Trends in Plant Science

Review

Ecology of Plastic Flowers

Quint Rusman ,1,* Dani Lucas-Barbosa ,1,2 Erik H. Poelman ,1 and Marcel Dicke ,1

Plant phenotypic plasticity in response to herbivore attack includes changes in Highlights

flower traits. Such herbivore-induced changes in flower traits have conse- Herbivore-induced changes in flower
quences for interactions with flower visitors. We synthesize here current knowl- traits are to a large extent herbivore
edge on the specificity of herbivore-induced changes in flower traits, the species-specific.

underlying molecular mechanisms, and the ecological consequences for Differences in the specificity of induced
flower-associated communities. Herbivore-induced changes in flower traits changes in plant traits between leaves
seem to be largely herbivore species-specific. The extensive plasticity observed and flowers can be explained by tissue-

in flowers influences a highly connected web of interactions within the flower- specific and plant ontogeny-specific
gene expression and regulation.
associated community. We argue that the adaptive value of herbivore-induced
plant responses and flower plasticity can be fully understood only from a com- Herbivore-induced plasticity in flowers in-
munity perspective rather than from pairwise interactions. fluences a highly connected web of inter-
actions within the flower-associated
community.

Flowers and Their Environment Community-wide effects of flower plas-

The angiosperms, which comprise the majority of plant species, are characterized by an enormous ticity likely vary considerably depending
on local or systemic induction.
diversity in flowers that range from the tiny millimeter-long flowers of duckweed (Lemna minor) to
the giant meter-wide corpse flower (Rafflesia arnoldi) [1,2]. The diversity of shapes, colors, and
scents of flowers is largely a result of evolution alongside an even more diverse flower-associated
community [3,4]. This community includes mutualists such as pollinators (see Glossary), preda-
tors, and parasitoids, as well as antagonists such as herbivores and pathogens [5–7]. Because
flowers are the reproductive organs of the plant, their displays are under selection to maximize re-
production under varying environmental conditions. Flowering plants are therefore expected to re-
spond to the environment, including visitation by antagonists and mutualists. Indeed, flower traits
readily change in response to herbivory or pathogen attack, even when these antagonists attack
plant organs other than flowers [8,9]. Flower traits also change in response to the activity of mutu-
alists such as pollinators [10–13] and beneficial microbes [14–16]. We focus here on herbivore-
induced changes in flower traits because, despite accumulating data on this topic, we still have im-
portant knowledge gaps. It is unclear if there are general patterns in flower phenotypic plasticity
in response to different types of herbivores, and whether such plasticity is adaptive for the plant.
Alternatively, herbivore-induced changes that are nonadaptive may merely change as pleiotropic
effects of herbivore-induced plant responses via resource trade-offs and genetic, biochemical, or
functional linkage (Box 1). Because changes in flower traits mediate interactions with multiple
flower-associated community members, we need to adopt a community approach to understand
flower trait plasticity [6,7,17,18]. In this review we discuss; (i) current knowledge on the specificity
of induction and to what extent herbivore-induced changes affect flower traits, (ii) the underlying
1
Laboratory of Entomology, Wageningen
University, PO Box 16, 6700 AA
molecular mechanisms of flower plasticity, (iii) the adaptive value of herbivore-induced flower plas-
Wageningen, The Netherlands
ticity in mediating interactions with flower visitors, and (iv) the ecological consequences of floral 2
Current address: Bio-Communication
plasticity for flower-associated communities. and Ecology, Eidgenössische
Technische Hochschule (ETH) Zürich,
Schmelzbergstrasse 9, 8092 Zürich,
Specificity of Herbivore-Induced Floral Plasticity
Switzerland
The influence of herbivore-induced changes on flower traits and the consequences for flower vis-
itors were first recognized more than two decades ago [19] (Figure 1, Key Figure). Since then it
has become apparent that floral plasticity in response to herbivory differs among plant species, *Correspondence:
and ranges from limited to extensive plasticity [17,20,21] (Table 1). For example, folivory by [email protected] (Q. Rusman).

Trends in Plant Science, August 2019, Vol. 24, No. 8 https://doi.org/10.1016/j.tplants.2019.04.007 725
© 2019 Elsevier Ltd. All rights reserved.
Trends in Plant Science

Box 1. Physiological Links between Defense and Reproduction

Glossary
Plant defense and reproduction appear to be linked because defensive and reproductive traits are correlated [90,108] and
Adaptive: enhancing fitness, in other
the expression of flower traits changes in response to herbivore attack. Various mechanisms have been proposed to ex-
words the contribution of an individual to
plain these links [17,18]. All traits share resources from the limited nutrient pool of the plant, and reallocation of resources to
the gene pool of the next generation.
defense can impair reproduction [109]. Defense and reproductive traits share phytohormonal signaling pathways including
Autogamous selfing: self-pollination
jasmonic acid (JA) and salicylic acid (SA) [26,29,110–112], and herbivore-induced increases or decreases in any of these
within a flower.
phytohormones potentially alter the expression of flower traits [111]. Downstream from phytohormonal signaling, the ex-
Florivores: consumers of developing
pression of both flower and defense traits is controlled by genetic and biochemical pathways. Genetic pleiotropy via gene
flower buds or mature flowers before the
regulatory networks and shared transcription factors, or individual genes involved in multiple regulatory pathways, can
development of the seed coat, including
connect defense and reproduction [113–118]. Biochemical pleiotropy can occur via changes in pools of shared precur-
consumers of bracts, sepals, petals,
sors or enzymes [77,78,119,120]. Other physiological constraints can lead to coexpression of defense and reproductive
stamens, pistils, pollen, and ovules [7].
traits, such as the passive diffusion of defensive metabolites from the phloem to flower organs [18,121], or herbivory-in-
The term also includes sap-feeding
duced changes in the chemical environment of the cell, which are important for phytohormone signaling [29], but can also
consumers that feed from floral stalks.
change the redox state of pigments, leading to a shift in light absorbance [122]. Finally, individual traits can have multiple
Folivores: consumers of aboveground
functions and be involved in both defense and reproduction [24,123]. For example, flowering plants use floral volatiles to
vegetative tissues.
attract not only pollinators but also natural enemies of herbivores [24]. Redirection of flower traits to a defensive function
Nectar robber: a flower visitor that
upon herbivore attack with associated changes can render these traits suboptimal for reproductive functions.
damages flowers while collecting nectar,
but does not contribute to pollination
[126].
Trichoplusia ni affects flower morphology of Campanula rotundifolia, and floral volatile blend com-
Nectar thief: a flower visitor that
position of Heterotheca villosa, but does not affect flower traits of Phacelia hastata or Potentilla collects nectar without damaging
recta [21]. In Brassica nigra, a range of herbivore species that commonly attack this species flowers, but does not contribute to
change multiple traits simultaneously, including floral morphology, volatiles, color, and nectar pollination [126].
Nonadditive selection: the selective
and pollen production [8]. Some flower traits are more plastic than others, either because their ex-
effect of a community member depends
pression is more plastic [22,23] or because they are more closely connected to physiological reg- on the presence of other community
ulation of plant defenses (Box 1). Flower phenology, morphology, and volatile emission in members [104].
particular seem to be plastic in response to herbivory, whereas nectar production changes in Phenotypic plasticity: the capacity of
a single genotype to display different
some cases but not in others (Table 1). Few studies have investigated a set of multiple flower
phenotypes in response to different
traits; for example, flower color, pollen production, and pollen chemistry are rarely investigated. environments [127].
When measured, color and pollen production show changes in response to herbivory, but not Plant ontogeny: the development of a
necessarily so [8,12,24,25]. For instance, the reflectance spectrum of flowers might not change plant from seed to mature seed-
producing plant [57].
in response to herbivory [20,24], or changes by reflecting higher or lower intensities of specific Pollinator: a flower visitor that
wavelengths such as yellow and UV [8]. When pollen production changes in response to herbiv- contributes to successful pollination, in
ory, this is mostly by a reduction in the amount of pollen produced [8,25]. other words to the transfer of pollen from
one flower to another conspecific flower.
In addition to variation in plant species and traits, herbivore-induced changes in flower traits vary Specificity of induction: differential
changes in phenotype in response to
with herbivore identity and feeding behavior (Table 1). Because herbivore-induced plant re-
different inducers.
sponses often contain a systemic component, and defense and reproduction are physiologically
linked (Box 1), we expect similar patterns of specificity in the induction of flower traits as for de-
fense traits. The specificity of induction of foliar defense traits often includes a general component
based on the feeding mode and feeding site of the herbivore, and a more specific component
based, among others, on herbivore identity [26–29]. The limited direct evidence available sug-
gests that changes in flower traits are to a large extent herbivore species-specific (Table 1). Indi-
rect evidence gained by comparing different studies supports this. For example, application of
jasmonic acid (JA) to Brassica nigra leaves, which induces the general component of
herbivore-induced plant responses against chewing herbivores, reduced nectar production in
the flowers, whereas folivory by Pieris rapae or Pieris brassicae caterpillars, two chewing herbi-
vores that induce JA, increased nectar production or had no effect [30,31]. To explain the differ-
ences in specificity of induction between foliar and floral tissues, we need to consider the
underlying physiological mechanisms.

