Darwinism renewed: contemporary studies of plant adaptation
One hundred fifty years after On the origin of species was first published (Darwin, 1859), biologists continue to share Darwin's fascination with the adaptive traits of organisms, and to theoretically and empirically expand his initial formulation of natural selection as the process that shapes those traits. In this special issue of the journal, New Phytologist recognizes this continuing legacy by bringing together current ideas and findings about plant adaptation from an accomplished, multi-national group of researchers studying a wide array of plant systems. Many of these studies make use of the formidable set of research tools now available to examine the molecular and genomic bases of adaptive traits and their selective dynamics, and these tools have certainly provided critical insights. But the most exciting aspect of contemporary work on adaptation is not these new techniques per se, but rather the way researchers are combining diverse tools in robust, richly informative interdisciplinary approaches. Genomic data from model organisms are being used in new ways to inform studies of naturally evolved systems, and a rigorous phylogenetic context has become standard across sub-disciplines to refine population, species, and higher-level comparisons.
The studies presented here address questions at all levels from the molecular to the macroevolutionary, drawing on information ranging from transgenic functional assays, transcriptomics, and quantitative genetics trait matrices to GIS-based niche modeling, field explant studies of recombinant inbred lines, and distribution data from herbarium sheets. This new work makes clear how these sophisticated interdisciplinary approaches are illuminating some of the most fundamental and long-standing questions about plant adaptation and the process of selective change. It also reflects an increasing awareness that the study of adaptation has important bearing on biodiversity conservation, invasion biology, and potential constraints on adaptive evolution in the face of rapidly changing environments. Here, a brief overview of the feature points to some of the key findings with respect to several areas of shared focus: co-evolutionary interactions; floral and mating system evolution; geographic patterns of adaptive evolution; genetic architecture of adaptation; and evolution of functional traits.
The interactions of plants with their pathogens, parasites and pollinators can generate reciprocal selection pressures and consequently a distinctive co-evolutionary process. L. G. Barrett et al. (this issue, pp. 513–529) analyze this process for the complex spectrum of associations between plants and their microbial pathogens. In contrast to the specific, highly virulent pairwise interactions that have been emphasized to date in co-evolutionary models, they argue that host-pathogen associations form a continuum from specialists to generalists, with widely varying fitness effects (and therefore selective outcomes) depending on environmental and genetic factors that jointly influence the expression of both microbial virulence and host resistance genes. These factors can interact at several levels to affect selection pressures, for instance in the case of co-infection by multiple pathogens that interact within the host to modify its defence gene expression. Godsoe et al. (pp. 589–599) focus on a very different aspect of plant co-evolutionary dynamics, a well-studied pollination mutualism. Using GIS-based niche modeling and well-resolved phylogenetic data, they determine that the specialized biotic interaction between Yucca brevifolia and its moth pollinators, rather than climatic variables, have led to evolutionary divergence in this system.
Floral and mating system evolution
Floral features such as pollination syndromes and organ placement provide clear examples of adaptation. Harder & Johnson (pp. 530–545) comprehensively review evidence as to whether floral and inflorescence traits fit a Darwinian model of gradual, consistent selective change. Although data from manipulation experiments generally confirm that these ‘beautiful contrivances’ are indeed adaptations that evolved for pollen transfer, the results of phenotypic selection analyses indicate that directional selection on these traits may be relatively weak and inconsistent in natural populations. Instead, adaptive floral features may be shaped largely during episodes of strong selection rather than by continual gradual selection as Darwin envisioned. This meta-analysis reveals limits to phenotypic selection analysis as a definitive test for adaptive function. Armbruster et al. (pp. 600–617) pose a complementary question about the evolution of floral adaptations: does floral diversification at the macroevolutionary level reflect adaptation to pollinators? They determine an optimal-fitness ‘adaptive surface’ relating pollination performance to anther and stigma position, and test against this prediction the realized distribution of floral traits in three phylogenetically diverse genera. Their analysis reveals clade-specific departures from the optimum reflecting both lack of floral integration and conflicting selection pressures for other outcrossing features. Fenster et al. (pp. 502–506) draw our attention to vertical versus horizontal floral orientation as a significant influence on pollinator effectiveness that has been largely ignored for the past century. Their initial dataset shows that hummingbird approach behavior and floral contact vary in response to differently oriented artificial flowers, a result that argues for future pollination studies to include this aspect of floral adaptation.
