Illuminated behaviour: phytochrome as a key regulator of light foraging and plant anti-herbivore defence



    1. IFEVA, Consejo Nacional de Investigaciones Científicas y Técnicas, and Universidad de Buenos Aires, Avenida San Martín 4453, C1417DSE Buenos Aires, Argentina
    Search for more papers by this author

C. L. Ballaré. Fax: 00 54 11 4514 8730; e-mail:


In many ecological scenarios, the success of an individual plant is defined by the behavioural decisions that it makes when confronted with the risks of competition with other plants, and biomass losses to insect herbivores. These decisions involve expression of shade avoidance responses and induced chemical defences. Because these responses are costly, they frequently engender resource allocation dilemmas. In this review, I discuss the mechanisms that trigger adaptive responses to competitors and herbivores, highlighting the role of phytochromes as central organizers of the overall resource allocation strategy of plants. Phytochromes sense the reduction in the red to far-red (R : FR) ratio of sunlight caused by the proximity of other plants. Shade-intolerant plants respond to low R : FR ratios with shade avoidance behaviours and reduced investment in defence. Pfr depletion leads to increased stability of growth-promoting phytochrome-interacting factors (PIFs), and results in the production of auxins and gibberellins, degradation of DELLA proteins, which are repressors of PIFs, and reduced sensitivity to jasmonates. Thus, phytochrome appears to fulfil its organizational role by regulating the relative strength of the signalling circuits controlled by growth-related and defence-related hormones. I point out cases of signalling redundancy and discuss the significance of recent work on hormone signalling for our understanding of the mechanisms that control adaptive plant behaviour.


far-red radiation


high irradiance response




photosynthetically active radiation


FR-absorbing, active form of phytochrome


phytochrome-interacting factor


red light


Plant growth is the direct product of CO2 fixation by photosynthesis. The basic mechanism of CO2 assimilation is relatively conserved among terrestrial plants, with relatively few variants associated with the C3/C4/CAM strategies to regulate the concentration of CO2 around ribulose 1·5-bisphosphate carboxylase/oxygenase (Rubisco). In contrast, the strategies used by plants to distribute their chloroplasts in space and time so as to maximize light utilization (foraging for light), and to defend their biomass from the threat posed by herbivores, are extraordinarily diverse. These strategies use behavioural responses based on information-acquiring mechanisms that read the environment and relay the information to signal transduction cascades that ‘interpret’ the information and produce the appropriate functional adjustments (Aphalo & Ballaré 1995; Trewavas 2003, 2009). Behavioural responses may involve anything from subtle modifications of leaf chemistry and cellular organization, to radical changes in developmental patterns that are much more profound that any equivalent developmental adjustment in the animal kingdom. The extraordinary diversity and efficiency of these mechanisms allow plants to reconcile their sessile nature with the overwhelming variability in the physical and biotic parameters of their natural environment.

In this paper, I focus on the behavioural responses that plants use to optimize the capture of light in patchy canopies and to defend their biomass from the attack of insect herbivores. After a presentation of the general patterns and mechanisms, I discuss the central role played by the photoreceptor phytochrome in the perception of light signals, the assessment of actual and potential competition and the modulation of the overall resource allocation strategy.

Important developments have occurred in the last few years in the understanding of the mechanisms responsible for optimizing light capture and biomass defence, particularly in mapping the links between signal perception and the activation of hormonal cascades that elicit the terminal responses. As it may be expected, these mechanisms share multiple features with the processes that control other behavioural responses in plants. These common features are summarized in Box 1.

Box 1. Common features of networks that control plant responses to informational signals.

Signalling networks involved in the control of plant phenotypic plasticity are commonly characterized by: (1) redundant signalling; (2) pathway activation by targeted destruction of repressor elements; and (3) response modulation by negative feedback mechanisms.

Redundancy is frequently seen both at the levels of signal perception (different environmental signals trigger similar plant responses) and signalling circuits (the same signal activates parallel response channels). Redundancy and partial overlap of signal function are present in many (all?) the mechanisms that regulate essential physiological responses, and appear to be important to generate robust and finely tuned behavioural responses in complex natural scenarios.

Targeted destruction of negative regulators, via the 26S proteasome (Stone & Callis 2007), appears to be a conserved mechanism to allow rapid responses of the plant to changes in environmental and internal conditions by de-repressing temporally inactivated but otherwise fully functional signalling circuits (Huq 2006).

Negative feedback mechanisms have been described for nearly all foraging and defence mechanisms, and ensure that the response circuit is shut down when the initial signal disappears, or when the production of a long-term response is unlikely to be beneficial.



The light environment in plant canopies is complex and highly dynamic. In the simplest scenario (e.g. a mono-specific, even-aged stand), there is a steep light gradient between the canopy understorey and the top of the canopy, where PAR may vary from essentially zero below the canopy, to values that are three times higher than the light saturation point of a C3 plant above the canopy. This variation occurs over a distance of several metres in a forest, or just a few centimetres in a herbaceous canopy (Grime & Jeffrey 1965; Ballaréet al. 1988). The light environment is also heterogeneous over the horizontal plane, with steep transitions across the boundaries of canopy gaps. Within their biomechanical constraints, and the limitations imposed by nutrient and water availability, shade-intolerant plants often develop architectures and distributions of photosynthetic capacity that tend to match the distribution of the light resource (e.g. Ackerly & Bazzaz 1995; reviewed in Valladares & Niinemets 2007). Architectural plasticity is considered to be a decisive factor for competitive success in dense canopies (Küppers 1994), and it is controlled by a complex interaction of photo- and gravitropic responses, stem elongation patterns and changes in branching, developmental timing and leaf morphology. This complex syndrome of responses is commonly referred to as foraging for light, or shade avoidance (Hutchings & de Kroon 1994; Ballaréet al. 1995; de Kroon et al. 2009).


