SPECIAL FEATURE

Phytohormonal Ecology1

Essentially all of the energy that flows through food webs is generated by the action of pho-
tosynthesis in green plants. Save decomposition, the conduit from this primary production to the
rest of the food web is via consumption of plants. Where the abiotic environment imposes stress
or favorable conditions, the quality of plants changes, adaptive responses may ensue, and eco-
logical effects ripple through the web of interactions. Given that hormones are universally em-
ployed by plants to coordinate responses to the biotic and abiotic environment, we introduce the
term phytohormonal ecology to recognize the central role that hormones may have in the
ecology and evolution of species. These often small and simple molecules are now known to
trigger plant immune responses to pathogens and defenses against herbivores, and to coordinate
plant adaptations to environmental stress. Although the lack of mobility and a central nervous
system certainly predisposes plants to rely disproportionately on hormones, hormones are also
gaining attention in animal ecology.
Phytohormones are molecular signals that trigger and regulate germination, growth, devel-
opment, defense, responses to stress, and reproduction in plants. Plant hormones have been studied
for decades, and preexisting knowledge of hormonal physiology provided a strong foundation
from which recent ecological studies have emerged. In the past 20 years, an explosion has occurred
in the number of studies demonstrating the role of phytohormones as fundamental communication
signals involving plants, their abiotic environment, and the surrounding ecological community.
New and accessible analytical tools that revealed heretofore hidden biochemical and genetic
processes have facilitated much of this activity. Discoveries have been made exposing phyto-
hormones as mediators of novel ecological and physiological functions, complex molecular syn-
ergies and antagonisms, and strong effects on other species. Indeed, phytohormonal processes
even mediate interactions between organisms in distinct kingdoms.
The more deeply we investigate phytohormones, the broader their definition becomes. As
knowledge about the numerous roles of phytohormones growsseemingly exponentiallyso
does our awareness of their omnipresence in virtually every aspect of plant ecology. Our goal
in this Special Feature was to highlight some of the hot spots of research activity into phyto-
hormonal ecology. These papers use diverse analytical, experimental, and genetic methods to
investigate hormonal processes in model (Arabidopsis, Lycopersicon) and non-model (Iris, Rhi-
zophora, Rumex, Quercus) species, and they encompass an ecological spectrum ranging from
within-species ecophysiology to diverse community patterns.
The opening paper by Elizabeth Farnsworth introduces the field of phytohormonal ecology,
and integrates traditional hormonal physiology of plant growth and development with eco-
logical concepts of life-history traits, phenotypic plasticity, and evolution. Rens Voesenek and
colleagues next illustrate how hormones link individual growth physiology, responses to envi-
ronmental cues, and large-scale patterns of plant species distributions. In a new analysis, they
argue that the ethylene-induced stem elongation capacity of plant species determines their dis-
tribution in a floodplain. In the third paper, Don Cipollini demonstrates the potential for hormones
to interact, as plants are faced jointly with competitors and herbivores. Although studying these
diverse stresses has been a classic area of plant ecology, Cipollini sheds new light on the topic
and produces testable mechanistic hypotheses for how and why competition and herbivory should
interact. The interaction between abiotic environmental stress and attack of plants is emphasized
in the following two papers. Susan Mopper and colleagues explore plant hormonal responses to
salinity stress and its potential consequences for herbivores in field experiments. Jennifer Thaler

1
Feature accepted 11 May 2003. Reprints of this 75-page Special Feature are available for $11.25 each, either
as pdf files or as hard copy. Prepayment is required. Order reprints from the Ecological Society of America,
Attention: Reprint Department, 1707 H Street, N.W., Suite 400, Washington, DC 20006 USA.

3
 and Richard Bostock complement this approach, examining the complexity and functional di-
versity of phytohormones as plants react simultaneously to abiotic stress, herbivores, and path-
ogens. The interactions among different trophic levels is further elucidated by Paul Hatcher and
his colleagues, who outline how hormonal signaling within plants may mediate interactions
between herbivores and fungal pathogens and describe some of the challenges in merging mech-
anistic and ecological approaches. The final contribution by Jack Schultz and Heidi Appel presents
the broadest perspective and proposes that hormonal signals represent shared interactive languages
among plants, microbes, and insects, the evolution of which can only be understood by examining
their similarities as well as their differences.
The success of phytohormonal ecology will depend on the collaboration among, and joint
training of, physiologists, field biologists, and molecular ecologists and the interplay between
model and novel study systems. Ultimately, the biodiversity and interactions that occur in a
community can probably be linked to some phytohormonal process. The signal-to-noise ratio,
representing the relative importance of phytohormonal ecology compared to other factors, will
determine its staying power.

SUSAN MOPPER,
Guest Editor
University of LouisianaLafayette
ANURAG A. AGRAWAL,
Special Features Editor
University of Toronto

Key words: abiotic stress; allocation; herbivory; interspecific interactions; molecular signals; patho-
Special Feature

gens; phenotypic plasticity; plant competition; plant development; salinity; signalling pathways; systemic
induced defense.
q 2004 by the Ecological Society of America
Ecology, 85(1), 2004, pp. 515
q 2004 by the Ecological Society of America

HORMONES AND SHIFTING ECOLOGY THROUGHOUT

PLANT DEVELOPMENT
ELIZABETH FARNSWORTH1
New England Wild Flower Society, 180 Hemenway Road, Framingham, Massachusetts 01701 USA

Abstract. Hormones modulate complex suites of ecologically relevant behaviors

through interactive cascades of signal transduction; evolutionary changes in the function
of a single hormone can result in multiple changes in plant traits. Sites of hormone action
and tissue sensitivity change throughout plant ontogeny, as embryos, seedlings, and re-
productively mature plants cope with shifting suites of environmental variables and resource
availability. Phenotypic plasticity and correlations and trade-offs between life history traits
(such as resource use efficiency and allocation to growth and differentiation of meristems)
also change as plants age; hormonal changes are central to these shifts. I synthesize evidence
from the molecular and physiology literature and present novel data on mangrove propa-
gules, seedlings, saplings, and trees. Together, these data demonstrate that several ecolog-
ically important traits are hormonally mediated (both in the short term and over evolutionary
time) in the diversification of plant lineages. I focus on two hormones with contrasting
action: (1) abscisic acid (ABA), which regulates internal plant osmotic stability, membrane
integrity, seed dormancy, and stomatal conductance; and (2) cytokinins (zeatin and allied
adenine derivatives), which promote cell division, stimulate growth, delay leaf senescence,
enhance the capacity of tissues to act as N sinks, and help to transduce signals of light and
nutrient availability. ABA and cytokinin levels are correlated with patterns of osmotic
tolerance, photosynthesis, growth, and leaf longevity, and their concentrations in specific

Special Feature
tissues change throughout plant development. Ecologists and evolutionary biologists are
in a strong position now to creatively comprehend, predict, and potentially engineer plant
strategies and adaptive trade-offs over plant lifetimes in light of emerging knowledge of
hormones.
Key words: abscisic acid; allocation; cytokinins; evolution; growth; mangroves; maturation;
phenotypic plasticity; phytohormones; plants; seedlings; zeatin.

INTRODUCTION emerging, integrative field of ecodevo (sensu Gil-

bert 2001), is beginning to flourish as our understand-
Hormones run most of the important aspects of life,
ing of hormonal evolution and controls on development
including the major processes that demand the attention
improves.
of many ecologists: resource capture, growth, and sex.
Meanwhile, plant physiologists have been busily un-
Physiological ecologists have long recognized the im-
tangling the means by which phytohormones function
portance of hormones in integrating environmental sig-
during plant morphogenesis and throughout life (Da-
nals, controlling flexible physiological responses, reg-
vies 1995). While it is true that plant and animal hor-
ulating phenotypic plasticity, conferring maternal ef-
mones function differently in terms of their production
fects, and constraining life history strategies. A decade
sites and action (well summarized by Weyers and Pat-
ago, Feder and Block (1991) exhorted animal physi-
erson [2001], Kushiro [2003]), hormones have much
ologists to join forces with molecular geneticists to
potential to explain major evolutionary innovations and
study the evolution of hormonal physiology in an adap-
novel traits in plants. Thus, the ecological and evolu-
tive ecological context. They have risen to the charge.
tionary significance of hormonal physiology in plants
With the advent of new techniques for manipulating
is becoming clearer as we come to understand signaling
endogenous hormone levels in animals, increasingly
pathways, mechanisms of tissue sensitivity, genetic ba-
elegant experiments have demonstrated the ways in
ses for hormonal action, and synergistic and antago-
which hormones coordinate complex suites of behav-
nistic interactions among hormones. Recent compara-
iors that are sensitive to strong natural selection (see
tive studies, for example, have shown that similar evo-
recent reviews by Sinervo and Svensson [1998], Ket-
lutionary changes in the function of a single hormone
terson and Nolan [1999], and Dufty et al. [2002]). The
within multiple plant lineages can be manifested in
convergent changes in multiple plant traits that may in
Manuscript received 21 October 2002; revised 23 February part explain trade-offs and correlations among char-
2003; accepted 22 March 2003. Corresponding Editor: A. A.
Agrawal. For reprints of this Special Feature, see footnote 1, p. acters (e.g., Farnsworth 2000). Plants constitute ex-
3. cellent systems for studying proximate hormonal ef-
1 E-mail: efarnswo@mtholyoke.edu
fects on physiology and fitness and ultimate impli-
5
 6 ELIZABETH FARNSWORTH
cations for evolution of correlated traits. A wealth of
well-characterized mutant lines with differing hormon-
al sensitivities and levels of production exist across a
variety of species, and more are emerging all the time
(McCourt 1999). The existence of rapid-cycling cul-
tivars and annual species also permit studies of char-
acter change over several generations (e.g., Mitchell-
Olds 2001). Ecologists are also moving beyond Ara-
bidopsis, transcending the traditional crop and herba-
ceous study systems to investigate the hormonal
physiology of plants growing in a variety of complex
environmental situations (e.g., Voesenek and Blom
1996, Farnsworth and Farrant 1998, Mercier and En-
dres 1999, Wang et al. 2001, Zotz et al. 2001).
Plant form, physiology, and sites of hormone action
and tissue sensitivity also change throughout plant on-
togeny, as plants cope with shifting suites of environ-
mental variables and resource availability (Trewavas
1986, Evans and Poethig 1995, Amzallag 2001). Cor-
relations and trade-offs between life history traits (such
as resource use efficiency and allocation to growth and
differentiation of meristems) also change as plants age.
Hormonal changes are central to these shifts, but mech-
anisms are only now becoming clearer. These mecha-
Special Feature
nisms are likely to involve an interplay of complex
signal transduction pathways, secondary messengers,
differential synthesis and sensitivity of receptors, rel-
ative control by hormones and extrinsic factors, and
variable phenotypic expression of hormone effects
(Paterson et al. 2001, Weyers and Paterson 2001).
While both ecologists and physiologists recognize that
growing plants are dynamic systems, ecologically rel-
evant traits and hormone action have rarely been stud-
ied in an explicitly dynamic developmental context,
spanning the course of development from embryo to
tree.
Here, then, is an opportunity to understand plant
ecology more fully by integrating among approaches
from evolutionary, molecular, and developmental bi-
ology. Predicting variability in plant responses over
time, measuring fitness costs of phenotypic plasticity,
evaluating which life history phases are most vulner-
able to stress, and scaling estimates of plant produc-
tivity from leaf to forest all depend on a stronger un-
derstanding of physiological controls on development.
Here, I explore some linkages among suites of traits
that change over ontogeny and that are known to be
controlled at least in part by changing hormonal action.
I will focus on two illustrative phytohormones with
contrasting action: abscisic acid (ABA) and cytokinins,
using examples from the literature and new data from
studies of mangroves. Mangroves (a term applied to
over 50 species in 15 families of woody halophytes
restricted to tropical saline coasts, sensu Tomlinson
[1994]) are a promising group in which to study con-
vergent evolution of hormonal mechanisms involved
in the maintenance of specialized traits (such as salt
tolerance, viviparous reproduction, photosynthetic ac-
Ecology, Vol. 85, No. 1
climation, etc.) that enable existence in challenging
environments.
Major life history transitions in plants
Plants undergo recognizable ontogenetic phase
shifts, which take the general trajectory of (1) rapid
growth and tissue differentiation immediately follow-
ing germination; (2) exponential growth to an asymp-
totic maximum size; (3) stabilization of woody support
structures; (4) maturation and attainment of reproduc-
tive competence; and (5) senescence (Jones 1999).
Studies have elucidated the physiological changes that
can occur within a single growing leaf (e.g., Wait et
al. 1998) or branch (Emery et al. 1998). However, the
notion of whole-plant developmental stages and their
corresponding physiological states have not often been
integrated conceptually into long-term ecological stud-
ies of plants. With some exceptions (e.g., Honkanen
and Haukioja 1998, Kearsley and Whitham 1998, Kar-
ban and Thaler 1999), most investigations of linkages
among plant traits and plastic responses to the envi-
ronment focus on plants at a single developmental
phase, or measure growth variables (e.g., relative
growth rate, net assimilation rate) that integrate across
growth phases during a single season. Laboratory,
greenhouse, and shadehouse studies of long-lived
plants focus on more tractable seedlings or potted
shoots. Herbaceous species tracked in the laboratory
and field tend to be measured at reproductive maturity.
Likewise, many comparative field studies of correlated
traits focus only on mature plants, saplings, or small
understory trees. Observations across plant lifetimes
would afford insights into the origins of correlations
among traits and selective pressures maintaining these
correlations. McConnaughay and Coleman (1999)
pointed out that observed changes in traits over time
or between experimental treatments can arise from
ontogenetic drift (sensu Evans 1972) as much as
from extrinsic factors. That is, when allometric rela-
tionships among traits vary non-isometrically as plants
grow, such patterns will change with plant age regard-
less of environmental conditions. Thus, they caution
that ecologists consider ontogenetic drift when making
static comparisons among populations.
Studies tracking herbaceous plants from seedlings to
senescence have demonstrated that rates of photosyn-
thesis, patterns of biomass allocation, relative growth
rates, construction costs, leaf longevity, root : shoot ra-
tios, plant architecture, levels of plasticity, and sensi-
tivity to stress covary with plant age, and have reported
varying levels of ontogenetic drift (Poorter and Poth-
mann 1992, Coleman et al. 1994, Romero and Maranon
1996, Lyons and Barnes 1998, Kikuzawa and Ackerly
1999, Harmens et al. 2000, Davies 2001). The few
studies that have attempted to systematically compare
juvenile, adult vegetative, and reproductively compe-
tent plants have demonstrated that ranges of juvenile
traits can be decoupled from those of more mature
January 2004 PHYTOHORMONAL ECOLOGY
plants. For example, Clark and Clarks (2001) long-
term measurements of saplings of rainforest trees re-
vealed few correlations between adult distribution and
juvenile growth rates. Similarly, cotyledons and mature
leaves, as well as juvenile and older shoots, differ in
their resistance to herbivory (Kearsley and Whitham
1998, Karban and Thaler 1999).
Other studies have found higher levels of phenotypic
flexibility, plasticity (sensu stricto), niche breadth
across resource gradients, or responsiveness to envi-
ronmental variables among younger woody plants than
among older or reproductively mature plants and be-
tween lineages that differ in ontogenetic trajectory
(Farnsworth and Ellison 1996, Pigliucci 1997, Eriksson
2002). This finding accords with an emerging hypoth-
esis that early phenotypic plasticity may constrain plas-
ticity later in life (Weinig and Delph 2001). To what
degree are these adult differences due to changing ex-
trinsic conditions vs. internal hormonal mechanisms?
Roles of phytohormones at different life stages
Extrinsic environmental signals trigger phase change
in meristem cells of woody and herbaceous species,
while intrinsic physiological cues regulate responsive-
ness to these signals as well as the length of time a
meristem remains in a particular phase (Poethig 1990).
Hormones may help plants control meristem identity
as they track cumulative information about the distance
of the apex from the roots and numbers of vascular
branches (Kearsley and Whitham 1998, Jones 1999,
Sachs 1999). Specific genes are implicated in the trans-
formation of meristem identity, i.e., from vegetative to
flowering (e.g., Weigel and Milsson 1995, Ratcliffe et
al. 1999), but hormonal actions linked to these genes
need to be elucidated fully. Several phytohormones
play critical roles in maturation, biomass partitioning,
growth, and stress responses of plants (Davies 1995).
Among the best studied are abscisic acid (ABA) and
cytokinins (zeatin [Z], zeatin riboside [ZR], isopenten-
yl adenine [IPA], and related molecules). These two
phytohormones act independently and often oppositely
on plant cells, exerting major influences plant devel-
opment and growth.
ABA, seeds to senescence.ABA is a low-molecu-
lar-weight acid that modulates plant development,
seed dormancy, germination, cell division, and cellular
responses to environmental stresses such as drought,
cold, salt, pathogen attack, and UV radiation (Rock
2000:358). ABA also can inhibit cell elongation during
responses to other stresses such as flooding (Hoffman-
Benning and Kende 1992). Studies of both ABA-de-
ficient and ABA-insensitive mutants indicate that ABA
is a general inhibitor that slows metabolic processes sues. Although propagules of Rhizophora mangle and
and prepares plant cells for conservative stress avoid-
ance (Chapin 1991, Leung and Giraudat 1998, McCourt
1999).
In the seed, ABA contributes to the upregulation of
dehydrin proteins that act as membrane chaperones dur-
7
ing the embryo desiccation that precedes dormancy
(Ingram and Bartels 1996). ABA also suppresses ger-
mination-specific genes and counteracts the action of
gibberellins that induce germination. ABA-deficient
and -insensitive mutants and seeds of naturally occur-
ring plants that lack ABA exhibit high sensitivity to
freezing and drying (recalcitrance), low dehydrin pro-
tein production, lack of dormancy, and, frequently, pre-
cocious germination (Box 1; Farrant et al. 1996, Farns-
worth and Farrant 1998, Rock 2000). In the growing
plant, ABA promotes stomatal closure by inducing ion
efflux from guard cells to slow transpiration in response
to drought, salinity, prolonged flooding, and other os-
motic stress. It remains unclear whether ABA is itself
a signal conveyed from the roots (which sense changes
in soil osmoticum), or whether the hormone acts in-
directly by stimulating gene transcription for proteins
involved in a larger signal transduction pathway (Wil-
kinson and Davies 2002). However, it is clear that ABA
is involved in short-term, plastic responses to changing
water potential and nutrient availability, as ABA levels
in drought-stressed tissues rise considerably within
hours. Even tissues that do not normally produce ABA
can be rapidly induced to produce the hormone or to
Special Feature
release its activated (unconjugated) form from storage
compartments (Box 1). Short-term boosts in ABA pro-
duction in response to osmotic stress have been doc-
umented in mangroves (Box 1), Suaeda maritima (a
halophyte) (Clipson et al. 1988), and Iris hexagona (a
glycophyte) (Wang et al. 2001).
Interspecific differences in ABA levels suggest that
plant populations subjected to strong selection under
osmotic stress should exhibit microevolutionary dif-
ferentiation in hormone production. Findings from
mangroves growing in contrasting salinity environ-
ments show little such differentiation (Box 2). These
findings point to a homeostatic mechanism in this spe-
cies in which growing seeds (and viviparous propa-
gules in mangroves) are shielded from varying external
conditions. In contrast, Wang et al. (2001) found con-
siderably higher levels of ABA in tissues of young Iris
hexagona Walter seeds exposed to high salinity relative
to a low-salinity population. Unlike the mangroves
studied, however, this species produces dormant seeds.
Differences in sensitivity to salinity declined with age
of the seed, much as with Rhizophora mangle. Impa-
tiens capensis, a non-halophyte, does show apparent
population differentiation for abscisic acid responsive-
ness, as well as a trade-off between survivorship in
saline conditions and reproductive output (Heschel and
Hausmann 2001).
ABA production is compartmentalized among tis-
other viviparous mangroves exhibit low ABA levels
throughout development, maternal pericarp and foliar
tissues show significantly higher ABA concentrations,
in line with those found for other salt-tolerant plants
(Fig. 1). Wang et al. (2001) demonstrate similar com-
 8 ELIZABETH FARNSWORTH Ecology, Vol. 85, No. 1

BOX 1. Dynamics of ABA Levels in Developing Embryos of Viviparous and Nonviviparous Species

This study compared levels of the phytohormone abscisic acid in related groups of species that differ
in terms of viviparity, and determined whether ABA production or release could be induced in viviparous
mangrove propagules subjected to osmotic manipulation in the field. Nondormant, viviparous embryos
(propagules sensu Tomlinson 1994) of 10 species across four disparate families of mangroves maintain
unusually low levels of ABA throughout their development on the parent plant relative to their nonvi-
viparous, upland relatives (Fig. 1). However, short-term increases in ABA levels can be induced by applied
osmotic stress. Detached mangrove embryos of three different growth stages subjected to desiccation
produce elevated levels of ABA relative to controlslevels comparable to related nonviviparous non-
mangroves (Farnsworth and Farrant 1998). Increases in hormone levels can also be induced in situ on
propagules that are still attached to mature trees. Four sets of 24-mo-old (7-cm long) embryos of the
red mangrove, Rhizophora mangle L., on mature trees growing in ;20 salinity at the Fairchild Tropical
Garden, Miami, Florida, USA, were selected. Ten propagules per set were randomly subjected to one of
four treatments: (1) injection with 1 mL distilled water; (2) injection with 1 mL of 2% NaCl solution
approximating measured embryo tissue concentrations of salt, serving as a solution control; (3) piercing
needles without injection, serving as an injection control; and (4) comparably aged, unmanipulated controls.
Propagules were collected 48 h following treatment and analyzed for ABA content by the methods of
Farnsworth and Farrant (1998) and Weiler (1980). Mangrove propagules with osmoticum altered by fresh-
water injection produced ABA in concentrations three times higher than controls (Fig. 2), levels that are
comparable to those found in embryos of nonviviparous non-mangroves and in maternal vegetative tissues
(leaves and roots) of Rhizophora mangle (Fig. 1). These findings indicate that ABA is either produced in
or translocated to embryonic tissues experiencing osmotic alteration, or that ABA is converted from a
Special Feature

conjugated (inactive) to a free (metabolically active) form in response to this stress.

partmentalization of hormone production and action
among maternal and embryonic tissues of Iris hexa-
gona. In mangroves, the time spent growing on the
maternal tree is a desalinating process for the prop-
agule (sensu Zheng et al. 1999), in which levels of
solutes, salt, and hormones are closely regulated. Fol-
lowing abscission, however, the mangrove propagule
is released directly into seawater and must quickly gain
competence to cope with changeable osmotic condi-
tions. At a critical point during dispersal, the rooted
seedling must upregulate ABA production in vegetative
tissues, as other hormonal changes are occurring (Smith
et al. 1996). As Amzallag (1999) demonstrated for Sor-
ghum bicolor, acclimation to salinity can only occur

BOX 2. Comparison of Baseline ABA Levels in Mangrove Propagules Growing in Contrasting

Salinity Environments

While ABA levels can change rapidly in response to osmotic stress, ABA normally appears to remain at
low concentrations in mangrove propagules growing in a range of salinities. A comparative study illustrated
little differentiation among three populations of Rhizophora mangle L. growing in contrasting salinity con-
ditions. I harvested six propagules from trees of each of (1) a hypersaline lagoon, West Pond, on Twin Cays,
Belize (salinity of soil interstitial water measured with refractometer 51); (2) a tidally flushed, fringing
mangrove stand on Wee Wee Cay (salinity 30); and (3) a riverine, mainland, mangrove stand along the
Sittee River (salinity 510). ABA concentrations were measured in tissues of three ages: (1) ,30 d after
pollination (dap); (2) 24 mo dap; and (3) .6 mo dap. ABA levels were highest among propagules from
the hypersaline populations, especially in the youngest cohort of propagules (Fig. 3). However, ABA levels
in older propagules converged among all sites, remaining low throughout ontogeny relative to concentrations
in maternal tissues. The apparent lack of physiological differentiation among populations of R. mangle
growing in different salinities may, in part, explain Smith and Snedakers (1995) finding that R. mangle
seedlings from hypersaline and intermediate salinity sites did not differ in early growth ratesthat is, that
hypersaline propagules did not show higher acclimation or tolerance for salinity.
January 2004 PHYTOHORMONAL ECOLOGY 9

Special Feature
FIG. 1. Contents of abscisic acid (ABA; upper panel) and cytokinins (zeatin [Z] and zeatin riboside [ZR]; lower panel)
in embryos, maternal pericarp, and mature foliar tissues of 10 viviparous mangrove species (gray bars) and four nonviviparous,
upland species (white bars). Means across species (N 5 9 [three samples of three propagules per age or leaves per species])
61 SE are shown. Cytokinins and ABA were quantified using methods of Farrant et al. (1993). Viviparous species were
Bruguiera exaristata Ding Hou, B. gynmorrhiza (L.) Lamk., Ceriops decandra (Griff.) Ding Hou, C. tagal (Perr.) C. B.
Robinson, Kandelia candel (L.) Druce, Rhizophora mangle L., R. stylosa Griff., Aegiceras corniculatum (L.) Blanco, Nypa
fruticans (Thunb.) Wurmb., and Aegialitis annulata R. Brown. Nonviviparous species were Cassipourea elliptica (Sw.) Poir.,
Ardisia escallonioides Schiede & Deppe, Limonium perigrinum (Berg.) R. A. Dyer, and Phoenix reclinata (Jacq.). Members
of the mangrove Rhizophoraceae (Bruguiera, Rhizophora, Ceriops, and Kandelia) showed high interspecific variation in
cytokinin levels; thus, incorporating a correction for phylogenetic relatedness by pooling data across the Rhizophoraceae did
not change the results. An asterisk indicates a significant difference (P , 0.05 by ANOVA within tissue type and age,
comparing family means) between viviparous and nonviviparous species.

during developmental competence windows that re-
flect ontogenetic adjustments in hormone production
and sensitivity. Disparities between levels of ABA
among different tissues indicate tight control and early
maternal effects on localized deployment of this phy-
tohormone in these plants.
It is of interest that such compartmentalization has
apparently occurred multiple times during the indepen-
dent evolution of mangrove lineages, and that low ABA
levels and consequent precocious germination charac-
terize many species of flooded and saline sites (Farns-
worth 2000). The evolutionary reasons for this conver-
gent switch are not entirely clear, although Elmqvist and
Cox (1996) posit plausibly that once such vivipary aris-
es, the phenotype may enjoy a selective advantage in
heterogeneous environments where suitable sites for ger-
mination and establishment may be few and far between.
More studies of hormonal change in a phylogenetic con-
text become feasible as our understanding of evolution-
ary relationships improves. Clearly, selection pressures
act differently on young and mature phases. Do low
levels of ABA in mangrove seeds, for example, permit
the rapid growth of the embryo even under highly saline
conditions experienced by the maternal tree? Do com-
paratively high levels of ABA in adult foliar and root
tissue enable the mature tree to maintain the conser-
vative photosynthetic and growth rates typical of man-
groves (Ball 1988, Farnsworth and Ellison 1996)?
 10 ELIZABETH FARNSWORTH
Special Feature
FIG. 2. Contents of free ABA (upper panel) and cytoki-
nins (Z and ZR; lower panel) in propagules of Rhizophora
mangle subjected to osmotic manipulation while still on the
parent tree. Bars show mean values (61 SE) for three samples
of 10 propagules per treatment. ANOVA with post-hoc Schef-
fe test was performed (N 5 10 propagules per treatment, df
5 3). ABA levels were significantly different (P , 0.001)
only in the freshwater injection, which altered salt contents
of embryonic tissue relative to the other three treatments.
Cytokinin levels were significantly lower (P , 0.01) in the
saltwater- and freshwater-injection treatments relative to the
two control treatments.
Field studies indicate that ABA may figure in slow-
ing leaf growth (by reducing cell wall extensibility)
and photosynthesis, especially in species with long leaf
lifespans in osmotically challenging environments. Ev-
ergreen and semideciduous tree species show higher
xylem ABA levels during the dry season of the Aus-
tralian savannah than do deciduous species (Thomas
and Eamus 2002). Mangroves, with high foliar ABA
levels, exhibit long leaf lifespans (estimated range
Ecology, Vol. 85, No. 1
across 13 species is 942 mo) relative to upland ev-
ergreen tropical trees (summarized by Ellison and
Farnsworth 1996). Similarly, net photosynthesis re-
mains conservative (essentially flat) across a range of
leaf longevities in mangroves, causing the allometric
relationship between Amax (light-saturated rate of net
CO2 assimilation) and leaf lifespan to depart signifi-
cantly from that established by Reich et al. (1999) for
105 upland plant species (Ellison 2002). Maximal pho-
tosynthetic rates for Rhizophora mangle are highest and
most variable among light environments in seedlings;
responsiveness of Amax to light is lower in saplings and
declines still further in mature trees (Farnsworth and
Ellison 1996). ABA may be implicated in this onto-
genetic shift, as saplings and mature trees increasingly
utilize this hormone.
In the short term, however, leaf senescence and trans-
location of scarce internal resources by some plant spe-
cies can be accelerated by ABA spikes in response to
pulses of stress, especially during reproduction (Yang
et al. 2002). This mechanism may partially explain why
mangroves defoliate when exposed to freezing tem-
peratures, especially if ABA upregulation is not ac-
companied by increased chaperonin protein production
(Farrant et al. 1996). The interaction between multiple
stressors on hormone regulation merits more explora-
tion.
Cytokinins, seeds to senescence.Cytokinins, com-
mon adenine derivatives, promote cell growth and di-
vision (Thimann 1987). The four naturally occurring,
physiologically active, cytokinin compounds are zeatin
(Z), zeatin riboside (ZR), dihydrozeatin, and isopen-
tenyl adenine (IPA). The considerable challenges of
developing nonlethal mutants (Miklashevichs and Wal-
den 1997) have hindered progress in understanding cy-
tokinin signal transduction, but recent work has begun
to overcome these limitations (Vogel et al. 1998,
Hutchison and Kieber 2002). Genetic studies indicate
that these pathways are complex, redundant, and in-
teractive with other hormonal pathways (Schmulling et
al. 1997, DAgostino and Kieber 1999). Cytokinins ac-
cumulate in actively dividing meristems, and are most
abundant in plant compartments that are rapidly grow-
ing, such as elongating stems of germinating seedlings,
proliferating roots, and nondormant embryos. They ac-
crue early in the growing season in vegetative tissues,
attain high levels in young plants, and decline in ma-
ture, fruiting individuals (Mercier and Endres 1999).
The types of cytokinins producedIPA-type vs. Z-
typecan be indicators of particular maturation phases
(Valdes et al. 2002). Sensitivity to cytokinins also
changes as plants age and cope with new stressors such
as salinity (Amzallag 2001, Yang et al. 2002). Cyto-
kinins regulate vascular tissue differentiation (Ye 2002)
and, acting in opposition to auxin, overcome apical
dominance to promote the formation of lateral buds
(Emery et al. 1998, Cowan et al. 1999).
January 2004 PHYTOHORMONAL ECOLOGY 11

Cytokinins increase cell wall plasticity and protect

cell membranes by inhibiting formation of and hasten-
ing breakdown of free radicals (Thompson et al. 1987).
Cytokinins also delay leaf senescence (e.g., Jordi et al.
2000), both by reducing free radicals and by height-
ening the capacity of plant tissues to act as sinks for
phloem transport of sugars and nutrients (principally
N). Cytokinins promote photosynthesis by upregulat-
ing production of two important photosynthetic pro-
teins: the small subunit of rubisco (rbcS) and the major
chlorophyll a/b-binding polypeptide of the light-har-
vesting complex in chloroplasts. They also appear to
stimulate plant growth under high light by way of ad-
ditive and independent pathways from phytochrome
(Fankhauser 2002). In promoting growth and photo-
synthesis and creating nitrogen sinks in sun leaves rel-
ative to shaded leaves, cytokinins also function in light
sensing and allocation under a canopy (Pons et al.
2001).
Cytokinins also figure importantly in the formation
of plant mutualisms with symbiotic bacteria and my-
corrhizae. Cytokinin production by symbionts stimu-
lates promoter activity of nodulation genes and facil-
itates nutrient transfer between the plant and its partner

Special Feature
(Ng et al. 1982).
In seeds, cytokinin levels usually peak during early
histodifferentiation, when the embryo is undergoing
rapid cell division. In nondormant seeds produced by
viviparous mangroves and recalcitrant species, for ex-
ample, cytokinins peak in early and intermediate de-
velopmental stages (Fig. 1, Fig. 3, and Farrant et al.
1993); however, mangrove species show only margin-
ally to significantly higher levels of cytokinins (Z and
ZR) in young and intermediate-aged embryos and ma-
ternal (pericarp) tissues relative to related non-man-
groves (Fig. 1). Variance among viviparous taxa sug-
gests that these differences may be more reflective of
responses to osmotic stress than to the evolution of
FIG. 3. Contents of free ABA (upper panel) and cytoki-
viviparity per se. Rather, cytokinins may be linked with nins (lower panel) in propagules of Rhizophora mangle har-
the import of sugars (which act as compatible solutes) vested at three developmental stages. Shown are means (61
into embryonic cells. For example, cytokinin concen- SE ) for six propagules harvested from a low-salinity, riverine
trations dropped in the experimentally manipulated R. site (open circles, dotted lines), a fringing site with inter-
mediate salinity (gray circles, dot-dash lines), and a hyper-
mangle propagules described above, most precipitously saline lagoon (black circles, solid lines). By ANOVA com-
in those injected with fresh water (Box 1, Fig. 2). Sim- paring populations within each developmental stage, only cy-
ilarly, mangrove propagules from the low-salinity riv- tokinin levels were significantly higher among young and
erine site showed the lowest levels of cytokinins of the intermediate-aged propagules from the hypersaline site rel-
three populations studied throughout ontogeny (Box 2, ative to the riverine and fringing sites.
Fig. 3), while the youngest propagules from the hy-
persaline site showed high concentrations.
ABA and cytokinins have long been regarded as an- stress (Schmulling et al. 1997). When ABA levels are
tagonists, showing opposing effects on cell division, high in tissues, cytokinin levels tend to be low, and
resource allocation, and stress reactions and interfer- vice versa. Such is the case for the seeds studied here
ence in each others synthetic pathways (Jacqmard et (Box 1, Fig. 1, Fig. 3). However, an exception exists
al. 1995, Cowan et al. 1999, Wilkinson and Davies in mangrove leaves, where both cytokinins and ABA
2002). ABA counteracts cytokinin-dependent activities concentrations are higher relative to non-mangrove
of several genes, including those involved with tran- species (Fig. 1) and may be able to coexist in different
scription of photosynthetic proteins, transitions be- cellular compartments. As cytokinins delay leaf senes-
tween C3 and C4 photosynthesis, and responses to salt cence, they could be implicated in the very long life-
 12 ELIZABETH FARNSWORTH Ecology, Vol. 85, No. 1
Special Feature

FIG. 4. Conceptual model linking actions of ABA, cytokinins, and their cofactors (calcium and sugars, respectively) to
environmental responsiveness, development, and fitness. Arrows indicate a positive (black) or negative (white) effect of
hormones on a given process; dotted lines indicate a temporal as well as physiological link between processes. Processes
are arrayed vertically according to the time scale over which they occur.

