An emerging focus on plant ecological development

Long regarded as uninteresting ‘noise’, the influence of environmental conditions on development has emerged over the past decade as a central, shared focus of biological research. Ecological development investigates how organisms develop in their natural contexts – in the ‘real world’ (Gilbert, 2001), which is an inherently variable and sometimes unpredictable place. Eco-devo research addresses how naturally varying environmental factors affect developmental processes at a mechanistic level, how phenotypic expression varies in response to these factors (i.e. phenotypic plasticity), and how the phenotypic variation this creates in natural populations affects ecological and ultimately evolutionary processes (Gilbert & Bolker, 2003; Sultan, 2003a,b). This common focus thus integrates a stunningly broad range of research disciplines that share an empirical basis in plasticity, from developmental genetics, through functional, population and community ecology, to evolutionary biology. These disciplines are featured in this issue of New Phytologist, with research and review articles that contribute to exciting empirical, analytical and conceptual directions in plant ecological development. ‘Plant eco-devo integrates molecular/development approaches with ecology in increasingly sophisticated conceptual frameworks for understanding how plants live and evolve in the real world’ Plant biologists have led the way in this renascent focus on understanding the causes and consequences of development in its environmental context. Decades after A.D. Bradshaw's path-breaking 1965 article on plant phenotypic plasticity (which met with a puzzled reception by its contemporary neo-Darwinian audience) (Bradshaw, 1965), studies of plasticity and life-history strategies have compellingly shown that in all kinds of organisms environmentally mediated development can include the expression of adaptive (as well as maladaptive) phenotypes in response to varying conditions (reviewed by Schlichting, 1986; Stearns, 1989; Travis, 1994; Sultan, 2000; Pigliucci, 2001; DeWitt & Scheiner, 2003). Due in part to the continuous organogenesis that makes plant development extraordinarily flexible at all levels (Walbot, 1996), the environmental contingency of phenotypes in both controlled and natural settings has been most intensively studied in plants (Schlichting, 2002). Studies of plant plasticity have revealed not only immediate morphological and allocational responses to environmental factors, but complex environmental effects on ontogenetic trajectories (Diggle, 1994; Schlichting & Pigliucci, 1998; Weinig & Delph, 2001) and offspring traits (Donohue & Schmitt, 1998; Agrawal et al., 1999; Lundgren & Sultan, 2005). Although testing the functional and fitness consequences of environmentally mediated variation remains a difficult challenge, Schmitt and colleagues have shown that robust tests of adaptive plasticity are possible, by either manipulating environmental cue systems or using transgenic or mutant plants with disabled plastic responses to generate, experimentally, individuals which do and do not alter their phenotypes in contrasting conditions (reviewed by Schmitt et al., 1999, 2003; see also Galen et al., 2004; Huber et al., 2004). These and several other recent studies have built on the well-studied system of developmental responses to red : far-red light ratio mediated by the phytochrome gene family. To date, this system is the one plant case in which the entire ecological development pathway is reasonably well understood, from perception and transduction of the environmental cue to effects on several genes that regulate phenotypic expression (references in Callahan et al., 1997; Schlichting & Smith, 2002). Whereas the genetic basis of other aspects of plasticity remains largely unknown, quantitative trait locus (QTL) mapping is now being used as an initial step to investigate these systems (Pigliucci, 1996; Weinig & Schmitt, 2004). Researchers in diverse aspects of plant ecological development are building rapidly upon this foundation both empirically and conceptually. Along with a growing body of data on plant developmental responses to specific environmental variables are newly available tools for investigating the genetic bases and transduction pathways of those responses, as well as increasingly sophisticated analytical, comparative, and theoretical frameworks for understanding their ecological and evolutionary consequences. In this issue, articles by Callahan et al. (pp. 129–139), Cronk (pp. 39–48), and van Kleunen & Fischer (pp. 49–60) highlight the range of approaches currently available to study the genetic basis of plasticity at increasingly precise levels, from QTL mapping analysis and cDNA microarrays to candidate gene studies in model systems such as Arabidopsis and Populus. Stanton & Thiede (pp. 319–337) and Rowe & Speck (pp. 61–72), respectively, offer statistical and biomechanical insights to the fitness consequences of eco-devo variation at the population and macroevolutionary levels. Several articles present new data and/or innovatively refined conceptual frameworks for testing environmental effects on plant development in an adaptive, ecological context (de Kroon et al. pp. 73–82; Donohue, pp. 83–92; Galloway, pp. 93–99; Griffith & Sultan, pp. 141–147; Moriuchi & Winn, pp. 149–158; Wong & Ackerly, pp. 159–172). The role of plasticity in evolution is explored by de Jong (pp. 101–117) and Delph & Wolf (pp. 119–128), while Brock et al. (pp. 173–183) examine its relation to ecological invasiveness. The genetic basis of phenotypic response to environment is inherently an evolutionary as well as a mechanistic issue. Callahan et al. and van Kleunen & Fischer argue convincingly that information on specific loci involved in phenotypic response (‘plasticity genes’) is essential to understand both the pathway from gene expression to phenotype and the resulting selective dynamics on genetic variation for plasticity. van Kleunen & Fischer offer a thought-provoking critique of our presently limited knowledge of evolutionary constraints on adaptive plant plasticity. They discuss promising new tools to determine the genetic architecture of plasticity, its variation, and its potential costs and