In contrast to other foraging processes in plants, for which we only know the patterns, we have a rather robust description of at least some of the mechanisms used by plants to forage for light. Genetic and molecular tools were adopted early on in ecological studies to identify which receptors play significant roles and to test the functional consequences of altered foraging behaviours (Ballaré 2001).

Multiple environmental factors are affected by the proximity of vegetation, and, in theory at least, all of these factors might serve as cues for plant shoots in their quest for sunlight. For example, the increased elongation of stems and petioles that is often seen in response to increased population density might be a response to reduced light quantity, altered light quality, reduced mechanical stress, increased relative humidity or accumulation of organic volatile compounds such as ethylene, among others. Indirect evidence from physiological experiments does indeed support a role for all of these variables (reviewed in Ballaré 1994), suggesting a high level of signalling redundancy for the critical process of foraging for light. However, only for the case of light, there is direct evidence, from manipulative physiological experiments in real canopies and genetic experiments, indicating that light signals are key modulators of adaptive plant morphogenesis.

There are essentially two types of light signals that plants read to gather information about the distribution of the light resource in their vicinity: signals associated with photon flux density (light quantity) and signals associated with spectral balance (light quality).

Light quantity signals

These signals include: changes in the flux of PAR, changes in the intensity of blue light and changes in the intensity of the R and far-red (FR) components of total radiation. In most cases, a change in PAR causes a concomitant change in CO2 fixation, and hence biomass accumulation, which may have an impact on the ability of plants to capture additional resources: the bigger the plant, the bigger its resource exploitation area. However, in addition to this size-dependent effect, changes in PAR may lead to altered plant behaviour by affecting critical parameters of plant homeostasis, such as the excitation pressure of photosystem II (PSII) and the concentration of carbohydrates in plant tissue. PSII excitation pressure is a signal that controls the expression of mechanisms involved in the definition of leaf shape and other morphological characters that are important for light harvest and utilization (Huner, Oquist & Sarhan 1998). Changes in carbohydrate levels, on the other hand, affect the expression of a myriad of genes (Koch 1996; Smith & Stitt 2007; Baena-Gonzalez & Sheen 2008), and may ultimately lead to adjustments in shoot morphologies and leaf biochemical parameters that modify the efficiency of light capture and photosynthesis. For example, in stoloniferous grasses that display procumbent growth habit in high light, sucrose depletion caused by shading leads to the production of vertically oriented shoots, thereby increasing their ability to compete for light (Willemoës, Beltrano & Montaldi 1988). At the biochemical level, changes in the carbohydrate status modify the expression of several genes involved in photosynthesis, carbon storage and utilization, with important implications for growth and resource exploitation in dynamic light environments (Koch 1996; Gonzali et al. 2006; Smith & Stitt 2007).

Variations in the levels of blue light are perceived by two families of dedicated blue light receptors: the phototropins (Briggs & Christie 2002) and the cryptochromes (Cashmore et al. 1999). Phototropins use information on the directional distribution of the blue light field to elicit phototropic movement of plant stems, leaves, roots and chloroplasts. For example, plant stems bend towards blue light, and roots show the opposite phototropic behaviour. It has been shown by means of physiological experiments that these responses are important components of the mechanisms used by seedlings to establish themselves during emergence from soil (Galen, Huddle & Liscum 2004), and established plants to find and colonize canopy gaps (Ballaréet al. 1992, 1995). The spatial distribution of chloroplasts within the mesophyll cells is also modulated by phototropins, which increase chloroplast exposure under low light and reduce it under high-light intensities (Jarillo et al. 2001; Kagawa et al. 2001; Sakai et al. 2001). These intracellular movements contribute to adjust the light-harvesting capacity of the leaf to the prevailing light environment (for a review, see Wada, Kagawa & Sato 2003). In addition to phototropic adjustments at various levels, plants respond to reductions in blue light irradiance with increased stem elongation (Ballaré, Scopel & Sánchez 1991) and, at least in some species (Pierik et al. 2004b), leaf hyponasty (but see Mullen, Weinig & Hangarter 2006). The elongation responses, which are important determinants of morphological plasticity in crowded stands, are likely to be mediated by cryptochromes, but this hypothesis has not been explicitly tested.

Reductions in the intensity of the R and FR components of solar radiation, even if there are no changes in the relative intensities of R and FR (i.e. at constant R : FR ratio), are also interpreted by the plant as a signal of shading. The photoreceptor responsible for detecting these changes in R + FR photon flux has not been directly investigated, but evidence from studies with Arabidopsis suggests that phytochrome A (phyA) is likely to operate as fluence rate sensor in de-etiolated plants (see ‘Phytochrome and foraging for light’).