span of mangrove leaves. Indeed, measured Z and ZR
concentrations in leaves of five mangrove species cor-
relates positively with leaf lifespan (r2 5 0.954, linear
regression of leaf lifespan on measured mean Z and
ZR levels; Ellison 2002; unpublished data). However,
this correlative link needs to be explored in more detail
with experimental and comparative work.
DISCUSSION
Although measurement of hormone levels in plants
is highly suggestive of the roles they may play in the
autecology of plants, the story of hormone action is a
complicated one. Data on levels by themselves tend to
be correlative and should be interpreted with caution.
Quantifying levels alone will contribute to new hy-
potheses, but will not be sufficient in the long run to
elucidate hormone roles. Thus, plant ecologists and
physiologists must collaborate both to understand hor-
monal pathways fully and to comprehend the ecological
significance of molecular studies. Weyers and Paterson
(2001) aptly review the history of hormone concepts
and caution that simple doseresponse relationships do
not always hold because of lags, feedbacks, differing
rates of anabolism and catabolism in synthetic path-
ways, complex actions of cofactors (such as calcium),
differing sensitivities among plant tissues, and the in-
teraction of stimuli and hormones in inducing signal
cascades. The early data from field studies reviewed
here must be accompanied by controlled experiments
that demonstrate: replicable correlations of hormones
with responses; deletion and reinstatement of hormone
action; specificity; isolation of biochemical responses
from confounding factors; distinguishable roles of de
novo production vs. translocation and activation of
compartmentalized hormones; relevance to actual per-
formance of plant phenotypes in the field; and gener-
ality among species.
Chapin (1991) developed a cohesive model by which
ABA serves to reduce photosynthesis, stomatal con-
ductance, and plant relative growth rate in conditions
of drought, flooding, and low nitrogen availability. Lat-
er, this model was extended to encompass ABA and
cytokinins (Chapin et al. 1993). In a similar vein, Lesh-
em and Kuiper (1996) proposed a hypothetical gen-
eral adaptation syndrome for plants, which incorpo-
rates new understanding about signal transduction
pathways, chaperonin proteins, and other phytohor-
mones. Most recently, ubiquitous and evolutionarily
January 2004 PHYTOHORMONAL ECOLOGY
conserved heat shock proteins have been found to mod-
ulate both short-term responses to stress and to stabilize
development of the phenotype as a whole in both plants
and animals (Queitsch et al. 2002). Subsequent exper-
imental work has contradicted certain elements of
Chapins findings, which may reflect the complexity of
hormone action discussed above and points to the need
for testing these hypotheses (Coleman and Schneider
1996).
Together, this new information and findings from
comparative studies allow us to elaborate on the Chapin
et al. (1993) model, incorporating a sense of both
changing selection pressures and shifting hormone ac-
tion over time, encompassing both temporary accli-
mation and longer-term ontogeny (Fig. 4). This model
can ultimately incorporate synergistic effects of mul-
tiple hormones, cofactors, and environmental drivers
to provide more mechanistic explanations of plant be-
havior, linked traits, and evolution. For example, as
experiments elucidate the influence of multiple stress-
ors, such as competition, on the costs and benefits of
inducible defenses (e.g., Van Dam and Baldwin 2001,
Cipollini 2002) and light responsiveness (Donohue and
Schmitt 1999, Weinig and Delph 2001), it is becoming
clearer that phytohormones can help explain mecha-
nisms of plasticity in complex environments. A better
understanding of hormones may clarify several life his-
tory conundrums, which, to date, have been received
more attention in animals than in plants (Ricklefs and
Wikelski 2002). For example, how do life history con-
straints arise; do hormonal mechanisms permit only
certain combinations of physiological and architectural
states to occur in plants? Are correlated traits (such as
reduced seed dormancy, conservative photosynthesis,
longer leaf life spans, and salt tolerance) linked through
the pleiotropic (sensu McCourt 1999) nature of genes
that control hormones and their tissue-specific effects?
Age-specific studies of hormones may also help iden-
tify constraints of resource harvesting and biome-
chanics on plant growth and development (Enquist et
al. 1999:908) that shape the allometry of basic plant
traits as they grow.
ACKNOWLEDGMENTS
I thank Susan Mopper for the invitation to submit an article
for this Special Feature; J. M. Farrant, P. B. Tomlinson, E.
A. Kellogg, N. M. Holbrook, D. Stein, C. Baskin, J. Baskin,
P. Nel, and F. A. Bazzaz for inspiration; A. M. Ellison, S.
Mopper, A. Agrawal, and two anonymous reviewers for help-
ful comments on manuscript drafts; and staff of the Harvard
University Herbaria, Stellenbosch University, Possum Point
Biological Station, and Fairchild Tropical Garden for logis-
tical support. Research and writing were supported by Na-
tional Science Foundation grants IBN-9623313, DGE-
9714522, and DGE-0123490, the Deland Fund of the Arnold
Arboretum, and the New England Wild Flower Society.
LITERATURE CITED
Amzallag, G. N. 1999. Adaptive nature of the transition phas-
es in development: the case of Sorghum bicolor. Plant, Cell
and Environment 22:10351041.
13
Amzallag, G. N. 2001. Developmental changes in effect of
cytokinin and gibberellin on shoot K1 and Na1 accumu-
lation in salt-treated Sorghum plants. Plant Biology 3:319
325.
Ball, M. C. 1988. Salinity tolerance in the mangroves Ae-
giceras corniculatum and Avicennia marina. I. Water use
in relation to growth, carbon partitioning, and salt balance.
Australian Journal of Plant Physiology 15:447464.
Chapin, F. S., III. 1991. Integrated responses of plants to
stress. BioScience 41:2936.
Chapin, F. S., III, K. Autumn, and F. Pugnaire. 1993. Evo-
lution of suites of traits in response to environmental stress.
American Naturalist 142:S78S92.
Cipollini, D. F. 2002. Does competition magnify the fitness
costs of induced responses in Arabidopsis thaliana? A ma-
nipulative approach. Oecologia 131:514520.
Clark, D. A., and D. B. Clark. 2001. Getting to the canopy:
tree height growth in a neotropical rain forest. Ecology 82:
14601472.
Clipson, N. J. W., D. R. Lachano, and T. J. Flowers. 1988.
Salt tolerance in the halophyte Suaeda maritima L. Dum.:
abscisic acid concentrations in response to constant and
altered salinity. Journal of Experimental Botany 39:1381
1388.
Coleman, J. S., K. D. M. McConnaughay, and D. D. Ackerly.
1994. Interpreting phenotypic variation in plants. Trends
in Ecology and Evolution 9:187191.
Coleman, J. S., and K. Schneider. 1996. Evidence suggesting
that ABA may not regulate changes in growth and biomass
partitioning in response to low soil resource availability.
Special Feature
Oecologia 106:273278.
Cowan, A., A. Cairns, and B. Bartels-Rahm. 1999. Regu-
lation of abscisic acid metabolism: towards a metabolic
basis for abscisic acidcytokinin antagonism. Journal of
Experimental Botany 50:595603.
DAgostino, I. B., and J. J. Kieber. 1999. Cytokinin signaling
and the plant cell cycle. Current Opinions in Plant Biology
2:359364.
Davies, P. J. 1995. Plant hormones: physiology, biochemis-
try, and molecular biology. Kluwer Academic Publishers,
Dordrecht, The Netherlands.
Davies, S. J. 2001. Tree mortality and growth in 11 sympatric
Macaranga species in Borneo. Ecology 82:920932.
Donohue, K., and J. Schmitt. 1999. The genetic architecture
of plasticity to density in Impatiens capensis. Evolution 53:
13771386.
Dufty, A. M., Jr., J. Clobert, and A. P. Mller. 2002. Hor-
mones, developmental plasticity and adaptation. Trends in
Ecology and Evolution 17:190196.
Ellison, A. M. 2002. Macroecology of mangroves: large-
scale patterns and processes in tropical coastal forests.
Trees 16:181194.
Ellison, A. M., and E. J. Farnsworth. 1996. Spatial and tem-
poral variability in growth of Rhizophora mangle saplings:
links with insolation, herbivory and local sedimentation
regimes on Belizean coral cays. Journal of Ecology 84:
717731.
Elmqvist, T., and P. A. Cox. 1996. The evolution of vivipary
in flowering plants. Oikos 77:39.
Emery, R., N. Longnecker, and C. Atkins. 1998. Branch de-
velopment in Lupinus angustifolius L. II. Relationship with
endogenous ABA, IAA and cytokinins in axillary and main
stem buds. Journal of Experimental Botany 49:555562.
Enquist, B. J., G. B. West, E. L. Charnov, and J. H. Brown.
1999. Allometric scaling of production and life-history
variation in vascular plants. Nature 401:907911.
Eriksson, O. 2002. Ontogenetic niche shifts and their impli-
cations for recruitment in three clonal Vaccinium shrubs:
Vaccinium myrtillus, Vaccinium vitis-idaea, and Vaccinium
oxycoccos. Canadian Journal of Botany 80:635641.
 14 ELIZABETH FARNSWORTH
Evans, G. C. 1972. The quantitative analysis of plant growth.
University of California Press, Berkeley, California, USA.
Evans, M. M. S., and R. S. Poethig. 1995. Gibberellins pro-
mote vegetative phase change and reproductive maturity in
maize. Plant Physiology 108:475487.
Fankhauser, C. 2002. Light perception in plants: cytokinins
and red light join forces to keep phytochrome B active.
Trends in Plant Science 7:143145.
Farnsworth, E. J. 2000. The ecology and physiology of vi-
viparous and recalcitrant seeds. Annual Review of Ecology
and Systematics 31:107138.
Farnsworth, E. J., and A. M. Ellison. 1996. Sunshade adapt-
ability of the red mangrove, Rhizophora mangle (Rhizo-
phoraceae): changes through ontogeny at several levels of
biological organization. American Journal of Botany 83:
11311143.
Farnsworth, E. J., and J. M. Farrant. 1998. Reductions in
abscisic acid are linked with viviparous reproduction in
mangroves. American Journal of Botany 85:760769.
Farrant, J. M., N. W. Pammenter, P. Berjak, E. J. Farnsworth,
and C. W. Vertucci. 1996. Presence of dehydrin-like pro-
teins and levels of abscisic acid in recalcitrant (desiccation-
sensitive) seeds may be related to habitat. Seed Science
Research 6:175182.
Farrant, J. M., N. W. Pammenter, J. G. M. Cutting, and P.
Berjak. 1993. The role of plant growth regulators in the
development and germination of the desiccation-sensitive
(recalcitrant) seed of Avicennia marina. Seed Science Re-
search 3:5563.
Feder, M. E., and B. A. Block. 1991. On the future of animal
Special Feature
physiological ecology. Functional Ecology 5:136144.
Gilbert, S. 2001. Ecological developmental biology: devel-
opmental biology meets the real world. Developmental Bi-
ology 233:112.
Harmens, H., C. M. Stirling, C. Marshall, and J. F. Farrar.
2000. Is partitioning of dry weight and leaf area within
Dactylis glomerata affected by N and CO2 enrichment?
Annals of Botany 86:833839.
Heschel, M. S., and N. J. Hausmann. 2001. Population dif-
ferentiation for abscisic acid responsiveness in Impatiens
capensis (Balsaminaceae). International Journal of Plant
Sciences 162:12531260.
Hoffman-Benning, S., and H. Kende. 1992. On the role of
abscisic acid and gibberellin in the regulation of growth in
rice. Plant Physiology 99:11561161.
Honkanen, T., and E. Haukioja. 1998. Intra-plant regulation
of growth and plantherbivore interactions. Ecoscience 5:
470479.
Hutchison, C. E, and J. J. Kieber. 2002. Cytokinin signaling
in Arabidopsis. Plant Cell 14(Supplement):S47S59.
Ingram, J., and D. Bartels. 1996. The molecular basis of
dehydration tolerance in plants. Annual Review of Plant
Physiology and Plant Molecular Biology 47:377403.
Jacqmard, A., C. Houssa, and G. Bernier. 1995. Abscisic acid
antagonizes the effect of cytokinin on DNA-replication or-
igins. Journal of Experimental Botany 46:663666.
Jones, C. S. 1999. An essay on juvenality, phase change, and
heteroblasty in seed plants. International Journal of Plant
Sciences 160:S105S111.
Jordi, W., A. Schapendonk, E. Davelaar, G. M. Stoopen, C.
S. Pot, R. De Visser, J. A. van Rhijn, S. Gan, and R. M.
Amasing. 2000. Increased cytokinin levels in transgenic
PSAG12-IPT tobacco plants have large direct and indirect
effects on leaf senescence, photosynthesis, and N parti-
tioning. Plant, Cell and Environment 23:279289.
Karban, R., and J. S. Thaler. 1999. Plant phase change and
resistance to herbivory. Ecology 80:510517.
Kearsley, M. J. C., and T. G. Whitham. 1998. The devel-
opmental stream of cottonwoods affects ramet growth and
resistance to galling aphids. Ecology 79:178191.
Ecology, Vol. 85, No. 1
Ketterson, E. D., and V. Nolan Jr. 1999. Adaptation, exap-
tation, and constraint: a hormonal perspective. American
Naturalist 154:S4S25.
Kikuzawa, K., and D. D. Ackerly. 1999. Significance of leaf
longevity in plants. Plant Species Biology 14:3945.
Kushiro, T., E. Nambara, and P. McCourt. 2003. The key to
signalling. Nature 422:122.
Leshem, Y. Y., and P. J. C. Kuiper. 1996. Is there a GAS
(general adaptation syndrome) response to various types of
environmental stress? Biologia Plantarum 38:118.
Leung, J., and J. Giraudat. 1998. Abscisic acid signal trans-
duction. Annual Review of Plant Physiology and Plant Mo-
lecular Biology 49:199222.
Lyons, T. M., and J. D. Barnes. 1998. Influence of plant age
on ozone resistance in Plantago major. New Phytologist
138:8389.
McConnaughay, K. D. M., and J. S. Coleman. 1999. Biomass
allocation in plants: ontogeny or optimality? A test along
three resource gradients. Ecology 80:25812593.
McCourt, P. 1999. Genetic analysis of hormone signalling.
Annual Review of Plant Physiology and Plant Molecular
Biology 50:219243.
Mercier, H., and L. Endres. 1999. Alteration of hormonal
levels in a rootless epiphytic bromeliad in different phe-
nological phases. Journal of Plant Growth Regulation 18:
121125.
Miklashevichs, E., and R. Walden. 1997. Plant mutants with
altered responses to cytokinins. Physiologia Plantarum 100:
528533.
Mitchell-Olds, T. 2001. Arabidopsis thaliana and its wild
relatives: a model system for ecology and evolution. Trends
in Ecology and Evolution 16:693700.
Ng, P. P., A. L. Cole, P. A. Jameson, and J. A. Mcwha. 1982.
Cytokinin production by ectomycorrhizal fungi. New Phy-
tologist 91:5762.
Paterson, N. W., L. Herdman, and J. D. B. Weyers. 2001.
Relative control potential of abscisic acid, carbon dioxide,
and light in responses of Phaseolus vulgaris stomata. Phy-
siologia Plantarum 111:412418.
Pigliucci, M. 1997. Ontogenetic phenotypic plasticity during
the reproductive phase in Arabidopsis thaliana. American
Journal of Botany 84:887895.
Poethig, R. S. 1990. Phase change and the regulation of shoot
morphogenesis in plants. Science 250:923930.
Pons, T. L., W. Jordi, and D. Kuiper. 2001. Acclimation of
plants to light gradients in leaf canopies: evidence for a
possible role for cytokinins transported in the transpiration
stream. Journal of Experimental Botany 52:15631574.
Poorter, H., and P. Pothmann. 1992. Growth and carbon econ-
omy of a fast-growing and a slow-growing grass species
as dependent on ontogeny. New Phytologist 120:159166.
Queitsch, C., T. A. Sangster, and S. Lindquist. 2002. Hsp90
as a capacitor of phenotypic variation. Nature 417:618
624.
Ratcliffe, O. J., D. J. Bradley, and E. S. Coen. 1999. Sepa-
ration of shoot and floral identity in Arabidopsis. Devel-
opment 126:11091120.
Reich, P. B., D. S. Ellsworth, M. B. Walters, J. M. Vose, C.
Gresham, J. C. Volin, and W. D. Bowman. 1999. Generality
of leaf trait relationships: a test across six biomes. Ecology
80:19551969.
Ricklefs, R. E., and M. Wikelski. 2002. The physiology/life-
history nexus. Trends in Ecology and Evolution 17:462
468.
Rock, C. D. 2000. Pathways to abscisic acid-regulated gene
expression. New Phytologist 148:357396.
Romero, J. M., and T. Maranon. 1996. Allocation of biomass
and mineral elements in Melilotus segetalis (annual sweet-
clover): effects of NaCl salinity and plant age. New Phy-
tologist 132:565573.
January 2004 PHYTOHORMONAL ECOLOGY
Sachs, T. 1999. Node counting: an internal control of bal- Voesenek, L. A. C. J., and C. W. P. M. Blom. 1996. Plants
anced vegetative and reproductive development. Plant, Cell
and Environment 22:757766.
Schmulling, T., S. Schafer, and G. Romanov. 1997. Cyto-
kinins as regulators of gene expression. Physiologia Plan-
tarum 100:505519.
Sinervo, B., and E. Svensson. 1998. Mechanistic and selec-
tive causes of life history trade-offs and plasticity. Oikos
83:432442.
Smith, S. M., and S. C. Snedaker. 1995. Salinity responses
in two populations of viviparous Rhizophora mangle L.
seedlings. Biotropica 27:435440.
Smith, S. M., Y. Y. Yang, Y. Kamiya, and S. C. Snedaker.
1996. Effect of environment and gibberellins on the early
growth and development of the red mangrove, Rhizophora
mangle L. Plant Growth Regulation 20:215223.
Thimann, K. V. 1987. Plant senescence: a proposed integra-
tion of the constituent processes. Pages 119 in W. W.
Thompson, E. A. Nothnagle, and R. C. Huffaker, editors.
Plant senescence: its biochemistry and physiology. Amer-
ican Society of Plant Physiologists, Rockville, Maryland,
USA.
Thomas, D. S., and D. Eamus. 2002. Seasonal patterns of
xylem sap pH, xylem abscisic acid concentration, leaf water
potential and stomatal conductance of six evergreen and
deciduous Australian savanna tree species. Australian Jour-
nal of Botany 50:229236.
Thompson, J. E., R. L. Legge, and T. F. Barber. 1987. The
role of free radicals in senescence and wounding. New
Phytologist 105:317344.
Tomlinson, P. B. 1994. The botany of mangroves. Cambridge
University Press, Cambridge, UK.
Trewavas, A. J. 1986. Resource allocation under poor growth
conditions. A major role for growth substances in devel-
opmental plasticity. Pages 3177 in D. H. Jennings and A.
J. Trewavas, editors. Plasticity in plants. Company of Bi-
ologists, Cambridge, UK.
Valdes, A. E., M. L. Centeno, S. Espinel, and B. Fernandez.
2002. Could plant hormones be the basis of plant matu-
ration indices in Pinus radiata? Plant Physiology and Bio-
chemistry 40:211216.
Van Dam, N. M., and I. T. Baldwin. 2001. Competition me-
diates costs of jasmonate-induced defences, nitrogen ac-
quisition and transgenerational plasticity in Nicotiana at-
tenuata. Functional Ecology 15:406415.
15
and hormones: an ecophysiological view on timing and
plasticity. Journal of Ecology 84:111119.
Vogel, J. P., K. Woeste, A. Theologis, and J. J. Kieber. 1998.
Recessive and dominant in the ACC synthase 5 gene of
Arabidopsis result in cytokinin-insensitivity and ethylene
overproduction respectively. Proceedings of the National
Academy of Sciences (USA) 95:47664771.
Wait, D. A., C. G. Jones, and J. S. Coleman. 1998. Effects
of nitrogen fertilization on leaf chemistry and beetle feed-
ing are mediated by leaf development. Oikos 82:502514.
Wang, Y. Y., S. Mopper, and K. H. Hasenstein. 2001. Effects
of salinity on endogenous ABA, IAA, JA, and SA in Iris
hexagona. Journal of Chemical Ecology 27:327342.
Weigel, D., and O. Nilsson. 1995. A developmental switch
sufficient for flower initiation in diverse plants. Nature 377:
495500.
Weiler, E. W. 1980. Radioimmunoassays for the differential
and direct analysis of free and conjugated abscisic acid in
plant extracts. Planta 148:262272.
Weinig, C., and L. F. Delph. 2001. Phenotypic plasticity early
in life constrains developmental responses later. Evolution
55:930936.
Weyers, J. D. B., and N. W. Paterson. 2001. Plant hormones
and the control of physiological processes. New Phytologist
152:375408.
Wilkinson, S., and W. J. Davies. 2002. ABA-based chemical
signalling: the co-ordination of responses to stress in plants.
Plant, Cell and Environment 25:195210.
Yang, J., J. Zhang, Z. Wang, Q. Zhu, and L. Liu. 2002. Abs-
Special Feature
cisic acid and cytokinins in the root exudates and leaves
and their relationship to senescence and remobilization of
carbon reserves in rice subjected to water stress during
grain filling. Planta 215:645652.
Ye, Z. H. 2002. Vascular tissue differentiation and pattern
formation in plants. Annual Review of Plant Biology 53:
183202.
Zheng, W. J., W. Q. Wang, and P. Lin. 1999. Dynamics of
element contents during the development of hypocotyls and
leaves of certain mangrove species. Journal of Experimen-
tal Marine Biology and Ecology 233:247257.
Zotz, G., V. Thomas, and W. Hartung. 2001. Ecophysiolog-
ical consequences of differences in plant size: abscisic acid
relationships in the epiphytic orchid Dimerandra emargin-
ata. Oecologia 129:179185.
 Ecology, 85(1), 2004, pp. 1627
q 2004 by the Ecological Society of America

PLANT HORMONES REGULATE FAST SHOOT ELONGATION UNDER

WATER: FROM GENES TO COMMUNITIES
L. A. C. J. VOESENEK,1,3 J. H. G. M. RIJNDERS,2 A. J. M. PEETERS,1 H. M. VAN DE STEEG,2 AND
H. DE KROON2
1Department of Plant Ecophysiology, University of Utrecht, Sorbonnelaan 16, 3584 CA Utrecht, The Netherlands
2 Department of Experimental Plant Ecology, University of Nijmegen, Toernooiveld, 6525 ED Nijmegen, The Netherlands

Abstract. Flooding affects the abundance and distribution of plant species worldwide.
Many plants are damaged or even killed by flooding events due to the associated oxygen
deprivation in cells. Stimulated shoot elongation is an important adaptive mode that can
restore contact of leaves with the atmosphere above the water surface. This strongly im-
proves inward diffusion of oxygen and the rate of photosynthesis. Fast elongation of sub-
merged petioles of the model plant Rumex palustris involves the integrated action of the
plant hormones ethylene, auxin, gibberellin, and abscisic acid. The closely related Rumex
acetosa is unable to switch on petiole elongation when submerged. In a comparative study
of these two Rumex species, we found that the response to the gaseous phytohormone
ethylene, which accumulates in plant tissues during submergence, explains their contrasting
elongation behavior. In order to study the importance of this shoot elongation response in
the distributional patterns of plants in natural floodplains, we quantified the ethylene-
induced elongation response of 22 plant species occurring in the Rhine River floodplain.
These results were compared with the results of a multivariate analysis based on 84 veg-
etation surveys performed in the same area. The species compositions of the surveys were
Special Feature

grouped along two environmental gradients: flooding duration and soil dehydration after
the floodwater subsided. If we superimpose the ethylene-induced elongation capacity on
these vegetation data, it becomes clear that the capacity to elongate upon exposure to
ethylene positively correlates with flooding duration and negatively with soil dehydration.
Based on this analysis, we conclude that the capacity to elongate is an important selective
trait in field distribution patterns of plants in flood-prone environments. Fast shoot elon-
gation under water seems to be a favorable trait only in environments with shallow and
prolonged flooding events, while costs associated with this response prevent its expression
in sites with deep floods, sites with floods short in duration, or in sites in which flood water
recedes rapidly. The approach outlined in this paper may be more widely applicable in
ecological studies that aim to understand the functional relationship between plant traits
and species distributions along environmental gradients.
Key words: abscisic acid; auxin; costbenefit analysis; elongation growth; ethylene; flooding
duration; gibberellin; plant species distribution; Rumex; soil dehydration; submergence; upscaling.

INTRODUCTION which hamper plant performance (Voesenek et al.

Flooding can occur in many ecosystems worldwide,
including river floodplains, salt marshes, bogs, dune
slacks, and swamp forests (Maltby 1991). A typical
feature of these environments is their episodically an-
aerobic nature. Plants need molecular oxygen to syn-
thesize large amounts of ATP. Shortage of oxygen is,
therefore, considered to be a major factor that nega-
tively affects plant growth in flood-prone habitats.
Flooding also leads to accumulation of toxic soil com-
ponents, reduction in light availability when plants are
completely submerged, and the production of free rad-
icals upon reaeration when floodwater subsides, all of
Manuscript received 25 November 2002; revised 17 April
2003; accepted 19 April 2003. Corresponding Editor (ad hoc):
S. Mopper. For reprints of this Special Feature, see footnote 1,
p. 3.
3 E-mail: L.A.C.J.Voesenek@bio.uu.nl
16
1992b, Crawford et al. 1994).
Adaptations to flooding
Some plants species avoid flooding stress through
life history tactics and survive a period of flooding as
dormant seeds or quiescent perennating organs. Con-
comitantly, growth and reproduction are facilitated
within the window of favorable circumstances (Menges
and Waller 1983, Blom and Voesenek 1996). Other
species exploit amelioration and true tolerance tactics
to survive the flooding stress. Amelioration implies that
plants are exposed to environmental disturbances, but
the adverse effects are alleviated. Examples in relation
to flooding include (1) fast elongation of shoot organs
to allow contact of a snorkel (often leaf blades) with
the atmosphere to reestablish gas exchange (Kende et
al. 1998, Voesenek and Blom 1999, Peeters et al. 2002;
see Plate 1). (2) development of longitudinally inter-
January 2004 PHYTOHORMONAL ECOLOGY 17

connected gas-filled channels to facilitate interorgan

gas diffusion (Visser et al. 1997, Jackson and Arm-
strong 1999, Colmer 2003), (3) development of a bar-
rier to radial oxygen loss to prevent excessive loss of
diffusing oxygen to the anaerobic soil (Colmer et al.
1998), (4) initiation and outgrowth of adventitious
roots (Visser et al. 1996), and (5) continuation of pho-
tosynthesis under water that generates oxygen and car-
bohydrates (He et al. 1999, Vervuren et al. 1999, Rijnd-
ers et al. 2000). True tolerance refers to situations in
which oxygen deprivation is not dispelled, but adaptive
metabolic pathways are switched on to ensure enhanced
survival rates. These comprise (1) the generation of
energy without oxygen via continued glycolysis and
fermentation (Perata and Alpi 1993, Vartapetian and
Jackson 1997), (2) a reduction of metabolic rates to
levels that just sustain maintenance (so-called anaer-
obic dormancy), (3) avoidance of cytoplasmic acidosis
(Saglio et al. 1999), and (4) protection against post-
anoxia injury via antioxidants or enzymes that scav-
enge oxygen radicals (Smirnoff 1995).
Temporary floods differ considerably in seasonal
timing, duration, depth, and frequency (Vervuren et al.
2003). Variation in these factors creates a continuum

of flooding regimes in flood-prone habitats. The ability
of some plant species to elongate petioles and/or stems
upon submergence correlates with the type of flooding
regime in their natural environment, because there are
costs involved with underwater elongation and because
the benefits will not always be apparent. In this paper,
we focus on stimulated shoot (petiole or stem) elon-
gation as a trait in which plant hormone sensitivity
strongly interacts with the ecology of plant species in
flood-prone environments. This flooding-induced elon-
gation response, also described as depth accommo-
dation (Ridge 1987), returns the shoot to the atmo-
sphere above the water surface, needed for long term
survival.
Costs of elongation
Increased cell division rates and the synthesis of new
cell walls during fast cell elongation require substantial
amounts of energy and carbohydrates. This is illus-
trated by work of Setter and Laureles (1996) in which
elongating rice varieties had a lower submergence tol-
erance compared to nonelongating varieties. When the
elongation capacity was reduced artificially by the gib-
berellin biosynthesis inhibitor paclobutrazol, survival
increased significantly (Setter and Laureles 1996), sug-
gesting that the resources saved by refraining from
elongation were available for survival. Other costs of
submergence-induced elongation may become apparent
when floodwaters subside. Elongated petioles and
leaves of some species (e.g., Ranunculus repens, Cal-
tha palustris) are brittle and weak, desiccate rapidly
compared to nonsubmerged control leaves (Ridge
1987), and the plant parts typically have high mortality
rates after desubmergence (Voesenek and Blom 1989 a).
Special Feature
PLATE. 1. Several Rumex species have the capacity to
stimulate petiole elongation once submerged. This growth
response reestablishes contact of leaves with the atmosphere
and thus enhances plant survival. Photo credit: F. Millenaar.
Fast elongation will only persist as a strategy if these
costs are outweighed by benefits such as improved aer-
ation and restored rates of photosynthesis (Ridge
1987). Enhanced elongation rates during submergence
increase plant survival, only if contact between leaf
blades and the atmosphere is restored. Thus, fast elon-
gation under water will be a high-risk strategy when
flooding events are of short duration and when the sub-
mergence depths go beyond the elongation capacity of
the plant (Ridge 1987). We therefore hypothesize that
plant traits associated with submergence-induced shoot
elongation predominantly occur in plant species from
flood-prone habitats characterized by prolonged, but
shallow, submergence.
To test this hypothesis we formulated two research
questions:
1) Which mechanisms are responsible for shoot
growth stimulation in certain species upon submer-
gence and lack of this response in others?
2) Can we relate these opposing elongation responses
of plant species upon submergence to their distribu-
tional patterns in natural flood-prone environments?
 18 L. A. C. J. VOESENEK ET AL.
Special Feature
FIG. 1. Submergence-induced hyponastic growth and petiole elongation in Rumex palustris, shown as a time series.
To answer the first question, we unraveled the signal
perception and transduction cascade involved in sub-
mergence-induced petiole elongation in the dicot Ru-
mex palustris. Fast elongation of submerged petioles
of R. palustris involves the integrated action of at least
four plant hormones: ethylene, auxin, gibberellin, and
abscisic acid. These hormones are main players in the
signal perception and transduction network. It is our
aim to shed light on key steps in the transduction net-
work that make the difference between species that are
able to switch on fast elongation (e.g., R. palustris)
and those that are not (e.g., the closely related species
Rumex acetosa).
In order to answer the second question, we scaled
up from a plant trait (enhanced shoot elongation) to
the level of the river floodplain community. The elon-
gation capacity of 22 river floodplain species was ex-
perimentally tested and the results were superimposed
on a vegetationenvironment relationship using 84
vegetation surveys from floodplains along the Rhine
River in the Netherlands. By this approach, we quantify
the relationship between the capacity to elongate and
the plant species distribution along an environmental
gradient.
SUBMERGENCE-INDUCED SHOOT ELONGATION
IN RUMEX PALUSTRIS
In natural river floodplains, Rumex palustris has a
biennial or short-lived perennial life cycle. Shortly af-
Ecology, Vol. 85, No. 1
ter germination it develops a rosette. Vegetative ro-
settes as well as bolting plants can be exposed to natural
floods and both life cycle phases have the capacity to
respond to submergence with enhanced elongation of
petioles and stems, respectively (Voesenek and Blom
1989a, Van der Sman et al. 1991, 1992).
Most work on Rumex has focused on submergence-
induced elongation of rosette leaves. Rosette leaves
assume a horizontal position, close to the substrate un-
der nonsubmerged conditions. The orientation of leaves
changes dramatically into an almost vertical position
after ;6 h of submergence (Fig. 1; Peeters et al. 2002,
Cox et al. 2003). This so-called hyponastic growth is
caused by elongation of several cells located at the
basal abaxial side of the petiole (M. C. H. Cox, un-
published data). When the angle between the petiole
and the horizontal is larger than 40508, the entire pet-
iole starts to elongate (Cox et al. 2003). This elongation
growth is distributed equally over the petiole (Rijnders
et al. 1996) and is responsible for 80% of the whole-
leaf (petiole 1 blade) elongation response, indicating
that the contribution of the leaf blade in underwater
elongation is very restricted (Voesenek and Blom
1989b). Elongation of existing petioles in Rumex is
predominantly driven by cell elongation; cell division
is not important (Voesenek et al. 1990). Submergence-
induced elongation in R. palustris varies between
leaves, with the strongest elongation responses in the
youngest leaves (Fig. 1; Voesenek and Blom 1989 a).
January 2004 PHYTOHORMONAL ECOLOGY
SENSING SUBMERGENCE
Ethylene
The gaseous phytohormone ethylene plays a key role
in the perception of environmental changes induced by
submergence. Nearly all cells of higher plants produce
ethylene in low amounts (Abeles 1992). Endogenous
levels are controlled by a subtle balance between pro-
duction rates, conjugation rates, and diffusion to the
outside environment (Voesenek and Blom 1999). When
a plant is surrounded by air, ethylene is vented off quite
rapidly, due to the relatively high diffusion coefficients
for ethylene in air and the steep concentration gradient
from the interior of the cell to the air outside the plant.
The situation changes dramatically when a plant is sub-
merged completely in water. Gases diffuse ;10 000
times slower in water than in air, resulting in fast in-
crease of the endogenous ethylene concentration, as-
suming that the production rate is unchanged (Jackson
1985, Ridge 1987).
A 20-fold increase in ethylene concentration in
shoots of R. palustris has been observed within 1 h of
submergence (Banga et al. 1996), resulting in an en-
dogenous ethylene concentration (1 mL/L) that is high
enough to saturate ethylene-induced elongation in Ru-
mex under nonsubmerged conditions (Voesenek et al.
1996). The importance of ethylene accumulation is fur-
ther illustrated by the observation that silver ions,
which interfere with ethylene perception and thus in-
hibit the action of this hormone, prevent hyponastic
growth and stimulated petiole elongation (Voesenek
and Blom 1989a).
Ethylene binds specifically to members of a family
of high affinity receptor proteins belonging to the so-
called two-component regulators (Bleecker and Kende
2000). Mutants of Arabidopsis thaliana containing one
amino acid substitution at position 65 in the ethylene
receptor molecule are unable to bind ethylene and are
insensitive to applied ethylene in a broad spectrum of
responses (Bleecker and Kende 2000). A cDNA clone
(Rp-ERS1) similar to an ethylene receptor gene of Ar-
abidopsis thaliana has been isolated for R. palustris
(Vriezen et al. 1997). The expression of this gene is
strongly up regulated by submergence, ethylene, and
low oxygen concentrations; down-regulation occurs
upon desubmergence (Voesenek et al. 1997, Vriezen et
al. 1997). These data suggest that R. palustris plants
respond to submergence by increasing the number of
their ethylene receptors. Since ethylene receptors are
active in the absence of ethylene and negatively reg-
ulate ethylene responses, more receptor molecules re-
sult in a desensitization of the cells to ethylene (Hua
and Meyerowitz 1998). The loss of sensitivity to eth-
ylene during submergence has no effect on the elon-
gation response of R. palustris. This is probably related
to the over saturating concentrations of ethylene inside
plants that are submerged for several hours (Voesenek
et al. 1993a).
19
Low oxygen
In addition to ethylene, hyponastic growth and en-
hanced petiole elongation in R. palustris can also be
induced by subambient oxygen concentrations of 2
10% (Voesenek et al. 1997). A similar observation has
been made for coleoptiles of rice (Horton 1991). When
applied together (low oxygen 1 ethylene), a stronger
elongation response was observed than under separate
treatments with ethylene and low oxygen. Low levels
of oxygen exert their effect on petiole elongation in
Rumex not independent of ethylene. When low oxygen
is applied together with L-alpha-(2-aminoethoxyvi-
nyl)-glycine (AVG), an inhibitor of ethylene biosyn-
thesis, the elongation response is completely abolished.
Addition of 1-aminocyclopropane-1-carboxylic acid
(ACC), a precursor of ethylene, restores the low oxygen
response, indicating that the effect of AVG is specific
(Voesenek et al. 1997). Low oxygen levels do not affect
the ethylene production rate, but cause an increase in
the sensitivity of the petiole towards ethylene. The ex-
act nature of this shift is still unknown.
We suggest that the very fast elongation that occurs
in R. palustris when low oxygen and elevated ethylene
levels are both present, functions as a mechanism to
Special Feature
ensure quick leaf blade emergence. The oxygen-lim-
iting condition can occur in nature when plants are
submerged in turbid water, and therefore, produce little
photosynthetic oxygen. Under less turbid conditions,
photosynthetically derived oxygen improves not only
the oxygen status of the whole plant (Rijnders et al.
2000), but also increases its chances of survival con-
siderably (Nabben et al. 1999). A lower petiole elon-
gation rate driven only by ethylene under these con-
ditions would be less harmful in terms of survival.
TRANSDUCTION OF SUBMERGENCE SIGNALS
Auxin and gibberellins
Thirty years ago, Musgrave and colleagues (1972)
stated that at least one other plant hormone in addition
to ethylene is required for fast underwater elongation.
These authors studied stem elongation in Callitriche
platycarpa and demonstrated that gibberellin was the
additional hormone. A similar ethylenegibberellin re-
quirement is also described for deepwater rice (Kende
et al. 1998). In other plant species (e.g., Ranunculus
sceleratus) that extend rapidly when submerged, the
second plant hormone was auxin. It was even suggested
that all three hormones mentioned, ethylene, gibber-
ellins, and auxin, are required for sustained elongation
under water. Auxin might play a role in the initiation
of rapid elongation, whereas gibberellins might be re-
sponsible for longer-term maintenance of fast growth
(Ridge 1987).
Auxin is predominantly produced in shoot apices and
in young expanding leaf blades (Ljung et al. 2001).
This plant hormone is transported in a polar manner to
petioles, stems, and roots, mediated by influx and efflux
 20 L. A. C. J. VOESENEK ET AL.
carrier proteins located in the membrane (Muday and
DeLong 2001). Removal of leaf blades, the putative
source of auxin, abolishes hyponastic growth in sub-
merged R. palustris plants for many hours. Similarly
N-naphthylphthalamic (NPA), an inhibitor of auxin ef-
flux, delays hyponastic growth. The response to leaf
blade removal could be rescued by 1-naphthaleneacetic
acid (1-NAA), an artificial auxin that freely diffuses
into cells (M. C. H. Cox and L. A. C. J. Voesenek,
unpublished data). We conclude that polar auxin trans-
port from the leaf blades to the basal side of the petioles
is required for submergence-induced hyponastic
growth. These results further point toward an inter-
action between ethylene that accumulates in submerged
petioles and auxin. The precise nature of this plant
hormone interaction is not yet known.
Inhibitors of gibberellin (GA) biosynthesis inhibit
partially submergence-induced petiole elongation in R.
palustris. This inhibition could be rescued completely
upon application of various GAs, indicating that the
effect of the inhibitors is specific and nontoxic (Rijnd-
ers et al. 1997). We conclude that, as for deepwater
rice, a certain endogenous concentration of GA is re-
quired for underwater shoot elongation in Rumex. GA
Special Feature
levels and the sensitivity of petiole tissues of R. pal-
ustris towards GA are regulated by submergence. The
concentrations of both GA1 and its precursor GA20 is
increased during 24 h of submergence (Rijnders et al.
1997). Furthermore, submergence induces an increase
in the sensitivity of Rumex petioles towards GA. The
effects of submergence on GA concentration and sen-
sitivity can by mimicked completely by exposing plants
to elevated ethylene levels, indicating that ethylene in-
teracts with GA in this response (Rijnders et al. 1997).
Abscisic acid (ABA)
ABA plays a major role in seed maturation and ger-
mination, as well as in adaptations to abiotic stress
(Buchanan et al. 2000). Besides its important role in
drought avoidance ABA also seems to be important in
the regulation of shoot elongation in submerged plants.
In both deepwater rice and Scirpus micronatus, ABA
inhibits submergence-induced shoot elongation. Fur-
thermore, a fast decline in endogenous ABA concen-
trations in shoots is observed when these species were
submerged or exposed to elevated levels of ethylene
(Hoffmann-Benning and Kende 1992, Lee et al. 1996).
It has been assumed that the sensitivity to GA of deep-
water rice internodes is decreased by ABA and that an
ethylene-induced down-regulation of ABA increases
this sensitivity. In this respect, ABA can be seen as a
negative regulator of GA action (Hoffmann-Benning
and Kende 1992, Ritchie and Gilroy 1998).
An inhibition of submergence-induced shoot elon-
gation by ABA is also observed in the dicot R. pal-
ustris. The ABA concentration in this species declined
80% within 1 h of submergence or exposure to elevated
ethylene levels. Submergence of R. palustris plants pre-
Ecology, Vol. 85, No. 1
treated with 1-MCP, an inhibitor of ethylene perception,
prevented this fast down-regulation of ABA (J. J. Ben-
schop and L. A. C. J. Voesenek, unpublished data).
Taken together, all these results suggest that down-reg-
ulation of ABA, mediated by ethylene, is a prerequisite
for fast shoot elongation under water in both monocots
and dicots.
VARIATION IN UNDERWATER ELONGATION: AN
ANALYSIS WITH A BOOLEAN NETWORK
The signaling network resulting in enhanced petiole
elongation and hyponastic growth in R. palustris when
completely submerged is summarized in Fig. 2. We
applied the Boolean language, which is a digital (nu-
meric) formalism that allows a qualitative description
of signal transduction networks (Genoud and Metraux
1999, Genoud et al. 2001, Cox et al. 2003). The Bool-
ean language converts signaling elements into binary
state elements, that either have an on (1) or an off
(0) state. Interfering signals can be seen as inputs
that join in a Boolean gate or operator. The combination
of signals results in a specific output signal (Genoud
and Metraux 1999). Complete submergence of R. pal-
ustris always results in an enhanced endogenous eth-
ylene concentration. Furthermore, when flooding co-
incides with reduced light levels it can induce a strong
decline of the endogenous oxygen concentration in the
petioles. Subsequently, low oxygen conditions increase
the sensitivity of petioles towards ethylene. The sub-
mergence-oxygen-ethylene interaction is characterized
by two NOT gates, indicating that in both cases the
output signal is negatively regulated by the input. The
oxygenethylene pathway joins with the ethylene-
alone pathway in an OR gate, indicating that an input
signal generated by at least one of the pathways is
required to induce an output signal. This output signal
downregulates the ABA concentration, upregulates
both the concentration of and the sensitivity to GA,
and acts as an input signal leading to hyponastic
growth. However, this differential growth response will
only take place when the petiole angle is smaller than
808. The ethylene output signal and the angle ,808
signal join in a so-called AND operator, indicating that
both signals are absolutely required to evoke an output.
Preliminary evidence indicates that this output signal
modifies local auxin concentrations and/or auxin sen-
sitivities, leading to a differential growth stimulus with
a hyponastic curvature as a result. Enhanced petiole
elongation takes place when the signal associated with
the reduced ABA level and the GA signal join in an
AND operator. However, it will only lead to elongation
growth when the petiole angle is above 40508, dem-
onstrating that hyponastic growth interacts with petiole
elongation. From this, we hypothesize that the petiole
angle generates a signal that is required as an input
signal for both hyponastic growth and petiole elon-
gation. The nature of this signal is yet unknown.
January 2004 PHYTOHORMONAL ECOLOGY
FIG. 2. Model of signal-transduction network involved in submergence-induced hyponastic growth and petiole elongation
of Rumex palustris.
Given the signaling network as revealed to date, we
can now begin to answer one of the central questions
of this paper: why are certain species able to elongate
and others not? Is it possible to trace back a fundamental
difference in growth response to certain steps or even
single genes in the signal-transduction cascade? To an-
swer these questions, we compared the signal transduc-
tion network of R. palustris with that of the closely
related species Rumex acetosa. Rumex acetosa L. is a
perennial species from grasslands. In river floodplains,
it rarely occupies flooded sites. Under experimental con-
ditions it is less tolerant to complete submergence than
R. palustris is. This is true for seeds (Voesenek et al.
1992a), seedlings (Voesenek et al. 1993b) and adult
plants (Voesenek 1990). Petioles of R. acetosa do not
demonstrate hyponastic growth and fast rates of petiole
elongation when exposed to complete submergence
(Voesenek and Blom 1989a). Despite the absence of an
elongation response, ethylene does accumulate to con-
centrations that induce responses in R. palustris (Voe-
senek et al. 1993a). Moreover, an inhibition of petiole
elongation is observed when R. acetosa was exposed to
21
Special Feature
gas mixtures containing elevated ethylene concentra-
tions (Voesenek and Blom 1989a).
In contrast to R. palustris, the GA and ABA concen-
trations do not change when R. acetosa is submerged or
exposed to elevated ethylene levels (Rijnders et al. 1997;
J. J. Benschop and L. A. C. J. Voesenek, unpublished
data). Furthermore, this species exhibits a decrease in
GA sensitivity during submergence and in the presence
of ethylene, rather than an increase (Rijnders et al.
1997). Based on these data, we conclude that the decline
in the ABA concentration and the increase in the GA
concentration/sensitivity typical for submerged R. pal-
ustris plants is absent in submerged shoots of R. acetosa.
Interestingly, if we expose R. acetosa to elevated GA
concentration in a nonsubmerged (air) environment, a
significant stimulation of petiole elongation is observed
(Rijnders et al. 1997).
We conclude that in R. acetosa the downstream sig-
nal transduction pathway from GA to petiole elongation
is intact and that it can be switched on by the appro-
priate signal. However, from our results, it is clear that
the phytohormone ethylene is unable to switch on this
 22 L. A. C. J. VOESENEK ET AL.
downstream cascade. We speculate that in R. acetosa
other input signals than ethylene will be able to induce
enhanced elongation. One such signal may be low light
and/or decreased red to far red ratios. These signals,
perceived by phytochrome proteins, generate longer
stems and petioles in a large number of plant species
as part of the shade avoidance response (Schmitt et al.
1999, Smith 2000).
The main difference explaining the variation in sub-
mergence-induced petiole elongation between R. pal-
ustris and R. acetosa is now reduced to essentially one
step in the network: the interaction between ethylene-
generated signals with downstream targets such as the
growth promoting hormone GA. We hypothesize that
ethylene in R. palustris interacts with genes whose
products are able to regulate ABA and GA in such a
way that elongation growth is stimulated. This specific
interaction is different or absent in R. acetosa.
In order to scale up this result to riverine plant com-
munities, we must ask the question whether the differ-
ences in response between the two Rumex species are
indicative for a larger number of plant species, and wheth-
er the essential difference in a plant trait is predictive for
the distribution of the species in the field. To this aim,
Special Feature
we now first analyze the species composition of herba-
ceous plant communities along the lower river Rhine in
the Netherlands, as an example of a particular gradient
along which flooding intensity varies strongly. Next, for
a group of 22 species from these communities, including
both Rumex species, we examine the shoot elongation
response to elevated levels of ethylene.
METHODS
Field survey of plant species distribution
and hydrological characteristics
Floodplains in The Netherlands are characterized by
multiple variations in elevation, both natural and hu-
man made, resulting in a variety of habitats including
river banks, natural levees, floodplain channels, back
swamp areas with floodplain channels, summer flood-
ing protecting embankments, and flood preventing
dikes. Elevational variation in combination with runoff
pattern (free or obstructed) results in the formation of
flooding gradients perpendicular to the river and a clear
vegetation zonation (Sykora et al. 1988, Blom et al.
1990, 1996).
We examined variation in floodplain vegetation, and
its underlying hydrological factors, by a set of 84 re-
levees (vegetation surveys) from floodplains along riv-
er Rhine. Most relevees were from the Waal distribu-
tary, the major branch of the Rhine River delta system
in The Netherlands. Hydrologically, this part of the
river belongs to the upper, nontidal part of the Rhine
River system, characterized by large water-level fluc-
tuations of several meters. Main flooding events occur
in winter and early spring. Surveys were made between
Rhine kilometers 863 to 950 (measured from Konstanz,
Ecology, Vol. 85, No. 1
Germany) at different positions (sites) along the flood-
ing gradient. The relevees date from the period 1972
1991.
Vegetation relevees were made according to the tech-
nique of the Zurich-Montpellier School (see Shimwell
1971) using species cover estimates with plot sizes
between 4 m2 and 25 m2. The scores for species abun-
dance were transformed according to Van der Maarel
(1979) to an ordinal scale of 1 through 9, reducing the
effect of dominant species. Ordinal species abundances
were processed following standardized ordination tech-
niques. A detrended correspondence analysis (DCA;
Hill and Gauch 1980), using the CANOCO program
package (Ter Braak 1987), generated scores for sites
and species at a first and a second ordination axis. For
classification of the sites, the TWINSPAN package
(Hill 1979) was used.
The elevations of the sites at the places of the re-
levees were determined by leveling or estimated from
detailed topographical maps (scale 1:25 000). Mea-
sured NAP-levels (1 Normal Amsterdam Level) were
transformed to Lobith levels by correcting for the de-
cline of the river water level (Survey Department of
the Dutch Ministry of Public Works). At Lobith, a
gauging station records daily water levels of the Rhine
River. Data for the period 19811990 (Waterdata desk,
Dutch Ministry of Public Works, Lelystad, The Neth-
erlands), together with the Lobith-level elevation data,
were used to compute the mean flooding incidence per
year for each of the sites.
The geomorphology of the Rhine River system is
complex and has lead to the formation of riverine
dunes, depressions in old branches, dikes, embank-
ments, and the like. Soil types vary concomitantly,
from coarse river sand on dunes to heavy clay in
blocked depressions. Due to the dynamic nature of the
river, the hydrology of the sites may have altered in
history so that combinations of sand and clay may be
found. For soil dehydration rate, two factors are rele-
vant. First, the soil texture, with clay drying out slowly
and sand drying out fast. Second, the drainage con-
ditions of the site, i.e., whether free runoff of flood
water is prevented as is the case behind small artificial
embankments or in depressions. Based on these con-
siderations and the variation encountered in the field,
each of the sites was assigned to one out of five ordinal
classes for dehydration rate: very quick (sand, free run-
off), quick (sand on clay, free runoff), intermediate
(clay, free runoff), slow (clay, blocked), and very slow
(mud, blocked).
Ethylene-induced shoot elongation
Species were selected for ethylene-induced shoot
elongation when they were generally common in the
Waal River floodplains and when they occurred at re-
stricted elevations within the flooding gradient. These
criteria led to a selection of 22 grass and herbaceous
dicot species. Seeds of all screened species were col-
January 2004 PHYTOHORMONAL ECOLOGY
lected in Waal River floodplains and germinated on
moist filter paper in Petri dishes for 10 d in a germi-
nation cabinet (16 h photoperiod, day/night tempera-
ture of 258/108C, photosynthetic photon flux density,
PPFD, of 30 mmolm22s21 from fluorescent lamps).
Seedlings were planted in plastic pots (60 mL) filled
with a mixture (1:1 by volume) of sand and potting
compost and grown for 4 wk in a growth chamber (16
h photoperiod, day/night temperature of 218/158C,
PPFD of 95 mmolm22s21 from fluorescent lamps).
For the first experiment, similar sized 4-wk-old
plants of 11 species were selected and placed in 500-
mL pots with a mixture (1:1 by volume) of sand and
potting compost. Replicate plants were randomly dis-
tributed over two treatments (ethylene exposure and
control, n 5 89 per treatmentspecies combination).
In a second experiment, plants of 22 species were se-
lected and were allowed to grow for an additional pe-
riod of 3 wk after repotting. Thereafter, three to four
replicate plants per species were randomly assigned to
one of the two treatments.
Plants were treated for 5 d with air or air containing
10 mL/L ethylene in a gas-tight compartment (2.3 m3,
16 h photoperiod, day/night temperature of 258/208C,
PPFD of 95 mmolm22s21 from fluorescent lamps). At
this concentration ethylene-induced shoot elongation is
saturated in all plant species so far studied (Ridge 1987,
Voesenek and Blom 1999). At the beginning and the
end of the experimental period, for each plant, the lon-
gest leaf, youngest leaf, youngest petiole, and the youn-
gest internode were marked and measured. For mono-
cots, the youngest internode, longest and youngest leaf,
and leaf sheath were measured. The result of the most
responsive organ to the ethylene treatment was used in
subsequent analyses.
PLANT COMMUNITIES AND HYDROLOGY
IN RHINE R IVER F LOODPLAINS
The results of a multivariate analysis of the species
composition of the 84 sample sites are shown in Fig.
3. The first two ordination axes could be explained from
two major environmental variables. The first variable
is flooding duration that correlated strongly with the
first axis site scores (r 5 0.93, Pearson productmo-
ment coefficient, P , 0.001). The second variable, soil
dehydration rate, was significantly negatively corre-
lated with the first axis (r 5 20.61, Spearman rank
coefficient, P , 0.001), and positively with the second
axis (r 5 0.61, P , 0.001).
The ordination diagram of Fig. 3 suggests that two
environmental gradients can be distinguished that both
run from the seldom-inundated sites on dike slopes
towards the long flooded sites. These gradients were
further apparent from a TWINSPAN classification of
the field survey data resulting in eight different veg-
etation classes depicted in Fig. 3. At the left side of
the ordination diagram, both gradients start with classes
C1 and C2, for which flooding was nearly absent (on
23
FIG. 3. Two-axis ordination diagram of the species com-
position of 84 sample sites in Rhine River floodplains in The
Netherlands, based on a detrended correspondence analysis
(DCA). A TWINSPAN classification of the site data produced
eight classes (C18), indicated with different symbols. The
Special Feature
species screened for ethylene-induced shoot elongation (see
Fig. 4) are superimposed according to their position in the
ordination. Abbreviated species names: Acm, Achillea mil-
lefolium; Acp, A. ptarmica; Ags, Agrostis stolonifera; Ans,
Anthriscus sylvestris; Are, Arrhenatherum elatius; Bif, Bidens
frondosa; Chr, Chenopodium rubrum; Elr, Elymus repens; Fea,
Festuca arundinacea; Inb, Inula britannica; Lop, Lolium per-
enne; Pll, Plantago lanceolata; Pol, Polygonum lapathifolium;
Por, Potentilla reptans; Raa, Ranunculus acris; Ras, R. sce-
leratus; Rua, Rumex acetosa; Ruc, R. crispus; Rum, R. mar-
itimus; Rup, R. palustris; Trr, Trifolium repens; Vec, Veronica
catenata.
average less than 2 d/yr). These vegetations include
seldom-flooded sites on sand or sandy clay, dominated
by perennial grasses like Festuca rubra, Poa pratensis,
and Arrhenaterum elatius, and the dicot Rumex ace-
tosa.
From these high-elevated sites one gradient directs
towards open and well-drained sites (the upper right
corner in the ordination diagram). Vegetation class C4
is an intermediate, consisting of sites on very quickly
drying sandy substrate exposed to floods of short du-
ration (mean 35 d/yr), and characterized by Potentilla
reptans and Achillea ptarmica as well as several grass
species. With means of about 120 d of flooding per
year, flood duration is substantial for the extreme sites
(classes C5 and C6) along this gradient. These vege-
tations are situated at riverbanks with very quickly dry-
ing sandy soils dominated by short-lived plant species
such as Chenopodium rubrum, Atriplex prostata, and
Polygonum lapathifolium. Many of these species avoid
these flooding events through life cycle timing.
The other gradient extends towards more blocked
(free runoff is restricted) and slowly drying depressions
 24 L. A. C. J. VOESENEK ET AL. Ecology, Vol. 85, No. 1

on mudflats (the lower right corner of Fig. 3; classes

C7 and C8), with even longer flood durations of around
170 d/yr. These sites were dominated by annuals, bi-
ennials, and ephemeral perennials, but the species com-
position was very different from the classes C5 and
C6. Typical species include Rumex palustris, Limosella
aquatica, and Rorippa amphibia. For this gradient, the
moist clay sites of class C3 are intermediate, charac-
terized by short flooding (mean 33 d/yr) but dehy-
drating more slowly than the sites of class C4. A typical
species for class C3 is the perennial grass Agrostis
stolonifera.
ETHYLENE-INDUCED SHOOT ELONGATION IN
RELATION TO PLANT SPECIES DISTRIBUTION
Species from the floodplain communities described
above differ markedly in flooding induced shoot elon-
gation. For example, Rumex acetosa from high ele-
vations did not significantly increase in shoot height
under submerged as compared to drained conditions,
whereas low-elevated Rumex palustris increased three-
fold after 4 d of submergence (Voesenek and Blom
1989a). A qualitatively similar difference exists be-
tween two Ranunculus species. R. acris from high-
Special Feature

