nodes without lateral branches, leaving only stems
with nodes each bearing a branch. Thus a mature vine will have the
maximum number of fruiting branches with profuse leaf development, as a
flower spike can only develop in a leaf axil. During the first 2 years, flower
spikes are stripped combined with selective leaf removal to encourage
branching.
Weed control in pepper plantations is essential as the shallow roots will be
in direct competition with weeds for plant nutrients. Although mechanical
weed control is possible, pepper roots are easily damaged by too-deep
cultivation; where a ground cover is desirable, mowing or slashing is effec-
tive. In most pepper plantations, hand weeding is normal and this too must
be shallow; deep-rooted grasses such as Imperata cylindrica are easily and
cheaply eliminated with a herbicide wand or spot spraying. Mulching can
also greatly reduce weed growth in addition to conserving moisture, and
significantly increase berry yield. Mulching with rice husks, hay or sawdust
at 3 t/ha in Brazil proved sawdust to be the most effective, with a seed yield
of 3600-7800 kgiha compared to rice, second at 3500-7300 kgiha (Kato et
Piperaceae 375
al., 1980). Herbicides are little used but, as pepper is very shallow rooted,
soil inactivated compounds are to be preferred.
Harvesting
Pepper vines are normally harvested 18-20 months after planting out.
Picking in India is in January-March in the hills, November-January on the
plains; in Sarawak, May-August with the heaviest yield in July-August; in
Indonesia, August-November. The main harvest in Sri Lanka is October-
January with a minor crop in June-July, and in peninsular Malaysia, in
August-September and March-April. Small farmer vines are usually har-
vested year-round, and other factors may locally affect time of harvest; in Sri
Lanka for instance berries are often picked earlier than the optimum time to
reduce loss through theft. Time of picking (i.e. maturity of berry) affects
berry composition (Tables 10.2) and berry oil (Table 10.4). Picking for
white pepper begins when one to three berries on a fruiting spike turn red,
and for black when berries are still green, and picking continues every 2-3
weeks until end of harvest. Tripod ladders are used to reach higher branches
and avoid damage to vines. On intensively managed plantations, branches
are stripped of leaves except for one or two terminal leaves when all fruiting
branches have been removed. Damaged and diseased branches are also
removed.
The yield of green berries per vine varies widely, from a few grams to a very
high 10 kg from specially selected and treated plants. This range indicates the
potential yield possible and, in general, any improvement in cultural tech-
niques by small producers quickly increases yield. Green berry yield from
well managed plantations in Sarawak averages 8000-9000 kgiha from the
first picking rising to 18,000 kg/ha at the sixth to seventh harvest; maintained
to the tenth year. In India, yield is generally lower at 100-400 kgiha since
Table 10.4. Effect of maturity on some selected oil constituents in Sri Lanka.
Distillation
Distillation is uncomplicated and no special equipment or technique neces-
sary. Pepper berries are reduced to a coarse powder and on steam distillation
yield 1.O-4.0% oil, but specific varieties or grades to 9%. Whole berries can
vines is that the flowering and fruiting period is condensed. Thus the time
spikes are exposed to insect attack is also reduced and the opportunity for
effective control enhanced.
Stem borers may cause considerable damage where they occur, but are
seldom of major importance; most often reported in India are Pterolophia
annulata and Diboma procera. The greyish-green larva of the small brown
moth, Thosea sinensis can cause serious damage to foliage, and has been
reported on pepper from India through Indonesia, South East Asia to China.
Larvae of another moth Cricula trifenestrata, the mango hairy caterpillar, can
become a major pest in India especially where mango trees are used to
support or shade vines; vines can be completely defoliated. Caterpillars of
both these moths bear urticating hairs, and when present in numbers make
berry picking almost impossible. Scale insects and mealy bugs debilitate
vines and can cause severe damage if not controlled. Most often mentioned
are the cosmopolitan Fem'sia virgata and Saissetia coffeae; the pepperscale
Lepidosaphes piperis in India and Sri Lanka; and Planococcus citri in Indo-
China, especially Kampuchea.
Roots are host to various nematode species including the root-knot
nematodes, Meloidogyne javanica and M . incognita, and the burrowing
nematode Radopholus similis. The extent of damage has not always been
accurately determined, nor the exact relationship between root disease and
local nematode species. An association between R. similis and incidence of
vine wilt was established in India, and control of the nematode reduced
damage to vines (CPCRI, 1985). A relationship between nematodes and
incidence of pepper yellows caused by Fusarium oxysporum has also been
observed in Indonesia (Sitepu, 1993). In Brazil, treatment of cuttings with a
nematicide significantly increased growth, and different mulches had vary-
ing effects on growth of vines and soil nematode populations; a combination
of both is recommended (Ichinohe, 1985). Nematicides also control soil
nematodes in India (Mohandas and Ramana, 1987) and in Indonesia, but
some were toxic to vines.
Stored berries are often attacked, most commonly by Sitodrepa panicea,
Pyralis fannalis and Lasioderma sem'corne. Application of berry oil to pepper
and other whole spices slows the rate of population growth and inhibits
larval development. Washing and redrying to reduce microbiological con-
tamination is also effective in removing most surface insect contamination,
but fumigation may be necessary in serious infestations.
A number of pathogenic fungi have been reported from P. nigrum, but
only two or three are of major importance and generally only one locally.
Probably the most economically important is a root rot caused by Phyto-
phthorapalmivora, known as quick wilt in India and sudden death in Sarawak
and Indonesia because of its effect on vines. Initial infection in the field
begins in minor roots a few centimetres below soil level, then spreads to
major roots and the stem. When an appreciable portion of the underground
380 Chapter 10
stem has been attacked wilt symptoms appear, usually 50-60 days after
infection. The first above-ground symptom is a slight droop over the whole
vine, followed by yellowing of leaves which usually fall within 7-14 days
leaving the branches bare, and the vine dies. In wet weather, leaves may be
directly attacked causing brown circular lesions with a fimbriate edge;
abscission occurs before infection can spread to the stem. Artificial inocula-
tion of root cuttings produced typical symptoms after 6-9 days, indicating
the ease and rapidity with which the pathogen can spread. Giant African
snails and nematodes are disease vectors, and control measures are effective
(Sadanandan, 1989).
Resistance to Phytophthora palmivora varies substantially, with Piper spp.
from South East Asia generally highly susceptible, but American spp.
resistant or slightly to moderately susceptible; Peperomia spp. are considered
resistant. The damage caused by quick wilt in India is substantial; about 120
t of pepper is lost annually in one district of Kerala State (Balakrishnan et al.,
1986; Nair and Sasikumaran, 199 1). In Brazil, the disease can be devastat-
ing on susceptible varieties; in Bahia Province plantings were reduced by half
prior to the routine application of fungicides (Tsuda, 1989). Another local
species, P. colubrinum, was resistant to Phytophthora palmivora and Fusarium
solani f.sp. piperis and used as rootstock for susceptible P. nigrum, but grafts
deteriorated after the fourth year. In Brazil and Puerto Rico, root-rot
epidemics caused by Phytophthora palmivora usually occurred when vines
began fruiting, but the incidence was low in new areas. Research into control
of P. palmivora is a major project at the Malaysian Agricultural Research and
Development Institute (MARDI) Biotechnology Centre which is also pro-
ducing pathogen-free plants by in vitro culture.
Phytophthora capsici is the most important pathogen in Indonesia where it
causes sudden death disease, now the major constraint on expansion of
pepper production, The disease is difficult to control or its spread limited
because most pepper is grown by small farmers few of whom apply either
pesticides or fungicides. Resistance at various levels may exist in some
cultivars (Mozzetti et al., 1995), and a biological control programme has
proved of value in India (Anandaraj and Sarma, 1994). Root rots or wilts
caused by the following are locally damaging: Fusarium spp., Diplodia spp.,
Rosellinia bunodes and Sclerotium rolfii in various regions of India; Fomes
lignosus, Ganoderma lucidum and Rhizoctonia solani in Sarawak, and various
Pythium spp. in Indo-China. The importance of root nematodes in spread-
ing root and other diseases has still to be determined.
Diseases which can originate in the canopy are often due to Cortzcium
salmonicolor, pink disease, or Marasmius scandens, thread blight, and both
may cause almost complete necrosis of leaves and branches. They can be
distinguished from root rot as the symptoms differ. In pink disease, white
pustules occur on the bark or lateral branches, followed by the pink encrust-
ing mass of the fungus. In thread blight, cobweb-like strands grow over the
fiperaceae 38 1
Component %
Alpha-pinene 5.8
Camphene 0.1
Beta-pinene 10.4
Delta-3-carene 20.2
Limonene 17.1
Gamma-terpinene 0.2
p-Cymene 0.8
Terpinolene 1.o
Delta-elemene 2.5
Alpha-copaene 2.3
Beta-elemene 0.4
Beta-caryophyllene 27.8
Alpha-humulene 1.4
Delta-cadinene 0.8
Caryophyllene oxide 0.6
Source: Analyst 109, 1359.
essential to ascertain the method used (Weaver et al., 1988; Wood et al.,
1988). The isomers of piperine and relative pungency have been stated as:
piperine (tram-tram) 5 ; isopiperine (cis-tram) 1; isochavicine (tram-cis) 1;
chavicine (cis-cis) 2 . Interestingly the pungency of piperine is only 200,000
Scoville units (SU), compared to 30 million SU for capsaicin from Capsicum
species.
The oleoresin is available as a liquid, dissolved in an edible solvent (e.g.
polyoxyethylene sorbitan fatty acid esters), emulsified with gum arabic and
dispersed on an edible dry powder carrier (salt or sugar), encapsulated in
gum arabic or gelatine. One kg of oleoresin obtained from 8 kg of black
pepper and dispersed on an inert base can replace up to 25 kg of the spice for
flavouring purposes. The EOA specification for oleoresin is shown in Table
10.8, and adopted by many countries. Hazardous solvent residues some-
times occur as contaminants and prohibit use of the oleoresin in foods.
The principal use for oleoresin is in processed and canned meats, pickles,
dressings of all kinds and a variety of other food products where the spice
enhances their flavour. A decolorized oleoresin is available for special use.
An extract of black pepper, ultrasonic, is normally free of harmful solvent
residues and commercially available for use in processed foods. White
pepper oil is occasionally produced using the same methods as for black, but
is normally more expensive due to the higher cost of raw material and lower
yield. There is apparently no basic difference between the two oils when
produced from similar quality material.
Cubeb pepper
Piper cubeba is indigenous to Malaysia and Indonesia and as tailed pepper
was one of the first peppers to reach Europe. The fruits have long been used
in their native home by local peoples as a condiment and in herbal medi-
cines, and powdered dry fruits widely used to treat amoebic dysentery
(Burkhill, 1966). These uses were maintained by Arabs who traded with the
region and named the fruit kabab. Used mainly as a spice in Classical
Europe, it later became more widely used in medicine, but by the end of the
Botany
Piper cubeba L. is commonly cubeb or tailed pepper in English; Arabic
kababa; usually kemukus in Malaysia and Indonesia; and in Hindi kabab
chini. It is generally a perennial climber, occasionally a shrub, of tropical
forests. The stem is greyish, smooth and flexuous. The leaves are glabrous,
entire, petiolated, acuminate, oblong to ovate-oblong, coriaceous and ob-
liquely cordate. The inflorescence is a spike bearing clusters of sometimes 50
or more dioecious flowers. The fruit is a small to 6 mm diameter, reddish-
yellow drupe, drying to black, with a coarsely reticulated surface. The base
of the pericarp extends into a pedicel-like formation and seed is usually free
from the pericarp. The histology of fruit and seed has been discussed (Parry,
1962).
For use as a spice the fruit is harvested when mature but unripe together
with a portion of stem, hence the common name of tailed pepper. Dried
fruits resemble peppercorns, and good quality fruit has a pleasant aromatic
odour and a bitter, pungent taste resembling allspice. The fruit yields 4-30%
essential oil and this very wide range indicates not only differing distilling
methods but considerable regional variation in fruit oil content. Indonesian
fruit locally distilled seldom yields above 10% oil, but the yield can be up to
20% from Malaysian and Sri Lankan fruits distilled in Europe. The main
constituents of cubeb oil are the sesquiterpenes cadinene and dipentene,
plus small amounts of starch, calcium and magnesium malate; the main
characteristics of steam-distilled oil are shown in Table 10.9.
Cultivation
Cubeb is wholly a small farmer crop and unlikely to be grown on a plantation
scale, but methods successful for black pepper would be equally suitable.
Pests, diseases and control methods are also similar, with some small
regional variation to suit particular conditions. Preparation of dried spice
and distillation for oil is similar to black pepper. Distilling time affects not
only yield but specific gravity, optical rotation and acid value; prolonged
distillation in Indonesia increased oil yield and raised values of the three
386 Chapter 10
A B C
Specific gravity (25°C) 0.905 0.9167 0.8994
0.924 (15°C)
Optical rotation (25%) -25O -29.9O - 10.7'
-40°
Refractive index (20°C) 1.490 1.4894 1.4885
1.496 (25°C) (25%)
Acid value ng ng 0.3
Ester value ng ng 15.2
Saponification number ng 0.5 ng
Solubility (alcohol 90%) up to 10 1:5 nQ
A, Good quality commercial oil; B, C, commercial Indian oils; ng, not given. Figures in columns are
range.
characteristics mentioned. Weight of charge can also affect acid and saponi-
fication values (Rusli and Soepandi, 198 1).
Betel Pepper
Piper betle is native to Malaysia, Sumatra and probably also Java, but was
introduced into many countries of Asia, South East Asia and Africa for its
Piperaceae 387
from one to several hundred vines, but there are no large commercial
plantations. Successful cultivation of vines requires close and constant
attention and this is only possible because of the strong and sustained
demand for fresh leaves. It is most unlikely that the vine will ever be
cultivated to produce essential oil, and for this reason will not be discussed,
although it is fully described elsewhere (Balasubrahmanyam et al., 1994).
The essential oil is obtained by steam distilling fresh, sometimes partially
dried leaves with a yield of 0.5- 2.0%; young leaves give a higher yield than
those fully mature. There are substantial regional differences in leaf oil
content and oil composition from local cultivars. Those from India have a
higher leaf oil content and more pungent taste, and for these reasons have
been introduced to many countries in preference to Malaysian or Javan
cultivars.
Betel oil is produced on a very small scale from fresh leaves and is seldom
used by European or North American manufacturers. The main use is in
Asia and supplies the betel taste and flavour where leaves are not easily
obtained. Betel oil is usually a light to medium brown, occasionally bright
yellow, depending on the type of leaf used, i.e. young or old, fresh or dried.
The odour is sharp, distinctly phenolic and reminiscent of creosote or strong
tea, the taste acrid, bitter and burning. The specific gravity of an Indonesian
oil was 1.02-1.04 (15"C), a Malaysian 0.9-1.02 and an Indian 1.00-1.06,
temperature not specified. The major constituents are phenols to SO%,
mainly eugenol, and the main characteristics and constituents of Indian
betel oils are shown in Table 10.10, which indicates the wide regional
variation. The oil has significant fungicidal and nematicidal activity (Evans
et al., 1984).
References
Anandaraj, M. and Sarma, Y.R. (1994) Biological control of black pepper diseases.
Ind. Cocoa, Arecanut SpicesJ. 18(1), 22-23.
Arctander, S. (1960) Perfume and Flavour Materials of Natural Origin. Mrs G .
Arctander, 6665 Valley View Blvd, Las Vegas, Nevada, USA.
Balakrishnan, R. et al. (1986) Estimates on the extent ofloss due to quick wilt disease
of black pepper in Calicut district of Kerala. J. Plant. Crops 14(1), 15-18.
Balasubrahmanyam, V.R. et al. (1 994) Betelvine - Piper betle L. National Botanical
Research Institute, Lucknow, India.
Bavappa, K.V.A., Gurusingha, P. de A. and Anadaswamy, B. (1 98 1) Pepper. Tech.
Bull. No. 4. Dept of Minor Export Crops, Colombo, Sri Lanka.
Boxer, C.R. (1965) The Dutch Seaborne Empire 1600-1800. Hutchinson, London,
UK.
Brandao, A.L. de, Tafani, R.R. and Lima, L.P. (1 978) Economic viability of black
pepper cultivation in the cacao region of Bahia. Bole. Tech. Cent. de Pesqziisas do
cacau. N o . 64. Bahia, Brazil.
Burkill, I.H. (1 996) A Dictionay of the Economic Products of the Malay Peninsula, 2
vols, Govt. Printer, Kuala Lumpur, ihlalaysia.
CPCRI (1985) Quick and slow wilt diseases of black pepper. Research Highlights
1985. Central Plantation Crops Research Institute, Kerala, India.
Darlington, C.D. and Wylie, A.P. (1955) Chromosome Atlas of Flowering Plants. 2nd
edn. Allen & Unwin, London, UK.
Ekundaya, 0. et al., (1988) Essential oil constituents of P. guineense. J. Agric. Food
Chew 36(5), 880-882.
Evans, P.H., Bowers, W.S. and Funk, E.J. (1984) Identification of fungicidal and
nematicidal components in the leaves of P. betle. J. Agric. Food Chem. 32(6),
1254-1256.
Geetha, C.K. Aravindakshan, M. and Wahid, P.A. (1993) Influence of method of
fertilizer application on nutrient absorption of black pepper vines. S. Ind. Hort.
41(2), 95-100.
Geus, J.G. de (1973) Fertilizer Guidefor the Tropics and Subtropics, 2nd edn. Centre
d'Etude Azote, Zurich, Switzerland.
390 Chapter 10
Gopalkrishnan, M. et al. (1993) Analysis and odour profiles of four new Indian
genotypes of P. nigrum. J. Ess. Oil.Res. 5(3), 247-253.
Ichinoho, M . (1985) Integrated control of root-knot nematode M . incognita on black
pepper plantation in the Amazonian region. Agric. Ecosystems Envir. 12(4),
27 1-283.
Jansz, E.R. et al. (1983a) Pepper. Monograph No. 2. CISIR, Colombo, Sri Lanka.
Jansz, E.R. et al. (1983b) Liberation of volatile oil from pepper. Proc. 12th Ann. Sess.
Int. Chem. Committee. (Unpaginated). Colombo, Sri Lanka.
Jansz, E.R. et al. (1984) Effect of maturity on some chemical constituents of pepper.
3. Sci. FoodAgric. 35, 41-46.
Jose, J. and Sharma, A.K. (1984) Chromosome studies in the genus Piper L. J. Indian
Bot. SOC.63(3), 313-319.
Kamruddin, A.T. and Wenur, F. (1994) Drying of black pepper by solar energy. In:
Proc 9th Int. Dying Sym. Vol B, 887-895. Gold Coast, Australia.
Kanakamany, M.T. et al. (1985) Key for identification of different cultivars of
pepper. Ind. Cocoa, Arecanut, Spices J. 9(1), 6-1 1.
Kandiannan, K. et al. (1994) Growth regulators in black pepper production. Znd.
Cocoa, Arecanut. SpzcesJ. 18(4), 119-123.
Kato, O.R. and Albuquerque F.C. (1980) Relationship between support size and
black pepper productivity. Pesquisa em Andamento. Alt. 4, 3-6.
Kato, O.R. et al. (1980) Effect of type of mulch on black pepper. Pesquisa em
Andamento. Alt. 3, 3-6.
Korikanthimath, V.S. and Peter, K.V. (1992) Coffee based pepper cropping system.
Planters Chronicle 87(3), 121-139; also 141-143.
Kumar, T.P. and Nair, M.R.G.K. (1987) Effect of some planting conditions on
infestation of black pepper by L. nigripennis. Znd. Cocoa, Arecanut, Spices J.
10(4), 83-84.
Malijan, H.V. (1982) Parninta in our Forests. Agroforestry Research Centre, The
Philippines.
Mathai, C.K. (1986) Growth and yield analysis in black pepper varieties. MysoreJ.
Agric. Sci. 18(4), 317-319.
Mathai, C.K. (1986) Growth and yield analysis in Black Pepper varieties. MysoreJ.
Agric. Sci. 18(4), 317-319.
Mathew, A.G. (1990) Oil and oleoresin from Indian spices. In: Proc. 11th Int. Cong.
Essen. Oils &Flavours. pp. 189-195. Aspect Pubs, London, UK.
Mathew, A.G. (1993) Green and white pepper. Znt. Pepper News Bull. 17(3),
10-13.
Mathew, F.M. (1974) Karyomorphological studies in P.nigrum. J . Plant. Crops 1,
15-18.
Milanez, D. et al. (1987) Culture ofBluck Pepper. Bull. No. 33, EMCAPA, Campinas,
Brazil.
Milanez, D. and Ventura, J.A. (1987) Methods f o r Producing Black Pepper Cuttings.
Bull. No. 42, EMCAPA, Campinas, Brazil.
Mohandas, C. Ramana, K.V. (1987) Slow wilt disease of black pepper and its
control. Znd. Cocoa, Arecanut, Spices J. 1 1(l), 10-1 1.
Mozzetti, C. et al., (1995) Variation in enzyme activities in leaves and cell suspen-
sions as markers of incompatibility in different Phytophthoraipepper inter-
actions. Physiol. Molec. Plt. Path. 46, 95-107.
Piperaceae 39 1
Nair, P.K.U. and Sasikumaran, S. (1991) Effect of some fungicides on quick wilt of
black pepper. Znd. Cocoa, Arecanut, SpicesJ. 14(3), 95-96.
Nuryani, Y.H. (1984) Effect of planting material in black pepper. P.P. T. Zndustri
(Bogor) 9(50), 8- 12.
Nybe, E.V. and Nair, P.C.S. (1987) Nutrient deficiency symptoms in black pepper.
Agric. Res. J. Kerala 25(1), 52-65; 24(2), 132-150.
Orjala, J. et al. (1993) Derivatives with antimicrobial and molluscicidal activity from
P. aduncunz leaves. Planta Medica 59(6), 546-551.
Parry, J.W. (1962) Spices, Their Morphology, Histology and Chemistry. Chem. Pub.
Co., New York, USA.
Pino, J. (1990) Chemical and sensory properties of black pepper oil (P. nigrum).
Nahrung 34(6), 555-560.
Pohlan, J. and Heynoldt, J. (1986) Investigations on cultivation of pepper in
Vietnam. Beit. Trop. Land. Veter. 24(1), 13-24.
Pruthy, J. S. (1993) Technological innovations in pepper technology; 2-super-critical
fluid extraction. Int. Pepper News Bull. 17(3), 14-24.
Purseglove, J.W. et al. (198 1) Spices 2 vols, Longman, Harlow, UK.
Putnam, G.G. (1924) Salem Vessels and their Voyages. The Essex Institute, Salem,
Massachusetts, USA.
Radhakrishnan, T.C. and Naik, B.J. (1983) Varietal reaction and loss of yield in
pepper due to fungal pollu under coconut shade. Ind. Cocoa, Arecanut Spices J.
6(4), 91-92.
Ravindran, P.N. and Babu, K.N. (1994) Chemotaxonomy of south Indian pepper.
J . Spices Aromatic Crops 3(1), 6-13.
Ravindran, P.N. and Nair, 1M.K. (1984) Pepper varieties. Znd. Cocoa, Arecanut,
Spices?. 7(3), 67-69.
Reddy, B.N., Sivaraman, K. and Savandan, A.K. (1991) High plant density ap-
proach to boost black pepper production. Znd. Cocoa, Arecanut, Spices J. 15(4),
35-36.
Rosengarten, F. (1969) The Book of Spices. Livingstone Pub. Co., Wynnewood,
USA.
Rusli, S. and Nurdjannah, N. (1993) Handling and processing of pepper. Indonesian
Agric. Res. Dev. J. 15(2), 38-43.
Rusli, S. and Soepandi, I.Y. (1981) Effect of raw material weight and distillation
time on oil yield and oil properties. Pemb. Littri. 7(40), 9-15.
Sadanandan, A.K. (1989) Management of wilt affected black pepper. J. Plant. Crops
16 (SUPP), 287-291.
Salvi, B.R. and Desai, A.G. (1989) Effect of application of growth regulators on
maturity and yield of black pepper. J. Plant. Crops. 17(1), 44-49.
Samuel, M.R. and Bavappa, R.V.A. (1981) Chromosome numbers in the genus
Piper. CUP.Sci. 50(4), 197- 198.
Seibert, B. (1988) Living poles for pepper in East Kalimantan (Indonesia). Samar-
inda No. 8, 9-15. Mulanvarman University, Indonesia.
Semler, U. and Gross, G.C. (1988) Distribution of piperine in vegetative parts of P.
nigrum. Phytochem. 27(5), 1566-1567.
Seneviratne, K.G.S. et al. (1985) Influence of shade on rooting and growth of black
pepper. J. Plant. Crops. 13(1), 41-43.
392 Chapter 10
Sim, E.S. (1971) Dry matter production and major nutrient content of black pepper
in Sarawak. Malay Agric. J. 48, 73-93.
Sitepu, D. (1993) Disease management of pepper. Ind. Agric. Res. Dev. J. 15(2),
31-37.
Sumathikutty, M.A. et al. (1984) Analysis of oils from commercial grades of pepper.
Ind. Perfum. 28(314), 119-22.
Sumathikutty, M.A. et al. (1989) Quality criteria of pepper grades from pure
cultivars. Ind. Perfum. 33(2), 147-150.
Tsuda, M. (1 989) The production and diseases of pepper in Brazil. Agric. and Hortic.
64(10), 1159-1166.
Une, M.Y. (1988) Black pepper in Para State, Brazil. B r a d de Geografia 50(2),
75-79.
Verghese, J. (1989) On the isolation of oleoresin black pepper by steam distillation-
cum-solvent extraction and tailoring of oleoresin. Perfum. FZuv. 14(6), 33-38.
Waard, P.W.F. de (1964) Pepper cultivation in Sarawak. World Crops 16(3),
24-30.
Wahid, P. and Zaubin, R. (1993) Crop improvement and cultivation of black
pepper. Indon. Agric. Res Dev. J. 15(2), 27-30.
Watt, G. (1908) The Commercial Products of India. John Murray, London, UK.
Weaver, K.M. et al. (1 988) Liquid chromatographic method for determining piper-
ine in P. nigrum. J. Ass. Off. AnaZyt. Chemists 71(1), 53-55.
Wood, A.B. et al. (1988) Piperine determination in pepper and its oleoresins-a
reverse phase high-performance liquid chromatographic method. Flav. Frag. J.
3(2), 55-64.
Yufdi, M.P. (1986) The effect of topping of plant material on the rooting of pepper
cuttings. P.P. T . Industri (Bogor) 11(3/4), 45-50.
Zhao, J. and Cranston, P.M. (1995) Microbial decontamination of black pepper by
ozone and the effect of the treatment on volatile oil constituents of the spice.
J.Sci.FoodAgric. 6 8 , 11-18.
Rosaceae 11 -
The Rosaceae is a large family of some 115 genera and about 3200 species,
with the genus Rosa L. containing approximately 150 species, naturally
distributed throughout the Northern Hemisphere except the tropics. Roses
can now be found virtually worldwide so popular are they with gardeners,
and for their medicinal value and religious significance (Touw, 1982,
Ueyama et al., 1990, Phillips and Rix, 1993). The Mediterranean R. gallica
L., one of the many species from which modern cultivars are derived, has
been cultivated since antiquity and reached Central Europe in the early
Middle Ages. The double-flowered Provence rose was developed in medie-
val times via R. centifolia L. from the Caucasus. The inaccurately named
Chinese rose, R. indica L. reached Europe in a cultivated form and its
introduction to France is credited to a Mr Parsons who brought four plants
with him from an eastern voyage at the end of the eighteenth century; R.
chinensis Jacq. arrived in London in 1789. A type of R. damascena Mill.
reached Europe via Asia Minor in ancient times, and its modern descen-
dants are the most important essential oil producers. Rosa indica and R.
damascena crosses produced the Bourbon rose on the Indian Ocean island of
Bourbon, now Reunion. Later crosses with the Bourbon rose resulted in the
so-called perpetual varieties with greatly extended flowering. Rosa multiflora
Thunb. was brought to Europe from Japan, and crosses with R. chinensis
produced the polyantha roses. The European name for the Damascus rose
apparently dates from the Crusades, but is now incorporated in its botanical
name. The first use of R. damascena in literature was in the sixteenth century
(Shepherd, 1954), and its history and utilization have been reviewed (Widr-
lechner, 1981; TOUW, 1982).
Today there are thousands of rose varieties with very different types of
growth, degree of thorniness and scent, but with the basic flower colours of
white, yellow and red. Commercial rose breeders, especially in France, have
produced flowers with very different scents to those of traditional cultivars,
including anise, lemon and other citrus, and more musky notes, aimed
directly at the fragrance industry (cf. jasmine).
393
394 Chapter 11
Rose
The history of roses began in the fossil record, expanded with the introduc-
tion of writing and development of printing, and there is a mass of published
material including beautifully illustrated books and informative video
tapes.
Fossils in Oligocene Period rocks from Oregon and Colorado, USA,
indicate that roses occurred in that region 35 million years ago. Roses were
popular among the ruling classes of Mesopotamia, as a cuniform tablet in
the British Museum apparently refers to a rose and rose-water. T o the
Medes and Persians the rose was a religious symbol, in the same region
modern followers of the Sufi sect consider the rose denotes perfection. The
plants and their derivatives were widely admired and used by the Greeks and
Romans. In the 9th century BC, Homer in his Iliad wrote of the body of
Hector being anointed with rose oil before burial. Ovid and Virgil cultivated
roses, Pliny claimed to have ‘improved’ them, and noted bushes should be
pruned before March and hot water applied to induce flowering; he also
described a hundred-petalled rose. Plutarch recommended interplanting
rose plants with garlic to improve their scent! Theophrastus (372-287 BC)
described two types in his Enquiry into Plants, the kynosbaton (dog rose) and
rhodon (cultivated rose), and also noted that ‘sesame oil receives rose
perfume better than other oils’. Herodotus (490-420 BC),gave details of a
60-petalled, strongly scented rose brought to Greece by the exiled King
Midas of Phrygia. The Roman emperor Nero showered his guests with rose
petals, and vast rose gardens were established at Peastum, south of Naples,
to supply the ever-growing demand.
The famous 3500-year-old Blue Bird fresco at Knossos, Crete, showes a
light pink, five petalled rose with three leaflets. The very earliest preserved
rose specimen dating from the beginning of the present era and believed to
be R. sancta (syn R. richardiz] is in the Petrie Museum London, and is from
a tomb he excavated at the Hawara cemetery in the Faiyum. According to
the Islamic prophet Muhammad’s somewhat immodest account ‘When I
was taken up to heaven some of my sweat fell to earth and from it sprang the
rose; who smells the scent of the rose smells Muhammad!’ These are a few
examples from a multiplicity of historical references.
The first modern book on roses was by Mary Lawrence, A Collection of
Roses from Nature, 1799. Probably the most famous illustrations are those in
Pierre-Joseph Redoute’s three volume Les Roses (1 8 17-1 924), which depicts
roses in the collection of the Empress Josephine of France. Many plants
unrelated to the Rosaceae have rose added to their name, including Guelder
rose (a virburnum), Rose apple (a eugenia), Rose Box (cotoneaster), Christ-
mas and Lenten roses (hellebores) etc.
Rosaceae 395
When rose petals were first extracted is unknown, but the earliest written
reference is in the ninth century when production was well established, for
30,000 bottles of attar of rose were delivered annually between AD 810 and
817 as tribute from the Iranian province of Faristan to Maman, the then
Caliph of Baghdad. Rose-water was also traded along the Silk Road to
China, Distillation probably began independently in Europe, the Middle
and Far East. An Italian, Geronimo Rossi of Ravenna is credited as the first
European to do so in 1574, describing his methods and oils produced in De
Distillatione Liber, 1582.
The Chinese Li Shih-Chin in his work the Pen T’sao (1578) noted that
essence of rose is made by distilling the flowers of R. rugosa and he also
mentions essence of R. indica. Distillation began in Persia in 1612 and
production of rose oil quickly became a major industry, with local cultivars
and their processing spreading eastwards and westwards along the Silk
Road. This distribution can be followed in the name for rose oil, attar of
roses; attar from the Turkish is derived from the Farsi, aettr, which then
meant a fat. In Hindi itr described a liquid produced by absorbing in
sandalwood oil a distillate of rose flowers. In modern Turkish the attar is
known as gulyagi from the Farsi gul for rose flower.
Cultivation of R. damascena and production of rose oil began in what was
then the Turkish province of Eastern Roumelia (now Kazanlik) in the
fifteenth century. It was not until 1894 that R. damascena was re-introduced
to Turkey from Kazanlik, and large-scale production recommenced in the
south-western provinces of Burdur and Isparta. The first modern export
record is 1110 kg attar from the then Turkish Empire to England and France
in 1860. Turkey is now the world’s second largest producer of rose oil.