Physiological Regulation of Flower Plasticity

Herbivore-induced changes in plant leaves are well known to be mediated by phytohormonal sig-
nal transduction pathways [32]. Emerging evidence indicates that both flowers and leaves

726 Trends in Plant Science, August 2019, Vol. 24, No. 8

Trends in Plant Science

Key Figure
Illustration of the Potential Effects of Herbivore-Induced Plant Responses
on Flower Traits and the Consequences for Flower-Associated Organisms

Microbes Predators
Florivores Pollinators and parasitoids

Flower traits

Defense Reproduction

SA JA Others
Flower sap-feeding
herbivores Flower-chewing
herbivores

Leaf-chewing
herbivores
Leaf sap-feeding
herbivores

Root-feeding
herbivores

Trends in Plant Science

Figure 1. Plant responses to herbivory involve systemic activation of phytohormonal signaling pathways. Overall patterns
illustrate that above- and belowground chewing herbivores induce the jasmonic acid (JA) pathway, whereas sap-feeding
herbivores induce the salicylic acid (SA) pathway. Both defense and reproduction are mediated by phytohormones such
as JA and SA, but also others. Moreover, defense and reproduction are physiologically linked via various mechanisms.
Both defense- and reproduction-related processes affect the expression of flower traits. Flower traits mediate interactions
with flower-associated community members such as floral microbes, florivores (including seed predators, and nectar or
pollen thieves and robbers), pollinators, predators, and parasitoids. Photograph credits: Dani Lucas-Barbosa, Jitte
Groothuis, Erik Poelman, and Quint Rusman.

Trends in Plant Science, August 2019, Vol. 24, No. 8 727

 Trends in Plant Science
728

Table 1. Phenotypic Plasticity of Flowers in Response to Herbivory with Its Effects on Subsequent Flower Visitation by Other Organismsa
Flower morphology
Trends in Plant Science, August 2019, Vol. 24, No. 8

Feeding guild or site Insect herbivore Plant Flower Flower Petal/corolla Petal/corolla Petal/corolla Petal Flower Anther Style Herkogamy
speciesb abundance phenology size length width roundness display length length
size
Chewing herbivores Athalia rosae Bn O . O − O − O . . .
Depressaria pastinacella Ps − . . . . . − . . .
Manduca sextra Sp O . . . . . O O − −
Phaedon cochleariae Sa O +/O . . +/O . O . . .
Pieris brassicae Br . + . . . . O . . .
Pieris brassicae Bn . . . . . . . . . .
Pieris brassicae Bn . . . . . . . . . .
Pieris rapae Rr O − . − − . . . . .
Plutella xylostella Bn O . O O + − + . . .
Spodoptera littoralis Br . − . . . . O . . .
Spodoptera littoralis Sa O − − . . . . − . .
Spodoptera littoralis Sl O . . O O . . . . .
Trichoplusia ni Cr O . . . . . − . . .
Trichoplusia ni Hv O . . . . . O . . .
Trichoplusia ni Ph O . . . . . O . . .
Trichoplusia ni Pr O . . . . . O . . .
Zeromastax selenesii Pa . . . O . . . O − .
Sap-feeding herbivores Brevicoryne brassicae Bn O . O − O − O . . .
Lipaphis erysimi Bn O . + O + − O . . .
Philaenus spumarius Mg O . − − O . O . . +
Root-feeding herbivores Agriotes sp. Sa O O O . . . . O . .
Delia radicum Bn O . + O + − O . . .

(continued on next page)

 Trends in Plant Science
Table 1. (continued)Phenotypic Plasticity of Flowers in Response to Herbivory with Its Effects on Subsequent Flower Visitation by Other Organismsa
Flower volatiles Flower color Petal quality
Feeding guild or site Total volatile emission Volatile blend Reflectance spectrum Yellow reflectance UV reflectance Defensive chemistry Nutritional chemistry

Chewing herbivores O C C − O . .
+ C . . . . .
. C . . . . .
O/? O/? . . . . .
− ? O . . O .
. C O O O C .
. C . . . . .
. . . . . . .
O C C O O . .
− ? O . . + .
. . . . . . .
?/+ ? . . . . .
O O . . . . .
O C . . . . .
O O . . . . .
O O . . . . .
O O . . . . .
Sap-feeding herbivores O C C O + . .
O C C O + . .
. . . . . . .
Root-feeding herbivores . . . . . . .
Trends in Plant Science, August 2019, Vol. 24, No. 8

O C C O O . .

(continued on next page)

729
 Trends in Plant Science
730

Table 1. (continued)Phenotypic Plasticity of Flowers in Response to Herbivory with Its Effects on Subsequent Flower Visitation by Other Organismsa
Flower rewards Behavioral responses Refs
Trends in Plant Science, August 2019, Vol. 24, No. 8

Feeding guild or site Nectar Nectar sugar Nectar Pollen Pollen Syrphid fly Butterfly Hummingbird Bee Predators and Florivores
volume concentration chemistry grains chemistry behavior behavior behavior behavior parasitoids

Chewing herbivores − . . . . O O . . . . [8]

. . . . . − . . . . . [106]
. . . . . . . . − . − [55]
. . . . . . . . . . . [20]
. . . . . . . . − + . [24]
. . . . . − O . . . . [12]
O C O . . − − . − . . [30]
− . . − . − . . − . . [19]
. . . . . O − . . . . [8]
. . . . . . . . − . . [24]
O . . . . . . . . . . [25]
O O . . . . . . . . . [67]
. . . . . . . . O . . [21]
. . . . . . . . − . . [21]
. . . . . . . . O . . [21]
. . . . . . . . O . . [21]
− + . . . . . − O . . [73]
Sap-feeding herbivores − . . . . − − . . . . [8]
O . . O . + + . . . . [8]
. . . . . O . . O . . [107]
Root-feeding herbivores O . . . . . . . . . . [25]
O . . − . + O . . . . [8]

a
Compared with uninfested plants, herbivory may increase (+), decrease (−), or have neutral (O) effects on plant traits and flower visitors, or herbivory may change traits in a specific direction (C). A dot (.) indicates
that the trait/insect response was not assessed for the respective herbivore. A question mark indicates that no exact data were presented or accessible, or that no proper statistical analysis was carried out. We
have only included studies that assessed three or more trait groups.
b
Plant species: Bn, Brassica nigra; Br, B. rapa; Cr, Campanula rotundifolia; Hv, Heterotheca villosa; Mg, Mimulus guttatus; Pa, Palicourea angustifolia; Ph, Phacelia hastata; Pr, Potentilla recta; Ps, Pastinaca
sativa; Rr, Raphanus raphanistrum; Sa, Sinapis arvensis; Sl, Silene latifolia; Sp, Solanum peruvianum.
Trends in Plant Science

respond to herbivory by activating phytohormonal signaling. Increased expression of the JA bio-