elevated sites hardly increased its petioles upon sub-

mergence (1.2-fold as compared to drained controls),
while flood-tolerant R. sceleratus increased 3.5-fold
(He et al. 1999). Ridge (1987) showed that for 27 her-
baceous species with marked elongation underwater,
the hormone ethylene was involved in the response in
21 cases. For the remaining six species, ethylene sen-
sitivity had not been studied at that time. This infor-
mation is consistent with our current knowledge on the
regulation of elongation given above, and highlights
ethylene as the key plant hormone needed to initiate
and continue fast shoot elongation under water. It calls
for a systematic investigation of ethylene-induced
shoot elongation in relation to species distribution in
the field.
The results of a comparative screening showed that
species differ markedly in response to ethylene, both
quantitatively and qualitatively (Fig. 4). From the 11
species tested in the first experiment, the four species
from rarely inundated sites, at the left side of the first
ordination axis (Class C1 and C2 in Fig. 3), produced
on average shorter shoots upon exposure to ethylene
and for two of these species (Rumex acetosa, Trifolium
repens) this response was significant (Fig. 4A). The
three species from long-flooded and slowly dehydrating
sites (Rumex palustris, Rumex maritimus, and Ranun-
culus sceleratus; class C8) were characterized by a sig-
nificant ethylene-induced increase in stem elongation.
The intermediate four species produced either shorter
or longer shoots in response to ethylene, but none of
these responses were statistically significant. Notably,
the two species that tended to form shorter shoots (Po-
tentilla reptans and Chenopodium rubrum) were from
sites with relatively good dehydration (classes C4 and
FIG. 4. Ethylene-induced shoot elongation in herbaceous
species from floodplains of the Rhine River. (A) Mean shoot
elongation rates (61 SE, n 5 89) of 11 species relative to
the mean elongation under control conditions (standardized
at 1). Species abbreviations are as in Fig. 3. Asterisks indicate
significant differences (P , 0.05) between ethylene and con-
trol treatments. (B) Shoot elongation rates of individual plants
of the 22 species listed in Fig. 3, under ethylene exposure
(open symbols) and control conditions (closed symbols), rel-
ative to the mean elongation rates under control conditions
(standardized at 1). The nonlinear regression line (dashed) is
statistically significant at P , 0.001 (F2,63 5 9.51). In both
panels, species are ordered on the abscissa according to their
position on the first DCA axis of Fig. 3.
C5 in Fig. 3), while the two species that marginally
increased in length (Agrostis stolonifera and Rumex
crispus) were typically from more slowly drying sites
(class C3).
The results of the second experiment confirmed the
results of the first experiment (Fig. 4B). Over all 22
species, there was a significant positive correlation be-
tween ethylene-induced elongation values and the
scores on the first ordination axis (r 5 0.481, P ,
0.001). Nonlinear regression analysis showed a ten-
dency for plants from rarely inundated sites at the left
side of the diagram to produce shorter shoots upon
ethylene exposure, while the shoots of plants from
long-flooded sites may double or more in length under
ethylene as compared to controls. The mid group of
January 2004 PHYTOHORMONAL ECOLOGY
species in Fig. 4B showed a variable response, which
may be explained by the negative correlation between
ethylene-induced elongation and the position of the
species along the second ordination axis (r 5 20.624,
P , 0.001). As in the first experiment, species of riv-
erbanks with relatively good dehydration (classes C4
and C5) tended to produce shorter shoots upon ethylene
exposure while the reverse was true for species of the
more slowly drying sites of classes C3 and C8. Hence,
ethylene-induced elongation was negatively related to
soil dehydration rate scores of the species.
We suggest that the match between elongation ca-
pacity and the distribution of the species along the
flooding gradient is an expression of the costs and ben-
efits associated with fast shoot elongation underwater.
The results are consistent with the hypothesis that elon-
gation is only beneficial in areas characterized by long-
term, but shallow, flooding. The blocked, slowly drain-
ing sites, where shallow, temporary pools are formed
at which the plants have a fair chance of reaching the
water surface, contain the species with the most pro-
nounced ethylene-induced shoot elongation capacity.
These sites retain floodwater much longer than the
quickly drying exposed sites. For plant species in these
latter sites the increase in shoot elongation was small
or absent. In species of dry grassland sites, the response
to ethylene exposure was even the opposite (i.e., shoot
length reduction), suggesting that there are costs in-
volved with a rapid elongation upon flooding.
The results of the screening also suggest that the
critical difference in elongation response between the
contrasting species couple Rumex palustris and R. ace-
tosa, as outlined earlier in this paper, may well apply
to a large group of herbaceous species, both monocots
and dicots. Sensitivity to ethylene varies widely among
species and may largely explain the large differences
in shoot elongation capacity under submergence. In-
terestingly, no phylogenetic patterns are apparent.
Large differences in response occur for species of the
same genus, while species of different families may
respond similarly, suggesting that the ability to elon-
gate underwater has easily changed in the course of
evolution. In addition, the two Rumex species occupy
contrasting positions along one of the major environ-
mental gradients identified in the vegetation surveys,
i.e., from rarely flooded river dike grasslands ( R. ace-
tosa) to long-flooded, slowly drying mudflats (R. pal-
ustris), and may therefore serve as model species for
the plant communities in which they occur.
PERSPECTIVES AND CONCLUSION
Our vegetation analysis suggests that flood duration
and soil dehydration are major environmental variables
shaping the species composition of the plant commu-
nities along flooding gradients of the Rhine River. Rap-
id shoot elongation, regulated at least by four plant
hormones, is an effective remedy to counteract the
problems of flooding, but only when flooding is pro-
25
longed and shoots can reach the water surface. Species
growing under such conditions in the field have a high
shoot elongation capacity, but not the species from
more elevated sites that are rarely flooded or species
from riverine dunes where the water resides rapidly,
probably because of the major costs involved with the
elongation response. The signal transduction pathway
analysis of two Rumex species suggests that hormone
sensitivity to ethylene may be a key factor explaining
the range in elongation responses. In interaction with
the environmental gradients, this single trait and its
associated costs and benefits may thus act as an im-
portant structuring force for plant species distribution
and community composition in the riverine floodplain.
Knowledge of the crucial components that regulate
submergence-induced shoot elongation allows direct
experimentation into the occurrence of the trait among
genotypes, populations, or species. Next, with knowl-
edge of the regulation mechanism, original tests of the
hypotheses generated by these comparisons may be de-
veloped. Once the genes have been discovered that are
responsible for the elongation, transgenic plants with
altered expressions may be constructed allowing for a
direct experimental test of the costs and benefits. The
Special Feature
nature of these costs can be further typified with field
transplant experiments and thus it may be tested to what
extent they are responsible for structuring the flood-
plain species composition. We suggest that the ap-
proach outlined in this paper may be more widely ap-
plicable to research aimed at a functional understanding
of species distributions along environmental gradients.
ACKNOWLEDGMENTS
The work was supported by a PIONIER grant (800.84.470)
of the Dutch Science Foundation to L. A. C. J. Voesenek and
an EU grant (Contract SCI-CT91-0739-TSTS) to J. G. H. M.
Rijnders. We thank Joris Benschop, Jordi Bou, Marjolein
Cox, Henri Groeneveld, Frank Millenaar, and Robert Vree-
burg for access to their primary data and Rob Welschen for
technical assistance. Josef Stuefer gave valuable statistical
assistance.
LITERATURE CITED
Abeles, F. B. 1992. Ethylene in plant biology. Academic
Press, London, UK.
Banga, M., E. J. Slaa, C. W. P. M. Blom, and L. A. C. J.
Voesenek. 1996. Ethylene biosynthesis and accumulation
under drained and submerged conditions: a comparative
study of two Rumex species. Plant Physiology 112:229
237.
Bleecker, A. B., and H. Kende. 2000. Ethylene: a gaseous
signal molecule in plants. Annual Review of Cell and De-
velopmental Biology 16:118.
Blom, C. W. P. M., G. M. Bogemann, P. Laan, A. J. M. van
der Sman, H. M. van de Steeg, and L. A. C. J. Voesenek.
1990. Adaptations to flooding in plants from river areas.
Aquatic Botany 38:2947.
Blom, C. W. P. M., H. M. van de Steeg, and L. A. C. J.
Voesenek. 1996. Adaptive mechanisms of plants occurring
in wetland gradients. Pages 91112 in G. Mulamoottil, B.
G. Warner, and E. A. McBean, editors. Wetlands: environ-
mental gradients, boundaries and buffers. CRC Press, Boca
Raton, Florida, USA.
 26 L. A. C. J. VOESENEK ET AL.
Blom, C. W. P. M., and L. A. C. J. Voesenek. 1996. Flooding:
the survival strategies of plants. Trends in Ecology and
Evolution 11:290295.
Buchanan, B. B., W. Gruissem, and R. L. Jones. 2000. Bio-
chemistry and molecular biology of plants. American So-
ciety of Plant Physiologists, Rockville, Maryland, USA.
Colmer, T. D. 2003. Long-distance transport of gases in
plants: a perspective on internal aeration and radial oxygen
loss from roots. Plant, Cell and Environment 26:1736.
Colmer, T. D., M. R. Gibberd, A. Wiengweera, and T. K.
Tinh. 1998. The barrier to radial oxygen loss from roots
of rice (Oryza sativa L.) is induced by growth in stagnant
solution. Journal of Experimental Botany 49:14311436.
Cox, M. C. H., F. F. Millenaar, Y. E. M. de Jong van Berkel,
A. J. M. Peeters, and L. A. C. J. Voesenek. 2003. Plant
movement. Submergence-induced petiole elongation in Ru-
mex palustris depends on hyponastic growth. Plant Phys-
iology 132:282291.
Crawford, R. M. M., J. C. Walton, and B. Wollenweber-Ratz-
er. 1994. Similarities between post-ischaemic injury to an-
imal tissues and post-anoxia injury in plants. Proceedings
of the Royal Society of Edinburgh 102B:325332.
Genoud, T., and J. P. Metraux. 1999. Crosstalk in plant cell
signaling: structure and function of the genetic network.
Trends in Plant Science 4:503507.
Genoud, T., M. B. T. Santa Cruz, and J. P. Metraux. 2001.
Numeric simulation of plant signaling networks. Plant
Physiology 126:14301437.
He, J. B., G. M. Bogemann, H. M. v.d. Steeg, J. G. H. M.
Rijnders, L. A. C. J. Voesenek, and C. W. P. M. Blom.
Special Feature
1999. Survival tactics of Ranunculus species in river flood-
plains. Oecologia 118:18.
Hill, M. O. 1979. TWINSPAN, a Fortran program for ar-
ranging multivariate data in an ordered two-way table by
classification of the individuals and attributes. Cornell Uni-
versity, Ithaca, New York, USA.
Hill, M. O., and H. G. Gauch. 1980. Detrended correspon-
dence analysis: an improved ordination technique. Vege-
tatio 42:4758.
Hoffmann-Benning, S., and H. Kende. 1992. On the role of
abscisic acid and gibberellin in the regulation of growth in
rice. Plant Physiology 99:11561161.
Horton, R. F. 1991. The effect of ethylene and other regu-
lators on coleoptile growth of rice under anoxia. Plant Sci-
ence 79:5762.
Hua, J., and E. M. Meyerowitz. 1998. Ethylene responses
are negatively regulated by a receptor gene family in Ar-
abidopsis thaliana. Cell 94:261271.
Jackson, M. B. 1985. Ethylene and the responses of plants
to excess water in their environment: A review. Pages 241
265 in J. A. Roberts and G. A. Tucker, editors. Ethylene
and plant development. Butterworths, London, UK.
Jackson, M. B., and W. Armstrong. 1999. Formation of aer-
enchyma and the process of plant ventilation in relation to
soil flooding and submergence. Plant Biology 1:274287.
Kende, H., E. v.d. Knaap, and H.-T. Cho. 1998. Deepwater
rice: a model plant to study stem elongation. Plant Phys-
iology 118:11051110.
Lee, T. M., Y. J. Shieh, and C. H. Chou. 1996. Abscisic acid
inhibits shoot elongation of Scirpus mucronatus. Physiol-
ogia Plantarum 97:14.
Ljung, K., R. P. Bhalerao, and G. Sandberg. 2001. Sites and
homeostatic control of auxin biosynthesis in Arabidopsis
during vegetative growth. Plant Journal 28:465474.
Maltby, E. 1991. Wetlands. Pages 321 in M. B. Jackson, D.
D. Davies, and H. Lambers, editors. Plant life under oxygen
deprivation. SPB Academic Publishing, The Hague, The
Netherlands.
Menges, E. S., and D. M. Waller. 1983. Plant strategies in
relation to elevation and light in floodplain herbs. American
Naturalist 122:454473.
Ecology, Vol. 85, No. 1
Muday, G. K., and A. deLong. 2001. Polar auxin transport:
controlling where and how much. Trends in Plant Science
6:535542.
Musgrave, A., M. B. Jackson, and E. Ling. 1972. Callitriche
stem elongation is controlled by ethylene and gibberellin.
Nature 238:9396.
Nabben, R. H. M., C. W. P. M. Blom, and L. A. C. J. Voesenek.
1999. Resistance to complete submergence in Rumex spe-
cies with different life histories: the influence of plant size
and light. New Phytologist 144:313321.
Peeters, A. J. M., M. C. H. Cox, J. J. Benschop, R. A. M.
Vreeburg, J. Bou, and L. A. C. J. Voesenek. 2002. Sub-
mergence research using Rumex palustris as a model; look-
ing back and going forward. Journal of Experimental Bot-
any 53:391398.
Perata, P., and A. Alpi. 1993. Plant responses to anaerobiosis.
Plant Science 93:117.
Ridge, I. 1987. Ethylene and growth control in amphibious
plants. Pages 5376 in R. M. M. Crawford, editor. Plant
life in aquatic and amphibious habitats. Blackwell Scien-
tific Publications, Oxford, UK.
Rijnders, J. G. H. M., W. Armstrong, M. J. Darwent, C. W.
P. M. Blom, and L. A. C. J. Voesenek. 2000. The role of
oxygen in submergence-induced petiole elongation in Ru-
mex palustris: in situ measurements of oxygen in petioles
of intact plants using micro-electrodes. New Phytologist
147:497504.
Rijnders, J. G. H. M., G. W. M. Barendse, C. W. P. M. Blom,
and L. A. C. J. Voesenek. 1996. The contrasting role of
auxin in submergence-induced petiole elongation in two
species from frequently flooded wetlands. Physiologia
Plantarum 96:467473.
Rijnders, J. G. H. M., Y.-Y. Yang, Y. Kamiya, N. Takahashi,
G. W. M. Barendse, C. W. P. M. Blom, and L. A. C. J.
Voesenek. 1997. Ethylene enhances gibberellin levels and
petiole sensitivity in flooding-tolerant Rumex palustris but
not in flooding-intolerant R. acetosa. Planta 203:2025.
Ritchie, S., and S. Gilroy. 1998. Gibberellins: regulating
genes and germination. New Phytologist 140:363383.
Saglio, P., V. Germain, and B. Ricard. 1999. The response
of plants to oxygen deprivation: role of enzyme induction
in the improvement of tolerance to anoxia. Pages 373393
in H. R. Lerner, editor. Plant responses to environmental
stresses: from phytohormones to genome reorganization.
Marcel Dekker, New York, New York, USA.
Schmitt, J., S. A. Dudley, and M. Pigliucci. 1999. Manipu-
lative approaches to testing adaptive plasticity: phyto-
chrome-mediated shade-avoidance responses in plants.
American Naturalist 154:S43S54.
Setter, T. L., M. Ellis, E. V. Laureles, E. S. Ella, D. Senadhira,
S. B. Mishra, S. Sarkarung, and S. Datta. 1997. Physiology
and genetics of submergence tolerance in rice. Annals of
Botany 79:6777.
Setter, T. L., and E. V. Laureles. 1996. The beneficial effect
of reduced elongation growth on submergence tolerance of
rice. Journal of Experimental Botany 47:15511559.
Shimwell, D. W. 1971. The description and classification of
vegetation. Sidgwick and Jackson, London, UK.
Smirnoff, N. 1995. Antioxidant systems and plant response
to the environment. Pages 217243 in N. Smirnoff, editor.
Environment and plant metabolism. Flexibility and accli-
mation. Bios Scientific Publishing, Oxford, UK.
Smith, H. 2000. Phytochromes and light signal perception
by plantsan emerging synthesis. Nature 407:585591.
Sykora, K. V., E. Schepers, and F. van der Zee. 1988. In-
undation and the distribution of plant communities on dutch
river dikes. Acta Botanica Neerlandica 37:279290.
Ter Braak, C. J. F. 1987. CANOCO, a FORTRAN program
for canonical community ordination by partial detrended
canonical correspondence analysis, principal components
January 2004 PHYTOHORMONAL ECOLOGY
analysis and redundancy analysis. Version 2.0. TNO, ITT,
Wageningen, The Netherlands.
Van der Maarel, E. 1979. Transformation of cover-abundance
values in phytosociology and its effects on community sim-
ilarity. Vegetatio 39:97114.
Van der Sman, A. J. M., C. W. P. M. Blom, and H. M. Van
de Steeg. 1992. Phenology and seed production in Che-
nopodium rubrum, Rumex maritimus, and Rumex palustris
as related to photoperiod in river forelands. Canadian Jour-
nal of Botany 70:392400.
Van der Sman, A. J. M., L. A. C. J. Voesenek, C. W. P. M.
Blom, F. Harren, and J. Reuss. 1991. The role of ethylene
in shoot elongation with respect to survival and seed output
of flooded Rumex maritimus L. plants. Functional Ecology
5:304313.
Vartapetian, B. B., and M. B. Jackson. 1997. Plant adaptation
to anearobic stress. Annals of Botany 79:320.
Vervuren, P. J. A., S. M. J. H. Beurskens, and C. W. P. M.
Blom. 1999. Light acclimation, CO2 response and long-
term capacity of underwater photosynthesis in three ter-
restrial plant species. Plant, Cell and Environment 22:959
968.
Vervuren, P. J. A., C. W. P. M. Blom, and H. de Kroon. 2003.
Extreme flooding events on the Rhine and the survival and
distribution of riparian plant species. Journal of Ecology
91:135146.
Visser, E. J., C. W. P. M. Blom, and L. A. C. J. Voesenek.
1996. Flooding-induced adventitious rooting in Rumex:
morphology and development in an ecological perspective.
Acta Botanica Neerlandica 45:1728.
Visser, E. J. W., R. H. M. Nabben, C. W. P. M. Blom, and L.
A. C. J. Voesenek. 1997. Elongation by primary lateral
roots and adventitious roots during conditions of hypoxia
and high ethylene concentration. Plant, Cell and Environ-
ment 20:647653.
Voesenek, L. A. C. J. 1990. Adaptations of Rumex in flooding
gradients. Dissertation. University of Nijmegen, Nijmegen,
The Netherlands.
Voesenek, L. A. C. J., M. Banga, J. G. H. M. Rijnders, E. J.
W. Visser, and C. W. P. M. Blom. 1996. Hormone sensi-
tivity and plant adaptations to flooding. Folia Geobotanica
Phytotaxonomica 31:4756.
Voesenek, L. A. C. J., M. Banga, R. H. Thier, C. M. Mudde,
F. J. M. Harren, G. W. M. Barendse, and C. W. P. M. Blom.
27
1993a. Submergence-induced ethylene synthesis entrap-
ment, and growth in two plant species with contrasting
flooding resistances. Plant Physiology 103:783791.
Voesenek, L. A. C. J., and C. W. P. M. Blom. 1989a. Growth
responses of Rumex species in relation to submergence and
ethylene. Plant, Cell and Environment 12:433439.
Voesenek, L. A. C. J., and C. W. P. M. Blom. 1989b. Ethylene
and flooding responses of Rumex species. Pages 245253
in H. Clijsters et al., editors. Biochemical and physiological
aspects of ethylene production in lower and higher plants.
Kluwer Academic Publishers, Dordrecht, The Netherlands.
Voesenek, L. A. C. J., and C. W. P. M. Blom. 1999. Stimulated
shoot elongation: a mechanism of semiaquatic plants to
avoid submergence stress. Pages 431448 in H. R. Lerner,
editor. Plant responses to environmental stresses: from phy-
tohormones to genome reorganization. Marcel Dekker, New
York, New York, USA.
Voesenek, L. A. C. J., M. C. C. de Graaf, and C. W. P. M.
Blom. 1992a. Germination and emergence of Rumex in
river flood-plains. II. The role of perianth, temperature,
light and hypoxia. Acta Botanica Neerlandica 41:331343.
Voesenek, L. A. C. J., F. J. M. Harren, C. W. P. M. Blom,
and A. J. M. Van der Sman. 1992b. An amalgamation
between hormone physiology and plant ecology: a review
on flooding resistance and ethylene. Journal of Plant
Growth Regulation 11:171188.
Voesenek, L. A. C. J., P. J. M. Perik, C. W. P. M. Blom, and
M. M. A. Sassen. 1990. Petiole elongation in Rumex during
submergence and ethylene exposure: the relative contri-
butions of cell division and cell expansion. Journal of Plant
Special Feature
Growth Regulation 9:1317.
Voesenek, L. A. C. J., F. J. M. M. Van Oorschot, A. J. M.
Smits, and C. W. P. M. Blom. 1993b. The role of flooding
resistance in the establishment of Rumex seedlings in river
flood plains. Functional Ecology 7:105114.
Voesenek, L. A. C. J., W. H. Vriezen, M. J. E. Smekens, F.
H. M. Huitink, G. M. Bogemann, and C. W. P. M. Blom.
1997. Ethylene sensitivity and response sensor expression
in petioles of Rumex species at low O2 and high CO2 con-
centrations. Plant Physiology 114:15011509.
Vriezen, W. H., C. P. E. Van Rijn, L. A. C. J. Voesenek, and
C. Mariani. 1997. A homologue of the Arabidopsis thaliana
ERS gene is actively regulated in Rumex palustris upon
flooding. Plant Journal 11:12651271.
 Ecology, 85(1), 2004, pp. 2837
q 2004 by the Ecological Society of America

STRETCHING THE LIMITS OF PLASTICITY: CAN A PLANT DEFEND

AGAINST BOTH COMPETITORS AND HERBIVORES?
DON CIPOLLINI1
Wright State University, Department of Biological Sciences, 3640 Colonel Glenn Highway, Dayton, Ohio 45435 USA

Abstract. Plants exhibit numerous adaptive responses to competitors and herbivores

mediated by hormones, including induction of the shade avoidance syndrome in response
to neighbor shade, and induction of chemical defenses in response to herbivore damage.
Although largely studied in isolation, interactions between plastic responses to competitors
and herbivores can occur and may be mediated by external and internal resource availability,
direct interactions between hormones controlling plastic responses, or other pleiotropic
effects. Such interactions may have positive or negative consequences for plant fitness and
could serve as an important ecological and evolutionary constraint on the expression of
these responses. In this article, I illustrate the potential for plastic responses in plants to
physiologically interact, with particular reference to Arabidopsis thaliana. I also suggest
avenues for future study utilizing characterized plant material, tools to manipulate the
expression of plastic responses, and ways to monitor physiological and ecological conse-
quences that are available using this and other plant species.
Key words: adaptive phenotypic plasticity; Arabidopsis thaliana; auxin; competition; gibberellic
acid; herbivores; induced defenses; jasmonic acid; peroxidase; phytochrome; salicylic acid; shade
avoidance responses.
Special Feature