Commercial rose growing to produce oil was well established in what is
now the Kazanlik region of Bulgaria by the end of the seventeenth century,
although the plant had been grown for many centuries prior to its commer-
cial exploitation. So dominant did Bulgarian attar become on world markets
that rose cultivars originating there became universally known as Kazanlik
roses. On-going selection and breeding produced higher oil-yielding culti-
vars with more flowers blooming over a longer period and Bulgaria remains
the leading producer of rose oil (Staikov et al., 1975; Irintchev and Oghnen-
ski, undated). Bulgaria collectivized all rose plantations after the Second
World War, and nationalized the distilling and marketing sectors. Between
1938 and 1948 the area under roses fell from 6300 ha to 2000 ha, and later
ranged between 2000 and 2500 ha, with annual exports of attar and concrete
around 2000 kg each. Since the fall of communism, the Bulgarian rose
plantations have been returned to private ownership. Despite the disruption
this caused, reliable reports estimated total annual production in 1994 and
1995 as the equivalent of 750-1000 kg rose oil.
The Kazanlik cultivars have also been introduced to southern Russia, but
the author was told on a visit that an important local Crimean rose was
396 Chapter 11
Botany
T h e Caninea group are native to Europe, western Asia and North Africa,
and characteristically have an erect, arching habit, medium sized often
greyish-green leaves, stems with numerous hooked thorns and flowers in
small clusters. Garden types include the albas, dog roses, and sweet briars.
Rosa canina L., the common European wild dog rose, is a vigorous shrub to
3 m, with extremely prickly stems. T h e scented, usually pale pink flowers are
solitary or in clusters of two to three, and produce abundant fruit. Also in
this group is R. centifolia L. the cabbage or Provence rose, considered to be
a complex hybrid. It is a medium lax shrub to 2 m, with abundant grey-
green, coarsely-toothed leaves, and stems with several types of thorns. T h e
rich pink, strongly scented flowers are cabbage-shaped, double, singly or in
clusters; summer flowering. It produces no fruit.
The Chinensis group is based on two wild roses from China, R. chinensis
and R. gigantea Coll. Growth is variable but usually upright, the flowers are
borne in small clusters, and flowering is recurrent. Garden types include the
tea roses, noisettes, bourbons, hybrid perpetuals, and hybrid climbing tea
roses.
T h e Gallicanae group is native to Europe, Turkey and Iraq and character-
istically have large leaves composed of five leaflets and flowers solitary or in
threes and fours on long stems. Garden types include cabbage roses,
Rosaceae 397
damasks, mosses and Portlands. R. gallica L. is the group’s major wild rose,
and its widespread occurrence is partially due to its ability to grow on soils of
low fertility. R. gallica is a vigorous spreading shrub to 1.5 my with dense
foliage and few thorns. The fragrant, semi-double flowers are light red with
prominent yellow stamens, and major flowering is in late summer. It is
considered to be the oldest Gallica species in cultivation.
Rosa damascena Mill. (main syn. R. bifera Pers.; R. calendarum Borkh) is
commonly known as rose in English and many other languages. The
evolution of R. damascena is obscure and one of the first major rose
taxonomists (Lindley, 1820) stated ‘Our knowledge of European roses has
become by the extraordinary attention they have received, so extensive that
it is impossible to doubt that limits between what are called species do not
exist.’ A situation somewhat less confused today but still unresolved. The
basic chromosome number of the Rosaceae is x = 7 , for Rosa 2n = 14, but
most members of the Rosa are tetraploids with some interesting exceptions;
a diploid R. damascena and a triploid R. centifolia,while a pentaploid from an
Iranian cultivar has been found (2n = 3 5 ) .
It is generally agreed that R. damascena should be included in the genus
Rosa (Gallicanea), together with the polymorphic species R. gallica and its
closely related forms which may be allied species or hybrids; the relationship
between R. damascena and R. gallica is extremely involved. It is probable that
R. damascena cultivars have a major input from R. gallica, R. phoenicia Boiss.
and other species now lost.
The emergence of modern R. damascena cultivars resulted initially from
selection over a very long period, plus cross-breeding with locally available
species. In Turkey, for example, there are 30-40 Rosa spp. and the Anatolian
region is postulated as the centre of origin of R. damascena, although it has
not been collected as a truly wild plant (Nilsson, 1972). A proposed
genealogy of the various major Rosa spp. is shown in Fig. 11.1. The modern
essential oil cultivars, the Kazanlik roses, were subjected to an extensive
breeding programme and designated R. damascena forma trigintipetala
Dieck. Similarities between the composition of their flower wax and wax
from R. gallica and R. centifolia suggest these could be closely involved in
development of the Kazanlik rose.
Commercially cultivated R. damascena is a perennial bush rose, much
branched, thorny, with large, highly scented, roseate flowers. Although
other plant parts are aromatic, only the flowers are normally used to produce
an essential oil. Mature plants have well-developed and extensive roots, and
rate of root growth and extent is governed mainly by soil type or depth of
cultivation. Alluvial meadow soils in Kazanlik, for instance, allow plants to
produce a root system twice as extensive as on plants growing in adjacent
alkaline brown forest soils. Multiple stems arise from the root crown, and in
Bulgaria are topped when four to six leaves have developed to ensure
maximum branching. Such plants become very bushy, 1.2-2.0 m, and form
398 Chapter 11
( 1 600 BC)
R. damascena
R. damascena R. centifolia
R. alba
R. canine
(600 BC) -
R. chinensis
Bourbon rose
I
(1700AD)
7
R. moschata -+rose
Modern rose
ethephon increased flowering and also flower oil content (Farooqi and
Sharma, 1990). Reports of trials elsewhere with these chemicals and also
maleic hydrazide produced conflicting data and there may be a cultivar
interaction. Gamma irradiating R. damascena stem cuttings in India did not
stimulate increased flowering or oil content, although both improved in a R.
borboniana cultivar (Lata and Gupta, 1971).
Flower-bearing stems form in spring from adventitious buds in the crown
of annual branches. Weather conditions at this time are very important for
bud formation, with the optimum nightiday temperatures 3-4°C and 15-
16°C respectively; colder nights or warmer days adversely affect bud forma-
tion (Topalov, 1964); the author also discusses bud organogenesis in detail.
A period of winter dormancy is essential to induce bud development, as was
demonstrated at the Kazanlik Research Institute. Plants kept in a green-
house continued to make vegetative growth throughout the winter, but
failed to flower the following year. Flowers of the most favoured cultivars are
large, rosy pink to rosy red, possessing 20-30 normal, 5-10 deformed petals,
and 100-120 stamens. Flowers are very fragrant and the scent varies in
strength and odour characteristics between cultivars. An attempt to differ-
entiate and thus typify the various flower scents has been proposed (Naka-
mura, 1987), and the seven types chosen are shown in Table 11.1.
Anthesis is normally completed in 20-25 days, and is directly affected by
climatic conditions immediately prior to and during flowering, with an
optimum air temperature of 1 5 2 0 ° C and relative humidity above 60%.
High temperature or hot dry wind can induce premature bud opening, and
reduce the flowering period and essential oil content of flowers. Irradiating
pollen from Rosa spp. produced a general increase in fertility, but the dosage
at which this occurred and the level of increase varied between species
(Tsvetkov, 1978). All flower parts contain essential oil which generally
reaches its maximum when the petals become cup-shaped and the stamens
bright yellow; designated stage five of the six flower development stages (Fig.
11.2). The rate of oil synthesis varies and is highest in crown petals. Whole
1 2 3 4 5 6
Fig. 11.2. Stages in rose flower development.
flowers are normally picked at stage five for rose oil, but young buds are
selected for highest yield of concrete.
The essential oil content of flowers varies with the cultivar, as does the
weight of flowers harvested per hectare. Flower yield ranges from
1000-3500 kgiha, and 2000-4000 kg are required to produce 1 kg oil. The
main oil constituent is geraniol to 80%, but the ratio of three main terpene
alcohols, citronellol to geraniol and nerol (the C:GN ratio), is of major
importance in determining rose oil quality. The effect of various growth
stimulants on flower production has been noted; their effect on oil content
and characteristics has produced conflicting results, with a particular culti-
var grown in different countries reacting differently to the same chemical or
level of application. While the small amount of data indicates that most
chemicals adversely affect oil content, the overall level is not significant and
their use is of little practical value (Farooqui et aZ., 1993). The main
characteristics of rose oil are shown in Table 11.2.
The fruit is light to dark red, fleshy, elongated, ovoid, (commonly rose-
hips in English), which originates from the cup-like receptacle bearing the
sepals, petals and stamens on its rim, and enclosing the carpels. Carpels are
attached to the base but otherwise free. The majority of fruit is shed before
ripening. The fruit can be processed to make a syrup, and fruit in some rose
species has a very high vitamin C content, particularly R. rugosa cultivars. A
tonic is made by steeping fruit in boiling water for 5 minutes, and drunk hot
to alleviate the effects of a cold. The fruit contains one to three brownish
seeds (nutlets), rounded-triangular, 3-5 mm depending on cultivar. Seeds
are viable, but seldom used for propagation.
Ecology
Rosa damascena is adaptable to a wide range of environments and it is
basically non-climatic factors which have limited its commercial cultivation
for oil to very specific regions, as discussed later. A mild climate with no
extremes of temperature, long periods of warm, sunny weather and a regular
well-distributed rainfall is preferred. Within a region there may be consider-
able annual variation in the weather pattern and this directly affects the
flower yield; over 8 years in the Kazanlik region the average annual flower
yield was 1130-2700 kg/ha, with an overall average of 1700 kgiha (Lozzi,
1970).
The most important rose-growing areas are in Bulgaria, southern Russia,
Turkey and Morocco, where the environment is similar. Since Bulgaria is
the most successful, a description of the climate in the Kazanlik region gives
an excellent guide to the requirements necessary for commercial rose
growing, with temperature and humidity the most important factors influ-
encing flowering. It is locally considered that the lengthy period in spring
when mean temperature is 5 1 5 ° C has a decisive effect on flower produc-
tion, ensuring large numbers of high oil content flowers per plant. A low
night temperature of 10-12°C during flowering inhibits, but up to 2OoC
increases oil synthesis. Regular rainfall also ensures plants do not suffer
moisture stress, especially in spring and early summer when the oil content
of flowers is determined. During this period, May- June, the daily humidity
averages 70% which also contributes to oil accumulation; however, rain
during main blooming damages flowers and reduces oil content.
Long sunny periods encourage flowering and over 20 years in the Ka-
zanlik region sunshine averaged 2300 hours annually, but hot dry periods
during flowering rapidly reduce oil content. Shade is inimical to flowering
and the Kazanlik plantations are free of trees but in India, where the
temperature in direct sun is much higher, shade is necessary. At Kanpur,
Uttar Pradesh, oil yield from flowers grown in direct sunlight was half that of
flowers grown in light shade or with artificially increased humidity. Strong
winds are physically destructive to flowers, and hot, dry or cold winds
adversely affect oil content. These winds are absent from the Kazanlik
Valley, naturally protected from the very cold northern winds by the Balkan
hills. Rose flourishes below 1000 m and most commercial plantations are
402 Chapter 11
Cultivation
The main operation prior to establishing a rose plantation is deep ploughing,
not only to fracture a pan or compacted layer which would restrict root
growth, but to expose plant debris and similar material, which must be
collected and burnt. Plantations are normally established using cuttings or
404 Chapter 11
splits obtained from mature rose plants, but grafting or budding are simple
procedures widely practised whenever roses are grown. Micropropagation to
produce large numbers of plants from a specific parent has yet to be
introduced on a commercial scale. Research on suitable methods and
techniques continues (Kirichenko et al., 1993: Kornova and Michailova,
1994).
Splits are obtained by dividing mature rose plants, often from exhausted
plantations, leaving each split with a section of root attached; in Bulgaria 1
ha produces sufficient splits to plant 2 ha, often placed directly in prepared
trenches where they are to grow. Suitable cuttings from mature plants taken
in the autumn in Bulgaria are also laid in trenches in a special nursery and
covered. Shoots appear the following spring and plants remain in the nursery
for 18-24 months. Cuttings can also be directly rooted in nursery beds,
which must be specially prepared, fertilized and preferably fumigated.
Cuttings 20-30 cm long are obtained from mature plants during pruning
but, for more intensive production in special nurseries or greenhouses, two-
node cuttings taken when shoots begin to lignify are most suitable, and only
lower leaves removed (Ivanova and Gladun, 1986; Decheva et al., 1989).
Cuttings should be treated with a rooting compound, then inserted in the
bed up to half their length and under good management can be planted out
after 12 months; in India, a bed 20 X 5 m will accommodate 10,000 cuttings
sufficient for 1 ha.
Transplanting can be either manual or mechanical, and the method and
equipment depend on the type of planting material. Planting can be fully
automated, with cuttings prepared to suit the transplanter used. Splits or
divided plants are more difficult and this material is usually planted by semi-
manual equipment; but where labour is freely available or cheap, manual
planting remains common. Pre-trenching or digging individual planting
holes to receive cuttings or splits is still practised, but full automation is
inevitable on large scale commercial plantations, and may then require a
complete revision of the local system of preparing planting material. Hedge
planting remains the most popular, with rows 2.0-2.5 m apart, 1.O-1.5 m
between plants. Spacing basically varies according to soil type and cultivar
and closer rather than wider in-row spacing is preferred to ensure an
unbroken hedge to achieve maximum flower production.
Young bushes are pruned annually in early summer to promote uniform
branching, and in the first and sometimes the second year all flower buds are
removed. In the third year plants are allowed to flower and regular picking
commences. Weed control is essential and may be manual or mechanical,
but should be as shallow as possible to reduce root damage. Black plastic
sheets may also be used in new plantations to retain soil moisture and inhibit
weed growth. Herbicides can be used and the following compounds have
been applied in rose plantations, usually with no effect on flowers or oil:
atrazine, dalapon, glyphosate, propanil, and simazine. Some herbicides may
Rosaceae 405
adversely affect oil yield and composition but, as many factors are in-
volved, only local testing can determine whether a particular chemical, or
combination, is suitable.
Irrigation may be required to ensure sufficient soil moisture and thus
growth in the important bud formation and early flowering period, and to
maintain humidity. Over-watering should be avoided especially when plants
are surface irrigated. Maintaining soil moisture at 85% capacity during the
growing period produced highest flower yield in Russia, 5360 kglha com-
pared to 3750 kg/ha without irrigation (Korshunov and Umanets, 1984).
Sprinklers were more efficient in water use, gave a higher flower yield per
unit of water applied and did not increase pest and disease damage in both
Russia and Bulgaria. Regular irrigation is necessary in January-March in
northern India and considered essential after pruning. In Tunisia and
Morocco, where rain may be sufficient in total but erratic in timing,
irrigation is usually essential; in Egypt, roses are grown in the Delta wholly
under surface irrigation. A nitrogenous top-dressing should precede irriga-
tion or coincide with rainfall. A more general interaction between fertilizer
application and irrigation may occur; over several years in Bulgaria irrigation
increased flower yield by 4-5 t/ha (40%) at the same level of fertilizer
compared with no irrigation (Staikov et al., 1975).
Annual pruning is essential to keep bushes in shape and promote branch-
ing, and seasonal pruning to remove dead or diseased shoots. Older plants
may be cut-back severely to promote new and vigorous growth; this opera-
tion rejuvenates bushes, ensures continued high flower production and is
normally necessary once in a plantation’s life. Selective pruning is usually
manual, but overall annual or rejuvenation pruning is more often mechanical
and a range of suitable machines is available. The extent of pruning depends
on the system of management, region and cultivar. Pruning to 15 cm from
ground level in India gave higher flower yield than moderate or light pruning
(Singh and Ram, 1987). Light to moderate pruning in the Bulgarian spring
(early March) before growth commenced increased flower yield and flower
oil content, but heavy pruning reduced both; pruning once every three years
was acceptable in terms of cost and flower yield (Aztadzhov et al. , 1986).
Prunings should be burnt and the ash spread as fertilizer, or used as distillery
fuel. Roses are seldom underplanted or intercropped, since few crops are
compatible with commercial rose oil production (Tajuddin et al. , 1993). The
economic life of a plantation is 10-12 years, with flower yields per hectare
normally increasing in the first 5-7 years, then falling gradually until the
plantation becomes unprofitable. Judicious pruning and fertilizing can thus
maintain flower production and extend plantation life.
The wide variation in flower oil content between cultivars is reflected in
the average oil yield per hectare from different regions. Considerable varia-
tion in oil content and composition between individual bushes within a
cultivar can occur, although this may not be very obvious in areas where
406 Chapter 11
Harvesting
Flowers are generally picked manually and it was this operation which
determined the feasibility of establishing new rose plantations; if there was
insufficient labour available at this season the project was non-viable (Fig.
11.3). Hand-held mechanical pickers could be more widely used since the
slight damage to flowers is unimportant provided they are promptly distilled.
Any loss of oil would be offset by the substantial reduction in harvesting cost.
Picking should be between 5 and 10 a.m. as it is during this period that the
flowers open and the oil content is highest; in Italy the optimum time was
8.30-9.00 a.m. and oil content declined thereafter. It was also highest in the
middle of the picking season at 1 kg oil per 3000 kg flowers, compared to 1
kg from 4000 kg at beginning of the season (Garner0 and Buil, 1976).
Whole, fully open flowers should be selected, not buds or the previous day’s
flowers, i.e. only those flowers in stage five (Fig.11.2). An experienced
worker can pick 25 kg flowers in 4 hours.
Flower yield varies widely and generally indicates the standard of manage-
ment as there is a direct correlation between the two; a high level of
husbandry producing high flower and oil yield. Flower yield in Bulgaria is
2000-3000 kg/ha; in Turkey, 2000-2500 kgiha; in Russia, 1500-2000 kgiha;
in India, 1000-1500 kgiha. Yields however may not be directly comparable
since different cultivars are involved; in Morocco, R. centifolia averages 2500
kg/ha; in Egypt, 2500-4000 kg/ha from R. gallica.
Oil composition varies substantially over the flower stages shown in Fig.
11.2 due to differences in rate of synthesis of the various components. In the
first three stages, flowers contain more stearoptene, alpha-pinene, myrcene
and less of the important alcohols citronellol, geraniol and nerol. In sub-
sequent stages, the alcohol content rapidly increases to around 60%; oil
content increases to stage five and then rapidly decreases. Pathways involved
in essential oil production during flower development have been discussed
(Pogorel’skaya et al., 1980). The oil content and ratio of constituents in the
various flower parts differs; in Kazanlik cultivars it is highest in stamens and
internal petals. The ratio between oil components is shown in Table 11.3.
Rosaceae 407
(Garner0 and Buill, 1976). Thus the stage at which flowers are picked has a
major influence not only on oil yield but also oil characteristics. Since flower
petals of R. damascena var. sempe$orens have similar characteristics, these
data are probably relevant to all oil cultivars.
Flowers should be transported to the distillery without delay, but storing
bagged flowers in clean cold water retains oil content and quality for up to 3
days. Spreading flowers for short periods in layers in cool shade retains most
of the oil without loss of quality, but storage beyond 24 hours causes
significant loss. Temperature above 25°C during storage can decrease
flower oil content by half, increase citronellol and stearoptene but decrease
nerol and geraniol content; temperatures of 4-1 0°C increased oil content,
stearoptene decreased, but nerol and geraniol increased (Staikov et al.,
1974).Petal oil content increases after picking due to enzymatic activity, and
also during storage under anaerobic conditions using various gases (Tyu-
tyunnik and Ponamaryava, 1977).
Distillation
Rose oil is produced by a two-stage distillation of whole flowers; the initial
stage producing decant or first oil, then redistilling the water to obtain the
second oil. Stills vary in construction and method of operation, and their
products differ considerably in composition, quality and yield of oil from the
same weight of flowers. Thus 1 kg oil is obtained from approximately 300 kg
of flowers in Bulgaria, 380 kg in Russia, 350 kg in India, 300-350 kg in
Turkey. From R. gallica in Egypt, 1000 kg flowers yield 2.2 kg concrete and
52% absolute; from R. centijolia in Morocco 1000 kg flowers yielded 1.8 kg
concrete and 50% absolute, while extraction of a Bulgarian R. canina
cultivar gave a concrete yield from flowers of 0.51% with a 50% absolute
content. Much higher oil yields are obtained in the laboratory from selected
bushes, indicating the potential for a considerably higher oil yield when
modern processing plants are used and high-yielding cultivars planted.
Following distillation, the oil should be placed in full, tightly closed,
opaque containers and kept in cool storage. Oil exposed to air and light
quickly deteriorates. Rose-water is more important than rose oil in India,
and distillate is not returned to the still but sold directly as rose-water.
Depending on the ratio between weight of flowers and water, distillate is sold
as 20140 or 20/80, i.e. 20 litres of distillate from 40 kg flowers, or 20 litres
from 80 kg. Spent flowers can be solvent extracted, dried and used as an
ingredient in incense, confectionery, or returned to plantations as mulch,
but are often discarded.
The gross factors affecting yield of flowers and thus oil have already been
discussed. Improved methods of distillation can also substantially increase
oil yield, and are described in the relevant literature. Probably the most
important agronomic factor which directly influences yield and oil type is the
stage of flower development at picking, and the various stages of flower
Rosaceae 409
Root Rhizococcus spp., root mealy bugs; Maculolachnus spp., root apids;
Diplolepsis spp., root gall wasps; Paria spp., rootworms; Cicadidae, cicadas;
Pratylenchus spp., nematodes; Pantomorus spp., beetles.
Stems and branches Agrilus spp. and Oberea spp., stem borers and girders; Stictocephala spp.,
treehoppers; Paris spp. (adults);Aulacaspis spp., scales; Syrista spp., stem
sawflies.
Leaves Argyrotaenia spp,, Macrodactylus spp., Sibine spp., Parasa spp. Archips spp.
and Orgyia spp., larvae feed on leaves; Endelomyia spp., rose slugs;
Edwardsiana spp. and Typhlocyba spp., leaf hoppers; Maculolachnus spp.
and Macrosiphum spp., aphids; Dasineura spp. and Wachtliella spp.,
midges: Bulgarialeurodes spp. and Trialeurodes spp., whiteflies;
Blennocampa spp. and Arge spp., sawflies; Tetranychus spp., spider mites.
Flowers Rhynchites spp., cerculios; Macrodactyius spp.and Cetonia spp., chafer
beetles: Nodonota spp., rose beetles; Forficula spp., earwigs; Dasineura
spp., midges; Thrips spp., thrips; Arge spp., sawflies.
India, China, Morocco and previously France, differs from Bulgarian and
Turkish otto, and is normally sold as rose oil followed by the country of
origin, i.e. rose oil Morocco. Rose oil is also produced in Morocco and
France from R. centifolia, in Egypt from R. gallica var. aegyptiaca, in China
from R. rugosa and other Rosa spp. These oils are also sold as rose otto, but
experienced perfumers can readily identify them. Only Moroccan oil is
normally available on the world market in commercial quantities, although
Chinese oil is becoming more frequently available, is cheap, and will
probably increase to meet a growing demand. A comparison of a Chinese
rose oil with oils from other origins has been published (Ohno and Tanaka,
1977).
Because of its importance to formulators and its relatively high price, rose
oil has been exhaustively studied. There is thus an extensive literature on its
use, characteristics and composition. A review of the oil and absolutes to
1976 has been published (Garnero, 1976). Analyses have identified some
150 constituents, many in minute proportions. T o the perfumer, most
important are the aromatic terpene alcohols, citronellol, geraniol and nerol,
but other minor constituents are important modifiers (Omata et al., 1991),
and a reason why synthetic otto is seldom equal to the natural product. An
interesting discussion on which constituents contribute to the characteristic
rose odour has been published (Buccellato, 1980).
The oil is expensive and commercial samples are frequently adulterated
with an extensive range of similar natural compounds or synthetics. Most
alter either the oil’s natural odour or persistence. Rose oil is virtually
indispensable in perfumery, cosmetics and similar products either as a
primary scent or modifier, since it blends with most other floral oils and can
be used in extreme dilution. The oil has a slightly bitter, very sharp taste,
most pleasant when diluted. Its main use in flavourings is to modify a
primary taste, and for this reason it is used in a very wide range of sweets,
confectionery and tobacco products. The oil has a minor use in medicine
and has insecticidal and fungicidal properties.
Rose oil (otto of rose) Bulgarian is a pale yellow to yellowish green; fresh
oil is often more green due to the presence of azulenes which gradually
decompose over time; when cooled below 20°C white or colourless crystals,
stearoptene, are deposited; the liquid is elaeoptene. Stearoptene is odourless
and constitutes 15-25% of the oil. On cooling below 16°C the oil may
solidify into a translucent mass with surface crystals. The odour is warm,
deep-floral, slightly spicy, very rich with traces of honey, tenacious and
intense. Oil from Japanese tea roses is humid, fresh green, flowery, slightly
phenolic and spicy, soft and refreshing rather than sweet. Rose oil is
considered not to reflect the true scent of rose flowers since certain aroma
components, in particular phenylethyl alcohol, are lost during distillation.
T o correct this loss it has been suggested that rose flowers should be distilled
in water saturated with the alcohol.
412 Chapter 11
High quality oil should have a specific gravity approaching 0.86, refractive
index 1.46, optical rotation -3" to -4O, congealing point 18-22 " C , and
total alcohol content of 66-70%. Rose oil is very slightly soluble in water,
sparingly so in alcohol, but soluble in fatty oil and chloroform. Turkish otto,
Anatolian rose oil, is similar in characteristics and odour to Bulgarian oil, for
which it is generally a direct substitute (Anac, 1984). A comparison of the
main constituents of Bulgaria and Turkish rose oils is shown in Table
11.5.
Rose oil (otto of rose) Morocco obtained from R. centifolia is colourless to
pale yellow, sometimes with a greenish tinge when fresh; odour deep, sweet,
warm, rich, but less spicy than Bulgarian or Turkish oils, tenacious. The
main uses are similar to these oils to which it is often added as an extender,
but it is not a substitute. Rosa centifolia oils from France and Spain resemble
the Moroccan oil, but that from India differs. None are regularly available.
Rose concretes are generally similar in appearance, but differ in their
odour to reflect the scent of the original flowers, and approach more closely
the true flower scent than oil. The method of extraction also affects the
important odour constituents (Moates and Roverts, 1991). T o avoid un-
necessary descriptions, concretes and their absolutes will be discussed in
general. Rose concretes are waxy solids, dark yellow to dark brownish
orange, melting point 45-55"C, and are used in similar products to oils, but
mainly extracted to produce absolutes. The most easily obtainable are rose
concrete centifolia, rose concrete damascena from various origins but
mainly from Morocco, Bulgaria and Turkey (Anac, 1984), and rose con-
crete gallica from Egypt (Nofal et al., 1982).
Rose absolutes are orange-yellow to orange-brown, sometimes with a
reddish or greenish tinge, viscous liquids, and tend to have a richer odour
than concretes (Tables 11.6 and 11.7). Because absolutes are so concen-
trated they are normally used at very high dilution rates in end-products;
main uses are the same as for concretes and oils. Absolutes are becoming
increasingly important since their production in the country of origin adds
value to the local rose industry, and usage is expected to continue to rise as
manufacturers adopt the material. Similarly to concretes, absolutes are
identified as rose absolute centifolia, etc. However, absolute from Morocco
Moroccoa Turkeyb
Ethanol 0.38 ng
Camphene t heptanal 0.01 0.02(camphene only)
Beta-pinene 0.03 ng
Linalool 0.40 0.28(t rose oxides)
Phenylethyl alcohol t decanal
t terpinen-4-01 74.06 72.28
Citronellol 8.77
Nerol 2.52
Citronellol t nerol 14.04
Geraniol t neral 5.18 4.21
Citronellyl acetate 0.23 0.10
Neryl acetate 0.10 0.17(t cinnamaldehyde)
Geranyl acetate 0.38 0.15
Eugenal t trans-beta-damascenone 0.89 0.53
trans-trans-Farnesol 1.30 0.20
Unidentified 5.70 ng
ng, not identified under compound name.
',
Source: ', Buccellato 1980; Anac (1984).
414 Chapter 11
French Bulgarian
(R. centifoliai (R. darnascena)
Colour Orange-brown Greenish-brown
Specific gravity (15°C) 0.960-0.990 0.970-0.990
Refractive index (20") 1.490-1.520 1.500-1.515
Acid index < 20 5-1 1
Ester index 15-20 20-30
Source: Commercial Standard Specifications.
References
Anac, 0. (1984) Gas chromatographic analysis of Turkish rose oil, absolute and
concrete. Perjiun. Flav. 9(1), 2-4, 6-7, 10-14.
Astadzhov, N., Kamburova, K. and Peshov, I. (1986) Changes in flower and
essential oil productivity of the Kazanlik rose caused by single pruning of shoots.
Rust. Nauki 23(12), 63-68.
Bayrak, A. and Akgul, A. (1994) Volatile oil composition of Turkish rose (Rosa
damascena). J. Sci. Food. Agric. 64, 441-448.
Buccellato, F . (1980) An anatomy of rose. Pe$um. Flav. 5(2), 29-32.
Decheva, R. et al. (1989) T h e effect of leaves on rooting of softwood cuttings of
Kazanlik rose. Rust. Nauki 26(2), 36-41.
Dimov, N. and Tsoutsoulova, A. (1987) Pattern recognition methods for discrim-
ination of essential oils (rose oils) by their gas chromatograms. Pe@m. Flav.
12(6), 45-50.
Farooqui, A.H.A. and Sharmia, S. (1990) Effect of growth retardants on flowering of
R. damascena. In: Proc. Znt. Cong. Plt. Physiol. Vol. 2, pp. 1369-1372.
Rosaceae 415
Farooqui, A.H.A. et al. (1993) T h e effect of kinetin on flower and oil production in
R. damascena. 3. Ess. Oil Res. 5(3), 305-309.
Garnero, J. (1976) Chemical composition of the essential oils and absolutes of R.
damascena. Cosmet. Aerosol. 58(10), 548-552.
Garnero, J. and Buil, P. (1 976) Evolution of the composition of rose essential oils
and concrete during a production campaign. Riv. Ital. Ess. Prof. Peante OfJic.
Aromic, Sap. Aerosol. 58(10), 537-540.
Irintchev, I. and Oghnenski, G. (undated) Histoire de la Culture et de l’lndustrie de la
Rose en Bulgarie. Bulgaria.
Ivanova, Z . Ya and Gladun, 1hI.1. (1986) Propagation of essential oil rose. Byull.
Gosod. Nikit. Botan. Sada. 60, 72-76.
Kirinchenko, E.B. et al. (1993) Features of organogenesis in essential oil roses during
clonal micropropagation. Byull. Glavnogo Botan. Sada 167, 96-1 02.
Kornova, K.M. and Michailova, J. (1994) Study of the in vitro rooting of Kazanlik
oil-bearing rose (R. damascena). 3. Ess. Oil Res. 6(5), 485-492.
Korshunov, V.A. and Umanets, E.A. (1984) Productivity of irrigated essential oil
rose. Maslich. Kul. 3, 40-42.
Koseva, D. et al. (1 978) Effect of fertilizers on the utilization and removal of N P K by
Kazanlik rose. Rust. N a u k i . 15(3), 107-118.
Lata, P. and Gupta, N. (1971) Effects of gamma rays on stem cuttings of essential
oil-bearing Rosa species. Flavour Inc. 2(7), 421-425.
Lindley, J. (1820) Rosarum Monographia-A Botanical History of Roses. J. Ridgeway,
London.
Lozzi, C . (1970) Roses and otto of roses in Bulgaria. Flavour Ind. 1, 797-798.
Ma, Z., Hou, S. and Chang, Y . (1985) Effects of ethephon and gibberellin on
flowering period and flower yield of Kushu rose. Acta Hort. Sinica 12(2),
125-130.
Mihailova, J., Atanasouva, R. and Balinova-Tsvetkova, A. (1977) Direct gas chro-
matographs of the essential oil in separate flower parts of the flower of Kazanlik
rose. In Proc. 7th Int. Cong. Ess. Oils, Kyoto, Japan, pp. 219-221.
Moates, G.K. and Roverts, J. (1991). Comparison of rose extracts produced by
different extraction methods. J. Ess. Oil. Res. 3(5), 289-294.
Nakamura, S. (1987) Scent and component analysis of the hybrid tea rose. Perj%nz.
and Fluu. 12(4), 43-45.
Nazarenko, L. (1 982) Development of reproductive shoots in essential oil rose.
Byull. Glav. Botan. Sada. 126, 62-63.
Nikolov, N. et al. (1977) Bulgarian rose oil and its quality. In: Proc. 8th Inst. Cong.
Ess. Oils.Kyoto, Japan, pp. 159-163.
Nilsson, 0. (1972) Rosa. In P.H. Davis (ed.) Flora of Turkey and the Aegean Islands,
Vol. 4, University Press, Edinburgh, UK, pp. 104-128.
Nofal, M A . , Ho, C . T . and Chang, S.S. (1982) Major volatile constituents of
Egyptian rose absolute. Pe$um. Flav. 7(4), 23-26.
