Ecological implications of plants’ ability to tell the time


  • Víctor Resco,

    Corresponding author
    1. Department of Renewable Resources, University of Wyoming, Laramie, WY 82071, USA
      *Correspondence and present address: Departamento de Ciencias Ambientales, Universidad de Castilla-La Mancha, Toledo, 45071, Spain.
    Search for more papers by this author
  • James Hartwell,

    1. School of Biological Sciences, University of Liverpool, Bioscience Building, Crown Street, Liverpool, L69 7ZB, UK
    Search for more papers by this author
  • Anthony Hall

    1. School of Biological Sciences, University of Liverpool, Bioscience Building, Crown Street, Liverpool, L69 7ZB, UK
    Search for more papers by this author

*Correspondence and present address: Departamento de Ciencias Ambientales, Universidad de Castilla-La Mancha, Toledo, 45071, Spain.


The circadian clock (the endogenous mechanism that anticipates diurnal cycles) acts as a central coordinator of plant activity. At the molecular and organism level, it regulates key traits for plant fitness, including seed germination, gas exchange, growth and flowering, among others. In this article, we explore current evidence on the effect of the clock for the scales of interest to ecologists. We begin by synthesizing available knowledge on the effect of the clock on biosphere–atmosphere interactions and observe that, at least in the systems where it has been tested, the clock regulates gas exchange from the leaf to the ecosystem level, and we discuss its implications for estimates of the carbon balance. Then, we analyse whether incorporating the action of the clock may help in elucidating the effects of climate change on plant distributions. Circadian rhythms are involved in regulating the range of temperatures a species can survive and affects plant interactions. Finally, we review the involvement of the clock in key phenological events, such as flowering time and seed germination. Because the clock may act as a common mechanism affecting many of the diverse branches of ecology, our ultimate goal is to stimulate further research into this pressing, yet unexplored, topic.


The circadian clock, the endogenous mechanism that anticipates diurnal cycles, is currently considered a central ‘director’ of plant physiological activity that coordinates the multitude of processes plants undergo through development to survive, grow and reproduce in the face of, sometimes unpredictable, changes in the environment. Plant scientists have favoured research into the model C3 species Arabidopsis thaliana. About 90% of the transcriptome in Arabidopsis is expressed rhythmically under some condition or other (Michael et al. 2008). Although the molecular underpinnings of the clock have been mostly explored in Arabidopsis, the clock has been present and conserved for at least c. 450 million years of land plant evolution (Rensing et al. 2008). In the wild species where the clock has been studied, a large degree of similarity with that of Arabidopsis has been observed (Ramos et al. 2005; Murakami et al. 2007; Slotte et al. 2007; Yakir et al. 2007). The clock drives the rhythmic expression of ‘output’ genes. These clock-regulated outputs include key traits for plant fitness, such as fluxes of water, carbon and volatile organic compounds (Dodd et al. 2005; Wilkinson et al. 2006), nutrient uptake (Harmer et al. 2000), growth (Nozue et al. 2007), flowering (Slotte et al. 2007) and seed germination (Saloméet al. 2008). Indeed, it has been experimentally and repeatedly observed that plants with impaired clock function show reduced fitness (Highkin & Hanson 1954; Green et al. 2002; Dodd et al. 2005). Despite the high relevance of the clock for coordinating plant physiological activity, the potential ecological relevance of this process has not yet been studied. In this article, we explore whether the role of the clock in driving plant physiological activity is relevant for the higher levels of organization studied by plant ecologists. We are particularly interested in exploring the potential of the clock as a mechanism of common interest for ecologists working in biosphere–atmosphere interactions and in plant distributions under a changing climate.

In the first section of this article we briefly define and describe the structure of circadian clocks. In the second section, we explore the relevance of the clock in driving the fluxes of water, CO2 and biogenic volatile organic compounds (BVOCs), such as isoprenoids. We review current evidence on whether these fluxes are under circadian control and explore the diverging views in the literature regarding whether circadian control of gas exchange is relevant in the field (Williams & Gorton 1998; Doughty et al. 2006). We also discuss whether circadian control on gas exchange matters at higher levels of organization (from the leaf to the ecosystem and beyond).