ADAPTIVE PHENOTYPIC PLASTICITY IN RESPONSE
TO H ERBIVORES AND COMPETITORS
A central assumption in evolutionary ecology is that
organisms make trade-offs in allocating limited re-
sources to different functions. Life-history trade-offs
are thought to have selected for the plastic expression
of traits that allow an organism to maximize fitness in
changing environments (Sultan 2000). In plants, the
well-known fitness effects of herbivore damage have
apparently selected for a vast taxonomic diversity of
proteins, chemical compounds, and physical structures
with defensive function (e.g., Rhoades 1979). Most
models to explain the evolution, distribution, and di-
versity of antiherbivore defenses assume that allocation
of resources to resistance represents a trade-off with
growth (e.g., Rhoades 1979, Coley et al. 1985, Herms
and Mattson 1992). In turn, current theories suggest
that inducible defenses against herbivores in plants are
cost-saving mechanisms because they are induced only
after herbivore attack when the benefit of the defense
can counterbalance the costs of defense production
(Karban and Baldwin 1997). As such, inducible de-
fenses are thought to be a form of adaptive phenotypic
plasticity, allowing a plant to closely match an appro-
priate phenotype with a selective environment (Agra-
wal 1998, Baldwin 1998, 1999, Cipollini 1998, Agra-
wal et al. 1999).
Manuscript received 7 October 2002; revised 5 February 2003;
accepted 22 March 2003. Corresponding Editor: A. A. Agrawal.
For reprints of this Special Feature, see footnote 1, p. 3.
1 E-mail: don.cipollini@wright.edu
28
A great deal of work has been done to elucidate
mechanisms of inducible defense responses to herbi-
vores and other attackers. A multitude of evidence in-
dicates that the lipid-derived hormone, jasmonic acid
(JA), and the phenolic hormone, salicylic acid (SA),
are key signals mediating the induction of genes in-
volved in defense responses to herbivores, pathogens,
and other stresses (Creelman and Mullet 1997, Mauch-
Mani and Metraux 1998, Glazebrook 2001). The in-
duced response itself can vary greatly by plant and
species of attacker, but it often includes the production
of numerous proteins, compounds, and physical struc-
tures that, through mechanisms such as alteration in
tissue quality and recruitment of predators, tend to limit
damage by herbivores and other attackers (reviewed in
Karban and Baldwin 1997). For example, analysis of
the transcriptome of A. thaliana L. after induction of
systemic acquired resistance (mediated largely by sal-
icylic acid) revealed that 413 of 7000 genes analyzed
exhibited changes in expression of 2.5-fold or more
(Maleck et al. 2000). Such changes typically result in
enhanced resistance of A. thaliana to a diverse array
of pathogens (Delaney 1997).
Plant competitors limit the availability of light, wa-
ter, and nutrients to individuals in a population, factors
well known to independently and interactively affect
fitness (Grace and Tilman 1990). In addition, compet-
itors can alter soil composition, local temperatures, ex-
posure to wind, and the interaction of individuals with
their biotic environment (Reader 1992, Retuerto et al.
1996, Ball et al. 2002). Like herbivore damage, the
fitness effects of competitors have apparently selected
for a number of putatively adaptive morphological and
January 2004 PHYTOHORMONAL ECOLOGY
physiological responses. A well-studied response of
plants to neighbors or canopy shade is the phyto-
chrome-mediated shade avoidance response, charac-
terized in many plants by increases in hypocotyl and
internode elongation and alterations in leaf production
and expansion that enable individuals to escape shading
by neighbors and maximize photosynthesis (e.g., Bal-
lare et al. 1990, Ballare and Scopel 1992, Schmitt 1997,
Dorn et al. 2000, Weinig 2000b). This response is trig-
gered by light with a low ratio of red to far-red wave-
lengths (R:FR) typical of canopy shade or light reflect-
ed from neighbors, and involves several light-respon-
sive phytochrome genes, and hormones such as gib-
berellic acid and auxin (Morgan and Smith 1978,
Sharrock and Quail 1989, Smith and Whitelam 1997,
Neff et al. 2000). In A. thaliana, shade-avoidance re-
sponses include increased internode elongation, re-
duced branching and leaf production, and early flow-
ering, and at least some of these responses appear to
be adaptive under low R:FR light (Dorn et al. 2000).
For example, A. thaliana genotypes that exhibit a taller
and less branched architecture in the presence of high
competition for light are selectively favored over ge-
notypes that are shorter and more branched (Duffy et
al. 1999).
Less well characterized, but undoubtedly important,
responses to competitors include adaptive alterations
in root development and physiology that enable plants
to exploit patchily distributed resources and maximize
water and nutrient uptake. For example, at low phos-
phorus levels, A. thaliana exhibits alterations in root
hair production and development (Ma et al. 2001) that
apparently allow wild-type plants to outcompete root-
hairless mutants (Bates and Lynch 2001). Individual
soybean plants will exhibit greater root proliferation in
soils shared with neighbors than when grown alone,
even when soil volume and nutrients per individual are
held constant (Gersani et al. 2001). A recent microarray
experiment demonstrated that 115 of 1280 genes de-
rived from tomato roots were inducible by nitrate
(Wang et al. 2001a). The differential induction of these
genes among individuals or species in locally nitrate-
enriched patches (e.g., Robinson et al. 1999) could be
a mechanism of root competition. Signaling mecha-
nisms that regulate root responses to neighbors are not
well understood (Callaway 2002).
Adaptive phenotypic plasticity in plants to compet-
itors and herbivores has been examined in ever-in-
creasing mechanistic and evolutionary detail. Quanti-
tative genetic variation is known to exist for both shade
avoidance responses and for inducible defense respons-
es to herbivores (Schmitt 1997, Agrawal et al. 2002),
and populations exhibit ecotypic variation in the ex-
pression of these responses (Zangerl and Berenbaum
1990, Dudley and Schmitt 1995, Dixon et al. 2001).
Quantitative trait loci controlling plasticity in the in-
duction of specific defenses have been identified (Klie-
benstein et al. 2002). At the mechanistic level, signal-
29
ing networks that control plastic responses are increas-
ingly being elucidated and several mutants affected in
responses to herbivores or neighbors have been char-
acterized (Ballare and Scopel 1992, Delaney 1997, Ber-
ger 2002). Several genes that mediate each of these
responses have been cloned and manipulated in trans-
genic plants (e.g., Delaney et al. 1994, Palecanda and
Sharrock 2001). Exogenous application of key hor-
mones that mediate plasticity has enabled the explo-
rations of costs and benefits of plastic responses (e.g.,
Baldwin 1998, Agrawal et al. 1999, Cipollini and
Schultz 1999, Cipollini 2002). Using these and other
phenotypic manipulation techniques, the adaptive val-
ue of shade avoidance responses and of inducible de-
fenses has been demonstrated in the field (e.g., Dudley
and Schmitt 1996, Baldwin 1998).
PLASTIC RESPONSES TO COMPETITORS
AND HERBIVORES MAY INTERACT
Adaptive plant responses to competitors and to her-
bivores have largely been studied in isolation, but
plants must respond to changes in numerous environ-
mental factors simultaneously in nature. Competitors
and herbivores, for example, are commonly present at
Special Feature
the same time in any plant population. Favorable abi-
otic conditions created by high plant density may even
increase herbivore and pathogen populations (Reader
1992), making herbivore pressure oftentimes more se-
vere as plant density increases. Regardless of the role
of plastic responses in maximizing fitness in such sit-
uations, extrinsic resource limitation can passively lim-
it the ability of a plant to respond optimally to each of
these stressors, as it could for any combination of
stressors (Callaway 2002). Not surprisingly, a recent
meta-analysis of multifactorial field experiments con-
firmed that competitors and predators (including her-
bivores) often interactively affect fitness (Gurevitch et
al. 2000), which may often be the result of passive
resource-mediated interactions. However, physiologi-
cal interactions between plastic responses themselves
may also affect the ability of a plant to respond opti-
mally to multiple stressors. For example, induction of
shade-avoidance responses in Abutilon theophrasti ear-
ly in development constrained both elongation respons-
es to low R:FR and morphological responses to wind
later in development (Weinig and Delph 2001, Henry
and Thomas 2002). On the other hand, exposure to high
temperatures increased the responsiveness of A. theo-
phrasti to low R:FR (Weinig 2000a), while induction
of shade avoidance responses in Impatiens capensis
with low R:FR light had no effect on susceptibility to
UV damage (Dixon et al. 2001).
INTERACTIONS CAN BE MEDIATED BY INTERNAL
RESOURCE ALLOCATION TRADE-OFFS
Physiological interactions between plastic responses
to competitors and herbivores might be mediated by
internal resource availability. For example, evidence
 30 DON CIPOLLINI
suggests that induction of shade avoidance responses
can affect both plant tolerance and resistance to her-
bivory through resource allocation trade-offs. By sim-
ulating plastic morphological changes to neighbors and
shade with gibberellic acid application, artificially
elongated individuals of common bean were shown to
be less tolerant of simulated herbivory than non-elon-
gated individuals in both the laboratory and in the field
(Cipollini 1997a, Cipollini and Schultz 1999). Because
external resources were not limiting in these studies,
this effect was thought to reflect an opportunity cost
of increased allocation to nonproductive stem material,
reduced leaf area, and reduced root biomass in artifi-
cially elongated individuals (or other pleiotropic ef-
fects of gibberellic acid treatment). Other studies find-
ing that plants expressing shade avoidance responses
were less tolerant of defoliation were confounded by
variation in resource levels (e.g., light levels; Lentz
and Cipollini 1998). In Brassica napus, increasing the
density of neighbors around target plants led to si-
multaneous increases in the expression of shade avoid-
ance responses and decreases in the expression of con-
stitutive and induced levels of defensive trypsin inhib-
itors in the target plants (Cipollini and Bergelson
Special Feature
2001). Reductions in trypsin inhibitor expression may
have been due to external nutrient or light limitation
caused by increasing neighbor density, or due to in-
trinsic resource allocation trade-offs between expres-
sion of adaptive responses to competitors and to her-
bivores. Fertilizing competing plants restored consti-
tutive and induced levels of trypsin inhibitors in target
plants to levels found in noncompeting plants, but had
little effect on growth. This suggests that resources
were prioritized to adaptive growth responses to com-
petitors in unfertilized plants at the expense of adaptive
responses to herbivores, and only when fertilizer was
added could competing plants both grow and defend
(sensu Siemens et al. 2002). The observation that com-
petition in greenhouse-grown plants could affect de-
fense expression may have field consequences. B. na-
pus individuals grown in plots with weed competitors
exhibited shade avoidance responses to a greater degree
and experienced greater herbivory than individuals in
plots without weed competitors, a result that may have
been associated with reduced expression of defenses
in plants expressing the shade avoidance response (Ci-
pollini and Bergelson 2002). A similar mechanism may
explain why the fitness effects of rust infection in Im-
patiens capensis stands were more severe at high den-
sity than at low density (Lively et al. 1995).
INTERACTIONS MAY RESULT FROM
HORMONAL INTERACTIONS
The production of some hormones and chemical re-
sistance traits is mechanistically linked in some plants,
and provides another way in which plant responses to
competitors and herbivores may interact. Adaptive
changes in stem elongation or bending characteristic
Ecology, Vol. 85, No. 1
of shade avoidance responses, for example, require lo-
cal increases in auxin content in plant stems that stim-
ulate cell elongation (Behringer and Davies 1992, Mo-
relli and Ruberti 2000). The production of auxin is
known to interact directly with the production of cer-
tain inducible defenses in plants. In A. thaliana, syn-
thesis of the indole glucosinolates (known to be in-
ducible by herbivores and JA in brassicaceous plants)
competes with auxin (specifically indole-acetic acid,
IAA) synthesis for a shared pathway intermediate
(Mikkelson et al. 2000, Bak et al. 2001, Celenza 2001).
Consequently, IAA-overexpressing genotypes of A.
thaliana have reduced levels of indole glucosinolates,
and vice versa. This mechanistic linkage suggests that
the induction of indole glucosinolates by herbivores
(or increased constitutive levels in different genotypes)
could compromise the ability of a plant to express IAA-
mediated changes in stem elongation and/or tropistic
bending, at least temporarily. Conversely, induction of
shade avoidance or other tropistic responses in which
elevated IAA levels play a role could inhibit the sub-
sequent ability of plants to synthesize indole glucosi-
nolates in response to herbivore attack, which could
compromise resistance.
In addition to directly interacting with indole glu-
cosinolate production through substrate competition,
IAA is known to regulate the expression of other in-
ducible defenses. For example, exogenously applied
IAA is known to inhibit the wound-induced expression
of nicotine (Baldwin et al. 1990), and the wound- and
JA-induced expression of proteinase inhibitor genes in
tobacco (Thornburg and Li 1991, Ishikawa et al. 1994).
In turn, wounding results in a two- to threefold decline
in endogenous IAA levels in tobacco, the kinetics of
which correlate perfectly with proteinase inhibitor gene
induction (Thornburg and Li 1991). Rather than com-
peting for substrate in such cases, IAA appears to keep
expression of some defenses within developmental
range limits, beyond which costs to growth and fitness
of overexpressing defense responses would far ex-
ceed their benefits and could even be autotoxic (Bald-
win and Callahan 1993). IAA regulation of defense
may ensure that resources are prioritized to growth
rather than defense production in situations where plas-
tic morphological responses to competitors may have
greater fitness benefits than defense production. A re-
cent study has suggested that the jasmonate response
protein JAR1 and its homologs in A. thaliana can reg-
ulate active levels of both JA and IAA, thus providing
another mechanism through which plant responses re-
quiring each of these hormones might interact (Stas-
wick et al. 2002).
Other hormones involved in adaptive responses to
competitors can also potentially affect induction of re-
sistance to herbivores and pathogens. Gibberellic acid,
a key mediator of phytochrome-mediated stem elon-
gation (Peng and Harberd 1997), can increase the se-
verity of white mold disease in bean and cucumber
January 2004 PHYTOHORMONAL ECOLOGY
when applied exogenously as GA3 (Al-Masri et al.
2002). Conversely, methyl jasmonate treatment can de-
crease disease severity in the same plants. Although
the mechanism causing these effects was unknown,
GA3 can inhibit peroxidase activity when applied ex-
ogenously in some plants (Jupe and Scott 1992). Per-
oxidase has diverse roles in phenolic metabolism, lig-
nification, and cell wall deposition and is characterized
by having a general lack of substrate specificity and
thus capable of having many pleiotropic effects (Campa
1991). It is broadly inducible by environmental stimuli,
including herbivore and pathogen attack, and JA ap-
plication (Cipollini and Sipe 2001). A negative asso-
ciation between gibberellic acid levels and peroxidase
activity would suggest that plants expressing shade
avoidance responses (or high levels of gibberellic acid
for other reasons) could have a reduced ability to ex-
press peroxidase as an inducible defense to herbivores
or other stresses. On the other hand, a plant expressing
high peroxidase activity as a result of induction may
be unable to express an elongated phenotype in re-
sponse to shade as a consequence of increased cell wall
rigidity imparted by peroxidase activity. Peroxidase is
also involved in auxin metabolism, having IAA-oxi-
dase activity (Normanly 1997, Jansen et al. 2001). To-
bacco plants overexpressing anionic peroxidases have
reduced levels of auxin, and have shorter and thicker
stems with higher lignin content than normal plants
(Lagrimini 1991, 1999, Lagrimini et al. 1997, Jansen
et al. 2001). Interestingly, bean seedlings with high
levels of peroxidase and lignin as a result of wind-
induced mechanical stimulation were not only shorter
and thicker than control seedlings, but also were more
resistant to a mite herbivore and a fungal pathogen
(Cipollini 1997b). This further suggests that plants ex-
pressing high levels of peroxidase as a result of in-
duction may be unable to respond to competitors with
changes in stem elongation, although resistance to her-
bivores may increase. Exogenous gibberellic acid and
auxin are both known to decrease the abundance of a
RNA in tomato with sequence similarity to wound-
inducible proteinase inhibitors, and full induction of
this RNA requires less severe wounding in a gibber-
ellin-deficient mutant than in wild-type plants (Jacob-
sen et al. 1996). Negative effects of gibberellic acid
and auxin on the expression of proteinase inhibitors
may help to explain the suppression of wound-induced
trypsin inhibitors in B. napus plants exhibiting shade
avoidance responses (Cipollini and Bergelson 2001).
Negative mechanistic links between the induction of
shade avoidance responses and the induction of defense
to herbivores could serve as an important ecological
and evolutionary constraint on the expression of these
responses. Fig. 1 illustrates a hypothesized relationship
between plant growth responses to decreasing R:FR
light and herbivore resistance based on known mech-
anistic interactions between these responses.
31
FIG. 1. Hypothetical relationship between induction of
the shade avoidance response (characterized by stem exten-
sion rate) with decreasing R:FR light and herbivore resis-
tance. Nutrient and water availability are assumed to be con-
stant across the light quality gradient in this example. The
logarithmic stem extension rate is known to decrease linearly
with increasing R:FR light ratio in plants such as Chenopo-
dium album (Morgan and Smith 1978). Decreases in herbivore
resistance with increasing stem extension rate are hypothe-
sized based on the known effects of gibberellins and auxins
involved in phytochrome-mediated stem extension on induc-
Special Feature
ible defenses in plants. The herbivore resistance curve is as-
ymptotic, indicating that there are basal levels of resistance
unaffected by light quality at low R:FR light, and develop-
mental range limits on defense production at high R:FR light.
MUTUAL BENEFITS OF PLASTIC RESPONSES TO
COMPETITORS AND HERBIVORES
All interactions between plant responses to compet-
itors and herbivores need not be negative. Interestingly,
a possible positive interaction between these responses
is evident in studies of costs of resistance in plants.
Costs of resistance are hypothesized to increase under
resource-limiting conditions, including the presence of
competitors (Bazzaz et al. 1987, Herms and Mattson
1992, Herms 1999, Weis and Hochberg 2000). How-
ever, experimental attempts to demonstrate this effect
have produced examples of increases, decreases, and
no effects of competition on costs of resistance (Ber-
gelson 1994a, b, van Dam and Baldwin 1998, 2001,
Baldwin and Hamilton 2000, Cipollini 2002, Siemens
et al. 2002). Studies finding either no effect or decreas-
es in costs under competition have been difficult to
interpret thus far on the basis of resource allocation.
One intriguing hypothesis to explain such results, how-
ever, is that some traits that influence herbivore resis-
tance and certain aspects of competitive ability may be
positively linked in some plant species, counterbal-
ancing the presumed trade-off between these traits (Ci-
pollini 2002, Siemens et al. 2002). Positive linkages
between plant responses to competitors and to either
herbivores and pathogens or both would seem to be
selectively favored in some environments on the basis
of cost effectiveness.
 32 DON CIPOLLINI
One way in which these responses could be posi-
tively linked is if some signaling or metabolic pathways
essential to the induced production of defense also me-
diate (or at least benefit) some plant responses to com-
petitors or other stresses. In A. thaliana, the role of JA
and SA in induced resistance has often been demon-
strated by examining the resistance of plants that are
altered in their accumulation of these hormones, and
by otherwise manipulating their presence (e.g., Dela-
ney et al. 1994, McConn et al. 1997). In addition to
showing enhanced susceptibility to certain pathogens
and insects, however, mutant or transgenic A. thaliana
lacking full competence in inducible defense pathways
mediated by these hormones are poorer competitors
with wild-type A. thaliana than the wild-type itself (Ci-
pollini 2002). In this study, the nahG line (engineered
to degrade SA in planta, Delaney et al. 1994), the
nim1-1 line (lacking a functional NIM1 protein and
sensitivity to SA, Delaney et al. 1995), and the jar1-1
line (lacking a functional JAR1 protein and sensitivity
to endogenous JA, Staswick et al. 1992, 2002) pro-
duced more seed than their respective wild-type parent
ecotypes when plants were grown alone in pots with
adequate water and nutrient inputs. This may relate to
Special Feature
the fact that these lines tended to express lower con-
stitutive levels of defense than their wild-type parent
ecotypes. However, seed production in these mutant
lines was more greatly reduced by the presence of six
wild-type competitors grown in the same pot than in
their respective wild-type parent ecotypes. Thus, le-
sions in these inducible response pathways negatively
affected competitive ability, although the mechanism
was unknown. Aside from marker defense production,
no analysis of physiological status or growth and ar-
chitecture was performed in this study. Another jas-
monate response mutant of A. thaliana that is altered
in the production of the JAR1 protein (fin219) exhibits
a hypersensitive stem elongation response after expo-
sure to far-red light (Hsieh et al. 2000, Staswick et al.
2002). Expression of this aspect of the shade avoidance
response by the jar1-1 mutant cited above, however, is
apparently no different than wild-type plants (Staswick
et al. 2002), indicating that some other aspect of com-
petitive ability is apparently disrupted in this mutant.
Since the identification of the roles of JA and SA in
inducible defense responses, it has become clear that
the roles of these compounds in plant growth and de-
velopment can be quite broad. In addition to its sig-
naling role in defense responses to insects, wounding,
and pathogen attack, JA has roles in processes ranging
from photosynthesis to pollen fertility in plants, many
of which could affect overall competitive ability or
growth under stress (Creelman and Mullet 1997). One
particular stress that is often imposed by competitors,
to which JA may play a defensive role, is water deficit
(Grace and Tilman 1990, Creelman and Mullet 1997).
Jasmonic acid levels increase in roots and shoots of
maize under water deficit (Xin et al. 1997), and water
Ecology, Vol. 85, No. 1
deficit, wounding, and JA induce some of the same
genes in shoots of A. thaliana (Reymond et al. 2000).
In barley, JA is involved in the accumulation of abscisic
acid and proline in response to water deficit (Bandorska
and Stroinski 2001). Such evidence supports the in-
volvement of JA in plant responses to water deficit (or
other stresses) potentially imposed by competitors, but
this effect has never been quantified. Conversely, in-
creases in JA under water deficiency may explain why
drought-stressed plants can sometimes exhibit in-
creased chemical defense levels and resistance to her-
bivores (English-Loeb et al. 1997, Thaler and Bostock
2004). On the other hand, JA treatment can reduce
nitrate uptake, chlorophyll synthesis, CO2 uptake, and
growth in young Brassica napus (Rossato et al. 2002),
and has an inhibitory effect on primary root elongation
in A. thaliana (Staswick et al. 1992, 2002). While such
JA-induced responses can be costly, alternative benefits
could counterbalance at least some of their costs under
stressful conditions. For example, treatment of pea with
JA will diminish the inhibitory effect of NaCl stress
on CO2 fixation, chlorophyll fluorescence, and proline
accumulation (Velitchkova and Fedina 1998). JA is
known to increase naturally in salt-stressed Iris plants
as IAA and SA levels decline (Wang et al. 2001b).
Salicylic acid can have alternative benefits as well. In
addition to its signaling role in the induction of sys-
temic acquired resistance (Delaney 1997), SA can in-
crease heat and drought tolerance in A. thaliana (Lar-
kindale and Knight 2002), and both SA and JA can
increase chilling tolerance in tomato (Ding et al. 2002).
Induction of JA- and SA-mediated pathways with ex-
ogenous sprays was costly to fitness in individual A.
thaliana grown in pots in the greenhouse (presumably
due to physiological costs of defense production), but
costs of defense induction was no greater in target
plants surrounded by six uninduced intraspecific com-
petitors than in target plants grown alone (Cipollini
2002). This suggests that induction with JA and/or SA
had enough (unknown) positive benefits to competing
plants to counterbalance the additional costs that were
predicted to accrue under competition. Baldwin and
Hamilton (2000) showed that MeJA induction of to-
bacco reduced the ability of plants to compete for ni-
trogen, resulting in an increase in fitness costs, but their
experiments were conducted using a hydroponic design
where benefits to protection from water deficiency
could not be seen. Outside of work on phytochrome-
mediated responses to light competitors, signaling in
the induction of adaptive responses specific to com-
petitors has not been studied enough to know if and
how JA and/or SA might be involved. However, the
roles that these defense hormones can play in plant
responses to other environmental stresses (particularly
water deficiency), suggests that induction of these path-
ways could benefit competitive ability in plants, even
if only through positive side-effects on general
stress tolerance. Whether the outcomes of interactions
January 2004 PHYTOHORMONAL ECOLOGY
between plastic responses to competitors and herbi-
vores are positive or negative will likely depend upon
the limiting resource driving competition and the par-
ticular biochemical and morphological changes re-
quired to adaptively respond to the particular environ-
mental stressors.
BENEFITS OF SPECIFIC CHEMICAL DEFENSES
TO COMPETITIVE ABILITY
In addition to the roles of signaling hormones and
the pathways that they mediate in competitive ability,
some inducible defense proteins and secondary metab-
olites can directly affect competitors. Involvement of
chemical defenses in competitive interactions seems to
be a particular possibility for plants such as A. thaliana,
that produce glucosinolates in a constitutive and in-
ducible fashion (Kliebenstein et al. 2002). Glucosi-
nolates are known to affect feeding and growth of insect
herbivores in the Brassicaceae (Mauricio 1998, Stowe
1998). In addition, glucosinolates can possess allelo-
pathic properties (Brown and Morra 1995, Siemens et
al. 2002), and may benefit a plant by suppressing the
growth of neighboring vegetation if they are released
from the roots and hydrolyzed into bioactive isothio-
cyanates. Such benefits could obscure the known costs
of their production (e.g., Mauricio 1998, Stowe 1998,
Siemens et al. 2002) when expressed in certain eco-
logical contexts. This hypothesis has been dubbed the
defense-stress-benefit hypothesis (Siemens et al. 2002).
Alternative benefits of resistance traits are likely to be
species specific and dependent upon the mechanism of
resistance, as is suggested by the fact that uninduced
tobacco plants benefit from JA induction of their neigh-
bors (van Dam and Baldwin 1998, 2001, Baldwin and
Hamilton 2000). Different species may be better than
others at usurping resources or may be more or less
sensitive to the morphology and biochemistry of neigh-
bors of another species (Marriot and Zuazua 1996). In
the case of glucosinolates, the enzyme myrosinase is
required to convert glucosinolates into toxic byprod-
ucts than can act in allelopathy (Siemens et al. 2002).
Plants in the same family are more likely to contain
this enzyme and are more likely to be affected by glu-
cosinolates released by their neighbors than individuals
of species that do not contain this enzyme, or are not
growing in soils where microbial forms of this enzyme
are present (M. Morra, personal communication).
Plants that closely share resource requirements, in-
cluding plants in the same family, would be an appro-
priate target for allelopathy (Gersani et al. 2001,
Callaway 2002). Other competing species may not be
directly affected by glucosinolates, but could be indi-
rectly affected through negative effects of glucosino-
lates on their mycorrhizal associates (Roberts and An-
derson 2001). The release of specific volatile organic
compounds by plants, including A. thaliana, in re-
sponse to herbivore damage is being increasingly ap-
preciated (e.g., van Poecke et al. 2001). The possibility
33
exists that some of these herbivore-inducible com-
pounds could directly suppress the growth of neigh-
boring plants (Chuiha et al. 2002). Benefits of the ex-
pression of specific JA- and/or SA-mediated defenses
in environments including variation in nutrients, water,
or competitors may counterbalance some of their costs
seen in studies of single plants grown under relatively
benign conditions.
SUGGESTIONS FOR FUTURE STUDY
Despite some interesting observations and some
data, the interaction of adaptive plant responses to com-
petitors and induced responses to herbivores and path-
ogens has been little studied. Physiological evidence
of both negative and positive interactions between
these responses is available for plants such as A. thal-
iana, which shares features of phytochrome- and jas-
monate-signaling with a wide representation of plant
species from different families. Although information
derived from the use of A. thaliana may not always
fully extend to other plant-environment interactions, it
is a logical place to start. Wild accessions of A. thaliana
are known to vary in both phytochrome-mediated shade
avoidance responses (Botto and Smith 2002) and in
Special Feature
JA-inducible glucosinolate accumulation (Kliebenstein
et al. 2002). Genetic correlations between plasticity in
these responses could be examined over a similar range
of ecotypes to reveal important interactions that may
have ecological effects. Numerous mutant and trans-
genic A. thaliana genotypes also exist that are altered
in their ability to express adaptive responses to com-
petitors or herbivores, including the phytochrome-B-
deficient mutant, phyB (Sharrock and Quail 1989), and
the JA-response mutant jar1-1 (Staswick et al. 1992).
An examination of the phenotypic effects of such mu-
tations in an ecological context that includes compet-
itors, herbivores, and other stresses will broaden our
view of the functional ecology of these genes and the
responses that they mediate. In addition, selection ex-
periments in related species (e.g., Brassica rapa) have
produced genotypes that express varying levels of spe-
cific defenses (such as glucosinolates). Such plants can
be used to examine consequences of the expression of
certain defenses on shade-avoidance responses and oth-
er aspects of competitive ability (Stowe 1998, Siemens
et al. 2002).
In detecting mechanisms that underlie interactions
between plastic responses, one challenge will be to
determine the degree to which interactions occur due
to hormonal linkages between the expression of adap-
tive plant responses or due to passive resource-driven
processes. This can by accomplished in part by ma-
nipulating these responses independently of resources
by using exogenous hormones or other elicitors to mim-
ic natural induced responses (Cipollini and Schultz
1999, Cipollini 2002). In turn, external resource levels
could be manipulated along with plant phenotypes to
 34 DON CIPOLLINI
examine the passive resource control of interactions
between plastic responses.
At the gene-expression level, microarray analysis
could be used to examine coordinate regulation of
genes responsive to herbivores and competitors more
broadly than possible using other methods. For ex-
ample, expression of phytochrome-regulated and JA-
regulated genes could be compared to identify groups
of genes that share common regulatory patterns that
may scale up to have important phenotypic effects (as
in Maleck et al. 2000). In turn, transcripts could be
compared in plants simultaneously treated with jas-
monate (or herbivore feeding) and low R:FR light to
reveal up- or downregulation of specific genes under
simultaneous pressure from both competitors and her-
bivores. Microarray analysis has already revealed that
many phytochrome A-regulated genes in A. thaliana
encode putative transcriptional regulators, which could
affect a diverse array of downstream target genes, in-
cluding genes specific for defense (Tepperman et al.
2001). Cheong et al. (2002) used microarrays to show
that wounding can simultaneously induce many de-
fense-related genes and suppress many auxin-regulated
genes in A. thaliana. Microarray analysis has some
Special Feature
methodological limitations and the interpretation of
gene-expression patterns can be somewhat subjective,
but these issues continue to be refined and the tech-
nology has become cheaper and more accessible to
ecologists. While microarray analysis has been applied
primarily to A. thaliana, it is increasingly being applied
to new species (including rice, tomato, maize, and
spruce) as genome sequencing projects progress.
The best studies of interactions between plastic re-
sponses will link biochemical and molecular evidence
of interactions, revealed using techniques such as mi-
croarray analysis, with ecological consequences in a
range of environments. Data generated by physiolo-
gists, biochemists, and molecular geneticists focusing
on plants such as A. thaliana, can guide studies of the
functional ecology of genes involved in plastic re-
sponses to competitors and herbivores, regardless of
plant system (e.g., Pigliucci 1998, Schmitt et al. 1999,
Kliebenstein et al. 2002). If important mechanistic in-
teractions between these responses exist across plant
families, then models to explain costs, benefits, and
selection on these plastic traits will need to be refined.
ACKNOWLEDGMENTS
Research in my lab is currently supported by USDA NRI-
CGP grant 01-02783, the Ohio Plant Biotechnology Consor-
tium, and Wright State University. Comments by Anurag
Agrawal, Susan Mopper, and two anonymous reviewers great-
ly improved this manuscript.
LITERATURE CITED
Agrawal, A. A. 1998. Induced responses to herbivory and
increased plant performance. Science 279:12011202.
Agrawal, A. A., J. K. Conner, M. T. J. Johnson, AND R.
Wallsgrove. 2002. Ecological genetics of an induced plant
defense against herbivores: additive genetic variance and
costs of plasticity. Evolution 56:22062213.
Ecology, Vol. 85, No. 1
Agrawal, A. A., S. Y. Strauss, and M. J. Stout. 1999. Costs
of induced responses and tolerance to herbivory in male
and female fitness components of wild radish. Evolution
53:10931104.
Al-Mazri, M. I., M. S. Ali-Shtayeh, Y. Elad, A. Sharon, P.
Tudzynski, and R. Barakat. 2002. Effect of plant growth
regulators on white mould (Sclerotinia sclerotiorum) on
bean and cucumber. Phytopathologie Zeitschrift 150:481
487.
Bak, S., F. E. Tax, K. A. Feldmann, D. W. Galbraith, and R.
Feyereisen. 2001. CYP83B1, a cytochrome P450 at the
metabolic branch point in auxin and indole glucosinolate
biosynthesis in Arabidopsis. Plant Cell 13:101111.
Baldwin, I. T. 1998. Jasmonate-induced responses are costly
but benefit plants under attack in native populations. Pro-
ceedings of the National Academy of Sciences (USA) 95:
81138118.
Baldwin, I. T. 1999. Inducible nicotine production in native
Nicotiana as an example of adaptive phenotypic plasticity.
Journal of Chemical Ecology 25:330.
Baldwin, I. T., and P. Callahan. 1993. Autotoxicity and chem-
ical defense: nicotine accumulation and carbon gain in so-
lanaceous plants. Oecologia 94:534541.
Baldwin, I. T., and W. Hamilton. 2000. Jasmonate-induced
responses of Nicotiana sylvestris are costly due to impaired
competitive ability for nitrogen. Journal of Chemical Ecol-
ogy 26:915952.
Baldwin, I. T., C. L. Sims, and S. E. Kean. 1990. Repro-
ductive consequences associated with inducible alkaloidal
responses in wild tobacco. Ecology 71:252262.
Ball, M. C., J. J. G. Egerton, J. L. Lutze, V. P. Gutschik, and
R. B. Cunningham. 2002. Mechanisms of competition:
thermal inhibition of tree seedling growth by grass. Oec-
ologia 133:120130.
Ballare, C. L., and A. L. Scopel. 1992. Phytochrome sig-
nalling in plant canopies: testing its population-level im-
plications with photoreceptor mutants of Arabidopsis.
Functional Ecology 11:441450.
Ballare, C. L., A. L. Scopel, and R. A. Sanchez. 1990. Far-
red radiation reflected from adjacent leaves: an early signal
of competition in plant canopies. Science 247:329332.
Bandorska, H., and A. Stroinski. 2001. Investigation of jas-
monic acid participation in accumulation of abscisic acid
and proline and its role in response to water deficit in barley
genotypes. Acta Physiologia Plantarum 23:77.
Bates, T. R., and J. P. Lynch. 2001. Root hairs confer a com-
petitive advantage under low phosphorus availability. Plant
and Soil 236:243250.
Bazzaz, F. A., N. R. Chiarello, P. D. Coley, and L. F. Pitelka.
1987. Allocating resources to reproduction and defense.
BioScience 37:5867.
Behringer, J. F., and P. J. Davies. 1992. Indole-3-acetic acid
levels after phytochrome-mediated changes in the stem
elongation rate of dark- and light-grown Pisum seedlings.
Planta 188:8592.
Bergelson, J. 1994a. Changes in fecundity do not predict
invasiveness: a model study of transgenic plants. Ecology
75:249252.
Bergelson, J. 1994b. The effects of genotype and the envi-
ronment on costs of resistance in lettuce. American Nat-
uralist 143:349359.
Berger, S. 2002. Jasmonate-related mutants of Arabidopsis
as tools for studying stress signaling. Planta 214:497504.
Botto, J. F., and H. Smith. 2002. Differential genetic variation
in adaptive strategies to a common environmental signal
in Arabidopsis accessions: phytochrome-mediated shade
avoidance. Plant, Cell and Environment 25:5363.
Brown, P. D., and M. J. Morra. 1995. Glucosinolate-con-
taining plant tissues as bioherbicides. Journal of Agricul-
tural and Food Chemistry 43:30703074.
January 2004 PHYTOHORMONAL ECOLOGY
Callaway, R. M. 2002. The detection of neighbors by plants.
Trends in Ecology and Evolution 17:104105.
Campa, A. 1991. Biological roles of plant peroxidases:
known and potential function. Pages 2550 in J. Everse,
K. E. Everse, and M. B. Grisham, editors. Peroxidases in
chemistry and biology. Volume II. CRC Press, Boca Raton,
Florida, USA.
Celenza, J. L. 2001. Metabolism of tyrosine and trypto-
phannew genes for old pathways. Current Opinion in
Plant Biology 4:234240.
Cheong, Y. H., H. S. Chang, R. Gupta, X. Wang, T. Zhu, and
S. Luan. 2002. Transcriptional profiling reveals novel in-
teractions between wounding, pathogen, abiotic stress, and
hormonal responses in Arabidopsis. Plant Physiology 129:
661677.
Chuiha, K., F. Hu, and X. Xi. 2002. Allelopathic potential
of chemical constituents of volatiles from Ageratum co-
nyzoides under stress. Journal of Chemical Ecology 28:
11731182.
Cipollini, D. F. 1997a. Gibberellic acid treatment reduces the
tolerance of field-grown common bean to leaf removal.
Journal of Plant Growth Regulation 16:123127.
Cipollini, D. F. 1997b. Wind-induced mechanical stimulation
increases pest resistance in common bean. Oecologia 111:
8490.
Cipollini, D. F. 1998. Induced defenses and phenotypic plas-
ticity. Trends in Ecology and Evolution 13:200.
Cipollini, D. F. 2002. Does competition magnify the fitness
costs of induced responses in Arabidopsis thaliana? A ma-
nipulative approach. Oecologia 131:514520.
Cipollini, D. F., and J. Bergelson. 2001. Plant density and
nutrient availability constrain the constitutive and wound-
induced expression of trypsin inhibitors in Brassica napus
L. Journal of Chemical Ecology 27:593610.
Cipollini, D. F., and J. Bergelson. 2002. Interspecific com-
petition affects growth and herbivore damage of Brassica
napus L. in the field. Plant Ecology 162:227231.
Cipollini, D. F., and J. C. Schultz. 1999. Exploring cost con-
straints on stem elongation in plants using phenotypic ma-
nipulation. American Naturalist 153:236242.
Cipollini, D. F., and M. Sipe. 2001. Jasmonic acid treatment
and mammalian herbivory differentially affect chemical de-
fense expression and growth of Brassica kaber. Chemoe-
cology 11:137143.
Coley, P., D. Bryant, and F. S. Chapin. 1985. Resource avail-
ability and plant antiherbivore defense. Science 230:895
899.
Creelman, R. A., and J. A. Mullet. 1997. Biosynthesis and
action of jasmonates in plants. Annual Review of Plant
Physiology and Molecular Biology 48:355381.
Delaney, T. P. 1997. Genetic dissection of acquired resistance
to disease. Plant Physiology 113:512.
Delaney, T. P., L. Friedrich, and J. A. Ryals. 1995. Arabi-
dopsis signal transduction mutant defective in chemically
and biologically induced disease resistance. Proceedings of
the National Academy of Sciences (USA) 92:66026606.
Delaney, T. P., S. Uknes, B. Vernooij, L. Friedrich, K. Wey-
mann, D. Negrotto, T. Gaffney, M. Gut-Rella, H. Kess-
mann, E. Ward, and J. Ryals. 1994. A central role of sal-
icylic acid in plant disease resistance. Science 266:1247
1250.
Ding, C.-K., C. Wang, K. C. Gross, and D. L. Smith. 2002.
Jasmonate and salicylate induce the expression of patho-
genesis-related protein genes and increase resistance to
chilling injury in tomato fruit. Planta 214:895901.
Dixon, P., C. Weinig, and J. Schmitt. 2001. Susceptibility to
UV damage in Impatiens capensis (Balsaminaceae): testing
for opportunity costs to shade avoidance and population
differentiation. American Journal of Botany 88:14011408.
35
Dorn, L. A., E. H. Pyle, and J. Schmitt. 2000. Plasticity to
light cues and resources in Arabidopsis thaliana: testing
for adaptive value and costs. Evolution 54:19821994.
Dudley, S. A., and J. Schmitt. 1995. Genetic differentiation
in morphological responses to simulated foliage shade be-
tween populations of Impatiens capensis from open and
woodland sites. Functional Ecology 9:655666.
Dudley, S. A., and J. Schmitt. 1996. Testing the adaptive
plasticity hypothesis: density-dependent selection on ma-
nipulated stem length in Impatiens capensis. American Nat-
uralist 147:445465.
Duffy, N. M., S. P. Bonser, and L. Aarssen. 1999. Patterns
of variation in meristem allocation across genotypes of spe-
cies in monocarpic Brassicaceae. Oikos 84:284292.
English-Loeb, G., M. S. Stout, and S. S. Duffey. 1997.
Drought stress in tomatoes: changes in plant chemistry and
potential non-linear consequences for insect herbivores. Oi-
kos 79:456468.
Gersani, M., J. S. Brown, E. E. OBrien, G. M. Maina, and
Z. Abramsky. 2001. Tragedy of the commons as a result
of root competition. Journal of Ecology 89:660669.
Glazebrook, J. 2001. Genes controlling expression of defense
responses in Arabidopsis2001 status. Current Opinion in
Plant Biology 4:301308.
Grace, J., and D. Tilman. 1990. Perspectives on plant com-
petition. Academic Press, New York, New York, USA.
Gurevitch, J., J. A. Morrison, and L. V. Hedges. 2000. The
interaction between competition and predation: a meta-
analysis of field experiments. American Naturalist 155:
435453.
Special Feature
Henry, H. A. L., and S. C. Thomas. 2002. Interactive effects
of lateral shade and wind on stem allometry, biomass al-
location, and mechanical stability in Abutilon theophrasti
(Malvaceae). American Journal of Botany 89:16091615.
Herms, D. A. 1999. Physiological and abiotic determinants
of competitive ability and herbivore resistance. Phyton 39:
5364.
Herms, D. A., and W. J. Mattson. 1992. The dilemma of
plants: to grow or defend. Quarterly Review of Biology
67:283335.
Hsieh, H. L., H. Okamoto, M. Wang, L. H. Ang, M. Matsui,
H. Goodman, and X. W. Deng. 2000. FIN219, an auxin
regulated gene, defines a link between phytochrome A and
the downstream regulator COP1 in light control of Ara-
bidopsis development. Genes and Development 14:1958
1970.
Ishikawa, A., T. Yoshihara, and T. Nakumurak. 1994. Jas-
monate-inducible expression of a potato cathepsin D in-
hibitor GUS gene fusion in tobacco cells. Plant Molecular
Biology 26:403414.
Jacobsen, S. E., and N. E. Olszewski. 1996. Gibberellins
regulate the abundance of RNAs with sequence similarity
to proteinase inhibitors, dioxygenases, and dehydrogenas-
es. Planta 198:7886.
Jansen, M. A. K., R. E. van den Noort, M. Y. A. Tan, L. M.
Lagrimini, and R. N. F. Thorneley. 2001. Phenol-oxidizing
peroxidases contribute to the protection of plants from ul-
traviolet radiation stress. Plant Physiology 126:10121023.
Jupe, S. C., and I. M. Scott. 1992. Gibberellin and the pro
gene suppress peroxidase activity in elongating tomato (Ly-
copersicon esculentum) stem tissue. Annals of Botany 69:
3337.
Karban, R., and I. T. Baldwin. 1997. Induced responses to
herbivory. University of Chicago Press, Chicago, Illinois,
USA.
Kliebenstein, D. J., A. Figuth, and T. Mitchell-Olds. 2002.
Genetic architecture of plastic methyl jasmonate responses
in Arabidopsis thaliana. Genetics 161:16851696.
Lagrimini, L. M. 1991. Wound-induced deposition of poly-
phenols in transgenic plants overexpressing peroxidase.
Plant Physiology 96:577583.
 36 DON CIPOLLINI
Lagrimini, L. M. 1999. The role of peroxidase in auxin me-
tabolism. Abstracts peroxidase 99. International sympo-
sium for plant peroxidases. Columbus, Ohio, USA.
Lagrimini, L. M., V. Gingas, F. Finger, S. Rothstein, and T.-
T. Y. Liu. 1997. Characterization of antisense transformed
plants deficient in the tobacco anionic peroxidase. Plant
Physiology 114:11871196.
Larkindale, J., and M. R. Knight. 2002. Protection against
heat stress-induced oxidative damage in Arabidopsis in-
volves calcium, abscisic acid, ethylene, and salicylic acid.
Plant Physiology 128:682695.
Lentz, K., and D. F. Cipollini. 1998. Effect of light and sim-
ulated herbivory on growth of endangered northeastern bul-
rush, Scirpus ancistrochaetus Schuyler. Plant Ecology 139:
125131.
Lively, C. M., S. G. Johnson, L. F. Delph, and K. Clay. 1995.
Thinning reduces the effect of rust infection on jewelweed
(Impatiens capensis). Ecology 76:18591862.
Ma, Z., T. C. Walker, A. Ma, and J. P. Lynch. 2001. Mor-
phological synergism in root hair length, density, initiation,
and geometry for phosphorus acquisition in Arabidopsis
thaliana: a modeling approach. Plant and Soil 236:221
235.
Maleck, K., A. Levine, T. Eulgem, A. Morgan, J. Schmid, K.
A. Lawton, J. Dangl, and R. A. Dietrich. 2000. The tran-
scriptome of Arabidopsis thaliana during systemic acquired
resistance. Nature Genetics 26:403410.
Marriot, C. A., and M. T. Zuazua. 1996. Tillering and par-
titioning of dry matter and nutrients in Lolium perenne
growing with neighbors of different species: effect of nu-
Special Feature
trient supply and defoliation. New Phytologist 132:8795.
Mauch-Mani, B., and J.-P. Metraux. 1998. Salicylic acid and
systemic acquired resistance to pathogen attack. Annals of
Botany 82:535540.
Mauricio, R. 1998. Costs of resistance to natural enemies in
field populations of the annual plant Arabidopsis thaliana.
American Naturalist 151:2028.
McConn, M., R. A. Creelman, E. Bell, J. Mullet, and J.
Browse. 1997. Jasmonate is essential for insect defense in
Arabidopsis. Proceedings of the National Academy of Sci-
ences (USA) 94:54735477.
Mikkelsen, M. D., C. H. Hansen, U. Wittstock, and B. A.
Halkier. 2000. Cytochrome P450 CYP79B2 from Arabi-
dopsis catalyzes the conversion of tryptophan to indole-3-
acetaldoxime, a precursor of indole glucosinolates and in-
dole-3-acetic acid. Journal of Biological Chemistry 275:
3371233717.
Morelli, G., and I. Ruberti. 2000. Shade avoidance responses:
driving auxin along lateral routes. Plant Physiology 122:
621626.
Morgan, D. C., and H. Smith. 1978. The relationship between
phytochrome photoequilibrium and development in light-
grown Chenopodium album L. Planta 142:187193.
Neff, M. N., C. Fankhauser, and J. Chory. 2000. Light, an
indicator of time and place. Genes and Development 14:
257271.
Normanly, J. 1997. Auxin metabolism. Physiologia Plantar-
um 100:431442.
Palecanda, L., and R. A. Sharrock. 2001. Molecular and phe-
notypic specificity of an antisense PHYB gene in Arabi-
dopsis. Plant Molecular Biology 46:8997.
Peng, J., and N. P. Harberd. 1997. Gibberellin deficiency and
response mutations suppress the stem elongation phenotype
of phytochrome-deficient mutants of Arabidopsis. Plant
Physiology 113:10511058.
Pigliucci, M. 1998. Ecological and evolutionary genetics of
Arabidopsis. Trends in Plant Sciences 3:485489.
Reader, R. J. 1992. Herbivory as a confounding factor in an
experiment measuring competition among plants. Ecology
73:373376.
Ecology, Vol. 85, No. 1
Retuerto, R., L. Rochefort, and F. I. Woodward. 1996. The
influence of plant density on the responses of Sinapis alba
to CO2 and windspeed. Oecologia 108:241251.
Reymond, P., H. Weber, M. Damond, and E. E. Farmer. 2000.
Differential gene expression in response to mechanical
wounding and insect feeding in Arabidopsis. Plant Cell 12:
707719.
Rhoades, D. F. 1979. Evolution of plant chemical defense
against herbivores. Pages 354 in G. A. Rosenthal and D.
H. Janzen, editors. Herbivores: their interactions with sec-
ondary plant metabolites. Academic Press, New York, New
York, USA.
Roberts, K. J., and R. C. Anderson. 2001. Effect of garlic
mustard (Alliaria petiolata (Beib. Cavara & Grande)) ex-
tracts on plants and arbuscular mycorrhizal (AM) fungi.
American Midland Naturalist 146:146152.
Robinson, D., A. Hodge, B. S. Griffiths, and A. H. Fitter.
1999. Plant root proliferation in nitrogen-rich patches con-
fers competitive advantage. Proceedings of the Royal So-
ciety of London Series B 266:431435.
Rossato, L., J. H. MacDuff, P. Laine, E. LeDeunff, and A.
Ourry. 2002. Nitrogen storage and remobilization in Bras-
sica napus L. during the growth cycle: effects of methyl
jasmonate on nitrate uptake, senescence, growth, and VSP
production. Journal of Experimental Botany 53:1131
1141.
Schmitt, J. 1997. Is photomorphogenic shade avoidance
adaptive? Perspectives from population biology. Plant, Cell
and Environment 20:826830.
Schmitt, J., S. A. Dudley, and M. Pigliucci. 1999. Manipu-
lative approaches to testing adaptive plasticity: phyto-
chrome-mediated shade avoidance responses in plants.
American Naturalist 154:4354.
Sharrock, R. A., and P. H. Quail. 1989. Novel phytochrome
sequences in Arabidopsis thaliana: structure, evolution, and
differential expression of a plant regulatory photoreceptor
family. Genes and Development 3:17451757.
Siemens, D. H., A. S. Garner, T. Mitchell-Olds, and R. M.
Callaway. 2002. Cost of defense in the context of plant
competition: Brassica rapa may grow and defend. Ecology
83:505517.
Smith, H., and G. G. Whitelam. 1997. The shade avoidance
syndrome: multiple responses mediated by multiple phy-
tochromes. Plant, Cell and Environment 20:840844.
Staswick, P. E., W. Su, and S. H. Howell. 1992. Methyl jas-
monate inhibition of root growth and induction of a leaf
protein are decreased in an Arabidopsis thaliana mutant.
Proceedings of the National Academy of Sciences (USA)
89:68376840.
Staswick, P. E., I. Tiryaki, and M. L. Rowe. 2002. Jasmonate
response locus JAR1 and several related Arabidopsis genes
encode enzymes of the firefly luciferase superfamily that
show activity on jasmonic, salicylic, and indole-3-acetic
acids in an assay for adenylation. Plant Cell 14:14051415.
Stowe, K. 1998. Experimental evolution of resistance in
Brassica rapa: correlated response of tolerance in lines
selected for glucosinolate level. Evolution 52:703712.
Sultan, S. E. 2000. Phenotypic plasticity for plant develop-
ment, function, and life history. Trends in Plant Science 5:
537542.
Tepperman, J. M., T. Zhu, H. S. Chang, X. Wang, and P. H.
Quail. 2001. Multiple transcription-factor genes are early
targets of phytochrome A signaling. Proceedings of the
National Academy of Sciences (USA) 98:94379442.
Thaler, J. S., and R. M. Bostock. 2004. Interactions between
abscisic-acid-mediated responses and plant resistance to
pathogens and insects. Ecology 85:4858.
Thornburg, R. W., and X. Li. 1991. Wounding Nicotiana
tabacum leaves causes a decline in endogenous IAA. Plant
Physiology 96:802805.
January 2004 PHYTOHORMONAL ECOLOGY
van Dam, N. M., and I. T. Baldwin. 1998. Costs of jasmonate-
induced responses in plants competing for limited resourc-
es. Ecology Letters 1:3033.
van Dam, N. M., and I. T. Baldwin. 2001. Competition me-
diates costs of jasmonate-induced defenses, nitrogen ac-
quisition and transgenerational plasticity in Nicotiana at-
tenuata. Functional Ecology 15:406415.
van Poecke, R. M. P., M. A. Posthumus, and M. Dicke. 2001.
Herbivore-induced volatile production by Arabidopsis thal-
iana leads to attraction of the parisitoid Cotesia rubecula:
chemical, behavioral, and gene-expression analysis. Jour-
nal of Chemical Ecology 27:19111928.
Velitchkova, M., and I. Fedina. 1998. Response of photo-
synthesis of Pisum sativum to salt stress as affected by
methyl jasmonate. Photosynthetica 35:8997.
Wang, Y. H., D. F. Garvin, and L. V. Kochian. 2001a. Nitrate-
induced genes in tomato roots. Array analysis reveals novel
genes that may play a role in nitrogen nutrition. Plant Phys-
iology 127:345359.
Wang, Y., S. Mopper, and K. H. Hasenstein. 2001b. Effects
of salinity on endogenous ABA, IAA, JA, and SA in Iris
hexagona. Journal of Chemical Ecology 27:327342.
37
Weinig, C. 2000a. Limits to adaptive plasticity: temperature
and photoperiod influence shade-avoidance responses.
American Journal of Botany 87:16601668.
Weinig, C. 2000b. Plasticity versus canalization: population
differences in the timing of shade-avoidance responses.
Evolution 54:441451.
Weinig, C., and L. F. Delph. 2001. Phenotypic plasticity early
in life constrains developmental responses later. Evolution
55:930936.
Weis, A. E., and M. E. Hochberg. 2000. The diverse effects
of intraspecific competition on the selective advantage to
resistance: a model and its predictions. American Naturalist
156:276292.
Xin, Z.-Y., X. Zhou, and P. E. Pilet. 1997. Level changes of
jasmonic, abscisic, and indole-3yl-acetic acids in maize
under desiccation stress. Journal of Plant Physiology 151:
120124.
Zangerl, A. R., and M. R. Berenbaum. 1990. Furanocoumarin
induction in wild parsnip: genetics and population varia-
tion. Ecology 71:19331940.
Special Feature
 Ecology, 85(1), 2004, pp. 3847
q 2004 by the Ecological Society of America

IRIS HEXAGONA HORMONAL RESPONSES TO SALINITY STRESS,

LEAFMINER HERBIVORY, AND PHENOLOGY
SUSAN MOPPER,1 YONGYIN WANG,2 CECIL CRINER, AND KARL HASENSTEIN
Department of Biology, 300 East Saint Mary Boulevard, University of Louisiana, Lafayette, Louisiana 70504-2451 USA

Abstract. We used field, common garden, and laboratory studies to investigate the
influence of environmental stress on a native wetland species, Iris hexagona (Iridaceae),
and an iris leafminer, Cerodontha iridiphora (Diptera: Agromyzidae). There were strong
effects of salinity, herbivory, and phenology on jasmonic acid (JA) and salicylic acid (SA),
two important signal phytohormones. Within 48 h, foliar JA increased an order of magnitude,
and SA declined 50% in plants grown under salinity stress compared to freshwater controls.
Leafminer performance declined on plants exposed to the salinity treatment. Both herbivory
and salinity stress induced JA and suppressed SA, indicating that plants can respond to
abiotic and biotic stress through shared hormonal pathways. Surprisingly, leafminer pupae
contained JA and SA, and we discuss the implications of this novel discovery. Unrelated
to salinity, there was a sharp decline in foliar JA between April and September. Despite
its strong impact on hormonal signaling in plants, the abiotic environment is often ignored
in ecological studies of induced plant defenses. Importantly, natural seasonal cycles and/
or environmental stress can induce stronger changes in phytohormones than herbivory, and
should be incorporated into the design of ecological experiments whenever possible, for a
balanced understanding of phytohormonal ecology.
Key words: abiotic stress; Cerodontha iridiphora; cross-kingdom signal sharing; Iris hexagona;
Special Feature

jasmonic acid; leaf nitrogen; phytohormones; plant defense; plantinsect interactions; salicylic acid;
signaling molecules.

INTRODUCTION environment (Berger 2002). JA and SA regulate sep-

Plants are exposed to a multitude of environmental
challenges they must identify and respond to. Major
advances are occurring in the biology of plant molec-
ular signals, the pathways they initiate, and the adap-
tations that ensue. Much research is focused on phy-
tohormones, which are regulatory molecules that act
at very low concentrations and have numerous anti-
stress functions (Hasegawa et al. 2000, Knight and
Knight 2001, Weyers and Paterson 2001). The first step
in a cascade of events is the production of one or more
molecular signals that link environmental stimuli with
the appropriate physiological responses (Karban and
Baldwin 1997). These processes are not linear and spe-
cific as previously believed, but are complex and in-
teractive networks (Stout and Bostock 1999, Knight
and Knight 2001).
Jasmonic acid (JA) and salicylic acid (SA) are small,
simple hormones crucial in many plant responses to
ecological challenges (Agrawal et al. 1999, Stotz et al.
2002). The concentration and ratio of these and related
compounds are important determinants of ecological
interactions between plants, natural enemies, and the
Manuscript received 12 November 2002; revised 11 March
2003; accepted 22 March 2003; final version received 28 April
2003. Corresponding Editor: A. A. Agrawal. For reprints of this
Special Feature, see footnote 1, p. 3.
1 E-mail: mop@louisiana.edu
2 Present address: Mount Sinai Medical Center, New York,
New York 10029 USA.
38
arate biochemical pathways, and have distinct func-
tions in plant development, reproduction, and physi-
ology (Karban and Baldwin 1997). The JA- and SA-
signaling pathways function independently or can in-
teract through crosstalk, a process in which signal
molecules affect each others outcome in additive or
negative ways (Chao et al. 1999, Knight and Knight
2001, Stotz et al. 2002). JA and SA have opposite
effects on PR genes in tobacco (Niki et al. 1998) and
tomato (Thaler et al. 1999). SA inhibits JA and its
precursors in flax (Harms et al. 1998) and Arabidopsis
(Gupta et al. 2000). The crosstalk that occurs between
hormonal signals and their metabolic pathways (Moons
et al. 1997, Bostock et al. 2001, Knight and Knight
2001) creates complex networks of activity that vary
spatially and temporally (Hatcher et al. 2004, Voesenek
et al. 2004).
Jasmonic acid (JA) and its methyl ester (JAMe) vary
as a function of plant tissue type, developmental stage,
and external stimuli. The highest concentrations typi-
cally occur in flowers and reproductive tissues (Creel-
mann and Mullet 1995). JA is elicited by herbivory
and mechanical wounding (McCloud and Baldwin
1997, Baldwin 1998), water deficit (Creelman and Mul-
let 1997), microbial activity (Blechert et al. 1995), and
plant-signaling peptides (Bergey et al. 1996). The sys-
temic induction of JA by wounding or pathogens has
been attributed to damaged cell membranes that release
fatty acids, which then metabolize via lipoxygenase to
January 2004 PHYTOHORMONAL ECOLOGY
JA (Sembdner and Parthier 1993). Jasmonates and their
derivatives trigger the expression of plant chemical de-
fenses such as proteinase inhibitors, which disrupt her-
bivore digestive processes (Farmer et al. 1992, Karban
and Baldwin 1997). Exogenous application of JA to
plants induces systemic accumulation of compounds
with antiherbivore properties, and elicits volatile emis-
sions that attract herbivore natural enemies (Thaler
1999, Thaler et al. 2001).
Salicylic acid, a phenolic biosynthesized from ben-
zoic acid and phenylalanine, is an essential constituent
in the signal transduction pathways of systemic plant
defenses against herbivores and pathogens (Raskin
1992, Hammerschmidt and Smith-Becker 1999). Sal-
icylic acid initiates pathogen induced systemic ac-
quired resistance (SAR) and triggers the expression of
pathogenesis related (PR) genes, which code for de-
fensive proteins (Klessig and Malamy 1994). SA is the
precursor of methyl salicylate (MSA), a primary in-
gredient in pollinator-attracting floral scent emissions
(Raguso and Light 1998, Pott et al. 2002). Salicylates
are induced by herbivory and can reduce herbivore
performance and deter feeding (Pierpoint 1994). Sa-
licylates also occur in green leaf volatiles that attract
herbivore natural enemies to damaged plants (Van Peo-
cke et al. 2001).
Many new insights into the functions of phytohor-
mones have arisen through exhaustive work on model
organisms like Arabidopsis in controlled studies that
eliminate as much environmental variation as possible
(Berger 2002). Ecologists studying nonmodel, native
plant species in natural settings are challenged to make
sense of plant signaling processes in the face of tre-
mendous environmental variability (Farnsworth 2004).
This has produced a dichotomy in phytohormonal ecol-
ogy, in which biotic factors such as herbivory are in-
vestigated separately from abiotic stress (e.g., Kessler
and Baldwin 2002), despite the powerful effects and
simultaneous occurrence of both. Recent ecological
studies have addressed these concerns by developing
new conceptual models of plant-signal responses and
adaptations to diverse natural enemies and environ-
mental conditions (Cippolini 2004, Hatcher et al. 2004,
Schultz and Appel 2004, Thaler and Bostock 2004, and
Voesenek et al. 2004).
To understand plant-signal responses to the environ-
ment, it is important to integrate controlled laboratory
and ecological field studies. We are investigating the
physiological ecology of Iris hexagona (Iridaceae), a
native wetland perennial, and its interactions with Cer-
odontha iridiphora (Diptera: Agromyzidae), an iris
leafminer. Salinity stress is a formidable and escalating
environmental threat to wild and cultivated plants (Yeo
1999, Rogers and McCarty 2000). Elevated sodium
weakens or kills plants by disrupting cellular, meta-
bolic, and genetic processes (Levitt 1980, Hasegawa et
al. 2000), and coastal populations are particularly vul-
nerable (Pezeshki et al. 1990, Howard and Mendels-
39
sohn 1999). One example in our study system is Hur-
ricane Lili, which in October 2002 submerged coastal
Louisiana under a 23-m storm surge that raised en-
vironmental salinity in freshwater wetlands more than
10-fold.
Even small increases in salinity can dramatically af-
fect freshwater plants. For example, 24 NaCl re-
duced I. hexagona biomass by 40% and tripled seed
production (Van Zandt et al. 2003), delayed flowering
phenology by several days (Van Zandt and Mopper
2002), and significantly altered concentrations of JA,
SA, abscisic acid (ABA), and auxin, four phytohor-
mones with critical regulatory functions (Wang et al.
2001). The salicylates and jasmonates that are so im-
portant in plantenemy interactions also regulate plant
responses to salinity and other stresses (Iuchi et al.
1996, Tsonev et al. 1998, Chao et al. 1999, Hoyos and
Zhang 2000, Wang et al. 2001), and both JA (Creelman
and Mullet 1997) and SA (Yen and Yen 1999) can
ameliorate the effect of salt on plant cells.
In this paper, we present laboratory, common garden,
and field studies that were conducted in 1998 and 1999
with Iris hexagona (Iridaceae), and an iris leafminer,
Cerodontha iridiphora (Diptera: Agromyzidae). We fo-
Special Feature
cused on JA and SA because both are induced by her-
bivory and salinity stress. The experiments presented
here were designed to test independent questions; each
contributes unique information, and as a whole they
illustrate why long-term studies in variable environ-
ments are necessary to understand the ecological roles
of plant signal hormones.
METHODS
Background on study system
Iris hexagona (Walters), in the family Iridaceae, is
native to Louisianas freshwater and brackish wetlands.
The species is a long-lived perennial that reproduces
sexually and clonally. The I. hexagona populations in-
vestigated in this study are patchily distributed
throughout 300-km2 Marsh Island (298379 N, 918549
W) that was a freshwater wetland (Orton 1959) before
it separated from the Mississippi Delta mainland rough-
ly 20005000 years ago (Coleman 1988, Tornqvist et
al. 1996). Marsh Island salinity varies seasonally from
215 (Van Zandt and Mopper 2002) and is consid-
erably higher than mainland I. hexagona habitats. Is-
land I. hexagona populations are small and isolated
compared with freshwater populations, which grow in
large and continuous aggregations of thousands of in-
dividuals (see Plate 1).
Cerodontha iridiphora Spencer (Agromyzidae: Dip-
tera) is an iris leafminer that establishes mines on the
upper leaf surface in May through June. Mines contain
multiple larvae that pupate in the basal portion of the
mined leaf. Leaf damage is severe and mined leaves
senescence sooner than unmined leaves (L. Schile and
S. Mopper, unpublished data). Little is known about
 40 SUSAN MOPPER ET AL.
causes of C. iridiphora mortality, but plant defenses
are an important source of mortality in other leafminer
species, particularly in the early instars (Connor and
Taverner 1997, Mopper et al. 2000).
Leafminer density and pupal mass in natural
I. hexagona populations
Special Feature
We used airboats to access three remote Marsh Island
I. hexagona populations in 1999. Each population con-
sisted of ;250 individual ramets, each with 525
leaves. We examined an average of 18 ramets per pop-
ulation for evidence of leafminer feeding, and maxi-
mized the distance between plants to reduce the chance
of resampling the same genotypes. Density estimates
were calculated as the mean number of mines per leaf.
On 9 July, we collected 19, 22, and 91 leafminer pupae
from three I. hexagona populations that were separated
by at least 1 km. We also determined the interstitial
soil salinity at each site (Orion 125 conductivity meter;
Orion Research, Beverly, Massachusetts, USA). Pupae
were returned to the laboratory and weighed.
The common garden
We established a common garden in the winter of
1997 with plants collected from 10 I. hexagona natural
populations. Irises were grown in 30 227-L Rubber-
maid mesocosms (Rubbermaid, Wooster, Ohio, USA)
in a 30 3 30-m plot at the University of Louisiana
Ecology Center. Each mesocosm contained ;10 I. hex-
agona plants with a total of 250 leaves. Control (n 5
10) and salinity (n 5 10) treatments were randomly
assigned to mesocosms and a solution of Instant Ocean
(Aquarium Systems, Mentor, Ohio, USA) was applied
as necessary to maintain salinity. Interstitial salinity
averaged 0.01 and 6 NaCl, in the freshwater control
and salinity treatment, respectively.
Leafminer performance in the common garden
In September 1998, we collected I. hexagona plants
from Marsh Island for an unrelated experiment and
Ecology, Vol. 85, No. 1
PLATE 1. A natural population of Iris hex-
agona in a saline habitat. Populations in fresh-
water marsh contain thousands of individuals,
but brackish marsh populations, such as this
one, are small and discrete and can withstand
substantially elevated salinity. Photo credit: S.
Mopper.
distributed them among the established common gar-
den experimental mesocosms. The new plants con-
tained C. iridiphora leafminer pupae. When adults
emerged in the spring, they oviposited on the common
garden irises, and eggs hatched in May. In June, we
counted the number of leaves and the number of mines,
and measured leaf area. We collected a total of 164
pupae from five freshwater and six salinity mesocosms
in the common garden and weighed them in the lab.
Leaf nitrogen and carbon analyses
We used a Carlo Erba NC-2500 Elemental Analyzer
(CE Elantech, Lakewood, New Jersey, USA) to deter-
mine the total nitrogen (N2) and total carbon (CO2)
content of irises growing in the control and salinity
treatments. In June 1999, we removed a 5-cm portion
of leaf tissue from six plants in each experimental me-
socosm. We dried the tissue at 608C for 5 d, then ground
it to a fine powder in an IKA A-10 analytical mill (IKA
Works, Wilmington, North Carolina, USA). We used
six atropine (N 5 4.84%, C 5 70.6%) samples of 0.1,
0.5, 1.0, 2.0, 3.5, and 5.0 mg to establish a standard
curve. Samples of 4 6 0.5 mg ground plant tissue were
placed into individual tin sample containers. Each leaf
was analyzed at least twice and only samples with #5%
difference between replicates were included in the anal-
ysis. We ran atropine controls at 20-sample intervals
to ensure machine calibration.
Analysis of JA and SA in iris leaves
and leafminer pupae
We analyzed phytohormone concentration in plants
grown under different salinity levels in the laboratory
and common garden. In spring 1998, we filled 20 4-L
plastic containers with deionized water and mixed in
sufficient NaCl to produce 0, 100, 200, and 400 mmol/
L concentrations (equivalent to 0, 5.8, 11.2,
23.6; see methods in Wang et al. [2001]). Four I.
hexagona were placed in each container. Two plants
each were exposed to salinity for 24 or 48 h. We col-
January 2004 PHYTOHORMONAL ECOLOGY 41

Special Feature
FIG. 1. Induction of JA and suppression of SA in I. hexagona leaves by different NaCl levels (after Wang et al. 2001).
The four salinity levels are 0 (control), 100 mmol/L (5.6), 200 mmol/L (11.6), and 400 mmol/L (23.2).

lected foliage and measured JA and SA concentration. Statistical analysis

Monthly, from April through September, we collected
leaf samples from I. hexagona growing in the control
and salinity mesocosms in the common garden (meth-
ods in Wang et al. [2001]). In June 1999, we collected
foliage from unmined and mined irises growing in three
freshwater mesocosms. Plant material was placed on
ice in the field and stored in the lab at 2708C until
extraction. We ground ;1 g of plant tissue into fine
powder under liquid N2 using a mortal and pestle. We
processed and analyzed undamaged portions of mined
and unmined leaves, as well as 70 leafminer pupae that
were collected from two freshwater mesocosms. Leaf-
miner pupae were ground fresh in liquid nitrogen. We
used high performance liquid chromatography fol-
lowed by gas chromatography with mass spectroscopy
in selected ion mode to extract, purify, and quantify
endogenous JA and SA in plants and leafminer pupae
(detailed methods in Wang et al. [2001]).
We conducted the statistical analyses with SAS Ver-
sion 8 software (SAS 1999). Plant performance and
leafminer data were averaged within individual me-
socosms, and each was considered a replicate. It was
not possible to obtain data for each comparison from
all mesocosms, so sample sizes vary. We used PROC
UNIVARIATE to examine data normality and variance
and applied the appropriate transformations to meet
parametric assumptions. All proportional data were
arcsine square root transformed prior to analysis and
back-transformed for presentation. We ran correlation
and multivariate analyses for potentially correlated
plant and leafminer variables. We used PROC GLM
for analysis of variance, multivariate analysis of var-
iance, and repeated-measures analysis of variance to
assess plant characteristics, leafminer performance,
and phytohormone concentrations. We used Tukey ad-
justed pairwise comparisons of least-squares means.
 42 SUSAN MOPPER ET AL. Ecology, Vol. 85, No. 1

F1,13 5 9.5, P 5 0.001). In addition, the C/N ratio was

FIG. 2. Long-term pattern in foliar JA production in com-
mon garden I. hexagona plants growing in control (no salin-
ity) and salinity (6) treatments. Bars show means 6 1 SE
(after Wang et al. 2001)
RESULTS
Short term effect of salinity on JA and SA in leaves
In the laboratory experiment, salinity rapidly altered
JA and SA concentration in iris leaves (Fig. 1, from
Special Feature
22% lower in the salinity treatment (not shown, F1,13
5 8.6, P 5 0.01). The number of leaves produced by
control and salinity plants did not vary ( F1,19 5 0.41,
P 5 0.528).
Effect of salinity on leafminer performance
in the common garden and field
Experimental salinity in the common garden nega-
tively affected leafminer performance (Fig. 4, Wilks
lambda, F7,3 5 7.73, P 5 0.06). The four response
variables were not correlated in the MANOVA analysis
(Table 1). In the salinity treatment, 60% fewer larvae
survived to pupation than in the control treatment (Fig.
4A, F1,9 5 5.4, P 5 0.046), and salinity treatment plants
contained 43% fewer pupae per mine than controls
(Fig. 4B, F1,9 5 21.0, P 5 0.001). Leafminer density
(Fig. 4C) and pupal mass (Fig. 4D) were lower in the
salinity treatment, but the differences between control
and salinity were not significant (F1,19 5 2.76, P 5
0.11, and F1,15 5 1.89, P 5 0.19, respectively).