Ohno, Y. and Tanaka, S. (1977) O n the constituents of a Chinese rose oil. Agric.
Biol. Chem. 41(2), 399-401.
Omata, A. et al. (1991) New sulphur components of rose oil. Flav. Frag. 3. 6(2),
149-152.
Orlova, L . M . (1 984) Principles of essential-oil rose nutrition determined by leaf
analysis. Re$ Zhur. (1985) 5(55), 569.
416 Chapter 11
Phillips, R. and Rix, M. (1993) The Quest for the Rose, BBC Worldwide Publishing,
London, UK.
Pogorel’skaya, A.N. et al. (1980) Physiological aspects of essential oil accumulation
in rose petals. Fiziol. Rastenii. 27(2), 356-362.
Sharma, M.L. (1982) Cultivation of rose and manufacture of its products in India. In
Cultivation and Utilization of Aromatic Plants. CSIR, Jammu-Tawi, India.
Shepherd, R.E. (1954) Histoy ofthe Rose. Macmillan, New York, USA.
Singh, D.V. and Ram, M . (1987) Effect of spacing, pruning and nutrients on yield of
oil bearing rose in sub-tropical India. Acta Hort. 208, 83-86.
Staikov, V. et al. (1974) Rose flower storage conditions, and influence on the
quantity and quality of rose oil. In Proc. 6th Int. Congress Ess. Oils. San Francisco,
USA.
Staikov, V.M. et al. (1975) Bulgaria: County ofRoses andEssentia1 Oils. Ag. Acad. ‘G.
Dimitrov’, Kazanlik, Bulgaria.
Sudha, S. and Koshy, P.K. (1987) Nematodes associated with roses. Tech. DOC.137.
Asia and Pacific Plt. Prot. Comm., FAO, Rome.
Tajuddin, S.M.L., Yaseen, M. and Hussain, A. (1993) Productivity of rose (R.
damascena) with intercrops under temperate conditions. J. Ess. Oil Res. 5(2),
191-198.
Topalov, V. (1 964) Studies sur la formation des fleurs de la rose de Kazanlik. In Proc.
3rd Int. Ess. Oil Conf. Plovdiv, Bulgaria, pp. 246-252.
TOUW, M. (1982) Roses in the Middle Ages. Econ. Bot. 36(1), 71-83.
Tsvetkov, R. (1978) Radiosensitivity ofpollen from essential oil bearing roses. Genet.
i. Selek. 11(4), 265-270.
Tyutyunnik, V.I. and Ponamaryova, N.G. (1977) Change of essential oil content
and quality depending on conditions of storage. In Proc. 7th Int. Cong. Ess. Oils.
Kyoto, Japan, pp. 2221-2223.
Ueyama, Y. et al. (1990) Volatile constituents of Shi Mei (R. davurica) flower
concrete from China. Flav. Frag. J. 5(2), 115-120.
Velioglu, Y.S. and Mazza, G. (1991) Characteristics of flavones in petals of R.
damascena, by HPLC and spectral analysis. J. Agric. Food Chew. 39(3),
453-455.
Widrelchner, M.P. (198 1) History and utilization of R. damascena. Econ. Bot. 35( l),
42- 58.
Rutaceae 12-
The Rutaceae is a large family containing 130 genera in seven subfamilies,
with the Aurantioideae important fruit and oil producers. Within the Aur-
antioideae, the Citreae contains Citrus (1 6 spp.), Fortunella (four spp.) and
Poncirus (one sp.). Fortunella spp. are native to southern China and have
prolonged cold-weather dormancy to which they probably owe their exist-
ence, and are classified on essential oil composition (Katayama et al., 1994).
Poncirus trifoliata L., native to north and central China, is cold resistant and
deciduous, The remaining genera are Microcitrus, Eremocitrus and Clymenia,
the last considered the most primitive. Citrus, derived from Latin, was
originally applied to Cedrus and apparently transferred to citron probably in
the first century AD, reportedly owing to confusion between the woods
known as lignurn citreunt and lignum cedreum. Citrus consists o f two sub-
genera, Papeda, containing six spp. whose very acid fruit is unpalatable, and
Eucitrus of ten spp., eight cultivated, several on so large a commercial scale as
to be industries in their own right (Jatosti and Spina, 1992). Certain
members of the Rutaceae produce oil similar to sandalwood, and are
included in the chapter santalaceae for convenience.
The taxonomy of Citrus L. is confused, exacerbated by the proliferation of
hybrids and numberless cultivars, but chemotaxonomy could assist classifi-
cation and result in further reorganization (Ulubelde and Mendilicioglu,
1985; Ortiz and Marcide, 1986; Mizuno et al., 1991; Zhong and Ye, 1993;
Fang et al., 1993; Sawamura et al., 1994). The basic chromosome number is
x = 9 (Swingle, 1967; Simmonds, 1976). Most species are diploid, although
a few tetraploids occur naturally and many others produced experimentally.
Interspecific hybrids are common, including tangor, a mandarin-sweet
orange cross; intergeneric hybrids can be induced, while mutation has
produced valuable cultivars including Washington navel orange, Shamouti
orange and Marsh grapefruit. Periclinal chimaeras occur; one of the most
interesting is the grapefruit cultivar Thompson pink. Many citrus species are
rare or little known, and it is possible there exists in these species genes or
417
418 Chapter 12
peel, pulp etc.; additionally the major named cultivars are so well known it is
considered unnecessary to place them in inverted commas.
Botany
All the citrus species discussed are evergreen shrubs or small trees, with
extensive but frequently shallow roots, normally a single stem, very hard
wood, and extensive branching. The flowers are fragrant, the leaves aro-
matic and the fruit frequently palatable. The leaves are alternate, apparently
simple but derived from an original pinnate leaf by reduction to a single
leaflet. The petioles are large-winged, usually articulated with the leaf blade.
Single spines occur in the leaf axils on younger branches and twiglets. The
leaves are gland dotted, and from some species an essential oil is obtained.
Leaves may remain on the tree for 2-3 years, and a healthy full-grown tree
has 150,000-200,000 leaves, with a total area of about 200 mZ;a typical leaf
is shown in Fig. 12.1.
The usually white, sometimes pink or purple-tinged, fragrant flowers are
carried singly in leaf axils or in small clusters on short racemes. The flowers
are generally bisexual and insect pollinated. A typical flower is shown in Fig.
12.1. The flowers usually have five petals, subtended by five small greenish-
coloured sepals. The perianth segments are variable in number. The sta-
mens always more than four times the number of petals and collected into
five groups, the filaments of each group being partially united at the base.
The ovary is superior, composed of eight to 15 united carpels, each loculus
containing several ovules carried in two rows in axil placentation. The ovary
is surmounted by a short deciduous style.
Citrus and other subtropical evergreen trees differ from deciduous trees in
the time of flower induction and period to complete flower differentiation
and development, with temperature being the main factor. In the subtropics,
where the temperature may fall below 15°C for a period (winter), citrus
tends to flower profusely in the local spring and set only one major crop. In
tropical climates with no cool season, flowering may be induced by a
420 Chapter 12
/ Y W E J-- -Articulation
Winged petiole’
\w
Spine
- Petals
Flavedo , Albedo
Flesh
Fig. 12.1. Citrus. (a) Grapefruit leaves and flower; (b) sections of flower; (c) cross section of fruit.
Rutaceae 42 1
moisture stress but where there is ample soil moisture trees flower through-
out the year. Approximately 30% of flower buds drop before reaching full
bloom, of those remaining only 1-4% set fruit which will reach full maturity.
Although most citrus cultivars are self pollinated, some mandarin and
mandarin hybrids set more and larger fruit when cross-pollinated. A critical
review of the factors affecting flower development and flowering has been
published (Krajewski and Rabe, 1995).
Chemicals, including antimetabolites, growth retardants, abscissants and
white oils, can inhibit or induce flower induction, depending on concentra-
tion and time of application. Harvesting also affects time of flower induction
in the following year; early picking has no effect but late picking inhibits to
some degree, A comprehensive review of many aspects of citrus flower
development has been published (Davenport, 1990). Chemicals applied to
citrus to influence growth, flower or fruit production have been studied for
their effect on the plant part under test, but little has been noted of their
effect on oil yield or composition.
The fruit is a special kind of berry, a hesperidium, having a tough epicarp
protecting fluid-filled trichomes which fill the locule of each carpel to form
the characteristic segments. Since the fruit, mainly the peel, contains an
essential oil it is discussed in detail. A diagrammatic section of a typical
citrus fruit is shown in Fig. 12.1. Since the rate of fruit development is
important to essential oil producers, a summary of the main stages is
apposite. Cell division begins at full bloom and continues to maturity; peel
develops from the ovary wall and juice sacs form in pulp segments. Cell
elongation also begins at full bloom, and fruit size increases as juice sacs
enlarge and contents increase. The peel changes colour. Maturation varies
with species, but fruit colour usually continues to change, soluble solids
increase and acidity decreases. There may also be changes in peel oil
composition between early and late maturing fruits.
Flavedo (peel) has an outer layer consisting of a thick cuticle, and stomata
in varying numbers occur in the epidermal layer of cells. The exocarp is a
thicker layer of irregular parenchymatous cells rich in chloroplasts and
photosynthetic in ripening fruit. Completely enclosed in this tissue are
numerous glands (oil sacs) which fill with essential oil under considerable
pressure (turgor pressure) which develops in surrounding cells. Albedo
(mesocarp) is a thicker, spongy white mass of tissue surrounding the fleshy,
juicy interior of the fruit; it is rich in vitamin C, sugars, cellulose, and pectins
which are commercially extracted in large citrus-processing plants. The fruit
centre consists of the developed carpels of the ovary, arranged around the
central pithy axis as a number of closely packed fleshy segments, developing
from a single carpel and surrounded by a thin transparent membrane.
Segments are filled with a mass of multicellular hairs which grow out from
the carpel wall, in which juice is produced as the fruit develops.
422 Chapter 12
Seeds in axil placentation are embedded in juice sacs close to the central
axis. As the fruits ripen, chlorophyll contained in the plastids of the exocarp
breaks down; the colour thus changes during ripening from green to yellow-
orange as carotenoid pigments, xanthophyll and carotene, present in the
plastids gradually become dominant.
The average composition of various citrus fruit is shown in Table 12.2, but
there can be wide variation in size, and between proportions of the various
components within a species, indeed between individual trees. Because of this
variation, it is essential to ensure samples taken from plantations to determine
fruit composition are truly representative. Not only must a minimum number
of trees be sampled, but fruits must be taken from different tree levels and
aspects, as there are significant differences in fruit composition, oil content
and oil characteristics from fruit growing in different positions. With certain
exceptions which will be noted later, essential oil production from citrus fruits
is generally a by-product of large-scale canned fruit and juice production by
utilizing its wa.ste products. About 45% by weight of fruit remains as cannery
waste after juice extraction, consisting of peel, pulp, rag (internal membranes
and core tissue) and seeds. Where oil is not extracted prior to juicing, it is this
material that is further processed to obtain the oil.
Solid residues remaining after juice and oil extraction are normally con-
sidered waste products, but can be treated to extract individual constituents or
processed into stockfeed or fertilizer. About 10 t of waste produces 1 t of dried
material for processing and may contain, in per cent: protein 6, crude fat 6,
crude fibre 14, N-free extract 60. Analyses ofvarious citrus seed and cakes are
shown in Table 12.3. Aged or unwanted citrus trees are usually burnt in situ
but can be used to fuel stills or for gas production (Churchill et al., 1984).
Ecology
The majority of cultivated citrus are probably indigenous to the humid
tropics or subtropics, and this origin is supported by a general mesophytic
character. The trees normally have extensive shallow root systems very
susceptible to a soil moisture shortage, while the flat evergreen leaves and
succulent bark on young shoots are also unsuited to moisture stress. Citrus
is today cultivated throughout the tropics and subtropics mainly between
40"N and S latitudes; the specific geographical limit is formed by the 15°C
isotherm for the coldest month. Main commercial production areas are in
subtropical regions between 20"N and 20"s with approximately 80% in the
southern USA, Central America and the Mediterranean, 10% in South
America, the remainder in Australia and East Asia. Brazil is probably the
world's largest producer of tropical citrus.
Citrus can be grown from sea-level to about 2000 m near the equator and,
at the higher level, average temperature is approximately 10°C lower at
around 17°C than at sea-level. Lower temperature at higher altitude has
other effects; in Kenya for instance, a most serious disease of citrus,
greening, is severe above 700 m but absent at sea-level as the insect vector is
intolerant of high temperature. Degree of cold hardiness in citrus and
methods of selecting strains with greater frost tolerance are important in
countries when summer temperatures are suitable for citrus production, but
very low winter temperatures kill trees (Goncharova, 1989; Guei et al., 1992;
Zhong and Chen, 1993; Roose and Kupper, 1994). A comprehensive review
with an extensive bibliography of cold hardiness in citrus has been published
(Yelenosky et al., 1987). Citrus species in order of decreasing cold resistance
were rated as follows: Poncirus trifoliata, Fortunella spp., Citrus aurantium, C.
reticulata, C. sinensis, C. paradisi, C. limon and, least resistant, C. medica.
Techniques including retention of snow cover and materials used to protect
citrus trees from low or freezing temperatures have been investigated (Rarn-
ishvili et al., 1988; Rieger et al., 1988).
The following is a general review of climatic factors affecting citrus growth
and development. The optimum temperature is 25-30°C, above and below
this growth diminishes and virtually stops at 38OC and 13°C. Citrus may
424 Chapter 12
can retard the reproductive cycle and induce an imbalance in other nu-
trients. It may also increase insect damage; aleyrodid (whitefly) population
for example (Onillon, 1988).
Superphosphate is best applied in bands 10-20 cm deep alongside tree
rows just under the canopy where irrigation and rainfall runoff are concen-
trated. Phosphorus frequently remains available for long periods and for this
reason up to 3 years’ requirements may be applied in one operation.
Phosphorus can also be applied through low-level irrigation systems using
water-soluble monoammonium phosphate and, although usually expensive,
the uptake by trees is rapid. Potassium fertilizers are usually expensive in
comparison to nitrogen and phosphate but, since potassium is often present
in soils, an annual application may not be necessary. When required,
potassium can be broadcast, injected with nitrogen through low-level irriga-
tion systems, or banded as a phosphate-potash mixture; potassium sulphate
is generally preferable to potassium chloride. Leaf analysis is the best
method of determining if potassium is required and the amount.
Trace elements play a vital role in tree health, and a deficiency as
determined by soil or leaf analysis must be corrected. Copper, magnesium,
zinc, boron and manganese deficiencies are the most common, and fre-
quently produce symptoms similar to certain diseases, including die back,
leaf discoloration and premature abscission. The small amounts required
are usually applied as part of a compound fertilizer, or injected via sprinkler
irrigation. Toxicity has also been recorded and may be artificially induced,
copper for example (Alva and Chen, 1995). Chelates are a convenient
method of supplying many trace elements, but are usually too costly for
smallholder use. Many trace elements also occur in fruit, but are seldom
transmitted to distilled oils although detected in cold-pressed oils. Residues
in fruit juice are common and have been determined in most developed
countries, Greece for example (Soulis et al., 1988). Legislation also exists in
many countries limiting the level of such compounds in juice.
Organic manures generally have low nutrient content, large quantities are
needed to satisfy tree requirements and it is also important to balance the
nutrient content by adding those in deficit. These manures are unpopular
with commercial growers who consider they increase the incidence of
Rutaceae 427
disease, pests or weeds. Where they are easily available and cheap, poultry
and cattle manure, sterilized human waste and blood and bone mixtures are
becoming popular on light sandy soils to increase organic content.
Cultivation
Necessary operations have been widely described in published works, but
some are still undergoing change, in particular weed control, pruning and
harvesting techniques (Whitney et al., 1994; Blandini, 1994). Grove man-
agement in general is also becoming more sophisticated, not only by using
computer models (Ben-Mechilia and Carroll, 1989; Regev and Raber,
1993), but also by incorporating data from satellite photographs (Edwards et
al., 1988). One aspect of citrus growing which affects the cost of establish-
ment is the long period to maturity of citrus seedlings. Thus techniques
which reduce the period between seeding and profitable fruit production are
of great economic value to commercial growers. The shorter the nursery
period for instance, the earlier seedlings can be planted out. Nursery
conditions and techniques are thus important; shading is one and in South
Africa 40% shading compared to full sun produced the fastest and most
healthy growth (Muller, 1988). Some factors affecting choice of rootstock
have been previously discussed, but the trend to dwarfism and high density
planting of some citrus is of increasing influence.
Chemical weed control remains important and although a particular
compound may be innocuous, there is a danger that a combination may
cause damage through chemical reactions between them; for instance a
phytotoxic interaction between bromacil and the insecticide dimethoate.
The following herbicides have been used, either singly or in combination:
secbumeton, terbumeton, dichlobenil, fluazifop-butyl, and combinations of
bromacil+diuron, bromacil+simazine, ametryn+secbumeton. Glyphosate
and MSMA have produced conflicting results causing either damage to trees
or depressing yield, while others have caused changes in oil characteristics or
composition; including bromacil and diuron (citronellol, alpha-terpineol,
nerol), terbacil (gamma-terpinene, citronellol), simazine (citronellol, nerol,
alpha-terpineol) .
The cost of hand pruning trees has become so high that some degree of
mechanical pruning has been forced on growers, varying from manually
operated trimmers to self-propelled machines whose size has also required
plantations to be established with sufficient room for them to manoeuvre.
An important secondary effect has been a change from a standard single-
stemmed rounded canopy tree to a more bushy type suitable for hedging
techniques, or single stemmed, low pruned, closely planted orchards. In
some circumstances, prunings can be distilled or extracted to obtain an
essential oil and in Cuba, lime, lemon and mandarin prunings yielded twice
the oil volume of orange and grapefruit prunings (Leyva and Alvarez, 1984,
1992). Prunings from 10-year-old plantations average 4.6 tiha.
428 Chapter 12
Irrigation water will inevitably become more expensive and suffer from
competition between end users, thus methods of utilizing it more economic-
ally, or the development of new sources will become necessary; recycled
waste water is one method (Zekri and Koo, 1993).
Harvesting
The increase in mechanical methods of harvesting citrus may have an
adverse effect on oil quality; such techniques are generally unselective with
as much fruit as possible gathered in one pass. Thus fruit in various stages of
maturity may be bulked for processing. Variation in oil content and compo-
sition at different stages of fruit maturity is well documented and, with
certain exceptions, inclusion of immature or mature but unripe fruit pro-
duces inferior quality oil. There is little effect on juice, the main product,
that cannot be remedied by judicious blending or use of additives. While it is
possible similarly to treat oil, selective hand-picking will generally produce a
superior oil. One aspect of citrus harvesting should be noted; in some
countries, official standards define when fruit is mature and may be sold, and
these requirements can determine a harvesting period. The general require-
ments for the development of an automatic harvesting system have been
defined (Rabatel et al. , 1995).
Distillation
Citrus oils are frequently obtained in a secondary operation following
expression of citrus juice. Lime, lemon and orange juice are all major
432 Chapter 12
Products
Although unprocessed citrus oils are frequently used by manufacturers,
most are concentrated, especially as flavourings. The main chemical compo-
nents responsible for flavour are the aldehydes, although generally account-
ing for only a small proportion by volume. The bulk is terpenes which
possess a relatively low flavour value, are insoluble in water and in some
cases chemically unstable. Removal of all or some improves oil odour,
increases solubility in water and enhances stability. The most concentrated
Rutaceae 433
citrus oils are those with all terpenes removed, terpeneless oils. At slightly
higher terpene levels are concentrated oils or more commonly folded oils; a
ten-fold oil contains only 10% of the oil’s original terpenes, but in practice
maximum strengths are generally five-fold for lime oil, ten-fold for lemon oil
and twenty-fold for sweet orange oil. Such oils are expensive, but are so
concentrated that only very small amounts are required.
The concentration or deterpenation processes may involve distillation,
solvent or poroplast extraction (Fleisher, 1994). Solvent extraction removes
virtually all the wax which is a important stabiliser of whole unprocessed oil,
but is considered a disadvantage in concentrated oils. Most citrus oils
deteriorate on standing or with age, and suitable storage and containers are
necessary to protect the required properties, prevent contamination or allow
oil to be transported (Table 12.7).
A most important aspect of concentration is the large volume of terpene
residue, widely used in cheap flavourings and perfumery; lime, lemon and
orange terpenes are generally available. Residues of solvent extraction are of
better quality and higher value than those from distillation, and referred to as
washed citrus oils as distinct from terpenes. Washed citrus oils may be used
in their own right, but are widely used to extend citrus oils to produce a
cheap but good-quality flavouring. The residue or effluent from fruit pro-
cessing contains useful ingredients which can be recovered when their sale or
utilization is profitable, or where effluent discharge content is legally con-
trolled. The major constituents are limonene, various sugars, yeasts and a
small proportion of oil (Parish et al., 1986).
Citrus oils have a very wide range of uses and research is continually
finding additional applications. Major outlets for citrus oils are in foods,
beverages, perfumery, cosmetics, soap, pharmaceuticals and condiments,
with secondary uses in the masking of unpleasant odours in domestic and
industrial insecticides, washing liquids, textiles and rubber products. The
most important market is the flavour industry which consumes large
amounts of natural oils and concentrates. In the industrial field, new uses are
regularly developed: waterless hand cleaners based on solvent properties of
terpenes present in citrus oil, or as a solvent in animal shampoo and similar
products, while considerable amounts of terpenes from citrus oils are used in
resins and adhesives, and orange oil is a source of d-limonene.
Citrus oil derivatives, such as distilled oil, essence oils and various types of
terpeneless oils are used where citrus flavour or odour is required without
the disadvantage of colour, or to blend with or enhance other oils. The
terpene, d-limonene, has a wide range of uses, one of the best known is in
production of synthetic spearmint oil flavour. Use of synthetic citrus com-
pounds in flavouring is legally restricted in many countries. It is, however,
legally permissible to use compounds consisting of blends of a whole citrus
oil, a washed oil and perhaps other components of the natural oil, such
compounds being variously known as lime oil WONF, lemon oil WONF,
434 Chapter 12
etc; the suffix means with other natural flavours. Washed oils are technically
more suitable for use in composite oils than extracted terpenes, although the
latter are more generally available and widely used. The growing trend by
consumers worldwide to pressure manufacturers to use only natural ad-
ditives and colourants in food and related products is an important factor in
Bergamot Less reactive than high high high low high Cool inert
other citrus headspace:
protect from light.
Lemon Rapid deterioration high moderate high moderate high Cool inert
on exposure to heat, headspace;
air and light; protect from light;
increase in viscosity antioxidants
upon ageing
Orange (Bitter) Rapid deterioration high high high moderate high Cool inert
on exposure to heat, headspace;
air and light; protect from light;
increase in viscosity antioxidants
upon ageing
Orange (Sweet) Rapid deterioration high high high moderate high Cool inert
on exposure to heat, headspace:
air and light; protect from light;
increase in viscosity antioxidants.
upon ageing
Lime (distilled) Rapid deterioration high moderate high low moderate Cool inert
on exposure to heat, headspace;
air and light; protect from light;
increase in viscosity antioxidants
upon ageing
(expressed) high moderate high moderate high
Grapefruit Rapid deterioration high high high moderate high Cool inert
on exposure to heat, headspace:
air and light; protect from iight;
increase in viscosity antioxidants.
upon ageing
Mandarin Rapid deterioration high high high high high Cool inert
Tangerine on exposure to heat, headspace;
air and light; protect from light;
increase in viscosity antioxidants
upon ageing
~
Lime
Botany
Citrus aurantifolia Swingle (syn. C. medica var. acida). Certain authorities
include under C. aurantifolia only the Key lime group, with the Persian
group classified as C. latifolia Tanaka. The latter may also be a hybrid
between Key lime and lemon or citron (Sacco and Calvarano, 1981). Thus
reference to cv. Tahiti for example as a lime is botanically incorrect, but is
common usage and generally accepted. In this section both groups are
discussed as lime, with any major divergence noted.
Lime is a small spreading tree usually to 4-6 m, canopy diameter to 7 m,
with many irregularly spaced often drooping branches which usually have
many short, stiff, very sharp spines, although some cultivars are spineless.
The canopy is usually rounded, dense and symmetrical in the Persian group,
more upright in the Key group, but the natural growth habit of both is
normally modified by regular pruning when cultivated. The leaves are 5-10
X 1.5-2.5 cm, glossy, dark green above, much paler below; elliptic-ovate to
obovate-ovate, round-toothed; the petioles are usually narrowly winged but
there is very wide variation. An essential oil is obtained by distilling leaves
but is of no commercial importance.
Table 12.8. Main characteristics of lime oils.
The remainder of this section will refer primarily to the acid lime group,
and lime, lime oils, lime products without qualification refer to acid lime.
Ecology
Lime is one of the least hardy citrus species and grown mainly in the tropics
and subtropics. An average annual temperature of 25530°C is suitable for
commercial production, with no prolonged periods of very high tempera-
ture. Persian limes are generally more tolerant of low temperature than Key
but, in Florida, where there are extensive commercial plantations, a major
constraint on fresh lime production is the regular occurrence of frost.
However, Tahiti has probably the greatest potential of all acid citrus fruits
for extended commercial production in the American tropics (Campbell and
Malo, 1980). Trees grow well from sea-level to higher frost-free inland areas
but most commercial plantations are at low altitudes.
An average rainfall of 3500 mm is the optimum well distributed over the
year; 2000 mm the minimum for successful commercial cultivation supple-
mented by irrigation; below 2000 mm irrigation is essential. It may be
possible to introduce increased resistance to water stress in cultivated lime
since related species are adapted to drier conditions, including Australian
Microcitrus spp. which produce edible fruit. Although most trees are topped
and pruned, high winds can cause considerable damage to long, drooping
branches. Hot dry winds quickly shrivel flowers and desiccate young foliage
and fruit. Shelter belts may be necessary in some windy regions, with tree
rows also aligned to reduce wind damage.
Oil characteristics from different countries vary considerably as will be
frequently noted. This is partly due to environment, partly to local cultivar.
A comparison of the main characteristics of selected distilled oils is shown in
Table 12.9. This variation can be substantial even within a relatively small
region, Table 12.10, and there is little published data on which environmen-
tal components affect fruit-oil content or composition.
Table 12.9. Main characteristics of distilled lime oils from selected countries.
India
Bombav Central Africa Westa Sri Lanka Cuba
Specific 0.8533 0.8610 0.861 0.8726 0.856
gravity (25") 0.8547 0.865
Refractive 1.4710 1.4750 1.476 1.4755 1.474
index (25') 1.4720 (30') 1.478
Optical t38"54' t 5V65' t 42'3' t58"02' t34"
rotation (25') t45'22 t 45O
Aldehydes 0.94 0.70 1.3 1.9 1.5
(as citral %) 4.24
Solubility (viv 95% ng 1 :2 1 :2
alcohol)
a All at 20°C; ng. not given. Figures in columns are range.
amount of nitrogen required is usually stated per tree and, unless there are
specific local recommendations, a good general guide is 0.5 kg N per mature
tree annually, but a slightly higher rate may be an advantage up to 4 years
after planting out, when growth is fast. Phosphate is best applied in planting
holes and, although an annual application is seldom necessary, an additional
amount may be required to maintain yield. Potassium has produced very
erratic results in terms of increased fruit yield, and only in commercial
plantations using high-yielding trees is this nutrient profitable. Normally one
application is required when planting out seedlings, and additional applica-
tions as indicated by foliar analysis.
Trace element deficiencies usually become apparent as discoloured or
distorted foliage or fruit, and this is particularly evident with iron, magne-
sium and zinc. A small sulphur deficiency can often be corrected by using
single instead of triple-superphosphate. Toxicities may also occur, boron
being one of the most common. This can be partially corrected by improved
cultural practices or increased water application on free-draining soils but, in
general, soil with a high boron content should be avoided.
Cultivation
The operations necessary to establish and maintain any commercial citrus
plantation are equally suitable for lime. The Key lime group produces viable
seed and seedlings are normally obtained by sowing seeds in containers or
seedbeds; there are commonly 5000-6000 seeds per kg (Tahiti is virtually
seedless). Seed viability is often low, and is usually further reduced when
stored at ambient conditions, although there is usually no dormancy period
(Dharmalingam and Vijayakumar, 1987). Spacing for lemon trees is gen-
erally suitable for limes in the same region, but profitability depends on
determining optimum tree number per hectare. In Mexico, spacings of 6 X
440 Chapter 12
Distilled Expressed
Mexican Cuba Caribbean Caribbean Florida
Component A B C C A
Alpha-pinene 1.3-1.9 0.8 1.69 2.47 2.7-3.1
Camphene 0.4-0.6 0.4 0.37 0.17 0 .O-0.5
Beta-pinene 1.7-5.3 1.2 9.49 24.23 4.3-23.0
Sabinene ng ng ng ng 0,442
Myrcene 0.7-1.1 1,l 0.86 1.40 0.8-1.6
Alpha-terpinene 0.8-1 ,I 1.9 ng ng 0.1-0.4
Limonene 42.8-52.3 46.9 47.25 44.17 43.7-56.2
1 ,8-Cineole 1.8-3.0 ng 0.29 0.26 0.3-0.5
Gamma-terpinene 6.8-1 1.6 8.9 8.86 11.88 2.5-16.6
p-Cymene 1.5-7.5 3.5 0.21 ng 0.1-1.3
Terpinolene 2.9-7.5 7.3 7.24 0.83 0.5-0.7
Linalool 0.2-0.3 0.3 0.12 0.25 0.1-0.2
Alpha-fenchyl alcohol 0.4-1 .O 1.4 0.32 ng ng
Bergamotene t 0.6-1.4 1.7 3.00 1.66 0.7-1.3
terpinen-4-01
Neral 0.2-0.3 ng 0.13 1.13 0.5-1 .O
Beta-terpineol 0.5-1 .O 0.9 0.26 ng ng
Alpha-terpineol 6.1-9.1 9.0 6.81 ng 0.1-0.4
Geraniol 0.2-0.4 ng 0.31 1.70 0.349
Beta-bisabolene t ng 1.5 2.12 3.18 ng
neryl acetate 1.3-2.1 ng ng 0.07 1.5-2.3
Geranyl acetate 0.6-1.4 ng 0.56 0.16 0.1-1.8
ng, Not given. Figures in columns are range in per cent.
Source: A, Alessandro et al. (1985); 6, Analytical Methods Committee (1984);C, Azzous eta/. (1976).
Rutaceae 44 1
Distillation
Distilled lime oil is obtained by two basic methods: steam distilling a slurry
of crushed whole limes and finely chopped peel, or distilling juice from fruit
pressing. In the latter, juice suitable for canning is drawn off after a storage
period and the residue of juice, pulp and dissolved oil is steam distilled. An
improved technique combines steam and vacuum low-temperature distilla-
tion, which allows direct use of halved, de-juiced fruits without adding
water; this reduces the danger of affecting oil flavour, inherent in treating a
highly acid mixture with boiling water.
Lime oil composition is extremely variable as noted, and a comparison of
oils produced in the Caribbean region is shown in Table 12.10, which well
illustrates the effect of regional variation even within a relatively small
geographical area. In general, recovered oil from steam-distilled whole fruit
juice is usually 10-15% less than is present in unprocessed fruit.
Oil recovery from lime-juice production is low on an industry basis; in
Florida, the annual average is below 0.5 kgit from crushing thousands of
tonnes of Persian limes. Peel-oil content from random samples gave yields of
3.0-4.7 kgit, average 3.6 kgit. However, a certain amount of oil was
deliberately retained in the juice to impart its specific flavour. The average
yield of crude oil in Bombay, India, is 3.0 kgit but quality is poor and oil
must be reprocessed before use in food or drinks. Caribbean and Mexican
yield is generally below 2.5 kgit, but in modern factories it is up to 5 kg/t from
steam-distilled juice from whole pulped fruit (Mentonelli, 1994). There is
currently no major commercial production of lime oil in sub-Saharan Africa,
although pilot plantings in the Gambia yielded 0.27-0.82%, averaging
0.48% from local limes (about 5 kgit fruit), and in Kenya experimental
plantings near the coast yielded to 8 kg oilit from selected trees.
Pressed oil is used in high-grade men's toiletries, while distilled oil has a
particular application in shaving creams. In comparison with distilled oil,
use of cold-pressed oil in soft drinks is low. Lime oil is often concentrated
before use as a flavouring and terpene residues also have many applications.
Use of the oil in perfumery, especially cold-pressed oil, suffered a severe
setback during the mid 1970s when it was discovered that coumarins could
have phototoxic effects on human skin. Distilled oil presented no problem as
coumarins are not carried over during distillation, however it was affected by
the general uncertainty. Rectification of cold-pressed oil to wholly or par-
tially remove coumarins largely overcomes the problem, but inevitably
increases cost.