In the third section, we assess the clock whether the range of environmental conditions influences a species can survive (the fundamental niche) and its role in mediating plant interactions (the realized niche). Throughout this section, we constantly draw links to encourage research into how the clock may modulate the effects of climate change on species distributions. For instance, one of the most important axes in the multidimensional space that constitutes the fundamental niche is temperature, which strongly constrains the distributions of species worldwide (Whittaker 1975). In the section on fundamental niche, we explore the potential role of the circadian clock in determining the range of temperatures and photoperiod where a species is able to inhabit. The bulk of this knowledge has been developed in laboratory settings and with model species, while field tests of this concept are still rare. Therefore, we first outline available evidence in favour or against this hypothesis at the molecular level. Then we draw hypotheses as to whether the role of the circadian clock in determining the range of temperatures a plant is able to tolerate is relevant to levels of interest for ecologists, and what are the potential implications. Similarly, in the section on plant interactions we begin by providing a brief outline of the molecular control underlying plant interactions, and then move to investigating the potential implications of this control in determining the realized niche.

In the last part of this section, we focus on the potential involvement of the clock in regulating different aspects of phenology. Our current understanding of how global change factors affect phenology is mainly based on experimental or observational studies (Menzel et al. 2006), but we are still far from understanding the mechanisms and elaborating accurate process-based models. Phenological processes, such as flowering time, leaf drop, seed germination, fruiting, winter dormancy, flower opening and scent emission are controlled by photoperiod. In turn, all photoperiod-driven processes studied to date appear to be controlled by the clock (Yakir et al. 2007). Because of space constraints, in the section on the clock regulation of phenology we mainly focus on flowering time and seed germination, two key stages for plant fitness.

Knowledge of circadian biology has increased drastically over the last few years. Therefore, we do not aim at a comprehensive review of the ecological relevance of all circadian-controlled processes, but arbitrarily select a few processes, key to disentangling ecological dynamics. Our ultimate goals are to explore the ecological relevance of the circadian clock as a mechanistic driver of field patterns to act as a common theme across all areas of ecology, and to encourage further research into this, relatively new, field. In other words, we aim at synthesizing current evidence on whether and when circadian regulation matters for higher levels of organization, from population and community dynamics, to biosphere–atmosphere interactions and biogeography.

Definition and description of the circadian clock

A biological rhythm, to be considered ‘circadian’, has to (McClung 2006): exhibit a c 24 h period (time to complete a cycle) of around 24 h; be endogenously generated such that it persists under ‘free-running’ conditions (constant levels of light or darkness, and temperature); and maintain a constant period, or circadian resonance, over a broad range of temperatures (temperature compensation).

A circadian system is composed of environmental cues that regulate the period of the ‘central oscillator’ of the clock. This oscillator generates circadian rhythmicity in various biological processes, which affect the expression of the outputs. The basic oscillator responsible for rhythmicity is studied in different model organisms, and involves a series of highly conserved transcriptional–translational positive and negative feedback loops. Recent advances in this area have been summarized by many, including Hartwell (2006), McClung (2006, 2008) and Hotta et al. (2007).

Clock effects on biosphere–atmosphere interactions

Evidence towards circadian regulation of gas exchange

Rhythmicity in the regulation of stomata was documented over 100 years ago by the pioneering work of Darwin (1898). During the 1970s and 1980s, scientists began to observe circadian regulation of leaf stomatal conductance (gs) and photosynthesis (A) under free-running conditions (Hennessey & Field 1991), with maximum values occurring around noon and minimum near midnight (Fig. 1a). It is now well established that circadian rhythms in photosynthetic gas exchange are also temperature-compensated across a broad range of temperatures in plants from different photosynthetic pathways (Webb 2003; Hartwell 2006; Hotta et al. 2007).

Figure 1.

