Mind the gap: the emerging synthesis of plant ‘eco-devo’


New directions in plant ecological development: the 14th New Phytologist Symposium, London, UK, January 2006

The past 10 years have witnessed great advances in developmental genetics and molecular techniques, as well as new conceptual insights into organisms as dynamic systems interacting with their environments. While researchers in the field of plant phenotypic plasticity have largely aimed to characterize phenotypic variation in response to environment, those in the field of ecological development, or ‘eco-devo’, aim for a more mechanistic understanding of this response process at all levels: the perception of environmental signals; their transduction by the organism at molecular, cellular, and hormonal levels; the phenotypic expression that results from these multiple, interacting regulatory processes; and the consequences of those phenotypic outcomes for functioning individuals, ecological communities, and evolving plant populations. The broad goal of eco-devo, then, is to close the gap between mechanisms and ecology, and provide a genuinely integrated view of plant development in its real environmental context. The 14th New Phytologist Symposium (http://www.newphytologist.org/eco-devo), held recently at the Royal Society in London, addressed this exciting goal under the theme of ‘New directions in plant eological development’ (see also the recent feature on this topic; Sultan, 2005). The symposium was attended by participants from 11 countries and a diversity of research disciplines, at all levels from graduate students and postdocs to eminent senior researchers, notably Professors Anthony D. Bradshaw and Harry Smith. Here we highlight the presentations of the speakers under two broad themes that emerged through the presentations and ensuing discussions.

Anthony Bradshaw opened the symposium with a brief and stimulating plenary, providing an historical perspective on the study of plasticity and raising some of the issues that led to his pioneering 1965 paper (Bradshaw, 1965; see accompanying article in this issue, Bradshaw, pp. 644–648). He argued the fundamental point that to understand organisms ‘knowing a character's mean isn't good enough, we must know how it is expressed in actual environments.’ In this light, Professor Bradshaw reminded the participants of the continuing importance of understanding the mechanisms and significance of phenotypic stability as well as plasticity, and the relationship between the two. His remarks set the stage for the ensuing presentations, as the dissection of mechanisms underlying plant development is rapidly shedding new light on these questions.

‘… the complementary notion of “ecological annotation”: how much of the observed phenotypic variation among life history stages, individuals, populations, or species can be attributed to allelic or regulatory variation in a particular gene’

Genes and gene function in ecological context

Developmental genetics examines the regulatory pathways that transform environmental or ontogenetic signals into cascades of gene transcription, translation, and phenotypic outcomes from the molecular to morphological levels. In an ecological context, there has been great progress in understanding the genetic architecture underlying ‘model traits’ such as photomorphogenesis (Chen et al., 2004) and flowering time (Boss et al., 2004). Eco-devo research critically extends our understanding of these mechanistic models by studying them in the context of naturally occurring genetic and environmental variation. For example, J. Schmitt (Brown University, Providence, RI, USA) and colleagues have shown that geographically widespread natural populations of Arabidopsis thaliana show genetic variation for the FRIGIDA locus and its repressor flowering locus c (FLC), key flowering genes that interact epistatically (Caicedo et al., 2004). Under common environments, these genetic variants exhibit different flowering times mirroring their latitude of origin (Stinchcombe et al., 2004). Allelic variation at these loci is maintained within populations in part through antagonistic fitness effects in winter vs spring annual cohorts. A central theme of the symposium was the question ‘How can detailed knowledge of developmental genetics provide essential insights into ecological phenomena?’ One answer that emerged is that the genetic architecture of a trait imposes constraints (at least in the short term) on phenotypic responses to the environment, analogous to the constraints imposed by antagonistic genetic correlations on microevolutionary responses to selection. In the flowering time example, understanding variation in interacting signaling pathways may inform predictions about the responses of different populations to climate change. Short-term responses will be shaped by current developmental mechanisms. The variation among populations in these mechanisms demonstrates that they may be rapidly adjusted by evolutionary change, so the time-scale of these constraints will be linked to generation times and the strength of selection on the resulting traits.