synthesis gene LIPOXYGENASE 2 (LOX2) can be induced in various floral tissues by exogenous
methyl jasmonate [33], and leaf herbivory can induce the expression of JA biosynthesis genes
such as ALLENE OXIDE SYNTHASE (AOS) in the flowers [34,35], with subsequent induction of
JA [34,36,37]. In contrast to leaves, concentrations of floral salicylic acid (SA) have so far not
been shown to change in response to herbivory, whereas for abscisic acid (ABA) the evidence
suggests that herbivory either has no effect or results in increased concentrations of ABA in
flowers [34,36]. Despite the apparent conservation of phytohormonal signaling in leaves and
flowers (Box 2), important tissue-specific patterns in the accumulation and regulation of signaling
pathway components have been identified [38]. This may explain the discrepancy in the specific-
ity of herbivore induction between foliar and floral tissues. So far, the accumulation and regulation
of foliar and floral tissue-specific components have mostly been documented for the JA pathway.
The constitutive accumulation of ALLENE OXIDE CYCLASE (AOC) mRNA and proteins was
found to be higher in tomato flowers compared with leaves, and also differed between flower
tissues [39]. The accumulation of AOC mRNA and proteins was accompanied by tissue-
specific increases in the concentrations of JA, 12-oxo-phytodienoic acid (OPDA), and
jasmonoyl-L-isoleucine (JA-Ile). When AOC was overexpressed in tomato, this increased JA
and OPDA concentrations in buds and flowers, but not in leaves [40]. The increase in JA and
OPDA concentrations differed among floral tissues, resulting in specific ratios of these com-
pounds in various flower organs. Biosynthesis of JA in the stamens of Arabidopsis thaliana starts
with the expression of DEFECTIVE IN ANTHER DEHISCENCE 1, which is not expressed in other
flower tissues [41]. In pea (Pisum sativum), a LOX gene has been identified which is predominantly
expressed in the flowers, and shows differential expression in flower tissues, with highest expres-
sion in petals and carpels [42]. Together, these results suggest differential regulation of JA biosyn-
thesis between leaves and flowers, and even different flower organs, via tissue-specific
accumulation of conserved pathway components and the presence of unique pathway compo-
nents. Such tissue specificity is also present in the JA-induced part of the signaling pathway. Spe-
cific types of JASMONATE ZIM-DOMAIN (JAZ) and MYB proteins are found in flowers, and these
types are not expressed in leaves [43–46]. These specific JAZ proteins are important for the ac-
cumulation of constitutive defenses [43]. Moreover, JAZ proteins that occur both in leaves and
flowers can serve different functions that have different regulatory targets [47]. In particular, JAZ
proteins and transcription factors (TFs) such as MYCs, MYBs, and APs may be important for
tissue-specific regulation of particular processes such as flower development and defense [46,
48–50]. For example, in Arabidopsis flowers, JAZ1, 3, 4, and 9 proteins interact with the TFs
TOE1 and TOE2 to regulate CORONATINE INSENSITIVE 1 (COI1)-dependent flowering, but
not COI1-dependent defense gene expression [48]. Moreover, MYB and WRKY TFs, as well as
multiple CYP94 genes – that are important in feedback mechanisms of JA-Ile – are differentially
expressed in leaves and flowers, and even in different floral tissues [45,51,52].

Regarding the SA pathway, leaves and flowers can contain different concentrations of free and total
SA, and this results in tissue-specific transcriptional responses of SA-regulated defense genes [36,

Box 2. Defense Regulation in Flowers: What Can We Learn from Leaves?

Research on flower development suggests that the gene regulatory networks for various phytohormones, and for jasmonic
acid (JA) in particular, are conserved in leaves and flowers [52,111]. Both flowers and leaves express multiple JA biosynthesis
genes and products such as LIPOXYGENASE (LOX), ALLENE OXIDE SYNTHASE (AOS), 12-oxo-phytodienoic acid (OPDA),
and OPDA REDUCTASE (OPR), and produce JA locally [43,44,52,124]. Later steps in JA signaling also seem to work sim-
ilarly in leaves and flowers, with essential roles for CORONATINE INSENSITIVE 1 (COI1) and JASMONATE ZIM-DOMAIN
(JAZ) proteins, and WKRY and MYB TFs [45,48]. In addition, leaves and flowers have similar expression patterns for three
NAC genes encoding JA regulatory proteins [52], and also for some JAZ genes: JAZ5 and JAZ7 [44,125]. Thus, the back-
bone of phytohormonal signaling, and for JA in particular, appears to be conserved in leaves and flowers.

Trends in Plant Science, August 2019, Vol. 24, No. 8 731

Trends in Plant Science

53]. Moreover, SA concentrations and gene transcription levels also differ between sepals and
petals, although the differences are smaller compared with leaves. The TF HAHB10 is induced in
sunflower leaves after SA treatment and pathogen attack, but is repressed after wounding [54].
This TF increases the expression of multiple genes involved in flowering, and represses the expres-
sion of multiple genes involved in JA- and SA-mediated defense. Interestingly, this TF is mostly
expressed in mature leaves, and is almost absent from floral tissues except the carpels [54].

This suggests that the regulation of development and defense is different in vegetative and floral tis-
sues of flowering plants. Tissue-specific gene expression patterns and regulatory components likely
lead to different plant responses in leaves versus flowers. Indeed, floral and foliar herbivory induce
different changes in the plant [35,55], and plant responses to foliar herbivory differ for foliar and floral
tissues [12,30,37]. Flowers even respond differently to attack on different systemic tissues. For ex-
ample, root and foliar herbivory induce different changes in flower traits [8,56] (Figure 1). Moreover,
the plant as an integrated phenotype can adopt different defense strategies depending on plant
ontogeny [57], and specifically between vegetative and flowering stages [58,59]. The timing of
herbivory over plant ontogeny can therefore result in different patterns of herbivore-induced plant
responses [60]. Taken together, tissue-specific and ontogeny-specific expression of genes and
regulatory components with resultant differential expression of plant responses likely explain the
differences in specificity of herbivore-induced changes in plant traits between leaves and flowers.
Such knowledge about the underlying mechanisms allow us to manipulate the plant phenotype,
while preserving the context of these complex interactions, to test the effects of floral plasticity on
other organisms present in the environment, as well as on plant fitness.

Adaptiveness of Floral Plasticity

Flower traits that commonly change in response to herbivory are hypothesized to be adaptive by
mediating interactions that maximize reproductive output, thus benefiting plant fitness (Figure 2). Flo-
ral volatiles commonly change in response to herbivory and these are exploited by natural enemies of
herbivores [24,61,62]. Although plant fitness benefits through the attraction of natural enemies of
herbivores are intuitive, and have been shown for plant interactions with herbivores and natural en-
emies in the vegetative stage [63], these remain largely elusive for flowering plants (but see [64]).
Herbivore-induced changes can increase the attraction of pollinators [8,65,66], potentially via
changes in floral volatile emission [67]. An increased attraction of pollinators can increase reproduc-
tive output [67,68], but not necessarily [65], and this likely depends on conditions such as pollen and
resource limitations. Herbivore-induced changes can lead to increased resistance to florivores,
seed predators, and nectar thieves [66,69] through changes in flower chemistry [7,37,70,71].
Leaf herbivore-induced resistance to seed predators benefits plants of Oenothera biennis by reduc-
ing seed predation to a large extent, whereas leaf herbivory itself had little impact on reproductive
output [37]. Hence, herbivore-induced changes in flower traits that mediate interactions with mutu-
alistic or antagonistic flower visitors may be adaptive for the plant.