Light quality signals

The most important and best characterized light quality signal is the reduction in the R : FR ratio. Because the combination of a strong absorption of R light and strong reflection of FR radiation seems to be a unique property of chlorophyll-containing tissues, a reduction in R : FR ratio is a reliable indicator of the proximity of green vegetation. Humans routinely monitor the ratio between R and FR components of back-scattered solar radiation to detect the presence of canopy cover using airborne instrumentation. Plants use the phytochromes to monitor the R : FR ratio and obtain information about the proximity and spatial distribution of other plants in the neighbourhood. Detection of low R : FR ratios triggers strong shade avoidance responses in shade-intolerant plants. These responses typically include morphological changes that redirect growth towards areas of the canopy where the R : FR ratio is higher, and therefore the risk of competition for light is minimized: phototropic movements away from areas with low R : FR, increased height growth by means of stems and petiole elongation and the production of erect leaves, reduced branching and accelerated leaf senescence in areas of low R : FR (Aphalo, Ballaré & Scopel 1999; Ballaré 1999; Smith 2000). Experiments using mutants under natural conditions of radiation have shown that phytochrome B (phyB) plays a key role in the elicitation of these responses to reduced R : FR (see ‘Phytochrome and plant defence’).

Reductions in light quantity and in R : FR ratio frequently trigger similar, redundant shade avoidance responses. Light quality perception seems to play a central role as an early signal of competition, because of the ability of phytochrome to detect FR radiation reflected from neighbouring plants (Ballaré, Scopel & Sánchez 1990). However, when the degree of shading among neighbours increases (e.g. when the leaf area index of the canopy is greater than one), light quantity signals become more intense and may play a large role as drivers of shade avoidance responses (Ballaréet al. 1991).



Plants must continuously confront the risk of losing part of their biomass to consumer organisms. In most terrestrial ecosystems, phytophagous insects represent one of the principal threats, and constitute a major link between primary producers and other trophic levels (Schoonhoven, van Loon & Dicke 2005). However, only a small fraction of the potential pool of herbivorous species actually succeeds in obtaining food from a given plant species. The reason for this is that plants use a battery of effective, dynamic defence traits to protect themselves against insect attack. Defence responses are therefore of paramount importance in the repertoire of adaptive behaviours used by plants to deal with changes in their biotic environment (Karban 2008; Metlen, Aschehoug & Callaway 2009). Direct defences are those that have an immediate impact on the herbivore, whereas indirect defences usually work by attracting natural enemies of the target herbivore. Direct defences involve the coordinated production of plant secondary metabolites and defence proteins that have repellent or toxic effects on the attacking insects (Walling 2000; Kessler & Baldwin 2002). Indirect defences are usually based on the release of volatile compounds, which attract predators and parasitoids that associate the specific plant-derived compounds with the presence of their prey (Heil 2008; Dicke 2009).

From an ecological perspective, competition and herbivory are often seen as antagonistic selective forces for plants. This is because behavioural responses activated to cope with each of these sources of stress may have a reciprocal cost in terms of the plant's ability to respond to the other. Resource allocation to competition may limit investment in defences, thereby increasing vulnerability to herbivores; and allocation to defence may reduce competitive ability against neighbouring plants. This allocation compromise, between growth and defence, is known as the ‘dilemma’ of plants (Herms & Mattson 1992; Cipollini 2004).


Like any inducible behaviour, the activation of plant defences involves the initial recognition of herbivory followed by the activation of signalling cascades that lead to the expression of the final phenotypic response.

Plants use several signals to recognize a herbivory event (for a review, see Howe & Jander 2008). These signals may include: (1) exogenous molecular elicitors that are present in the insects' oral secretions, such as fatty acid–amino acid conjugates (Alborn et al. 1997; Halitschke et al. 2001); (2) endogenous plant molecules that are modified by the activity of the insect, such as disulphide-bonded peptides (inceptins) (Schmelz et al. 2006); or (3) other endogenous signals produced by the distressed cells when the plant tissues are mechanically damaged. There are no known receptors for any of these elicitors, and the connection between these molecules and the activation of the plant defence pathways remains to be elucidated (Howe & Jander 2008).

Whereas the description of the early events that lead to the expression of the defence behaviour lags behind our understanding of the activation of shade avoidance responses (e.g. we do not know the receptors of the herbivory signals), major strides have been made in the last few years in our progress towards the elucidation of some of the key downstream regulators of plant resistance to herbivore insects. In particular, there has been significant progress in the elucidation of the roles of jasmonates in the orchestration of plant defence responses (Browse & Howe 2008; Chico et al. 2008; Katsir et al. 2008).

Jasmonates are a group of lipid regulators synthesized from linolenic acid via the octadecanoid pathway (Wasternack 2007; Howe & Jander 2008). Following attack by herbivore insects, jasmonate production is rapidly up-regulated (within minutes). Increased jasmonate production is the hormonal signal that regulates the expression of the majority of genes involved in plant defence against herbivorous insects (Reymond et al. 2000).

Recent work has revealed the mechanism of jasmonate perception (Chini et al. 2007; Thines et al. 2007). In the absence of jasmonates, positive regulators of jasmonate-inducible genes, such as the basic helix-loop-helix (bHLH) MYC2 transcription factor, are repressed by JASMONATE ZIM-domain (JAZ) transcriptional repressors. Jasmonates produced in response to herbivory are first conjugated with amino acids to form bio-active jasmonates (Staswick & Tiryaki 2004). These molecules facilitate the interaction between JAZs and the F-box protein CORONATINE-INSENSITIVE1 (COI1), promoting degradation of JAZ repressors through the activity of the E3 ubiquitin ligase SCFCOI1 and the 26S proteasome (Chini et al. 2007; Thines et al. 2007).