Wang et al. 2001). Salinity increased JA an order of

magnitude from 0.2 to almost 2 mg/g dry mass (Fig.
1A, F3,16 5 14.0, P 5 0.0001). There was a significant
time effect on JA production (F1,16 5 77.6, P 5 0.0001).
The 100-mmol/L NaCl treatment produced no signif-
icant change after 24 or 48 h of exposure, but the 200-
mmol/L and the 400-mmol/L treatments caused sig-
nificant increases after 48 h (P 5 0.0026 and P 5
0.0008, respectively). In contrast to the induction of
JA, salinity reduced SA levels significantly (F3,16 5
17.9, P 5 0.0001, Fig. 1B). Within 24 h of treatment
with 200 mmol/L NaCl, SA had dropped by 50% (P
5 0.0001).

Long-term changes in phytohormones

In the common garden, foliar JA declined an order
of magnitude from April to September (Fig. 2, F5,15 5
25.0, P 5 0.001), but there was no overall salinity
effect on JA (F1,3 5 0.0, P 5 0.974), and no salinity-
by-month interaction (F5,15 50.24, P 5 0.798). Uni-
variate probabilities for monthly differences in foliar
JA between salt-stressed and control plants are: PApril
5 0.971, PMay 5 0.265, PJune 5 0.081, PJuly 5 0.069,
PAugust 5 0.764, PSeptember 5 0.024. We did not determine
foliar SA levels, but it declined significantly in the
seeds of salt-stressed irises over a 2-mo period (Wang
et al. 2001).

Effect of salinity on Iris hexagona

Salinity reduced common garden I. hexagona leaf
area by 10% (Fig. 3A, F1,19 5 9.8, P 5 0.01). Salinity FIG. 3. Effect of freshwater and 6 NaCl on (A) leaf
had no effect on carbon (Fig. 3B, F1,13 5 1.9, P 5 area, (B) leaf carbon, and (C) leaf nitrogen in common garden
0.19), but increased total leaf nitrogen 22% (Fig. 3C, I. hexagona plants. Bars show means 6 1 SE.
January 2004 PHYTOHORMONAL ECOLOGY 43

Special Feature
FIG. 4. Performance of iris leaf miners on plants in freshwater and 6 NaCl treatments. (A) Total number of pupae that
survived, (B) number of mines per leaf, (C) number of pupae within individual mines, and (D) pupal mass. Bars show means
6 1 SE.

Leafminer densities averaged 40.9 6 16% (mean 6
1 SE) mines per leaf in three Marsh Island I. hexagona
populations. Pupal mass varied significantly among iris
populations (Fig. 5, F2, 129 5 4.45, P 5 0.014) and was
negatively correlated with interstitial salinity at the
population site (r 520.99, P 5 0.017, n 5 3). Leaf-
miner density was not correlated with interstitial sa-
linity (r 5 0.04, P 5 0.97, n 5 3).
Effect of salinity on JA and SA in mined and
unmined leaves and leafminer pupae
Jasmonic acid and salicylic acid varied significantly
among mined and unmined iris leaves and leafminer
pupae (Wilks lambda, F4,8 5 5.24, P 5 0.0227). There
was a nonsignificant negative correlation between JA
and SA (r 5 20.412, P 5 0.418). Mined leaves contain
almost twice as much JA as unmined leaves (Fig. 6A),
but the difference was not significant (P 5 0.13). Leaf-
miner pupae contained four times more JA than control
leaves (P 5 0.0014), and twice as much JA as mined
leaves (P 5 0.006).
Foliar SA was substantially higher than JA (Fig. 6B).
The greatest amount of SA occurred in control leaves,
and did not differ from mined leaves (P 5 0.536).
Leafminer pupae contained significantly less SA than
control leaves (P 5 0.038), but not mined leaves (P 5
0.116). There are no data for mined, salt-stressed
plants, but mining and salinity appear to have additive,
not inhibitory, effects, and both induce JA and suppress
SA production.
DISCUSSION
Jasmonic acid (JA) has been viewed as a key com-
ponent of plant antiherbivore defense, whereas salicylic
acid (SA) is best known for its importance in plant
systemic responses to pathogen attack. These distinc-

TABLE 1. Partial correlation coefficients and probabilities (in parentheses) from the MANOVA error SSCP matrix of
leafminer performance on common garden I. hexagona.

Parameter No. mines/leaf Total no. pupae No. pupae/mine Pupal mass (mg)
No. mines/leaf 1.000000 0.898645 (0.0004) 20.274579 (0.4426) 0.127476 (0.7256)
Total no. pupae 0.898645 (0.0004) 1.00000 0.037525 (0.9180) 0.163474 (0.6518)
No. pupae/mine 20.274579 (0.4426) 0.037525 (0.9180) 1.000000 0.348716 (0.3234)
Pupal mass (mg) 0.127476 (0.7256) 0.163474 (0.6518) 0.348716 (0.3234) 1.000000
 44 SUSAN MOPPER ET AL. Ecology, Vol. 85, No. 1

Different types of herbivory induce specific hor-

FIG. 5. Pupal mass of leafminers (bars, left axis) and in-
terstitial salinity ( symbols, right axis) from three Marsh
Island I. hexagona populations. Bars show means 6 1 SE,
letters above indicate (P # 0.05) differences based on Tukey
adjusted pairwise comparisons. Correlation between pupal
mass and salinity at the population site: r 5 20.99, P 5
0.017.
monal responses. For example, piercing and sucking
insects inflict much less tissue damage than chewing
insects, and elicit SA-triggered pathogen-like systemic
acquired resistance rather than JA-induced antiherbi-
vore defenses (Walling 2000). Leafminers elicit JA as
do other leaf chewers, but since leafminer feeding is
internal, the response may be localized rather than sys-
temic. This would create a JA gradient, with the highest
concentrations contained in the insects themselves, as
we observed (Fig. 6A). We were surprised and in-
trigued to discover JA and SA in leafminer pupae (Fig.
6). To our knowledge, these plant-signaling molecules
have not yet been reported in insects. Larvae evacuate
their guts prior to pupation, thus the JA and SA we
detected in pupae was not from undigested leaf ma-
terial, but was probably sequestered or incorporated
into insect cells or tissue. Interestingly, in the same
GC/MS analysis (not shown), auxin (indole-3-acetic
acid) and abscisic acid (ABA), were not detected in
leafminer pupal tissues, although they occurred in iris

tions are increasingly blurred as we learn about their

Special Feature

molecular biology (Walling 2000, Schultz and Appel

2004), and new models portray plant stress signals as
multidimensional interactive networks (Knight and
Knight 2001). Clearly, these hormones have interactive
roles in plant growth, ontogeny, and in shaping and
directing plant responses to abiotic and biotic chal-
lenges (Cipollini 2004, Farnsworth 2004, Hatcher et
al. 2004, Voesenek et al. 2004). Our work with I. hex-
agona demonstrates that strong temporal and spatial
plasticity in hormone concentrations occur, and in lab-
oratory and common garden experiments, hormones
varied significantly across brief and extended time pe-
riods (Figs. 1 and 2).
Time course studies reveal short-term JA inductions
and rapid declines in Nicotiana, Arabidopsis, and other
species in response to wounding or herbivory (Mc-
Cloud and Baldwin 1997, Niki et al. 1998, Kessler and
Baldwin 2002). We observed a sharp decline of JA from
spring to fall in common garden irises that were not
exposed to experimental stress and herbivory. This il-
lustrates the inherent variation that occurs during nor-
mal phenological or developmental changes, unrelated
to salinity or leafminers. Similar temporal variation in
JA and SA occurred in reproductive tissues (Fig. 4 in
Wang et al. 2001), but other hormonal signals produced
different patterns. For example, abscisic acid (ABA)
declined gradually from April to August in common
garden I. hexagona, then abruptly increased in Septem-
ber (Fig. 1 in Wang et al. 2001). This natural lability
exceeds herbivore- and stress-induced responses. More FIG. 6. Concentration of (A) jasmonic acid and (B) sal-
long-term studies are needed so we can compare hor- icylic acid in common garden I. hexagona plants growing in
freshwater without leafminers (control), in freshwater with
monal signals under stressful and benign conditions, leafminers (mined leaf), and in leafminer pupae. Bars show
and distinguish stress responses from normal physio- means 6 1 SE; letters above indicate significant (P # 0.05)
logical cycles. differences based on Tukey adjusted pairwise comparisons.
January 2004 PHYTOHORMONAL ECOLOGY
TABLE 2. Effect of salinity and leafminer herbivory on endogenous hormones in Iris hexagona
and their presence or absence in Cerodontha iridiphora pupal tissue.
Jasmonic
Parameter acid
Foliate
Salinity stress present
Leafminer herbivory present
Pupae present
leaves (Y. Wang, S. Mopper, and K. Hasenstein, un-
published data).
The retention of phytohormones in leafminer pupae
is a novel finding. It may have no adaptive significance,
and merely indicate insect-specific metabolism and se-
questration of plant hormones. Alternatively, leafmi-
ners may have evolved a mechanism to utilize phyto-
hormones. Schultz and Appel (2004) discussed the evo-
lutionary significance of such cross-kingdom sharing
of (plant) signaling systems. One proposed adaptive
function that may apply to leafminers is the suppression
of host plant defense responses. Empirical support for
this theory is presented in a recent study (Li et al.
2002). Helicoverpa zea feeding on celery (Apium grav-
eolens) used plant hormones to activate cytochrome
P450 enzymes that subsequently detoxified the plant
allelochemicals induced by herbivory. In these exper-
iments, induction of the counter-defense P450 enzymes
specifically required JA and SA, and was not triggered
by closely related molecules. The retention of JA and
SA in leafminers may indicate a similar ability to use
plant hormones. The Li et al. (2002) study did not
measure JA or SA in Helicoverpa zea herbivores, but
we can speculate that insects acquire sufficient con-
centrations of these hormones to trigger expression of
detoxification enzymes.
Little is currently understood about these reciprocal
response strategies (eavesdropping) to counteract
plant defenses (Schultz and Appel 2004), but it is an
emerging and exciting field in hormonal ecology. The
retention of plant hormones by insects could also en-
hance protection against enemies beside plants. Sali-
cylic acid is an important component of signal trans-
duction in antifungal defenses of both plants and mam-
mals (Pierpoint 1994) and leafminers are highly resis-
tant to fungal attack (Connor and Taverner 1997). The
potential use of SA by insects to resist pathogens would
be another example of different kingdoms using the
same plant compounds for defensive purposes (Schultz
and Appel 2004).
Adaptation to abiotic stress and/or natural seasonal
patterns in phytohormone production can have con-
sequences for herbivores. The relationship between SA
and salinity stress has not been clearly established, but
the decline of SA in salt-stressed iris is consistent with
an adaptive response. In tomato (Lycopersicon escu-
lentum), a decline in SA occurred in salinity-adapted
plant cells compared to nonadapted cells (Molina et al.
45
Salicylic Abscisic
acid acid Auxin
absent present absent
absent unknown unknown
present absent absent
2002), and SA can block the expression of genes as-
sociated with salinity protection (Chao et al. 1999).
However, supplemental SA reduced NaCl damage to
cells in the ice plant, Mesembryanthemum crystallinum
(Yen and Yen 1999). Salinity induction of JA in iris
foliage may increase resistance to herbivory, but the
reduced SA could hinder plant defenses against path-
ogen attack. Plant natural enemies may adaptively syn-
chronize to hormone phenological cycles to exploit the
shift in concentration. For example, the greatest leaf-
mining activity occurs in June, when foliage JA is low
relative to April (Fig. 6). Leaves begin senescing in
July (L. Schile and S. Mopper, unpublished data), and
a narrow window exists when low foliar JA coincides
Special Feature
with availability of leaves that are sufficiently healthy
to support leaf miners.
Although they operate through separate pathways,
both JA and SA were affected by salinity and herbiv-
ory, which caused similar patterns of induction and
suppression (Table 2). The JA elicited by salinity stress
(Fig. 1) and herbivory (Fig. 6), and the defense re-
sponses that are known to ensue (Thaler et al. 2001)
may provide a mechanism for poor leafminer perfor-
mance on salt-stressed I. hexagona (Figs. 4 and 5),
despite the greater foliage nitrogen (Fig. 3). It remains
to be tested if simultaneous salinity and herbivory
would have additive effects (Knight and Knight 2001).
When different challenges elicit similar signals
(Schultz and Appel 2004), as is the case with salinity
and leafminer herbivory, the cumulative effect could
reduce energetic costs and enhance plant recovery from
different types of stress. As the plant responds physi-
cally to one stress factor, it may increase tolerance for
another. However, when signals elicit opposite respons-
es, the strongest or earliest signal prevails and plant
defense is compromised (Niki et al. 1998, Stotz 2002,
Thaler and Bostock 2004). These crosstalk interactions
are determined by environmental variation and must
be studied in biologically realistic settings.
ACKNOWLEDGMENTS
We thank Anurag Agrawal for his editorial advice, which
improved the manuscript. Two anonymous reviewers had
helpful suggestions. Elizabeth Milligan, Lisa Schile, Jennifer
Singleton, Josh Smith, and Pete Van Zandt provided valuable
field and laboratory assistance. We could not have conducted
this research without the assistance of Edmond Mouton and
Scott Schales, biologists/airboat pilots with the Louisiana De-
partment of Wildlife and Fisheries. This research was sup-
 46 SUSAN MOPPER ET AL.
ported by The National Science Foundation grants DEB-
9632302 and DEB-0124901.
LITERATURE CITED
Agrawal, A. A., S. Tuzun, and E. Bent. 1999. Induced plant
defenses against pathogens and herbivoresbiochemistry,
ecology, and agriculture. American Phytopathological So-
ciety Press, St. Paul, Minnesota, USA.
Baldwin, I. T. 1998. Jasmonate-induced responses are costly
but benefit plants under attack in native populations. Pro-
ceedings of the National Academy of Science (USA) 95:
81138118.
Berger, S. 2002. Jasmonate-related mutants of Arabidopsis
as tools for studying stress signaling. Planta 214:497504.
Bergey, D. R., G. A. Howe, and C. A. Ryan. 1996. Poly-
peptide signaling for plant defensive genes exhibits anal-
ogies to defense signaling in animals. Proceedings of the
National Academy of Science (USA) 93:1205312058.
Blechert, S., W. Brodschelm, S. Holder, L. Kammerer, T. M.
Kutchan, M. J. Mueller, Z. H. Xia, and M. H. Zenk. 1995.
The octadecanoic pathway: signal molecules for the reg-
ulation of secondary pathways. Proceedings of the National
Academy of Science (USA) 92:40994105.
Bostock, R. M., R. Karban, J. S. Thaler, P. D. Weyman, and
D. Gilchrist. 2001. Signal interactions in induced resis-
tance to pathogens and insect herbivores. European Journal
of Plant Pathology 107:103111.
Chao, W. S., Y. Q. Gu, V. Pautot, E. A. Bray, and L. L.
Walling. 1999. Leucine aminopeptidase RNAs, proteins,
Special Feature
and activities increase in response to water deficit, salinity,
and the wound signals systemin, methyl jasmonate, and
abscisic acid. Plant Physiology 120:979992.
Cippolini, D. 2004. Stretching the limits of plasticity: Can a
plant defend against both competitors and herbivores?
Ecology 85:2837.
Coleman, J. M. 1988. Dynamic changes and processes in the
Mississippi River delta. Geological Society of America
Bulletin 100:9991015.
Connor, E. F., and M. P. Taverner. 1997. The evolution and
adaptive significance of the leaf-mining habit. Oikos 79:
625.
Creelman, R. A., and J. E. Mullet. 1995. Jasmonic acid dis-
tribution and action in plants: regulation during develop-
ment and in response to biotic and abiotic stress. Proceed-
ings of the National Academy of Science (USA) 92:4114
4119.
Creelman, R. A., and J. E. Mullet. 1997. Biosynthesis and
action of jasmonates in plants. Annual Review Plant Phys-
iology Plant Molecular Biology 48:355381.
Farmer, E. E., R. R. Johnson, and C. A. Ryan. 1992. Reg-
ulation of expression of proteinase inhibitor genes by meth-
yl jasmonate and jasmonic acid. Plant Physiology 98:995
1002.
Farnsworth, E. 2004. Hormones and shifting ecology
throughout plant development. Ecology 85:515.
Gupta, V., M. G. Willits, and J. Glazebrook. 2000. Arabi-
dopsis thaliana EDS4 contributes to salicylic acid (SA)-
dependent expression of defense responses: evidence for
inhibition of jasmonic acid signaling by SA. Molecular
PlantMicrobe Interactions 13:503511.
Hammerschmidt, R., and J. A. Smith-Becker. 1999. The role
of salicylic acid in disease resistance. Pages 3753 in A.
A. Agrawal, S. Tuzun, and E. Bent, editors. Induced plant
defenses against pathogens and herbivoresbiochemistry,
ecology, and agriculture. American Phytopathological So-
ciety Press, St. Paul, Minnesota, USA.
Harms, K., I. Ramirez, and H. PenaCortes. 1998. Inhibition
of wound-induced accumulation of allene oxide synthase
transcripts in flax leaves by aspirin and salicylic acid. Plant
Physiology 118:10571065.
Ecology, Vol. 85, No. 1
Hasegawa, P. M., R. A. Bressan, J. K. Zhu, and H. J. Bohnert.
2000. Plant cellular and molecular responses to high sa-
linity. Annual Review of Plant Physiology and Plant Mo-
lecular Biology 51:463499.
Hatcher, P. E., J. Moore, J. E. Taylor, G. W. Tinney, and N.
D. Paul. 2004. Phytohormones and plantherbivorepath-
ogen interactions: integrating the molecular with the eco-
logical. Ecology 85:5969.
Howard, R. J., and I. A. Mendelssohn. 1999. Salinity as a
constraint on growth of oligohaline marsh macrophytes. I.
Species variation in stress tolerance. American Journal of
Botany 86:785794.
Hoyos, M. E., and S. Q. Zhang. 2000. Calcium-independent
activation of salicylic acid-induced protein kinase and a
40-kilodalton protein kinase by hyperosmotic stress. Plant
Physiology 122:13551363.
Iuchi, S., K. YamaguchiShinozaki, T. Urao, and K. Shinozaki.
1996. Characterization of two cDNAs for novel drought-
inducible genes in the highly drought-tolerant cowpea.
Journal of Plant Research 109:415424.
Karban, R., and I. T. Baldwin. 1997. Induced response to
herbivory. The University of Chicago Press, Chicago, Il-
linois, USA.
Kessler, A., and I. T. Baldwin. 2002. Plant responses to insect
herbivory: the emerging molecular analysis. Annual Re-
view of Plant Biology 53:299328.
Klessig, D. F., and J. Malamy. 1994. The salicylic acid signal
in plants. Plant Molecular Biology 26:14391458.
Knight, H., and M. Knight. 2001. Abiotic stress signalling
pathways: specificity and cross-talk. Trends in Plant Sci-
ence 6:262267.
Levitt, J. 1980. Responses of plants to environmental stress-
es. Volume II. Academic Press, New York, New York,
USA.
Li, X., M. A. Schuler, and M. R. Berenbaum. 2002. Jasmonate
and salicylate induce expression of herbivore cytochrome
P450 genes. Nature 419:712715.
McCloud, E. S., and I. T. Baldwin. 1997. Herbivory and
caterpillar regurgitant amplify the wound-induced increas-
es in jasmonic acid but not nicotine in Nicotiana sylvestris.
Planta 203:430435.
Molina, A., P. Bueno, M. C. Marin, M. P. Rodrigues-Rosales,
A. Belver, K. Venema, and J. P. Donair. 2002. Involvement
of endogenous salicylic acid content, lipoxygenase and an-
tioxidant enzyme activities in the response of tomato cell
suspension cultures to NaCl. New Phytologist 156:409
415.
Moons, A., E. Prisen, G. Bauw, and M. V. Montagu. 1997.
Antagonistic effects of abscisic acid and jasmonates on salt
stress-inducible transcripts in rice roots. Plant Cell 92:243
259.
Mopper, S., P. Stiling, K. Landau, D. Simberloff, and P. Van
Zandt. 2000. Spatiotemporal variation in leafminer pop-
ulation structure and adaptation to individual oak trees.
Ecology 81:15771587.
Niki, T., I. Mitsuhara, S. Seo, N. Ohtsubo, and Y. S. Ohashi.
1998. Antagonistic effect of salicylic acid and jasmonic
acid on the expression of pathogenesis-related (PR) protein
genes in wounded mature tobacco leaves. Plant Cell Phys-
iology 39:500507.
Orton, E. W. 1959. A geological study of Marsh Island, Iberia
Parish, Louisiana. Technical Report of the Louisiana Wild-
life and Fisheries Commission, New Orleans, Louisiana,
USA.
Pezeshki, S. R., R. D. DeLaune, and W. H. Patrick, Jr. 1990.
Flooding and saltwater intrusion: potential effects on sur-
vival and productivity of wetland forests along the US Gulf
Coast. Coastal Forest Ecology Management 33/34:287
301.
January 2004 PHYTOHORMONAL ECOLOGY
Pierpoint, W. S. 1994. Salicylic acid and its derivatives in
plants: medicines, metabolites, and messenger molecules.
Advances in Botanical Research 20:163235.
Pott, M. B., E. Pichersky, and B. Piechulla. 2002. Evening
specific oscillations of scent emission, SAMT enzyme ac-
tivity, and SAMT mRNA in flowers of Stephanotis flori-
bunda. Journal of Plant Physiology 159:925934.
Raguso, R. A., and D. M. Light. 1998. Electroantennogram
responses of male Sphinx perelegans hawkmoths to floral
and green-leaf volatiles. Entomologia Experimentalis et
Applicata 86:287293.
Raskin, I. 1992. Roles of salicylic acid in plants. Annual
Review Plant Physiology Plant Molecular Biology 43:439
463.
Rogers, C. E., and J. P. McCarty. 2000. Climate change and
ecosystems of the mid-Atlantic region. Climate Research
14:235244.
SAS Institute. 1999. SAS Version 8. SAS Institute, Cary,
North Carolina, USA.
Schultz, J. C., and H. M. Appel. 2004. Cross-kingdom cross-
talk: hormones shared by plants and their insect herbivores.
Ecology 85:7077.
Sembdner, G., and B. Parthier. 1993. The biochemistry and
physiology and molecular actions of jasmonates. Annual
Review Plant Physiology Plant Molecular Biology 44:569
586.
Stotz, H. U., T. Koch, A. Biedermann, K. Weniger, W. Boland,
and T. Mitchell-Olds. 2002. Evidence for regulation of re-
sistance in Arabidopsis to Egyptian cotton worm by sali-
cylic and jasmonic acid signaling pathways. Planta 214:
648652.
Stout, M. J., and R. M. Bostock. 1999. Specificity of induced
responses to arthropods and pathogens. Pages 183209 in
A. A. Agrawal, S. Tuzun, and E. Bent, editors. Induced
plant defenses against pathogens and herbivores: biochem-
istry, ecology and agriculture. American Phytopathological
Society Press, St. Paul, Minnesota, USA.
Thaler, J. S. 1999. Jasmonate-inducible plant defences cause
increased parasitism of herbivores. Nature 399:686688.
Thaler, J. S., and R. M. Bostock. 2004. Interactions between
abscisic-acid-mediated responses and plant resistance to
pathogens and insects. Ecology 85:4858.
Thaler, J. S., A. L. Fidantsef, S. S. Duffy, and R. M. Bostock.
1999. Trade-offs in plant defense against pathogens and
herbivores: a field demonstration of chemical elicitors of
47
induced resistance. Journal of Chemical Ecology 25:1597
1609.
Thaler, J. S., M. J. Stout, and R. Karban. 2001. Jasmonate-
mediated induced plant resistance affects a community of
herbivores. Ecological Entomology 26:312324.
Tornqvist, T. E., T. R. Kidder, W. J. Autin, K. van der Borg,
A. F. M. de Jong, C. J. W. Klerks, E. M. A. Snijders, J. E.
A. Storms, R. L. van Dam, and M. C. Wiemann. 1996. A
revised chronology for the Mississippi River. Science 273:
16931696.
Tsonev, T. D., G. N. Lazova, Z. G. Stoinova, and L. P. Popova.
1998. A possible role for jasmonic acid in adaptation of
barley seedlings to salinity stress. Journal of Plant Growth
Regulation 17:153159.
Van Poecke, R. M. P., M. A. Posthumus, and M. Dicke. 2001.
Herbivore-induced volatile production by Arabidopsis thal-
iana leads to attraction of the parisitoid Cotesia rubecula:
chemical, behavioral, and gene-expression analysis. Jour-
nal of Chemical Ecology 27:19111928.
Van Zandt, P. A., and S. Mopper. 2002. Delayed and carry-
over effects of salinity on flowering in Iris hexagona (Ir-
idaceae). American Journal of Botany 89:364383.
Van Zandt, P. A., M. A. Tobler, E. Mouton, K. H. Hasenstein,
and S. Mopper. 2003. Positive and negative consequences
of salinity stress for the growth and reproduction of the
clonal plant, Iris hexagona. Journal of Ecology, in press.
Voesenek, L. A. C. J., J. H. G. M. Rindjers, A. J. M. Peters,
H. M. van de Steeg, and H. de Kroon. 2004. Plant hormones
regulate fast shoot elongation under water: from genes to
communities. Ecology 85:1627.
Special Feature
Walling, L. L. 2000. The myriad plant responses to herbi-
vores. Journal of Plant Growth Regulation 19:195216.
Wang, Y., S. Mopper, and K. H. Hasenstein. 2001. Effects
of salinity on endogenous levels of ABA, IAA, JA, and
SA in Iris hexagona. Journal of Chemical Ecology 27:325
340.
Weyers, J. D. B., and N. W. Paterson. 2001. Plant hormones
and the control of physiological processes. New Phytologist
152:375407.
Yen, H. C. E., and S. K. Yen. 1999. Effect of high salinity
on tracheal element differentiation in light-grown callus of
Mesembryanthemum crystallinum. Plant Cell Tissue and
Organ Culture 58:5965.
Yeo, A. R. 1999. Predicting the interaction between the ef-
fects of salinity and climate change on crop plants. Scientia
Horticulturae 78:159174.
 Ecology, 85(1), 2004, pp. 4858
q 2004 by the Ecological Society of America

INTERACTIONS BETWEEN ABSCISIC-ACID-MEDIATED RESPONSES AND

PLANT RESISTANCE TO PATHOGENS AND INSECTS
JENNIFER S. THALER1,3 AND RICHARD M. BOSTOCK2
125 Willcocks Street, Department of Botany, University of Toronto, Toronto, Ontario, Canada M5S 3B2
2Department of Plant Pathology, University of California, One Shields Avenue, Davis, California 95616 USA

Abstract. One goal of phytohormonal ecology is to study the interactions between

biotic and abiotic stress at hierarchical levels of biological organization. From an ecological
perspective, exposure to one stress may alter the plants probability of being exposed to
another stress. From a mechanistic perspective, hormonal and biochemical signaling in-
teractions between responses to each stress may influence the severity or ability to adaptively
respond to the subsequent stress. In this article, we consider the relationship between plant
water and salt stress and attack by pathogens and herbivores. Empirical data suggest that
water stress and the probability of attack by pathogens and herbivores are correlated between
habitats. Biochemical interactions between plant responses to water and salt stress and
insect and pathogen attack are also interrelated. Initial biochemical models indicated that
abscisic acid (ABA), an important hormone in responses to water and salt stress, had a
synergistic positive role with jasmonate-induced defenses against herbivores and an an-
tagonistic role with salicylate-based resistance to some pathogens. Based on this back-
ground, we developed predictions about how water and salt stress would alter plant resis-
tance to insects and pathogens and tested the predictions using tomato plants as a model
system. We used polyphenol oxidase activity as a marker of the jasmonate response and
pathogenesis-related protein P4 as a marker of the salicylate response. First, we examined
Special Feature

levels of chemical defense in wild-type and ABA-deficient plants and the ability of these
plants to resist insect and pathogen attack. In the second experiment, we exposed plants
to short-term salinity stress and tested their subsequent resistance to a chewing insect
Spodoptera exigua and the bacterial speck pathogen Pseudomonas syringae pv. tomato. We
have two key findings. First, ABA-deficient plants had higher levels of salicylate-mediated
responses and were more resistant to bacterial speck disease, consistent with the proposed
role of salicylate in defense against pathogens. This suggests linkage between water avail-
ability to the plant and salicylate action in pathogenesis through ABA signaling. ABA-
deficient plants had reduced resistance to the insect Spodoptera exigua, suggesting a positive
correlation between responses to water stress and herbivory. The lack of difference in
chemical expression of the jasmonate (JA) response (polyphenol oxidase activity) between
wild-type and ABA-deficient plants did not support the proposed mechanism of synergism
with the jasmonate response. Second, salt stress reduced the chemical induction (e.g.,
pathogenesis-related protein P4) of the salicylate response, but this did not affect resistance
to the pathogen. Salt stress did not alter resistance to the herbivore Trichoplusia ni, but
did alter the negative signal interaction between the jasmonate and salicylate responses.
Under control conditions, the jasmonate and salicylate responses are antagonistic to one
another, with induction of one response reducing the inducibility of the other. Under salt
stress conditions, the negative effect of salicylate on the jasmonate response was reduced.
Thus, complex interactions occur between ABA, JA, and SA, hormones that are important
regulators of abiotic and biotic stress responses. Phytohormonal ecology is attempting to
link ecological and hormonal interactions to develop a predictive framework for how and
why plants coordinate responses to the environment.
Key words: abscisic acid; induced responses; jasmonic acid; phytohormonal ecology; salicylic
acid; tomato.