limits. They also point to timely research strategies to build our surprisingly limited knowledge of adaptive plasticity, including selection experiments on plastic traits (see Scheiner, 2002) and studies of environmental variation at different scales in relation to plasticity. Callahan et al. describe an innovative way to combine developmental and quanititive genetics methods to investigate the genetic basis of plasticity. By using recombinant inbred lines of Arabidopsis that are polymorphic at the FRIGIDA locus (which influences flowering response to vernalization regime), they detected an environment-specific plasticity cost. They also unexpectedly found evidence for maladaptive plasticity for delayed flowering in response to partial vernalization, compared with genotypes that constitutively flowered earlier. These data suggest the intriguing possibility that such plasticity may be selectively maintained as a result of pleiotropic positive effects of loci influencing environmental response, indicating greater than expected complexity to the issue of plasticity costs. Cronk highlights the potential value of Populus tricharpa as a model system for molecular eco-devo investigations, now that the complete genome sequence of this widespread, ecologically important tree is available. Because this species has ecologically diverse but caryoptypically similar congeners, this technical achievement may allow comparative work at the genome level to study how nucleotide variation influences ecologically important aspects of plant development. Studies of potentially adaptive plant responses to environment are beginning to extend beyond straightforward growth and allocational effects to more complex aspects of the phenotype such as architecture, functional anatomy, life history and even effects on offspring traits. Galloway and Rowe & Speck synthesize their recent studies to explore how these diverse aspects of ecological development can contribute to adaptation. Galloway asks whether maternally transmitted environmental effects can provide a means of fine-scale local adaptation in natural plant populations. Her ongoing work on Campanula americana, a herb native to patchy light habitats, demonstrates how the maternal plant's light environment influences offspring life-history schedule. Preliminary field data confirmed an adaptive hypothesis: for offspring produced by genetically similar parents grown in either canopy gap or understory sites, rates of fall germination were enhanced when the offspring's light environment matched that of the maternal parent. From a larger perspective, Rowe & Speck argue that both ecological and phylogenetic effects on development are key to understanding growth form diversity in plants. They consider how little-studied environmental effects such as mechanical perturbation can alter the heterochronic expression, architecture, and wood anatomy of self-supporting vs climbing plant growth forms through specific effects on the cambium. Rowe & Speck propose that this capacity for ontogenetic, biomechanic and hydraulic plasticity in itself may be one of the key innovations provided by the evolution of the bifacial vascular cambium, and discuss how phylogenetic constraints on these traits may influence potential plastic responses to these and other environmental factors. Stanton & Thiede provide a very different type of insight to the adaptive significance of ecological development: a critique of standard statistical methods for assessing the fitness of phenotypes produced in different environments. Because favorable conditions increase both mean reproductive output and its variance (especially in indeterminate organisms such as plants), researchers usually log-transform fitness data to meet homoscedasticity assumptions. Yet this standard method can distort fitness results and lead to incorrect evolutionary predictions. The authors’ clear demonstration of this neglected problem and their proposed alternatives constitute a useful step toward adjusting statistical techniques to match the particular goals and quirks of eco-devo studies (see also Scheiner et al., 2002; Stinchcombe et al., 2002). Understanding ecological development demands a way of framing the organism–environment relationship, with its complex feedbacks, so as to allow precise, testable hypotheses incorporating multiple levels of investigation. Evolutionary ecologists are continuing to incorporate new ideas into an emerging body of conceptual approaches to this challenging problem (references in Schlichting & Pigliucci, 1998; Lewontin, 2001; Nijhout, 2001; Schlichting & Smith, 2002; Sultan & Stearns, 2005). In this issue, several authors contribute novel and perhaps controversial perspectives to this ongoing synthesis. de Kroon and colleagues present a modular rather than whole-plant framework for understanding eco-devo response, critically reviewing the literature on plasticity at the level of functional modules such as individual leaves and roots. When heterogeneous conditions occur within the plant individual, the developmental integration of these modular responses can profoundly affect whole-plant resource uptake and growth. As a result, they argue, mechanisms of modular integration can be seen as a key aspect of ecological development. Donohue focuses instead on life-history plasticity as a fundamental aspect of eco-devo variation from which a distinctive formulation of the organism–environment relationship emerges. Through developmental and phenological plasticity, plants alter not only their own phenotypic expression but the environments they subsequently encounter. This ‘niche construction’ is most dramatic in the case of life-history traits such as reproductive timing, dispersal and germination, which directly influence the seasonal and competitive environment of offspring and consequently their selective regime. Donohue synthesizes new results from naturally evolved Arabidopsis ecotypes to show how niche-constructing traits can interact to both produce novel life histories in the field, and constrain adaptive evolutionary change in these and related traits. Moriuchi & Winn address a knotty interpretive question regarding the developmental basis of phenotypes expressed in contrasting environments: do plants in poor vs rich conditions