In a clear example of the principle of pathway activation by targeted degradation of negative regulators (Huq 2006; Stone & Callis 2007) (see Box 1), JAZ protein degradation provides a mechanism for rapid activation of defence responses by jasmonates. On the other hand, JAZ genes are among the early genes induced by jasmonate. This induction of a negative regulator by the circuit-activating molecule (jasmonate) has been interpreted as the starting point of a negative feedback loop that ensures the shutting down of jasmonate signalling cascade when the production of jasmonate ceases (e.g. when the risk of herbivory has passed; Thines et al. 2007).

Jasmonates are essential for plant resistance to a wide variety of insect herbivores. Mutants impaired in jasmonate synthesis (Howe et al. 1996; McConn et al. 1997), conjugation (Caputo, Rutitzky & Ballaré 2006; Kang et al. 2006), perception (Abe et al. 2008) or JAZ degradation (Chung et al. 2008) have been shown to be impaired in the production of anti-insect defences and/or to be more susceptible to herbivore attack in bioassays. One of the key features of the defence system in plants is that defence traits, such as production of proteinase inhibitors or secondary metabolites, are induced systemically upon herbivore attack. Jasmonates are an integral component of the signal that transmits information from the site of herbivore attack to the unattacked tissues and, therefore, are required for long distance, interorgan communication (Schilmiller & Howe 2005; Glauser et al. 2008).

In addition to their role in the induction of chemical defences and systemic communication, jasmonates are inhibitors of cell division and elongation (Yamane et al. 1980; Ueda & Kato 1982; Yan et al. 2007; Zhang & Turner 2008). Arrested plant growth may be necessary to save resources for defence or to shift allocation of carbon and nutrients away from the attacked plant organs (Babst et al. 2005; Zavala & Baldwin 2006; Henkes et al. 2008). The involvement of jasmonates in the control of both defence and growth responses suggests that they are important players in the allocation decisions that plants make when confronted, simultaneously, with the threats of competition with other plants and attack by consumer organisms. This point is addressed in the last section of this paper.


Role of phyB

The phytochromes are photochromic plant photoreceptors (Bae & Choi 2008) encoded by a family of divergent genes. There are five phytochrome genes in Arabidopsis (PHYA-E) (Clack, Mathews & Sharrock 1994; Mathews & Sharrock 1997). The shade avoidance responses to low R : FR are primarily mediated by the photoreceptor phyB (for reviews, see Ballaré 1999; Franklin 2008). The active (Pfr) form of phyB, which dominates the photoequilibrium when the plant is exposed to light of high R : FR ratio, inhibits the expression of the shade avoidance phenotype (Smith 1982, 2000). When the plant experiences a drop in the R : FR ratio, for example because it receives FR reflected by other plants (Ballaréet al. 1987, 1990), the active form of phyB is depleted. Inactivation of phyB permits accumulation of a group of growth-promoting transcription factors, thereby unleashing the expression of the shade avoidance syndrome (see next section). That phyB is the principal photoreceptor involved in shade avoidance has been demonstrated for elongation, leaf angle and FR-induced phototropic responses by means of experiments carried out in real canopies under natural sunlight levels (for reviews, see Ballaré 1999; Franklin 2008). Phytochrome D (PhyD) and phytochrome E (phyE) have been shown in physiological experiments under laboratory conditions to act redundantly with phyB in the suppression of the shade avoidance under light of high R : FR (Franklin 2008), and presumably play important roles in fine-tuning the magnitude of the response to R : FR ratio in complex light environments.

PhyA is present at very low levels in plant tissues exposed to sunlight, because of rapid light-induced proteolytic degradation of the photoreceptor in the Pfr form (Sharrock & Clack 2002). PhyA has important functions in the control of seed germination, de-etiolation and day-length perception, but has a less prominent role in neighbour detection in plant canopies. In fact, because phyA signalling is enhanced at low R : FR ratios, which reduce the rate of phyA degradation (Hennig, Buche & Schäfer 2000), phyA, acting in the so called ‘HIR’ mode, might to some extent antagonize the expression of the shade avoidance syndrome (Salter, Franklin & Whitelam 2003). PhyA is likely to be responsible for the elongation responses to attenuation of total irradiance in the R + FR waveband mentioned earlier (Ballaréet al. 1991), because higher irradiances are predicted to preserve increased levels of phyA Pfr from degradation (Hennig et al. 2000), even under light qualities that establish relatively high Pfr % (Franklin, Allen & Whitelam 2007). Tobacco plants that over-express PHYA transgenes have been shown to be unable to mount normal shade avoidance responses to neighbour proximity, and stem elongation remained inhibited even in plants grown at high density (Ballaréet al. 1994; Schmitt, McCormac & Smith 1995). This aberrant behaviour is consistent with the idea that down-regulation of phyA is important for the proper expression of proximity responses controlled by phyB and phyB-like phytochromes (McCormac, Whitelam & Smith 1992).

Phytochrome targets and downstream processes

Although the functional significance of phyB in the mechanisms of neighbour perception and foraging for light has been established some time ago, the downstream processes that connect phyB inactivation with the elicitation of shade avoidance responses have only recently began to be uncovered (Fig. 1).