INTRODUCTION plant responses to environmental stress. This includes

A major goal of phytohormonal ecology is to un-
derstand the role of plant hormones in determining
Manuscript received 12 November 2002; revised and accepted
21 April 2003. Corresponding Editor (ad hoc): S. Mopper. For
reprints of this Special Feature, see footnote 1, p. 3.
3 E-mail: thaler@botany.utoronto.ca
48
the role that each hormone plays in responding to in-
dividual stresses as well as the coordination of multiple
hormones in responses to multiple stresses. One par-
ticular challenge is to unite the knowledge of the bio-
chemistry of hormones with the patterns of ecological
correlations between stresses. Why does phytohor-
monal ecology aim to study the interactions between
January 2004 PHYTOHORMONAL ECOLOGY
plant responses to multiple stresses? First, exposure to
one stress may alter a plants vulnerability to others
(e.g., Garcia-Guzman and Dirzo 2001). Second, hor-
monal and biochemical signaling interactions between
responses to each stress may influence the severity or
the ability of plants to adaptively respond to the second
stress (Chapin 1991, Genoud and Metraux 1999, Bos-
tock et al. 2001, de Bruxelles and Roberts 2001).
Patterns and correlations between plant stresses
Some biotic and abiotic stresses covary in nature.
For example, observational and experimental data sug-
gest positive environmental correlations between water
availability and the probability of attack by plant path-
ogens (Agrios 1997). This may be because water plays
a major role in the movement, germination, and leaf
entry of many pathogens. For example, tomato plants
had higher levels of infection by Septoria when they
were exposed to rainfall compared to when they were
protected (Parker et al. 1995). Water availability may
also influence plant osmotic stress, which may in turn
affect success of attack by pathogens. Plant osmotic
stress may directly alter the microenvironment of the
plant (humidity, temperature), which may be particu-
larly important for pathogens invading through the leaf
surface.
Many researchers have investigated the relationship
between plant stress and herbivory. This research led
to the plant stress hypothesis that states that herbivores
should have increased performance on mildly stressed
plants (White 1974). The logic of this argument is that
stressed plants invest less in defense and may be of
higher nutritional value. This hypothesis was tested in
a metanalysis of 70 experiments that manipulated abi-
otic stress, and the results showed support for the plant
stress hypothesis for some feeding guilds of insects but
not others (Koricheva et al. 1998). In particular studies,
both of these patterns have been found, depending on
the insect species and the defensive compounds in-
volved (Gershenzon 1984, Larsson 1989, Waring and
Cobb 1992). For instance, English-Loeb et al. (1997)
demonstrated that increasing drought in tomato plants
had nonlinear effects on plant chemistry and resistance
to herbivores. There is also evidence for an ecological
association between moisture and herbivory. Chase et
al. (2000) found a positive association between rainfall
and herbivory in different habitats. This is probably in
large part due to the positive effect of rainfall on plant
productivity, but nevertheless generates a large-scale
environmental correlation between water availability
and levels of damage. Similar patterns have been found
for the correlation between net primary productivity
and levels of herbivory (Cyr and Pace 1993); net pri-
mary productivity being positively correlated with pre-
cipitation. If this correlation is real and common, plants
experiencing water stress may be resistant to insects.
There may be direct effects of the plant microen-
vironment on insect performance. Stomatal closure fol-
49
lowing water stress causes plants to reduce transpira-
tion, which in turn can alter the leaf environment for
insects (Holtzer et al. 1988). This reduction in humidity
reduces hatching of lepidopteran eggs (Godfray and
Holtzer 1991), spider mite population growth (Perring
et al. 1984), and growth of many species of insect
larvae (Mattson and Scriber 1987). Insect wounding
itself can also cause cellular damage that causes leaf
dehydration, which alters the leaf environment and
causes plant dehydration responses (Reymond et al.
2000). Thus, abundant evidence suggests that abiotic
stresses, such as water availability, may covary with
plant attackers. Most plants in natural and managed
systems are likely faced with water stress as well as
attack by pathogens and herbivores. Thus, the ecolog-
ical correlations between stresses can be used as a
framework to examine phytohormonal interactions that
regulate plant responses to the respective stresses.
Phytohormonal models of plant stress interactions
When plants experience stress, activation of highly
phylogenetically conserved phytohormonal pathways
are thought to mediate adaptive responses. Current bio-
chemical and molecular models of plant hormonal sig-
Special Feature
naling outline three phytohormones and their signaling
pathways as predominantly responsible for plant resis-
tance to water and salt stress (abscisic acid, ABA; Wal-
ton 1980, Zeevaart and Creelman 1988), pathogen at-
tack (salicylic acid, SA; Delaney et al. 1994), and insect
attack (jasmonic acid, JA; Creelman and Mullet 1997).
These lines of distinction are not absolute and there is
a high degree of interaction among these three hor-
mones. Recent work using microarrays to analyze ex-
pression of many ABA-, JA-, and SA-regulated genes
emphasizes the interrelatedness of responses mediated
by these three hormones (Reymond et al. 2000,
Schenck et al. 2000). For instance, abscisic acid is in-
volved in an important portion, but not all, of the plant
responses to salt stress (Grillo et al. 1995, Wei et al.
2000). Jasmonate has been shown to be involved in
responses to salt stress in rice roots in a manner an-
tagonistic with the ABA-regulated responses to salt
stress (Moons et al. 1997, see also Wasternack and
Parthier 1997). On the other hand, salicylic acid can
interfere with regulation of genes by salt stress and
ABA (Chao et al. 1999) and disrupt the maintenance
of plant water balance. In a study by Barkowsky and
Einhelling (1993), exogenous salicylic acid interfered
with transpiration and maintenance of water potential
and caused a decrease in soybean seedling growth.
There can also be negative interactions between the SA
and JA signaling pathways, such that activation of one
of these pathways results in a reduced ability to induce
the other pathway (Felton et al. 1999, Thaler et al. 1999,
Paul et al. 2000). Although, the jasmonate response
does provide resistance to some pathogens including
Pseudomonas syringae (Thomma et al. 1998, Stout et
al. 1999). Ethylene, another important hormone in
 50 J. S. THALER AND R. M. BOSTOCK
FIG. 1. Diagram of proposed relationship between abscisic acid, salicylic acid, and jasmonic acid signaling pathways.
Solid arrows indicate a positive regulating relationship between stresses and hormones. Dashed arrows indicate a positive
(1) or negative (2) relationship between the hormonal signaling pathways.
stress responses often correlated with the jasmonate
response, will not be considered here.
Initial biochemical models suggested that ABA had
a synergistic role with jasmonate-mediated induced re-
sponses to herbivory (Pena-Cortes et al. 1989, 1991,
Wasternack and Parthier 1997) (Fig. 1). Whether a bas-
Special Feature
al level of endogenous ABA is all that is required for
the induction of jasmonate induced responses (Birken-
meier and Ryan 1998, Herde et al. 1999), or whether
ABA levels quantitatively affect JA responses, remains
controversial and appears to be dependent on the par-
ticular response (Chao et al. 1999). Water stress, sa-
linity, and ABA treatment all result in up-regulation of
several, but not all, wound-inducible genes (Chao et
al. 1999). Conversely, circumstantial evidence has im-
plicated ABA as an antagonist to salicylate-mediated
induced responses. Increases in endogenous ABA have
been demonstrated following pathogen infection (Ca-
hill and Ward 1989, Kettner and Dorffling 1995); in
some cases the fungus may be producing abscisic acid
itself (Crocoll et al. 1991). Exogenous application of
ABA has frequently been demonstrated to decrease
plant resistance to fungal pathogens (Henfling et al.
1980, Li and Heath 1990, McDonald and Cahill 1999),
but has been also been shown to cause decreased sus-
ceptibility in other plantfungus and plantvirus inter-
actions (Whenham et al. 1986, Ward et al. 1989). Fur-
ther evidence for the role of ABA in resistance to path-
ogens is that drought and salt stress have been shown
to increase the severity of diseases caused by Phyto-
phthora parasitica infecting tomato plants (Ristaino
and Duniway 1989). Predisposition to disease can oc-
cur as a result of relatively minor water stress that the
plant would recover from in the absence of the path-
ogen (Ristaino et al. 1988). Moreover, there is evidence
that this predisposition results from suppressed plant
resistance (Swiecki and MacDonald 1988). Recent ev-
idence demonstrates that ABA interferes with salicy-
late-mediated resistance to Botrytis cinerea in tomato
plants (Audenaert et al. 2002).
Ecology, Vol. 85, No. 1
This information yields two predictions about how
ABA-mediated responses to drought and salt stress may
influence salicylate- and jasmonate-mediated resis-
tance to pathogens and insects. First, ABA-mediated
responses may interact with jasmonate responses re-
sulting in an increase in plant resistance to insects.
Second, ABA-mediated responses may interfere with
the deployment of the salicylate response and decrease
resistance to pathogens.
We tested these predictions using two approaches.
First, we used plants with genetically manipulated
ABA levels by using mutant tomato plants that are
deficient in their ability to produce ABA. The mutants
are analogous to an ecological press experiment
(Bender et al. 1984), i.e., the alteration in ABA occurs
for the entire life of the plant. Since ABA is a major
hormone involved in maintaining water balance, mu-
tant plants should be under mild water stress even when
well watered (see Results). The interaction between
hormonal pathways was assessed by assaying wild-type
and ABA-deficient mutant plants for their resistance to
insects and pathogens. Second, we phenotypically ma-
nipulated levels of abscisic acid by subjecting plants
to short-duration salt stress and then measured their
ability to induce jasmonate- and salicylate- mediated
resistance to herbivores and pathogens. These manip-
ulations are analogous to a pulse experiment, where-
by ABA levels are changed for only a short period of
time. Our purpose in these experiments was to impose
a short-term salinity stress on the plants to avoid the
extreme reworking of the plants physiology that can
occur in response to severe water limitation (Mattson
and Haack 1987, Bray 1993).
We employed the phenotypic and genetic approaches
because each has advantages and limitations. Our two
methods of manipulating endogenous ABA levels al-
lowed for independent tests of the relationship between
ABA, JA, and SA signaling. The advantage of using
phenotypic manipulations is that the manipulation can
be conducted in one genetic background and plants can
January 2004 PHYTOHORMONAL ECOLOGY
be grown identically until the treatment is imposed.
Elicitation of the jasmonate and salicylate response us-
ing chemical elicitors is a useful approach because the
treatments are a singular cue, accurate, and repeatable;
in some respects these elicitors mimic actual herbivory
and pathogen attack while in other respects they do not
(Baldwin 1996, Thaler et al. 1996). A limitation of
utilizing a phenotypic manipulation like salt stress is
that, in addition to increasing ABA levels, other factors
change as well, including responses to the toxicity of
the salt (Munns 2002). The advantage of a genetic ma-
nipulation approach is that ABA levels in the plant have
been quantitatively altered (Taylor 1987). A limitation
of this approach is that genetic changes may alter the
nonstress functions of ABA as well as have pleiotropic
effects on other plant processes. Our purpose is not to
directly compare these two approaches, but to manip-
ulate endogenous ABA levels in two ways to assess
the potential roles of ABA in plant resistance to her-
bivores and pathogens. Because water stress and salt
stress can both induce endogenous ABA, these results
will be relevant for projecting the effects of these two
abiotic stresses on plant resistance.
METHODS
General methods
Plant growth.Tomato plants were grown in a
greenhouse in 500-mL pots for ;1 mo, until the four-
leaf stage, and watered as needed to maintain soil mois-
ture and prevent wilting.
Bioassays
Herbivore growth rate bioassay.Newly hatched
Spodoptera exigua or Trichoplusia ni caterpillars were
reared on artificial diet (Southland Products, Lake Vil-
lage, Alaska, USA) for 35 d before the bioassays.
These herbivores have been shown to respond in a sim-
ilar manner to induction of the jasmonate and salicylate
responses (Thaler et al. 2002a, b). Caterpillars were
placed individually on a single lateral leaflet from the
third leaf in 90-mm Petri dishes lined with moist filter
paper. The caterpillars fed for 23 d at room temper-
ature. The caterpillars were weighed at the beginning
and end of the experiment. At the beginning of the
experiment, the caterpillars were in the second instar
and at the end they were in the third instar. The relative
growth rate ([final mass-initial mass]/[initial mass 3
number of days]) (RGR) was calculated (Waldbauer
1968) and used as a measure of herbivore performance.
Bacterial speck disease assay.Pseudomonas syr-
ingae pv. tomato (Pst) isolated from field-grown tomato
plants (isolate Pst23, gift of D. Cooksey, Department
of Plant Pathology, University of California, Riverside,
California, USA) were incubated at 278C for 4872 h
on Kings B medium (King et al. 1954) and colonies
were suspended in sterile water. An aqueous suspension
of 1 3 107 colony-forming units per milliliter was gent-
51
ly painted on the attached terminal leaflet of the fourth
leaf of intact plants with a cotton applicator, and the
plants were incubated in the greenhouse. Seven days
later, the number of lesions on the inoculated leaflet
was counted as a measure of disease. There is a strong
correlation between the number/amount of lesions and
bacterial population density (colony forming units/cm 2)
(C. Richael, personal communication).
Biochemical assays.Levels of the hallmark indi-
cators of the jasmonate and salicylate pathways were
also assayed in the ABA-deficient and wild-type plants.
We measured pathogenesis-related protein 4 mRNA
(P4) as a marker of salicylate-mediated resistance to
pathogens, and polyphenol oxidase (PPO) activity as
a marker for jasmonate-mediated resistance to herbi-
vores. P4 is consistently induced by the salicylate path-
way and is associated with resistance to pathogens in
tomato plants (Fidantsef et al. 1999). Polyphenol ox-
idase activity was chosen as a marker of the jasmonate-
induced response because of its consistent systemic
pattern of induction following herbivore damage and
jasmonic acid spray and because PPO activity corre-
lates well with herbivore performance in tomato plants
(Felton et al. 1989, Orozco-Cardenas et al. 1993, Thaler
Special Feature
et al. 1996).
P4 mRNA levels.Abundance of P4 mRNA was de-
termined by analysis of total RNA (20 mg) prepared
from tomato leaves (Choi et al. 1994) according to the
methods in Fidantsef et al. (1999). Quantitation of hy-
bridization was measured by exposing the probed blots
to a screen that was scanned using a phosphorimaging
system (Storm 860 scan, Molecular Dynamics, Inc.,
Sunnyvale, California, USA). Equal loading of RNA
samples was verified by hybridization of the blots with
a tomato cyclophilin DNA probe or by staining of
rRNAs with ethidium bromide.
Polyphenol oxidase activity.We extracted the en-
zyme from weighed leaflets that were homogenized in
ice-cold buffer and the homogenate was centrifuged to
obtain a clarified extract for enzyme analyses. The su-
pernatant was added to a caffeic acid solution and ab-
sorbance read at 470 nm (Thaler et al. 1996).
Experiments
Genetic manipulation.In the first set of experi-
ments, we examined the levels of chemical defense in
ABA-deficient tomatoes and the ability of these plants
to resist insect and pathogen attack. Two independent
lines of ABA-deficient plants (cv. sitiens and flacca)
were grown in the greenhouse and compared to wild-
type plants (cv. Rheinlands Ruhm). These two single-
gene mutations have been partially characterized and
segregate to separate loci (Neill and Horgan 1985, Mer-
lot and Giraudat 1997). Sitiens plants contain between
8% and 12% and flacca plants contain 1621% of the
ABA present in the upper shoot tissues of five-to-six-
week-old wild-type plants (Tal and Nevo 1973, Taylor
1987). Salt application has been shown to increase the
 52 J. S. THALER AND R. M. BOSTOCK
level of ABA twofold in wild-type tomato plants,
whereas salt does not increase ABA levels in sitiens
or flacca plants (Grillo et al. 1995; R. Bostock, J. Mac-
Donald, and J. Duniway, unpublished data).
Because ABA-deficient plants cannot perform ABA-
regulated functions, they are likely to be under slight
water stress, even when well watered. To determine the
degree of water stress experienced, leaf water potential
was measured using a Scholander pressure chamber
(Soil Moisture Equipment, Santa Barbara, California,
USA) (Tyree and Hammel 1972). The third leaf was
excised with a razor and immediately placed into the
pressure chamber leaving the cut petiole exposed. The
pressure was then slowly increased and the point where
a bubble of liquid on the cut stem became visible was
recorded. Three replicates of each plant type were con-
ducted during the predawn time period (04300530
hours) and five replicates of each plant type were con-
ducted in the midday (14301530 hours). Predawn and
midday water potential results were analyzed using
ANOVA with plant type as the independent variable
followed by adjusted Tukey contrasts between plant
types.
Bioassays and chemical assays were conducted on
Special Feature
the plants when they were in the four-leaf stage. We
assayed levels of disease caused by Pst, level of P4
mRNA, and PPO activity in wild-type and both ABA-
deficient plant lines. In a separate experiment, we ex-
amined S. exigua growth rate and PPO activity on wild-
type and one ABA-deficient line (cv. sitiens). Ten to
twelve replicates per treatment were assayed for the
disease, S. exigua performance and PPO analysis.
Three to four replicates per treatment were assayed for
P4 mRNA levels. P4 mRNA data were log transformed
to improve normality. Results were analyzed using one-
way ANOVA with plant type as the independent var-
iable, followed by a Tukey test to differentiate between
the three plant types.
Phenotypic manipulation.In the second set of ex-
periments, we exposed plants to short-term salinity
stress and tested the effects of this on the ability of the
plant to induce the jasmonate and salicylate response
and on performance of herbivores and pathogens. Wild-
type plants (cv. Bonnie Best) were divided into two
groups of 3640 plants each. One group was given a
short duration (24 h) salt stress, while the other group
was a control group. Salt-stressed plants were watered
with 600 mL of a 0.2 mol/L NaCl/0.02 mol/L CaCl2
solution (11.7 g/kg NaCl, 2.2 g/kg CaCl2) (Chen and
Plant 1999). The control plants were watered with 600
mL of tap water to mimic the flush of water without
the additional salt. The salt-stressed plants wilted im-
mediately but regained turgor pressure within 2 h. After
24 h, we rapidly alleviated the salt stress by flushing
all plants (control and stressed) with excess water five
times. Immediately after alleviating the salt stress,
plants were treated with jasmonic acid or the synthetic
salicylate mimic benzothiodiazole (BTH) to determine
Ecology, Vol. 85, No. 1
whether the salt stress treatment influenced the induc-
ibility of the jasmonate- or salicylate-regulated re-
sponses. BTH and salicylic acid (SA) are known to
induce systemic acquired resistance (SAR) in plants
through the same biochemical signaling pathway (Law-
ton et al. 1996) and therefore we refer to the responses
induced by BTH as SA-mediated induced responses.
Plants in control and salt-stress treatment groups were
subdivided into four elicitor treatments: (1) control, (2)
1.5 mmol/L JA, (3) 1.2 mmol/L BTH and, (4) simul-
taneous 1.5 mmol/L JA and 1.2 BTH (eight treatments,
n 5 810 replicates per treatment). JA and BTH were
applied to the third leaf of plants using hand held at-
omizers. The JA was prepared in aqueous suspension
using acetone to help disperse it in water (1% acetone
in water). BTH was dissolved in water. Control plants
were sprayed with an equal amount of water with 1%
acetone. The solutions/suspensions were sprayed to
runoff onto the desired portion of the plant, shielding
the rest of the leaves.
Two days following elicitor treatment, these plants
were used for chemical assays and bioassays. The ter-
minal leaflet of the fourth leaf was collected for the
PPO assay, one lateral leaflet of the fourth leaf was
collected for the P4 assay, and the other lateral leaflet
of the fourth leaf was collected for the T. ni perfor-
mance experiment. Removing individual leaflets with
a razor does not result in a systemic induced response
(see Agrawal et al. 1999). The attached, terminal leaflet
of the fifth leaf was inoculated with Pst. This experi-
ment was conducted three times for a total of at least
25 replicates in each of the eight treatments. Three-
way ANOVA was used for analyses of PPO activity,
performance of T. ni, and disease caused by Pst with
salt treatment, jasmonate treatment, and BTH treatment
as the independent variables. Each P4 replicate consists
of tissue pooled from four plants; however, we only
measured a single replicate per treatment. We used one-
way ANOVA to test for an effect of salt stress on P4
levels (by comparing the four elicitor treatments in the
non-stressed control plants with the four elicitor treat-
ments in the salt-stressed plants). Comparisons be-
tween the particular elicitor treatments are included
simply as a guide to see if the mean responses were
consistent with our hypotheses.
RESULTS AND INTERPRETATION
Although there was no visual difference between the
wild-type and ABA-deficient plants in terms of leaf or
stem turgor under our growth conditions, ABA-defi-
cient plants were under some water stress. They had
3050% lower water potential compared to wild-type
plants in both the predawn and midday measurements
(predawn, F2,7 5 13.4, P 5 0.004; midday, F2,12 5 14.4,
P 5 0.001; sitiens and flacca were both lower than
wild-type [P , 0.01] and no difference was found be-
tween sitiens and flacca [P . 0.5]; Fig. 2).
January 2004 PHYTOHORMONAL ECOLOGY
FIG. 2. Leaf water potential in wild-type (Rhinelands Rhum) and ABA-deficient plants (sitiens, flacca). Water potential
was measured on excised leaves of 4-wk plants predawn as well as during midday. Values are mean 6 1 SE . Different letters
above bars indicate significant differences between treatments based on Tukey contrasts.
Genetic manipulation
Levels of pathogenesis-related proteins were four to
10 times higher in ABA-deficient plants compared to
wild-type plants (F2,8 5 5.2, P 5 0.011; sitiens vs. wild-
type, P 5 0.008; flacca vs. wild-type, P 5 0.21; Fig.
3a). There was no difference in the activity of PPO in
wild-type and ABA-deficient plants (plant, F2,82 5
0.329, P 5 0.720; trial, F2,82 5 5.0, P 5 0.009; plant
3 trial, F4,76 5 1.8, P 5 0.143; Fig. 3b); the significant
trial effect simply represents differences between the
mean PPO expression in the three trials. These results
are consistent with the prediction that ABA is antag-
onistic to the deployment of the salicylate response,
but inconsistent with the predicted synergism with the
jasmonate response. ABA-deficient mutants had lower
levels of disease (plant, F2,29 5 4.3, P 5 0.024; sitiens
vs. wild-type, P 5 0.35; flacca vs. wild-type, P 5
0.013; sitiens vs. flacca, P 5 0.38; Fig. 3c); but sup-
ported higher growth rates of caterpillars (F1,22 5 4.9,
P 5 0.038; Fig. 3d) in comparison to wild-type plants.
This is consistent with the prediction that ABA inter-
feres with resistance to disease and interacts (or is re-
quired at basal levels) with resistance to insects. Taking
the biochemical assays with the bioassays, we find sup-
port for the prediction that elevated ABA results in
interference with SA-mediated resistance to the disease
Pst. However, even though insect growth was reduced
on the ABA-deficient plants, our measure of the jas-
monate response was unaffected.
Phenotypic manipulation
Overall, the salt stress lowered the levels of P4
mRNA, a response mediated by the SA pathway, re-
53
Special Feature
gardless of elicitation of plants by JA or BTH (mean
counts across salt treatments 6 1 SE: control, 108.3 6
14.76; salt stressed, 62.4 6 6.88; F1,6 5 7.9, P 5 0.03).
BTH (an SA mimic) induced P4 in control plants but
not in salt-stressed plants (Fig. 4a). The salt-stress
treatment did not alter resistance to the pathogen Pst
(Fig. 4b, Table 1). JA and BTH elicitation reduced Pst
lesion formation on salt stress and control plants. So,
overall, the salt-stress manipulation attenuates the SA
response, although this did not affect resistance to Pst.
By itself, salt stress marginally increased the relative
growth rate of T. ni caterpillars (P 5 0.06; Fig 4c,
Table 2). On control plants, caterpillar growth rate was
decreased 50% by jasmonate induction (P , 0.001)
and increased almost 50% by SA induction (P 5
0.024). On control plants, herbivores in the BTH and
JA treatment grew three times the amount of herbivores
feeding on plants treated only with JA, indicating that
BTH attenuated the jasmonate response when plants
were treated with both elicitors. On salt stressed plants,
JA decreased and BTH increased caterpillar growth
rate. However, on the salt-stressed plants, induction of
the SA- response did not attenuate the jasmonate re-
sponse (salt 3 JA interaction, P 5 0.045). While not
exactly what we predicted, this result is consistent with
our prediction that ABA antagonizes the SA response
and is required for the jasmonate response. Following
salt stress, the jasmonatesalicylate biochemical inter-
action may be weighted in favor of expression of the
jasmonate response.
DISCUSSION
We originally predicted that ABA would enhance
resistance to herbivores but diminish resistance to path-
 54 J. S. THALER AND R. M. BOSTOCK
Special Feature
FIG. 3. Chemical and biological assays of resistance in
wild-type and ABA-deficient (sitiens and flacca) plants: (a)
pathogenesis-related protein 4 (P4) mRNA levels; (b) poly-
phenol oxidase (PPO) activity; (c) number of lesions caused
by the bacterial pathogen Pseudomonas syringae pv. tomato;
(d) relative growth rate of Spodoptera exigua caterpillars
([final mass 2 initial mass]/[initial mass 3 no. days]). Values
are means 6 1 SE. Letters above bars indicate significant
differences between treatments based on Tukey contrasts.
ogens. We conclude that ABA plays an important role
in resistance to insects and pathogens but not neces-
sarily through the originally proposed biochemical
mechanisms. ABA-deficient plants produced more PR
proteins and had decreased levels of disease than wild-
type plants, consistent with the hypothesis that ABA
antagonizes the SA response pathway. This result is
similar to that of Audenaert et al. (2002), who dem-
onstrated increased resistance of ABA-deficient tomato
plants to Botrytis, a necrotrophic fungal pathogen. The
influence of ABA on resistance to insects was more
complicated. Consistent with the hypothesis that ABA
positively affects JA-mediated responses; insect per-
formance, as measured by growth rate, was higher on
the ABA-deficient plants. However, our marker of the
jasmonate response, polyphenol oxidase (PPO), was
not different between wild-type and ABA-deficient
plants. It is possible that ABA differentially influences
JA-regulated compounds (e.g., LapA vs. proteinase in-
hibitor) (Chao et al. 1999). It is also possible that the
resistance to herbivores on ABA-deficient plants was
Ecology, Vol. 85, No. 1
due to a non-jasmonate-mediated defensive change or
a nutritional change in the plant. However, other re-
searchers using the same mutant lines have found that
ABA-deficient plants have a reduced jasmonate re-
sponse following wounding (Pena-Cortes et al. 1989).
One limitation of our experiments is that we only as-
sayed PPO on undamaged plants; an assay following
damage may have revealed differences in inducibility
in wild-type and ABA-deficient plants. Thus, the insect
performance result is consistent with other research and
the idea that a threshold level of ABA is required for
expression of resistance to insects.
In our phenotypic manipulations, salt stress reduced
the inducibility of the SA response but did not influence
disease caused by the pathogen. This is consistent with
ABA inhibiting the SA responses; in these experiments,
we do not know why the pathogenesis-related protein
was not a good marker for disease resistance. Salt stress
marginally increased the performance of insects (mean
across all control treatments 6 1 SE, 0.314 6 0.043;
across all salt treatments, 0.427 6 0.042; P 5 0.06;
Fig. 4c, Table 1). This is in contradiction to our pre-
diction that salt stress would be synergistic with the
expression of the jasmonate response. It could be that
non-jasmonate-related changes in the plant were more
important determinants of herbivore performance than
any synergism with jasmonate.
While salt stress by itself did not alter the inducibility
of the jasmonate responses as measured by herbivore
performance, it did influence the interaction between
the jasmonate and SA response. In this study, as in
others (Thaler et al. 1999, 2002a, b) the expression of
the jasmonate and SA responses is antagonistic. This
can be seen in our phenotypic manipulations, where
jasmonate treatment reduced the performance of the
herbivore, but plants treated with jasmonate and BTH
had equal resistance to the control plants (Fig. 4c). This
negative interaction between the jasmonate and salic-
ylate pathways was not detected in salt-stressed plants.
The key point from these experiments is that salt stress
alters the interaction between salicylate and jasmonate
responses. This result is consistent with the findings of
Wang et al. (2001) that found that ABA and JA in-
creased following salt stress while SA levels declined.
Thus, the phytohormonal interactions that determine
plant resistance are certainly complex and involve sev-
eral pathways.
We did not find consistent results when we compared
our different ways of manipulating plant ABA levels.
This is not surprising; given that the two approaches
both alter ABA levels but also fundamentally distinct
physiological processes. The greater effect of the ge-
netic manipulation compared to the phenotypic manip-
ulation is consistent with the proposal that a basal level
of ABA is required for the expression of the jasmonate
(and perhaps salicylate response) and that only a subset
of the compounds require higher levels of ABA for
their expression (Chao et al. 1999). However, other
January 2004 PHYTOHORMONAL ECOLOGY 55

FIG. 4. Chemical and biological assays of resistance on control (unstressed) and salt-stressed plants when these plants Special Feature
are treated with jasmonate alone (JA), benzothiodiazole alone (BTH), jasmonate and benzothiodiazole (BTH&JA), or water
control. (a) Levels of pathogenesis-related protein P4. Bars indicate the level in one sample pooled from four plants. (b)
Disease caused by the bacterial pathogen Pseudomonas syringae pv. tomato (mm2 of infected tissue). Values are means 6 1
SE . (c) Relative growth rate of Trichoplusia ni caterpillars ([final mass 2 initial mass]/[initial mass 3 number of days]).
Values are means 6 1 SE. Results from three trials are combined in (b) and (c).

differences between the genetic and phenotypic ma-
nipulation could have generated these differences. The
possibility of whether threshold or higher levels of
ABA are required for the SA response warrants further
exploration (Audenaert et al. 2002).
These results highlight the pros and cons of using
different techniques for manipulating plant responses
to stress. For example, we can compare the usefulness
of genetically modified plants, hormonal manipula-
tions, and natural attack as ways to manipulate the plant
hormonal level. As both genetic modification and hor-
monal manipulation currently tend to be blunt tools,
causing extreme depression or induction of phytohor-
mones, it is important to (1) use a combination of ap-
 56 J. S. THALER AND R. M. BOSTOCK Ecology, Vol. 85, No. 1

TABLE 1. The effect of salt stress, jasmonate (JA), and ben- TABLE 2. The effect of salt stress, jasmonate (JA), and ben-
zothiodiazole (BTH) treatment on lesion formation (mm2) zothiodiazole (BTH) treatment on the relative growth rate
by the pathogen Pseudomonas syringae pv. tomato. ([final mass 2 initial mass]/[initial mass 3 number of
days]) of Trichoplusia ni caterpillars.
Effect df MS F P
Effect df MS F P
Salt 1 331.7 1.732 0.19
JA 1 1265.2 6.606 0.011 Salt 1 0.693 3.57 0.06
BTH 1 1474.9 7.702 0.006 JA 1 2.688 13.84 ,0.001
Salt 3 JA 1 135.2 0.706 0.402 BTH 1 1.007 5.182 0.024
Salt 3 BTH 1 0.018 0.000 0.992 Trial 2 4.335 22.319 ,0.001
JA 3 BTH 1 0.001 0.000 0.998 JA 3 BTH 1 0.025 0.13 0.719
Salt 3 JA 3 BTH 1 13.21 0.069 0.793 Salt 3 JA 1 0.792 4.075 0.045
Error 213 191.2 Salt 3 BTH 1 0.479 2.465 0.118
Salt 3 JA 3 BTH 1 0.213 1.096 0.296
Error 206 0.194

proaches and (2) couple these manipulations with ex-

periments that manipulate the hormonal level in a re-
alistic manner. These extreme manipulations define
what interactions are possible, more subtle manipula-
tions using weaker genetic modifications, or less potent
elicitors, describe the phenomenon in a real physio-
logical background and natural manipulations dem-
onstrate what actually occurs given a specific set of
conditions and organisms.
We can consider several hypothetical scenarios de-
scribing the interactions between stresses. For example,
Special Feature
plants mature (Farnsworth 2004). This pattern could
result from unrelated processes, physiological con-
straints, or an ontogenetic shift in the ecological im-
portance of the different stresses. Similarly, there is
seasonal variation in JA content of plants (Mopper et
al. 2004). In conclusion, we have clearly shown inter-
actions between ABA-, JA-, and SA-mediated signal-
ing responses. This demonstrates that a plants hor-
monal response to one stress influences its responses
to additional stresses. Phytohormonal ecology is thus

in a microenvironment with low levels of water avail-
ability or saline soils, ABA levels will be high in the
plant. Therefore, the SA response pathway will be sup-
pressed and plant resistance to pathogen attack will be
low. If, in general, levels of pathogen attack are lower
in dry microenvironments, plants may not often ex-
perience a fitness decrement from ABA-suppressed dis-
ease resistance. In environments with sufficient water,
the ABA pathway will not be elevated and, while dis-
ease pressure may be high, the SA response pathway
should be fully functional (unless basal levels of ABA
suppress SA signaling). However, if the positive cor-
relation between moisture content in the microhabitat
and disease incidence does not hold, then the inter-
actions between ABA and SA pose a greater problem
for the plant because plants will be attacked by path-
ogens when their defenses are compromised.
Future directions
The probabilities of multiple stresses simultaneously
affecting plants in nature will determine the conse-
quences of overlapping regulation between ABA-me-
diated and SA- or JA-mediated defense. Our long-term
goal should be to examine the congruence between
hormone-signaling interactions and the ecological pat-
tern of association of stresses. If water stress and path-
ogen stress are negatively associated in the environ-
ment, then negative signaling interactions between the
hormones responsive to these stresses may represent
adaptive coordination of responses. Age-related shifts
in hormonal signaling may also affect ecological in-
teractions. For instance, jasmonate and salicylate re-
sponses become less inducible as plants age (Karban
and Baldwin 1997) whereas ABA levels increase as
emerging as an integrative approach by testing the role
of interactions between hormonally signaled responses
on ecologically important traits.
ACKNOWLEDGMENTS
We thank A. Agrawal, T. Holtzer, R. Karban, S. Mopper,
C. Muis, D. Viswanathan, and two anonymous reviewers for
helpful discussion and comments and K. Boege and A. Fi-
dantsef for help with the experiments. This research was sup-
ported by USDA-NRI grant 98-02362 to R. M. Bostock and
R. Karban, and an NSERC operating grant to J. S. Thaler.
LITERATURE CITED
Agrawal, A. A., S. Y. Strauss, and M. J. Stout. 1999. Costs
of induced responses and tolerance to herbivory in male
and female fitness components of wild radish Evolution 53:
10931104.
Agrios, G. N. 1997. Plant pathology. Fourth edition. Aca-
demic Press, San Diego, California, USA.
Audenaert, K., G. B. De Meyer, and M. M. Hofte. 2002.
Abscisic acid determines basal susceptibility of tomato to
Botrytis cinerea and suppresses salicylic acid-dependent
signaling mechanisms. Plant Physiology 128:491501.
Baldwin, I. T. 1996. Methyl jasmonate-induced nicotine pro-
duction in Nicotiana attenuata: inducing defenses in the
field without wounding. Entomologia Experimentalis et
Applicata 80:213220.
Barkosky, R. R., and F. A. Einhellig. 1993. Effects of salicylic
acid on plant water relationships. Journal of Chemical Ecol-
ogy 19:237247.
Bender, E. A., T. J. Case, and M. E. Gilpin. 1984. Perturbation
experiments in community ecology: theory and practice.
Ecology 65:113.
Birkenmeier, G. F., and C. A. Ryan. 1998. Wound signaling
in tomato plants. Plant Physiology 117:687693.
Bostock, R. M., R. Karban, J. S. Thaler, P. D. Weyman, and
D. Gilchrist. 2001. Signal interactions in induced resis-
tance to pathogens and insect herbivores. European Journal
of Plant Pathology 107:103111.
Bray, E. A. 1993. Molecular responses to water-deficit. Plant
Physiology 103:10351040.
January 2004 PHYTOHORMONAL ECOLOGY
Cahill, D. M., and E. W. B. Ward. 1989. Rapid localized
changes in abscisic acid concentrations in soybean in in-
teractions with Phytophthora megasperma F Sp glycinea or
after treatment with elicitors. Physiological and Molecular
Plant Pathology 35:483493.
Chao, W. S., Y.-Q. Gu, V. Pautot, E. A. Bray, and L. L.
Walling. 1999. Leucine aminopeptidase RNAs, proteins,
and activities increase in response to water deficit, salinity,
and the wound signals systemic, methyl jasmonate, and
abscidic acid. Plant Physiology 120:979992.
Chapin, F. S. 1991. Integrated responses of plants to stress.
BioScience 41:2936.
Chase, J. M., M. A. Leibold, A. L. Downing, and J. B. Shurin.
2000. The effects of productivity, herbivory, and plant spe-
cies turnover in grassland food webs. Ecology 81:2485
2497.
Chen, C. C. S., and A. L. Plant. 1999. Salt-induced protein
synthesis in tomato roots: the role of ABA. Journal of Ex-
perimental Botany 50:677687.
Choi, D., R. M. Bostock, S. Avdiushko, and D. F. Hildebrand.
1994. Lipid-derived signals that discriminate wound- and
pathogen-responsive isoprenoid pathways in plants: methyl
jasmonate and the fungal elicitor arachidonic acid induce
different 3-hydroxy-3-methylglutaryl-coenzyme A reduc-
tase genes antimicrobial isoprenoids in Solanum tuberosum
L. Proceedings of the National Academy of Sciences (USA)
91:23292333.
Creelman, R. A., and J. E. Mullet. 1997. Biosynthesis and
action of jasmonates in plants. Annual Review of Plant
Physiology and Plant Molecular Biology 48:355381.
Crocoll, C., J. Kettner, and K. Dorffling. 1991. Abscisic acid
in saprophytic and parasitic species of fungi. Phytochem-
istry 30:10591060.
Cyr, H., and M. L. Pace. 1993. Magnitude and patterns of
herbivory in aquatic and terrestrial ecosystems. Nature 363:
148150.
de Bruxelles, G. L., and M. R. Roberts. 2001. Signals reg-
ulating multiple responses to wounding and herbivores.
Critical Reviews in Plant Sciences 20:487521.
Delaney, T. P., S. Uknes, B. Vernooij, L. Friedrich, K. Wey-
mann, D. Negrotto, T. Gaffney, M. Gut-Rella, H. Kess-
mann, E. Ward, and J. Ryals. 1994. A central role of sal-
icylic acid in plant disease resistance. Science 266:1247
1250.
English-Loeb, G., M. J. Stout, and S. S. Duffey. 1997.
Drought stress in tomatoes: changes in plant chemistry and
potential nonlinear consequences for insect herbivores. Oi-
kos 79:456468.
Farnsworth, E. 2004. Hormones and shifting ecology
throughout plant development. Ecology 85:515.
Felton, G. W., K. Donato, R. J. Del Vecchio, and S. S. Duffey.
1989. Activation of plant foliar oxidases by insect feeding
reduces nutritive quality of foliage for noctuid herbivores.
Journal of Chemical Ecology 15:26672694.
Felton, G. W., K. L. Korth, S. V. Wesley, D. V. Huhman, M.
C. Mathews, J. B. Murphy, C. Lamb, and R. A. Dixon.
1999. Inverse relationship between systemic resistance of
plants to microorganisms and to insect herbivory. Current
Biology 9:317320.
Fidantsef, A. L., M. J. Stout, J. S. Thaler, S. S. Duffey, and
R. M. Bostock. 1999. Signal interactions in pathogen and
insect attack: expression of lipoxygenase, proteinase in-
hibitor II, and pathogenesis-related protein P4 in the to-
mato, Lycopersicon esculentum. Physiological and Molec-
ular Plant Pathology 54:97114.
Garcia-Guzman, G., and R. Dirzo. 2001. Patterns of leaf-
pathogen infection in the understory of a Mexican rain
forest: Incidence, spatiotemporal variation, and mecha-
nisms of infection. American Journal of Botany 88:634
645.
57
Genoud, T., and J. P. Metraux. 1999. Crosstalk in plant cell
signaling: structure and function of the genetic network.
Trends in Plant Science 4:503507.
Gershenzon, J. 1984. Changes in the levels of plant secondary
metabolites under water and nutrient stress. Pages 273320
in B. N. Timmermann, C. Steelink, and F. A. Loewus, ed-
itors. Phytochemical adaptations to stress. Plenum, New
York, New York, USA.
Godfrey, L. D., and T. O. Holtzer. 1991. Influence of tem-
perature and humidity on European corn-borer (Lepidop-
tera, Pyralidae) egg hatchability. Environmental Entomol-
ogy 20:814.
Grillo, S., A. Leone, Y. Xu, M. Tucci, R. Francione, P. M.
Hasegawa, L. Monti, and R. A. Bressan. 1995. Control of
osmotin gene expression by ABA and osmotic stress in
vegetative tissues of wild-type and ABA-deficient mutants
of tomato. Physiologia Plantarum 93:498504.
Henfling, J., R. Bostock, and J. Kuc. 1980. Effect of abscisic
acid on rishitin and lubimin accumulation and resistance
to Phytophthora infestans and Cladosporium cucumerinum
in potato tuber tissue-slices. Phytopathology 70:1074
1078.
Herde, O., H. P. Cortes, C. Wasternack, L. Willmitzer, and J.
Fisahn. 1999. Electric signaling and Pin2 gene expression
on different abiotic stimuli depend on a distinct threshold
level of endogenous abscisic acid in several abscisic acid-
deficient tomato mutants. Plant Physiology 119:213218.
Holtzer, T. O., T. L. Archer, and J. M. Norman. 1988. Host
plant suitability in relation to water stress. Pages 111137
in E. A. Heinrichs, editor. Plant stressinsect interactions.
Special Feature
John Wiley and Sons, New York, New York, USA.
Karban, R., and I. T. Baldwin. 1997. Induced response to
herbivory. University of Chicago Press, Chicago, Illinois,
USA.
Kettner, J., and K. Dorffling. 1995. Biosynthesis and metab-
olism of abscisic acid in tomato leaves infected with Bo-
trytis cinerea. Planta 196:627634.
King, E. O., M. K. Ward, and D. E. Raney. 1954. Two simple
media for the demonstration of procyanin and flourescin.
Journal of Laboratory and Clinical Medicine 44:301307.
Koricheva, J., S. Larsson, and E. Haukioja. 1998. Insect per-
formance on experimentally stressed woody plants: a meta-
analysis. Annual Review of Entomology 43:195216.
Larsson, S. 1989. Stressful times for the plant stressinsect
performance hypothesis. Oikos 56:277283.
Lawton, K. A., L. Friedrich, M. Hunt, K. Weymann, T. De-
laney, H. Kessmann, T. Staub, and J. Ryals. 1996. Ben-
zothaidoazole induces disease resistance in Arabidopsis by
activation of the systemic acquired resistance signal trans-
duction pathway. Plant Journal 10:7182.
Li, A., and M. C. Heath. 1990. Effect of plant growth reg-
ulators on the interactions between bean plants and rust
fungi non-pathogenic on beans. Physiological and Molec-
ular Plant Pathology 37:245254.
Mattson, W. J., and R. A. Haack. 1987. The role of drought
stress in provoking outbreaks of phytophagous insects.
Pages 365407 in P. Barbosa and J. C. Schultz, editors.
Insect outbreaks. Academic Press, San Diego, California,
USA.
Mattson, W. J., and J. M. Scriber. 1987. Nutritional ecology
of insect folivores of woody plants: nitrogen, water, fiber,
and mineral considerations. Pages 105146 in F. Slansky
and J. G. Rodriguez, editors. Nutritional ecology of insects,
mites, spiders and related invertebrates. Wiley, New York,
New York, USA.
McDonald, K. L., and D. M. Cahill. 1999. Influence of abs-
cisic acid and the abscisic acid biosynthesis inhibitor, nor-
flurazon, on interactions between Phytophthora sojae and
soybean (Glycine max). European Journal of Plant Pathol-
ogy 105:651658.
 58 J. S. THALER AND R. M. BOSTOCK
Merlot, S., and J. Giraudat. 1997. Genetic analysis of abscisic
acid signal transduction. Plant Physiology 114:751757.
Moons, A., E. Prisen, G. Blaw, and M. V. Montagu. 1997.
Antagonistic effects of abscisic acid and jasmonates on salt
stress-inducible transcripts in rice roots. Plant Cell 9:2243
2259.
Mopper, S., Y. Wang, C. Criner, M. Tobler, and K. Hasenstein.
2004. Iris hexagona hormonal responses to salinity stress,
leafminer herbivory, and phenology. Ecology 85:3847.
Munns, R. 2002. Comparative physiology of salt and water
stress. Plant Cell And Environment 25:239250.
Neill, S. J., and R. Horgan. 1985. Abscisic acid production
and water relations in wilty tomato mutants subjected to
water deficiency. Journal of Experimental Botany 36:1222
1231.
Orozco-Cardenas, M., B. McGurl, and C. A. Ryan. 1993.
Expression of an antisense prosystemin gene in tomato
plants reduces resistance toward Manduca sexta larvae.
Proceedings of the National Academy of Science (USA)
90:82738276.
Parker, S. K., M. L. Gleason, and F. W. Nutter. 1995. Influence
of rain events on the spatial distribution of Septoria leaf
spot of tomato. Plant Disease 79:148152.
Paul, N. D., P. E. Hatcher, and J. E. Taylor. 2000. Coping
with multiple enemies: An integration of molecular and
ecological perspectives. Trends in Plant Science 5:220
225.
Pena-Cortes, H., J. J. Sanchez Serrano, R. Mertens, L. Will-
mitzer, and S. Prat. 1989. Abscisic acid is involved in the
Special Feature
wound-induced expression of the proteinase inhibitor-ii
gene in potato and tomato. Proceedings of the National
Academy of Sciences (USA) 86:98519855.
Pena-Cortes, H., L. Willmitzer, and J. J. Sanchez-Serrano.
1991. Abscisic acid mediates wound induction but not de-
velopmental-specific expression of the proteinase inhibitor
II gene family. Plant Cell 3:963972.
Perring, T. M., T. O. Holtzer, J. L. Toole, J. M. Norman, and
G. L. Myers. 1984. Influences of temperature and humidity
on pre-adult development of the banks grass mite (Acari,
Tetranychidae). Environmental Entomology 13:338343.
Reymond, P., H. Weber, M. Damond, and E. E. Farmer. 2000.
Differential gene expression in response to mechanical
wounding and insect feeding in Arabidopsis. Plant Cell 12:
707719.
Ristaino, J. B., and J. M. Duniway. 1989. Effect of preinocu-
lation and postinoculation water-stress on the severity of
Phytophthora root-rot in processing tomatoes. Plant Dis-
ease 73:349352.
Ristaino, J. B., J. M. Duniway, and J. J. Marois. 1988. In-
fluence of frequency and duration of furrow irrigation on
the development of Phytophthora root rot and yield in pro-
cessing tomatoes. Phytopathology 78:17011706.
Schenk, P. M., K. Kazan, I. Wilson, J. P. Anderson, T. Rich-
mond, S. C. Somerville, and J. M. Manners. 2000. Co-
ordinated plant defense responses in Arabidopsis revealed
by microarray analysis. Proceedings of the National Acad-
emy of Sciences (USA) 97:1165511660.
Stout, M. J., A. L. Fidantsef, S. S. Duffey, and R. M. Bostock.
1999. Signal interactions in pathogen and insect attack:
systemic plant-mediated interactions between pathogens
and herbivores of the tomato, Lycopersicon esculentum.
Physiological and Molecular Plant Pathology 54:115130.
Swiecki, T. J., and J. D. MacDonald. 1988. Histology of
chrysanthemum roots exposed to salinity stress and Phy-
tophthora crypotgea. Canadian Journal of Botany 66:280
288.
Ecology, Vol. 85, No. 1
Tal, M., and Y. Nevo. 1973. Abnormal stomatal behavior and
root resistance, and hormonal imbalance in three wilty mu-
tants of tomato. Biochemical Genetics 8:291300.
Taylor, I. B. 1987. ABA-deficient tomato mutants. Pages
197217 in H. Thomas and D. Grierson, editors. Devel-
opmental mutants in higher plants. Cambridge University
Press, Cambridge, UK.
Thaler, J. S., A. L. Fidantsef, and R. M. Bostock. 2002a.
Antagonism between jasmonate- and salicylate-mediated
induced plant resistance: effects of concentration and tim-
ing of elicitation on defenserelated proteins, herbivore,
and pathogen performance in tomato. Journal of Chemical
Ecology 28:11311159.
Thaler, J. S., A. L. Fidantsef, S. S. Duffey, and R. M. Bostock.
1999. Trade-offs in plant defense against pathogens and
herbivores: a field demonstration of chemical elicitors of
induced resistance. Journal of Chemical Ecology 25:1597
1609.
Thaler, J. S., R. Karban, D. E. Ullman, K. Boege, and R. M.
Bostock. 2002b. Cross-talk between jasmonate and salic-
ylate plant defense pathways: effects on several plant par-
asites. Oecologia 131:227235.
Thaler, J. S., M. J. Stout, R. Karban, and S. S. Duffey. 1996.
Exogenous jasmonates simulate insect wounding in tomato
plants (Lycopersicon esculentum) in the laboratory and
field. Journal of Chemical Ecology 22:17671781.
Thomma, B. P. H. J., K. Eggermont, I. A. M. A. Penninckx,
B. Mauch-Mani, R. Vogelsang, B. P. A. Cammue, and W.
F. Broekaert. 1998. Separate jasmonate-dependent and sa-
licylate-dependent defense-response pathways in Arabi-
dopsis are essential for resistance to distinct microbial path-
ogens. Proceedings of the National Academy of Science
(USA) 95:1510715111.
Tyree, M. T., and H. T. Hammel. 1972. The measurement of
the turgor pressure and the water relations of plants by the
pressure-bomb technique. Journal of Experimental Botany
23:267282.
Waldbauer, G. B. 1968. The consumption and utilization of
food by insects. Advances in Insect Physiology 5:229289.
Walton, D. C. 1980. Biochemistry and physiology of abscisic
acid. Annual Review of Plant Physiology and Plant Mo-
lecular Biology 31:453489.
Wang, Y., S. Mopper, and K. H. Hasenstein. 2001. Effects
of salinity on endogenous ABA, IAA, JA, and SA in Iris
hexagona. Journal of Chemical Ecology 27:327241.
Ward, E. W. B., D. M. Cahill, and M. K. Bhattacharyya. 1989.
Abscisic acid suppression of phenylalanine ammonia lyase
activity and mNA, and resistance of soybeans to Phyto-
phthora megasperma f. sp. glycinea. Plant Physiology 91:
2327.
Waring, G. L., and N. S. Cobb. 1992. The impact of plant
stress on herbivore population dynamics. Pages 167226
in L. Bernays, editor. Insectplant interactions. Volume 4.
CRC Press, Boca Raton, Florida, USA.
Wasternack, C., and B. Parthier. 1997. Jasmonate signalled
plant gene expression. Trends in Plant Science 2:302307.
Wei, J. Z., A. Tirajoh, and J. Effendy, A. L. Plant. 2000.
Characterization of salt-induced changes in gene expres-
sion in tomato (Lycopersicon esculentum) roots and the role
played by abscisic acid. Plant Science 159:135148.
Whenham, R. J., R. S. S. Fraser, L. P. Brown, and J. A. Payne.
1986. Tobacco mosaic virus induced increase in abscisic
acid concentration in tobacco leaves. Planta 168:592598.
White, T. C. R. 1974. A hypothesis to explain outbreaks of
looper caterpillars with special reference to populations of
Selidosema suavis in a plantation of Pinus radiata in New
Zealand. Oecologia 16:279301.
Zeevaart, J. A. D., and R. A. Creelman. 1988. Metabolism
and physiology of abscisic acid. Annual Review of Plant
Physiology and Plant Molecular Biology 39:439473.
Ecology, 85(1), 2004, pp. 5969
q 2004 by the Ecological Society of America

PHYTOHORMONES AND PLANTHERBIVOREPATHOGEN

INTERACTIONS: INTEGRATING THE MOLECULAR
WITH THE ECOLOGICAL
PAUL E. HATCHER,1,3 JASON MOORE,2 JANE E. TAYLOR,2 GLENDA W. TINNEY,2,4 AND NIGEL D. PAUL2
1Schoolof Plant Sciences, The University of Reading, Reading RG6 6AU, UK
2Department of Biology, IENS, Lancaster University, Lancaster LA1 4YQ, UK

Abstract. Current research into indirect phytopathogenherbivore interactions (i.e.,

interactions mediated by the host plant) is carried out in two largely independent directions:
ecological/mechanistic and molecular. We investigate the origin of these approaches and
their strengths and weaknesses. Ecological studies have determined the effect of herbivores
and phytopathogens on their host plants and are often correlative: the need for long-term
manipulative experiments is pressing. Molecular/cellular studies have concentrated on the
role of signaling pathways for systemic induced resistance, mainly involving salicylic acid
and jasmonic acid, and more recently the cross-talk between these pathways. This cross-
talk demonstrates how interactions between signaling mechanisms and phytohormones
could mediate plantherbivorepathogen interactions. A bridge between these approaches
may be provided by field studies using chemical induction of defense, or investigating
whole-organism mechanisms of interactions among the three species. To determine the role
of phytohormones in induced resistance in the field, researchers must combine ecological
and molecular methods. We discuss how these methods can be integrated and present the
concept of kaleidoscopic defense. Our recent molecular-level investigations of inter-

Special Feature
actions between the herbivore Gastrophysa viridula and the rust fungus Uromyces rumicis
on Rumex obtusifolius, which were well studied at the mechanistic and ecological levels,
illustrate the difficulty in combining these different approaches. We suggest that the choice
of the right study system (possibly wild relatives of model species) is important, and that
molecular studies must consider the environmental conditions under which experiments are
performed. The generalization of molecular predictions to ecologically realistic settings
will be facilitated by middle-ground studies concentrating on the outcomes of the in-
teractions.
Key words: herbivores; insects; jasmonic acid; molecular ecology; plant pathogens; Rumex
obtusifolius; salicylic acid; systemic induced resistance.