 Circadian regulation of gas exchange. (a) Diurnal pattern of photosynthesis, stomatal conductance and isoprene emissions (gas exchange) in wild-type plants with a carefully optimized clock (continuous line), and in arrhythmic mutants (dashed line) under constant levels of light and temperature (free-running). The period that under normal conditions would have been dark, is indicated by bars on the x-axis. This conceptual figure has been derived from the experimental work of Hennessey & Field (1991), Dodd et al. (2005), Wilkinson et al. (2006) and Doughty et al. (2006), which all show the same pattern for the different components of gas exchange. (b) Experimental work has shown a reduction in photosynthesis and in growth of up to 40% in the wild-type as compared with arrhythmic mutants (Dodd et al. 2005). (c) The effect of the clock is relevant to the ecosystem level. Net ecosystem exchange (continuous line, relative units) increased fivefold from 6 to 10 am in an Amazonian forest, even in periods when photosynthetic photon flux density was between 500 and 600 μmol m−2 s−1 (modified from Doughty et al. 2006).

Substantial progress was made during the 1990s and 2000s to understand the molecular mechanisms by which the circadian clock controls diel fluctuations of A and gs in the model C3 species Arabidopsis thaliana (Liu et al. 1996; Harmer et al. 2000), and in the model CAM plants Kalanchoe fedtschenkoi and Mesembryanthemum crystallinum (Hartwell et al. 1999; Boxall et al. 2005). The molecular control of photosynthetic gas exchange has been studied by Webb (2003), Hartwell (2006) and Hotta et al. (2007). We are unaware of any detailed reports on the molecular control of photosynthetic gas exchange in C4 species.

The most extensive study to date testing whether the circadian clock regulates observed patterns of photosynthetic gas exchange in the field was conducted by Doughty et al. (2006). They monitored A and gs in free-running conditions in 17 tropical species from different functional groups (large-canopy or subcanopy trees, understory shrubs and lianas) by keeping the leaf within a cuvette at constant levels of light and temperature. They concluded that the clock drives the diel pattern of photosynthetic gas exchange in the field in 15 of the 17 tropical species studied (Fig. 1a).

Diel fluctuations in the level of isoprenoid emissions have been traditionally related to changes in photosynthetically active radiation and temperature. This has resulted in a surprising paucity of data on circadian regulation of these secondary metabolites. For instance, the first test on circadian control over isoprene emissions was performed only 2 years ago (Wilkinson et al. 2006). The authors observed that isoprene emission rates in the tropical tree Elaeis guineensis (oil palm) under free-running conditions showed circadian regulation for several days, with maximum and minimum emission rates around noon and midnight, respectively (Fig. 1a). This circadian rhythmicity in isoprene emissions was temperature-compensated between 25 and 38 °C. Similarly, Wang et al. (2007) observed that the emission rate of the isoprenoid cis-ocimene showed circadian oscillations under free-running conditions in the rubber tree (Hevea brasiliensis).

Although there are currently no estimates on how widespread the circadian control of isoprene emissions is, it could be hypothesized that it is quite a widespread phenomenon because: (1) the activity of the methylerythritol (MEP) pathway, which provides substrates for the synthesis of numerous isoprenoids, undergoes circadian regulation (Dudareva et al. 2005); (2) the genetic structure of the clock is highly conserved and its control over isoprene emissions seems to operate at the transcriptional level (Wilkinson et al. 2006); and (3) circadian regulation over BVOC emissions has been confirmed in phylogenetically distant species, growing in contrasting environmental conditions (Lu et al. 2002; Wilkinson et al. 2006; Loivamaki et al. 2007).

Potential implications of the circadian regulation of gas exchange at higher levels of organization

Photosynthesis and water use efficiency directly influence the carbon balance of the plant, and a mechanistic link between A and growth has been described (Smith & Stitt 2007; Resco et al. 2008). Hence, it could be hypothesized that circadian regulation of photosynthesis affects growth.