Expressed oil was formerly obtained by manual ecuelling described ear-
lier. A mechanical method presses whole fruits, and the extract is then
passed through high-speed centrifuges quickly to separate the oil. Although
this is technically an expressed oil, its characteristics, flavour and odour
differ from ecuelled oil, and they are not always interchangeable. Expressed
oil is medium yellow to dark green and may be so dark as to render
determination of optical rotation impossible. Fresh oil has a rich sweet
odour, lemon-like but more mellow and nearer that of crushed limes but not
lime juice; it is of outstanding tenacity, far exceeding that of distilled oil and
superior even to terpeneless oils. The spicy-herbaceous and sweet balsamic
notes are especially persistent. Although expressed oil has the true peel
flavour, it is unpopular in flavourings, and is frequently blended with other
citrus oils when used in beverages or confectionery. The oil is widely used in
perfumery and also as a modifier or blender. Since expressed oil is produced
only in small quantities and is relatively expensive, it is frequently adult-
erated; commonly with distilled lime and lemon oils, citrus oil terpenes, and
certain synthetic compounds. Analyses of pure and adulterated lime oils and
additives have been published (Alessandro et al., 1985).
444 Chapter 12
Bitter Orange
The bitter, or sour, orange is native to South East Asia, and was brought by
traders to the Middle East and thence to Europe, where it rapidly became
established in the Mediterranean region generally, particularly Spain, hence
the common name of Seville orange. Bitter orange trees may live to a great
India USA
Colour Dark browna Dark brown' Yellow Clear
Specific gravity (25") 1.021 0.975 0.959 0.8694
Refractive index 1.526 1.510 1.4825 1.474
Optical rotation -0.9lC ng -0.9" t 33"55'
Aldehydes as citral(%) 19.98 23.90 6.09 0.71
Ester number 17.06 14.83 5.42 2.42
Acid number 7.99 6.34 5.65 1.36
Solubility 1 :3 1 :3 1 :3 1 :2
(viv 95% alcohol)
aWaxed; Dewaxed; ng, not given.
Data are from commercial oils.
Rutaceae 445
Botany
Citrus aurantium Swingle is commonly known as the bitter, sour or Seville
orange. Bergamot orange is now classified as C. aurantium. subsp. bigaradia
and petitgrain orange as C. aurantium. subsp. amara. There are numerous
sour orange hybrids with other citrus species, some doubtfully allocated
species status (Tanaka, 1954; Samson, 1986; Gocorcena and Ortiz, 1989)
and many named cultivars. Bitter orange is source of a number of important
products, although these are not usually produced from the same subspecies
(Zhengkul et al., 1988; Boelens, 1991; Boelens and Operto, 1991). Most
important are leaf oils chiefly petitgrain, flower oils mainly neroli, peel and
fruit oils.
Bitter orange is a medium sized tree to 10 m, usually less under cultiva-
tion, normally with a central stem bearing a number of stout branches. The
446 Chapter 12
branches are spiny, the spines on young branches being long and slender but
more stout and stiffer on older branches. The wood is strong, takes a fine
polish and is used for furniture. The leaves are large (8-10 X 4-5 cm) ovate
to ovate-oblong, with the base wedge-shaped and are wavy edged or slightly
toothed. They are glossy green above, paler below and the petiole very
broadly winged. All species of bitter orange bear aromatic leaves, which yield
an essential oil on distillation, but that known as petitgrain oil derived from
C. aurantium. subsp. amara is the most important.
Leaf oil is obtained as a secondary product from trees grown to produce
another oil, or grown especially to produce leaves for processing, but the
method of obtaining oil is essentially similar once leaves have been picked.
The most important constituent of leaf oil and that which defines its quality
is ester content, calculated as linalyl acetate, and oil is sold commercially
based on the content.
The highly scented white flowers are borne singly in leaf axils or in small
clusters on axillary racemes and are larger than those of sweet orange with
20-24 prominent stamens. An essential oil, neroli oil, can be obtained by
distillation or enfleurage. A form of sour orange with scentless flowers and
lacking oil glands in leaves has been noted in Cuba (Valle et al., 1985).
Neroli oil is produced from flowers of several Citrus spp; that obtained from
bitter orange flowers is neroli bigarade, from sweet orange neroli Portugal,
and from lemon neroli citronier.
The fruit is roughly spherical, 6-8 cm in diameter, may be compressed
longitudinally and often develops a hollow core at maturity. The skin is
normally rough or warty, but smooth in some cultivars. The fruit is a
medium to dark green, and may become more yellow when fully ripe; the
pulp is very acid. An essential oil is obtained from skins and to a minor extent
from pulp. A major use for whole fruit is in (Seville) marmalades and the
juice is a valuable source of vitamins A and B. One common form of bitter
orange, bittersweet, has fruit which contains 3% citric acid compared with
an average of 1.5% for true bitter orange. There may be a correlation
between degree of acidity and oil content of leaves, since in Paraguay leaf-oil
content was lowest in those forms with the least acid fruit. The number of
seeds per fruit varies between cultivars, and they are usually viable, with no
dormancy period.
Ecology
Bitter orange is basically a warm-region species and flourishes where the
minimum temperature remains above 0°C except for very short periods. An
annual average temperature of 20-25°C is considered most suitable, but up
to twice this will be tolerated provided soil moisture is adequate. Best growth
and flowering occur in bright sunny conditions. Frost normally causes
extensive damage to small branches, and a severe frost or icing is frequently
fatal. There is apparently a relationship between high levels of polyamine
Rutaceae 447
Cultivation
Bitter orange is seldom the first crop on newly cleared land as weed control
is essential if seedlings are to survive, and weeding is always difficult in the
first year after clearing. An interval of at least one year is recommended when
replanting an exhausted plantation, all trees and residue burnt and, if
possible, the land should be ploughed and left fallow. Cultivation should aim
at maximum weed reduction. Seeds are sown in pots or seedbeds, where
448 Chapter 12
Harvesting
Under normal conditions first harvest is from 3-year-old trees which are cut
back to 12-15 cm above ground level. All material is transported to stills,
where leaves and twigs are stripped, and larger branches used as fuel. When
bitter orange is grown for flowers or fruit, the annual pruning supplies leafy
material for distillation. Trees can be cut at approximately 9-month intervals
and regular cutting maintains a leafy bush which produces the maximum
foliage. When the world price of oil is low or demand falls, plantations in
Paraguay are virtually abandoned until harvesting is again profitable. Trees
which have been left uncut for several years can be recoppiced without harm,
and normal cutting resumed. There is thus a resilience in this type of
production eminently suited to local conditions. Cutting may be at any time;
in Paraguay mainly in October-March; pruning is generally in July-October
in France, February-April in Italy. On Haiti, leaves can be harvested
throughout the year, but highest quality oil is obtained in May-October after
the flower harvest. Oil yield in Paraguay increases steadily to the fifth year,
then remains constant for decades under good management (Baker,
1980).
The most important factors affecting leaf-oil content are seasonal, varietal
and to a lesser extent geographical. The second and third are to a great
extent complementary, since one or at most two cultivars are grown in a
particular region, and there are few data on the effect of growing one in
different locations; Paraguay is an exception. Annual and seasonal variation
is high since the local climate directly influences leaf oil content and oil
characteristics. In Paraguay, the highest oil content generally occurs in
January-June, the lowest in August-October; in Italy, yield rose as the
weather became warmer from February to April; in Brazil, the highest oil
content is in warm months but linalyl acetate content is frequently highest in
cool months. Two major operations affecting oil yield are directly under
Rufaceae 449
leaves and petioles produces an off note. Flowers are usually stored over-
night in thin layers in special stores before being distilled.
Distillation
Leaf and oil yield vary widely; in Paraguay individual trees provide 10-1 5 kg
leaves annually and 200-300 kg yields about 1 kg oil; in Haiti, 300-400 kg
about 1 kg oil; in France, 500 kg for only 1 kg of very high quality oil. A charge
normally consists of leaves, twiglets and sometimes small branches. Still
operators consider that inclusion of a reasonable amount of twigs stops
consolidation and channelling thus ensuring more efficient distillation.
While time of cutting in terms of leaf maturity effectively controls potential
oil yield, treatment after cutting can directly affect oil per tonne of material.
Distilling fresh leaves gives the highest yield; in Italy, fresh leaves yielded
0.28%, after 3 days storage 0.23%. Prolonged distillation reduces linalyl
acetate content through ester hydrolysis, especially when direct-fired stills
are used and material remains overlong in contact with water; however, in
Paraguay, the odour of this type of oil is locally preferred.
Most stills in South America and the Caribbean are unsophisticated, and
an interesting description of local Paraguayan stills is that of Baker (1980),
but this is more a factor of economic development and plantation profitabil-
ity than lack of knowledge. The Paraguayan government ‘actively supports
methods of increasing the efficiency of stills where considered necessary to
upgrade oil quality and was possible within the existing rural infrastructure’,
in practice where the local people could afford to do so on their own!
Those factors affecting oil content also basically affect oil composition,
but incorrect or prolonged distillation has a greater adverse effect on ester
content. In commercial oils, ester content as linalyl acetate is, in per cent
(average in brackets): Paraguayan 45-55 (49-52); Italian 45-79 (50-60);
French 55-70 (60-65); Brazilian 55-90 (60-65); laboratory analyses
quoted Indian (25), French Guinea ( 5 5 ) and Algerian (36). T h e very high
variation indicates there is considerable scope for selection for high leaf oil
content, and this also applies to ester content, as in Brazil linalyl acetate
ranged from 57 to 90% (Craveiro et al., 1981).
Flowers are now generally distilled for oil, while solvent extraction of
flowers and orange water yields various absolutes. Distillation differs from
that of leaf oil as flowers are placed directly into a still body containing water
which is then indirectly heated. A significant amount of oil is dissolved in the
water and recovered by solvent extraction. Thus two products are obtained,
Neroli oil and orange flower water. When only unadulterated flowers are
distilled about one gram of oil is obtained per kilogram of flowers, while
3000 kg of solvent extracted distillation water yields about 1 kg orange water
absolute.
Peel oils are obtained from all types of bitter orange fruit, and although it
is not strictly an essential oil it is widely considered as such under the name
Rutaceae 451
bitter orange oil, but for ease of reference as peel oil hereafter. Peel oil is
mainly produced in the Caribbean, particularly Jamaica, the Dominican
Republic, Haiti, and some Latin American countries especially Brazil; in
Europe, Sicily is the most significant producer, while Russia, China and
some North African countries produce small amounts. West Indian oils are
considered superior, while those from Sicily are often the most expensive.
Probably the most important factor affecting oil quality and type is method
of expression and subsequent treatment.
Peel oil is obtained by rupturing oil-containing cells in peel either by
pressure or abrasion, and the method of expression affects total yield, since
some producers only partially express peel which is then sold for other
purposes. Expression may be manual or mechanical, should be cold, and the
fruits as uniformly mature as possible. Inclusion of unripe fruit causes a
proportionally greater reduction in oil quality than inclusion of those more
ripe. Fallen and diseased fruit should be rejected, since these produce off
notes in oil. Peel may be distilled following expression to recover a poor
quality residual oil. Peel oil is more expensive than sweet orange oil as there
is no significant use for juice. Important constituents are citronellal, linalyl
acetate, 1,8-cineole, and non-volatile coumarins to 0.25% not found in sweet
orange oil.
period in storage, the container, and often by the extent and type of
adulteration. Crude or bulk Paraguayan oil can be very variable in quality, is
seldom consistent, often contains water, and should be filtered and dried
before evaluation. The Paraguayan government has established export stan-
dards for oil and shipping containers, which only a few major exporters
observe. This oil is extensively used in soaps and similar products, and in
inexpensive perfumes. Petitgrain Haiti is even more variable, mainly due to
type of material distilled. Its uses are similar.
Terpeneless petitgrain oil is normally derived by vacuum distillation of
Paraguayan oil with a volume loss of 20-35%. Terpeneless oil contains
mainly linalyl acetate, linalool and methyl anthranilate, and is almost colour-
less. The extent to which terpenes are removed basically determines the oil’s
odour, which is often the individual preference of a perfumer or company.
Deterpened oils are important in perfumery where their greater solubility is
an asset; in food and drinks they are used in conjunction with natural or
synthetic fruit flavours.
Petitgrain bigarade surjleurs d’oranger is a pale to very pale yellow, with a
sweet orangy odour. It is produced in small quantities mainly in the Grasse
region of France by distilling bigarade oil over flowers from the same type of
bitter orange. The ratio of oil to flowers is not fixed, thus there is a wide
variation in the final product. The odour of this oil is considered by
perfumers to be superior to bigarade oil, and is extensively used to enhance
other flower oils in fine perfumes; the oil suffers from numerous inferior
imitations with the same name. A similar oil is produced by mixing bigarade
oil with neroli oil, but is of little value to perfumers. Another speciality oil is
obtained by distilling together leaves, twigs and flowers, whose character-
istics and odour differ as the proportion of the three plant parts varies.
Petitgrain bergamier is obtained from leaves of C. auruntium subsp.
bigarudia and has an odour similar to Paraguayan petitgrain. The oil is
seldom used alone but commonly as an adulterant or extender of other citrus
oils. A number of other leaf oils have been produced from various Citrus spp.
growing in the Philippines, East and West Africa and the Caribbean, but
none are of economic importance and unlikely to become so. Some have
specific characteristics noted in the relevant literature. The further develop-
ment of these oils probably depends on their use by a particular manu-
facturer seeking a novel product.
Flower oils are generally known as neroli oil, and usually differentiated by
a country or quality suffix. They suffer from the same problems of adultera-
tion and inferior substitutes. Neroli bigarade is the highest quality, obtained
by steam or water distilling flowers of the true bitter orange. Haitian neroli is
steam distilled from flowers of various species of citrus, and is quite different
in its characteristics and odour. Neroli oil can also be obtained by hot
454 Chapter 12
enfleurage, but this operation has now virtually ceased. Flower oils are
produced mainly in the Mediterranean region, the Caribbean, Latin Amer-
ica, Southern Africa and China and although manufacturers frequently
prefer a specific origin, the highest quality oils are French and Tunisian. The
major components are linalyl acetate, limonene, linalool, methyl anthrani-
late and geraniol.
Neroli bigarade oil is a clear, pale to very pale yellow mobile liquid,
becoming viscous and darker with age. Fresh oil has a strong, attractive
orangy-floral odour, but little persistence. The oil quickly darkens and the
odour deteriorates unless kept in well-sealed containers in a cool dark store.
The main characteristics of neroli bigarade of various origins are shown in
Table 12.15, but there can be substantial regional, indeed seasonal varia-
tion. Distilling methods and techniques also have a major effect on the oil,
particularly the extent to which certain alcohols are hydrolysed. A large
Spanish producer uses copper rather than stainless steel equipment, claim-
ing its product has a superior odour. At least 40 components have been
identified in neroli oil and analyses of two oils shown in Table 12.16.
A very interesting review of the contribution various constituents make to
oil odour is that of Buccellato (1981), who states that hydrocarbons provide
the bright, fresh citrus character, alcohols the floral body, esters the fruity
component, while phenols and nitrogenous compounds modify the whole.
Neroli bigarade normally contains a high proportion (approximate per-
centage in brackets) of hydrocarbon monoterpenes (35); linalool (30);
aldehydes (1 2 ) ; linalyl acetate (1 0); and nerolidol ( 7 ) . Although many
analyses of neroli oil have been published, they often fail to state the origin
and are thus of limited value, A characteristic of fresh neroli oil is the blue
Commercial
France North Africa Sicily Spain Haiti range (USA)
Specific 0.87 0.87 0.87 0.86 0.86a 0.863a
gravity (20°C) 0.89 0.88 0.88 0.88 0.87 0.880
Refractive 1.47 1.47 1.4 1.4 1.47 1 .oa
index (20") 1.48 1.48 1.5 1.6 1.49 2.2
Optical t? t6" t5 O t8" t2ya t 1 .ya
rotation (20°C) t 7$ t9" 7".t +loo t 25" t9.IC
Solubility 1 12-3 112-4 1 :I-2 111.5 1:1-1.3 1 I1
(viv 70% alcohol) (80%) -2.0 (80%) 1.5
Acid number < 2.0 < 2.0 < 2.0 < 3.5 < 3.0 < 3.0
Linalyl acetate (%) 10-15 15-25 8-1 0 15-25 12-15 ng
Linalool (%) 25-40 25-45 55-65 25-35 ng ng
a Figures in columns are the average range; the range can be much greater; 25"; ng, not given.
Rutaceae 455
Table 12.18. Major components of bitter orange and bergamot peel oils.
ng, Not given, Figures in columns are range in per cent. The very wide range of individual components
is noteable.
Source: Published data from Italian commercial samples.
Rutaceae 459
USA
~
Bergamot Orange
This subspecies of C. aurantium has had a somewhat chequered taxonomic
career and is now known as C. aurantium subsp. bergamia. Argument
continues regarding its origin, status, and whether it is a mutation or hybrid.
Development of the toilet water known as eau-de-cologne, originally made
in Italy by the Feminis family in the sixteenth century, stimulated bergamot
oil production, as the oil is an essential ingredient. Expressed oil was well
established in Italy by the mid-eighteenth century, and commercially traded.
Bergamot orange has a much more restricted distribution than bitter orange,
and was originally confined to the south-west coast of Calabria, southern
Italy. Although bergamot has been introduced elsewhere, only the Ivory
Coast, Guinea, Morocco and Corsica produce oils comparable with Cal-
abrian oils, and Italy remains the most significant producer (Giacomo,
1987; Dugo, 1994).
Botany
Bergamot is an upright tree to 12 m with a single trunk to 25 cm DBH and
numerous branches. Under cultivation, trees are pruned to 4-5 m and
branches cut back to maintain a tree diameter of 5 m. The leaves and flowers
are typical of the species. An essential oil can be obtained from leaves by
distillation but is of no commercial importance (Lawrence, 1993). The fruit
is generally globoid to 7.5 cm, the peel thin, tough, smooth and initially
green becoming yellow when ripe. The very regular, spherical shape of
bergamot fruits and the evenness of peel thickness greatly facilitated the
introduction of mechanical processing equipment. Bergamot oil is obtained
460 Chapter 12
from the peel, the juice is of minor value, and residues after processing are
used as stockfeed. In Calabria, three bergamot varieties are recognized,
Feminello, Castagano and Inserto, and the last is locally considered a hybrid
of the first two and by far the most popular. All three were tested in North
and West Africa where Inserto proved most adaptable. The rootstock used
for bergamot has a significant influence on growth, yield and disease
resistance, and remains to be fully evaluated.
Ecology
It was widely publicized that bergamot required quite specific climatic
conditions if it was to flourish and its fruits to contain oil of the required type.
The most important producing regions of Calabria have a mild equable
climate where temperature only very rarely falls below 10°C or exceeds
30°C. The success of the West African plantations with a very different
climate indicates that agronomic skill is probably more important. Frost
causes serious damage to trees and, if prolonged, is often fatal. Low
temperatures at flowering or when fruit is setting drastically reduce yield.
Strong, hot winds have a similar effect and shelter belts are recommended
where there are seasonally high winds. An annual rainfall of 500-700 m m is
adequate if augmented by irrigation. Plantations are normally established
below 400 m in Europe, at higher levels only on southern slopes or sheltered
valleys. Detailed statistics on climatic conditions in the North and West
African bergamot regions have been published (Huet and Dupuis, 1967)
and can be used as a guide when considering the suitability of establishing a
bergamot plantation.
Cultivation
In general, cultivation as for other citrus is suitable, and spacing between
trees governed by the agricultural system employed. Bergamot seeds are
viable, but most seedlings are budded or grafted onto other citrus rootstocks.
Sour orange is widely used, but other species may prove to be more disease
resistant. Older growers in the Calabria area believed lemon and citron stock
gave an oil of finer quality, but there is no data indicating rootstock
influences oil composition or characteristics. Tree density can vary from 450
to 600 per hectare in older plantations, 350-400 in plantations where
mechanical equipment is used.
Many plantations are intercropped initially to reduce establishment cost;
with jasmine in North Africa and occasionally in Italy. Irrigation is normal
since trees flourish in warm dry conditions. Surface irrigation is common in
Calabria partly from tradition but also because installing sprinklers or
rainguns is unprofitable. Drip irrigation which conserves moisture and
ensures regular growth, is probably the most viable system where it can be
employed. Weed control is normally manual when family or cheap labour is
available, but herbicides are now becoming more popular as there are
frequently labour shortages at critical periods. An overall grass cover is
normal in older plantations established on hillsides or sloping land to resist
soil erosion, and this is either scythed or grazed by small livestock. A circular
area around tree trunks is often cleared to reduce damage by disease and
pests. This system is of value elsewhere, where contour planting on terraces
would be an advantage. Pruning is in February after harvesting is completed
and this is a skilled operation as each tree is pruned individually. Mechanical
pruners have been tested and are discussed in the sections on lemon and
sweet orange.
Bergamot trees mature slowly; a small harvest may be possible after 7
years, increasing annually to 12-1 5 years, and trees remain in bearing for
50-60 years with proper attention. Fruit yield from well-managed planta-
tions in Calabria is 200-300 kg per tree, and average oil yield 300-600 g per
tree. Individual trees have much higher fruit and oil yield and should be used
to upgrade local stock. Those factors affecting fruit size, etc. noted under
bitter orange are generally applicable to bergamot. Peel oil content and its
composition, however, are not only directly affected by agronomic factors
but also by time of picking, handling, transport and method of expression,
462 Chapter 12
which are not further discussed since these operations can be easily con-
trolled.
Peel oil content is extremely variable, and although this is partly genetic,
climate directly affects oil content. It is difficult to assess the effect of a
specific component; a high temperature for instance can be accompanied by
high humidity as in the West African rainy season, or low as during a sirocco
wind in Italy or the khumsin in North Africa. A moderately high temperature
during maturation apparently reduces linalool content and increases ester
content in North Africa and, although ester content increases when the
weather becomes warmer as the season progresses in Calabria, a drop in
temperature in the final stages of fruit maturation may further increase ester
content of peel oil. Growers still remember a sirocco wind which occurred in
July 1965 when the temperature rose very quickly to a high level causing
extensive fruit damage; oil produced that season had a low ester content and
a high rotation. In Calabria, linalool content is highest in early fruits and
decreases steadily to a minimum in March, while linalyl acetate content is
the reverse; early fruit in general has less terpenes than late maturing.
Flowering begins in West Africa at the beginning of the rainy season and
early maturing fruit has lower esters and alcohol content.
Harvesting
Manual harvesting is general and although tree-shakers have been used
where possible, it is unlikely for economic reasons that mechanical harvest-
ing will become common. Picking begins in November-December and
completed in February-March in Calabria and North Africa, somewhat
earlier in West Africa where seasonal rainfall and temperature have a major
effect on rate of fruit maturation. Fruit is harvested when mature but unripe,
carefully stored in the shade after picking and quickly transported to the
processing plant. Many producers in Calabria now sell their crop on the tree
to an independent contractor, or the processor arranges for it to be picked,
as young people have little interest in agriculture and labour is scarce and
costly. A result is that smaller, independent oil producers have gradually
been replaced by large processing plants. One of the biggest was the
Consorzio di Bergamotto (the Consortium) which not only processed fruit
but bought local oil which it further processed to a guaranteed standard of
quality and purity. The Consortium was subsequently reorganized and its
monopoly power reduced.
Distillation
Several oils can be obtained from bergamot fruit other than the standard oil,
including treatment of crude oil or residues. In Calabria, bergamot oil is
produced mainly by the Pelatrice process with only a minor amount by the
traditional Calabrese method. An oil, Bergamotelle, is distilled from young
fruit; another locally named Nero is expressed from ripe and fallen fruit; also
Rutaceae 463
a derived distillate Frecci, and Mollicata extract. These are normally only
produced to order, or by extractors for their own use. Most important is
Nero, highly aromatic and frequently used to enhance less fragrant oil. The
oil's main characteristics are apparent density 0.896-0.900; rotation < 10"
but turbidity may make accurate measurement difficult; esters 30%; alco-
hols 50%; non-volatile residues 15%. A Calabrian oil, fallen bergamot, is
obtained from peel of fallen fruits which are whole and not diseased or
rotten, and mainly used to extend true oil. An oil obtained by steam distilling
peel remaining after cold-pressing fruit has a low linalool and linalyl acetate
but high monoterpene content, and its main use is also to adulterate true oil.
Juice and pulp residue may also be distilled to recover an oil which does not
resemble peel oil in odour. The annual oil yield from factory-processed fruits
is 0.45-0.65% depending mainly on climatic conditions in a particular year,
averaging 0.50%.
Italy Ivory
USA (Calabria) Guinea Coast Morocco USSR
(EOA) A B C D E F
Specific 0.875 0.8795 0.8790 0.8730 0.875 0.862 0.877
gravity (20%) 0.880 0.8845 0.8867 0.8910 0.882 0.873 0.888
(25%) (15°C)
Refractive 1.4650 1.464 1.462 1.464 1.462 1.468 1.464
index (20°C) 1.4675 1.466 1.467 1.468 1.465 1.770 1.469
Optical t8 O t15O t5 O t12O t6" t38" ng
rotation (20°C) t 24" t 30" t3Ic t 33" t 22O t 60°
Solubility
(viv 90% alcohol) 1:l 1:l 1 :0.5 1:l ng ng 1:I
Ester contenta 35-45 30-60 30-45 32-45 32-44 16-33 3545
aAs linalyl acetate pi);ng, not given; EOA, Essential Oil Association, USA.
Source: A, Muller (1966); B, Calvarano (1965); C, Schwob (1953); D and E, Chapot (1962); F, Lozzi
(1963)I
has been intensively studied and there are a number of synthetic substitutes
available, but those which approach the true oil are also nearly as ex-
pensive.
Terpeneless bergamot oil is obtained by vacuum distilling whole oil, with
a yield of 66-75% depending on the product required. Its most important
assets are greater solubility and slightly stronger and finer odour. Since it is
more expensive, it is used in higher quality perfumes and similar products.
Bergamot concrete produced in small amounts by solvent extraction has an
excellent bergamot odour. Bergamot petitgrain is produced by steam distill-
ing leaves and twigs from annual pruning, and used as a partial substitute for
other petitgrain oils, but is normally too expensive or available only in small
amounts. This oil is a green to olive-yellow mobile liquid, with generally the
same odour characteristics as South American oils. The residues from oil
extraction, dried, wet, and as silage are used for stockfeed, or may be further
processed to obtain specific ingredients, The juice known as ugro, although
too bitter for use in beverages without further processing, is raw material for
extraction of various acids and vitamins, but frequently discarded as
waste.
Lemon
Lemon is of Asiatic origin but its native home remains unclear although the
Indo-Burmese region is most valid. The fruit reached Europe via Persia and
Rutaceae 465
the Middle East sometime in the twelfth century. Sicily is now the most
important European producer of lemon oils, Spain is of lesser importance.
Sicilian lemon oil was first mentioned in 1780 when Abbot Domenico
Sestini noted production of several thousand pounds. Christopher Colum-
bus carried seeds to the West Indies in the fifteenth century and lemon was
subsequently dispersed throughout South America by Catholic missionar-
ies; Argentina is now the largest lemon producer followed by Peru and
Brazil. Missionaries were also responsible for introducing lemons to Cal-
ifornia in the late 18th century, and this state is probably the world’s largest
single producer of lemon oil.
Seed was first carried to Australia in 1788 and lemon is grown commer-
cially and generally in gardens including that of the author, and rough lemon
has become widely naturalized. Lemon has also been more recently in-
troduced to many African countries, but only South Africa, Ivory Coast and
Guinea have significant commercial production. Lemons are seldom grown
commercially in Asia or South East Asia, although single trees are often
planted in gardens and smallholdings. A very interesting account of contem-
porary world lemon oil production was Ernest Guenther’s survey
(1963-64). There are many named cultivars of smooth and rough lemon
available commercially, and specialists mass-propagate rough lemon seed-
lings for rootstock.
Botany
Citrus limon (L.) Burm. (syn. C. medica subsp. limonium Hook., C. limoniunz
Risso), and the species are divided into two groups with either smooth or
rough skinned fruit. Some taxonomists regard the two as distinct species; the
smooth as C. limon, the rough (citronelle) as C. jumbhiri Hodg. Others
consider rough lemon a stabilized lime X citron hybrid since their centres of
origin overlap and generally coincide with lemon. Chemotaxonomy may
offer a solution, and also assist in classifying the many cultivars. Meyer
lemon (sometimes designated C. meyeri but usually C. limon X C. sinensis) is
considered to be a smooth lemon X orange hybrid, with a characteristic oil.
In this section C. limon will be discussed as smooth and rough lemon.
Smooth lemon is a medium-sized tree to 10 m but generally 4-6 m, with
a single trunk and many branches which are usually well, if irregularly,
spaced, giving the tree an open, spreading appearance. Small branches and
twigs bear short stiff spines, usually absent on mature wood. Rough lemon is
a large spreading tree to 12 m seldom cultivated for its fruit. It quickly
becomes naturalized as seed is viable and seedlings vigorous. Leaves on both
are pointed, elliptic-ovate, 5-10 X 2.5-5.0 cm, slightly serrate, with short
almost wingless petioles. They are normally light to mid-green in colour,
glossy, and less strongly scented than other Citrus spp. Steam distilled leaves
and twiglets of smooth lemon produce oil of petitgrain lemon with a citral
content to 50%; the main characteristics are shown in Table 12.21. Nearly
466 Chapter 12
100 compounds have been isolated, and the major components are limo-
nene, sabinene, gamma-terpinene, beta-ocimene and linalool (Lund et al. ,
1982; Lawrence, 1993); rough lemon leaves contain a very similar oil (Lund
et al., 1981; Lawrence, 1993).
The flowers are 3-5 cm diameter, white, the buds pink-tinged and the
reflexed petals retain the colour. They are usually borne singly or in pairs in
axils of leaves. The flowers are fragrant in both types but not processed for
oil. Flowering can continue year-round and thus flowers, immature and ripe
fruits can occur on the same tree. In certain cultivars, flowering can be so
profuse that growers thin to ensure large, even-sized fruit. Growth regulators
can enhance fruit set when applied at flowering, or substantially reduce
premature fruit drop, but some adversely affect peel thickness and/or oil
content. Neroli citronier is obtained by distilling lemon flowers, but is not a
substitute for neroli bigarade obtained from bitter orange blossoms, and is
produced only on demand.
Smooth lemon fruit is ovoid to oblong, 7-10 X 3.5-5.0 cm diameter,
tapered at both ends with a pronounced terminal nipple: rough lemon may
be considerably larger, but very variable in shape. The fruit is normally mid-
green becoming yellow to bright yellow when ripe. The peel is relatively thin,
smooth and tight; or rough, thicker and more loosely attached. The seed is
small, ovate, and may be absent in smooth lemon, numerous and viable in
rough, with no dormancy period. Seed from Eureka grown in Pakistan
yielded 28% oil, containing ten fatty acids, including palmitic 41 % and oleic
34% (Sattar et al., 1987).
Smooth lemon fruit is used mainly for juice and although very acid has a
multitude of uses. Fruit is seldom eaten although slices may be added to
many foods and beverages. Rough lemon fruit has virtually no commercial
use except as a seed source. Whole smooth lemons may contain to 1.6% oil
(2.2% in peel only) but cold-pressing normally returns only 66-75% of oil
Rutaceae 467
present, Steam distilling produces a higher yield, but the oil has different
characteristics and odour and is considered inferior. By contrast, Meyer peel
oil contains over 90% limonene, but also thymol to 6%. The major compo-
nents of cold-pressed peel oil are shown in Table 12.22.
Ecology
Lemon flourishes in a Mediterranean-type climate with an average tem-
perature of 20-30°C, but selected cultivars have greater cold or heat toler-
ance. While the climatic requirements for commercial lemon production
may be restrictive, lemon is very adaptable. In Australia, self-sown rough
lemon trees have become widely naturalized along creek banks in dry areas,
the fruit known as bush lemons. Mature trees tolerate low temperatures for
very short periods but frost causes extensive damage to young shoots, and
seedlings and young trees invariably succumb. Frost can limit production in
otherwise suitable areas, as in Spain where a late frost can drastically reduce
a crop and severe frost decimates groves. Selection within cultivars grown at
the northern limit of their range could produce types more frost resistant,
since there is a relationship between the amount of free proline present in
seedlings and their ability to withstand freezing (Yelenosky, 1985; Kushad
and Yelenosky, 1987; Zambakhidze et al., 1989). Lemons in frost-prone
USA
New
A B C Italy Hungary India Algeria Zealand Guinea USSR
Specific 0.849 0.845 0.846 0.849 0.880 0.852 0.850 0.854 0.849 0.856
gravity 0.855 0.852 0.851 0.858 0.886 0.860 0.857 0.855 0.862
(20°C)
Refractive 1.474 1.474 1.473 1.474 1.479 1.472 1.4754 1.474 1.473 1,471
index(20"C) 1.476 1.476 1.476 1.476 1.481 1.475 1.478 1.476 1.478
Optical t 57O t 59" +
70" t 57O t2O0 t 67' t 68O t 46" t 50" ng
rotation t 65O t 67O t 78" t 65' t 25O t 80" t 53" t70.2c
(20°C)
Solubility 1 :3 1:3 1 :3 1 :3
125 1 :3 1 :I
(95% alcohol) (85%)
Aldehyde 2.3 2.2 > 2.0 3.0 3.0 2.5 3.0 4.0 3.0
as citral (%) 3.0 3.6 4.5 4.0 2.8 4.2 4.5
Residue on 1.5 1.4 < 2.5 < 2.0 ng 2.0 ng 2.5 ng
evaporation (%) 1.8 3.0 2.4 2.7 S
A, California; B, Florida; C, Arizona; ng, not given.