INTRODUCTION host plant without necessarily directly affecting each

Given that both herbivores and phytopathogens can
parasitize the same part of the same plant at the same
time, it is not surprising that they should interact. The
importance of direct interactions between these phyla,
for example insects vectoring or feeding upon fungi
(e.g., Wilding et al. 1989) or insect vectors of viruses
(Nault 1997), is widely recognized. However, indirect
interactions between these organisms have received
much less attention and detailed studies have only ap-
peared in the last 1015 yr (reviewed in, for example,
Barbosa 1991, de Nooij et al. 1992, Hatcher and Paul
2001). These indirect interactions, termed tripartite,
whereby herbivores and pathogens affect their shared
other, bring the plant to center stage.
Current research into indirect herbivorepathogen
interactions is carried out in two largely independent
directions: ecological/mechanistic and molecular. In
this article, we explore these different research ap-
proaches for investigating plantpathogenherbivore
interactions. We then suggest how these approaches
might be integrated to elucidate the importance of phy-
tohormones within this field, and consider our expe-
rience in trying to apply molecular techniques to plant
pathogenherbivore interactions in Rumex, which have
been well studied ecologically.
TRIPARTITE INTERACTIONS: DIFFERENT SCALES,
DIFFERENT PERSPECTIVES

Manuscript received 18 November 2002; revised 28 April
2003; accepted 29 April 2003; final version received 10 June
2003. Corresponding Editor (ad hoc): S. Mopper. For reprints of
this Special Feature, see footnote 1, p. 3.
3 E-mail: p.e.hatcher@rdg.ac.uk
4 Present address: Trinity College, Carmarthen SA31 3EP,
UK.
59
Ecological studies into indirect plantpathogenher-
bivore interactions developed from the observation of
direct interactions, mainly of the role of insects as vec-
tors of plant diseases, which arose in the latter half of
the 19th century (Leach 1940). For much of the 20th
century, published studies of tripartite interactions fo-
 60 PAUL E. HATCHER ET AL.
cused on the observation that diseased plants were of-
ten more subject to herbivory (e.g., Leach et al. 1963,
Lewis 1979; others are reviewed in Hatcher [1995]).
Systematic studies of such interactions during the
1980s led to detailed investigation of the underlying
mechanisms (Ramsell and Paul 1990, de Nooij et al.
1992), and possible parallels with pathogenpathogen
interactions (Hallett et al. 1990). This research was
partly driven by the possible application of tripartite
interactions in weed biocontrol, the underlying hy-
pothesis being that interactions between potential in-
sect and pathogen biocontrol agents could produce a
synergistic amount of damage on the host plant. Field
experiments involving the release of potential biolog-
ical control agents (e.g., Baudoin et al. 1993, Kok et
al. 1996) have provided information about the potential
for synergistic effects on the host plants, but cannot
tell us about the occurrence of such effects in the wild.
Studies of the fundamental ecology of herbivore
phytopathogen interactions start with the key question
of whether such interactions occur in the wild, and
culminate in determining whether they are significant
in the dynamics of natural populations of the partici-
pants and in shaping their niches. The last decade has
Special Feature
seen the development of field studies correlating insect
and phytopathogen distribution with or without exper-
imental manipulation. This has involved individual her-
bivore and pathogen species (e.g., Ericson and Wenn-
strom 1997, Lappalainen and Helander 1997) or guilds
of herbivores (Kluth et al. 2001). However, a causal
relationship between herbivore and phytopathogen dis-
tributions is harder to determine since correlations
could be due solely to selection of suitable plant char-
acteristics. Recent experimental support for this view
comes from Ahlholm et al.s (2002) study comparing
the phenotypic and genetic correlations between fungal
frequency and invertebrate herbivore performance on
Betula pubescens half-sib progeny. They found little
evidence for direct interactions between fungi and her-
bivores; rather, their distribution was determined by
differing responses to genetically determined tree char-
acteristics. Correlative field studies are vital in eluci-
dating the occurrence of tripartite interactions in na-
ture, but fall short of the ultimate goal of understanding
the role of tripartite interactions in the dynamics of
natural populations of the participants. This issue has
yet to be adequately addressed, although experiments
to investigate the effects of interactions between nat-
ural or artificially increased levels of herbivore and
pathogen damage on the lifetime fitness of the host are
soon to be in hand (P. E. Hatcher, unpublished data).
Research into these tripartite interactions by cellular
and molecular biologists has developed from the bur-
geoning understanding of the signaling pathways in-
volved in defense against pathogens and herbivores,
and the scope for cross-talk between these pathways tion of the wound response genes, and ethylene and JA
(Maleck and Dietrich 1999, Bostock et al. 2001). Stud-
ies of systemic acquired resistance (SAR) against path-
Ecology, Vol. 85, No. 1
ogen infection (including salicylic-acid [SA]-mediated
pathways) and the wound response pathway (including
jasmonic-acid [JA]-mediated pathways), which pro-
vides resistance to insect feeding, initially developed
largely independently, but now provide several lines of
evidence that plant responses to insects and pathogens
can interact at the molecular/cellular level. Cross-talk
between SAR pathways and wound-response pathways
provides a phytohormonal mechanism for indirect in-
teractions between insects and plant pathogens (Bos-
tock et al. 2001). While some studies have suggested
that stimulation of one pathway facilitates the other
(e.g., Schweizer et al. 1998), others indicate inhibition
of one pathway by prior activation of the other (e.g.,
Felton et al. 1999, Thaler 1999, Bostock et al. 2001,
Thaler et al. 2002). This may be due to various dif-
ferences in the experiments, including differential sig-
nal transduction following insect feeding between the
plant species studied (i.e., monocots vs. dicots) (de
Bruxelles and Roberts 2001, Kessler and Baldwin
2002).
Cross-talk in the narrow sense of direct interactions
between components of signaling pathways is not the
only mechanism of interaction revealed by cell and
molecular studies. Interactions are also possible
through overlap in gene expression: Reymond and
Farmer (1998) put forward the concept of a tuneable
dial for the regulation of defense gene expression
mediated by the balance of signal regulators, which is
induced by a specific aggressor. We, too, suggested
(Paul et al. 2000) that, although there is specificity in
plant recognition of diverse attackers, the downstream
outcomes may show overlap in terms of the effects at
the whole-plant level. Support for such a model is pro-
vided by recent experiments using cDNA microarray
expression analysis, which showed that, in Arabidop-
sis, expression of 168 out of 2375 transcripts was reg-
ulated by multiple treatments or defense pathways
(Schenk et al. 2000).
As well as cross-talk between pathways or shared
gene expression, understanding of herbivorephyto-
pathogen interactions may also benefit from the rec-
ognition that the view of one signaling pathway for
pathogens (SA) and one for insect defense (JA) is over-
simplified. For example, microarray expression anal-
ysis and other global molecular approaches have
demonstrated the induction of phytopathogen-respon-
sive genes by insect feeding in Arabidopsis and wild
tobacco (Reymond et al. 2000, Hermsmeier et al.
2001). Conversely, Pseudomonas syringae induces not
only pathogenesis-related protein expression but also
proteinase inhibitor genes considered to be JA induced
(Bostock et al. 2001). Other phytohormones are known
to interact with the JA/SA pathways: for example, eth-
ylene acts synergistically with JA for complete induc-
may stimulate the biosynthesis of each other
(ODonnell et al. 1996, Penninckx et al. 1998). Other
January 2004 PHYTOHORMONAL ECOLOGY
induced-resistance pathways, some as yet poorly de-
fined, may differentially interact with the JA and SA
pathways, and affect the outcome of their interaction,
or may prove to be independent of SA or JA (Cameron
2000, Bostock et al. 2001, Cui et al. 2002). Thus, phy-
tohormonal pathways can no longer be considered as
linear signaling chains but rather as complex networks
interacting over a range of levels. The implications of
such signaling networks for plantpathogenherbivore
interactions in the field remain essentially unexplored.
Much of the work on interactions between defense
pathways has used chemically induced plants: JA or
MeJA application will induce resistance against insect
herbivores, and applications of SA or artificial mimics
such as benzothiadiazole (BTH/BION) (Oostendorp et
al. 2001) will induce the SA pathway against phyto-
pathogens. However, using insect herbivores and phy-
topathogens to produce the induction can lead to dif-
ferent results. For example, application of SA or BTH
to tomato can inhibit JA-mediated resistance and in-
crease the vulnerability of tomato foliage to herbivory
by some insects, but infection by either P. syringae or
insect feeding allows for the concurrent expression of
both the SA- and JA-signaling pathways in contrast to
the effects of exogenous JA treatment (Bostock et al.
2001). There may be similar limits in the use of sig-
naling mutants rather than real plant enemies (Cui et
al. 2002). Thus, while the use of chemical induction
has the great merit of allowing field studies that can
consider relatively broad ecological questions (e.g.,
Thaler 1999, Thaler et al. 2002) and seek to address
the implications of cross-talk, they require careful
interpretation in the context of ecological interactions.
Clearly, studies by ecologists on the one hand, and
molecular or cell biologists on the other, have ap-
proached tripartite interactions from fundamentally dif-
ferent perspectives. One bridge between these two bod-
ies of research may be the field studies using chemical
induction of defense. Another possible bridge could be
provided by the relatively small number of studies
looking at whole-organism mechanisms of interactions
between herbivores, phytopathogens and plants (re-
viewed by Hatcher and Paul [2001]). These have de-
veloped from field observations and have borrowed
ecophysiological approaches used in phytopathology
and insect ecology to explore two-way interactions.
Given their origins, these studies took a broad view
of the mechanisms by which herbivores and phyto-
pathogens might interact via their host. Changes in de-
fensive chemistry have been studied in some cases: for
example infection by the rust Uromyces rumicis alters
local concentrations of oxalate in Rumex spp. (Hatcher
et al. 1997), and the rust Puccinia lagenophorae causes
systemic reductions in the concentration of pyrrolizi-
dine alkaloids in Senecio vulgaris (Tinney et al. 1998).
However, these systemic changes were explained
through mechanisms such as altered photosynthate sup-
ply or patterns of transpiration, rather than invoking
61
any specific signaling pathway. Ecophysiological stud-
ies of tripartite interactions have also highlighted the
role of mechanisms other than defense, sensuo stricto.
For example, host nitrogen has a major influence on
both plantpathogen and plantherbivore interactions
and, in turn, is affected both by herbivory and disease.
Not surprisingly, several studies have demonstrated a
key role of nitrogen in tripartite interactions (Hatcher
1995), for example, insectfungus interactions on Ru-
mex obtusifolius (Hatcher and Paul 2001).
This holistic ecophysiological perspective on the
mechanisms of tripartite interactions is paralleled in
recent evidence that reprogramming of gene expression
following attack extends far beyond defense genes
(Schenk et al. 2000, Hermsmeier et al. 2001). Unfor-
tunately, to date, such related data have tended to con-
tinue in parallel rather than converging through any
effective integration across the traditional disciplines
of molecular/cell biology, ecophysiology, and ecology.
INTEGRATION ACROSS DISCIPLINES
We believe that the best route to understanding the
mechanisms of tripartite interactions, and the role of
phytohormones in such interactions, is from an inte-
Special Feature
gration of research approaches from the molecular to
the ecological. This requires the recognition of the im-
portance of each approach and a genuine dialogue be-
tween, and open-mindedness from, both ecologists and
molecular biologists.
HOW CAN MOLECULAR/CELLULAR STUDIES INFORM
POSSIBLE ECOLOGICAL RESEARCH?
Methodological
Recent changes of emphasis in molecular studies
away from individual signaling pathways towards in-
teractions between multiple elicitors such as JA and
SA have gone some way towards examining more com-
plex interactions. Likewise, genomics is increasingly
indicating that herbivore attack is more than just the
induction of wound-responsive genes, and that signals
that transduce herbivore defense may also influence the
molecular response to pathogen invasion. However, it
will still be a great leap to address such interactions at
the molecular level under field conditions. There is no
doubt that if we are to understand the consequences of
such pathways in an ecological context, where complex
interactions with a range of abiotic and biotic factors
may influence the outcome, then we have to be highly
imaginative and integrative in our approaches.
The choice of the right system will be crucial for the
first attempts. We are bringing these disparate fields
together in a molecular examination of the interaction
between the chrysomelid beetle Gastrophysa viridula,
the rust fungus Uromyces rumicis and their host plant
Rumex obtusifolius. This system is well-defined eco-
logically but previously untouched at the molecular
level. This has necessitated ecologists and molecular
 62 PAUL E. HATCHER ET AL. Ecology, Vol. 85, No. 1

BOX 1. Molecular Biology in a Non-Model System:

Applying Molecular Techniques to Rumex Tripartite Interactions

Working with species other than the standard models of molecular biology, especially when these are
grown in the field, may pose challenges in terms of the chemical composition of tissues. Many plants
contain secondary metabolites that interfere with the extraction of nucleic acids, and the concentration of
such metabolites may vary according to growing conditions. In our Rumex system, oxalic acid poses
particular problems. Leaf tissues of Rumex obtusifolius contain up to 80 mg/g dry mass of oxalate. This
compound co-extracts with RNA using standard extraction procedures and also has pronounced UV ab-
sorbance, so contamination leads to both poor yields of RNA and great difficulty in accurately quantifying
the precise yield (Tinney et al. 2002). Furthermore, oxalate concentrations vary according to growing
conditions (Hatcher et al. 1997), adding additional uncertainty to the efficiency of extraction. Thus, reliable
RNA extraction and quantification was a significant initial hurdle in our attempts to apply molecular
techniques to Rumex, until a satisfactory modification to extraction procedures was developed (Tinney et
al. 2002).
Even with extraction optimized, there was marked plant-to-plant variation in the quantity of RNA
extracted from Rumex leaf tissue, even in controlled environment experiments. For example, the yield of
RNA (mg RNA extracted/g fresh mass leaf) varied by 40%, even between replicate plants in the same
experiment. Such variation is less likely in many conventional molecular studies, which use plant material
that is genotypically well defined and uniform (a single genotype, strain, or cultivar), compared to ecological
studies which often have to use less uniform plants. Our experiments with Rumex use a consistent seed
population that shows no gross visual variation, but is not a single genotype, and, of course, any study
of natural field populations will be confronted with even greater plant-to-plant variation. Such variation
Special Feature

in extraction efficiency may well affect the interpretation of data (Box 2).

biologists learning to speak each others language
in terms of methods and approaches, and has high-
lighted the difficulty in finding a middle ground be-
tween very different disciplines (Boxes 1 and 2). Eco-
logical and ecophysiological investigators ensure the
rigor of their data by relying on adequate replication
and formal statistical analysis. Experiments may be re-
peated, but this is often constrained by the time scale
of many studies and, in the field, repeated experiments
need to be interpreted in the light of inevitable variation
in environmental conditions. This approach is funda-
mentally different from that adopted in molecular bi-
ology, where data are usually presented as gels that are
cited as representative of repeated experiments.
Within experiments, there may be several plants per
treatment, but these are typically pooled, so changes
evident in gels represent bulked changes, with no
indication of plant-to-plant variation. Where gels are
presented, differences are discussed on the basis of
visual difference in the area and/or intensity of a given
band. Even where data are quantitative (e.g., real-time
PCR), changes are often presented as the ratio of ex-
pression between control and treatment samples, with
differences being considered significant above a par-
ticular ratio: statistical analysis of such data remains
rare. In trying to apply molecular methods to ecological
experiments with Rumex and its parasites, we have
sought a middle ground (Box 2).
It is also possible to apply ecological approaches to
work with the model plant systems that are central to
molecular biology. This approach has provided real
gains in some areas of ecology (Jackson et al. 2002),
but less so in understanding the molecular ecology of
plant defense. This approach is further confounded by
the difficulty in moving from models such as Arabi-
dopsis (which were selected for molecular applications
on the basis of particular genetic and developmental
features, ease of growth in laboratory, etc.), to species
with very different ecological (including defensive)
strategies. Perhaps the ideal route forward, combining
the best of both approaches, is to study wild relatives
of model species. The use of Nicotiana attenuata by
Baldwin and colleagues (Baldwin 2001, Hermsmeier
et al. 2001, Schittko et al. 2001) demonstrates the pow-
er of this approach in plantherbivore interactions.
Conceptual
Regardless of the applicability of molecular tech-
niques in field ecology, the developing understanding
of phytohormonal control of plant responses to phy-
topathogens and herbivores, alone and in combination,
offers exciting new insights into the ecology of plant
defenses. We now need to progress from a mechanistic
perspective that simply aids the interpretation of spe-
cific interactions to a more general understanding that
might offer consistent rules across systems. Such rules
January 2004 PHYTOHORMONAL ECOLOGY 63

BOX 2. Experimental Design and Interpretation: Reconciling Molecular and Ecological Approaches
In applying molecular techniques to our Rumex system, we have had to adopt very different approaches
from those used in our previous ecophysiological studies. Using 1015 replicates per treatment, as in our
previous studies, was not feasible for our molecular studies. However, analysis was viable, if time consuming,
using fewer replicates and treating each separately throughout extraction. Below, we give an example of our
methodology, using Uromyces rumicis infection, that allowed us to assess how plant-to-plant variation and
the application of statistical analyses would influence the interpretation of gene expression data.
Methodology
Rumex obtusifolius plants were grown as described in Moore et al. (2003) and divided into two treatment
sets of four to five replicates: Uromyces rumicis-inoculated plants; and control, uninoculated plants. Treat-
ments were carried out as described in Moore et al. (2003) and each plant was harvested separately. Leaf
6 (the damaged leaf or control equivalent) and leaf 7 (this developed after treatment of leaf 6) were removed
from each plant, divided along the midrib and immediately frozen in liquid nitrogen. All leaves were stored
separately in a 2808C freezer. For comparison of expression in pooled vs. replicated samples, one half of
each treatment was subsequently processed as independent replicates and the other half was combined to
give a pooled sample. These were then used in northern analysis to compare expression patterns.
RNA was isolated as described in Tinney et al. (2002). Poly A1 RNA from the U. rumicis-inoculated
sample was used to produce a Clontech custom-made library constructed in lTriplEx 2 (Clonetech, Ba-
singstoke, UK). To make the libraries pooled RNA from leaves 5, 6, and 7 (leaf 6 being the treated leaf)
of three replicate plants was used after 0, 6, 24, and 72 h, and 7 and 10 d of treatment. These libraries were
differentially screened for herbivore- or pathogen-induced cDNAs, respectively. This was carried out using
replicate lifts and labeled cDNA probes made from RNA from either treated plants or nontreated controls
over similar time scales of sampling. Isolated clones were sequenced and the data deposited in GenBank.5

Special Feature
Northerns were carried out by applying 20 mg RNA to a 1.5% agarose denaturing gel using standard
procedures. The membranes were then probed with DNA (either PCR product or plasmid) made using the
rediprime II (Amersham Pharmacia Biotech, Buckingham, UK) random prime labeling system. After
hybridization was complete, the membranes were washed, the final wash using 0.1 3 SSC, 0.1% SDS at
658C. The membranes were placed on Hyperfilm (Amersham Pharmacia Biotech) in a film cassette at
2808C prior to developing. Northerns were quantified using a Pharmacia ImageMaster VDS.
Results and Discussion
Differential screening revealed a range of cDNAs that were differentially expressed following rust
infection. These included a range of defense genes such as phenylalanine ammonium lyase, chalcone
synthase, and cysteine proteinase inhibitior. Other nondefense-related genes such as Rubisco and chlo-
rophyll a/b binding protein (chlab) often showed complex patterns of up and down regulation depending
upon the sampling time after attack.
Changes in expression showed the same trends in pooled and replicated samples. However, while there
was qualitative agreement, there were often quantitative differences between the two approaches, as
illustrated by changes in chl a/b binding protein following rust infection (Fig. 1). In principal, if all
replicate plants contributed equally to the pooled sample such deviation should not occur within the limits
of sampling error. We believe that plant-to-plant variation in extraction efficiency in Rumex (Box 1) biased
the pooled samples, with samples with higher extraction efficiencies making a disproportionate contribution
to the pooled data. When using pooled samples this problem will be minimized by pooling sufficient
numbers of plants to minimize the effect of any outlier and, clearly, where experiments are repeated to
ensure good and equal loading then conclusions may be drawn with some certainty.
The most significant conclusion from this work is that relatively small changes in transcript level that
might be hard to detect by visual assessment of gels, and so would not be treated as real in many
molecular studies, can be statistically significant. This can be illustrated by considering the effects of rust
infection on chl a/b binding protein (Fig. 1). The expression of this gene was reduced by 30% in rusted
Rumex 48 h after inoculation, and this was significant at P , 0.01 (Fig. 1b). Experiments carried out in
this manner enable statistical analysis to be undertaken and undoubtedly give more confidence to an
assessment of the data. This approach also provides an objective assessment of subtle changes in gene
expression that might otherwise be ignored in conventional molecular studies, but which may have bio-
logical relevance. We would therefore argue that to effectively integrate molecular and ecological ap-
proaches in this kind of study such an approach should be adopted.

5 URL: ^www.ncbi.nlm.nih.gov/&
 64 PAUL E. HATCHER ET AL.
remain elusive, and there are inconsistencies in the mo-
lecular literature, but existing knowledge hints at some
possibilities, at least in model species. For example,
there are suggestions that defense against chewing her-
bivores is mostly JA mediated while responses to suck-
ing herbivores are controlled by JA 1 SA (Walling
2000). Similarly, biotrophic and necrotrophic patho-
gens use different combinations of signaling pathways
(Kunkel and Brooks 2002). On this basis, we might
predict that interactions between organisms that induce
the same pathway (e.g., chewer with chewer, chewer
Special Feature
with necrotrophic pathogens) will be different from
interactions between organisms that induce different
pathways (e.g., chewer with sucker, chewer with bio-
trophic pathogens). Given the broader view of mech-
anisms of tripartite interactions that we develop here,
we accept that such predictions will almost certainly
prove oversimplistic, but they at least offer a starting
point on which to build.
The developing view among molecular biologists
that defenses are fine-tuned to particular attackers,
i.e., that defenses are deployed together in specific
combinations to match the vulnerabilities of particular
consumers (Reymond and Farmer 1998, Kunkel and
Brooks 2002), poses intriguing ecological questions.
To achieve an ecological integration of studies of plant-
defense-related phytohormonal changes within studies
of tripartite plantpathogenherbivore interactions re-
quires consideration of induced defense over a range
of scales, both spatial (from within leaf to between
plant) and temporal (from a few hours to, in long-lived
species, years). The suggestion that a plant may fine-
tune induced defenses offers an interesting perspec-
tive on this issue, especially on shorter time scales. The
concept that defenses are induced in different combi-
nations is consistent with the known deployment of
constitutive defenses in combination (Wittstock and
Gerhenzon 2002). As with constitutive defenses, the
combination of induced defense mechanisms is likely
to confer both synergy in effect and greater durability
against counteradaptations (Wittstock and Gerhenzon
2002). With fine-tuning the induced permutation
would be consumer specificthe induction of most
individual components of resistance would not be. A
particular combination of mechanisms might be opti-
Ecology, Vol. 85, No. 1
FIG. 1. Chlorophyll a/b (Chlab) binding
protein expression in Rumex obtusifolius leaves
in response to pathogen infection of leaf 6. RNA
was extracted from untreated leaf 7, (a) 24 h
and (b) 48 h after inoculation with Uromyces
rumicis (leaf 6 controls were mock inoculated).
Levels of transcript, as detected by northern
blot, were calculated as a percentage of the 18S
ribosomal RNA transcript level to account for
variation in gel loading, and results are pre-
sented relative to expression at 0 h. Means 6
1 SE are shown.
** P , 0.01, compared to 0 h; otherwise P
. 0.05.
mal for defense against a particular consumer, but
would be expected to have some degree of activity
against a wide range of consumers (Paul et al. 2000).
This is relevant to the consideration of the costs of
induced resistance since the cost of induction by any
one organism would be offset by the benefit of in-
creased defense against a range of potential consumers:
field experiments have shown this benefit (reviewed by
Thaler [1999]). Thus, it is important that we should
not confuse fine-tuning with specificity.
Fine-tuning also merits examination in the context
of interactions between plants and multiple enemies,
which we would argue is the norm in the field, with
attack by a single consumer a rare exception. For ex-
ample, individual plants of Rumex obtusifolius are com-
monly attacked simultaneously by a phloem sucker
(Aphis rumicis), a chewer (Gastrophysa viridula), and
a fungal pathogen, which could be a biotroph ( Uro-
myces rumicis) or a necrotroph (Ramularia rubella).
How does a plant faced with such multiple attacks de-
ploy its fine-tuned defenses? One possibility, to
some extent a variation on the view that plants are
induced all the time (Karban and Myers 1989), would
be that all components are induced (i.e., that potential
fine-tuning is not actually expressed in nature). We
believe that a key point here (and indeed in the plants
are induced all the time argument) is the spatial and
temporal dynamics of systemic induced resistances. In
general, induction is not an irreversible change ex-
pressed uniformly across the whole plant. Molecular
studies have demonstrated that plant responses are ex-
tremely dynamic, varying on an hour-by-hour or day-
by-day basis (Reymond et al. 2000). Following attack,
induced defenses rise to a peak and typically decline
again, with different components rising and falling over
different time scales. Similarly, different defenses
show different spatial patterns of induction. Thus, fol-
lowing induction a plant becomes a mosaic of tissues
with different patterns of induction which change over
time. Both the spatial and temporal dynamics of in-
duction are likely to be even more complex in response
to multiple attacks: a kaleidoscope rather than a mosaic.
Thus, if a molecular perspective can be extrapolated
to the field, we can hypothesize that plant defense is
extremely dynamic on relatively short time scales, cer-
January 2004 PHYTOHORMONAL ECOLOGY
tainly hours to days in many herbaceous plants. This
leads to the further prediction that the specific profile
of defenses expressed by a plant at any given time will
be a function of the timing and combination of attacks
it has experienced, perhaps over a period of several
weeks. Indeed, the potential for other elements of the
environment to influence elements of defense (e.g.,
drought [Thaler et al. 2004], salinity [Mopper et al.
2004], nitrogen [Baldwin 2001], and light [Cipollini
2004]) may add extra complexity to patterns of ex-
pression. Overall, as new techniques increase our abil-
ity to resolve the components of defense we would
expect increasing evidence that defense is highly dy-
namic in the context of the biotic and abiotic history
of the individual plant.
This model of defense may be hard to demonstrate
experimentally, although it may be amenable to newer
techniques in molecular biology such as microarrays
and real-time PCR, which provide the technology to
examine and compare expression profiles with high
throughput, and makes the examination of complex
temporal and spatial patterns of gene expression an
achievable goal (at least in species where extraction of
nucleic acids is simple and rapid [Box 1]). Such mo-
lecular analyses could be integrated with measurements
at different scales of organization. For example, we
have used biochemical measurements to relate the time
course of the induction of cell wall peroxidases in R.
obtusifolius with changes in leaf growth and herbivore
resistance (Moore et al. 2003). While we could not
extend this study to patterns of gene expression in R.
obtusifolius, this should certainly be possible in model
species such as Arabidopsis, where microarrays have
highlighted that genes associated with the cell wall are
commonly affected by pathogens or herbivores
(Schenk et al. 2000). The relationships between dy-
namic changes at different scales of organization
would, admittedly, be largely correlative, but even this
would expand our understanding of the real links be-
tween altered gene expression and pathogen or herbi-
vore response.
However, this model of kaleidoscopic defense is, in
our view, very different from that taken by most ecol-
ogists. What might the consequences of such defense
be for the organisms involved? For the plant, there may
be benefits in that multiple components of induced de-
fense are primed for very quick increases in re- two-way and three-way interactions. In particular, we
sponse to subsequent attack (Conrath et al. 2002). From
the consumers perspective, kaleidoscopic defense
would be an expression of the hosts phenotypic plas-
ticity in the face of attack, presenting a moving tar-
get that may be difficult to counter over a range of
time scales (Adler and Karban 1994). In the long term,
the dynamic deployment of multiple defenses by the
host may constrain adaptation by consumers to counter
resistance (Karban and Baldwin 1997). Perhaps more
intriguing is that a highly dynamic defense strategy
may limit consumers own phenotypic plasticity to
65
counteract host resistances. Such plasticity is evident,
for example, in the ability of some insects to modify
digestive enzymes to reduce the impact of proteinase
inhibitors (Cloutier et al. 2000, Paulillo et al. 2000),
or the induction in some pathogens of enzymes that
degrade phytoalexins (Morrissey and Osbourn 1999),
induced defense mechanisms (Heath 2002), or indeed
to alter constitutive antifungal compounds and use
them to interfere with induced defense responses
(Bouarab et al. 2002). It is intriguing that recent evi-
dence (Li et al. 2002) shows that some herbivores in-
duce detoxification systems not in response to defen-
sive end products but to signaling molecules (SA and
JA). Sensing signaling molecules may provide not only
an earlier cue for consumer countermeasures, but per-
haps also a more consistent and reliable cue than spe-
cific end products. For herbivores, we can consider both
compensatory feeding (Cloutier et al. 2000, Paulillo et
al. 2000) and food selection behavior (Stadler 1992)
as elements of plasticity. Such interplay of host and
consumer plasticity has rarely been considered (but see
Agrawal 2001), and the perspective offered by molec-
ular data may provide a focus for ecologists.
HOW CAN ECOLOGY INFORM POSSIBLE MOLECULAR
Special Feature
AND CELLULAR R ESEARCH?
The concept that plants present a highly dynamic
target for consumers over quite short time scales is not
limited to induced defense mechanisms and phytohor-
monal communication. This understanding that de-
fense in nature is multifaceted and variable is well
founded in ecological research. For example, leaf ni-
trogen and water content not only influence defenses
but also directly influence consumers (Mattson and
Scriber 1987). Amounts of these within the plant can
vary temporally (even diurnally) and spatially during
the life of the plant and between old and young leaves
(Denno and McClure 1983) and are affected both by
insect and fungus attack (Hatcher 1995). Similarly,
constitutive resistance mechanisms against insects and
fungi are affected by both the abiotic and biotic en-
vironment (e.g., Tinney et al. 1998).
While we would not argue that molecular studies
should try to embrace such complexity, we do believe
that they would benefit from the ecological understand-
ing of the contribution of the broader environment on
feel that many molecular and cellular biologists need
to reconsider the environmental conditions under which
they perform their experiments. This would help the
extrapolation of data from such experiments to the field
(Box 3) and may also address those areas of cross-talk
that are truly relevant in the field, as opposed to those
that are artifacts of unrealistic growing conditions. The
light environment offers a key example here (Cipollini
2004). Many molecular and cellular studies use plants
grown under low light, conditions that might be ex-
pected to influence defense in many ways. Low light
 66 PAUL E. HATCHER ET AL. Ecology, Vol. 85, No. 1

BOX 3. Can We Apply Cellular Approaches Effectively

to Understand HostConsumer Interactions in the Field?
We have shown in controlled environment (CE) experiments that localized feeding by Gastrophysa
viridula on Rumex obtusifolius leads to a reduction in the subsequent growth of undamaged leaves produced
after grazing ends, and that this is a result of a reduction in final epidermal cell areas (Moore et al. 2003).
This reduction in growth as a consequence of herbivore grazing was not linked to a reduction in assimilation
as a result of loss of leaf material, but was associated with an increase in the activity of a cell-wall-bound
peroxidase in the systemic leaf. We hypothesise that increased peroxidase activity leads to crosslinking
within the cell wall matrix that leads to stiffening of the cell wall, which both contributes to induced
resistance and restricts cell wall expansion. Hence, in this system at least, it appears that herbivore-induced
resistance imposes a significant cost by limiting future vegetative growth of the plant.
To test whether the same systemic effects of G. viridula occurred in field-grown R. obtusifolius, we
carried out essentially the same experiment as that detailed in Box 2 and in Moore et al. (2003) but using
plants sown, and grown, in pots outside. Several seeds were sown in John Innes No. 2 compost (Keith
Singleton, Egremont, UK) in 12 cm diameter pots in an open, unshaded site at Lancaster University,
Lancaster, UK (548 N, 28 W). Seedlings were thinned at the first-leaf stage to one per pot. The herbivore
treatment was applied essentially as described in Moore et al. (2003). The experiment was repeated four
times. In all cases, G. viridula was applied just prior to the emergence of leaf 8, but the time at which
this occurred varied between experiments depending on overall plant growth rate. The field site was
isolated from natural Rumex populations by at least 20 m, but no other attempts were made to protect
plants from naturally occurring herbivores or pathogens.
Table 1 details measurements taken in the four experiments from leaf 8 of R. obtusifolius following
grazing of leaf 4 by G. viridula. There was marked variation in growth between experiments in both grazed
Special Feature

plants and controls, which was related to the difference in season and weather conditions. In experiments
2 and 4, there was a significant reduction in systemic leaf growth due to reduced cell size, and this was
associated with an increase in cell wall bound peroxidase activity. Thus, these two field experiments
confirmed the pattern of results seen from experiments carried out in a controlled environment. By contrast,
in experiments 1 and 3, herbivory did not lead to a significant reduction in subsequent leaf growth, nor
to any significant induction of peroxidase activity.
Thus, the mechanism of reduced leaf growth that occurs in CE experiments does act in the field. This
occurred despite a low level background of leaf damage due to wind movement and random herbivory,
and, inevitably in the field, marked day-to-day variation in temperature, incident radiation, rainfall, etc.
However, it is clear that the significant induction of peroxidase is not consistent under field conditions.
Whether herbivore effects were significant or not was not related to overall plant growth rates (growth
was much slower in experiments 1 and 2 than experiments 3 and 4). We are currently exploring whether
the presence/absence of significant herbivore effects can be correlated with any specific climatic variable.

TABLE 1. Effects of grazing by Gastrophysa viridula (beetle) on final leaf area and peak
peroxidase (PX) activity in the eighth leaf of Rumex obtusifolius across four field experiments.

Experiment 1 Experiment 2 Experiment 3 Experiment 4

Parameter July 2001 Sep. 2001 July 2002 July 2002
Final leaf area (cm2)
Control 40.15 6 0.37 36.11 6 0.30 132.83 6 0.70 101.11 6 0.48
Beetle 36.52 6 0.19 27.29 6 0.17* 144.20 6 0.72 85.25 6 0.27*
Peak PX activity (nmol tetraguaicolmg fresh mass21h21)
Control 0.04 6 0.02 0.60 6 0.15 1.76 6 0.39 0.31 6 0.35
Beetle 0.04 6 0.01 1.56 6 0.42* 1.75 6 0.54 0.92 6 0.22*
Note: Each value is the mean 6 1 SE of 11 replicates.
* P , 0.05 compared to control; otherwise P . 0.05.
January 2004 PHYTOHORMONAL ECOLOGY
TABLE 2. Complex interactions: a selection of other current multiple-organism fields of in-
vestigation involving herbivores, fungi, and plants. These studies have been carried out
largely at the ecological or ecophysiological level.
Organisms involved
Plantmycorrhizaphytopathogen
Plantmycorrhizainsect herbivore
Plantendophyteinsect herbivore
Plantmycorrhizaendophyteinsect herbivore
Plantphytopathogeninsect herbivoreinsect parasitoid
is known to influence constitutive defenses, for ex-
ample, constitutive levels of furanocoumarins in pars-
nip were reduced by plant shading, although their in-
ducibility by insect attack was not affected (Berenbaum
and Zangerl 1999). In addition, there is increasing ev-
idence that high irradiances within the environmental
range can induce local and systemic changes in the
expression of genes usually associated with defense,
and that such induction can interact with responses to
disease or wounding (Mullineaux et al. 2000). In ad-
dition, solar ultraviolet radiation (UVR), which is much
reduced in glasshouse and controlled-environment
studies, is known to regulate the expression of many
defense-related genes (Mackerness 2000). While we
understand that there may be constraints on lighting in
some molecular studies, mechanistic studies conducted
at more realistic PAR levels would be a valuable step
in linking molecular and ecological studies of tripartite
interactions.
CONCLUSIONS
In assessing the future development of studies of the
phytohormonal ecology of plant defense the ever-in-
creasing complexity of the topic is concerning. Mo-
lecular and cellular studies are pointing to an increasing
number of defense signaling systems, and this inherent
complexity is compounded by cross-talk between path-
ways. Ecological studies now demonstrate a different
order of complexity in terms of interactions between
the many groups of consumers that exploit plants (Ta-
ble 2) and the extension of phytohormone-mediated
signaling from within plants to between plants of the
same (Bruin and Dicke 2001) or different species (Kar-
ban et al. 2000). Ecology is often confronted with the
need to distill simplicity from apparent complexity.
This is certainly a key requirement if we are to advance
our understanding of plant interactions with multiple
enemies: there is a pressing need here to distill a few
simple generalizations from molecular biology that can
ultimately be used to inform ecological studies. We
have sought to identify some generalizations here, but
are constrained by the lack of a middle ground between
molecular/cellular studies and ecology. This middle
ground is clearly at the whole-plant level of organi-
zation, and would address such issues as the integration
of diverse components of defense, including aspects
such as nutrition, as well as more specific anticonsumer
67
Recent reference
Gernns et al. (2001)
Gange et al. (2002)
Saikkonen et al. (1998)
Vicari et al. (2002)
Cardoza et al. (2002)
products, into whole-tissue or whole-plant resistance
as perceived by the consumer. Understanding the ef-
fects of defense-related phytohormones at this whole
plant level is clearly central to such understanding, and
poses an exciting challenge for future research.
ACKNOWLEDGMENTS
We thank the Natural Environment Research Council of
Great Britain for funding our research into Rumex tripartite
interactions; and Susan Mopper, Anurag Agrawal, and two
anonymous referees for helpful, constructive comments on
drafts of this paper.
LITERATURE CITED
Adler, F. R., and R. Karban. 1994. Defended fortresses or
Special Feature
moving targetsanother model of inducible defenses in-
spired by military metaphors. American Naturalist 144:
813832.
Agrawal, A. A. 2001. Phenotypic plasticity in the interactions
and evolution of species. Science 294:321326.
Ahlholm, J., M. Helander, P. Elamo, I. Saloniemi, S. Neu-
vonen, S. Hanhimaki, and K. Saikkonen. 2002. Micro-
fungi and invertebrate herbivores on birch trees: fungal
mediated plantherbivore interactions or responses to host
quality? Ecology Letters 5:648655.
Baldwin, I. T. 2001. An ecologically motivated analysis of
plantherbivore interactions in native tobacco. Plant Phys-
iology 127:14491458.
Barbosa, P. 1991. Plant pathogens and nonvector herbivores.
Pages 341382 in P. Barbosa, V. A. Krischik, and C. G.
Jones, editors. Microbial mediation of plantherbivore in-
teractions. Wiley, New York, New York, USA.
Baudoin, A. B. A. M., R. G. Abad, L. T. Kok, and W. L.
Bruckart. 1993. Field evaluation of Puccinia carduorum
for biological control of musk thistle. Biological Control
3:5360.
Berenbaum, M. R., and A. R. Zangerl. 1999. Coping with
life as a menu option: inducible defenses of the wild pars-
nip. Pages 1032 in R. Tollrian and C. D. Harvell, editors.
The ecology and evolution of inducible defenses. Princeton
University Press, Princeton, New Jersey, USA.
Bostock, R. M., R. Karban, J. S. Thaler, P. D. Weyman, and
D. Gilchrist. 2001. Signal interactions in induced resis-
tance to pathogens and insect herbivores. European Journal
of Plant Pathology 107:103111.
Bouarab, K., R. Melton, J. Peart, D. Baulcombe, and A. Os-
bourn. 2002. A saponin-detoxifying enzyme mediates sup-
pression of plant defences. Nature 418:889892.
Bruin, J., and M. Dicke. 2001. Chemical information transfer
between wounded and unwounded plants: backing up the
future. Biochemical Systematics and Ecology 29:1103
1113.
Cameron, R. K. 2000. Salicylic acid and its role in plant
defense responses: What do we really know? Physiological
and Molecular Plant Pathology 56:9193.
 68 PAUL E. HATCHER ET AL.
Cardoza, Y. J., H. T. Alborn, and J. H. Tumlinson. 2002. In
vivo volatile emissions from peanut plants induced by si-
multaneous fungal infection and insect damage. Journal of
Chemical Ecology 28:161174.
Cipollini, D. 2004. Stretching the limits of plasticity: Can a
plant defend against both competitors and herbivores?
Ecology 85:2837.
Cloutier, C., C. Jean, M. Fournier, S. Yelle, and D. Michaud.
2000. Adult Colorado potato beetles, Leptinotarsa decem-
lineata compensate for nutritional stress on oryzacystatin
I-transgenic potato plants by hypertrophic behavior and
over-production of insensitive proteases. Archives of Insect
Biochemistry and Physiology 44:6981.
Conrath, U., C. M. J. Pieterse, and B. Mauch-Mani. 2002.
Priming in plantpathogen interactions. Trends in Plant Sci-
ence 7:210216.
Cui, J., G. Jander, L. R. Racki, P. D. Kim, N. E. Pierce, and
F. M. Ausubel. 2002. Signals involved in Arabidopsis re-
sistance to Trichoplusia ni caterpillars induced by virulent
and avirulent strains of the phytopathogen Pseudomonas
syringae. Plant Physiology 129:551564.
de Bruxelles, G. L., and M. R. Roberts. 2001. Signals reg-
ulating multiple responses to wounding and herbivores.
Critical Reviews in Plant Science 20:487521.
Denno, R. F., and M. S. McClure, editors. 1983. Variable
plants and herbivores in natural and managed systems. Ac-
ademic Press, New York, New York, USA.
de Nooij, M. P., A. Biere, and E. G. A. Linders. 1992. In-
teraction of pests and pathogens through host predisposi-
tion. Pages 143160 in P. G. Ayres, editor. Pests and path-
Special Feature
ogens: plant responses to foliar attack. BIOS, Oxford, UK.
Ericson, L., and A. Wennstrom. 1997. The effect of herbivory
on the interaction between the clonal plant Trientalis eu-
ropaea and its smut fungus Urocystis trientalis. Oikos 80:
107111.
Felton, G. W., K. L. Korth, J. L. Bi, S. V. Wesley, D. V.
Huhman, M. C. Mathews, J. B. Murphy, C. Lamb, and R.
A. Dixon. 1999. Inverse relationship between systemic re-
sistance of plants to microorganisms and to insect herbiv-
ory. Current Biology 9:317320.
Gange, A. C., P. G. Stagg, and L. K. Ward. 2002. Arbuscular
mycorrhizal fungi affect phytophagous insect specialism.
Ecology Letters 5:1115.
Gernns, H., H. von Alten, and H. M. Poehling. 2001. Ar-
buscular mycorrhiza increased the activity of a biotrophic
leaf pathogenIs a compensation possible? Mycorrhiza
11:237243.
Hallett, S. G., P. Hutchinson, N. D. Paul, and P. G. Ayres.
1990. Conidial germination of Botrytis cinerea in relation
to aeciospores and aecia of groundsel rust (Puccinia la-
genophorae). Mycological Research 94:603606.
Hatcher, P. E. 1995. Three-way interactions between plant
pathogenic fungi, herbivorous insects and their host plants.
Biological Reviews 70:639694.
Hatcher, P. E., and N. D. Paul. 2001. Plant pathogenher-
bivore interactions and their effects on weeds. Pages 193
225 in M. J. Jeger and N. J. Spence, editors. Biotic inter-
actions in plantpathogen associations. CAB International,
Wallingford, UK.
Hatcher, P. E., N. D. Paul, P. G. Ayres, and J. B. Whittaker.
1997. The effect of nitrogen fertilization and rust fungus
infection, singly and combined, on the leaf chemical com-
position of Rumex obtusifolius. Functional Ecology 11:
545553.
Heath, M. C. 2002. Hypersensitive response-related death.
Plant Molecular Biology 44:321334.
Hermsmeier, D., U. Schittko, and I. T. Baldwin. 2001. Mo-
lecular interactions between the specialist herbivore Man-
duca sexta (Lepidoptera, Sphingidae) and its natural host
Nicotiana attenuata. I. Large-scale changes in the accu-
Ecology, Vol. 85, No. 1
mulation of growth- and defense-related plant mRNAs.
Plant Physiology 125:683700.
Jackson, R. B., C. R. Linder, M. Lynch, M. Purugganan, S.
Somerville, and S. S. Thayer. 2002. Linking molecular
insight and ecological research. Trends in Ecology and Evo-
lution 17:409414.
Karban, R., and I. T. Baldwin. 1997. Induced responses to
herbivory. University of Chicago Press, Chicago, Illinois,
USA.
Karban, R., I. T. Baldwin, K. J. Baxter, G. Laue, and G. W.
Felton. 2000. Communication between plants: induced re-
sistance in wild tobacco plants following clipping of neigh-
boring sagebrush. Oecologia 125:6671.
Karban, R., and J. H. Myers. 1989. Induced plant responses
to herbivory. Annual Review of Ecology and Systematics
20:331348.
Kessler, A., and I. T. Baldwin. 2002. Plant responses to insect
herbivory: the emerging molecular analysis. Annual Re-
view of Plant Biology 53:299328.
Kluth, S., A. Kruess, and T. Tscharntke. 2001. Interactions
between the rust fungus Puccinia punctiformis and ectopha-
gous and endophagous insects on creeping thistle. Journal
of Applied Ecology 38:548556.
Kok, L. T., R. G. Abad, and A. B. A. M. Baudoin. 1996.
Effects of Puccinia carduorum on musk thistle herbivores.
Biological Control 6:123129.
Kunkel, B. N., and D. M. Brooks. 2002. Cross talk between
signaling pathways in pathogen defense. Current Opinion
in Plant Biology 5:325331.
Lappalainen, J. H., and M. L. Helander. 1997. The role of
foliar microfungi in mountain birchinsect herbivore re-
lationships. Ecography 20:116122.
Leach, C. M., E. A. Dickason, and A. E. Gross. 1963. The
relationship of insects, fungi and nematodes to the dete-
rioration of roots of Trifolium hybridum L. Annals of Ap-
plied Biology 52:371385.
Leach, J. G. 1940. Insect transmission of plant diseases. Mc-
Graw-Hill, London, UK.
Lewis, A. C. 1979. Feeding preference for diseased and wilt-
ed sunflower in the grasshopper, Melanoplus differentialis.
Entomologia Experimentalis et Applicata 26:202207.
Li, X., M. A. Schuler, and M. R. Berenbaum. 2002. Jasmonate
and salicylate induce expression of herbivore cytochrome
P450 genes. Nature 419:712715.
Mackerness, S. A.-H. 2000. Plant responses to ultraviolet-B
(UV-B: 280320 nm) stress: What are the key regulators?
Plant Growth Regulation 32:2739.
Maleck, K., and R. A. Dietrich. 1999. Defense on multiple
fronts: How do plants cope with diverse enemies? Trends
in Plant Science 4:215219.
Mattson, W. J., and J. M. Scriber. 1987. Nutritional ecology
of insect folivores of woody plants: nitrogen, water, fiber,
and mineral considerations. Pages 105146 in F. Slansky
and J. G. Rodriquez, editors. Nutritional ecology of insects,
mites, spiders and related invertebrates. Wiley, New York,
New York, USA.
Moore, J. P., J. E. Taylor, N. D. Paul, and J. B. Whittaker.
2003. Reduced leaf expansion as a cost of systemic induced
resistance to herbivory. Functional Ecology 17:7581.
Mopper, S., Y. Wang, C. Criner, and K. Hasenstein. 2004.
Iris hexagona hormonal responses to salinity stress, leaf-
miner herbivory, and phenology. Ecology 85:3847.
Morrissey, J. P., and A. E. Osbourn. 1999. Fungal resistance
to plant antibiotics as a mechanism of pathogenesis. Mi-
crobial Molecular Biology Reviews 63:708724.
Mullineaux, P., L. Ball, C. Escobar, B. Karpinska, G. Creis-
sen, and S. Karpinski. 2000. Are diverse signaling path-
ways integrated in the regulation of Arabidopsis antioxidant
defence gene expression in response to excess excitation
January 2004 PHYTOHORMONAL ECOLOGY
energy? Philosophical Transactions of the Royal Society of
London. Series B: Biological Sciences 355:15311540.
Nault, L. R. 1997. Arthropod transmission of plant viruses:
a new synthesis. Annals of the Entomological Society of
America 90:521541.
ODonnell, P. J., C. Calvert, R. Atzorn, C. Wasternack, H.
M. O. Leyser, and D. J. Bowles. 1996. Ethylene as a signal
mediating the wound response of tomato plants. Science
274:19141917.
Oostendorp, M., W. Kunz, B. Dietrich, and T. Staub. 2001.
Induced disease resistance in plants by chemicals. Euro-
pean Journal of Plant Pathology 107:1928.
Paul, N. D., P. E. Hatcher, and J. E. Taylor. 2000. Coping
with multiple enemies: an integration of molecular and eco-
logical perspectives. Trends in Plant Science 5:220225.
Paulillo, L. C. M. S., A. R. Lopes, P. T. Cristofoletti, J. R. P.
Parra, W. R. Terra, and M. C. Silva-Filho. 2000. Changes
in midgut endopeptidase activity of Spodoptera frugiperda
(Lepidoptera: Noctuidae) are responsible for adaptation to
soybean proteinase inhibitors. Journal of Economic Ento-
mology 93:892896.
Penninckx, I. A. M. A., B. P. H. J. Thomma, A. Buchala, J.-
P. Metraux, and W. F. Broekaert. 1998. Concomitant ac-
tivation of jasmonate and ethylene response pathways is
required for induction of a plant defensin gene in Arabi-
dopsis. Plant Cell 10:21032113.
Ramsell, J., and N. D. Paul. 1990. Preferential grazing by
molluscs of plants infected by rust fungi. Oikos 58:145
150.
Reymond, P., and E. E. Farmer. 1998. Jasmonate and salic-
ylate as global signals for defence gene expression. Current
Opinion in Plant Biology 1:404411.
Reymond, P., H. Weber, M. Damond, and E. E. Farmer. 2000.
Differential gene expression in response to mechanical
wounding and insect feeding in Arabidopsis. Plant Cell 12:
707719.
Saikkonen, K., S. H. Faeth, M. Helander, and T. J. Sullivan.
1998. Fungal endophytes: a continuum of interactions with
host plants. Annual Review of Ecology and Systematics
29:319343.
Schenk, P. M., K. Kazan, I. Wilson, J. P. Anderson, T. Rich-
mond, S. C. Somerville, and J. M. Manners. 2000. Co-
ordinated plant defense responses in Arabidopsis revealed
by microarray analysis. Proceedings of the National Acad-
emy of Sciences (USA) 97:1165511660.
Schittko, U., D. Hermsmeier, and I. T. Baldwin. 2001. Mo-
lecular interactions between the specialist herbivore Man-
69
duca sexta (Lepidoptera, Sphingidae) and its natural host
Nicotiana attenuata. II. Accumulation of plant mRNAs in
response to insect-derived cues. Plant Physiology 125:701
710.
Schweizer, P., A. Buchala, R. Dudler, and J.-P. Metraux. 1998.
Induced systemic resistance in wounded rice plants. Plant
Journal 14:475481.
Stadler, E. 1992. Behavioral responses of insects to plant
secondary compounds. Pages 4588 in G. A. Rosenthal
and M. R. Berenbaum, editors. Herbivores, their interac-
tions with secondary plant metabolites. Second edition.
Volume II: ecological and evolutionary processes. Aca-
demic Press, San Diego, California, USA.
Thaler, J. S. 1999. Jasmonic acid mediated interactions be-
tween plants, herbivores, parasitoids and pathogens: a re-
view of field experiments in tomato. Pages 319334 in A.
A. Agrawal, S. Tuzun, and E. Bent, editors. Induced plant
defenses against pathogens and herbivores: biochemistry,
ecology, and agriculture. APS Press, St. Paul, Minnesota,
USA.
Thaler, J. S., and R. Bostock. 2004. Interactions between
abscisic-acid-mediated responses and plant resistance to
pathogens and insects. Ecology 85:4858.
Thaler, J. S., A. L. Fidantsef, and R. M. Bostock. 2002. An-
tagonism between jasmonate- and salicylate-mediated in-
duced plant resistance: effects of concentration and timing
of elicitation on defense-related proteins, herbivore, and
pathogen performance in tomato. Journal of Chemical
Ecology 28:11311159.
Tinney, G. W., S. C. Pritchard, R. Gonzalez, N. D. Paul, P.
Special Feature
E. Hatcher, and J. E. Taylor. 2002. Elimination of oxalate
contamination in RNA isolation from Rumex obtusifolius.
Plant Molecular Biology Reporter 20:309.
Tinney, G., C. Theuring, N. Paul, and T. Hartmann. 1998.
Effects of rust infection with Puccinia lagenophorae on
pyrrolizidine alkaloids in Senecio vulgaris. Phytochemistry
49:15891592.
Vicari, M., P. E. Hatcher, and P. G. Ayres. 2002. Combined
effect of foliar and mycorrhizal endophytes on an insect
herbivore. Ecology 83:24522464.
Walling, L. L. 2000. The myriad plant responses to herbi-
vores. Journal of Plant Growth Regulation 19:195216.
Wilding, N., N. M. Collins, P. M. Hammond, and J. F. Webber,
editors. 1989. Insectfungus interactions. Academic Press,
London, UK.
Wittstock, U., and J. Gerhenzon. 2002. Constitutive plant
toxins and their role in defense against herbivores and path-
ogens. Current Opinion in Plant Biology 5:300307.
 Ecology, 85(1), 2004, pp. 7077
q 2004 by the Ecological Society of America

CROSS-KINGDOM CROSS-TALK: HORMONES SHARED BY

PLANTS AND THEIR INSECT HERBIVORES
JACK C. SCHULTZ1 AND HEIDI M. APPEL
Department of Entomology, Penn State University, University Park, Pennsylvania 16802 USA

Abstract. Dynamic defense responses plants exhibit when attacked by insects are prob-
ably as important ecologically as constitutive defenses were once thought to be. Plants
recognize and respond to interspecific differences among insects and microbes, and some
insects appear to manipulate plant signaling and response systems. These interactions are
based on an exchange of signals among plant, insect, and microbes. Many of these signals
are hormones, and a surprising number are common to both kingdoms. Fatty-acid-based
hormones provide a particularly good example, and preliminary data suggesting that plant
defense responses can be mediated by common insect hormones and signals (prostaglandins)
are presented. The evolutionary and ecological implications of shared signals and signal-
stealing are diverse and can greatly improve our understanding of plantpest interactions.
Key words: eicosanoids; fatty acids; forest tent caterpillar; gypsy moth; hormones; oak; oxylipins;
plantherbivore interactions; prostaglandins; signaling.