Analyses of the transcriptome and metabolome of Arabidopsis (Harmer et al. 2000; Edwards et al. 2006; McClung 2008) indicate that the circadian clock affects gas exchange by ‘anticipating’ cycles of dawn and dusk. For instance, proteins involved in the light-dependent photosynthetic reactions become upregulated before dawn and peak at the middle of the subjective day (Harmer et al. 2000; Edwards et al. 2006), as do transcript levels regulating the expression of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) (Liu et al. 1996). Similarly, in the CAM plants Kalanchoe and Mesembryanthemum, the peak in gene transcript abundance and activity of phosphoenolpyruvate (PEP) carboxylase (PEPc) kinase (PPCK) and in the Ki of PEPc for feedback inhibition by malate, occurs during the dark period under the direct control of the central circadian clock (Hartwell et al. 1999; Boxall et al. 2005).

Because the circadian clock is the mechanism by which plants ‘anticipate’ transitions from dawn to dusk, it could be hypothesized that plants without circadian resonance would show low efficiency in photosynthesis which, in turn, would result in lower plant growth (Smith & Stitt 2007). This hypothesis was tested by Dodd et al. (2005), who compared rates of A in wild-type Arabidopsis, with mutants showing impaired clock regulation. They observed a 40% decrease in net carbon fixation in arrhythmic mutants, as compared to the wild-type (Fig. 1b). In turn, aboveground biomass was also around 40% lower in arrhythmic mutants than in wild-type plants (Fig. 1b).

One of the classic paradigms in the physiological ecology of gas exchange is that leaves ‘track’ environmental changes. gs, for instance, is often considered to be tightly coupled with diel fluctuations in atmospheric conditions. These fluctuations affect the water potential of the cells surrounding the stomata, via changes in radiation, leaf-to-air vapour pressure deficit and leaf or stem hydraulics (Brodribb et al. 2003; Buckley 2005; Resco et al. 2009). gs typically increases with light during the morning until around noon, when atmospheric ‘demand’ of water increases and, consequently, gs drops. In turn, A is considered to be coupled with gs, because of the changes in Ci that accompany the variation in gs. This diurnal fluctuation in photosynthetic gas exchange is a common topic of interest for scientists working at the leaf and ecosystem scales, who aim at understanding the drivers in the diurnal cycle of fluxes, as well as for climate modellers, who often need to know the drivers of these fluxes to model their effect on future climate dynamics.

Photosynthetic gas exchange is, in general, strongly driven by environmental relationships. However, there is now convincing evidence that, at least to explain two particular and important cases, diurnal changes in A and gs cannot be interpreted based on environmental changes alone, but circadian regulation must also be included. The first of these cases is the observed pattern of variability in nighttime stomatal conductance. Nocturnal gs is not kept at a constant minimum throughout the night. Results from a synthetic analysis across phylogenetically distant wild species (Caird et al. 2007) show that gs increases shortly before dawn. Because this result matches the pattern of gs observed for model C3 species in free-running conditions (Fig. 1a, Webb 2003), it has been attributed to a circadian-mediated regulation (Caird et al. 2007).

The second case – where predicting photosynthetic gas exchange solely based on environmental conditions is difficult and the incorporation of circadian regulation is necessary to diminish the uncertainty – is to understand the morning increase in net ecosystem exchange (NEE, the balance between CO2 assimilation and respiration; Doughty et al. 2006). The most widely established method to measure ecosystem fluxes of CO2 and water is the eddy covariance technique. Doughty et al. (2006) analysed 16 138 half-hour eddy covariance measurements of ecosystem NEE between 2000 and 2004, and selected those periods when photosynthetic photon flux density remained constant between 500 and 600 μmol m−2 s−1. They observed that the uptake of CO2 between 06:45 and 09:45 hours increased from 3.6 to 16.1 μmol m−2 s−1 (Fig. 1c), and neither soil respiration rates nor other atmospheric conditions (such as leaf-to-air vapour pressure deficit) changed significantly (Fig. 1c). Doughty et al. (2006) also measured leaf photosynthesis in situ under free-running conditions. They concluded that this fivefold increase in NEE corresponded well with the measured increase in leaf photosynthesis under free-running conditions in their field experiment, as well as in laboratory experiments (Fig. 1a, Dodd et al. 2005; Doughty et al. 2006), and that this morning increase in NEE was particularly difficult to explain if gas exchange was solely driven by the physical environment.