Figure 1.

Simplified schematic of the functional connections between competition signals in the light environment (increased FR) and the expression of shade avoidance and defence responses. FR causes inactivation of phytochrome B (phyB) by photoconverting the active, Pfr form, into Pr. Removal of Pfr allows accumulation of phytochrome-interacting factors 4 and 5 (PIF4/5), which in turn activate several transcription factors such as ATHB2, ATHB4, etc (early SAS genes) and growth-promoting genes. These early responses are followed by activation of genes involved in the production of auxin (IAA) and bioactive gibberellins (GAs). Increased GA levels in turn facilitate degradation of DELLA proteins, which exert their biological function by inhibiting the transcriptional activity of PIFs. Phytochrome inactivation also leads to decreased sensitivity to jasmonates, and this phenomenon appears to be related to increased transcription of jasmonate signalling repressors (JAZ genes) and reduced expression of some positively acting transcription factors such as ERF1. Arrows indicate positive action (i.e. promotion, activation); blunt connectors indicate negative action (i.e. inhibition). Boxed arrows indicated protein destruction by the 26S proteasome. Dotted connectors indicate functional relationships where the molecular effectors are not firmly established.

The initial critical event for elicitation of shade avoidance responses following phytochrome photoconversion to the inactive Pr form under low R : FR ratios is the increased stability of a group of growth-promoting transcription factors known as PIFs (Lorrain et al. 2008). PIFs are members of the bHLH family of transcriptional regulators that interact specifically with the active Pfr form of the phytochrome molecule (for reviews, see Castillon, Shen & Huq 2007; Monte et al. 2007). At least two of these PIFs (PIF4 and PIF5) interact predominatly with phyB (Huq & Quail 2002; Khanna et al. 2004; Shen et al. 2007). If phyB is in the active Pfr form, it is located in the nucleus, binding PIF4 and PIF5 proteins and targeting them for degradation in the 26S proteosome; when the plant is exposed to low R : FR ratios, PhyB Pfr is photoconverted to Pr, which no longer binds PIFs and exits the nucleus, thereby permitting greater stability of PIFs (Lorrain et al. 2008). PIF4 and PIF5 promote the expression of genes involved in cell elongation (such as β expansin) (de Lucas et al. 2008) and the elicitation of shade avoidance responses (Lorrain et al. 2008). Therefore, an increase in PIF stability following phyB inactivation by low R : FR ratios results in increased growth promotion. Simultaneous mutation of PIF4 and PIF5 in Arabidopsis (pif4 pif5) reduced the hypocotyl elongation response to low R : FR ratio, and partially reversed the elongated phenotype of the phyB mutant. In contrast, over-expression of PIF5 resulted in seedlings with elongated hypocotyls even under conditions of high R : FR ratio (Lorrain et al. 2008). These physiological data are consistent with the idea that PIF4 and PIF5 are necessary for the elicitation of shade avoidance responses induced by low R : FR or by the phyB mutation. However, the observation that low R : FR ratios have a growth-promoting effect even in pif4 pif5 double mutants indicates that additional, redundant molecular circuits connect R : FR perception with the activation of shade avoidance responses (Lorrain et al. 2008). These redundant signalling pathways still remain to be elucidated.

A few genes that are rapidly (within minutes) and reversibly activated following transfer of Arabidosis plants from high to low R : FR ratio have been described. These include ATHB2, ATHB4, PIF3-like 1 (PIL1), PHY RAPIDLY REGULATED 1 (PAR1) and HFR1 (reviewed in Jiao, Lau & Deng 2007). All of them (except PAR1) encode transcription factors. The up-regulation of these early genes by low R : FR or by phyB mutation is attenuated in pif4 pif5 double mutants, and their levels of expression are increased in PIF5 over-expressors (Lorrain et al. 2008). ATHB2 and ATHB4 are members of the homeodomain leucine zipper family of transcription factors (Carabelli et al. 1993; Ciarbelli et al. 2008); ATHB2 promotes shade avoidance responses by acting as a transcriptional repressor (Steindler et al. 1999). Evidence for a functional role of ATHB2 in the induction of shade avoidance responses was obtained by studying the phenotypes of transgenic plants expressing increased or reduced levels of this gene (Schena, Lloyd & Davis 1993; Steindler et al. 1999). The underlying mechanism has not been elucidated, but it likely involves changes in auxin homeostasis (Morelli & Ruberti 2002). In a further example of the operation of negative feedback controls in phytochrome-mediated behavioural responses, the product of one of the early shade avoidance genes, HFR, down-regulates the expression of others (such as ATHB2, ATHB4, PIL1 and PAR1), and eventually attenuates the expression of the shade avoidance response (Sessa et al. 2005). This down-regulation is supposed to prevent the expression of exaggerated elongation phenotypes under conditions of prolonged shade.

The involvement of hormones, particularly auxin and gibberellins in the orchestration of shade avoidance and light foraging responses, has long been suspected on the basis of the critical roles that these hormones play in controlling physiological processes that form the core of many foraging behaviours, such as tropisms, apical dominance, stem elongation, etc. However, only recently has a map of molecular interactions between photoreceptors and hormone homeostasis begun to take shape.