Figures in columns are range. Sources: Published analyses.
468 Chapter 12
regions must be protected and a variety of methods and materials have been
used (Eshankulov et al., 1985; Makharadze et al., 1987). It is, however,
unlikely that under such conditions large-scale lemon production will be
profitable. Partial or temporary protection is more common; straw and black
plastic is used in southern Italy (Pilone, 1987).
A rainfall of 750-1 000 mm is the optimum but lemon will grow well in dry
almost arid regions under irrigation, as in California where the coastal region
has an average of 500 mm, further inland below 100 mm, and in the virtually
waterless Negev Desert, Israel. Lemons from very dry areas often have
thicker peel than those from wetter or cooler districts, the surface wax is
thinner, and this is reflected in wax content of the oils. In regions where
rainfall approaches the maximum or seasonally heavy storms occur, ridge-
planting is effective.
Commercial lemon production is generally below 300 m but trees will
grow well and fruit regularly at relatively high altitudes. The author grew
lemons between 1500 and 2000 m in the Kenya Highlands on the Equator
and it is common in Uganda gardens and upland areas of Tanzania. At
Mazoe, Zimbabwe, large commercial plantations are at 1200 m, and West
African plantations are located north of the humid coastal region. In the
USA, the main growing regions are on the Californian coast and further
inland, and in Arizona. In Sao Paulo, Brazil, most groves are around 500
m.
Published data frequently compares seasonal yield of fruit and oil, and the
wide variation indicates the direct effect of climate. Fruit processed in
Florida from 1968-1972 gave an average oil yield of 6-8 kg/t; California has
much larger production from a bigger geographical area and seasonal
variation in bulk oil is much less since one district has little influence.
Analyses of early, mid and late season oils showed total hydrocarbons
averaged 94% in California, 97% in Florida, while total aldehydes (as citral)
averaged 2.9 and 1.8 respectively (Staroscik and Wilson, 1982). Commer-
cial oil from lemons grown in the three major producing areas of California,
Florida and Arizona are quite distinct, with substantial variation in major oil
components, including beta-pinene, limonene, neral, geraniol, linalool and
geranyl acetate. Some differences are varietal or due to a combination of soil
type and irrigation water, but climatic differences are probably most influen-
tial (Staroscik and Wilson, 1982; Chanukavadze and Kharebava, 1990).
will be noted later. Highly alkaline soils are also generally unsuitable, but
rough lemon was successful as a rootstock on high p H corraline soils in the
Solomon Islands (Caiger, 1987). Soils with a high boron, zinc or magnesium
content are also unsuitable for commercial plantations.
The approximate uptake of nutrients by lemon trees is shown in Table
12.4 and is a guide to the general level of nutrients which must be available.
In New Zealand when unfertilized tree yield (control) was 100, yield from
phosphate alone was 123; phosphate and potassium 133, phosphate and
nitrogen 138, all three 145. Local trials to determine the fertilizers required,
and the type and amount are essential since lemon cultivars react very
differently to similar amounts. This applies not only to tree growth, fruit
yield and quality, but also to fruit oil content and composition.
Lemon generally has a greater nitrogen requirement than other citrus
growing in the same area, since the optimum level in lemon leaves usually
requires the application of 20-50% more nitrogen to produce the same level
as in orange leaves. The reason for this apparently low nitrogen recovery in
lemon remains unclear. Nitrogen usually increases fruit yield, but can also
increase peel thickness and oil content. In Florida 180,325 and 470 kg Niha
gave an oil yield of 9.0, 9.4 and 9.5 kgit fruit, equal to 176, 206 and 216 kgi
ha (Kesterton et al., 1974).
Potassium in balance with other nutrients produces larger fruit with
smooth peel (although peel thickness may be reduced) and more juice with
a higher acid content. Reduction in peel thickness is compensated in terms
of oil production by more and larger fruit. An excess of potassium invariably
reduces fruit and oil yield, and potassium should not be applied unless there
is a deficiency limiting fruit production. Potassium sulphate is preferred to
potassium chloride, since chlorine in the latter can build up to toxic levels
after repeated application.
Superphosphate is generally applied to provide the necessary amount of
phosphate, but where a small suphur deficiency exists, single-
superphosphate will normally correct this imbalance. Lemon trees quickly
show symptoms of trace element deficiency or toxicity, in particular boron,
zinc and magnesium; high nitrogen levels applied to sandy soils will often
increase, or may induce, a magnesium deficiency. In Australia, Eureka
lemon on citronelle rootstock readily shows magnesium deficiency symp-
toms. Other effects of fertilizers are discussed later.
Cultivation
The cultivation necessary to establish any commercial citrus orchard is
suitable for lemon. Pruning is essential to maintain trees in full bearing, its
frequency and extent are governed mainly by local factors. Cultivars such as
Lisbon have a denser canopy which requires more frequent thinning. Eureka
canopy density should be related tohours and intensity of sunlight either to
increase light penetration or provide shade. Hand pruning is common on
4 70 Chapter 12
Harvesting
Fruit yield varies widely; smallholders obtain 6-1 5 t/ha but a large commer-
cial plantation with a density of around 400 treesiha using selected strains
should average at least 50 t/ha annually. In Australia, irrigated Frost Nu-
cellar Lisbon, or sweet orange on Troyer citrange stock at a spacing of 7 X 4
myaveraged 75 tiha fruit (six fruit per kg). Yield increased from the fourth to
the eighth-ninth years and subsequently maintained this level; rising from
5 5 to nearly 200 kg per tree.
Except where labour is still plentiful or cheap, mechanical harvesting is
usual, especially where the majority of fruit is processed. Since fruit for
processing can be handled in bulk less carefully than for the fresh fruit
market, transport, storage and pre-processing operations have considerable
flexibility of timing. Factory storage of several weeks is common at peak
delivery periods, with no major detrimental effect on juice or oil.
Distilling
Lemon peel oil is normally cold-pressed and distilled oil mainly produced
from expressed peel. These oils have different characteristics and odour, and
distilled oil has little commercial importance except as an additive (adulter-
ant) of expressed oil. Leaf oil is obtained by distillation.
In Russia, Meyer peel oil and monoterpene content of oil rose as fruit
matured; in Florida, the highest aldehyde content was in late-season and
lowest in early-season fruit, while high aldehyde content was related to high
oil content, and this seasonal variation can be used to advantage by pro-
ducers. Early maturing lemons from local trees in Sicily, prirnojiore, are
bulked irrespective of variety to yield a top quality oil, sold a t a premium or
used to boost later, lower quality oils (Cotroneo et al., 1986). There is
considerable regional variation in local oil yield and characteristics, and also
within-season variation. Pressed oil yield is highest at first harvest,
0.5-0.6%, and falls to half this at last picking. An additional 0.05-0.1 % of oil
is obtained by steam distilling expressed peel. Accepting citral as the most
4 72 Chapter 12
Table 12.24. Main characteristics of cold-pressed lemon peel oil and terpeneless oil.
leaf oil, and a very detailed analysis is contained in Baaliouamer et al. (1985)
and other oils reviewed (Hortobagyi and Szabo, 1958; Lawrence, 1993).
The main constituents are normally, in per cent, limonene to 30, beta-
pinene to 25, geranial to 15, with a total aldehyde content calculated as citral
to 45, but these proportions may not occur in the same sample. Terpeneless
petitgrain oil is prepared by vacuum-distillation with a yield around 50%
and is normally produced only on demand. The main use is in flavourings,
where its rich freshness is an asset, and in perfumery.
Lemon essence is obtained by fractionating and condensing vapours from
initial stages of juice production, and has a pleasant fresh aroma character-
istic of the fruit. Although not widely utilized, essence can be used as a
flavour enhancer and masking agent in industrial and household aerosol
products such as insecticides and detergents. A large number of lemon
aromas or distillates are available commercially derived wholly or partially
from lemon fruit or oil, while others are synthetic products.
Meyer lemon juice is similar to smooth lemon and used in both lemon
concentrate and lemon drinks, but the oil is quite different. Meyer lemon oil
has a low aldehyde content with only traces of the important flavour
constituents geranial and neral, but up to 6% thymol, and is thus different in
odour and taste to smooth lemon oil. Thymol imparts an off-flavour to end-
products and consequently the oil has limited value in the traditional lemon
oil market. The oil could, however, be used as a raw material for thymol
extraction, or in products where a thymol flavour or odour is desired, since
potential oil production is high. Meyer leaf oil is not produced commercially;
the major component is limonene to 60%.
A lemon scented oil available commercially is verbena oil distilled from
foliage of Verbena tnphylla L‘Herit., which has a citral content to 30%, a
fragrant lemon odour, and blends well with other citrus oils. Another
‘verbena’ oil is distilled from foliage of Lipia citriodora Knuth., native to
South America. Most of the commercially available verbena oils are syn-
thetic blends, since the true oil is produced in very small quantities and very
expensive.
Grapefruit
Grapefruit is the only citrus species native to the New World and probably
originated on Barbados from a natural cross between introduced parents
sometime in the seventeenth century; it has since spread around the world
(Kumamoto et al., 1987). Hybrids between grapefruit and other citrus
produce essential oils of varying composition (Ruberto et al. , 1994).
4 76 Chapter 12
Botany
Citrus paradisi Macfad. (syn. C. decumana Murr.) is considered a stabilized
hybrid between C. maxima (C. grandis) pummelo, and C. sinensis sweet
orange. A proposed classification for C. maxima and C. paradisi by the
Vavilov Institute placed them in Cephalocitrus (Karaya, 1987). The original
confusion between grapefruit and pummelo (or shaddock) was due to
misidentification. The origin of the common name grapefruit has long been
disputed (Scora and Nicolson, 1986).
Grapefruit is a large, vigorous tree to 30 my with a single trunk, many
branches, and a round to blunt conical shape if left unpruned. The branches
are sturdy, the twigs angular and the young shoots slightly pubescent. The
leaves are large, blunt-ovate sometimes pointed, with the margins round-
toothed, and are a dark glossy green above, lighter below. The petioles are
broadly winged, the wings forming a small heart-shaped secondary blade
below the main blade apex where the leaf joins the stem. An oil obtained by
steam-distilling leaves had the following major constituents (per cent):
sabinene 55, ocimene 10, linalool 15, gamma-terpinene 2.5, citronellal 5 .
The relative proportions in leaf oil vary substantially between young and
mature leaves, and between leaves sampled progressively through the grow-
ing season.
The flowers are large, up to 5 cm diameter, white, singly or to 20 in
clusters on small racemes, with four to five slightly reflexed thick petals and
25 stamens. The flowers are usually very fragrant but the strength varies with
the cultivar, and flowers are distilled to obtain Neroli grapefruit oil, the
major constituents of which are sabinene, myrcene, limonene, beta-
ocimene, linalool and terpinen-4-01. In contrast, oil produced in Vietnam
reportedly from C. maxima flowers contained 35 components, including in
per cent: nerolidol 29.3, limonene 18.2, linalool 16.4, farnesol 15.7
(Nguyen et al., 1991).
The fruit is large to 15 cm diameter, generally spherical but often
compressed laterally, light yellow to orange, and shape and colour are a
cultivar characteristic. Fruit is normally borne in clusters and, as total weight
per branch can be high, heavily laden branches may break or fracture in
strong winds. Controlling the time and extent of flowering can reduce this
problem, which is particularly dangerous in young trees (Taron, 1992).
Fruit characteristics differ with the cultivar; the peel may be thick to very
thick, the juice slightly to very acid, sometimes bitter, and cultivars divide
into two groups with whitish or pigmented flesh. The fruit may be seeded or
seedless, and the number of seeds varies between cultivars. In general,
seedless cultivars take longer to mature. For the fresh fruit and juice market,
seedless types are preferred, although lack of seed may also be accompanied
by a loss in flesh flavour. Products obtained from grapefruit and orange
grown in Florida are shown in Table 12.25, but the ratios are very variable.
Processed dried peel and pulp residues are used as stockfeed. Dried seed
Rutaceae 477
contains, on average, 32.2% oil, 44.1% meal and 23.7% hulls (see also
Table 12.3).
The major constituent of peel oil is limonene, while nootkatone derived
from C. maxima gives the oil its characteristic taste. Over 100 compounds
have been detected with monoterpenes accounting for about 95%; including
d-limonene to 95%, alpha-pinene about 1% and octanal to 0.6%. The
percentage of non-volatiles is high compared with orange or mandarin oils.
Substantial variation exists in peel oil from cultivars grown in the same
region, and from the same cultivar grown in different regions. The best
documented is between Californian and Florida grapefruit, while Argentine
grapefruit oil reportedly lacks gamma-terpinene and p-cymene. Peel oil
content also varies between cultivars; in Florida the average per tonne of
processed fruit from cultivars Duncan, Marsh and Ruby Red was 2.56,2.82
and 2.95 kg over a 4 year period (Kesterton et al., 1978). In general, yield of
oil from grapefruit is approximately half that of Valencia oranges in the same
region. The main characteristics of peel oil are shown in Table 12.26.
Ecology
Grapefruit prefers a very warm to hot climate with an average annual
daytime temperature of 2 5 3 0 ° C and warm nights; low temperature re-
stricts or inhibits growth and a severe frost kills trees. A rainfall of 1200-
1500 mm is required for commercial production, with supplementary
irrigation below 1200 mm. Above 1500 mm or in areas of high humidity,
fruit diseases become more frequent and the value as fresh fruit reduced.
Grapefruit plantations are usually below 500 m, but the tree will flourish and
produce fruit to 1200 m in the high-altitude tropics where there is bountiful
sunlight and little frost. Marsh grapefruit grows in East Africa from the coast
4 78 Chapter 12
Cultivation
Cultivation of sweet orange in the same region is suitable for grapefruit
including rootstock and seedling care, harvesting, transport, and processing.
A major difference between grapefruit and sweet orange is at maturity;
grapefruit may be left hanging (held) on trees until required without loss of
quality. Potential yield of fruit per tree is usually high and a major constraint
on achieving this is incorrect plantation management. Since individual
grapefruit are heavy, weight per tree may not be as important as number of
fruit, since fresh fruit are usually sold individually and not by weight; the
converse applies when fruit is processed. The number of fresh fruit per
kilogram should not be less than three for thin-peeled or two for thick-peeled
cultivars, since half a fruit is the normal domestic serving per person,
irrespective of weight.
Smallholder yields may be only 10 tiha, superior commercial plantations
80 tiha but up to 100 tiha. Experimental stations have recorded individual
tree yield to 400 kg, with 440 treesiha yielding 176 tiha, an indication of the
potential for improvement. Trees produce a crop 4 years after planting out,
steadily increasing annual yield to 12-14 years, and yield may continue to
increase to the twentieth year. The rootstock has a significant effect on
general tree growth and health. In Australia, Marsh Seedless on Symons
sweet orange rootstock gave a greater annual yield per tree for the first 14
years than on Troyer citrange. The former then levelled off at 250 kg per tree
but the latter steadily increased to 300 kg per tree in the twentieth year,
equivalent to 120 tiha at the spacing used (Thornton and El-Zeftawi, 1983);
in Cuba there was a significant difference in resistance to disease (Simon et
al., 1994). Tree population has a direct effect on yield, as noted, but
cultivars differ in their rate of growth; some can be more heavily pruned or
skirted and planted closer or in double rows. Since fully laden grapefruit
branches may touch the ground, high pruning reduces the danger of con-
tamination by soil-borne disease, insect pests and snails.
Irrigation techniques not only affect tree growth, but their efficient use in
terms of applied water directly affects plantation profitability; the most
efficient are under-canopy systems. At Bet Dagan, Israel, time of application
and wetted area significantly affected water usage and fruit yield; when 632
mm or 803 mm was applied via drips or sprinklers between April and
November, yield was highest a t 89 and 96 t/ha in the sprinkler irrigated plots
(Bielorai, 1987). On mature trees in Florida, the wetted surface area was
more important than the system used at the same water rate (Zekri and
Parsons, 1989). Determination of water need also allows accurate timing; in
Malawi, a water deficit of 5 cm from field capacity was considered the
optimum to commence watering and watering from full bloom to fruit
maturity was the most effective (Chilembewe, 1985). Watering during the
dry season to 85% of field capacity over the same period in Cuba gave a
higher average fruit yield of 35 t/ha. The total water applied was 2359 m3 in
480 Chapter 12
seven surface irrigations (Pardo et al., 1983). The vast difference in total
water applied compared with Israel indicates the massive saving in water
usage possible. Reclaiming water is another method of conserving existing
supplies (Maurer and Davis, 1993).
Herbicides may adversely affect general tree growth, but with no reported
effect on oil content or composition. Pest and disease control chemicals may
affect fruit and peel directly by contact, or indirectly by controlling a
particular pest or disease; application of aldicarb significantly increased yield
and fruit size in Florida (Roux et al., 1991). Chemicals to promote or control
growth, thin fruit or prevent premature fruit fall, can also affect peel and peel
oil; for example GA may increase peel thickness.
Harvesting
Mechanized harvesting is now routine in developed countries but even in
less developed regions there is increasing need to reduce costs or provide an
alternative to a labour shortage, Somalia for instance (Elmi et al. , 1988).
Tree shape also influences the extent to which harvesting machinery can be
used, and a high skirt without loss of yield is a major contributor to
profitability.
The basic factors affecting oil yield are weight of fruit per hectare and its oil
content, generally under the control of plantation managers. The total yield
of oil per hectare is related to amount of peel and peel oil content, and there
are significant differences in both between cultivars; differences which can be
exacerbated by natural factors, or management techniques. Duncan cultivar
in Florida for instance, had an average oil contendtonne of fruit of 2.8, 2.7,
2.5 and 2.2 kg in four successive years, averaging 2.54 kg. The seasonal effect
on oil content is well known but, as grapefruit is not grown for its oil, this
variation is merely accepted as a fact of life by growers! Variation can be
substantial as shown by figures published by the Florida citrus industry;
average oil recovery per tonne of fruit in succeeding years was 2.5 kg to 3 kg.
Cultivars also differed in their reaction to seasonal variation, the greatest
variation in oil content is often in March Seedless. The oil content of
grapefruit diminishes significantly as fruit matures, and can be five times as
high in early picked fruit as in fruit fully mature and held on trees. There is
thus a conflict between oil and fresh fruit production which must be a
personal decision by individual growers.
Distillation
Oil is generally obtained by cold-pressing fruits, but also by steam distilling
juice or pulp; there are basic differences in the oils and the latter is
considered much inferior. The main characteristics of oils from various
origins are shown in Table 12.26. Remarks in this section without qualifica-
tion refer to cold-pressed oil. A basic difference exists between oil from white
Rufaceae 481
and red fleshed cultivars; the former generally has a higher aldehyde content
and lower evaporative residue than the latter, which also contains a small
amount of linalool. Where large numbers of both are grown and processed in
the same plant, they should preferably be treated separately and the oil also
stored separately and offered as an individual type. Physical and chemical
properties of oil from different cultivars varies within a fairly narrow range,
the most commonly grown cultivars varying least.
Oil composition is more important and can differ substantially between
cultivars and within a cultivar depending on fruit maturity. In Marsh for
instance, the major constituent limonene remains relatively constant, octa-
nal and decanal rise while linalool, undecanal, alpha-pinene and dodecanal
fall substantially. An important constituent is the sesquiterpene ketone,
nootkatone, which in Duncan fruit increased with maturity from 0.065 to
0.8 10%. Nootkatone content also differs between Russian cultivars,
0.10-0.35% (Abkhazava and Kharebava, 1989). Chemical treatment can
alter the rate of accumulation and total nootkatone of oil (Tomas et al.,
1993). For processors to whom the oil is a valuable by-product, mid to end
of season fruit thus contain an oil of higher quality and value. Leaf oil is
obtained by distilling fresh leaves, and its composition and characteristics
vary with leaf maturity. It is normally produced only on demand (Lawrence,
1993).
Mandarin
Botany
Citrus reticulata Blanco. (syn. C. nobilis var. deliciosa, C. deliciosa Ten).
Mandarin and tangerine oranges have been given species status, considered
the same species or merely varieties of one species. In this section, both are
treated as one species and the multiplicity of interspecific or intergeneric
hybrids ignored. The name tangerine is used generally in English-speaking
countries but mandarin elsewhere, due mainly to the methods of marketing
and advertising fresh fruit. The various mandarin essential oils are well
known, but tangerine is of no significance; thus all reference in this text
without qualification refers to mandarin and mandarin oil.
Rutaceae 483
Segments are also canned or bottled, but the juice has a much smaller
market. Fruit may be seeded or virtually seedless (satsuma) and the number
of seeds per fruit is generally a cultivar characteristic, although climate and
other factors may also affect the number. Seedlessness is important in the
fresh fruit market, and research continues to produce cultivars with the
minimum or no seeds (Watson et al., 1995). Seeds rapidly lose viability if
stored at ambient conditions. Apartial analysis of seed composition is shown
in Table 12.3. Some cultivars, including Emperor, are favoured rootstock
for other mandarins.
The peel contains oil secreted in glands in the exocarp, and a detailed
description of the glands, oil accumulation and biogenesis of oil components
has been published (Bosabalidis and Tsekos, 1986). There is wider variation
in mandarin oils than in other major commercial citrus species, and thus
general comparisons are of little value. Florida mandarin and Brazilian
mandarin oils are often very similar in their characteristics, although varying
somewhat in composition and odour, due to many Brazilian plantations
being basically established from US material, particularly Dancy. Brazilian
mandarin oils, however, are often mixed (or contaminated) with sweet
orange oil. Mandarin peel oil content is among the highest of commercial
citrus, can reach 17% in some seasons, and in fruit from selected trees
almost 22%. Average oil yield from commercially processed cultivar Dancy
in Florida over 5 years was, in kgper tonne offruit, 6.6,7.6,7.8,6.0 and 7.0,
averaging 7.0 kg (Kesterton et al., 1978).
Ecology
The satsuma group is the most cold-tolerant of all mandarins, and grown
mainly in Japan, China and Russia. There are a number of Japanese satsuma
cultivars, those with greatest cold tolerance grown in northern regions, with
a semi-tropical cultivar on Kyushu. Satsuma mandarin was introduced to
the southern CIS republics, is the most important citrus species and culti-
vars were reportedly undamaged at -2°C (Odanbashyan et al., 1988).
The common group has a low heat requirement and probably the best-
known is Clementine, widely planted in North Africa especially Morocco,
and in Russia selections from the cultivar were little damaged by -6°C.
Cultivar Dancy, intermediate between temperate and tropical types, is the
most important mandarin of the USA, and requires a relatively high humid-
ity to flourish. Its range could probably be extended, since it has been
introduced to several South American countries.
The willowleaf group has a slightly higher heat requirement than com-
mon, and is the preferred type in southern Italy and the Mediterranean coast
generally. The king group is well suited to tropical conditions and an
additional asset is it can be budded on bitter orange rootstock. Cultivar
Ponkan (Indian Nagpur) is probably the most successful truly tropical
mandarin.
Rutaceae 485
Cultivation
The cultivations necessary to establish any commercial citrus plantation are
equally suitable for mandarin. Because of their small size mandarin trees can
be planted at closer spacing, 5 X 5 m or 5 X 3.5 m in double rows is
common. Larger cultivars can be spaced at 7 X 7 m or 6.5 X 3.5 m in double
rows (c. 450 treesiha). In Georgia (CIS), a satsuma cultivar at 1250 treesiha
(4 X 2 m) gave the highest fruit yield (Pirtshalaishvili and Tsereteli, 1986),
but 2500 treesiha was profitable using the low-growing Wase cultivar.
Mechanical pruning is possible, its disadvantages being outweighed by its
generally lower cost, and could become more common with the growing lack
of skilled labour (Oren, 1988). A variety of methods and techniques,
together with their effect on fruit yield, quality or general tree health and
longevity have been reviewed (Tsereteli, 1989). Chemicals can retard
growth, especially out-of-season shoots, and Fosamine in China controlled
summer shoots with less damage to trees than hand pruning, and increased
fruit yield (Meng et al. , 1987).
The economics of weed control should be determined locally, since there
is often little difference in fruit yield between clean weeding, herbicides or
grassed plantations. Herbicides may be necessary to control persistent or
perennial weeds, but must be evaluated prior to regular use as some damage
tree roots (Sharma, 1989; Bajwa, 1990; Kalyan et al., 1995). Since there is
growing consumer demand for organically grown fruit, non-chemical weed
control could become more profitable.
Rutaceae 487
Harvesting
Mechanical harvesting is becoming more common, but some mandarin
cultivars must be hand harvested including the popular Early Imperial.
Fruits of these cultivars must not be pulled but clipped to prevent plugging,
caused when the button is pulled out leaving a hole in the peel through to the
flesh. Fruit yield per tree is directly related to standard of plantation
management and cultivar; thus the yield of satsuma mandarins in Japan is
usually very high, since the standard of plantation management is such that
trees are often identified and treated individually! The importance to Jap-
anese consumers of completely unblemished skins, symmetrical shape and
uniform colour ensures growers produce the highest quality fruit, sold at
very high prices. The growing shortage of rural workers however, is forcing
Japanese growers to mechanize, and a variety of methods and systems have
been introduced (Anon., 1991).
488 Chapter 12
Distillation
Mandarin oil is usually obtained by cold-pressing peel, or steam distilling
expressed peel and pulp, occasionally from pulped whole fruit. The resulting
oils differ in their characteristics, and steam-distilled oils are considered low
quality and frequently used to adulterate cold-pressed oils. Oils vary in their
main characteristics, composition and odour and there is also a substantial
difference between cultivars. Leaf oil is also obtained by distillation, and
composition can vary substantially according to the month leaves are
picked.
Although peel oil can be obtained from almost any mandarin cultivar, in
practice commercial oil production is limited to a very few regions which use
the most favoured local cultivar. In the USA, this is Dancy, in the Medi-
terranean, Clementine and Willowleaf, and in Japan, the satsuma group.
The main characteristics of oils from various origins are shown in Table
12.28.
Major differences exist between mandarin cultivars in the ratio of peel to
fruit, the oil content of peel and oil characteristics, and this genetic variation
is the most important natural factor affecting oil yield in a particular region.
Since specific cultivars are well adapted to a particular area and thus the
obvious choice, regional oils differ. Climate also has a modifying effect and
oil from the same cultivar grown in different countries, sometimes regions,
will also usually differ in certain characteristics or composition. Complicat-
ing oil comparisons are the large number of hybrids whose oils are often
designated mandarin (or tangerine) but which differ significantly from
Rutaceae 489
Table 12.29. Characteristics of mandarin peel oil from different cultivars and regions.
Sweet Orange
Botany
Citrus sinensis Osbeck (syn. C. aurantium varinensis L., C. aurantiurn Lour.),
commonly known as sweet orange, is usually divided into three main groups,
blood, navel and common, with some authorities adding Spanish (Medi-
terranean). Blood oranges have pink or red flesh caused by anthocyanin;
navel oranges have a small secondary fruit pushed to the top of the primary
fruit producing the characteristic appearance; all other cultivars will be
designated common. The most important common cultivar is Valencia
because of its adaptability; in navels it is Washington; but in blood oranges
there is no one cultivar occupying a similar predominant position, although
St. Michael and Maltese are well known. However, so important is orange
production to a particular region that most have developed a specific
cultivar.
Sweet orange is usually a medium-sized, dense, conically headed tree with
glossy green leaves, the golden fruit carried singly or in bunches. Tree height
between 8 and 15 m is a cultivar characteristic, but dwarf types have been
produced. There is normally a single trunk, much branched often from low
on the trunk. Branches may be more or less spined and this is a cultivar
characteristic, although on many cultivars young twigs may be spined
initially. Spines also vary from short to long, to 5 cm, sturdy or slender,
pointed or blunt, but on many cultivars are slender and blunt.
The leaves are glossy green above, lighter below, narrow, ovate-oblong,
acute, base rounded, sometimes slightly toothed, 7.5-10.5 cm X 2.0-3.0
cm, on short petioles with very narrow wings. Petitgrain oil is obtained by
492 Chapter 12
(Tables 12.3 and 12.25). Dried ground seed can be processed into stockfeed
and other products.
Ecology
The main groups of orange cultivars are also suited to different environ-
ments. The navel group is less vigorous, requires a dry warm temperate
climate, and is frequently grown under irrigation. In the tropics it will usually
not be successful below 1000 m, sometimes below 2000 m, and in East
Africa for instance it grew and fruited well in the highlands around 1500 m
but was a commercial failure at the coast. The lower limit for commercial
navel orange production in Colombia is 2000 m.
The common group has the highest heat requirement, but individual
cultivars have greatly extended the group’s range; Valencia for example is
very adaptable and can be grown successfully from sea-level to 1500 m,
occasionally higher in the tropics and semi-tropics. The highly successful
plantations established in Guinea, West Africa, were between 800 and 1200
m although orange trees can be found growing almost to sea level. The main
effect of a fall in average temperature or rise in altitude is to lengthen the
period to maturity. Some cultivars are successful in a relatively restricted
environment, Shamouti in coastal Israel and Egypt is an example; others
such as Hamlin in Florida and Brazil require more humidity. India, with its
wide range of environments, has a number of important cultivars; Pineapple
and Valencia in the cooler north, Mosambi in the centre and west, Sathgudi
in the more tropical south.
Blood oranges generally require semi-tropical to tropical conditions and,
where this orange is grown extensively, a local selection has generally proved
the most successful; Blood Red in India, Kwata in Surinam, Ruby more
widely. The degree of flesh colour in a particular cultivar frequently de-
creases as the mean maximum temperature increases; for example the
relatively pale-fleshed cultivar Kwata produced true blood-coloured oranges
when grown in California.
A major limiting factor is frost, and where this is common commercial
orange production is hazardous, although frost damage is less if trees are
healthy. In Florida, trees with adequate nitrogen levels were more frost
resistant, while a deficiency of nitrogen and potassium reduced resistance
(Koo, 1985). Thin-peeled cultivars are liable to sunburn which affects flesh
and oil content and, where sunburn is common, a heat-tolerant cultivar
should be grown and pruning on the sunward side of trees reduced.
Interplanting with tall but narrow-canopied trees can also be effective,
provided these are planted to give the greatest protection in the afternoon.
Total rainfall and its distribution directly affect fruit production, fruit and
oil composition. The general effect of humidity on a specific cultivar is to
produce fruit with thinner peel and higher juice content than in arid
conditions, as is to be expected. Peel oil content may also be higher, but oil
494 Chapter 12
produced in a very dry year may have a harsher note. Moisture stress affects
fruit development generally, the detrimental effect being greatest during the
period of fast growth from mid-maturity. Damage to peel in this period is
frequently irreversible. In Florida for instance, rainfall directly affected
aldehyde content of Valencia oil, and a 14 year graph of monthly rainfall was
almost similar to one showing oil aldehyde content. Other cultivars followed
this general pattern, but cultivars with a normally low aldehyde content
showed least variation. Less detailed data from other countries tends to
support the rainfall/oil composition relationship, but this may not be so clear
cut and include factors other than available soil moisture, such as tem-
perature or relative humidity.
Regional oils vary considerably in main characteristics and composition;
total aldehyde content from a particular cultivar grown in different regions of
the same country can also vary substantially, as can the ratio of various
aldehydes, esters and alcohols (Table 12.30). Time of harvesting also affects
aldehyde content and proportions, noted later, while seasonal conditions
usually modify all oil components.
Table 12.30. Aldehyde content of some orange oils from the USA.
Soils with a p H 5.5-6.5 are the optimum, to p H 9.0 have been successfully
utilized, but rootstock then becomes increasingly important. In Australia,
soils of p H 8.0-9.2 were planted to Valencia orange on three rootstocks, and
fruit yield compared over 11 years. Cleopatra mandarin and Rangpur lime
rootstocks showed the greatest tolerance and gave highest yield. These two
rootstocks also differ in their ability to extract soil nutrients, and this must
also be a factor in rootstock selection. Valencia oranges grown on Cleopatra
rootstock in the Caribbean had a leaf content high in calcium and magne-
sium but low in potash; Rangpur rootstock produced the opposite effect.
Saline soils are generally unsuitable, but saline or chlorinated water can be
used for irrigation provided the composition is determined, and special
management techniques applied (Cole, 1985; Dasberg et al., 1988).