We are not aware of any other study examining whether the clock affects NEE, and therefore the results from this experiment must be interpreted with caution. However, testing the effect of the clock on NEE is an urgent research need, which could have huge implications for eddy covariance studies and, by extension, for current estimates of the global carbon balance. It is not rare that 40%, or more, of the data collected by this technique has to be discarded because environmental conditions did not match the theoretical requirements of this method (e.g. low turbulence during the night; Scott et al. 2004). In order to fill these data gaps and compute an annual carbon balance, researchers typically rely on empirical tables based on environmental conditions associated with the missing data or nonlinear regression techniques (Moffat et al. 2007). This approach is clearly not appropriate when NEE may vary between 3.6 and 16.1 μmol m−2 s−1 as a function of time of the day, in a period with relatively constant atmospheric conditions, as those reported in Doughty et al. (2006).

Fluxes of CO2 are often scaled up, from leaves to canopy, by modelling A by using the ‘Farquhar, von Caemmerer and Berry’ model, which calculates maximum carboxylation capacity (and other biochemical parameters) from A/Ci curves (de Pury & Farquhar 1997; Cramer et al. 2001). Because of logistic difficulties, this measurement is often performed throughout the day in order to achieve sufficient replication. However, if photosynthetic gas exchange undergoes circadian oscillations in a wide range of species, different estimates of leaf biochemical capacity will be obtained depending upon the time of the day when the A/Ci curve was measured. Contrary to this hypothesis, Williams & Gorton (1998) argued that incorporating circadian regulation into current field models of gas exchange (e.g. Farquhar & von Caemmerer 1982; Leuning 1995) did not significantly improve the model fit. This result contradicts the experimental evidence on circadian regulation of gas exchange outlined above, and Webb (2003) hypothesized that it is an artefact that originates from the small data set evaluated in that model. Further research into this topic is necessary to clarify these opposing views of the relevance of circadian regulation of gas exchange in the field.

There are currently no estimates on how the control exerted by the clock on the emission of isoprenes influences rates at the ecosystem level. However, it is feasible that similar arguments as those proposed for the circadian regulation of photosynthetic gas exchange apply. Indeed, Beerling et al. (2007) considered the lack of circadian control of isoprene emissions in global trace gas models as a critical factor, potentially giving rise to large uncertainties in the predictions.

It seems reasonable to propose that the advantage in a circadian control of BVOC emissions is also related to the ‘anticipation’ hypothesis, proposed for photosynthetic gas exchange. In a recent Affymetrix oligoarray analysis where we compared the transcriptome of wild-type Arabidopsis with that in CCA1-ox arrhythmic mutants, we observed a downregulation in steady-state abundance of the transcripts associated with the production of secondary metabolites, such as terpenes, flavonoids and phenolic compounds, in the arrythmic mutants as compared with the wild-type 1 h before dawn (V. Resco, J. Hartwell, A. Hall unpublished data). Ongoing experiments will test whether these changes in transcript abundance correlate with changes in the abundance of these metabolites.

Clock effects on species distributions and climate change responses

Range of abiotic conditions a species can inhabit

A complex regulatory network at the genomic, transcriptomic and proteomic level is necessary to maintain plant biochemistry from succumbing to large diurnal and seasonal fluctuations in light and temperature. There is now a large body of literature proving that the circadian clock is tightly coupled to light and cold signalling. Three CBF (C-REPEAT BINDING FACTORS) transcription factors, induced by ICE-1 (INDUCER OF CBF EXPRESSION1), activate the expression of COLD RESPONSIVE (COR) genes, which lead to cold acclimation. In turn, the clock drives the rhythmic expression of the CORs through different feedback loops involving the PRRs (PSEUDO-RESPONSE REGULATOR) 1, 3, 5, 7 and 9 (Chinnusamy et al. 2007) and the CCA1 (CIRCADIAN CLOCK ASSOCIATED1) transcription factor (Fig. 2). Light signalling through the phytochrome controls the clock genes CCA1, LHY (LATE ELONGATED HYPOCOTYL), GI (GIGANTEA) and TOC1 (Alabadíet al. 2001;Hotta et al. 2007; Josse et al. 2008). The rhythmic expression of CBF transcription factors, that leads to cold acclimation, peaks late in the afternoon, and has been directly linked to the clock and to the phytochrome (Fowler et al. 2005; Franklin & Whitelam 2007).