INTRODUCTION invasive tactics of plant, and animal, pathogens are

Multicellular life could not have evolved without the quite similar (Wharam et al. 1995, Rossier et al. 1999,
chemical signaling necessary to coordinate interactions Cao et al. 2001).
Special Feature

among cells (Rasmussen et al. 1996, Stoka 1999, Kaiser
2001). Evidence is accruing to suggest that many in-
tercellular, eukaryotic signals (hormones) have been
adapted from signals originally used by prokaryotes to
coordinate interactions within populations of cells, for
example to regulate density or to locate nutrients (Stoka
1999, Blackstone 2000). In higher organisms, hor-
mones coordinate interactions among cells, tissues, and
organs, and produce whole-organism responses to the
environment. It is in this last sense that hormones have
great ecological significance, particularly in plantpest
interactions.
Dynamic defense responses exhibited by plants when
they are attacked by pests are probably as important
ecologically as constitutive defenses were once thought
to be (Kessler and Baldwin 2002, but see Gatehouse
2002). The specificity and importance of plant respons-
es to pathogens have been appreciated for a long time,
and are best represented by the gene-for-gene view
of plantpathogen compatibility, in which pathogen
identity and plant responses are based largely on pro-
tein differences (virulence factors) among the mi-
crobes (Ellis and Jones 1998, Bogdanov 2002, Fou-
longue et al. 2002, vant Slot and Knogge 2002). Plant
receptors (encoded by R genes) discriminate among
these proteins, signal the microbes identity, and elicit
the appropriate response from the plant (Ellis and Jones
1998). This kind of identification system is conserved
across kingdoms, in part because the biochemistry and
Manuscript received 12 November 2002; revised 28 February
2003; accepted 25 March 2002. Corresponding Editor: S. Mop-
per. For reprints of this Special Feature, see footnote 1, p. 3.
1 E-mail: ujq@psu.edu
70
Plants also recognize and respond differentially to
insects (Kessler and Baldwin 2002) using a complex,
interacting array of signal pathways (Gatehouse 2002).
This, plus the observation that some insects appear to
manipulate plant responses (Rhoades 1985, Eichenseer
et al. 1999, Felton and Korth 2000, Musser et al. 2002,
Schultz 2002), highlights the importance of signal ex-
change between plants and herbivores. Many of these
signals are hormones and a surprising number are com-
mon to both kingdoms. For some, a hormonal role is
only now being discovered in either plants or animals
or both.
We suggest that understanding ecological interac-
tions between plants and pests would be enhanced by
focusing attention on their commonalities, particularly
common signals, rather than their differences. While
plants, insects, and microbes are evolutionarily and bi-
ologically distant from one another, they have devel-
oped many interactive styles and signaling systems
that are quite similar. Moreover, all three groups often
interact simultaneously, as antagonists or sometimes as
allies, so that three-way communication is likely to be
a common event (e.g., Spiteller et al. 2000). The ability
to perceive, interpret, and manipulate each others sig-
nals is likely to have adaptive value, as is the ability
to deceive or mask signaling (Rhoades 1985). We sug-
gest that there is a limited set of interactive lan-
guages among plants, microbes, and insects, and that
the ecology and evolution of plantpest interactions can
be understood best by learning it.
PLANTS AND ANIMALS SHARE HORMONE SYSTEMS
Plants produce many chemicals that have physio-
logical function in animals. Often, it has been difficult
January 2004 PHYTOHORMONAL ECOLOGY
to identify primary roles for these molecules in the
plant, leading to use of the term secondary metabo-
lite (Rosenthal and Berenbaum 1991). For example,
all plants produce glutamate, a chemical messenger
with many important functions in animal nervous sys-
tems. In humans, functions of glutamate range from
acquiring and storing memories to suspected roles in
certain human mental health ailments. Since chemical
ecology developed in the 1960s, the common expla-
nation for the presence of such messengers in plants
has been that they function in interactions with her-
bivores or pathogens, i.e., that they are defensive (Ro-
senthal and Berenbaum 1991). This view was based on
the assumption that plants do not use such molecules
as signals. However, recent advances in molecular bi-
ology have led to the discovery of genes encoding glu-
tamate receptors in the Arabidopsis genome (and the
cognate protein in the plant), contradicting this view
(Chiu et al. 1999). Even though plants lack central
nervous systems, they are evidently using one of the
signals used in animal nervous systems. The impact of
glutamate or other plant signals on animals consuming
them may tell us how these putative defenses work in
plants (Tan and Ausubel 2000, Cao et al. 2001).
Even if plants and animals do produce related sig-
naling molecules, are they equipped to perceive them?
Many assume that plants, in particular, lack perceptual
abilities (e.g., see Lerdau 2002). But plants and animals
appear to share large families of extracellular receptors
that interact with many similar types of signals. Al-
though many of these receptors are well characterized
in animals, they are just beginning to be explored in
plants. Over 600 genes encoding putative receptor-like
kinases (RLKs) have been identified in the Arabidopsis
genome, and their structures appear to be conserved
across animal and plant kingdoms to a considerable
degree (Shiu and Bleecker 2001). Known functions in-
clude hormone perception (brassinosteroids, ethylene,
systemin), pathogen recognition, and self-recognition
(self-incompatibility) (Becraft 2002, Gomez-Gomez
and Boller 2002, Takayama and Sakagami 2002). It
seems likely that additional hormone receptors will be
found among these and related plant receptor families,
and that at least some will be able to perceive animal
hormones.
Similar hormonal systems in plant, animals, and mi-
crobes may be homologous or convergent (Bergey et
al. 1996). Plants and animals may have converged on
optimal hormonal systems that solve common sig- kingdom communication in shared fatty acid signaling
naling problems. Modeling optimal solutions to signal
transduction problems is a basis of rational drug design
(Hardcastle et al. 1995, Barril et al. 2002), and was
used to predict the chemistry of pheromonal signaling
among ants (Holldobler and Wilson 1990). Optimal
signals, signaling systems, or elements of signaling
systems may have evolved multiple times or been pre-
served in divergent organisms.
71
Whether homologous or convergent, sharing signal-
ing systems has emergent ecological and evolutionary
consequences; appreciating this may lead us to revise
some of our views of plantherbivore and plantpath-
ogen interactions. The need to recognize and interact
successfully with enemies and hosts is obvious. When
organisms using related hormonal systems interact, we
would expect individuals that can block or enhance
signal exchange to gain adaptive advantage. The co-
evolution of plants with pathogens is characterized by
ongoing modification of signal/recognition and should
include interference with plant perception or signaling
(Rhoades 1985). Some larval Lepidoptera activate dis-
ease-resistance signaling pathways in their host plants,
which has the effect of downregulating defenses
against insects (Felton and Korth 2000). Emission of
volatiles that attract parasitoids to plants attacked by
insect herbivores is cued by a diverse set of molecules
(e.g., volicitin) that act as signals in animals, plants,
and microbes (Krumm et al. 1995, Pohnert et al. 1999,
Spiteller et al. 2000, Turlings et al. 2000).
We suggest that two organisms sharing a common
signaling pathway are more likely to interact via that
pathway than by a mechanism unique to one or the
Special Feature
other. Evolutionary constraint, be it via common de-
scent or convergence on a particularly useful mecha-
nism, provides a predictive framework for understand-
ing the evolution and ecological interactions among
interacting species. Hence, a key to understanding eco-
logical interactions is to look for signaling common-
alities.
AN EXAMPLE: FATTY-ACID-BASED SIGNALING
Hormonal/signaling systems based on fatty acid ox-
idation products exemplify several of the points we
have raised. Their structure and some functions are
conserved from microbes through higher organisms,
possibly because they were originally derived from
plastid membranes of intracellular microbial symbionts
(Weber 2002). They have multiple functions, some of
which appear to be exaptations for interspecific
communication (Stoka 1999, Weber 2002). In plants,
the hormonal functions of fatty acid oxidation products
were discovered relatively recently and new roles are
still being discovered. Their impact on plantpest in-
teractions is so central that it now appears that herbi-
vores and pathogens may gain selective advantage by
manipulating their effects.
To understand the similarities and potential for cross-
pathways it is necessary to understand some of the
underlying chemistry. In plants, fatty acid signals are
called oxylipins (see Blee 2002, Fuessner and Was-
ternak 2002, Weber 2002 for helpful reviews) and are
produced via enzymatic oxidation of linoleic or lino-
lenic acid released from membranes by lipases (Creel-
man and Mullet 1997, Mueller 1997). One major prod-
uct of this pathway, jasmonic acid (JA), has received
 72 JACK C. SCHULTZ AND HEIDI M. APPEL Ecology, Vol. 85, No. 1

the bulk of attention by researchers because it and its

derivatives evidently modulate a wide range of eco-
logically important physiological events, including as-
pects of fruit ripening, production of viable pollen, se-
nescence, root growth and tuberization, stomatal resis-
tance, abscission, germination, tendril coiling, protein
synthesis and catabolism, and plant resistance to insects
and pathogens (Gardner 1995, Creelman and Mullet
1997, Kessler and Baldwin 2002). Many different oxy-
lipin structures are produced via oxidation of unsatu-
rated fatty acids by a suite of cytochrome P450, li-
poxygenase, and cyclo-oxygenase-like enzymes (Blee
2002). Any or all of these structures may be active
signals in plants, as may their fatty-acid conjugates and
a wide variety of synthetic structural variants (Gate-
house 2002, Weber 2002).
In animals, fatty acid signals are called eicosa-
noids and are functionally and structurally analogous
to plant oxylipins (Stanley 1999, Blee 2002). Eicosa-
noids play multiple roles in animal systems, including
regulating wound responses, reproductive physiology
and behavior, inflammation and immune responses,
platelet aggregation, cardiovascular function, and tu- FIG. 1. Structures of prostaglandin E2 (PGE2) from in-
mor growth, among others (van Ryn et al. 2000). They vertebrates used in this study, phytoprostane El (a plant pros-
Special Feature

taglandin), and jasmonic acid (a plant wound hormone).

include prostaglandins (PGs), prostacyclins, and
thromboxanes (all produced via cyclo-oxygenase
[COX] enzymes), as well as the lipoxygenase products
hydroperoxyeicosatetraenoic acids, hydroxyeicosate-
traenoic acids, and leukotrienes (Gardner 1995).
Fatty-acid-derived signals are good candidates for
plantinsect recognition and host plant manipulation
by herbivores. Volicitin is a fatty acid/amino acid con-
jugate produced in caterpillars and shown to elicit
plant-defense responses in some (Turlings et al. 2000),
but not all (Krumm et al. 1995), plants. Molecules of
this type are diverse and abundant in insect guts, where
their surfactant properties may aid nutrient absorption
(Deveau and Schultz 1992, Spiteller et al. 2000). The
structurally related prostaglandins are synthesized in
larval lepidopteran midguts (Buyukguzel et al. 2002)
and would likely be found in regurgitant (Fig. 1). The
potential for specific signaling between plants and her-
bivores via fatty acids is further suggested by the ob-
servations that volicitin does not elicit responses from
all plants, that many related structures may occur in
different herbivores, that diverse fatty acid-derived
structures (of any origin) can elicit diverse plant re-
sponses (Krumm et al. 1995, Pohnert et al. 1999), and
that some plants can use arachidonic acid (the basic
substrate for animal eicosanoid signals) to produce
analogous oxylipin signals (Yu et al. 1997). All studies
of fatty acid signaling between insect and plants thus
far have focused on elicitation of plant defenses, but
we suggest that the same family of signals could also
suppress defense responses in plants, as they do in
vertebrate hosts (Stanley 1999).
PROSTAGLANDINS AS CUES IN
PLANTHERBIVORE INTERACTIONS
Prostaglandins are diverse, well-studied, animal ei-
cosanoid hormones (Fig. 1). They are synthesized from
arachidonate in the cell membrane by the action of
phospholipase A2. Structure/activity variation is con-
ferred by substitutions to the pentane ring of the basic
20-carbon molecular skeleton. Two synthetic pathways,
the cyclo-oxygenase and lipoxygenase pathways, com-
pete with one another to form prostaglandins and
thromboxanes or leukotrienes, respectively. Because
they are lipid soluble, they can pass easily through cell
membranes.
Plants were not thought to make prostaglandins per
se until Imbusch and Mueller (2000b) recently discov-
ered a plant pathway synthesizing isoprostanes (Fig.
1). Isoprostanes are markers of oxidative stress in an-
imals, an observation now extended to plants (Imbusch
and Mueller 2000b). Other prostaglandins have been
found in plants (Groenewald and van der Westhuizen
1997, Imbusch and Mueller 2000a, b), but their sig-
naling activity there is unexamined.
Since animals produce a wide range of eicosanoid
structures that converge closely on plant hormones in
structure and function, we hypothesize that animal ei-
cosanoids, including prostaglandins, may convey her-
bivore identity to plants or mediate plant responses to
attack. This line of thought led us to carry out prelim-
inary experiments to see if a common animal eicosa-
noid is active in a plant system variably responsive to
insect feeding.
January 2004 PHYTOHORMONAL ECOLOGY
PROSTAGLANDINS AS POTENTIAL DEFENSE
MEDIATORS IN NORTHERN RED OAK
Herbivory comprises both wounding and chemical
signaling, and plant responses to the two are not the
same (Korth and Dixon 1997, Reymond et al. 2000,
Kessler and Baldwin 2002). Depending on the plant
and the insect, artificial wounding can produce stron-
ger, weaker, or the same response seen with live insects
(Lin et al. 1990, Korth and Dixon 1997, Kessler and
Baldwin 2002). The most logical explanation for these
differences is that insects provide some chemical cue
to the plant which either amplifies or suppresses the
plant response to wounding (Felton and Korth 2000).
Studies with salivary gland secretions (Musser et al.
2002), oral secretions (Halitschke et al. 2001), and day 7, all treated leaves were harvested by cutting at
regurgitant (Turlings et al. 2000) support this view.
Wounding elicits increased concentrations of proan-
thocyanidins in northern red oak (Quercus rubra L.)
leaves (Rossiter et al. 1988). On the other hand, de-
foliation by gypsy moth larvae (Lymantriidae: Lyman-
tria dispar L.) sometimes elicits similar responses
(Schultz and Baldwin 1982, Hunter and Schultz 1993,
1995) and sometimes does not (DAmico et al. 1998;
H. M. Appel and J. C. Schultz, unpublished data). Ros-
siter et al. (1988) found differential production of tan-
nins by oaks in response to tearing and to gypsy moth
chewing. Havill and Raffa (1999) found that poplar
responses to gypsy moth regurgitant differed from re-
sponses to actual herbivory.
Prostaglandin E2 (PGE2) (Fig. 1) is produced by the
midgut cells of Lepidoptera larvae (Buyukguzel et al.
2002) and is widely distributed among arthropods
(Stanley 1999). It is particularly abundant in the saliva
of blood-feeding species such as ticks and reduviids
(assassin bugs), where it suppresses the host wound
responses and promotes vasodilation (Stanley 1999). It
seemed reasonable to us that PGE2 might occur in re-
gurgitant or oral secretions of feeding caterpillars, 1999, Arnold and Schultz 2002, Stotz et al. 2002) and
where it could influence plant responses. Preliminary
ELISA assays suggest that PGE2 and related PGs can
be found in gypsy moth larval regurgitant (T. Lee, H.
M. Appel, and J. C. Schultz, unpublished data).
Methods
We carried out two experiments in which we eval-
uated the effect of PGE2 on phenolic defense responses
of wounded red oak seedlings and compared these re-
sponses with those to caterpillar regurgitant under sim-
ilar experimental conditions. For all three studies, we
grew half-sib red oak seedlings in the greenhouse in
20 3 50-cm pots containing MetroMix 250 (Scotts-
Sierra Horticultural Products, Marysville, Ohio, USA),
at a mean temperature of 258C under supplemental
lighting (peak 800 mE) with a 11:13 (light:dark) pho-
toperiod, until each seedling had four leaves about one-
half to three-quarters fully expanded. The soil con-
73
tained an initial fertilizer charge; we did not fertilize
further during the experiment.
To study PGE2 effects, leaves were damaged by run-
ning a pattern wheel once down each side of the midrib
every other day for 7 d, with subsequent woundings
added alongside each other towards the leaf margin
(Halitschke et al. 2001). A watch glass was held behind
the wheel to provide a firm surface against which to
wound. Immediately after wounding, each plant re-
ceived either water or an aqueous solution of PGE 2
(#P4172; Sigma, St. Louis, Missouri, USA). The water
and aqueous prostaglandin solution were applied to the
wound as five 1-mL droplets spaced evenly along its
length. All items contacting the plants were rinsed with
ethanol and dried between treatment of each plant. On
the petiole, pooled, and frozen in liquid nitrogen.
The two PG experiments differed in three ways. In
the first experiment (September 2001), 0.5 mmol/L
PGE2 was used and the droplets were left intact. In the
second experiment (June 2002), 1.0 mmol/L PGE2 was
used and the droplets were gently spread along the
wound using a fine camel hair brush, because we ob-
served in experiment 1 that little fluid appeared to enter
Special Feature
the wound. We have observed that induction of con-
densed tannins in Q. rubra is local; it occurs primarily
or only in wounded leaves (Hunter and Schultz 1995;
H. M. Appel and J. C. Schultz, unpublished data). In
the first experiment, half of the leaves on each plant
were treated; we performed statistical analyses on those
leaves and on all leaves combined from each seedling.
In the second experiment, all leaves on each tree were
treated and analyzed together. Sample sizes differed
within and between experiments (Fig. 2). Because we
hypothesized that PGs may be perceived by plants in
a manner similar to perception of octadecanoids, we
applied concentrations similar to those used in studies
in which JA has been applied (e.g., Havill and Raffa
considerably smaller total amounts per plant than many
(e.g., Agrawal et al. 1999, Thaler et al. 2001).
The regurgitant (RG) experiment was conducted ex-
actly as the second PG experiment, but five 1-mL drop-
lets of either full strength gypsy moth or forest tent
caterpillar (Lasiocampidae: Malacosoma disstria L.)
regurgitant were applied with a pipettor. There were 10
or 11 trees in each treatment class. Regurgitant from
fourth and fifth instar larvae fed red oak leaves was
collected by gently massaging caterpillars until they
regurgitated into microcentrifuge tubes on ice. Samples
were centrifuged at 48C to remove particulate matter
and stored at 2808C until use.
Plant samples were stored at 2208C, freeze-dried,
and ground in a UDY cyclone mill to reduce heating.
Tannins were extracted from the samples in a cold so-
lution of 70% acetone in 1 mmol/L ascorbate, and the
extracts measured for proanthocyanidins (condensed
tannins, butanol/HCl assay) as described in Appel et
 74 JACK C. SCHULTZ AND HEIDI M. APPEL
Special Feature
FIG. 2. Means (61 SE) of condensed tannin concentra-
tions in red oak leaves after 7 d untouched (CONTROL) or
after 7 d of wounding (WOUND) and wounding with the
addition of prostaglandin E2 (W 1 PGE2). (A) Experiment 1,
using 0.5 mmol/L PGE2. (B) Experiment 2, using 1.0 mmol/L
PGE2. (C) Condensed tannin concentrations in leaves of red
oak after 7 d untouched (CONTROL), with wounding
(WOUND), with wounding plus 5 mL of gypsy moth regur-
gitant (W 1 GMREG) or 5 mL of forest tent caterpillar re-
gurgitant (W 1 FTCREG). Bars with different letters are
statistically distinguishable at P 5 0.05; letters with asterisks
differ at P 5 0.058. Sample sizes are given at the bases of
bars.
al. (2001). Standards for these colorimetric assays were
prepared by purifying tannins from leaves of plants in
each treatment (Appel et al. 2001). Significant variation
in mean concentrations of each chemical class among
treated and control plants was characterized separately
Ecology, Vol. 85, No. 1
for each experiment using the GLM procedure (SAS
Institute, Cary, North Carolina, USA) followed by Tu-
keys hsd analysis to identify significant differences
among treatments.
Results
As expected, in both experiments, leaves of wounded
plants contained increased concentrations of condensed
tannins after 7 d, compared with leaves from unwound-
ed controls (Fig. 2A and B). In experiment 1, all leaves
on wounded plants had ;55% greater mean concen-
trations of condensed tannins after 7 d than did un-
wounded controls; concentrations in wounded leaves
on those plants increased ;78% above leaves on con-
trol plants. In experiment 2, condensed tannin concen-
trations were generally greater, and wounding increased
them ;50%. The wound effect was significant using
Tukeys hsd test at a level of P 5 0.05 for both sets
of leaves in experiment 1, and at P 5 0.058 in exper-
iment 2.
Wound responses were suppressed when PGE2 was
added to wounds in both experiments (Fig. 2A and B).
The significance estimate for overall treatment effects
(wounding and PGE addition) was below or near P 5
0.05 for all three data sets (experiment 1 all leaves, SS
5 0.013, F2,35 5 3.31, P 5 0.049; experiment 1 treated
leaves, SS 5 0.021, F2,22 5 4.81, P 5 0.019; experiment
2 all leaves, SS 5 0.040, F2,18 5 3.53, P 5 0.053). PGE2
application held wound-induced increases in mean con-
densed tannin concentrations to 31% (compared with
50%) in all leaves and to 52% (compared with 78%)
in treated leaves in experiment 1. Wound-induced con-
densed tannin increases were 2% (compared with 50%)
when PGE2 was applied in experiment 2. These sup-
pressive effects of PGE were significant at P 5 0.050
in experiment 1 (both leaf sets) and P 5 0.058 in ex-
periment 2 using Tukeys hsd test. The marginal P value
for the second experiment (despite greater numerical
impact) was probably due to its smaller samples sizes
(not enough seedlings grew sufficiently well for inclu-
sion as unwounded controls).
In experiment 3, leaves of wounded plants accu-
mulated condensed tannin concentrations 67% greater
than leaves of unwounded controls after 7 d (Fig. 2C).
Gypsy moth and tent caterpillar RG reduced those in-
creases to 18% and 2%, respectively. Leaves treated
with either gypsy moth or forest tent caterpillar regur-
gitant had condensed tannin concentrations that were
not significantly different from unwounded controls af-
ter 7 d, according to the Tukey hsd analysis (at P 5
0.05; full model, SS 5 0.04, F2,55 5 11.2, P , 0.0001).
These preliminary results suggest that something in
caterpillar regurgitant suppresses oak seedling wound
responses, at least in terms of phenylpropanoid syn-
thesis. This suppressive impact, especially of tent cat-
erpillar RG, was similar to that of similarly applied 1
mmol/L PGE2, although the small sample size of our
PGE application experiments reduced our ability to
January 2004 PHYTOHORMONAL ECOLOGY
achieve equivalent statistical confidence. These results
support the concept that the suppressive factor might
be one or more eicosanoids, among them prostaglan-
dins. These studies were designed merely to provide
support for this concept, and many details remain to
be addressed. Since we applied very little prostaglandin
signal (less than has been used in comparable experi-
ments with JA or other fatty acid signals, e.g., Krumm
et al. 1995) and because PGE2 has low solubility in
water, we suspect that very little actually entered the
plants. There was no visible evidence of overt toxicity
to any plant. Experiments with radiolabeled PGs will
eventually be necessary to determine in-plant titers. No
one has yet determined how much regurgitant is de-
posited by feeding caterpillars.
In our experiments, PGE2 attenuated the hosts
wound response, much as saliva containing PGE2 is
thought to do for blood-feeding insects and caterpillar
regurgitant did in experiment 3. We would expect her-
bivore regurgitant to contain a more complex mixture
of potential fatty-acid-based signals; this would make
the impact of any species mixture difficult to predict
from the action of one component. Moreover, inter-
actions between insect signals and plant chemistry
could influence fatty acid signal transmission (e.g., De-
veau and Schultz 1992). Regurgitant is also likely to
contain enzymes capable of inactivating fatty acid sig-
nals (J. H. Tumlinson, personal communication). Our
finding that PGE2 can modify wound responses in oak
leaves is consistent with the view that parasites and
hosts, herbivores and plants could interact using shared
fatty-acid-based signaling systems.
ECOLOGICAL AND EVOLUTIONARY IMPLICATIONS
There are as many potential consequences of cross-
kingdom hormonal interactions as there are functions
regulated by shared hormone classes. Plants and insects
(and often microbes) share peptide, arabinogalactan,
steroidal, mono- and polyamine, auxin-like, cytokinin/
nucleotide, reactive oxygen species, and nitric-oxide-
based signaling systems (Schultz 2002). These signals
control functions ranging from morphogenesis to se-
nescence, from immunity to reproduction, and provide
recognition of friend and foe. New examples of signals
shared by plants and animals continue to arise (e.g.,
Kuklin and Conger 1995, Dinan et al. 1999, Imbusch
and Mueller 2000b, Berna et al. 2002, Lindsey et al.
2002).
The fact that mechanisms underlying responsiveness
to the biotic and abiotic environment can be shared by
plants and animals (and microbes) means that each
brings to its interactions with the other(s) a form of
information that can be exploited. Since plants rec-
ognize their own signals, they may be able to recognize
similar ones in insects and use them to initiate appro-
priate defense responses. Li et al. (2002) recently found
that both SA and JA elicit cytochrome P450 activity
in lepidopteran herbivores, and interpreted this as sig-
75
nal-stealing by the insect. Successful symbioses also
require subtly altering the antagonistic signaling nor-
mally done between enemies (Kistner and Pamiske
2002). Avoiding detection and intercepting signals
ought to be valuable to plant, pest, and symbiont; co-
evolution does not comprise merely developing and
adapting to new classes of defenses.
If plants can regulate their own signaling systems
(e.g., producing inhibitors or promoters), they may be
able to manipulate the related signaling systems of her-
bivores. For example, plants modulate their own oc-
tadecanoid signaling with salicylate, a modified form
of which modulates eicosanoid signaling in animals
(Pierpoint 2002). Blocking PG production in insects
has the potential to interfere with immunity to patho-
gens, reproduction, and development (Stanley 1999).
Several phenolics, for example resveratrol, curcumin,
quercetin, and salicylate, block prostaglandin synthesis
by inhibiting cyclo-oxygenase activity (Moreno 2000,
Pierpoint 2002). Lysophospholipids such as lysophos-
phoethanolamine, which are common midgut surfac-
tants in insect herbivores, are also phospholipase in-
hibitors, and could deprive both the octadecanoid and
the eicosanoid biosynthetic pathways of needed sub-
Special Feature
strates (Ryu et al. 1997). Plant anacardic acids, which
are themselves fatty acid derivatives, inhibit lipoxy-
genase and prostaglandin endoperoxide synthase
(Grazzini et al. 1991). No doubt plants are feeding
insect herbivores many factors that may jam the
eicosanoid hormonal system.
We usually try to understand ecological and evolu-
tionary interactions between plants and herbivores in
terms of defensive secondary metabolites and adaptive
responses to them (e.g., Rosenthal and Berenbaum
1991), often with limited success. We suggest that
much of the interaction and adaptation between these
groupsas is true for plants and microbeshas more
to do with signaling, signal reception, and signal in-
terference. Many defensive secondary metabolites
can have a direct impact on enemies, but their existence
and impacts may be due as well to ancestral or con-
vergent roles as hormones. We can learn a lot about
how plants, animals, and microbes interact via com-
parative study of their signaling systems, but only if
we focus on similarities as well as differences. Plant
and animal signaling systems offer many avenues for
exploitation and resistance, many of which are not pres-
ent until elicited by an environmental stimulus or en-
emy and a conversation begins.
ACKNOWLEDGMENTS
This work was supported by NSF grant DEB-9974067 and
by the Defense Advanced Research Projects Agency (DAR-
PA). Thanks to Chrissy Cohen, James Ebertowski, Sharie
Ketcho, Sue Krzysik, and Theresa Lee for technical assis-
tance.
LITERATURE CITED
Agrawal, A. A., S. Y. Strauss, and M. J. Stout. 1999. Costs
of induced responses and tolerance to herbivory in male
 76 JACK C. SCHULTZ AND HEIDI M. APPEL
and female fitness components of wild radish. Evolution
53:10931104.
Appel, H. M., H. L. Govenor, M. DAscenzo, E. Siska, and
J. C. Schultz. 2001. Limitations of Folin assays of foliar
polyphenols in ecological studies. Journal of Chemical
Ecology 27:761778.
Arnold, T. M., and J. C. Schultz. 2002. Induced sink strength
as a prerequisite for induced tannin biosynthesis in devel-
oping leaves of Populus. Oecologia 130:585593.
Barril, X., S. G. Kalko, M. Orozco, and F. J. Luque. 2002.
Rational design of reversible acetylcholinesterase inhibi-
tors. Mini Revues in Medicinal Chemistry 2(1):2736.
Becraft, P. W. 2002. Receptor kinase signaling in plant de-
velopment. Annual Review of Cell and Developmental Bi-
ology 18:163192.
Bergey, D. R., G. A. Howe, and C. A. Ryan. 1996. Poly-
peptide signaling for plant defensive genes exhibits anal-
ogies to defense signaling in animals. Proceedings of the
National Academy of Science (USA) 93:1205312058.
Berna, A., F. Bernier, K. Scott, and B. Stuhlmuller. 2002.
Ring up the curtain on DING proteins. FEBS Letters 524:
610.
Blackstone, N. W. 2000. Redox control and the evolution of
multicellularity. BioEssays 22:947953.
Blee, E. 2002. Impact of phyto-oxylipins in plant defense.
Trends in Plant Science 7:315321.
Bogdanov, A. J. 2002. Pto update: recent progress on an
ancient plant defence signalling pathway. Molecular Plant
Pathology 3:283288.
Buyukguzel, K., H. Tunaz, S. M. Putnam, and D. Stanley.
Special Feature
2002. Prostaglandin synthesis by midgut tissue isolated
from the tobacco hornworm, Manduca sexta. Insect Bio-
chemistry and Molecular Biology 32:435443.
Cao, H., R. L. Baldini, and L. G. Cao. 2001. Common mech-
anisms for pathogens of plants and animals. Annual Revue
of Phytopathology 39:259284.
Chiu, J., R. DeSalle, H. M. Lam, L. Meisel, and G. Coruzzi.
1999. Molecular evolution of glutamate receptors: a prim-
itive signaling mechanism that existed before plants and
animals diverged. Molecular Biology and Evolution 16:
826838.
Creelman, R. A., and J. E. Mullet. 1997. Biosynthesis and
action of jasmonates in plants. Annual Revue of Plant Phys-
iology and Plant Molecular Biology 48:355381.
DAmico, Y., J. S. Elkinton, G. Dwyer, R. B. Willis, and M.
E. Montgomery. 1998. Foliage damage does not affect
within-season transmission of an insect virus. Ecology 79:
11041110.
Deveau, E. J. I., and J. C. Schultz. 1992. Reassessment of
the interaction between gut detergents and tannins in Lep-
idoptera and significance for gypsy moth larvae. Journal
of Chemical Ecology 18:14371453.
Dinan, L. S. Tamara, W. Pensri, and S. D. Sarker. 1999. Plant
natural products as insect steroid receptor agonists and an-
tagonists. Pesticide Science 55:331335.
Eichenseer, R., M. C. Mathews, J. L. Bi, J. B. Murphy, and
G. W. Felton. 1999. Glucose oxidase: multiple functions
for Helicoverpa zea L.? Archives of Insect Biochemistry
and Physiology 42:99109.
Ellis, J., and D. Jones. 1998. Structure and function of pro-
teins controlling strain-specific pathogen resistance in
plants. Current Opinion in Plant Biology 1:288293.
Felton, G. W., and K. L. Korth. 2000. Trade-offs between
pathogen and herbivore resistance. Current Opinion in
Plant Biology 3:309314.
Foulongne, V., S. Michaux-Charach, D. OCallghan, and M.
Ramuz. 2002. Type IV secretion systems and bacterial vir-
ulence. M S-Medicine Sciences 18:439447.
Fuessner, I., and C. Wasternack. 2002. The lipoxygenase
pathway. Annual Review of Plant Biology 53:275297.
Ecology, Vol. 85, No. 1
Gardner, R. W. 1995. Biological roles and biochemistry of
the lipoxygenase pathway. Hortscience 30:197205.
Gatehouse, J. A. 2002. Plant resistance towards insect her-
bivores; a dynamic interaction. New Phytologist 156:145
169.
Gomez-Gomez, L., and T. Boller. 2002. Flagellin reception:
a paradigm for innate immunity. Trends in Plant Science
7:251256.
Grazzini, R., D. Hesk, E. Heininger, G. Hildenbrandt, C. C.
Reddy, D. Cox Foster, J. Medford, R. Craig, and R. O.
Mumma. 1991. Inhibition of lipoxygenase and prostaglan-
din endoperoxide synthase by anacardic acids. Biochemical
Biophysical Research Communications 176:775780.
Groenewald, E. G., and A. J. van der Westhuizen. 1997. Pros-
taglandins and related substances in plants. Botanical Re-
vues 63:199220.
Halitschke, R., U. Schittko, G. Pohnert, W. Boland, and I. T.
Baldwin. 2001. Molecular interactions between the spe-
cialist herbivore Manduca sexta (Lepidoptera, Sphingidae)
and its natural host Nicotiana attenuata. III. Fatty acid ami-
no acid conjugates in herbivore oral secretions are neces-
sary and sufficient for herbivore specific plant responses.
Plant Physiology 125:711717.
Hardcastle, I. R., M. G. Rowlands, J. Houghton, I. B. Parr,
G. A. Potter, M. Jarman, K. J. Edwards, C. A. Laughton,
J. O. Trent, and S. Neidle. 1995. Rationally designed an-
alogues of tamoxifen with improved calmodulin antago-
nism. Journal of Medicinal Chemistry 38:241248.
Havill, N. P., and K. F. Raffa. 1999. Effects of elicitation
treatment and genotypic variation on induced resistance in
Populus: impacts on gypsy moth (Lepidoptera: Lymantri-
idae) development and feeding behavior. Oecologia 120:
295303.
Holldobler, B., and E. O. Wilson. 1990. The ants. Harvard
University Press, Cambridge, Massachusetts, USA.
Hunter, M. D., and J. C. Schultz. 1993. Induced plant de-
fenses breached? Phytochemical induction protects an her-
bivore against disease. Oecologia 94:195203.
Hunter, M. D., and J. C. Schultz. 1995. Fertilization mediates
within-tree phytochemical induction by the gypsy moth in
oaks. Ecology 76:12261232.
Imbusch, R., and M. Mueller. 2000a. Analysis of oxidative
stress and wound inducible dinor isoprostanes F-1 in plants.
Plant Physiology 124:12931303.
Imbusch, R., and M. J. Mueller. 2000b. Formation of iso-
prostane F-2 like compounds (phytoprostanes F-1) from
alpha linolenic acid in plants. Free Radical Biology in Med-
icine 28:720726.
Kaiser, D. 2001. Building a multicellular organism. Annual
Revue of Genetics 35:103123.
Kessler, A., and I. T. Baldwin. 2002. Plant responses to insect
herbivory: the emerging molecular analysis. Annual Revue
of Plant Biology 53:299328.
Kistner, C., and M. Parniske. 2002. Evolution of signal trans-
duction in intracellular symbiosis. Trends in Plant Science
7:511518.
Korth, K. L., and R. A. Dixon. 1997. Evidence for chewing
insect specific molecular events distinct from a general
wound response in leaves. Plant Physiology 115:1299
1305.
Krumm, R., K. Bandemer, and W. Boland. 1995. Induction
of volatile biosynthesis in the Lima bean (Phaseolus lu-
natus) by leucine and isoleucine conjugates of 1-oxo-and
1 hydroxyindan-4-carboxylic acid: evidence for amino acid
conjugates of jasmonic acid as intermediates in the octa-
decanoid signaling pathway. FEBS Letters 377:523529.
Kuklin, A. I., and B. V. Conger. 1995. Catecholamines in
plants. Journal of Plant Growth Regulation 14:9197.
Lerdau, M. 2002. Plants talkbut can they listen? Science
298:361.
January 2004 PHYTOHORMONAL ECOLOGY
Li, X., M. A. Schuler, and M. R. Berenbaum. 2002. Jasmonate
and salicylate induce expression of herbivore cytochrome
P450 genes. Nature 419:712713.
Lin, H., M. Kogan, and D. Fischer. 1990. Induced resistance
in soybean to the Mexican bean beetle (Coleoptera; Coc-
cinellidae): comparisons of inducing factors. Environmen-
tal Entomology 19:18521857.
Lindsey, K., S. Casson, and P. Chilley. 2002. Peptides: new
signalling molecules in plants. Trends in Plant Science 7:
7883.
Moreno, J. J. 2000. Resveratrol modulates arachidonic acid
release, prostaglandin synthesis, and 3T6 fibroblast growth.
Journal of Pharmacologic Experimental Therapy 294:333
338.
Mueller, M. J. 1997. Enzymes involved in jasmonic acid
biosynthesis. Physiologia Plantarum 100:653663.
Musser, R. O., S. M. Hum-Musser, H. Eichenseer, M. Peiffer,
G. Ervin, J. B. Murphy, and G. W. Felton. 2002. Herbivory:
caterpillar saliva beats plant defences. Nature 416:599
600.
Pierpoint, W. S. 2002. Can aspirin help identify leaf proteins
active in defence responses? Annals of Applied Biology
140:233239.
Pohnert, G., V. Jung, E. Haukioja, K. Lempa, and W. Boland.
1999. New fatty acid amides from regurgitant of lepidop-
teran (Noctuidiae, Geometridae) caterpillars. Tetrahedron
55:1127511280.
Rasmussen, L., S. T. Christensen, P. Schousboe, and D. N.
Wheatley. 1996. Cell survival and multiplication. The
overriding need for signals: from unicellular to multicel-
lular systems. FEMS Microbiology Letters 137:123128.
Reymond, P., H. Weber, M. Damond, and E. E. Farmer. 2000.
Differential gene expression in response to mechanical
wounding and insect feeding in Arabidopsis. Plant Cell 12:
707719.
Rhoades, D. F. 1985. Offensivedefense interactions between
herbivores and plants; their relevance in herbivore popu-
lation dynamics and ecological theory. American Naturalist
125:205238.
Rosenthal, G. A., and M. R. Berenbaum, editors. 1991. Her-
bivores, their interactions with secondary plant metabo-
lites. Second edition. Academic Press, San Diego, Cali-
fornia, USA.
Rossier, O. K., K. Wengelnik, K. Hahn, and U. Bonas. 1999.
The Xanthomonas Hrp type III system secretes proteins
from plant and mammalian bacterial pathogens. Proceed-
ings of the National Academy of Science (USA) 96:9368
9373.
Rossiter, M. C., J. C. Schultz, and I. T. Baldwin. 1988. Re-
lationships among defoliation, leaf phenolics, and gypsy
moth performance. Ecology 69:267277.
Ryu, S. B., B. H. Karlsson, M. Ozgen, and J. P. Palta. 1997.
Inhibition of phospholipase D by lysophosphatidylethanol-
amine, a lipid-derived senescence retardant. Proceedings of
the National Academy of Science (USA) 94:1271712721.
77
Schultz, J. C. 2002. Shared signals and the potential for phy-
logenetic espionage between plants and animals. Integra-
tive and Comparative Biology 46:454462.
Schultz, J. C., and I. T. Baldwin. 1982. Oak leaf quality
declines in response to defoliation by gypsy moth larvae.
Science 217:149151.
Shiu, S. H., and A. B. Bleecker. 2001. Receptor-like kinases
from Arabidopsis form a monophyletic gene family related
to animal receptor kinases. Proceedings of the National
Academy of Science (USA) 98:1076310768.
Spiteller, D., K. Dettner, and W. Boland. 2000. Gut bacteria
may be involved in interactions between plants, herbivores
and their predators: microbial biosynthesis of N acyl glu-
tamine surfactants as elicitors of plant volatiles. Biological
Chemistry 381:755762.
Stanley, D. W. 1999. Eicosanoids in invertebrate signal trans-
duction systems. Princeton University Press, Princeton,
New Jersey, USA.
Stoka, A. M. 1999. Phylogeny and evolution of chemical
communication: an endocrine approach. Journal of Molec-
ular Endocrinology 2:207225.
Stotz, H. U., T. Koch, A. Biedermann, K. Weniger, W. Boland,
and T. Mitchell-Olds. 2002. Evidence for regulation of re-
sistance in Arabidopsis to Egyptian cotton worm by sali-
cylic and jasmonic acid signaling pathways. Planta 214:
648652.
Takayama, S., and Y. Sakagami. 2002. Peptide signaling in
plants. Current Opinion in Plant Biology 5:382387.
Tan, M. W., and F. M. Ausubel. 2000. Caenorhabditis ele-
gans: a model genetic host to study Pseudomonas aeru-
Special Feature
ginosa pathogenesis. Current Opinion in Microbiology 3:
294.
Thaler, J. D., M. J. Stout, R. Karban, and S. S. Duffey. 2001.
Jasmonate-mediated induced plant resistance affects a com-
munity of herbivores. Ecological Entomology 26:312324.
Turlings, T. C. J., H. T. Alborn, J. H. Loughrin, and J. H.
Tumlinson. 2000. Volicitin, an elicitor of maize volatiles
in oral secretion of Spodoptera exigua: isolation and bio-
activity. Journal of Chemical Ecology 26:189202.
van Ryn, J., G. Trummlitz, and M. Pairet. 2000. COX-2 se-
lectivity and inflammatory processes. Current Medicinal
Chemistry 7:11451161.
vant Slot, K. A. E., and W. Knogge. 2002. A dual role for
microbial pathogen-derived effector proteins in plant dis-
ease and resistance. Critical Reviews in Plant Sciences 21:
229271.
Weber, H. 2002. Fatty acid-derived signals in plants. Trends
in Plant Science 7:217224.
Wharam, S. D., V. Mulholland, and G. P. C. Salmond. 1995.
Conserved virulence factor regulation and secretion sys-
tems in bacterial pathogens of plants and animals. European
Journal of Plant Pathology 101:113.
Yu, D., Y. Liu, B. Fan, D. F. Klessig, and Z. Chen. 1997. Is
the high basal level of salicylic acid important for disease
resistance in potato? Plant Physiology 115:343349.