In contrast to benefits, changes in flower traits may result in ecological costs [55,72]. The most
commonly reported ecological cost is a reduction in pollinator visitation [17,55] that is mediated
by herbivore-induced changes in floral volatiles, morphology, color, and/or rewards (Figure 2)
[8,24,55,73]. Such reductions in pollinator visitation can negatively affect plant reproductive out-
put [73], but not necessarily, again depending on conditions such as pollen and resource limita-
tions [74,75]. Herbivore-induced flower traits may also interfere with the optimization of
pollination. For example, plants use honest signaling to increase flower constancy and pollination
effectiveness [76]. With honest signaling, plants provide one or a few cues, such as volatile com-
pounds or flower size, that are reliable indicators of flower rewards. By altering flower volatile
emission, morphology, size, and nectar and pollen production, herbivores potentially interfere
with honest signaling [8,77]. Herbivory can also alter local changes in flower traits in response

732 Trends in Plant Science, August 2019, Vol. 24, No. 8

Trends in Plant Science

Trends in Plant Science

Figure 2. Schematic Representation of Herbivore-Induced Changes in Flower Traits That Mediate Interactions
with Flower-Associated Organisms. Unbroken lines indicate direct evidence, while broken lines reflect indirect evidence.
The sign in the circle represents the direction of effect, where + indicates positive effects, − indicates negative effects, and +/−
indicates both positive and negative effects, on the attraction, abundance, or performance of the insects and microbes. The
term florivore here includes seed predators as well as nectar and pollen thieves and robbers. Photograph credits: Erik
Poelman and Quint Rusman.

to visitation by flower visitors. Pollination often induces changes in flower volatile emission, color,
and morphology, resulting in reduced visitation by other pollinators to already pollinated flowers
[10–13]. Pollination-induced changes in, for example, floral volatiles can be dependent on
whether the plant is simultaneously attacked by herbivores [12], which might reduce pollination
effectiveness. Interestingly, the ecological costs of changes in flower phenotype in terms of re-
duced pollination can be compensated for by simultaneous changes in plant mating system
[78], for example, by increased autogamous selfing [79]. Herbivore-induced changes can in-
crease the attraction of florivores by changing apparency traits such as plant height, flower
color, and volatile emission [69], with potential negative effects on plant reproduction [80].
Thus, herbivore-induced changes in flower traits can be detrimental for plant reproduction by al-
tering interactions with flower visitors.

Herbivore-induced changes in interactions with flower visitors seem to be adaptive in some

cases, such as increased resistance to florivores, but maladaptive in others, such as reduced vis-
itation by pollinators (Figure 2). Despite these insights, we lack studies that specifically target fit-
ness effects of plant responses to herbivory [81] and link these to altered interactions with
flower visitors. To understand the adaptive value of flower plasticity with contrasting effects on
pairwise interactions, it is important to adopt a community perspective.

Consequences for Flower-Associated Communities

The flower-associated community is highly diverse and includes mutualists such as pollinators,
predators, parasitoids, and beneficial microorganisms, as well as antagonists such as herbivores

Trends in Plant Science, August 2019, Vol. 24, No. 8 733

Trends in Plant Science

and pathogenic microorganisms [5–7]. Flowers are visited by one or multiple pollinator species,
predominantly bees, flies, butterflies and moths, or beetles. The composition of the pollinator
community may be very specific for a given plant species, even when the plants are part of the
same plant community. Generalist predators such as crab spiders, true bugs, and social
wasps use flowers as hunting grounds [64,82], whereas specialist parasitoids may visit flowers
to find their specific hosts [6,24,62]. Herbivores visit flowers frequently, and as larvae or adults
can consume specific organs or the complete flower [7]. Florivores can feed exclusively on
flowers, but can also start feeding on leaves and then move to the flowers later in development,
or switch diet when flowers become available [83]. In addition to interactions with insects, flowers
contain a rich microbial community consisting of fungi, bacteria, archaea, and viruses [84]. These
microbes can be antagonists by destroying flower tissues, opportunists that exploit the transient
habitat without benefits or detriments to the flowers, or mutualists by competing for niche space
with antagonistic microbes. The floral microbial community shows considerable variation in
space, various flower organs, and time [84]. Some flower visitors use flower traits for host-plant
location, whereas the growth and survival of most of these organisms is affected by the chemical
composition of the flowers [7,84,85]. Therefore, we expect herbivore-induced changes in flower
traits to have a major impact on flower-associated community dynamics.

Herbivore-induced changes in flower traits have flower community-wide consequences. Pollina-

tor community composition can be different for herbivore-infested plants compared with
uninfested plants [20,68,86], but not necessarily [21,83], and this may depend on the identity
and feeding guild or feeding site of the herbivore [68]. Florivore community assembly can be af-
fected by early-season leaf herbivory [87]. Interestingly, the florivore community of Brassica
oleracea was affected differentially by the sequence of arrival of two leaf-feeding herbivores
early in the season [87]. Although there is no direct evidence that herbivore-induced changes in
flower traits affect floral microbial community composition, this is most likely [88]. Herbivore-
induced changes in flower traits include many characteristics that affect the abundance and di-
versity of floral microbes, such as nectar composition and volatile emission [84]. The chemical
properties of nectar are a prime determinant of microbial communities in nectaries. For example,
the composition of secondary metabolites such as pyridine-type alkaloids affects bacterial com-
munity richness, diversity, and composition in the nectar [89], and herbivory can increase nectar
alkaloid levels [90]. Therefore, we expect that changes in flower traits in response to herbivory af-
fect floral microbial communities.

Flower plasticity can link multiple interactions between flower visitors. Flowers are generally short-
lived [91]. Flower-associated organisms all interact with flowers during this short time-window.
Flower visitors that respond to herbivore-induced changes in flower traits, so-called receivers,
can become inducers themselves when their activities induce additional changes in flower traits
(Figure 3). Such flower visitor-induced changes in flower traits can subsequently affect other
flower visitors [12–16,66], and even feedback to leaf herbivores [69]. Changes in flower pheno-
type in response to each interacting flower visitor will result in multiple linked indirect interaction
units, where one interaction unit comprises an inducer, the mediator [the flower(s)], and a receiver
(Figure 3) [92]. For example, herbivory can induce changes in nectar composition and volatile
emission [30,90], which subsequently affect nectar microbial community composition and polli-
nator visitation [12,30,68,89]. Different nectar microbial communities may induce further changes
in nectar composition and volatiles, which affect parasitoid foraging and performance [14,16], as
well as pollinator visitation [15]. Indeed, artificial florivory, pollination, and nectar robbery have
been shown to affect multiple subsequent flower visitors at the same time, including pollinators,
nectar thieves and robbers, florivores, and predators [66,69]. Interestingly, most effects were
nonadditive, which suggests that the effect of one flower visitor can depend on the presence or

734 Trends in Plant Science, August 2019, Vol. 24, No. 8

Trends in Plant Science

(A)

(B)

(C)

(D)

Trends in Plant Science

Figure 3. Theoretical Framework of How Plant-Mediated Species Interactions Form a Network of Trait-
Mediated Species Interactions within Flowers (A), the Whole Plant (B), or Both (C,D). (A) A flower-feeding
herbivore that induces a local plant response (inducer) affects floral microbes (receiver) that in turn may affect the same or
different flower traits that are received by a pollinator. (B) A leaf-feeding herbivore that induces systemic plant responses
(inducer) affects a flower-feeding herbivore (receiver) that in turn also induces systemic plant responses and affects the
same or different flower traits that are perceived by a pollinator. (C) A leaf-feeding herbivore that induces systemic plant
responses (inducer) affects a pollinator (receiver) that in turn induces local plant responses and affects the same or
different flower traits that are received by a predator. (D) Local or systemic induced changes may affect multiple receivers
at the same time and have flower community-wide consequences. Photograph credits: Erik Poelman and Quint Rusman.

behavior of other flower visitors. Taken together, the extensive plasticity of flowers likely results in
a highly connected web of interactions within the complete flower-associated community.