Figure 2.

 Simplified representation of the interaction between the circadian clock, light signalling and cold acclimation (more comprehensive schemes are provided by Chinnusamy et al. 2007 and Hotta et al. 2007). Light activates the circadian clock through the phytochrome, in a rather complex manner (indicated by the lack of arrow ending). In turn, the central oscillator of the clock, represented by the triple feedback loop model, regulates the rhythmic expression of the CBFs, which, then activate the CORs, leading to cold acclimation (Chinnusamy et al. 2007). Complete details are given in the text.

The involvement of light and the clock in regulating the response to high temperatures has not been developed to the same depth as that for cold stress. However, Gould et al. (2006) show that an analogous mechanism may be operating for heat acclimation.

It has been clearly demonstrated that accurate and robust clock function enhances performance and fitness, and that clock impairment leads to poor signalling and even plant death, under some conditions. For instance, Dodd et al. (2005) compared different proxies for fitness across Arabidopsis genotypes with appropriate and impaired circadian resonance. They observed that those genotypes with impaired resonance showed a much worse phenotypic performance than those with accurate resonance, and even death in certain scenarios. Similarly, Nozue et al. (2007) observed that growth in Arabidopsis arrhythmic mutants was severely impaired as compared to that in the wild-type, because the circadian clock modulates (gates) the effect of light, through the phytochrome receptors, on growth. This evidence suggests that circadian resonance is an important feature in enhancing performance and fitness of plants at high and low temperatures, and under different photoperiods (Dodd et al. 2005; Gould et al. 2006).

The next stage to understand whether the circadian clock has ecological relevance in the determination of the range of abiotic conditions a species can survive, is to test whether the temperature range where robust and accurate clock function can be maintained could define the temperature range where a plant species or ecotype could survive, grow and reproduce. This is particularly important under recent predictions of global warming.

There is a growing number of studies on the effects of warming on plant community dynamics and growth (Peñuelas et al. 2007; Sebastià 2007; Reich & Oleksyn 2008). Unfortunately, the mechanisms underlying the response of different plant functional types to warming has not always been carefully examined, and a large degree of species-specific or site-specific differences are often observed (Luo 2007). The next step would be to test whether the structure and role of temperature compensation by the circadian clock should be incorporated to understand the mechanisms underlying this differential response in biomass, and diminish uncertainties in predictions of changes in species distributions after climate change. Preliminary results from ongoing experiments suggest that differences in fitness at high temperature increase between arrhythmic Arabidopsis mutants and the circadian-resonant wild-type (Costa et al. 2008). Understanding the potential role of the clock in buffering temperature stress in the wild is an impending research need.

Another potential application of the clock in determining the range of temperatures a species can tolerate is to develop heat-tolerant varieties of economically relevant species, or to preserve keystone species in geographical settings where, under warmer conditions, they will not be able to survive. Indeed, the Research Councils in Biotechnology and Biological Sciences and in Engineering and Physical Sciences in the United Kingdom funded in 2007 a 6 million sterling pounds project to produce crops resistant to heat stress (Halliday 2008). The theoretical basis of this project is the interlinked signalling network between temperature and light signalling with the clock described at the beginning of this section.

Circadian regulation of plant–plant interactions

Chlorophyll selectively absorbs blue and red wavelengths (400–500 and 600–700 nm, respectively), whereas far-red radiation (700–800 nm) is largely reflected or scattered. This decreases the proportion of red to far-red (R/FR) light from the top to the bottom of the canopy, changing the equilibrium between the Pr and Pfr forms of phytochromes B, D and E (Franklin et al. 2003). This provides a signal for the presence of potential competitors nearby, triggering the ‘shade-avoidance’ response, which aims at overgrowing neighbouring plants through the elongation of internodes and petioles, reductions in leaf growth and increased apical dominance (Smith & Whitelam 1997). When plants are exposed to low R/FR signal for prolonged periods of time, flowering is accelerated. We have previously described the strong links between light availability, phytochrome photoreceptors and the circadian clock (Fig. 2). Indeed, expression of all the plant photoreceptors, with the exception of PHYC, is clock-controlled. In turn, the clock gates this shade-avoidance response and mediates in the short-term responses to light availability, as well as in the accelerated growth associated with shade-avoidance (Salter et al. 2003).