express distinct developmental trajectories, or simply differences in developmental rate that are interpreted as distinct phenotypes when plants are measured at the same age? (See Coleman et al., 1994; McConnaughay & Coleman, 1999.) Moriuchi & Winn's data show that, in contrast to earlier work on annuals, perennial Viola plants grown at low resource levels were not simply developmentally delayed, but expressed distinct, adaptive ontogenetic patterns for traits such as storage allocation to rhizomes. In further contrast to annuals studied previously, these perennials maintained the ability to alter developmental trajectories beyond the first few weeks of growth in response to changing conditions. The interplay of ontogeny and plasticity in natural ecological contexts remains an important area for further study. Precise hypotheses about ecological development also require a clear focus on the environmental cues that initiate developmental responses in natural habitats. Griffith & Sultan distinguish perception of an environmental cue from subsequent developmental response as separate components of phenotypic plasticity, and present comparative data illustrating this decoupling in two closely related taxa. Because these components can have different genetic bases and evolutionary potential, they can differ among populations and species in ways that contribute to ecological differences. At present, little is known about either mechanisms of cue perception and transduction (Gilroy & Trewavas, 2001) or the correlations between specific cues and selective conditions that shape the evolution of adaptive response in natural populations (Tufto, 2000; Sultan, 2004). Wong & Ackerly approach the role of cue reliability in the selective maintenance of plasticity from a theoretical perspective. They present an optimality model testing variation in reproductive allocation in relation to predictable vs unpredictable season lengths. Their model and accompanying simulations confirmed that when environmental information poorly predicted selective conditions, plasticity for reproductive allocation was disadvantageous relative to a random bet-hedging strategy. Starting with Bradshaw (1965), it has been clear that ecological developmental variation in nature has profound implications for the process of adaptive evolution in plants. Two articles in this issue contribute, respectively, to empirical and theoretical knowledge of these broad-scale implications. Delph & Wolf consider how life-history plasticity may either promote or constrain the evolution of genetically dimorphic plant breeding systems. They review existing data to conclude that sex expression (relative seed or pollen output) in subdioecious species is often resource-dependent, a fact overlooked in models of breeding system evolution. Their overview of the literature shows that selection may act differently when the plastic morph is predominantly male rather than predominantly female: gender plasticity can prevent the evolution of complete dioecy from gynodioecy by reducing the cost of producing seeds, while plasticity can act to maintain gender dimorphism in androdioecious species that otherwise might evolve toward hermaphroditism. de Jong's provocative theoretical paper argues against the notion that plasticity may play a macroevolutionary role as a mechanism of adaptive species diversification, an idea which has lately sparked a great deal of interest among evolutionary biologists. According to this conceptual model (termed the ‘Baldwin effect’ by G.G. Simpson in 1953), the adaptive flexibility of genotypes produces phenotypes able to invade new habitats, so that new ecotypes and ultimately species result if these phenotypes become genetically fixed under subsequent selection in the new environment (West-Eberhard, 2003; Pigliucci & Murren, 2003; Schlichting, 2004). de Jong extends existing quantitative genetic approaches to propose instead that observed patterns of ecotypic variation can be explained by conventional, neo-Darwinian selective mechanisms operating on adaptive plasticity in the context of limited migration. The possibility of a unique role for plasticity in evolutionary diversification is likely to remain an area of controversy, particularly because empirical data on this intriguing issue are scarce. Environmentally mediated development also holds important ecological implications. Adaptive plasticity can enlarge a species’ environmental tolerances and hence contribute to broad, generalist distributions. Such adaptive plasticity can also allow the spread of an invasive into new habitats without a selective lag time. Brock et al. tested the relation of shade plasticity to invasiveness in two sympatric Taraxacum species. Their study confirms the key point that environmentally induced variation per se is not likely to explain invasiveness; rather, invasiveness may involve adaptive plasticity for traits that enhance fitness in relevant environments, along with constancy for traits such as high dispersal potential. These diverse avenues of research are part of an emerging synthesis in plant ecological development: the integration of molecular/development approaches with ecology in increasingly sophisticated conceptual frameworks for understanding how plants live and evolve in the real world. This synthetic approach accords with the growing recognition that, in general, gene interaction and regulation by both internal and external factors is of primary importance for phenotypic determination (Carroll, 2000; Nijhout, 2001). However, this synthesis is very much a work in progress. To achieve this integrated understanding of plant ecological development from concordant mechanistic, ecological and evolutionary points of view requires further research on the rich array of subjects addressed in this issue, as well as studies of higher-order effects on species’ interactions (Agrawal, 2001). It will also require continuing efforts to refine our empirical, statistical, and modelling techniques for assessing and interpreting environmentally mediated variation. In this brief overview I have not attempted to provide comprehensive references to this vast multidisciplinary literature. Please see individual papers in this issue and the listed references for more complete citations.