Although flower plasticity most likely has community-wide effects, broader patterns are difficult to
predict. We still lack detailed knowledge on flower plasticity in response to the community, for ex-
ample, how flowers respond to sequential induction by different interactors or multiple interactors
at the same time. Different interactors may vary in the magnitude of induced changes in flowers,

Trends in Plant Science, August 2019, Vol. 24, No. 8 735

Trends in Plant Science

with supposedly varying community-wide effects. Leaf and root herbivory induce systemic Outstanding Questions
changes in the plant, likely affecting all flowers in the inflorescences, with high potential to affect What are the molecular mechanisms that
temporal or spatially displaced flower visitors [68,73]. For floral herbivory, herbivore-induced underlie defense regulation in flower
changes can be local, which would restrict the effects to visitors of that specific flower or a specific tissues?

inflorescence. For example, changes in flower phenotype as a result of feeding damage by How tight is the link between physiologi-
florivores can be restricted to individual flowers, and damaged flowers subsequently receive cal regulation of defense and reproduc-
fewer pollinator visits [93,94]. Moreover, the tissue/organ that is damaged can influence how tion, which mechanisms are especially
the flower appears to flower visitors [95,96]. For example, nectar-guide removal in Alstroemeria important, and do these vary between
plant species or under different environ-
ligtu reduced pollinator visitation, whereas lateral red tepal removal did not [96]. Interestingly, mental conditions?
damaged flowers can also reduce visitation to undamaged flowers in the inflorescence [97] either
because of systemic changes in the whole inflorescence or because pollinators judge potential Can we identify patterns of specificity of
induction by flower visitors other than in-
food plants at the plant level and therefore avoid damaged plants. Pollinator- and microbe-
sect herbivores, such as pollinators and
induced changes in flower traits are expected to be mostly local. Some of these local effects, microbes, and what are the differences
such as pollinator-induced changes in flower longevity, affect overall plant appearance, and and similarities compared with specificity
thereby affect pollinator and florivore visitation to noninduced flowers [13,98]. Thus, the of induction by herbivores?
community-wide effects of flower plasticity likely vary considerably depending on the type of in-
To what extent do herbivore-induced
ducer, and on whether induction is local or systemic. changes in flower traits affect floral mi-
crobial community composition, and
It is also worth considering how such effects will vary depending on plant phenological traits such can we identify differential responses of
as flower longevity and abundance. Plant species can differ greatly in flower longevity and abun- microbial species and different floral
organs?
dance, from having only a few flowers that can last multiple months to having hundreds of flowers
that last for a few days or less [91]. Community-wide effects of local induction are probably much Can we identify broader patterns of
more extensive for plant species with a few long-lived flowers compared with plant species with herbivore-induced changes in flower
high floral turnover. The extent of floral plasticity can also vary depending on flower longevity and traits on flower-community assembly,
are these predictable, and are particular
abundance. Long-lived flowers are expected to be more plastic to allow continuous adaptation to
groups of flower-associated organisms
environmental variation [13,91]. Orchids with their relatively long-lived flowers, for example, alter more affected than others?
flower longevity depending on variation in the biotic environment, namely the presence of pollina-
tors which pollinate the flower [91]. Although potentially less affected by local pollinator-induced Do herbivore-induced changes in flower
traits drive eco-evolutionary dynamics?
changes, short-lived flowers still readily change in response to systemic induction by herbivores
[56]. Local microbe-induced changes might affect short- and long-lived flowers to the same ex-
tent because microbes modulate flower traits directly rather than indirectly via changes in plant
physiology [84,99]. However, the consequences of such microbe-induced changes will be
more apparent for plants with long-lived flowers because of the lower rate of flower turnover
and the smaller numbers of flowers [91]. Hence, flower longevity and abundance can be impor-
tant for community-wide effects of local and systemic induction.

Lastly, the responses of some groups of flower visitors to flower plasticity are better understood
than others, and we specifically lack fundamental knowledge about whether and how herbivore-
induced changes affect floral microbial community composition, the differential responses of mi-
crobe species, and differential effects on microbes inhabiting different floral organs. Although it
seems likely that flower plasticity has community-wide effects, there remain important gaps in
our knowledge with respect to flower plasticity, the extent to which such plasticity may be ex-
plained by plant life-history traits, and the consequences for flower visitors.

Concluding Remarks and Future Perspectives

To understand the ecology of flower plasticity we need to adopt a community context approach.
The evolution of flowers cannot be fully explained by focusing solely on pollinators, and is driven
by the combined selection imposed by the flower-associated community, including pollinators,
florivores, and microbes [3,66]. Likewise, the evolution of plant defenses cannot be fully explained
by focusing solely on herbivores, and theories on plant defense evolution will benefit from including
plant reproduction. For example, outcrossing Nicotiana species have lower nicotine concentrations

736 Trends in Plant Science, August 2019, Vol. 24, No. 8

Trends in Plant Science

in leaves, flowers, and nectar compared with selfing species [100]. Defensive traits may evolve from
pre-existing traits with reproductive functions, and hence reproductive evolutionary history may
help to explain the current defense phenotype [101]. Moreover, communities are characterized
by ample indirect interactions among their members. These indirect interactions are as important
as direct interactions in shaping evolution [72,102–104]. Hence, organisms such as herbivores
that are not directly associated with the flowers can contribute to flower evolution by indirectly af-
fecting interactions with flower visitors (nonadditive selection). The herbivore species-specific
nature of flower plasticity in response to herbivory suggests a large number of potential selective
agents, but also large temporal and spatial fluctuations in selection pressure exhibited by such in-
direct interactions. Particular plant traits such as flower longevity and abundance, or the ability to
separate foliar from floral plant responses, or defense and reproduction for the plant as a whole,
may determine the extent of nonadditive selection by indirect interactions. Separation of foliar
from floral plant responses, and defense and reproduction, can partly be achieved by tissue-
specific gene expression and regulatory components, as well as plant ontogenetic trajectories
in growth–defense–reproduction strategies. Nonetheless, it may be difficult for plants to
completely separate processes in leaves and flowers, and in defense and reproduction, owing
to diverse physiological links that allow plants to function as one integrated entity. Exploring the
ecology of plastic flowers will extend our understanding of the evolution of plant defense and
reproduction.

To better understand flower plasticity in response to herbivory, we need to deepen our knowl-
edge on the specificity of induction, for example, by using combinations of herbivores that
show patterns of induction of changes in foliar and/or root traits related to feeding guild, feeding
site, or host-plant specialization. Such studies should measure a multitude of flower traits be-
cause the specificity of induction can only be judged when considering the complete flower phe-
notype. We have only started to unravel the molecular mechanisms underlying specificity in flower
plasticity, especially how phytohormones, secondary metabolites, and defense genes are
expressed in the flowers [43,47,48,51,53]. More insight into such tissue-specific plant responses
will shed light on which of the potential links between defense and reproduction (Box 1) are com-
mon and important in flower plasticity in response to herbivory. Moreover, research on the spec-
ificity of induction will provide broader insights into how flowering plants deal with ecological
variation and optimize the attraction of mutualists while dealing with antagonists. The conse-
quences of floral plasticity should be investigated in a community context, and should consider
adaptive plasticity, with consequences for plant reproduction. An interesting approach would
be to focus on keystone herbivores: herbivores that have a large effect on the interaction network,
leading to associated fitness consequences for the plant [72]. Keystone herbivores that principally
affect the flower-visitor community can be compared with keystone herbivores that have large ef-
fects on foliar and/or root communities, as well as to non-keystone herbivores, to identify how
such herbivores drive selection in complex communities, and to determine the contributions of
flower plasticity and the flower-associated community. Another interesting approach will be to
compare the responses of flower-associated communities that vary in their overall degree of
host plant specialization [17] because these might differentially impact plant fitness, and
thereby open up different evolutionary trajectories. In addition to systemic induction, the role of
local flower plasticity in response not only to herbivory but also to pollination and microbial induc-
tion needs further attention (see Outstanding Questions). The community-wide effects of local
compared with systemic flower plasticity have so far not been explored, and it will be interesting
to compare flower plasticity of plants with particular traits, such as flower longevity and abun-
dance, in this perspective. Comparing plants with particular defensive flower traits, and applying
explicit hypotheses from theories on the evolution of plant defense, will reveal whether we can
generalize patterns of defenses in flowering plants as has been done for plants in the vegetative

Trends in Plant Science, August 2019, Vol. 24, No. 8 737

Trends in Plant Science

stage [105]. For example, do plants with apparent flowers invest more in constitutive defenses in
the flowers, whereas plants with unapparent flowers invest more in inducible defenses in the
flowers? Answers to such questions will facilitate the integration of evolutionary theories on
plant defense and reproduction and help to explain flower plasticity in response to herbivory
(see Outstanding Questions).