This circadian gating of the shade-avoidance response likely contributes to the strong differences that arise in the phenotypic performance and survival between Arabidopsis plants with a carefully optimized clock and those that show poor circadian resonance, when both genotypes were grown together in competition experiments (Dodd et al. 2005). Arabidopsis arrythmic mutants and wild-type plants where the endogenous clock did not match the external light–dark cycle, thus plants without optimised circadian regulation, showed lower chlorophyll content, leaf number, rosette diameter and aerial biomass than those accessions with circadian resonance in competition experiments (Dodd et al. 2005). Moreover, increased mortality rates were observed for accessions lacking circadian resonance (Dodd et al. 2005).

A carefully optimized clock could be essential to optimize the timing of resource uptake in the presence of competitors. For instance, plants with circadian resonance could out-compete neighbouring plants with less-optimal circadian control with respect to water use. An extreme example would be CAM plants, which have effectively moved their water requirements to the dark period under the control and optimization of the clock and thus do not have to compete with surrounding C3 plants which transpire in the light period (Hartwell 2006). Similar arguments are likely to be applicable for nutrient competition, with the added point that expression of genes involved in nitrogen uptake and assimilation is known to be clock-controlled, as is the expression of genes involved in the uptake and assimilation of a number of other key nutrients including phosphorus and sulphur (Harmer et al. 2000; Gutierrez et al. 2008).

Most, but not all, of the studied plant species growing in natural environments show accurate circadian resonance (Doughty et al. 2006). Plants without accurate optimization of resource uptake have been observed to be in a competitive disadvantage, as compared with those showing a highly optimized clock (Dodd et al. 2005). Competition is a strong evolutionary force that deeply modifies the structure of plant communities (Tilman 1988; Strauss & Irwin 2004). Thus, future research should establish whether circadian resonance is a strong selective agent, particularly under the strong competition interactions that plants must deal with in the field. Establishing the potential role of the clock in the evolution of plants under competitive scenarios may prove a key, yet unresolved, question in plant biology.

Predicting changes in phenology from anticipated global warming

Flowering is one of the most important phenological processes. It is directly related to fitness, it has been shown to respond to climate change and it is routinely measured in phenology databases. Slotte et al. (2007) studied differences in gene expression between two pairs of early- and late-flowering accessions of Capsella bursa-pastoris to identify pathways regulating the reported variation of flowering time in the wild. They observed a highly contrasting clock behaviour between the early- and late-flowering accessions. More specifically, the diurnal rhythm of CCA1 and TOC1, two important components of the clock ‘core’ (Fig. 2), was differentially expressed across accessions. These differences in the clock behaviour lead to the changes in gene expression responsible for variations in flowering time. Thus, it was concluded that CCA1 and TOC1 are ‘strong candidates for the evolution of adaptive flowering time variation’.

One of the most comprehensive analyses on changes in flowering time in the wild after climate change was performed by Menzel et al. (2006). They observed that almost 80% of all the records showed an advancement in flowering time in response to warming, across a large phenology network comprising 21 European countries. The circadian clock has a rather complex molecular structure, which allows for relatively fast adaptation or acclimation to environmental changes (Rand et al. 2006). As previously explained, the circadian clock is related to changes in flowering time between the early- and late-flowering accessions in C. bursa-pastoris, and these changes in flowering time were developed in response to the differences in temperatures in their sites of origin (Slotte et al. 2007). Therefore, a logical next step to understand the mechanisms underlying reported changes in flowering time after global change is to test whether: (1) the ‘functional genomics’ of the circadian clock has changed in those species after climate change, and (2) these changes in the clock are responsible for the advancement of flowering time.