A recent study demonstrated that increased auxin production through a previously uncharacterized biosynthesis pathway is necessary for the expression of early shade avoidance responses in Arabidopsis, such as the increased petiole elongation and leaf hyponasty responses that plants present when exposed to reflected FR radiation (Tao et al. 2008). The sav3 mutant was originally isolated because the seedlings had a normal phenotype in darkness and in white light, but failed to induce normal elongation responses when placed under low R : FR ratios. The mutant has normal levels of induction of early shade avoidance genes (such as ATHB2, PAR1 and HFR1), but lacks TAA1, an aminotransferase that catalyses the formation of indole-3-pyruvic acid from l-tryptophan, the first step in a novel auxin biosynthesis pathway. This pathway appears to be essential for the production of the high levels of auxin that are required to initiate shade avoidance responses (Fig. 1).

The connection between R : FR responses and gibberellins was established long ago, even before the term phytochrome was widely accepted by the scientific community. Lockhart (1959) demonstrated that the inhibitory effects of low-intensity light on internode elongation in pea, which were presumably mediated by the ‘R : FR pigment system’, were reversible by gibberellic acid. However, the functional architecture of this relationship has been unclear until very recently. Among the genes that are up-regulated by low R : FR are genes that encode enzymes responsible for the production of bioactive gibberellins, such as the GA20 oxidases (Hisamatsu et al. 2005). This is consistent with the observation that low R : FR ratios, which result in increased PIF accumulation (Lorrain et al. 2008), promote the accumulation of growth-promoting bioactive GAs in some species (Garcia-Martinez & Gil 2001; Kurepin et al. 2007) (Fig. 1). Gibberellins also exert their biological action by means of the mechanism of targeted degradation of negative regulators (Hirano, Ueguchi-Tanaka & Matsuoka 2008). Following a functional chain that has obvious parallels with the mechanisms of auxin (Dharmasiri, Dharmasiri & Estelle 2005; Kepinski & Leyser 2005) and jasmonate signalling (Chini et al. 2007; Thines et al. 2007), bioactive gibberellins stimulate the physical interaction between GID1 (the GA receptor) and proteins of the DELLA family (which are repressors of GA responses), causing their degradation through the SCFGID2/26S proteasome pathway (Ueguchi-Tanaka et al. 2005; reviewed in Hirano et al. 2008). Therefore, increased synthesis of bioactive GAs under low R : FR ratios are predicted to cause increased DELLA degradation, thereby de-repressing growth responses associated with shade avoidance (Fig. 1). DELLA abundance in fact decreases in response to low R : FR ratios and in response to increases in planting density in canopies, and the onset of DELLA degradation precedes the appearance of measurable shade avoidance responses (Djakovic-Petrovic et al. 2007). Decreased DELLA abundance is also seen when seedlings are transferred from light to darkness, presumably in response to an increase in bioactive GA levels (Achard et al. 2007)

The mechanisms used by DELLAs to inhibit growth have been elusive, in part because studies using chromatin immunoprecipitation have failed to show specific binding of DELLAs to the promoters of genes known to be responsive to GA. In an interesting twist of the phytochrome–GA connection, two research groups discovered that the PIF proteins are a principal target of DELLAs. Thus, when present in the nucleus, DELLA proteins bind to and prevent PIFs from stimulating the transcription of growth-promoting genes (Feng et al. 2008; de Lucas et al. 2008). These observations suggest that low R : FR ratios stimulate growth via at least two related mechanisms: (1) PIF stabilization by removal of PhyB Pfr (the PIF-de-stabilizing form of phyB); and (2) GA-mediated degradation of DELLA proteins, which are inhibitors of PIF transcriptional function (Fig. 1).

Ethylene is another plant hormone that appears to play important roles in shade avoidance. Low R : FR ratios promote ethylene biosynthesis in various plant species (Finlayson, Lee & Morgan 1998; Pierik et al. 2004a), and ethylene is known to cause changes in plant morphology that are reminiscent of the shade avoidance response (such as increased elongation and leaf hyponasty) (Pierik, Sasidharan & Voesenek 2007). Transgenic tobacco plants impaired in ethylene perception showed reduced shade avoidance responses when grown in density-gradient experiments (Pierik et al. 2003), and reduced response to blue light attenuation (Pierik et al. 2004b). The mechanisms that mediate the effects of ethylene defining the shade avoidance phenotype remain to be elucidated, but it is well established that ethylene promotes apoplastic acidification and up-regulates expansin gene expression and protein activity during petiole elongation in Rumex palustris (Vreeburg et al. 2005). Because ethylene is also known to be involved in the regulation of plant defence, showing positive and negative interactions with jasmonate signalling (Winz & Baldwin 2001; Lorenzo & Solano 2005), it is likely to be an important player in the functional decisions that plants implement to solve the dilemma of competition versus defence allocation (Izaguirre et al. 2006; Pierik et al. 2007).

Ecological consequences

Light-controlled foraging for light is adaptive for the individual plant (Schmitt et al. 2003). For example, transgenic tobacco plants that do not show stem elongation responses to low R : FR ratio, because they over-express PHYA transgenes and display a persistent HIR phenotype, are easily outcompeted by their neighbours when placed in a crowded neighbourhood of photomorphogenically normal plants (Schmitt et al. 1995). Mutant phyB plants of cucumber, which are impaired in phototropic responses and other biomechanical processes controlled by phyB, are much less efficient than wild-type plants at locating and deploying new growth into canopy gaps (Ballaréet al. 1995). Shade avoidance responses to low R : FR are more intense in species or ecotypes of open habitats than in those that occur in chronically shaded environments, where these responses are unlikely to result in increased PAR capture (Morgan & Smith 1979; Dudley & Schmitt 1995). Selection for plasticity to R : FR has been explicitly demonstrated in natural populations (Donohue et al. 2000).