Acknowledgments
This work was supported by the Earth and Life Science Council of the Netherlands Organisation for Scientific Research
(NWO-ALW; grant 831.14.004).

References
1. Soltis, P.S. and Soltis, D.E. (2014) Flower diversity and angio- visitor networks but not plant reproduction. Oikos 125,
sperm diversification. In Flower Development: Methods and 1241–1249
Protocols (Riechmann, J.L. and Wellmer, F., eds), pp. 21. Burkle, L.A. and Runyon, J.B. (2016) Drought and leaf herbiv-
85–102, Springer ory influence floral volatiles and pollinator attraction. Glob.
2. Moyroud, E. and Glover, B.J. (2017) The evolution of diverse Chang. Biol. 22, 1644–1654
floral morphologies. Curr. Biol. 27, R941–R951 22. Siebenkäs, A. et al. (2015) Phenotypic plasticity to light and nu-
3. Strauss, S.Y. and Whittall, J.B. (2006) Non-pollinator agents of trient availability alters functional trait ranking across eight pe-
selection on floral traits. In Ecology and Evolution of Flowers rennial grassland species. AoB PLANTS 7, plv029
(Harder, L.D. and Barrett, S.C.H., eds), pp. 120–138, Oxford 23. Rozendaal, D.M.A. et al. (2006) Plasticity in leaf traits of 38
University Press tropical tree species in response to light; relationships with
4. Schiestl, F.P. and Johnson, S.D. (2013) Pollinator-mediated light demand and adult stature. Funct. Ecol. 20, 207–216
evolution of floral signals. Trends Ecol. Evol. 28, 307–315 24. Schiestl, F.P. et al. (2014) Herbivory and floral signaling: pheno-
5. Junker, R.R. and Keller, A. (2015) Microhabitat heterogeneity typic plasticity and tradeoffs between reproduction and indirect
across leaves and flower organs promotes bacterial diversity. defense. New Phytol. 203, 257–266
FEMS Microbiol. Ecol. 91, fiv097 25. Poveda, K. et al. (2005) Floral trait expression and plant fitness
6. Lucas-Barbosa, D. et al. (2011) The effects of herbivore- in response to below- and aboveground plant–animal interac-
induced plant volatiles on interactions between plants and tions. Perspect. Plant Ecol. Evol. Syst. 7, 77–83
flower-visiting insects. Phytochemistry 72, 1647–1654 26. Thaler, J.S. et al. (2012) Evolution of jasmonate and salicylate
7. McCall, A.C. and Irwin, R.E. (2006) Florivory: the intersection of signal crosstalk. Trends Plant Sci. 17, 260–270
pollination and herbivory. Ecol. Lett. 9, 1351–1365 27. Ali, J.G. and Agrawal, A.A. (2012) Specialist versus generalist
8. Rusman, Q. et al. (2019) Floral plasticity: herbivore-species- insect herbivores and plant defense. Trends Plant Sci. 17,
specific-induced changes in flower traits with contrasting ef- 293–302
fects on pollinator visitation. Plant Cell Environ. Published on- 28. Johnson, S.N. et al. (2016) Roots under attack: contrasting
line January 18, 2019. https://doi.org/10.1111/pce.13520 plant responses to below-and aboveground insect herbivory.
9. McArt, S.H. et al. (2014) Arranging the bouquet of disease: flo- New Phytol. 210, 413–418
ral traits and the transmission of plant and animal pathogens. 29. Erb, M. et al. (2012) Role of phytohormones in insect-specific
Ecol. Lett. 17, 624–636 plant reactions. Trends Plant Sci. 17, 250–259
10. Rodriguez-Saona, C. et al. (2011) Variation in highbush blue- 30. Bruinsma, M. et al. (2014) Folivory affects composition of nec-
berry floral volatile profiles as a function of pollination status, tar, floral odor and modifies pollinator behavior. J. Chem. Ecol.
cultivar, time of day and flower part: implications for flower vis- 40, 39–49
itation by bees. Ann. Bot. 107, 1377–1390 31. Bruinsma, M. et al. (2008) Differential effects of jasmonic acid
11. Willmer, P. et al. (2009) Bidirectional flower color and shape treatment of Brassica nigra on the attraction of pollinators, par-
changes allow a second opportunity for pollination. Curr. Biol. asitoids, and butterflies. Entomol. Exp. Appl. 128, 109–116
19, 919–923 32. Pieterse, C.M.J. et al. (2012) Hormonal modulation of plant im-
12. Lucas-Barbosa, D. et al. (2016) Visual and odour cues: plant munity. Annu. Rev. Cell Dev. Biol. 28, 489–521
responses to pollination and herbivory affect the behaviour of 33. Jensen, A.B. et al. (2002) Fusion genetic analysis of jasmonate-
flower visitors. Funct. Ecol. 30, 431–441 signalling mutants in Arabidopsis. Plant J. 29, 595–606
13. van Doorn, W.G. (1997) Effects of pollination on floral attraction 34. Zhou, Y. et al. (2017) Insects (Thrips hawaiiensis (Morgan))
and longevity. J. Exp. Bot. 48, 1615–1622 change the stereochemical configuration of 1-phenylethanol
14. Lenaerts, M. et al. (2017) Nectar bacteria affect life history of a emitted from tea (Camellia sinensis) flowers. RSC Adv. 7,
generalist aphid parasitoid by altering nectar chemistry. Funct. 32336–32343
Ecol. 31, 2061–2069 35. Peng, Q. et al. (2018) Functional characterization of an allene
15. Rering, C.C. et al. (2018) Nectar-inhabiting microorganisms in- oxide synthase involved in biosynthesis of jasmonic acid and
fluence nectar volatile composition and attractiveness to a gen- its influence on metabolite profiles and ethylene formation in
eralist pollinator. New Phytol. 220, 750–759 tea (Camellia sinensis) flowers. Int. J. Mol. Sci. 19, 2440
16. Sobhy, I.S. et al. (2018) Sweet scents: Nectar specialist yeasts 36. Chrétien, L.T.S. et al. (2018) Caterpillars induce jasmonates in
enhance nectar attraction of a generalist aphid parasitoid with- flowers and alter plant responses to a second attacker. New
out affecting survival. Front. Plant Sci. 9, 1009 Phytol. 217, 1279–1291
17. Lucas-Barbosa, D. (2016) Integrating studies on plant– 37. McArt, S.H. et al. (2013) Leaf herbivory increases plant fitness
pollinator and plant–herbivore interactions. Trends Plant Sci. via induced resistance to seed predators. Ecology 94,
21, 125–133 966–975
18. Jacobsen, D.J. and Raguso, R.A. (2018) Lingering effects of 38. Schmid, M. et al. (2005) A gene expression map of Arabidopsis
herbivory and plant defenses on pollinators. Curr. Biol. 28, thaliana development. Nat. Genet. 37, 501–506
R1164–R1169 39. Hause, B. et al. (2000) Tissue-specific oxylipin signature of to-
19. Strauss, S.Y. et al. (1996) Foliar herbivory affects floral charac- mato flowers: allene oxide cyclase is highly expressed in dis-
ters and plant attractiveness to pollinators: implications for tinct flower organs and vascular bundles. Plant J. 24, 113–126
male and female plant fitness. Am. Nat. 147, 1098–1107 40. Miersch, O. et al. (2004) Constitutive overexpression of allene
20. Hoffmeister, M. et al. (2016) Herbivore-induced changes in oxide cyclase in tomato (Lycopersicon esculentum cv.
flower scent and morphology affect the structure of flower- Lukullus) elevates levels of some jasmonates and