In many countries, the choice of fertilizer is frequently influenced more by
price and availability than suitability, and in these circumstances fertilizer
trials will establish the optimum amount. Soil type or a change in cultivar
grown or rootstock used can affect a fertilizer programme. Farmyard or
other organic manure is often applied to orange orchards in less-developed
regions, but seldom in the main producing areas as its well-documented
disadvantages far outweigh the value of the nutrients supplied. High yielding
orange trees require substantial amounts of plant nutrients as indicated in
Table 12.4 and at least this level should be available. Fertilizers should
preferably be broadcast within the canopy area, but at the drip circle if
applied via irrigation equipment. Fertilizers have a direct influence on peel
oil content but no significant effect on composition at the levels normally
applied.
Nitrogen requirement is normally less than for lemon, grapefruit and
mandarin in the same area. An excess often produces thick-peeled fruit with
a lower juice content, but the degree to which this occurs is modified by
rootstock and cultivar. The type of nitrogen fertilizer applied is usually not
important; on mature double-planted and irrigated Navel and Valencia
oranges in Australia, urea (46% N) at 240 kg, ammonium nitrate (34% N)
a t 325 kg and ammonium sulphate (21% N) at 525 kg, all per ha, produced
the same fruit yield. On some soils, however, regular application of ammon-
ium sulphate can lower pH, reduce phosphate availability and thus adversely
affect fruit quality, peel becoming rough and puffy with lower oil content.
Slow-release nitrogenous compounds are becoming more popular especially
where leaching is high.
Potassium availability is frequently low in many orange growing soils, but
the amount applied should be sufficient only to correct a deficiency or
ensure a balance between applied nutrients, as was demonstrated in Egypt
(Fawzi et al., 1990). Lack of potassium results in small fruit; an excess
reduces fruit quality, especially in Valencia, may also reduce cold-hardiness,
and this is exacerbated if there is also a nitrogen deficiency. Orange trees are
496 Chapter 12
Cultivation
Orange production on a commercial scale is a major industry, and new
entrants or those wishing to upgrade their operations are advised to seek the
services of one of the many highly skilled and experienced consultants
available. It should be stressed that orange oil production is normally a by-
product of large-scale fruit processing and for this reason is unlikely to be
commercially viable in its own right, unlike sour lemon or lime.
Damage by pests or diseases to leaves or fruit can affect oil content,
sometimes oil composition. Fungal infections reduce leaf oil content mainly
by reducing leaf area, but may also affect oil components including linalool.
Susceptibility to the scale insect Aonidiellu aurantii in Egypt was correlated
with high hydrocarbon content of peel, while resistance increased with
higher terpene alcohol content (Salama and Saleh, 1984).
Probably the most important factor influencing peel oil content or compo-
sition is genetic via rootstock and scion (or bud), but as oranges are not
grown to produce oil but fruit or juice, selection is based on these characters.
Orange cultivars or selections whose fruit has yielded very high peel oil
content under experimental conditions have generally proved unsuitable for
commercial orange production. Therefore techniques or methods used to
improve oil yield must be applicable in commercial orange plantations
without significantly affecting fruit yield or quality. Consequently, most data
refer to methods of influencing peel oil content andior composition of the
most common cultivars.
The importance attached to oil yield is thus an indication of the relative
price of orange oil to other plant operations. Its major effect is to reduce cost
of orange juice extraction, and is of greatest significance in very competitive
markets. For this reason most data are from major juice-producing regions,
and may not be relevant elsewhere. Rootstock, bud or scion have a direct
effect on peel oil content, as demonstrated in Florida where oil content of 34
bud selections yielded 5.0-7.2 kg oilit fruit; nineteen different rootstocks
gave a much lower variation, indicating that budwood was more important
than the rootstock (Hendrickson et al., 1970). The rootstock, however, can
have more significant influence within a specific cultivar; when a clone of
Rutaceae 497
Valencia was budded on trifoliate orange, sweet orange or Rangpur lime, the
important aldehyde content of peel oil was 1.2, 1.44 and 1.73% respectively
(Kesterton and Braddock, 1977).
Harvesting
Maturity affects fruit composition, peel oil content and oil composition as
noted. Oil content increases as peel and contained oil cells develop, and total
oil increases as fruit size increases since there is a direct relationship between
surface area and number of oil cells. Oil composition changes, but is often
masked by techniques used to extract oil (Kesterton et al., 1971). This
applies generally to orange cultivars, irrespective of whether they are early,
mid-season or late maturing; a variation most clearly demonstrated at any
major orange-processing plant. Oil yield increases as the season progresses
with one or more peaks if several cultivars are successively processed. Over-
mature fruit remaining on trees retains its oil content but, when processed,
often gives a very low yield due to softening of peel which is more difficult to
process.
Peel oil yield varies substantially between cultivars, regions, and season-
ally. In Florida for instance the 5 year average for five cultivars was 3.5-6.2
kg oil per tonne of fruit, and oil yield of individual cultivars varied 20-30%
within a season, with Valencia generally outyielding all others. The average
commercial yield in Brazil was 4 kg from cultivar Pera; Sicilian yields
averaged 4-5 kg, elsewhere in Italy 4 kg; in Spain the average is 2.5 kg but
individual plantation yield is often twice this. When oil was manually
extracted the yield was frequently quoted per 1000 fruits; thus 1 kg oil in
Sicily required 2000 fruits, in French West Africa 1200-1500, in Tanzania
1500-2000.
Nearly 150 compounds have been detected in peel oil, some in minute
quantities. The main constituents are the terpene hydrocarbons at around
95% and aldehydes at 2%. Oil composition can change during harvesting as
noted, especially the important aldehydes (Table 12.31). Changes can also
occur in oil composition between harvesting and processing; fruit stored at
extraction plants for up to 7 days generally has substantially higher ester
content, slightly lower aldehyde content and higher evaporation residue.
Distillation
Oil is obtained by cold-pressing peel or whole fruit, or by steam distilling the
residues from segmenting or juicing, and the two oils are quite different in
composition and characteristics (Table 12.32). Peel oil content varies from
0.5-2.5%, but up to 3.5% in selected cultivars. Oil content, however, should
not be confused with oil yield from commercially processed fruit, which is
usually below two-thirds of peel oil content.
The maturity of the fruit arriving at a processing plant affects oil composi-
tion and quality, and changes in oil composition as fruit matures is shown by
498 Chapter 12
regions or countries. Analyses and other data quoted herein are from the
main orange-growing areas, and may not necessarily apply elsewhere. In
general, these oils conform to official standards or industry requirements,
and are a good guide to a saleable oil.
Neroli Portugal is obtained by steam distilling sweet orange flowers
mainly in Portugal and Spain, but is seldom available in any quantity since
the fruit is more important. Leaf oil is obtained by distilling fresh leaves, and
is also not normally available in commercial quantity.
Orange oils are used primarily in flavouring, especially soft drinks, and
many food products. Thus those oil components considered to influence
flavour are very important and their proportions in a specific oil normally
determines its price. Large extraction plants blend the various oils produced
during a season, and offer a standard proprietary product of guaranteed
purity and composition. Whole orange oil imparts a true orange flavour, but
in excess there is risk of product deterioration due to formation of rancid
odours, and a tendency to separate from other ingredients due to insolubility
in water. Addition of small quantities of butylated hydroxyanisole (BHA) or
butylated hydroxytoluene (BHT) controls these problems. Where whole oil
is unnecessary, concentration overcomes most technical problems. Cold-
pressed orange oil is frequently adulterated with distilled orange oil, other
distilled citrus oils or synthetics. A restraint on the level of adulteration is the
generally low price of orange oil compared to other oils. Interestingly, orange
oil is a potent ant-killer but the method of action is unclear.
Distilled orange oil is produced by steam distilling peels, expressed or not,
and juicing residues. The oil is colourless to very pale yellow with a fresh,
sweet, rather weak odour of low tenacity. It is used to adulterate expressed
oil or as a blender or a poor substitute for other citrus oils. A major use is as
raw material for production of d-limonene (noted later).
Terpeneless orange oil is obtained by various methods (Moyler and
Stephens, 1992), but is seldom available commercially in any quantity.
These oils are colourless to very pale yellow, with a sweet, refreshing,
orange-peel odour, have enhanced keeping qualities compared to whole oil,
and up to five times the flavour strength. The residual products of concen-
tration, which include orange terpenes known as washed orange oil, are
widely used as masking agents in household and industrial products such as
aerosols, scouring powders and dishwashing liquids.
Petitgrain orange oil is obtained by steam distilling fresh leaves and twigs
and varies from olive-green to green-orange, with a light fresh note followed
by a characteristic bitter undertone. The oil is produced in very small
quantities on demand, but in general cannot compete with petitgrain bitter
orange. Detailed analyses of the oil have been published (Lawrence,
1993).
A number of other products are derived from orange oil, usually produced
by manufacturers on demand or for a specific end use. The use of natural
instead of artificial colours in food and drink is increasing due to consumer
pressure; thus carotenoids which can be extracted from orange processing
waste are now more valuable, and an added asset is that their flavour is
compatible with orange flavoured foods and drinks. Orange wax is a by-
product of orange oil processing, and contains useful perfumery com-
pounds. A derivative, I-carvone, is an important source of a synthetic
spearmint flavour, and as the price of I-carvone from orange oil is usually
two-thirds of natural spearmint oil, this use will probably continue. As a
Rutaceae 501
Boronia
Boronia Sm. is a genus of about 95 species occurring throughout Australia,
with the one species native to New Caledonia differing greatly from all
Australian species. Western Australia has 49 species, 45 are endemic.
Boronia are dwarf to tall shrubs, rarely small trees. Branches are glabrous
and hairy. The leaves are opposite, simple or compound, and often aro-
matic. The inflorescence is a dichasium (cymose); flowers may be red, pink,
yellow-green, green, blue, purple, brown or white, valvate or imbricate in
bud. The flowers have four petals, a four-lobed spreading calyx, and eight
stamens. The fruit is a capsule of four carpels, separating to expel the ripe
seeds. The seed is hard, black, waxy or shiny. Only B. megastigma is currently
harvested for its oil.
Botany
Boronia megastigma Nees. (brown boronia) is an upright, dense, sometimes
straggling, small to medium shrub, 1.5 X 1 m in diameter, with numerous
often minutely hairy branches. The leaves are green, compound, trifoliate to
1.5 cm, very narrow, and soft, with a blunt tip. The peduncles are axillary,
short, pendant and profuse. The flowers are very fragrant, variable in colour
from dark to reddish brown or yellow-green outside, light yellow to yellow-
green inside, solitary, bell-shaped and about 1 cm in diameter. Main
flowering in Australia is late winter-early spring. Some factors affecting bud
and flower development have been determined (Davis and Menary, 1983;
Roberts and Menary, 1989a; Day et al., 1994, 1995). Since it became
popular as a garden plant, many forms have been selected for their flower
502 Chapter 12
colour and given varietal names. In this section boronia will be used to
designate B. megastigma, its products and derivatives.
Ecology
Boronia megastigma is a native of coastal areas of Western Australia but now
widely distributed as a garden plant, and currently also grown for its oil in
Tasmania and New Zealand (Smale, 1991). Plants occur in moist sites in
generally dry sclerophyll forest under medium or intermittent shade, or
where substantial leaf litter maintains soil moisture and ameliorates the
generally high temperature, as the roots are normally superficial. Mature
plants are frost resistant but young seedlings often succumb.
Temperature affects both rate of growth and flowering and, in general, a
temperature encouraging a high rate of vegetative growth was detrimental to
flower production, and vice versa. When plants were subjected to a range of
environmental conditions, the maximum flower numbers occurred with a
night temperature of 5-7OC, a short day of 10 hours, and under 50% to
100% sunlight (Roberts and Menary, 1989b).
Cultivation
Pre-planting operations should produce a clean seedbed suitable for hand or
mechanical planting. Seed has a prolonged dormancy period, a thick and
oily testa, is extremely difficult to germinate, and may take 6-10 weeks to
emerge. Thus plants are normally raised by cuttings from selected parent
stock, since there is great diversity of forms with varying flower oil content
and composition. Seedlings can be rooted in pots or a nursery bed and
Rufaceae 503
planted out when bushy and 5-8 cm high, in the southern spring. Seedlings
should not become rootbound in pots, since such plants fail to make
adequate root growth after planting out and usually die.
A spacing of 1.2 m X 7 5 cm giving 11,000 plantsiha will allow plants to
expand to fill the available space when about 3 years old. Since plants
normally grow in strong leaf litter or with similar protection, mulching is
essential to maintain soil moisture and reduce temperature in the root zone.
Weeding is essential in young plantations, should be frequent in mature
plantations, especially after the major pruning, and should be as shallow as
possible. Herbicides can damage the shallow roots and soil inactivated
chemicals are most suitable, but trials are necessary before regular use.
Plants are heavily pruned in the autumn after planting to produce an open-
hearted, well branched shrub, and again after each harvest to maintain this
shape.
Harvesting
Flowers are harvested manually in September-October by a comb raked
through bushes and, as this can be unselective, the inclusion of leaves and
stalks reduces oil quality, since leaves contain oil of very different composi-
tion. Combing is time-consuming and unpopular, and more efficient and
quicker harvesting methods are currently under trial. Harvesting is as early
in the day as possible, for rising temperature encourages a sticky exudate
which makes clean-combing of flowers difficult. Flowers must be trans-
ported to the distilling plant without delay, as they quickly ferment in bulk.
When delay is unavoidable, bags of flowers are placed in a mobile chiller at
5"C, and up to 12 days of such storage is possible without deterioration.
Boronia is generally free of major pests and diseases in the wild and, when
cultivated, only psyllids, mites and scale insects have required control.
Chemical pesticides should not be used once buds begin to expand, as
residues have occured in solvent extracted oil. Trials are essential to estab-
lish the most suitable chemical.
Distillation
Flowers and foliage can be steam-distilled but are usually extracted with
food-grade hexane to produce a concrete, and again extracted to produce an
absolute. Concrete yield is very variable and averages about 1% with an
absolute yield of around 60%. There is considerable variation in the yield of
flowers and concrete from individual plants, indicating that selection could
quickly increase yield of either or both.
The main characteristics of oil from wild plants growing in Western Aus-
tralia were: apparent density (40°C) 0.9091-0.9173, refractive index (38°C)
1.4572-1.4852, optical rotation -2" to -3.5", acid number 27.5-30.5,
ester number 106.7-109.3, melting point 35541°C. The considerable over-
all variation is attributed to seasonal conditions which can differ greatly
between years but composition and characteristics are less variable from
cultivated plants growing in more stable conditions in Tasmania and New
Zealand.
Absolute is produced by alcohol-washing concrete, which also separates
out the wax fractions. Absolute is a green viscous liquid, with a fresh, fruity-
spicy odour and a rich tenacious floral undertone. Its principal constituents
are ionone, eugenol, triacontane, phenols, ethyl alcohol and ethyl formate.
The main use is in food flavourings, with minor amounts used in perfumery.
Since the absolute is extremely expensive and available in very small quan-
tities, it is often adulterated except when obtained from primary pro-
ducers.
Amyris balsamifera L., source of West Indian sandalwood oil, is not related to
either true or Australian sandalwood, and occurs on Caribbean islands and
neighbouring countries of South America, but only in Venezuela, Haiti and
Jamaica has it been commercially exploited. There are no recognized sub-
species although oil obtained from trees growing in different localities was
stated to have slightly different characteristics. This is unlikely and any
differences are probably due to misidentification since the very similar A.
elemijera L. also occurs in the same region, known as bois chandelle, and
serves the same local purposes. Amyris balsamijera is known as bois chandelle
(candelwood) on Haiti, tigua in Venezuela, torchwood in Jamaica, and
balsam amyris in the USA. In most areas where the tree occurs and where it
is not distilled, local people used the wood to make torches and as firewood.
Amyris balsamijera is widely distributed in drier regions, most commonly on
limestone soils near the coast. When undisturbed it tends to develop
thickets, but is under pressure in more populated areas from browsing by
goats. The related A.elemifera also yields an essential oil by distillation of
heartwood, but was formerly used mainly for fence posts and furniture.
Amyrz's balsamifera is a small tree 3-6 m, 75-150 cm DBH, with smooth
bark greyish when young, becoming rougher and deeply fissured with age.
The inner bark is light brown. The trees have a rounded crown of pale green
foliage. The leaves are opposite, compound, with three to seven ovate
leaflets, 40-125 X 12-50 mm, glandular, and aromatic when crushed. The
flowers are small and white, in terminal and lateral clusters. The fruit is a
Rutaceae 505
References
Abkhazava, D.M. and Kharebava, L.G. (1989) The essential oil of grapefruit fruits
of different cultivars. Sub-trop. Kul’tuy. 6, 11 1-1 17.
Agusti, M . et al., (1994) Satsuma mandarin fruit size increased by 2,4-DP. HortS-
cience 29(4), 279-281.
Alessandro, R.T. et al. (1985) Gas chromatographic-mass spectrometric detection
of adulteration of natural lime oils. Assoc. OH. Anal. Chem. 68, 1154-1 159.
Allolli, T.B., Nalawadi, U.G. and Sullikeri, G.S. (1988) Influence of different forest
tree species on yield and yield attributes of Coorg mandarin. Myforest (India).
24(4), 233-240.
Alva, A.K. and Chen, E.Q. (1995) Effects of external copper concentrations on
uptake of trace elements by citrus seedlings. Soil Science 159(1), 59-64.
Analytical Methods Committee (1 984) Application of gas-liquid chromatography to
analysis of essential oils; Part 11. Analyst 109, 1343-1360.
Anon. (1984) Essential oil of lemon obtained by expression. Analyst 109(10),
1353-1 355.
Anon. (1991) New technology for mandarin farming. Farm Mech. (Japan) 2,
11-25.
506 Chapter 12
Katayama, Y., Nito, N. and Machino, J. (1994) Species identification of the genus
Fortunella by essential oil. Bull. Fac. Agric. Saga Univ. (Japan) 77, 37-45.
Kekelidze, N.A. and Lomidze, E.P. (1989) Analysis of terpene variation in leaves
and fruit of C. unshu during ontogenesis. Flan Frag. J. 4, 37-41.
Kekelidze, N.A. et al. (1984) Essential oil of Satsuma leaves. Khim. Prirod. Soedin. 5,
607-601 0.
Kekelidze, N.A. et al. (1987) Effect of rootstock on the essential oils of orange scion
leaves. Referat. Zhurnal7, 55-58.
Kesterton, J.W. and Braddock, R.J. (1977) Influence of cultural practices, scion and
rootstock selection on citrus peel oil production.In: Proc. Znt. Soc. Citricult. 3,
7 34-7 36.
Kesterton, T W. et al. (1 97 1) Florida citrus oils. Florida Agric. Exptl. Stn Tech. Bull.
749. Tallahassee, Florida, USA.
Kesterton, J.W., Braddock, R.J. and Koo, R.C.T. (1974) The effect of budwood,
rootstock, irrigation and fertilization on the yield of Florida lemon oil. Proc.
Florida St. Hort. Soc. 87 (Nov.), 6-9.
Kesterton, J.W., Braddock, R.I. and Crandall, P.G. (1978) Recovery of citrus by-
products and speciality products from Florida citrus. The Citrus Znd. 59(5),
20-25.
Koo, R.C.J. (1985) Effect of nitrogen and potassium fertilization on winter injury in
citrus trees. Proc. Florida St. Hort. Soc. 98, 53-56.
Koo, R.C.J. (1988) Citrus micronutrients in perspective. Proc. Soil Crop Sci. Soc.
Florida. 47, 9-12.
Krajewski, A.J. and Rabe, E.(1995) Citrus flowering: a critical evaluation. J . Hort.
S C ~70(3),
. 357-374.
Kumamoto, J. et al. (1 987) Mystery of the forbidden fruit: historical epilogue on the
origin of the grapefruit. Econ. Bot. 41(1), 97-107.
Kushad, M.B. and Yelenosky, G. (1987) Evaluation of polyamine and proline levels
during low temperature acclimatisation of citrus. Plant Physiol. 84, 692-695.
Lawrence, B. (1993) Progress in essential oils. PerJ?um.Fluv. 18(5), 43-46, 48-51,
56-67.
Lawrence, B. (1994) Progress in essential oils. Perfum. Flav. 19(3), 61-62, 64-69;
(4), 35-37; (6), 57-58.
Leo, De'P. Ghezzl, M. and Monzini, A. (1 985) Transformatzione e valorizzazionne
della frutta. In Proc 1st Conven. Tech. Alim. Medit. Bari, Italy, pp. 204-240.
Lev, J. (1994) A programme for using the sewage water of the coastal strip for
irrigating citrus groves of the Besor region. Hassadeh 74(9), 980-981.
Leyva, D. and Alvarez, E. (1984) Essential oil production from pruning residues
after mechanical pruning of citrus in Cuba. C. Y.T. en Za Agn'c. (Cit. & To.Fruit)
7(2), 87-95.
Leyva, D. and Alvarez, E. (1992) Production of essential oil from citrus tree
prunings. Cultivos Agroindust. 2(2), 17-22.
Lio, G.M., and di San, M. (1993) Integrated management of bacterial and fungal
diseases of citrus in the Mediterranean region. Bull. OZLBISROP 16(7),
132-1 4 1,
Lodge et al. (1984) The physico-composition of New Zealand lemon oil. J . Sci. Food
Agric. 35, 447-451.
Rutaceae 51 1
Lou, A.I.Q.and Yin, P.Z. (1986) Analysis of causes for low yield of citrus orchards on
beach land and effects of soil improvement on yield. Zhejiang Sci. 6, 270-272,
275.
Lozzi, C. (1963) Essential oils in the Soviet Union. P e Essen.
~ Oil Recd. 3, 179-183
and 1, 17-22 (1967).
Lund, E.D., Shaw, P.E. and Kirkland, C.L. (1981) Composition of rough lemon leaf
oil. J. Agric. Food Chem. 29, 490-494.
Lund, E.D., Shaw, P.E. and Kirkland, C.L. (1982) Components ofMeyer lemon leaf
oil. J . Agric. Food Chem. 30, 95-97.
Maas, E.V. (1993) Salinity and citriculture. Tree Physiol. 12(2), 195-216.
Makharadze, G.P., Kotovich, I.N. and Kutateladze, R.G. (1987) Optical and
temperature protective properties of some covering materials. Subtrop. Kul'tu y .
2, 130-138.
Marais, L.T. (1990) Citrus blight: world research review. Citrograph 75(5),
119-124.
Mathews, R.F. and Braddock, R.J. (1987) Recovery and applications of essential oils
from oranges. Food Tech. 41(1), 57-66.
Maurer, M.A. and Davis, F.S. (1993) Use of reclaimed water for irrigation and
fertigation of young Redblush grapefruit trees. Proc. Flor. St. Hort. Soc. 106,
22-30 (pub. 1994).
McHale, D. and Sheridan, J.B. (1988) Detection of adulteration of cold-pressed
lemon oils. Develop. Food Sci. 18, 525-535.
McHale, D . and Sheridan, J.B.' (1 990) The oxygen heterocyclic compounds of citrus
peel oils. In: Proc. 11th Znt. Con. Ess. Oils Frag. Flav. Vol. 5. Aspect Publishing,
London, pp. 63-7 1.
Meng, J.Z. et al. (1 987) Studies on the reduction by Fosamine of summer shoots and
fruit drop in mandarin. Shanghai Agri. Sci. Tech. 2, 7-9.
Mentonelli, N. (1994) Lime oils from Mexico. Perfum. Flaa. 19(5), 67-68.
Miersch, 0. and Cristobal, 1. (1983) Estimate of the content ofbergapten in fruits of
Persian lime. Ciencias d.1. Agric. 17, 22-24.
Mizuno, M. et al. (1991) Chemotaxonomy of the genus Citrus based on poly-
methoxyflavones. Chem. Pharma. Bull. 39(4), 945-949.
Mooney, P. et al. (1995) Development of New Zealand citrus budwood scheme. In:
Proc. 4th PaciJic Rim Biotechnology Con$ Monash University, LMelbourne, Aus-
tralia, p. 193.
Moyler, D.A. and Stephens, M.A. (1992) Counter current deterpenation of cold
pressed sweet orange peel oil. Perfum. Flav. 17(2), 37-38.
Muller, P.A. (1 966) The bergamot and bergamot oil. Perf. Essen. Oil Recd. 1, 18-25
and 6,368-371.
Muller, P. (1988) Optimum shade requirements in citrus nursery trees. CSFRZ
Inform. Bull. No. 12. South Africa.
Nath, J.C. and Sarma, R. (1992) Effect of organic mulches on growth and yield of
Assam lemon. Hort. J . 5(1), 19-23.
Nguyen, X . D et al. (199 1) The essential oil from flowers of C. maxima. J. Essen. Oil
Res. 3(5), 359-360.
Nigg, H.N. et al. (1988) Ethion distribution in Florida Valencia oranges. Bull.
Environ. Cont. Toxic. 4 1(l), 15 1-1 58.
512 Chapter 12
Nishida, R. and Acree, T.E. (1984) Isolation and characterisation of methyl epi-
jasmonate from lemon. J. Agric. Food Chem. 32, 1001-1003.
Odanbashyan, F.M., Otyrba, Z.M. and Pachkoriya, D.B.( 1988) Frost resistance of
various forms of mandarin. Subtrop. Kul’tuy 3, 90-94.
Onillon, J.C. (1988) Biological and integrated control in citrus groves in the
Mediterranean region. Entomophoga 33(4), 48 1-494.
Ono, S., Iwagaki, I. and Takahara, T. (1986) Relationships between root and leaf
distribution in citrus trees. Fruit Tree Res. Stn. Bull. 8, 23-56.
Oren, Y . (1988) Pruning clementine mandarin. In :Proc. 6th Int. Cit. Cong. 2,
953-956.
Ortiz, A. and Marcide, J.M. (1986) Nomenclatura botanica de 10s citricos. Fruits.
41(3), 199-209.
Ozeki, Y . et al. (1 995) Limonoid glucosides in fruit, juice and processing by-products
of satsuma mandarins. J. Food Sci. 60(1), 186-189.
Pardo, A. et al. (1 983) Preliminary study on the influence of different soil water levels
in Marsh grapefruit. Citn’cosy Otros Frutales. 6(2), 115-125.
Parish, M.E. Braddock, R.J. and Grumlich, T.R. (1986) Chemical microbial charac-
terization of citrus oilmill effluent. J. Food Sci. 51(3), 839-840 and (2),
431-433.
Paul, A. and Cox, P.A. (1995) An ethnobotanical survey of the uses for C. aurantium
in Haiti. Econ. Bot. 49(3), 249-256.
Pilone, h’. (1987) Influence of bioclimate on lemon cropping in the Sorrento and
Almafi penisula. In: Proc.2nd Inter. Cesena Agric. Con5 Inst. di. Colt. Arb.
Partici, Italy, pp. 455-458.
Pirtshalaishvili, S.K. and Tsereteli, G.A. (1986) Planting density in a young manda-
rin orchard. Subtrop. Kul’tuy 3, 94-98.
Rabatel, G. et al. (1995) Robotic harvesting of citrus. In: Proc. Con5 Harvest and
Postharvest Tech. Fruits. Veget. pp. 232-239. Amer. Soc. Agric. Eng., St Joseph,
USA.
Ramishvili, G.G. et al. (1988) Effect of snow cover on overwintering citrus trees.
Subtrop. Kul’tuy 1, 101-103.
Rao, D.U.R., Reddy, K.B. and Reddy, M.L.N. (1987) Preliminary screening of
certain citrus rootstocks for drought tolerance. Ind. J. Plt Physiol. 30(3),
303-304.
Regev, A. and Raber, D. (1992) A computer model to evaluate the profitability of
exchanging one citrus variety for another. Hassadeh 72(6), 672-676.
Rieger, M., Davis, F.S. and Jackson, L.K. (1988) Thermal properties of wraps used
for freeze protection of young trees. HortSci. 23(2), 329-332.
Roberts, N.J. and Menary, R.C. (1989a) A morphological study of bud types in B.
megastigma. J. Hort. Sci. 64(4), 459-501.
Roberts, N.J. and Menary, R.C. (1989b) Environmental interaction between day-
length, night temperature, and photon flux density on growth and flowering in
B. megastigma.J. Hort. Sci. 64(5), 597-604.
Roberts, N.J. and Menary, R.C. (1994) Effect of nitrogen on growth, flower yield, oil
composition and yield in B. megastima. J.Plant Nutr. 17(12), 2035-2052.
Roose, M. and Kupper, R. (1994) Effect of citrus rootstocks on freeze tolerance in
California. In: Proc. 7th.Congress Int.Soc. Citriculture Vol. 1. Florida, USA, pp.
256-258.
Rutaceae 513
ROUX,H.F. et al. (1991) Combined fosetyl-A1 trunk injection with soil application of
aldicarb for control of citrus decline. Plt. Disease 75(12), 1233-1236.
Ruberto, G. et al. (1 994) Profiles of essential oils of new Citrus hybrids. Flav. Frag. J .
8(4), 179-184.
Sacco, T. and Calvarano, M. (1981) Ricerche sul C.latifolia de Brasile. Ess. Der.
Agrum. 51,207-215.
Sakovich, N.J. and Post, S.E.C. (1986) Lemon root distribution in sprinkler and drip
systems. Citrograph. 71(7), 143-144.
Salama, H.S. and Saleh, M. (1984) Components of the essential oil of three citrus
species in correlation with their infestation with scale insects. Ziet. Ang. Entom.
97(4), 393-398.
Samadi, M. and Cochran, L.C. (1975) An unusual flowering and fruiting habit of
grapefruit in the Khuzestan area of Iran. HortSci. 10(3), 593.
Samson, J.A. (1986) Tropical Fruits, 2nd edn. Longman, London, UK.
Sattar, A., Mahmud, S. and Khan, S.A. (1987) Fatty acid composition of seed oil of
C. limon var. Eureka. Pak. J. Sci. Ind. Res. 30(a), 710-71 1.
Sawamura, M. et al. (1983) Aroma components of satsumas grown in vinyl houses
and in the open. J. Agric. Chem. Soc. Japan. 57(9), 683-871.
Sawamura, M. et al. (1 994) Multivariate analysis by measurement of peroxidase and
essential oil components in citrus flavedo. Biosc. Biotech. Biochem. 58(5),
874-877.
Schwob, R. (1953) L’essence de bergamote de la Guinee Francaise. Indust.d.l.Par-
fum. 8(10), 369-373, and 8(11), 406-409.
Scora, R.W. and Nicholson, D.H. (1986) The correct name for the shaddock, C.
maxima. J. Essen. Oil Res. 3(5), 359-360.
Shankaracharya, W.B., Anandaraman, S. and Natarajan, C.P. (1977) Studies on the
chromatographic deterpenation of lemon and lime oils. J. Food Sci. Tech. 14(3),
120-1 25.
Sharma, K.K. (1989) Control ofweeds in citrus 0rchards.J. Res. Punjab Univ. 26(3),
407-409.
Sharma, R., Sharma, E. and Purohit, A.N. (1995) Dry matter production and
nutrient cycling in agroforestry systems of mandarin grown in association with
Albizia and mixed tree species. Agrofor. Systems 29(2), 165-179.
Shaw, P.E. and Coleman, R.L. (1974) Quantitive composition of cold-pressed
orange oils. 3. Agric. Food Chem. 22, 785-787.
Simmonds, N.W. (ed.) (1976) The Evolution of Crop Plants. Longman, London,
UK.
Simon, A., Jimenez, R. and del Valle, N. (1994) Rootstocks for grapefruits (C.
paradisi] in Cuba. In: Proc 7th. Cong. Inter.Soc. Citriculture. Vol. 1. pp.
262-264.
Smale, P.E. (199 1) Investigation into the production of oil from flowers of B . megas-
tignza Nees. Hort. in hTewZealand. 2(1), 24-25.
Soulis, T., Kavlentis, E. and Arvgnitoyannis, I. (1988) Iron, copper, manganese and
zinc content of some processed and fresh citrus juices in Greece. J. Sci. Food.
Agric. 45(4), 373-377.
Staroscik, J.A. and Wilson, A.A. (1982) Seasonal and regional variations in quanti-
tive composition of coldpressed lemon oil. J. Agric. Food Chem. 30, 835-837.
514 Chapter 12
Stewart, I. (1985) Identification of caffeine in citrus flowers and leaves. J. Agric. Food
Chem. 33(6), 1163-1165.
Swaine, R.L. and Swaine, R.L., Jnr (1988) Citrus oils. Perj%uMz.Flav. 13(6), 1-20.
Swingle, W.T. (1967) The botany of citrus and its wild relatives. The Citrus Zndusty.
1, 190-430.
Syvertsen, J.P., Bonan, B. and Tucker, D.P.H. (1990) Salinity in Florida citrus
production.In: Proc. Flor. State Hort. Soc. 102, 61-64.
Tanaka, T. (1954) Species problems in Citrus. Jap. Soc. Prom. Sci. Tokyo, Japan.
Taron, Y . (1992) Early and unwanted flowering of grapefruit. Alon Hanotea 46(5),
340-343.