Knowledge of genetic mechanisms may also be essential to understand the degree of correlation or independence of plant responses to different environmental factors. There are several signal perception, transduction, and gene regulatory pathways (such as ethylene and abscisic acid (ABA) concentrations and phytochrome activation) that appear repeatedly in eco-devo studies of different species, traits, and environmental factors. Given these shared pathways, one intriguing research avenue is the degree to which plants have evolved to respond appropriately to distinct environmental signals. L. Voesenek (Utrecht University, Utrecht, the Netherlands) discussed his elegant studies of the mechanisms underlying stem elongation in response to flooding in the aquatic species Rumex palustris (Voesenek et al., 2004). Ethylene plays a key role, as endogenous production in plant cells leads to high ethylene concentrations in cells of submerged plants, as a result of the reduction in diffusion coefficients underwater. The build-up of ethylene inhibits ABA synthesis, and reduced ABA leads to higher gibberellic acid (GA) concentrations, promoting stem elongation (of course the full regulatory network is more complicated than this). Comparative studies of the closely related terrestrial species Rumex acetosa illuminate critical control points in the pathway that distinguish species with and without this submergence response (Benschop et al., 2005). Surprisingly, components of this same pathway are also involved in the elongation response to shade, although in this case phytochrome is employed in the mechanism of environmental perception. Voesenek asked whether a few ‘master switch genes’ are involved in diverse environmental responses, and if knowledge of these mechanisms would allow us to better understand and predict cross-responses and interactions among different signals.

The multifaceted role of phytochromes in plant development (Sullivan & Deng, 2003) continues to offer novel challenges to our understanding of signal transduction in plant responses to environment. Harry Smith (University of Nottingham, Nottingham, UK) synthesized his lifelong interest in the remarkable role of phytochromes as a ‘precise, graded, and unambiguous signal’ of neighbor density. Indeed, the evolution of phytochromes B, D and E, which are critical to shade avoidance responses such as shoot elongation, may have been a key step in the evolution of angiosperms and their dominance in highly productive vegetation (see Mathews, 2005). Recent studies demonstrate that woody plants differing in shade tolerance exhibit variable, phytochrome-mediated growth responses to the red:far-red ratio (R:FR), similar to those of herbaceous taxa (Gilbert et al., 2001). Microarray data suggest that the elongation response to far-red may involve a rapid transcription event inducing expression of numerous genes involved in growth within 10 min of the plant's perception of a far-red signal. These precise events as well as the localization of phytochromes on the plant body, the functional interactions among the different phytochromes, and the extent to which plants in certain habitats may grow toward light rather than away from shade remain to be determined.

C. Galen (University of Missouri, Columbia, MI, USA) offered a novel perspective on the role of blue-light perception and negative phototropism in root growth, in relation to the attenuation of light through the very top layers of the soil. Experiments with phototropin-disabled mutants (Galen et al., 2004) show that loss of function of these blue-light photoreceptors can lead to less downward-directed root growth, resulting in lower seedling establishment and total biomass in plants grown in dry soil. Concomitant effects on water use efficiency, which is higher in plants unable to sense blue light (an inducer of stomatal opening), raise the important question of how plants integrate changes in multiple functional traits in response to one or many environmental signals. Galen's research also exemplifies how a characteristic that may be perceived as ‘normal development’ (the downward growth of roots) may be dependent on active mechanisms of signal perception and transduction.

The theory of niche construction (Odling-Smee et al., 2003) emphasizes ecological and evolutionary feedbacks between organisms and their environments. K. Donohue (Harvard University, Cambridge, MA, USA) has applied this theory to the study of gene action and plasticity in plants (Donohue, 2005). The action of a gene early in the life history of a plant alters the phenotype, and thus effectively alters the environment experienced by that plant (or by parts of the plant) later in life. Thus, plant responses to the environment modify not only the phenotype, but also the future environment. These feedbacks can have important consequences for the mechanisms, impacts, and evolution of plasticity. For example, experiments with Arabidopsis ecotypes reveal that the timing of germination in spring vs fall is influenced in part by the action of phytochromes, via effects on light perception and temperature responsiveness. Germination timing will dramatically alter the growth environment for the rest of the life cycle, and this can be seen as an extended phenotypic effect of the genes controlling germination. These indirect pathways provide the mechanism to generate epistatic interactions among genes operating in different plant organs or at different times during the life history of the plant, and are thus critical to understanding the integrated effects of gene action on lifetime performance and fitness.