738 Trends in Plant Science, August 2019, Vol. 24, No. 8

Trends in Plant Science

octadecanoids in flower organs but not in leaves. Phytochem- shifting floral to defensive volatiles. Arthropod Plant Interact.
istry 65, 847–856 12, 751–760
41. Ishiguro, S. et al. (2001) The DEFECTIVE IN ANTHER DEHIS- 63. Schuman, M.C. et al. (2012) Herbivory-induced volatiles func-
CENCE1 gene encodes a novel phospholipase A1 catalyzing tion as defenses increasing fitness of the native plant Nicotiana
the initial step of jasmonic acid biosynthesis, which synchro- attenuata in nature. Elife 1, e00007
nizes pollen maturation, anther dehiscence, and flower open- 64. Knauer, A.C. et al. (2018) Crab spiders impact floral-signal evo-
ing in Arabidopsis. Plant Cell 13, 2191–2209 lution indirectly through removal of florivores. Nat. Commun. 9,
42. Rodríguez-Concepción, M. and Beltrán, J.P. (1995) Repres- 1367
sion of the pea lipoxygenase gene loxg is associated with car- 65. Poveda, K. et al. (2003) Effects of below-and above-ground
pel development. Plant Mol. Biol. 27, 887–899 herbivores on plant growth, flower visitation and seed set.
43. Li, R. et al. (2017) Flower-specific jasmonate signaling regu- Oecologia 135, 601–605
lates constitutive floral defenses in wild tobacco. Proc. Natl. 66. Soper Gorden, N.L. and Adler, L.S. (2018) Consequences of
Acad. Sci. U. S. A. 114, E7205–E7214 multiple flower–insect interactions for subsequent plant–insect
44. Reeves, P.H. et al. (2012) A regulatory network for coordinated interactions and plant reproduction. Am. J. Bot. 105, 1–12
flower maturation. PLoS Genet. 8, e1002506 67. Cozzolino, S. et al. (2015) Herbivory increases fruit set in Silene
45. Stitz, M. et al. (2014) Jasmonoyl-L-Isoleucine coordinates met- latifolia: a consequence of induced pollinator-attracting floral
abolic networks required for anthesis and floral attractant emis- volatiles? J. Chem. Ecol. 41, 622–630
sion in wild tobacco (Nicotiana attenuata). Plant Cell 26, 68. Rusman, Q. et al. (2018) Dealing with mutualists and antago-
3964–3983 nists: specificity of plant-mediated interactions between herbi-
46. Sherif, S. et al. (2015) A stable JAZ protein from peach medi- vores and flower visitors, and consequences for plant fitness.
ates the transition from outcrossing to self-pollination. BMC Funct. Ecol. 32, 1022–1035
Biol. 13, 11 69. Soper Gorden, N.L. and Adler, L.S. (2016) Florivory shapes
47. Oh, Y. et al. (2013) A jasmonate ZIM-domain protein NaJAZd both leaf and floral interactions. Ecosphere 7, e01326
regulates floral jasmonic acid levels and counteracts flower ab- 70. McCall, A.C. et al. (2018) Leaf herbivory induces resistance
scission in Nicotiana attenuata plants. PLoS ONE 8, e57868 against florivores in Raphanus sativus. Botany 96, 337–343
48. Zhai, Q. et al. (2015) Transcriptional mechanism of jasmonate 71. Li, R. et al. (2018) Jasmonate signaling makes flowers attrac-
receptor COI1-mediated delay of flowering time in Arabidopsis. tive to pollinators and repellant to florivores in nature.
Plant Cell 27, 2814–2828 J. Integr. Plant Biol. 60, 190–194
49. Yu, X. et al. (2018) The jasmonate ZIM-domain protein gene 72. Poelman, E.H. and Kessler, A. (2016) Keystone herbivores and
SlJAZ2 regulates plant morphology and accelerates flower the evolution of plant defenses. Trends Plant Sci. 21, 477–485
initiation in Solanum lycopersicum plants. Plant Sci. 267, 73. Chautá, A. et al. (2017) Leaf herbivory imposes fitness costs
65–73 mediated by hummingbird and insect pollinators. PLoS ONE
50. Wang, H. et al. (2017) The bHLH transcription factors MYC2, 12, e0188408
MYC3, and MYC4 are required for jasmonate-mediated inhibi- 74. Ghyselen, C. et al. (2016) Is there a missing link? Effects of root
tion of flowering in Arabidopsis. Mol. Plant 10, 1461–1464 herbivory on plant-pollinator interactions and reproductive out-
51. Bruckhoff, V. et al. (2016) Functional characterization of put in a monocarpic species. Plant Biol. 18, 156–163
CYP94-genes and identification of a novel jasmonate catabo- 75. Krupnick, G.A. and Weis, A.E. (1999) The effect of floral herbiv-
lite in flowers. PLoS ONE 11, e0159875 ory on male and female reproductive success in Isomeris
52. Wagstaff, C. et al. (2009) A molecular and structural character- arborea. Ecology 80, 135–149
ization of senescing Arabidopsis siliques and comparison of 76. Wright, G.A. and Schiestl, F.P. (2009) The evolution of floral
transcriptional profiles with senescing petals and leaves. Plant scent: the influence of olfactory learning by insect pollinators
J. 57, 690–705 on the honest signalling of floral rewards. Funct. Ecol. 23,
53. Ederli, L. et al. (2015) Arabidopsis flower specific defense gene 841–851
expression patterns affect resistance to pathogens. Front. 77. Junker, R.R. et al. (2018) Covariation and phenotypic integra-
Plant Sci. 6, 79 tion in chemical communication displays: biosynthetic con-
54. Dezar, C.A. et al. (2011) HAHB10, a sunflower HD-Zip II tran- straints and eco-evolutionary implications. New Phytol. 220,
scription factor, participates in the induction of flowering and 739–749
in the control of phytohormone-mediated responses to biotic 78. Johnson, M.T.J. et al. (2015) Evolutionary interactions between
stress. J. Exp. Bot. 62, 1061–1076 plant reproduction and defense against herbivores. Annu. Rev.
55. Kessler, A. and Halitschke, R. (2009) Testing the potential for Ecol. Evol. Syst. 46, 191–213
conflicting selection on floral chemical traits by pollinators and 79. Penet, L. et al. (2009) Florivory increases selfing: an experimen-
herbivores: predictions and case study. Funct. Ecol. 23, tal study in the wild strawberry, Fragaria virginiana. Plant Biol.
901–912 11, 38–45
56. Barber, N.A. et al. (2011) Effects of above-and belowground 80. Theis, N. and Adler, L.S. (2012) Advertising to the enemy: en-
herbivory on growth, pollination, and reproduction in cucum- hanced floral fragrance increases beetle attraction and reduces
ber. Oecologia 165, 377–386 plant reproduction. Ecology 93, 430–435
57. Barton, K.E. and Boege, K. (2017) Future directions in the on- 81. Erb, M. (2018) Plant defenses against herbivory: closing the fit-
togeny of plant defence: understanding the evolutionary ness gap. Trends Plant Sci. 23, 187–194
causes and consequences. Ecol. Lett. 20, 403–411 82. Lucas-Barbosa, D. et al. (2014) Caught between parasitoids
58. Diezel, C. et al. (2011) Mechanisms of optimal defense patterns and predators – survival of a specialist herbivore on leaves
in Nicotiana attenuata: Flowering attenuates herbivory-elicited and flowers of mustard plants. J. Chem. Ecol. 40, 621–631
ethylene and jasmonate signaling. J. Integr. Plant Biol. 53, 83. Lucas-Barbosa, D. et al. (2013) Reproductive escape: annual
971–983 plant responds to butterfly eggs by accelerating seed produc-
59. Desurmont, G.A. et al. (2015) Floral volatiles interfere with plant tion. Funct. Ecol. 27, 245–254
attraction of parasitoids: ontogeny-dependent infochemical dy- 84. Aleklett, K. et al. (2014) The microbial ecology of flowers: an
namics in Brassica rapa. BMC Ecol. 15, 17 emerging frontier in phyllosphere research. Botany 92, 253–266
60. Barton, K.E. and Koricheva, J. (2010) The ontogeny of plant 85. Dötterl, S. and Vereecken, N.J. (2010) The chemical ecology
defense and herbivory: Characterizing general patterns using and evolution of bee–flower interactions: a review and per-
meta-analysis. Am. Nat. 175, 481–493 spectives. Can. J. Zool. 88, 668–697
61. Dannon, E.A. et al. (2010) Effects of volatiles from Maruca 86. Steets, J.A. et al. (2006) Consequences of vegetative herbivory
vitrata larvae and caterpillar-infested flowers of their host for maintenance of intermediate outcrossing in an annual plant.
plant Vigna unguiculata on the foraging behavior of the parasit- Ecology 87, 2717–2727
oid Apanteles taragamae. J. Chem. Ecol. 36, 1083–1091 87. Stam, J.M. et al. (2018) Order of herbivore arrival on wild cab-
62. Silveira, T.A. et al. (2018) Phloem-feeding herbivory on bage populations influences subsequent arthropod community
flowering melon plants enhances attraction of parasitoids by development. Oikos 127, 1482–1493