Thornton, I.R. and El-Zeftawi, B.E. (1983) Culture oflmgated Citrus Fruits. Depart-
ment of Agriculture, Melbourne, Australia.
Tomas, A.O. et al., (1993) Influence of ethylene and ethephon on nootkatone
production in C. paradisi.J. Agric. Food Chem. 41(10), 1566-1569.
Tsereteli, G.A. (1989) Use of pruning in dense mandarin orchards. Subtrop. Kul-
' t u y . 2, 75-79.
Tucker, D.P.H. (1989) Planning, conduct, and evaluation of citrus weed control
programmes. Citrus Veg. Magazine 18, 48, 52, 54-55, 57.
Ulubelde, M. and Mendilicioglu, K. (1985) Studies on the identification of some
species in the genera Citrus and Prunus by their leaf phenolic compounds. Ege
Univer. Ziraat Fakul. Dergisi. 22(2), 43-61.
Valasco-Nunez, F. and Anguiano, H.S. (1983) Response of Mexican lime to
planting distances and fertilizer rates. In: Proc. Znt. Soc. Citnculture 1981
(pub.1983). V01.2, pp. 554-546
Valle, N., Campos, A. and Perez, R. (1985) Scentless sour orange - a new atypical
form. Centro Agricola. 12(1), 138-139.
Verzera, A. et al. (1992) On the genuineness of essential oils. Part XXX: Mandarin.
J. Ess. Oil Res. 4(3), 273-280.
Vora, J.D. et al. (1983) Preparation and chemical composition of orange oil concen-
trates. J . Food Sci. 48, 1197-1 199.
Watson, M. iMooney, P. and Harty, A. (1995) Somatic embryogenesis in citrus In:
Proc. 4th Pacijic Rim Biotechnology Con$ p. 156.
Watt, G.W. (1908) The Commercial Products of India. John Murray, London, UK.
Whitney, J.D. et al. (1 994) Optimizing orange grove factors for fruit production and
harvesting. Trans. A S A E 37(2), 365-371.
Wilson, C.W. and Shaw, P.E. (1984) Quantitation of individual and total aldehydes
in citrus coldpressed oils by fused silica capillary gas chromatography. J. Agnc.
Food Chem. 32, 399-40 1.
Yelenosky, A. (1985) Cold-hardiness in citrus. Hort. Reviews 7, 210-238.
Yelenosky, G . (1991) Minireview. Responses and adaptations of citrus trees to
environmental stress. Israel?. Bot. 40(3), 239-250.
Yelenosky, G . et al. (1987) Chemical bioregulation of growth and cold hardiness in
citrus. In: Plant Cold Hardiness. A.R. Liss Inc., New York, USA, pp. 299-321.
Yin, Y.Q. (1994) The current situation of occurrence and control of nematodes on
citrus in China. Tech. Exten. Plant Prot. 4, 24-28.
Yokoyama, H. et al. (1986) Bioregulation of essential oils of lemon. Food Tech.
40(11), 111-113.
Rutaceae 515
Sandalwood
Sandalwood's generic and common names are probably derived from the
Sanskrit chandana, and sandalwood is known as chandana, chandal and
sandal in India, plus many vernacular names (Watt, 1908). Sanskrit authors
distinguished various types using chandana as a collective name, srikhanda
for white and pitachandana for yellow sandalwood. Arabs, who became
acquainted with sandalwood via their trade with Indian merchants, called it
sandal and passed on the name to Europeans. It is laauala (fragrant wood) in
Hawaii, cendana generally in Indonesia, on Timor haumeni, but sandal on
Sumba Island. The sixteenth century traveller Garcia de Orta noted 'both
yellow and white kinds grow on the island, and a special red kind being
obtained in Tenasserim,'(Burma).
Sandalwood's original home is in dispute but is probably the Indo-Asia
region, as natural stands occur in India, Malaysia and Indonesia, but the tree
is today found chiefly in India and the Indonesian islands of Timor and
516
Santalaceae 517
Table 13.1. Heartwood content of sandal trees in Tamil Nadu State, India.
Table 13.2. Santalum species producing a timber (T) andior an oil (0).
al., 1988; Verghese et al., 1990; Wang et al., 1991). Standard characteristics
of the oil are shown in Table 13.1.
The sandalwood oil of commerce is mainly obtained from East Indian
sandalwood, S. album L. Other oils sold as sandalwood are derived from
other Santalum species or trees not members of the Santalaceae. In this
chapter the designation sandalwood without qualification refers to S. album,
and its oil as East Indian sandalwood oil. Santalum species producing a
timber and/or oil are listed in Table 13.2.
Species other than Santalum which produce an oil incorrectly designated
sandalwood oil, or oil used to adulterated genuine sandalwood oil, or a
timber substituted for true sandalwood are listed in Table 13.3.
Botany
Sandalwood, S. album, is a small evergreen tree to 18 my the trunk 2 . 5 m
DBH when mature. The tree is an obligate root parasite becoming less
dependent on other plants with maturity. A sandalwood seedling can
continue to grow only by becoming attached to roots of other plants. Initial
root growth is directed to this end, and unless seedlings are able strongly to
parasitize adjoining plants, mortality is high. Once parasitism is well estab-
lished, non-parasitic roots develop, and the young tree is then also able to
Table 13.3. Non-Santalum species producing 'sandalwood' timber (T) or oil (0).
obtain nutrients directly from the soil. The amount varies with individual
nutrients; the soil supplying most potash and calcium, the host nitrogen and
phosphate.
Root parasitism in sandalwood was first mentioned in 1871 by John Scott
then curator of the Royal Botanic Gardens, Lucknow, India, and has since
been studied in detail and reviewed (Morris, 1982-1983). Initial root
contact with host plant is by haustoria, and penetration by an intrusive
wedge of cells produced within the haustorium which invades the root core;
this acts as a direct link between the two plants, linkage being between xylem
systems (Kanjit, 1969) (Fig. 13.2). Some 300 plants have been recorded as
hosts to sandalwood, divided into those most and least suitable (Parthasar-
athi et al., 1973). Although sandalwood is main beneficiary of a link there is
some transfer in the opposite direction, and this can also occur by root
contact with non-parasitized plants. Examples quoted included transference
of the bitter principle present in leaves of Stychnos nux-vomica L. to leaves of
adjacent sandalwood trees, and the characteristic taste of sandalwood to
Eugenia spp. growing near by. Unusually, sandalwood can also be para-
sitized, commonly in India by Cuscuata reflexa Roxb.
When sandalwood trees are uprooted or above-ground parts destroyed by
natural causes, roots freely produce suckers often many metres from remains
of the original trunk. Young trees usually coppice from undamaged butts,
but mature trees seldom. Sandalwood roots contain the highest oil content
to 10% by weight, with the following characteristics: specific gravity (25°C)
0.9645; refractive index (30°C) 1.4930; optical rotation (25°C) - 15"; total
alcohols as santalol 95.9%.
There is usually one main stem on undamaged trees, but height and trunk
diameter at maturity are very variable, depending greatly on local condi-
tions. Under natural forest the mean annual girth increment was 2.7-4.6 cm
in Mysore and 1.9 cm in Coimbatore. The bark is generally reddish or
brownish but also greyish brown, sometimes nearly black, rough, and cracks
vertically on mature or aged trees. A characteristic of young trees exposed to
strong and prolonged sunlight or forest fires is cracking and peeling of bark.
A compound isolated from the bark produced hormone activity in insect
larvae by disturbing their physiological development. The compound also
has a chemosterilant effect which is being further investigated. At present,
bark has no commercial value and is usually discarded and burnt.
The oil is contained primarily in the heartwood and larger roots, thus rate
of heartwood formation is of prime importance; however, this occurs early in
some trees and late in others. Trees growing under favourable conditions in
India showed heartwood development at around 10 years of age when
saplings were 3 m and 24 cm DBH. Heartwood formation accelerates
rapidly from 20 years and is at its prime in trees 30-60 years old, 40-75 cm
DBH. The proportion of heartwood to outer wood and bark is 65-85% on
healthy mature trees, and the girth-heartwood ratio from three districts in
Santaiaceae 523
Fig. 13.2. Sketch of haustorium showing method of penetrating host, (CSIR, New Delhi.)
India is shown in Table 13.1. Trees affected by spike disease (cf.) usually
produce poor-quality wood and oil, with a heartwood yield below 25%.
Sapwood and heartwood are normally sharply demarcated, the former
unscented white to whitish yellow, the latter scented, light yellowish-brown
when freshly cut changing on exposure and age to dark reddish brown (see
also products and specifications). T h e environment has a substantial effect
on type of heartwood; from trees growing on dry rocky ground in India it is
harder with a higher oil content than from trees growing on more fertile soils,
whose heartwood is softer with a lower oil content but of greater volume age
for age.
Oil is obtained by hydro or steam distilling pulverized heartwood and
major roots, and its characteristics differ according to method used, types
and proportions of heartwood and roots in a charge. The most important
components are the sesquiterpene alcohols; alpha- and beta-santalols to
90%, alpha- and beta-santalenes to 3%, and other oxygenated sesqui-
terpenes to 4%, but not necessarily in a particular sample. Oil quality is
usually stated in total alcohols as santalol, and should not be less than 90%
(Shankaranarayana and Kamala, 1989).
The leaves are thin, glabrous, decussate, rarely alternate, elliptic-ovate to
ovate-lanceolate, entire, acuminate, and narrowed to a slender petiole. They
measure 1.5-8.0 X 1.6-3.2 cm and thus are veryvariable. They are generally
524 Chapter 13
mid to dark green and lighter below. Shape, size and colour of leaves on
individual trees vary considerably, since all stages of development can occur
on one tree (Fig. 13.3). Although the tree is a true evergreen, large numbers
of leaves may be shed during prolonged rain or dry periods in India, with the
main flush of new leaves appearing with the October monsoon, or light rains
in May. Fresh leaves yield a pale yellow wax, melting point 30°C, containing
approximately 75% unsaponifiables, largely n-octocosanol, n-tricontanol,
palmitone and d-10-hydroxy palmitone.
The flowers are unscented on short multi-flowered, terminal and lateral
paniculate cymes, usually shorter than the leaves. The flowers are initially
pale yellow, becoming blood or purplish red. T h e perianth is a bell-shaped to
obovate tube, four lobed, containing four short stamens, each with fine basal
hairs, the ovary half-inferior; the pedicel is usually similar or shorter than
perianth tube, seldom longer. The flowers are basically insect pollinated
(Bhaskar, 1993). Trees flower profusely from an early age, and in India main
flowering is February-April.
T h e fruit is a globose drupe, 1.0-2.0 cm diameter, purple-black when
ripe, the endocarp tough, ribbed, and containing one hard seed. The sweet
ripe fruits are a favourite food of birds and small animals, which are
important seed dispersal agents. The main fruiting season in India is July-
September with a secondary crop in November-April, and trees produce an
abundance of fruit and seed. T h e seed is red when physiologically mature,
but dark grey when over-mature and there are 4000-6000 per kg. Hot-
pressed ripe seed yield a dark red viscous fluid, to 60%, whose main local use
Fig. 13.3. Variation in leaf shape and size. 1, Sanfalum album; 2, S. yasi. (CSIR, New Delhi.)
Santalaceae 525
is as a drying oil, plus a range of minor industrial uses. The main character-
istics are: specific gravity (25°C) 0.9356; refractive index (30°C) 1.4891;
iodine value (Hanus) 153; saponification value 176; unsaponifiables 8.8%;
sterolic and santalbic acids; the oil thickens on exposure to sunlight or when
heated. Fatty acids in other Santalaceae have been determined (Hopkins
and Chisholm, 1969). Meal from whole seeds has a 2% nitrogen content.
Ecology
The natural distribution of sandalwood is between 30°N and 40"s and 40"E
and 80°W and, although this appears to cover a large area, sandalwood is
restricted in its occurrence. It is probably native to Indo-Malaysia and
occurs naturally from India through Asia to Indonesia, with species of
similar characteristics on many Pacific islands; in China it is most common
in the Lingnam region. Sandalwood could probably have been grown
successfully in many other countries if its semi-parasitic habit had been
known, since there are many records of such attempts.
An annual rainfall of 600-1600 mm is suitable, 850-1350 mm the
optimum but 2500 mm tolerated if the soil is free-draining. Below 600 mm
growth is slow and rate of heartwood formation commercially unviable. In
India, seasonal rainfall is superior to regular small falls. In low-rainfall
regions or during prolonged dry periods the water requirement is sub-
stantially augmented by moisture obtained through the host; thus moisture
stress is often initially noticed by the host wilting. Trees growing in areas of
rainfall above 1500 mm grow faster and form heartwood more quickly than
in lower rainfall regions, and although oil content may also be higher in the
former its quality is considered inferior. Trees are intolerant of waterlogging
when young but less affected when mature; permanently wet soil is unsuit-
able.
Sandalwood grows naturally from sea-level to 1500 m, but in India best
quality heartwood is produced between 600 and 900 m especially in Karna-
taka and Tamil Nadu states. Sandalwood has also been introduced into
many areas of India and, once trees become established, their high seed
production and ease of dispersal quickly results in a large natural population,
especially in government forests or areas protected from grazing and fire.
Sandalwood prefers a dry, sunny climate, but excessive heat or prolonged
periods of hot sunshine severely restrict growth and are frequently fatal to
seedlings; in older trees bark may split, form deep cracks and under severe
conditions may peel exposing the wood. Trees grow well in open scrub or
deciduous woodland with intermittent but not dense shade and are com-
monly found in Indian scrub forests among Lantana spp. and bamboo.
Seedlings benefit from shade, but mature trees seldom thrive under heavy
shade.
526 Chapter 13
Cultivation
Naturally occurring sandalwood is treated basically as a forest tree, and its
semi-parasitic nature is a major factor influencing the methods used. In
general, duplication or improvement of the natural system of reproduction
has proved most successful, summarized as dispersal of seeds, presence of
suitable hosts, protection from fire and grazing. At its simplest, this involves
closing sandalwood areas with protection against grazing and fire; improve-
ment involves seed collection and planting in established sandalwood areas
to increase tree numbers. Following the first showers of the monsoon, up to
five seeds treated with an insecticide are dibbled into the ground under a low
growing preferably thorny host (S.N. Rai, Conservator of Forests, Sandal
Research Centre, Bangalore, personal communication). Seed should not be
sown in heavily grassed areas, since uncontrolled burning is a major cause of
mortality in seedlings and young trees.
Santalaceae 52 7
Harvesting
Harvesting sandalwood for timber or oil involves felling the tree by uproot-
ing, not cutting the trunk. Only mature trees with a minimum 60 cm DBH,
trees affected with spike disease, or dead trees are harvested. Official
regulations governing selection, removal and disposal of sandalwood in
India are comprehensive and enforced. Regulations also control felling in
Australia, Fiji and Hawaii. Selected trees are marked by Indian Forestry
528 Chapter 13
Department officials, and uprooted at the end of the monsoon when soil is
wet to depth. Felling is normally by labourers who dig out trees which are
then winched over; large roots are also recovered. Once felled, those bran-
ches which have no heartwood are lopped and discarded, as are smaller
roots. Branches which have heartwood are removed by sawing as close to the
trunk as possible, to leave the trunk clear. The bark and outer wood is
removed leaving a 2-3 cm covering of white sapwood to protect heartwood
during transport. The trunk and major branches are then sawn into billets of
90 cm or multiples, the sawdust being carefully collected in sacks. Lop and
top, together with unwanted white wood, is usually burnt on site.
Billets are transported to the nearest government storage depot for certifi-
cation and weighing. This is a complicated procedure, as in some instances
the felled tree will be physically reconstructed from billets and branches to
ensure none has been stolen. Final dressing of billets is carried out at depots,
the remaining heartwood is accurately weighed and stamped. Billets are then
stored until sold or processed. The annual loss in storage averages about 2%
by weight of heartwood, 5% of roughwood (i.e. that with its covering of
whitewood). There are no data on how storage affects oil content, but it is
considered that loss under controlled storage is small, especially in rough-
wood billets, and has no effect on oil quality.
Sandalwood is so valuable in India that weighing is essential to establish its
value and, while large billets may be determined to the kilogram, small
pieces are weighed to the gram. Sandalwood was previously graded into
eight types of heartwood and rootwood, but today names usually denote
origin rather than type, i.e. Bombay billets or Malabar peria; Maccassar
sandalwood comes from Timor, Indonesia.
Distillation
The largest billets are normally reserved for wood sculptures and carvings,
only those with extensive cracks or unwanted irregularities plus branches
and roots are used for oil production. Water distillation was used in tradi-
tional stills with wood reduced to a coarse powder, mixed with water in a vat
and heated, the distillate being condensed and separated. One charge of
sandalwood could be distilled many times, sometimes continuously for
several weeks. A typical charge was 20-30 kg of finely divided or powdered
wood, soaked in water for the preceding 48 hours, yielding to 4.5% oil.
While it is claimed by local perfumers that such oil, attar of sandalwood, has
a more delicate fragrance than steam-distilled oil, there is no distinguishable
difference in quality or characteristics, although oil yield from a given
volume of timber is less with water distillation.
Most oil is currently produced in modern state-owned steam-distillation
plants; material to be distilled is progressively reduced in size from billets to
a free flowing uniform powder, to ensure the maximum amount of oil is
released. A charge of usually 750-1000 kg is distilled for 48-72 hours at a
Santalaceae 529
steam pressure of 1.4-2.8 kgicm’. In general, the higher the pressure the
greater the oil yield, but above 3 kg/cm2 the oil can have a harsher note
considered undesirable by buyers. The yield of oil per tonne of powder varies
according to the quality of original material; pure rootwood gives the highest
yield. Oil yield per charge of high-quality material is 4.5-6.5%; crude oil is
bulked for further processing to produce a standard grade. Extraction of
sandalwood using hydrocarbon solvents produces sandalwood concrete.
The main constraint on more widespread use was the high initial and
running costs and the process was abandoned; laboratory extraction gave a
yield of 7% (Adams et al., 1975).
shortening of the internodes with severe reduction in leaf size and the leaves
becoming yellow, then reddish and falling. Root tips and haustoria die.
Symptoms of spike normally indicate severe infection, since the disease can
remain undetected as symptoms may be masked (Parthasarathi and Ven-
katesan, 1982).
The major insect vectors are leafhoppers and jassids, and a very high
population is required to infect sandalwood, only sustained where there is a
high density of suitable host plants. Clearing rank vegetation around sandal-
wood or growing sandalwood in environments unsuitable for these insects
dramatically reduces incidence of spike disease, and is the recommended
method in India. Where spike disease is established it is essential not only to
control insect and plant vectors but to uproot attacked trees. Aerial photo-
graphy of sandalwood forests show circular areas of infection around a
diseased tree. Production of virus-free seedlings by micropropagation tech-
niques has been successful (Sita, 1991). A plant pest which has become
more important is the angiospermic parasite Cuscuta rejlexa, a classic in-
stance of the biter bit!
USA India UK
Specific gravity 0.969-0.975 (25°C) 0.962-0.976 (30°C) 0.971-0.983 (20'C)
Refractive index 1.505-1.508 (20°C) 1.499-1.506 (30°C) 1.505-1.51 0 (20")
Optical rotation -15" to -19'20' -15O to -20" -16" to -20"
Ester content (% 1.6-5.4 Min 2.0 Min 2.0
santalyl acetate)
Total alcohol j% 90.3-97.4 Min 90 Min 90
santalol)
Solubility (viv 70% 3 5-5 0
I I ng 1-5
alcohol)
ng, Not given. Figures in columns are range.
Source: Published Standards.
532 Chapter 13
West Indian sandalwood oil is dealt with in Chapter 12 as its source, Amyris
balsamifera, is a member of the Rutaceae and thus it is not related to either
true or Australian sandalwood.
Australian sandalwood
Several oils designated sandalwood oil or distilled from local Santalum spp.
have been produced in Australia, but only that from S. spicatum has achieved
commercial importance. Other common species are S. acuminatum A. DC.,
S. lanceolatum R. Brit. and S. murrayanum R. Brit. Development of the
South Pacific sandalwood trade as previously noted increased exploitation
of domestic sandalwood, and greatly contributed to the early development
of inland Western Australia (Donovan, undated; Talbot, 1983). Sandal-
wood has been exported since 1845 when 4 t was shipped to Colombo, and
between 1908 and 1930 the annual average was about 8000 t, reaching a
maximum of 15,000 t in 1920 but the market to China subsequently
collapsed and exports ceased. In 1945, the government sponsored Aus-
tralian Sandalwood Co. began exporting sandalwood logs and in recent
years exports have averaged about 1500 t. Harvesting of existing sandal-
wood stands is strictly controlled by the West Australian government
through a system of licensed contractors and quotas (Fig. 13.4).
Santalum spicatum (R. Br.) D C . (syn. Eucaya spicata Sp. & Sumn.) is
typically a small evergreen tree, 3-8 m, 10-30 cm DBH when mature,
greyish in appearance, somewhat umbrageous, leaves in the crown rather
sparse, and irregular branching of the crown is conspicuous. The bark is
rough, fibrous and furrowed on the lower trunk and grey or bluish and
smooth on the upper trunk and branches. The sapwood is pale, the heart-
wood dark brown. The leaves are erect on stout petioles, opposite, narrow
lanceolate, 3-6 X 1-2 cm, dull grey-green and leathery. The flowers are
Santalaceae 533
small, fragrant and numerous on axillary and terminal panicles. They are
green with a red interior and are bisexual. The perianth is in four fleshy
segments and the anthers are two-celled, with the filaments short and
incurved. The ovary is inferior, the style having two to three stigmas. The
fruit is a yellow to orange-red globoid drupe, 2.0-2.5 cm diameter. The
peduncle is short, the exocarp leathery, the endocarp smooth, and the
perianth and disc persist until fruit is ripe. The seed has a high santalbic acid
content. The ground under wild trees is usually littered with past seasons
seeds, known as quandong nuts, and used by Australian Aborigines to
alleviate coughs and colds.
Santalum spicatum occurs in drier inland regions of Western and South
Australia, generally between 22 and 34"S, usually below 500 m, with a mean
annual rainfall of 150-500 mm most falling in winter. Trees can be found on
a wide range of soils, including calcareous loams and sandy red loams. It
generally prefers level country, but can be found growing on hillsides. The
species' parasitic habit requires it to grow in association with other trees and
shrubs, and two common local hosts are Acacia acuminata Benth. and A .
aneura F. Muell. (Boland, 1984).
Sandalwood is harvested by pulling-out trees and roots, and the trunk and
branches above 2.5 cm diameter are retained. The trees make slow growth
and, in the Kalgoorlie district of Western Australia, the estimated time for
trees to reach the legal felling size is 50-100 years. T o preserve existing
stands the government has declared large reserves, with regeneration en-
couraged by seeding and protection against grazing. A Sandalwood Re-
search Institute has been established to develop methods of increasing the
number of wild trees, and establishing plantations. Commercial distillation
534 Chapter 13
leathery. The flowers are small, whitish and four petalled, in terminal
panicles. The fruit is an edible ovoid drupe, 1 cm in diameter and dark blue
when ripe. The oil yield of steam distilled heartwood is 1.75-2.5%, and its
major use is as an additive to local sandalwood oil to ensure compliance with
the optical rotation standard of pharmacopeias. The amount produced is
very small and quite different to sandalwood oil in its characteristics and
composition, since it contains no santalol and main constituent is a sesqui-
terpene alcohol, lanceol. Oil distilled from S. preissianurn Mig. timber
contains a long-chain sesquiterpene diol, the crystalline santal camphor.
Eremophila rnitchelli Benth., the bastard sandalwood, is a tall shrub or
small tree to 10 m, glabrous, viscid and strongly scented. The leaves are
lanceolate to 50 mm and the flowers are highly fragrant, solitary in axils. The
fruit is ovoid, containing four nuts each with one to two superposed seed. Oil
distilled from heartwood yields 1.5-3.0%, and is a viscous, dark yellowish-
brown liquid. It is neither a substitute nor an alternative to sandalwood oil.
Its main constituents are three related sesquiterpene ketones, eremophilone
being the most important. Fijian sandalwood oil is obtained from S. yasi and
contains 49% alpha-santalol and 42% beta-santalol (Smith and Morris,
1979).
African sandalwood
Various oils designated African sandalwood or East African sandalwood
have been described in the literature, but there is considerable doubt
regarding their botanical origin. Most were ascribed to Osyris tenuifolia Engl.
native to Tanzania where it is known as bastard sandalwood (msandali in
KiSwahili), south to Mozambique and perhaps also Madagascar. In Kenya,
Osyris compressa (Berg.) A. DC., now Colpoon compressum Berg., is known as
East African sandalwood, but as Cape sumach in South Africa where its
leaves and bark produce a tanning extract. The timber is not used to provide
an essential oil. In Kenya, C. compressum is listed by the Forestry Depart-
ment as a substitute for sandalwood but the author is unaware of sandal-
wood oil being produced. Osyris tenuijolia billets were exported from Tanza-
nia to India as East African sandalwood, but also not locally distilled. It is
thus unlikely that C. compressurn is the origin of East African or African
sandalwood oil, and that 0. tenuijolia is the source.
Osyris tenuijolia is a much-branched tree to 8 m with thin glabrous
lanceolate leaves, solitary, greenish-yellow inconspicuous flowers, and
berry-like red fruits. The oil yield from distilling wood chips averages 5%.
The only specifications located for East African sandalwood oil are probably
from 0. tenuifolia (Table 13.6). The oil is light brown, similar in odour to
that of S. lanceolatum. The main component is lanceol, and the oil is not an
alternative nor substitute for sandalwood oil. African sandalwood oil was not
536 Chapter 13
References
Adams, D.R., Bhatnagar, S.P. and Cookson, R.C. (1975) Sesquiterpenes of S.
album and S. spicatum. Phytochem. 14, 1459-1460.
Bapat, V.A. and Rao, P.S. (1988) Sandalwood plantlets from synthetic seed. Plant
Cell Reports. 7(6), 434-436.
Bhaskar, V. (1993) Pollination biology and fertilization in S. album. Flora (Jena)
187(1-2), 73-78.
Bigge, J.T. (1823) Report of the Commission of Inquiry on the State of Agriculture and
Trade in the Colony of New South Wales. Parliamentary Paper 136, p. 59.
London, UK.
Boland, D.J. (ed.) (1984) The Forest Trees ofAustralia. Thomas Nelson, Melbourne,
Australia.
Brunke, E.J. and Rojahn, W. (1980) Sandalwood oil. Dragoco Report. 5 , 67-75.
Dragoco Inc., Totowa, New Jersey, USA.
Brunke, E.J. and Tumbrink, L. (1986) First total synthesis of spirosantalol. Progress
in Essential Oil Research. W. de Gruyter & Co., Berlin, Germany, pp.
32 1-327.
Choudhuri, J.C.B. (1963) Sandalwood tree and its diseases. Indian Forest. 89(7),
456-462.
Sanfalaceae 53 7
Degener, 0. (1 945) Plants of the Hawaii National Park. Edwards Bros, Ann Arbor,
Michigan, USA.
Donovan, R.J. (undated) A History of the Sandalwood Industry of Western Australia.
Battye Library, Perth, Western Australia.
Hopkins, C.Y. and Chisholm, M.J. (1969) Fatty acid components of some Santala-
ceae seed oils. Phytochem. 8, 161-165.
Kanjit, J. (1969) The Biology of Parasitic Flowering Plants. University of California
Press, Berkeley, California, USA.
Mahdi, A. (1 986) The biology of S. album seed. Biotrop.Tech.Bull. 1(l), 1-9.
Morris, E.T. (1982-1983) Romantic sandalwood: Its history and uses. Dragoco
Report 4/5, 106-116 (1982); 2, 40-47. (1983). Dragoco (UK) Ltd, Ipswich,
UK.
Nagaveni, H.C. and Srimathi, R.A. (1980) Use of gibberellic acid to assist germina-
tion of sandal seeds. Ind. Forester 106(11), 792-799.
Narayana, K.H.S. and Parthasarathi, K. (1986) HESP - a new essential oil from the
acid hydrolysis of spent sandal heartwood. Perjtium. Flav. 10(6), 60-1.
Parthasarathi, K. and Venkatesan, K.R. (1982) Sandal spike disease. Current Sci.
5 1( 5 ) , 225-230.
Parthasarathi, K. et al. (1973) Hosts of sandalwood. Current Sci. 43(1), 20.
Rao, P.S. and Bapat, V.A. (1978) Vegetative propagation of sandalwood plants
through tissue culture. Can. Bot. J . 56, 1153-1 155.
Rao, P.S. and Ozias-Akins, P. (1985) Plant regeneration through somatic em-
bryogenesis in protoplast cultures of sandalwood. Protoplasma 124(1/2),
80-86.
Rumphius (1941) George Eberhand Rumpf. Herbarium Amboinense, 6 vols. Re-
printed 1971. Amsterdam, The Netherlands.
Scott, J. (1871) Root parasitism by sandalwood. J. Agri. Hort. Soc. India. 2, 287.
Sen-Sarma, P.K. (1 982) Sandalwood -its cultivation and utilization. In: Cultivation
and Utilization ofAromatic Plants. pp. 395-405. CSIR, Jammu-Tawi, India.
Shankaranarayana, K.H. and Parthasarathi, K. (1984a) Compositional differences
in sandal oils from young and mature trees. Znd. Perj'ium. 28(3/4), 138-141.
Shankaranarayana, K.H. and Parthasarathi, K. (1984b) Synthetic sandalwood
aroma chemicals. Perfum. Flav. 9(1), 17-20.
Shankaranarayana, K.H. and Kamala, B S. (1989) Fragrant products from less
obvious sandalwood oil. Perjtium. Flav. 14(1), 19-21.
Shineberg, D. (1967) They Came for Sandalwood. Melbourne University Press,
Victoria, Australia.
Sita, G.L. (1991) Tissue-cultured sandalwood. Current Sci. 61(12), 794-795.
Smith, R.M. and Morris, P.R. (1979) Composition of Fijian sandalwood oil. Znt.
Flav. FoodAdd. 10(2), 57.
Talbot, L. (1983) Wooden gold. Early days of the sandalwood industry. Forest Focus
30, 2 1-3 1. Western Australian Forest Department, Perth, Australia.
Uniyal, D.P., Thapliyal, R.C. and Rawat, M S. (1985) Vegetative propagation of
sandal by root cuttings. Znd. Forester 11 1(3), 145-148.
Verghese, J., Sunny, T . P and Balakrishnan K.V. (1990) Alpha-and beta-santalol (Z)
concentration, a new quality determinant in East Indian sandalwood oil. Flav.
Frag. J . 5 , 223-226.
538 Chapter 13
Walker, G.T. (1968) The chemistry of oil of sandalwood. Perf: Ess. Oil Record.
59(11), 778-781.
Wang, Z.H., Hong, X.K. and Bao, X.S. (1991) Comparative G C analysis of
essential oil in imported sandalwood. Zhongg. Zhongy. Zazhi. 16, 40-43.
Watt, G. (1908) The Commercial Products of India. John Murray, London, UK.
Yu, G.P. et al. (1988) Studies on the chemical constituents of Chinese sandalwood
oil. Yaoxue Xuebao 23, 868-872.
Zingiberaceae 14,
The Zingiberaceae includes some 47 genera, containing approximately 1500
species of perennial tropical and subtropical herbs mainly as ground flora of
lowland forests divided into two subfamilies, the Costoideae and Zingiber-
oideae (Tomlinson, 1956). The Zingiberoideae includes several important
spice producers: ginger, Zingiber officinale ROSC.,turmeric, Curcuma domestica
Val. and cardamom, Elettaria cardamomum Manton. A relative, Languas
officinarum (Hance) Fanvell, is known as ginger root or Chinese ginger. The
genus Zingiber Boehmer contains approximately 80 species but some are of
doubtful authenticity. The genus name is probably derived from the Sanskrit
singabera, via the Arabic zanzabil, and Greek zingiberi.
Ginger
The genus Zingiber is indigenous to tropical India and South East Asia,
Australia and Japan, with the main centre in Indo-Malaysia, although
Zingiber offkinale may not exist as a truly wild plant. Many species yield an
essential oil but only ginger in commercial quantity. Thus any reference to
ginger in the text without qualification refers to Z. oficinale. Some of the less
important locally cultivated species which produce spice or essential oil used
as substitutes for, or adulterants of, ginger and its products are Z. amada
Roxb., Z. elatum Roxb. and Z. zerumbet (L.) Smith in India, Z. cassumunar
Roxb. in India and South East Asia and Z. mioga Rosc. in Japan and Korea
(Taroeno, 199 1).