Because competition in crowded plant stands is essentially asymmetric [larger individuals obtain a disproportionate share of the light resource (for their relative size), and suppress the growth of smaller individuals] (Weiner & Thomas 1986), plant populations tend to develop increased size structuring in response to increased population density. This phenomenon has important implications for a number of ecological issues that range from the definition of effective population size in natural populations to the growth and yield of agricultural crops. Studies using photomorphogenic mutants (Ballaréet al. 1994; Ballaré & Scopel 1997) and detailed analysis of plant allometry in density-gradient experiments (Aphalo et al. 1999; Aphalo & Rikala 2006) demonstrate that phenotypic plasticity to R : FR contributes to buffer the population against the development of size structures at high density. Shade avoidance responses mounted by plants that are being suppressed by dominant individuals allow them to remain in the competition game and eventually produce seeds. In contrast, in canopies formed by plants that are unable to respond to low R : FR ratios, shaded plants become progressively more and more suppressed without reacting with adaptive changes in their allometry, providing a rather dramatic example of the importance of information in the definition of competitive interactions between neighbouring plants.



It is often reported in the agronomic literature that pest incidence increases with crop density (Yamamura 2002). Studies that measured the concentrations of defence compounds such as phenolic derivatives (Stamp et al. 2004) and proteinase inhibitors (Cipollini & Bergelson 2001) generally found a negative relationship between accumulation of these defences and crowding. On the other hand, studies that examined the costs of defence, using physiological (Baldwin 1998; Redman, Cipollini & Schultz 2001) and genetic approaches (Zavala et al. 2004; Zavala & Baldwin 2006; Yan et al. 2007), frequently found that activation of the defence programme has a penalty in terms of growth or competitive ability. These observations are broadly consistent with the idea that competition and herbivory represent antagonistic sources of stress for plants (Herms & Mattson 1992; Cipollini 2004).

Phytochrome effects

The reduced expression of plant defences in plants subjected to competition has often been interpreted as a simple consequence of resource limitation. In fact, some of the effects of crowding on the expression of induced defences are similar to the effects of nutrient limitation (Cipollini & Bergelson 2001).

However, studies in Nicotiana longiflora indicate that competition signals can elicit a down-regulation of plant defences even in the absence of competition. In fact, plants that are induced to express the shade avoidance syndrome, by exposing them to additional FR radiation, mimicking the effects of non-shading neighbouring plants, support increased insect growth, even if they are exposed to full sunlight and supplied with non-limiting quantities of water and nutrients (Izaguirre et al. 2006). In those experiments, the increased plant tissue quality in bioassays correlated with a down-regulation of induced plant defences by FR. Thus, wounding and addition of Manduca sexta oral secretions induced a marked increase in the content of leaf soluble phenolics. FR did not have any measurable effect on non-wounded plants, but completely eliminated the phenolic response induced by simulated herbivory.

Studies in cucumber (McGuire & Agrawal 2005), tomato (Izaguirre et al. 2006) and Arabidopsis (Moreno et al. 2009) suggest that the effects of FR down-regulating plant defence are at least partially mediated by phyB inactivation. Mutants that lack phyB are more vulnerable to herbivory (McGuire & Agrawal 2005; Izaguirre et al. 2006), support more insect growth in bioassays (Moreno et al. 2009) and display reduced levels of defences (Moreno et al. 2009) when compared to wild-type plants grown under identical conditions. These observations also reinforce the idea that it is the competition warning signal (low R : FR) that triggers a down-regulation of defence, not competition itself, as both wild-type and phyB plants in those experiments were grown in individual pots and exposed to unlimited light and soil resources.

It could be argued that the reduced investment in defence and increased tissue quality in FR-exposed plants, or phyB mutants, is an unavoidable consequence of the diversion of resources to shade avoidance responses (Cipollini & Schultz 1999; Cipollini 2004). Recent experiments with the sav3 mutant of Arabidopsis demonstrate that this is not necessarily the case. As described earlier, supplemental FR fails to elicit shade avoidance responses in this mutant because it is impaired in auxin biosynthesis (Tao et al. 2008). However, the effects of FR increasing tissue quality for insect herbivores are fully conserved in this mutant (Moreno et al. 2009).