The clock may also be involved in the early and key phenological stage of seed germination (Zhong et al. 1998). Seed germination is controlled by the photoperiod in a broad range of species from distant phylogenetic and geographic origins such as Diapensia lapponica, Chamaedaphne calyculata, Ledum decumbens, Saxifraga tricuspidata, Betula pubescens, Cyperus inflexus, Arabidopsis thaliana and many others (Baskin & Baskin 1998; Zhong et al. 1998; Yakir et al. 2007). The circadian clock has been shown to control all photoperiod-regulated processes, which suggests that the circadian clock could also be active during seed germination (Yakir et al. 2007). Indeed, water absorption (imbibition) sets the clock among seeds of Arabidopsis and Apium graveolens, and serves to synchronize the clock at the population level (Zhong et al. 1998; Thomas 2002). This clock synchronization then serves to time the buried seedlings with the local daily cycles before emergence (Saloméet al. 2008).

The timing of life cycle events is, in general, regulated by environmental cues of temperature and photoperiod (Ratchke & Lacey 1985; Jackson 2008), the two major entraining factors of the circadian clock. Research performed so far on photoperiod-regulated processes, such as leaf drop, fruiting, seed germination, winter dormancy, flowering time, flower opening and scent emission, has demonstrated that these processes undergo circadian regulation in Arabidopsis, as well as in ecologically and agronomically relevant species like chestnut, birch, tomato and tobacco (Roden et al. 2002; Ramos et al. 2005; Yakir et al. 2007).

Climatic and atmospheric conditions during the 21st century are expected to change dramatically, and at much faster rates than past climatic changes. Shifts in temperature, precipitation and nutrient availability are predicted to be widespread across the globe. In turn, changes in these factors have been reported to deeply impact species distributions and phenological processes (Sala et al. 2000; Peñuelas & Boada 2003; Peñuelas et al. 2004, 2007; de Dios et al. 2007; Menzel et al. 2006). We have described how temperature, water and nutrients have been reported to interact with the clock and modulate the range of environmental conditions Arabidopsis and other model species can tolerate, their competitive ability and phenology (Dodd et al. 2005; Gould et al. 2006; Slotte et al. 2007; Gutierrez et al. 2008; Saloméet al. 2008). To understand and predict the complexity of climate change effects on species fundamental and realized niches and on phenology, it is essential to understand the underlying mechanisms. Current evidence points towards the circadian clock as a strong candidate driving the response of phenology to changes in environmental cues such as temperature, water and nutrients in different species (Roden et al. 2002; Dodd et al. 2005; Ramos et al. 2005; Gould et al. 2006; Harmon et al. 2006; Slotte et al. 2007; Yakir et al. 2007; Gutierrez et al. 2008; Saloméet al. 2008) and in modulating important aspects associated with the niche these species occupy. Further research should be undertaken to elucidate the role of the clock as a control on plant distributions in non-model species, and on how to incorporate this information into process-based models of responses to climate change.


The circadian clock acts a central orchestrator of plant activity at the molecular level and is involved in regulating key aspects associated with plant fitness. Although the molecular structure of the clock is relatively well understood, its ecological relevance has been largely overlooked so far. In this study, we establish the biological basis for considering whether circadian rhythmicity matters at higher scales of organization, relevant for ecological problems. We observed that the circadian clock modulates biosphere–atmosphere interactions at different scales and across different systems, and that it may play a large role in controlling future plant species distributions. While it is still too early to generalize on the involvement of the clock in ecological dynamics, including the effects of circadian regulation of plant biology, it is potentially relevant in our search for general mechanisms linking the diverse branches of ecology. Studies performed to date, across a variety of environments and species, suggest that ecologists cannot afford a further lack of research into this topic.


VR was supported by a EU Marie Curie Early Stage Training Host Fellowship (MEST-CT-2005-020526) and received assistance from MG Jones. Comments by J. Chase, J. Peñuelas, I.R. Urbieta, D.A. Ramírez, C. Morales and three anonymous reviewers on a previous version of this manuscript were very instrumental in improving its quality.