The original use of ginger was as a spice and so ancient is this use that it
predates historical records, as the number of unrelated local names attest. In
India, it is mentioned in the earliest Sanskrit literature (Korla and Dohroo,
1991), but apparently not in the oldest Vedic works. In China, the first
known record is from Confucius (c. 500 BC) ‘who was never without ginger
when he ate’ (Breitschneider, 1870), and it was also used to treat rheuma-
tism, toothache and malaria. Ginger was introduced to Japan much later and
539
540 Chapter 14
cultivation is relatively recent. Dried ginger was traded from India via Arabia
to the Middle East, but whether it was known in dynastic Egypt before 500
BC is uncertain; it was certainly in use by the time of Alexander and later
became common in Greece and Rome. Ginger is mentioned in Dioscorides
De Materia Medica, and later by Pliny. Initially used mainly as a medicine to
treat stomach disorders in Europe, it gained popularity as a spice and was
well enough known to be included in most herbals from the ninth century; in
the thirteenth to fourteenth centuries it was in general use, together with
pepper and other oriental spices. The rhizomes of Z. cassumunar are widely
used in traditional medicine in the south east Asian region, and their anti-
inflammatory activity is apparently due to their curcuminoid content (Ma-
suda and Jitoe, 1995).
Potted ginger plants were carried on local vessels trading along the China
coast in the fifth century (Rosengarten, 1969) and, as plants are easily
transported, ginger was quickly introduced to countries along the maritime
trade routes of the Indian Ocean and South China Sea (Svendsen and
Scweffer, 1985). In the thirteenth century Arabs carried rhizomes on their
voyages to East Afnca, to plant at coastal settlements and on Zanzibar.
Probably the first European to see ginger growing in its natural habitat was
Marco Polo in the late thirteenth century. Chinese records show that ginger
was cultivated in the Malacca region (Malaysia) in 1416 and, following
European exploration in the region, live plants quickly became familiar to
European botanists. The Portuguese introduced ginger to their West African
colonies in the sixteenth century following Vasco da Gama’s voyage to India
and quickly established a thriving export trade in rhizomes sufficient to supply
their domestic market, becoming independent of other sources. The Spaniard,
Francesco de Mendoza, carried rhizomes to Mexico shortly after Columbus
and ginger subsequently spread throughout the Caribbean islands and Central
America. Spanish colonists exported 1000 t of Jamaican rhizomes in 1547 and
Jamaica remains a leading producer and exporter of high-quality ginger.
Botany
Zingiber oficinale Roscoe is commonly known as ginger in English, adrak in
Hindi, allam in Tamil, haliya in Malay, khing in Thai, kiang in Chinese (in
general) and shoga in Japane’se.The basic chromosome number of the genus
is x = 11; ginger is a diploid with 2n = 22, but tetraploids with 2n = 44 have
been chemically induced. There are many locally important cultivars, at
least 100 in India for instance and, as ginger is vegetatively propagated, local
cultivars often tend to be uniform and develop specific characteristics. The
highly mechanized ginger industry in Queensland, Australia, was founded
on a cultivar imported in the 1920s (Whiley, 198 1).
Zingiber oficinale is an erect, leafy perennial, usually cultivated as an
annual, with purple flowers and a robust branched rhizome growing hor-
izontally near the soil surface (Fig. 14.1). The root is a sympodial rhizome,
Zingiberaceae 54 1
Fig. 14.1. Zingiber o f h a l e , ginger. A, plant with rhizome; B, leaf; C, inflorescence; D, flower; E, flower
in longitidunal section (Purseglove et al., 1981).
542 Chapter 14
roots (Fig. 14.2). The rhizome is firm, the skin corky and small scaled, and
skin thickness and appearance are a cultivar characteristic. Rhizome shape,
size, degree of branching and extent of rooting are directly affected by soil
type and soil reaction. The skin colour varies from buff to very dark brown
almost black; the flesh colour from pale yellow to deep orange red, and both
are cultivar characteristics. Rhizome morphology and histology have been
described in detail (Parry, 1963; Nybe et al., 1979).
Th e rhizome is the raw material for the three main spice products, fresh
(green) ginger, dried whole or powdered ginger, and preserved ginger, with
methods of preparation described in detail (Purseglove et al., 198 1). Ginger
oil and oleoresin are obtained from fresh or dried rhizomes. The oil cells are
distributed in cortex and pith as are the gingerol-containing cells, but the
two are independent (Mangalakumari et al., 1985). Thus peeling has a
major effect on the amount and type of oil produced, as will be discussed. In
addition to the volatile oil, the dry rhizome contains (per cent): starch and
related compounds (4-60), crude protein (10-20), crude fibre (2-10) and a
small amount of minerals.
Within a local cultivar these proportions vary little, since vegetative
propagation limits major variation. Analyses of Indian rhizomes gave the
following percentages by dry weight (average in brackets): moisture 8.5-
16.5 (10.9), crude protein 10.3-15.0 (12.4), crude fibre 4.8-9.8 (7.2),
starch 40.4-59.0 ( 5 9 , water extract 14.4-25.8 (19.6), cold alcohol extract
3.6-9.3 (6.0), acetone extract 3.9-9.3 (6.5), volatile oil 1.0-2.7 (1.8), while
analyses of 30 cultivars gave a fibre content of 3.5-7.6% (Nybe et al., 1979).
A high fibre content is a major disadvantage when rhizomes are used as spice
or preserved.
Starch occurs as ovate to subrectangular granules, ta 45 X 24 microns,
which are normally absent from young rhizomes but occur in pith and to a
lesser extent inner cortex of mature rhizomes. Ginger distillation residue in
India contained 50-55% starch which could be recovered by processing.
Meat cooked with slices of fresh rhizome is often more tender due to the
action of a proteolytic enzyme, zingibain, and 1 g obtained from 40 g fresh
rhizome is sufficient to tenderize 10 kg meat; the antioxidant effect of
rhizome extracts on meat has been noted (Lee et al. , 1986). The rhizome is
further discussed at the end of this section and under harvesting and
distilling.
Shoots arise from the rhizome, and in China, Korea and Japan the small
shoots are especially relished and red-shooted ginger Z. rniogu is cultivated
specifically for this purpose; it is currently being grown in New Zealand for
export to Japan. The ontogeny of shoot apices has been studied in detail
(Shah and Raju, 1975). The effect of growth regulators on ginger has been
investigated (Furutani and Nagao, 1986) but their use is of little practical
value and may leave residues in plant parts Uayachandran and Sethumadha-
van, 1988). The stems (pseudostems) are slender, erect, to 100 cm but
generally 50 cm, 5-10 mm in diameter, glabrous except for short hairs near
the base of leaf blades, bearing 8-12 distichous leaves. The leaves are mid-
green above, paler below, measuring 5-25 X 1-3 cm. The lamina is thin,
subsessile, linear-lanceolate, the base obtuse or rounded, narrowing evenly
to a slender tip, with a pronounced midrib. The ligule is up to 5 mm, broad,
thin, glabrous and slightly bilobed.
The inflorescence is carried on a slim leafless stem, 15-25 cm, generally
arising directly from the rootstock. The scape is slender, 10-20 cm, the
upper sheaths with or without short leafy tips. The inflorescence is cylindri-
cal, cone shaped, 4-7 X 1.5-2.5 cm diameter. The bracts are appressed,
ovate or elliptic 2-3 X 1.5-2.0 cm, green with a pale submarginal band, the
lower with slender white tips. The flowers are axillary in each bract and
short-lived. The calyx is 1-1.2 cm long, thin, tubular, spathaceous and
three-lobed. The corolla tube is 2-2.5 cm long with three yellowish lobes,
the dorsal measuring 1.5-2.5 cm X 8 mm and curved over the anther. The
labellum is roughly circular, 1.O-1.5 cm, dull-purple with a cream-blotched
base; the side lobes are 6 X 4 mm wide, ovate-oblong, free almost to the
base, and coloured as the midlobe. The stamen filament is short and broad,
the anther cream-coloured, 9 mm. The dark purple 7 mm connective is
prolonged to a slender beak-like appendage, containing the upper section of
the style, and the stigma protrudes below the apex. The ovary is inferior and
trilocular with several ovules per loculus. Pollen morphology and structure
of a number of Zingiber spp. has been determined (Theilade et al., 1993).
Flowering is a cultivar characteristic; some flower rarely, others regularly,
544 Chapter 14
Indiana
Jamaican
Whole Peelings Malaysian (average) USA (EOA)
Specific 0.8905 0.870 0.878 0.871
gravity (“C) (30) 0.880 (23) 0.886 (25) 0.882 (25)
Refractive 1.4859 1.4870 1.489 1.488
index (“C) (30) 1.4880 (23) 1.496 (25) 1.494 (20)
Optical -5”12‘ -35” -30” -28”
rotation (“C) (30) -45”(23) -35’ (25) -45O (20)
Acid value 0.90 ng 3.5-4.9 ng
Ester value 6-1 0 ng 16-1 9 ng
Ester number 72.2 ng “g ng
after acetylation
Saponification < 20
value
aCochin; ng, not given: EOA, Essential Oil Association,
Figures in columns are range: there is very wide variation in analyses.
typical, and related to age of rhizome and period of exposure to air and light
during storage. The main oil constituents are, in per cent: zingiberenes to
40, alpha-circumene to 20, farnesene to 9 and sesquiterpene alcohols to 18;
monoterpenes represent about 1% and are important in determining the
final aroma.
The pungent constituents determining taste and flavour of ginger oil have
been widely investigated (Purseglove et al., 198 1; Lawrence, 1984). One
analysis gave the following (per cent wiw): 6-ginger01 11.88, 8-ginger01
1.67; 6-shogaol 0.08, 10-ginger01 2.38 (Chen et al., 1986). Analysis by
electrospray mass spectometry confirmed the leading role of 6-gingerol, and
that supercritical carbon dioxide extraction gave an oil which contained the
pungent and volatile compounds of the original rhizome (Bartley 1995).
These pungent compounds are present in well-prepared rhizomes, but can
be substantially reduced by poor post-harvest handling or improper distill-
ing. A pungency comparison of rhizomes from various origins based on these
constituents is shown in Table 14.3. The substantial differences in yield,
aroma, flavour and pungency between the various types of ginger is reflected
in the character of the resultant oleoresin (McHale et al., 1989).
Ecology
Ginger is native to moist tropical regions of South East Asia, and commer-
cial production is limited to regions lying mainly between the Tropics of
Capricorn and Cancer. So popular is the spice that ginger is now cultivated
as an annual much more widely than originally considered possible, but the
546 Chapter 14
bulk of rhizomes entering world trade still come from the tropics. A rainfall
of 2500-3000 mm well distributed over the year is the optimum, but
1500-2000 mm will produce good crops with supplementary irrigation.
Ginger seldom succeeds as a mainly irrigated crop since the necessary
humidity and soil moisture cannot usually be profitably maintained. Warm
sunny conditions are preferred, but in hot, dry periods especially when the
temperature exceeds 30"C, young unshaded plants may be scorched. A
temperature above 37OC without high humidity can cause severe foliar
damage at all growth stages and if the soil is also dry, rhizomes will be
damaged and may die.
When grown as a small farmer crop and rhus frequently interplanted
among taller plants, ginger benefits from their shade during very hot periods,
but it is generally considered shading per se is unnecessary. The short term
effect of high air or soil temperature is best ameliorated using other methods,
including sprinkler irrigation or heavy mulching. Australian growers apply
water at a low 2.5-3.0 mmihour via overhead rotary sprinklers from 10 am to
3 pm during October-December to protect young plants. A ground tem-
perature of 25530°C is the optimum for initial rhizome growth, and al-
though this occurs at the correct time of planting in tropical regions, it may
not in other ginger growing areas. In Australia, soil temperature at 10 cm
frequently does not reach the optimum until several months after the normal
planting season and delays full emergence by up to 6 weeks. Thus low soil
temperatures at or following planting could be an important constraint on
commercial ginger production where the time available to produce a crop is
limited. Irrigating immediately following planting may also have a similar
effect on soil temperature, but could be an advantage where soil temperature
is above the optimum. Frost at any stage of growth will kill foliage and
rhizomes at or near the soil surface. In areas where frost is common as in the
hill regions of India and China, annual ginger is harvested before frosts
occur.
Zingiberaceae 54 7
Mean moisture (%) Mean fibre (% DM) Mean ash (% DM) Mean ether extract (U DM)
Na 1971 1972 1973 1971 1972 1973 1971 1972 1973 1971 1972 1973
0 80.7 74.4 73.0 5.83 4.18 8.13 5.78 4.83 3.75 3.63 4.68 5.18
20 79.7 73.1 72.4 5.33 3.98 6.85 5.75 3.98 3.30 3.28 4.08 4.78
40 79.6 72.4 73.8 5.00 4.08 580 5.25 4.33 2.88 3.45 3.78 4.73
60 79.8 72.2 73.1 5.53 3.90 6.63 5.80 4.35 3.48 2.95 3.88 5.25
80 79.1 70.6 74.4 5.05 3.58 6.28 5.63 4.00 3.15 3.13 3.90 5.05
akgiarpent.
Source: Owadally eta/. (1981).
Cultivation
Ginger is grown from small farmer level to large-scale fully mechanized
operations, to service a range of markets. This section is concerned only with
ginger grown to produce whole dried rhizomes for oil and oleoresin.
When grown by small farmers, the land is cleared of major vegetation
which is dried and burnt, then hoed and weeds heaped round ginger hills.
On a larger scale, cleared land is first dug over or ploughed to control weed
growth then ridged, either manually or mechanically. These ridges normally
include a layer of manure or weeds at the base, and vary in size to suit the
locality. For the following season’s crop, weeds and manure are placed in the
channel and ridges on each side split to build the new ridge. Several rows of
rhizomes at whatever inter-plant distance is favoured are planted on each
ridge, and may be mulched to prevent over heating and suppress weeds. A
monthly weeding is normally required, and fertilizers applied according to
local experience and cash resources of the farmer.
Ginger is often underplanted in giant castor, coconut, orange and young
coffee and high, light shade can be beneficial; in India ginger planted under
6-year-old arecanut trees gave a higher rhizome yield than under 2-year-old
trees and in the open (Sankar and Muthuswamy, 1984). In mature tree
plantations, the beneficial effect of partial shade can, however, be offset by
root competition. Ginger is also frequently intercropped but this is not to be
recommended where it is a main cash crop. Benefits which may result from
the partial shading and soil cover are generally outweighed by a reduction in
rhizome yield. Interplanting ginger with maize and dwarf beans in South
Africa reduced rhizome yield by 80 and 60% respectively compared to
mulched ginger (Anon., 1979). Intercropping with Sesbania spp. to provide
light shade for young plants was beneficial in southern India, and the
Sesbania later cut to provide mulch. Growers also alternate several rows of
ginger and other crops: in West Africa, maize or yams; in Taiwan and China,
Zingiberaceae 55 1
sand beds in the corner of rooms and covered with more sand. Provided the
heaps are kept dry and examined regularly to remove those diseased, the
rhizomes remain sound until planting.
T o inhibit sprouting, weight loss or to control diseases during pre-sowing
storage, rhizomes can be waxed, chemically treated, or irradiated prior to
storage (Paul1 et al., 1988; Yusof, 1990; Wu and Yang, 1994). Cold storage
is not recommended as rhizome viability is gradually reduced and may be
zero after very short periods below 0°C (Ra et al., 1989). In Australia, good
quality rhizomes are kept in controlled storage to provide material to sow a
nursery in advance of main planting, and also to store superior clonal
material for multiplication. In general, setts 3-6 cm long or 30-60 g have
proved most successful, and many trials have shown a significantly higher
yield is obtained from larger setts (Table 14.5, which also shows the effect of
plant population on yield). There are distinct cultivar differences and in
some regions of India setts up to 150 g gave highest yield, but Jamaican
growers prefer setts with six buds irrespective of size.
Planting is usually manual by small farmers but is easily mechanized, and
a variety of small planters available. On large estates, lands are ridged or
bedded, fertilizer applied and setts placed by fully automatic equipment in
one operation. Depth of planting is not critical and 5-12 cm is common.
However, previous remarks on soil temperature are apposite, and fungicidal
treatment recommended. A row width of 25-30 cm but up to 50 cm
depending on bed size is normal, with in-row spacing of 15-35 cm. Row
spacing is also governed by harvesting machinery and thus three rows are
normally sown per bed and dug together. Accurate in-row spacing is not
necessary provided the optimum number of setts is sown, since ginger tends
to expand to fill the area available. About 8-10 t ha of setts are required on
Harvesting
Rhizomes are left to mature and develop their full aroma, flavour and
pungency when required for oil. Time to harvest thus varies with locality and
cultivar, and is based mainly on local experience. Leaving rhizomes too long
in the ground reduces oil and oleoresin content and may affect pungency;
thus the optimum period for harvesting in a particular locality must be
accurately determined. In southern India and Sri Lanka maximum oil and
oleoresin content occurred 245-265 days after planting; in northern India
180-220 days; in Taiwan and China 230-255 days; in Australia 270-290
days. In Hawaii, oleoresin was constant at 1% (fresh-weight basis) over 34
weeks of rhizome growth, but the important 6-ginger01principle was highest
on a dry-weight basis at 16 weeks (Baranowski, 1986).
The rate of oil increase and ratios of various constituents over time require
more investigation, but fibre development has received considerable atten-
tion and although the initial rate is slow, once rhizomes approach physio-
logical maturity it rapidly increases. In Australia, when green rhizomes reach
approximately 50% commercial fibre content by weight, daily fibre deposi-
tion became 1% by weight. So important is the fibre level in rhizomes to be
used as preserved or green ginger that Australian crops are monitored daily
to ensure the commercial limit of 40-45% fibre-free flesh is not exceeded
Zingiberaceae 555
(Fig. 14.3). Fresh weight of rhizomes gradually increases to 170 days after
planting in Fiji, but rose quickly thereafter to 224 days, with fibre content
also increasing faster over the latter period (Sivan, 1379). Oil content of
mature rhizomes generally tends to be in inverse ratio to fibre content, but
the rate of oil loss during maturation has not been determined. The propor-
tion of oil often tends to fall with the age of the rhizome, and not merely in
relation to total fibre. Both cultivar and season may have a direct influence
and there is evidence that these relationships can be very localized. In
Australia, the volatile oil content normally increases from 2 to 4.5% on a
dry-weight basis over the 4 month harvesting period, but remains fairly
constant at about 0.4% on a green weight basis.
Harvesting is usually manual, plants being carefully lifted from the soil to
prevent rhizome damage. At maturity much of the foliage has withered and
presents few problems. Harvesting in Australia is fully mechanized using
special equipment and this is becoming more common elsewhere. In India,
Table 14.6. Effect of drying method on yield and quality of ginger oil
may sound obvious, but the author has seen a high degree of management
skill wasted on low-yielding types when much higher yields could have been
obtained using another cultivar at lower total inputs! The reason usually
stated is tradition or local consumer preference. For example in Kerala,
India, dry rhizome yield from common local cultivars is 2-6 tiha, average
2-3 t/ha, but nearly 10 t/ha from high-yielding selections available from the
local agricultural experimental farm. The substantial differences in rhizome
yield and oil content between Indian cultivars is well documented, and
indicates the considerable potential to increase oil yield.
Yield from small farmer ginger crops is generally below 3 tiha dried
rhizomes, indeed in Africa and the Caribbean it seldom exceeds 2 tiha. As in
India, yields obtained a t local experimental stations and by more skilled
growers are frequently several times the local average, emphasizing the
potential for greatly increased production at little or no additional cost.
Australian specialist growers regularly obtain 10-1 5 t/ha dry rhizomes and
up to 150 t/ha green from selected strains; the ratio of green to dry ginger is
roughly 5 to 1. Similarly to yield, the oil or oleoresin content of rhizomes and
their characteristic odour and pungency are genetically controlled. Increases
can be achieved only by planting selected cultivars possessing the optimum
combination of these factors, but only in Australia has this been achieved on
a commercial scale.
Differences in composition of freshly harvested, whole rhizomes have
already been noted, and the range in commercial dried rhizomes is well
documented. Fibre, volatile oil content and pungency level are the most
important criteria in determining suitability for a particular product. Young,
tender rhizomes, about 5-7 months after planting are preferred for pre-
served ginger since the fibre content is negligible and the pungency mild. As
rhizomes mature, volatile oil, pungent constituents and fibre increase. The
volatile-oil content of Australian dried rhizomes increases from 1.8 to 4.4%,
with the mid-season crop preferred for oil or oleoresin, the late-season for
dried spice.
The extent to which rhizomes are cleaned prior to drying directly affects
volatile oil and fibre content. Removal of the cork skin reduces fibre content
but also enhances oil loss through rupture of the surface oil cells; thus cleanly
peeled Jamaican ginger generally has a lower oil and fibre content. Peeling
rhizomes also influences pungency, as the constituents concerned are mainly
located in the outer skin layers. The fibre content of unpeeled dry rhizomes
can be 10%, but in commercial dried ginger it is usually 1.5-6.0%, with the
pungent constituents, gingerols, in freshly dried rhizomes 1-2%.
Distillation
For steam distilling, dried rhizomes are comminuted to a coarse powder
immediately before being loaded into the still; in Sri Lanka the optimum
particle size was 1 mm, with a 1:15 ratio of ground ginger to water. The
558 Chapter 14
charge must be packed evenly, or the powder may be spread on several trays
to ensure complete recovery of the oil. Distilling is by live steam and,
depending on weight of charge and steam pressure, may take up to 20 hours.
Cohobation is normally necessary to obtain maximum oil yield. In Sri Lanka
and India, approximately 70-75% of oil was recovered from dried ground
rhizomes, a further 5-6% by cohobation; from green ginger 50-60%. Oil
yield generally is 1.5-3.0 averaging 2.0%, above 3.0 is exceptional but up to
6.0% can be obtained from selected strains at experimental stations. Chi-
nese oil distilled from dried, coated rhizomes yields 1.5-2.5%, but re-
portedly 4-5% from rhizomes produced in Hunan Province.
The distilled material, ginger marc, can be solvent extracted to produce an
inferior oleoresin, sometimes used as a diluent or adulterant of other
oleoresins or ginger oil. Marc is also dried, powdered and used as an inferior
spice and can be added to stockfeed and, although fibrous, is readily eaten.
Rhizome peelings left after preparing edible ginger should be distilled as
soon as possible, since delays of only a few hours can result in oil loss of
60-70% (Pieris, 1982). A yellow dye can be solvent extracted from ground
rhizomes (Popoola et aZ., 1994).
Ginger oleoresin is obtained by solvent extracting powdered, dried rhi-
zomes, with significant differences in yield, aroma, flavour and pungency
due to rhizome origin, age at harvest, solvent and extraction method.
Jamaican, Sierra Leone, Nigerian, Indian and Australian gingers are all
suitable, although manufacturers choose ginger according to the end use.
Jamaican ginger oleoresin is preferred by soft-drink manufacturers, the meat
industry favours heavier-bodied, more pungent African oleoresin. It is
important that rhizomes for extraction be mould-free, since mycotoxins may
be co-extracted with flavour principles (Pieris, 1982). Well-dried rhizomes
are ground or pulverized to a coarse powder and extracted with the selected
solvent. Using the b a x h counter-current method, the powder is packed in
stainless columns and extracted by cold percolation; compared with other
spices, percolation is notably slow. Fine grinding of rhizomes leads to caking
in the extraction column, resulting in an even slower percolation rate,
although resulting in a higher yield of oleoresin. Fractions of extract contain-
ing not less than 10% soluble solids are drawn off and distilled under
reduced pressure, to lower potential heat damage to the gingerols, since
excessive heat during extraction can degrade them to less pungent shogaols
or weakly pungent zingerone and aliphatic aldehydes.
Common commercial solvents are ethanol, acetone, trichloroethane or
dichloroethane, and the oleoresins produced differ in their characteristics,
odour and flavour. The moisture content of rhizomes may dilute water-
miscible solvents, thus the powder is effectively extracted with a solvent-
water mixture, which extracts non-flavour components including water-
soluble gums. This causes problems during solvent removal and oleoresin
concentration, and sludge may separate out during storage. Conversely, a
Zingiberaceae 559
with ginger. Resistance exists to some of these diseases, but whether this is
general resistance or merely to the local biotype has yet to be determined.
Rhizomes can also be infected in store, and once this has occurred it is
extremely difficult to control. Only healthy, dry rhizomes treated with a
fungicide should be stored, and regular inspection is essential during pro-
longed storage, especially for seed rhizomes. Bacterial soft rot caused by
Emoinia carotovora is responsible for serious loss of stored rhizomes in
Australia.
References
Anon. (1979) Ginger shade trial at Pinetown, South Africa. In$ Bull. Cit. Subtrop.
Fruit Res. Inst. (Nelspruit) 84, 5-8.
Babu, K.N. et al. (1993) Biotechnology in spices. Indian Hort. 38(3), 46-50.
Baranowski, J.D. (1986) Changes in solids, oleoresin and 6-ginger01 content of
ginger during growth. HortSci. 21(1), 145-146.
Bartley, J.P. (1995) A new method for the determination of pungent compounds in
ginger. J. Sci. Food Agric. 68, 215- 222.
Breitschneider, E. (1870) On the Study and Value of Chinese Botanical Works.
Foochow, China.
Chen, C.C. et al. (1986) Pungent compounds of ginger extracted by liquid carbon
dioxide. J. Agric. Food Chem. 34(3), 477-480.
Chew, W.Y. (1969) Performance oftapioca, sweet potato and ginger on peat. Planter
(Kuala Lumpur) 45, 445-45 1.
Connell, D.W. (1970) The chemistry of the essential oil and oleoresin of ginger.
Flav. Ind. 1, 677-693.
Das, T.M.P., Devadas, V.S. and Pillai, G.R. (1990) Efficacy of fungicides for
treatment against pre-emergence rhizome rot of ginger. Ind. Cocoa, Arecanut &
Spices?. 14(1), 13-15.
Dhirendra, iM.(1990) Seed protectant property of ginger oil. Ind. Pe$um. 34(4),
266-268.
Ekundayo, O., Laaksoi, I. and Hirtunen, R. (1988) Composition of ginger volatile
oils from Nigeria. Flav. Frag. J . 3, 85- 90.
Furutani, S.C. and Nagao, M.A. (1986) Influence of daminozide, gibberellic acid
and ethephon on flowering, shoot growth and yield of ginger. HortSci. 21 (3-1),
428-429; 20(3-l), 392- 393.
566 Chapter 74
Ra, S.W. et al. (1989) A study of cold injury to ginger. Res. Rept. Rural Dev. Ass.
Hurbo. 31(3), 39-42.
Rosengarten, F. Jr (1969) The Book of Spices. Livingstone Pubs. Co., Wynwood,
USA.
Sankar, C.R. and Muthuswamy, S. (1984) Dry matter production and recovery of
dry ginger in relation to light intensity. Ind. Cocoa, Arecanut & SpicesJ. 10(1),
4-6.
Shah, J.J. and Raju, E.C. (1975) Ontogeny of the shoot apex of ginger. Nomoeg. J.
Bot. 22(3), 227-236.
Sivan, P. (1979) Growth, spacing, time of lifting and production of early harvest
ginger. FijiAgric. J. 41(1), 37--43.
Steinegger, E. and Stucki, K. (1982) Trennung und quantitative Bestimung de
Hauptscharfstoffe von Zingiberis rhizoma. Pharm. Acta Helv. 57(3), 66-71.
Svendsen, A.B. and Scweffer, J.J.C. (1985) Essential Oils and Aromatic Plants.
Martinus NijhoffiDr W. Junk, Dordrecht, The Netherlands.
Taroeno, B.J.J. (1 99 1) Analysis of the essential oil of Z. cassumunar from Indonesia.
Flan. Frag. J . 6(2), 161- 163.
Theilade, I. et al., (1 993) Pollen morphology and structure in Zingiber. Grana. 32(6),
338-342.
Tomlinson, P.B. (1956) Studies of the systematic anatomy of the Zingiberaceae. J.
Linn. Soc. Bot. 60, 547-592.
Weiss, E.A. (197 1) Castor, Sesame and Safiower. Leonard Hill, London, UK.
Whiley, A.W. (1981) Effect of plant density on time to first harvest maturity etc. of
ginger grown in S.E. Queensland. Trop. Agric. (Trin.) 58(3), 245-251.
Willers, P. (1985) Efficient nematode control a step nearer for ginger producers. In5
Bull. Cit. Subtrop. Fruit Res. Inst. 160, 4-6.
Wu, J.J. and Yang, J.S. (1994) Effects of irridiation on the volatile compounds of
ginger rhizomes (Z. oficinale). J. Agric. Food Chem. 2(1 l), 2574-2577.
Xu, K., Zhao, D.W. and Jiang, X.M. (1993) Studies on the nitrogen absorption rate
in ginger using isotope-N, Acta Hort. S i n k 20(2), 161-165.
Yusof, N. (1990) Sprout inhibition by gamma irradiation in fresh ginger. J. Food
Process. Preserv. 14(2), 113- 122.
Distilling and
Extracting
Essential Oils 15 -
568
Distilling and Exfracting Essential Oils 569
Raw material 7
Extraction by
animal fat
Resinoid
G
1
Alcohol extraction
Absolute I
I
I 1
I
I Alcohol extractionI
Fig. 15.1. Basic flowsheet of the operations necessary to obtain an essential oil or derivative.
Distilling and Extracting Essential Oils 571
5 72 Chapter 15
E'
E
m
m
._
U
s
.-
I
-
-m
.-
I
.-U
v)
E
%
tj
d,
ii
Distilling and Extracting Essential Oils 5 73
'izzz
modifier, but solvent extracted oils are generally considered to reflect a
Condensors
0
-. Water circulation
Raw
material
Heating
jacket
=+- - n-7-r-
Vacuum pump
valve
L Dosing pump
+
2 Vacuum pump
plant’s natural odour more accurately than distilled oils (Fig 15.6). Com-
monly used are petroleum ether, hexane, toluene or other binary solvents.
Products from solvent extraction may be further processed, i.e. concretes
into absolutes (Fig. 15.7). Many countries have introduced stringent regula-
tions controlling the use and type of solvents, and recommendations govern-
ing solvents used in the food industry are published by the International
Organization of the Flavour Industry (IOFI).
Gaseous extraction uses liquid carbon dioxide which, under pressure and
regulated temperature, is passed through the material, then via a separator to
recover oil and gas. This method is considered superior to liquid solvents,
since it preserves important unstable or heat-sensitive components and
requires less energy. Carbon dioxide is safe, non-combustible, odourless,
tasteless, inexpensive and readily available, ideal properties for an extraction
solvent, while its low viscosity enables it to penetrate the material being
extracted and its latent heat of evaporation allows it to be easily removed
without residue. An important feature of this process is that by varying the
temperature and pressure of the gas, an extract with specific flavour or odour
specifications can be obtained. A less costly variation is the subcritical liquid
gas technique, whose products, however, may require further processing.
Since the introduction of gas extracted oils as ingredients for flavours and
fragrances in 1982, the range is increasing steadily. Some advantages of
carbon dioxide gas extracts are: (i) no organic solvent is employed; (ii) at low
evaporator
Fig. 15.7. Concrete to absolute unit.
Water -
Collector
(from water bath)
CO2 (1)
(from storage cylinder)
Continuous dense phase extraction apparatus
Water bath
References
Ames, G.R. and Mathews, W.S.A. (1968) The distillation of essential oils. Trop. Sci.
10, 136-148.
Bedoukian, P. (1993) Eau de Cologne. Perfum. Flav. 18(l), 37-4 1.
Boelens, M.H. et al. (1990) Ten years of hydrodiffusion oils. In: Proc. 11th. Znt. Cong
Ess. Oils.Frag. Flav. Vol. 4. pp. 121-126, Aspect Pubs, London, UK.
Chollot, R. (1985) History of Grasse. Perfum. Flav. 91(6), 72-77.
Forbes, R. (1 948) A Short Histo y of the Art of Distillation. Brill & Co., Leiden, The
Netherlands.
Guenther, E. (1949-1952) The Essential Oils, 5 vols. D . van Nostrand & Co., New
York, USA.
Jansz, E.R. et al. (1981) Distillation of leaf essential oils. CZSZR Res. Bull. No. 1.
Colombo, Sri Lanka.
Morris, E.T. (1981) Roots: The earliest history of the essential oil industry. Perfum.
F ~ u v 6(2),
. 54-58.
iMuller, P.A. (1967) The noses of Grasse. Perj! Ess. Oil Recd. 58(8), 531-534.
Swaine, R.L. and Swaine, R.L. Jr (1988) Citrus oils: processing, technology and
applications. Perfum. Flav. 13(6), 1-20.
Weiss, E.A. (1994) Oils from Outback. C T M S 6(1), 31-32.
Wilkinson, G . T . (1993) SFE extraction in essential oil recovery. In: Advances in
Supercritical Fluid Extraction (unpaginated). Food Research Institute, Werribee,
Victoria, Australia.
Glossary
579
580 Glossary
Whole oil. Specifically an oil free of all impurities, water and additives and
true to the original material; is often misused.
WONF. With Other Natural Flavours. An extract or flavour too weak to be
used alone, and reinforced with other natural flavouring material con-
taining similar or complementary components. This identification is a
legal requirement in certain countries including the USA.
Index
585
586 Index