The effects of FR as a down-regulator of plant defence are likely mediated by a reduction in the sensitivity to jasmonates. A recent study (Moreno et al. 2009) showed that Arabidopsis plants exposed to supplemental FR radiation have attenuated phenolic and defence gene expression responses to methyl jasmonate compared with control plants grown under normal R : FR ratio. A similar reduced sensitivity to jasmonates was evident in phyB plants. Desensitization towards jasmonates could be effected by several mechanisms, including reduction in the formation of bioactive conjugates or in the levels of the COI receptor, increased transcription or stability of JAZ repressors and down-regulation of some of the jasmonate-activated, down-stream signalling components, such as MYC2, ERF1, etc. The transcription of some JAZ genes was up-regulated under conditions of low R : FR, whereas that of ERF1 was down-regulated (Moreno et al. 2009). The up-regulation of some JAZ genes, including JAZ10 (Moreno et al. 2009), is particularly interesting in the light of recent evidence showing that one splice variant of JAZ10 (JAZ10.4) is highly resistant to jasmonate-mediated degradation and, when over-expressed, confers jasmonate insensitivity (Chung & Howe 2009). These observations suggest a major role of competition signals in controlling some of the key players of jasmonate perception in Arabidopsis (Moreno et al. 2009) (Fig. 1). Another study (Navarro et al. 2008) has additionally shown that quadruple DELLA mutants, or plants treated with GA to reduce DELLA stability, have reduced jasmonate sensitivity. As explained previously, low R : FR ratios reduce DELLA stability (Djakovic-Petrovic et al. 2007). Therefore, DELLAs may be important in connecting competition signals with the expression of plant defences, perhaps via their recently established role as negative regulators of PIF function (Feng et al. 2008; de Lucas et al. 2008) (Fig. 1).

Ecological consequences

The available evidence suggests that when increased levels of FR radiation signal a period of intense competition, shade-intolerant plants place their allocation priorities on maintaining their light-harvesting ability rather than on preventing biomass loss to herbivores. The plant seems to generate this response to the dilemma of competition versus defence allocation at least in part by using information on the risk of competition sensed by phytochrome to modulate its sensitivity to jasmonates.

Modulation of induced defences by phytochrome appears to be a widespread phenomenon (Kurashige & Agrawal 2005; McGuire & Agrawal 2005; Izaguirre et al. 2006; Moreno et al. 2009), with important implications for strategic resource allocation in plants. Selective desensitization to jasmonates (Moreno et al. 2009) could save plant resources by reducing the investment in defence and, at the same time, avoid the inhibitory effects of jasmonates on cell growth and division (Yan et al. 2007; Zhang & Turner 2008). If not suppressed by phytochrome signals, these growth-inhibitory effects of jasmonates may have negative impacts on plant fitness under conditions where the plant has to elongate rapidly to escape shading by its neighbours.

One aspect that has not been investigated is the distribution of the FR-induced defence down-regulation response throughout the plant. Because of the modular nature of plants, it is conceivable that the FR effect on sensitivity to herbivory cues is only expressed in certain parts of the shoot, such as basal leaves. Lower leaves have a lower potential to contribute to light capture and carbon fixation in a rapidly growing canopy, and are also the ones that experience the strongest shading and R : FR signals. Therefore, it would not be surprising to discover that these basal leaves are the first to be desensitized to jasmonate if the functional purpose of this desensitization is to save resources for investment in shade avoidance responses.


The role of phytochrome as a central regulator of the shade avoidance syndrome and foraging for light is now well established. The mechanisms that mediate this central component of plant behaviour in canopies have been unclear until very recently. Major experimental and conceptual advances in our understanding of the mechanisms of hormone action, and the connections between phyB inactivation and hormonal signalling (particularly auxin and gibberellins), are beginning to reveal the network of interactions that translates altered levels of phyB Pfr into changes in growth rate and overall morphology (Fig. 1).

The discovery of the significance of increased auxin biosynthesis for the expression of rapid shade avoidance responses, and the finding that PIFs are targets of DELLA proteins have contributed two fundamental pieces to solve the puzzle of shade avoidance elicitation in plants exposed to light of low R : FR ratio. This conceptual advance provides a functional framework to investigate the mechanisms that mediate the responses to other proximity signals, such as reduced total phytochrome-absorbable radiation and blue light levels.

A significant recent discovery is that phytochrome is an important regulator of the expression of anti-herbivore defence mechanisms under natural conditions, and therefore a key player in the behavioural decisions implemented by plants when confronted with the classic dilemma of distributing resources between growth and defence activities. Here again, the advances in our understanding of hormonal signalling, particularly jasmonate perception, have given us the tools to begin to elaborate a functional map of the relationships between the mechanisms of light perception and defence activation. The connections between molecular targets activated by Pfr depletion and elements of the jasmonate signalling pathway seem to be critical nodes for the control of the overall strategy of plants in canopies, as these connections are likely to regulate both the investment in defence and the effects of jasmonates as inhibitors of plant growth. Future work needs to address the regulation of jasmonate signalling elements by phytochrome and other informational photoreceptors.

Jasmonate signalling is also influenced by other hormones, such as ethylene, salicylic acid and gibberellins (Lorenzo & Solano 2005; Navarro et al. 2008). All of them are affected to some extent by phytochrome, and in turn have their own particular role as regulators of growth and defence against herbivores and pathogen microorganisms. Therefore, additional points of interaction between shade avoidance and defence will almost certainly be discovered in the near future.

Complex interactions in plant signalling networks were anticipated long ago, and the emerging body of data is revealing dense, intricate and flexible connections among signalling players in the cases of informational networks activated by light and hormones (Nemhauser, Hong & Chory 2006; Nernhauser 2008). Indeed, the available evidence for phytochrome responses suggests that we have barely begun to scratch the surface of a world of bewildering complexity, and are making the first steps in our attempts to draw a preliminary chart of mechanistic interactions among the key molecules and signalling components that regulate plant functional decisions in canopies.


I thank Miriam Izaguirre, Javier Moreno, Carlos Mazza and Amy Austin for many helpful discussions. Work in my laboratory is supported by grants from ANPCyT and UBACyT.