Trends in Plant Science, August 2019, Vol. 24, No. 8 739

Trends in Plant Science

88. Allard, S.M. et al. (2018) Insect exclusion limits variation in bac- for pollination and nectar robbing in a distylous plant. Am.
terial microbiomes of tomato flowers and fruit. J. Appl. J. Bot. 93, 64–72
Microbiol. 125, 1749–1760 109. Herms, D.A. and Mattson, W.J. (1992) The dilemma of plants:
89. Aizenberg-Gershtein, Y. et al. (2015) Pyridine-type alkaloid to grow or defend. Q. Rev. Biol. 67, 283–335
composition affects bacterial community composition of floral 110. Rivas-San Vicente, M. and Plasencia, J. (2011) Salicylic acid
nectar. Sci. Rep. 5, 11536 beyond defence: its role in plant growth and development.
90. Adler, L.S. et al. (2006) Leaf herbivory and nutrients increase J. Exp. Bot. 62, 3321–3338
nectar alkaloids. Ecol. Lett. 9, 960–967 111. Yuan, Z. and Zhang, D. (2015) Roles of jasmonate signalling in
91. Ashman, T.-L. (2004) Flower longevity. In Plant Cell Death Pro- plant inflorescence and flower development. Curr. Opin. Plant
cesses (Noodén, L.D., ed.), pp. 349–362, Academic Press Biol. 27, 44–51
92. Utsumi, S. et al. (2010) Linkages among trait-mediated indirect 112. Avanci, N. et al. (2010) Jasmonates are phytohormones with
effects: a new framework for the indirect interaction web. multiple functions, including plant defense and reproduction.
Popul. Ecol. 52, 485–497 Genet. Mol. Res. 9, 484–505
93. Cardel, Y.J. and Koptur, S. (2010) Effects of florivory on the 113. Rasmann, S. et al. (2018) Pleiotropic effect of the Flowering
pollination of flowers: an experimental field study with a peren- Locus C on plant resistance and defence against insect herbi-
nial plant. Int. J. Plant Sci. 171, 283–292 vores. J. Ecol. 106, 1244–1255
94. Muola, A. et al. (2017) Direct and pollinator-mediated effects of 114. van Es, S.W. et al. (2018) Novel functions of the Arabidopsis
herbivory on strawberry and the potential for improved resis- transcription factor TCP5 in petal development and ethylene
tance. Front. Plant Sci. 8, 823 biosynthesis. Plant J. 94, 867–879
95. Oguro, M. and Sakai, S. (2009) Floral herbivory at different 115. Hemm, M.R. et al. (2003) The Arabidopsis ref2 mutant is defec-
stages of flower development changes reproduction in Iris tive in the gene encoding CYP83A1 and shows both phenyl-
gracilipes (Iridaceae). Plant Ecol. 202, 221–234 propanoid and glucosinolate phenotypes. Plant Cell 15, 179–194
96. Botto-Mahan, C. et al. (2011) Floral herbivory affects female re- 116. Kaufmann, K. et al. (2009) Target genes of the MADS transcrip-
productive success and pollinator visitation in the perennial tion factor SEPALLATA3: integration of developmental and hor-
herb Alstroemeria ligtu (Alstroemeriaceae). Int. J. Plant Sci. monal pathways in the Arabidopsis flower. PLoS Biol. 7,
172, 1130–1136 e1000090
97. Sõber, V. et al. (2010) Florivores decrease pollinator visitation in 117. Nagpal, P. et al. (2005) Auxin response factors ARF6 and ARF8
a self-incompatible plant. Basic Appl. Ecol. 11, 669–675 promote jasmonic acid production and flower maturation.
98. Harder, L.D. and Johnson, S.D. (2005) Adaptive plasticity of Development 132, 4107–4118
floral display size in animal-pollinated plants. Proc. R. Soc. B 118. Pajoro, A. et al. (2014) The (r)evolution of gene regulatory net-
272, 2651–2657 works controlling Arabidopsis plant reproduction: a two-
99. Junker, R.R. et al. (2011) Composition of epiphytic bacterial com- decade history. J. Exp. Bot. 65, 4731–4745
munities differs on petals and leaves. Plant Biol. 13, 918–924 119. Fineblum, W.L. and Rausher, M.D. (1997) Do floral pigmenta-
100. Adler, L.S. et al. (2012) Reliance on pollinators predicts defensive tion genes also influence resistance to enemies? The W locus
chemistry across tobacco species. Ecol. Lett. 15, 1140–1148 in Ipomoea purpurea. Ecology 78, 1646–1654
101. Armbruster, W.S. et al. (2009) Macroevolutionary patterns of 120. Muhlemann, J.K. et al. (2014) The monolignol pathway contrib-
defense and pollination in Dalechampia vines: adaptation, utes to the biosynthesis of volatile phenylpropenes in flowers.
exaptation, and evolutionary novelty. Proc. Natl. Acad. Sci. New Phytol. 204, 661–670
U. S. A. 106, 18085–18090 121. Adler, L.S. (2000) The ecological significance of toxic nectar.
102. Guimarães Jr., P.R. et al. (2017) Indirect effects drive coevolu- Oikos 91, 409–420
tion in mutualistic networks. Nature 550, 511–514 122. Borghi, M. et al. (2017) The sexual advantage of looking, smell-
103. Siepielski, A.M. and Benkman, C.W. (2004) Interactions among ing, and tasting good: The metabolic network that produces
moths, crossbills, squirrels, and lodgepole pine in a geographic signals for pollinators. Trends Plant Sci. 22, 338–350
selection mosaic. Evolution 58, 95–101 123. Gronquist, M. et al. (2001) Attractive and defensive functions of
104. Haloin, J.R. and Strauss, S.Y. (2008) Interplay between ecolog- the ultraviolet pigments of a flower (Hypericum calycinum).
ical communities and evolution. Ann. N. Y. Acad. Sci. 1133, Proc. Natl. Acad. Sci. U. S. A. 98, 13745–13750
87–125 124. Sanders, P.M. et al. (2000) The Arabidopsis DELAYED DEHIS-
105. Agrawal, A.A. (2007) Macroevolution of plant defense strate- CENCE1 gene encodes an enzyme in the jasmonic acid syn-
gies. Trends Ecol. Evol. 22, 103–109 thesis pathway. Plant Cell 12, 1041–1061
106. Zangerl, A. and Berenbaum, M. (2009) Effects of florivory on 125. Chini, A. et al. (2016) Redundancy and specificity in jasmonate
floral volatile emissions and pollination success in the wild pars- signalling. Curr. Opin. Plant Biol. 33, 147–156
nip. Arthropod Plant Interact. 3, 181–191 126. Inouye, D.W. (1980) The terminology of floral larceny. Ecology
107. Ivey, C.T. and Carr, D.E. (2005) Effects of herbivory and in- 61, 1251–1253
breeding on the pollinators and mating system of Mimulus 127. Whitman, D.W. and Agrawal, A.A. (2009) What is phenotypic
guttatus (Phrymaceae). Am. J. Bot. 92, 1641–1649 plasticity and why is it important? In Phenotypic Plasticity of In-
108. Irwin, R.E. and Adler, L.S. (2006) Correlations among traits as- sects: Mechanisms and Consequences (Whitman, D.W. and
sociated with herbivore resistance and pollination: implications Ananthakrishnan, T.N., eds), pp. 1–63, Science Publishers

740 Trends in Plant Science, August 2019, Vol. 24, No. 8