Ultimately, the study of plants in their natural context is essential to obtain a meaningful answer to the question: what is the function of a gene? The traditional approach to functional annotation of the genome is based on knockout or overexpression of genes, preferably in an isogenic background, to determine the phenotypic effect of each gene. However, the dependence of the observed function on the particular background and laboratory environment is not always recognized. C. Weinig (University of Minnesota, St. Paul, MN, USA) introduced the complementary notion of ‘ecological annotation’: how much of the observed phenotypic variation among life-history stages, individuals, populations, or species can be attributed to allelic or regulatory variation in a particular gene (Weinig & Schmitt, 2004). In studies of Arabidopsis grown in the field, she has shown that different quantitative trait loci (QTLs) contribute to variation in growth and fitness in the presence and absence of competition. The contrast of functional vs ecological annotation presents an interesting analogy to the methods of prospective and retrospective sensitivity analysis in population biology (Caswell, 2001). These two scenarios contrast the sensitivity analysis of the engineer (‘how much will the system change in response to each component?’) with the decomposition of variance of ecology and evolution (‘how much of the observed variance is attributable to variation in each component?’).

A deeper understanding of phenotypic plasticity, and stability, will emerge from the synthesis of developmental, physiological and evolutionary approaches. F. Valladares (CCMA-CSIC, Madrid, Spain) discussed a broad range of mechanistic and adaptive factors that may contribute to species diversity in phenotypic plasticity. These approaches focus, respectively, on costs and constraints that may limit plasticity, such as energetic costs of signal perception and growth responses, vs environmental heterogeneity and ‘adversity’ as selective forces acting on plasticity (Valladares et al., 2002; van Kleunen & Fischer, 2005). Understanding how plants integrate and respond to contrasting environmental signals (Niinemets & Valladares, 2004) and novel biotic environments (Hobbs et al., 2006) is an immediate priority in the context of global environmental change.

Eco-devo and community-level interactions

Scaling up from molecules to organs, individuals, communities and ecosystems, the direct effects of gene action tend to become more diffuse and increasingly difficult to detect. Interactions among species, especially plants and their symbionts or herbivores, present one ecological arena in which eco-devo is providing fascinating insights. Below-ground, the infection of roots by nitrogen-fixing symbiotic bacteria involves a highly co-ordinated sequence of morphogenetic events, with reciprocal triggering of gene expression between the two partners leading to the production of nodules and their occupation by the bacterium (Miklashevichs et al., 2001). The genes involved in nodulation were apparently recruited, in evolutionary terms, from the ancestral mycorrhizal association, where the same genes participate in the process of fungal colonization of roots (Stracke et al., 2002). A. Hodge (University of York, York, UK) presented the results of her meticulous growth studies, demonstrating that reciprocal interactions between roots and arbuscular mycorrhizas influence the well-known proliferation response of roots in nutrient-rich soil patches. In fact, mycorrhizas also proliferate, enhancing decomposition and nitrogen mineralization in organic soil patches (Hodge et al., 2001). Importantly, roots that are colonized by mycorrhizas exhibit enhanced proliferation in response to nutrient-rich patches when they are grown in competition, and this leads to pre-emptive uptake of mobile ions, including nitrogen (Hodge, 2003). Much remains to be learned about the fascinating dynamics of these widespread plant–mycorrhiza interactions, including whether there are direct interactions among genetic regulatory networks and whether each partner contributes to proliferation in the other.

Moving above-ground, one of the key aspects of the plant environment (often overlooked by both developmental geneticists and functional ecologists) is the activity and phenotypic impact of insects and other herbivores. J. Schultz (Pennsylvania State University, University Park, PA, USA) discussed the transcriptional and regulatory overlap between self-regulated plant development and insect-induced responses. In plant galls induced by insects such as grape phylloxera (Daktulosphaira vitifoliae), the insect manipulates its host by altering the expression of floral induction loci such as LEAFY and APETALA1 (which are required for MADS box gene activity), possibly by direct injection of uric acid (which is known to have up-regulatory cytokinin-like effects). These same pathways are critically important for normal development of flower and fruit development, and the selective constraints imposed in that context create an opportunity for insects to co-opt the developmental program of the plant to their own advantage. A. Kessler (Cornell University, Ithaca, NY, USA) presented a very different aspect of the complex plastic responses of plants to insect attack, by showing how specific transcriptional changes induced by caterpillar attack create both direct and indirect plant defenses in Nicotiana. Interestingly, plants can maintain fitness despite such attack as a result of wound-induced up-regulation of genes important to primary metabolism. ‘Taking molecular tools out into nature’ can be particularly revealing. Kessler's experiments with a lipoxygenase (lox-3) Nicotiana knockout mutant (causing reduced induction of defensive nicotines, proteinase inhibitors, and volatile organics) explanted into a natural garden community showed that the loss of function of this gene actually attracted a new species of herbivore to the plants, a vivid demonstration of the importance of induced individual responses to the larger ecological community (Kessler et al., 2004).

A. Novoplansky (Ben-Gurion University of the Negev, Midreshet Ben-Gurion, Israel) gave a provocative presentation examining the ability of individual plants to distinguish self from nonself tissues in competitive arrays. He argued that environmental cues and physiological signals may provide plants with information about whether nearby leaves or roots, which are competing for limited resources, are part of the same plant (i.e. a physiologically integrated individual, of the same or different genotype). Roots that are spaced equidistant from self and nonself neighbors exhibit enhanced growth towards the nonself neighbors, presumably enhancing resource uptake relative to competitors (Falik et al., 2003; Gruntman & Novoplansky, 2004). The mechanisms of these responses are essentially unknown, although the available data suggest that they are dependent primarily on physiological integration within the plant, rather than biochemical mechanisms of allorecognition (as in many animals systems). This topic will certainly receive continued interest and scrutiny.

‘Fifty years ago, Bradshaw recalled, at a time when biologists were captivated by the New Synthesis approach, phenotypic responses to the environment were viewed as “an annoyance, a confusion that made it difficult to know what a genotype was … The froth of evolution”.’

Conclusions and prospects

Closing the gap between the study of mechanism and the study of ecological and evolutionary diversity is one of the great challenges of 21st century biology: indeed the interface of these domains can be seen as the major ‘new frontier’ in biology (Kafatos & Eisner, 2004). As plant biologists facing this challenge, we seek to know what developmental genetics brings to ecology, and vice versa. The use of ‘molecular tools in nature’ is growing steadily and will continue to bring fascinating new possibilities to the field. The performance of single gene mutants (gene alleles, knockout, overexpression, etc.) and recombinant inbred lines under field conditions provides insight into the ecological and evolutionary significance of specific genetic changes; luminescent reporter constructs linked to genes with dynamic expression across time and space allow us to ‘see’ gene expression in field settings, as in Weinig's studies of circadian transcription cycles; sampling of RNA constructs can be used as a physiological markers of plant responses, as in Voesenek's studies of ethylene production patterns; microarray analysis is just scratching the surface to reveal how thousands of genes may be involved in plant responses to abiotic and biotic factors. Ecology also has much to offer developmental genetics. More than anything, as Weinig cogently argued, the study of gene function must be carried out in nature as well in as the laboratory: the ‘function’ of a gene in natural environments varies depending on the environment and the genetic makeup of the population (Weinig & Schmitt, 2004). Genetic variation in natural populations also provides an important source of material for experimental study by developmental geneticists. Naturally occurring alleles may point the way towards functionally significant nucleotide substitutions, insertion–deletion (indel) events, or gene rearrangements that might not be anticipated by analysis of sequences or protein structures. More broadly, comparative studies incorporating interspecific diversity are a critical component of eco-devo research (e.g. Benschop et al., 2005; Griffith & Sultan, 2005), to reveal aspects of morphological and functional response variation as well as important phenomena that may be absent in our favored model taxa. Ecologists should take a proactive role in promoting the development of new model organisms that exhibit an ecologically interesting diversity of form and function, paralleling the efforts by phylogeneticists to select taxa spanning the tree of life.

Fifty years ago, Bradshaw recalled, at a time when biologists were captivated by the New Synthesis approach, phenotypic responses to the environment were viewed as ‘an annoyance, a confusion that made it difficult to know what a genotype was … The froth of evolution’. Following his pivotal 1965 paper, plasticity became a central focus of ecological and evolutionary research, bringing enormous insights into the realized phenotypic variation that shapes ecological interactions and selective change (Sultan, 2003). In many ways, we hope this symposium marks another turning point, as a newer mechanistic approach emerges linking advances in developmental genetics to the study of environmental responses. This newer approach views the genotype as a dynamic, integrated developmental system with flexible expression, and seeks to understand that system all the way from its molecular basis to its ecological and evolutionary roles. This ambitious undertaking will require collaborative, synthetic research to identify connections from molecular to higher levels of organization and among interacting regulatory pathways and traits.