Gender expression and the evolution of plant mating systems have long been a focus of theoretical and empirical interest. S. C. H. Barrett et al. (pp. 546–556) present a well-developed case study on the evolution of selfing based on phylogeographic analysis of multilocus nuclear DNA sequences (SNPs) from populations sampled across the Neotropical range of the sexually polymorphic species Eichornia paniculata. Their data reveal multiple independent transitions from outcrossing to selfing in populations of this colonizing species. Preliminary results also suggest that both genetic factors and environmental stress contribute to developmental instability in the early stages of selfing, which may facilitate this evolutionary transition. Randle et al. (pp. 618–629) investigate the ecological implications of self-fertilization: does selfing promote colonization and hence range expansion, or does the lower genetic diversity of selfing species instead restrict range size? They test the relation of selfing ability to size of realized geographic range in Collinsia, combining a phylogenetically controlled species-pair comparison with precise measures of floral form and self-fertilization activity. This elegant study establishes empirically that species with the highest proficiency for autonomous selfing also have significantly larger ranges, linking an individual reproductive trait with ecological distribution including, as predicted by Herbert Baker, the spread of weedy and invasive taxa. Two additional papers analyze the evolution of mating systems. In one, Mazer et al. (pp. 630–648) demonstrate that pollen:ovule ratios are temporally more constant in selfers than in outcrossers, consistent with the expectation that the optimal P:O ratio varies more temporally in outcrossers. Van Etten & Chang (pp. 649–660) test whether the Sex-Differential Plasticity hypothesis – which posits that plasticity in hermapthrodites allows them to reduce seed production in harsh environments, allowing the invasion of females – can explain variation among populations in whether females are present in addition to hermaphrodites in Geranium maculatum.
Geographic patterns of adaptive evolution
Patterns of geographic spread and diversification reflect the interplay of gene flow with selective and neutral evolutionary forces across ecological landscapes. Dispersal traits are of central importance to this process, but other aspects of plant phenotypes may play a key role as well. Levin (pp. 661–666) proposes a previously unrecognized connection between individual plasticity and adaptive divergence of populations. (The paper is graciously dedicated to the late Fakhri A. Bazzaz, whose work deeply enhanced our understanding of plasticity and its ecological implications.) Levin reviews the evolutionary ecology literature to establish that individuals encountering novel conditions often displace the timing of flowering. He argues that, when a population colonizes a new habitat, such an environmentally-induced phenological shift will lead to temporally assortative mating that effectively isolates the colonizing population from its source. As a result, phenological plasticity will facilitate local adaptation to the new habitat that would otherwise be impeded by gene flow from the source population. Along with intriguing implications for ecological range expansion, this paper adds a new dimension to contemporary ideas about the possible role of individual plasticity in evolutionary diversification.
Questions of dispersal and range also hold immediate implications for extinction risks as natural habitats are increasingly disrupted. For instance, effective dispersal can contribute to species persistence in fragmented landscapes by allowing re-colonization of habitat patches and outcrossing. However, if habitat fragmentation in itself imposes selection for reduced dispersal ability (as occurs in island populations), this evolutionary feedback will worsen its negative impact on species’ distributions and persistence. Riba et al. (pp. 667–677) find a negative correlation between the degree of landscape fragmentation and achene dispersal ability at both local and regional spatial scales in the wind-dispersed European herb Mycelis muralis. Together with a common garden experiment confirming a partial genetic basis to this trait, these results indicate that fragmentation may have negative evolutionary as well as ecological consequences.
Geographic patterns of variation can also provide insights into the spread of invasive species, including the possible role of selective adaptation to the introduced range. Keller et al. (pp. 678–690) document differentiation patterns in two weedy European Silene species that are rapidly spreading in North America. By characterizing patterns of variation for neutral genetic (AFLP) markers, the authors statistically account for random drift and colonization history to test for the presence of adaptive clinal patterns. In addition to selectively neutral genetic structure, the range expansion of introduced Silene spp. also reflects adaptive evolution in response to local environmental gradients. Murren et al. (pp. 691–701) examine selective response in non-native populations of Mimulus guttatus, finding evidence of selection for larger plant size compared with native populations. Analysis of transects across hybrid zones allows examination of the forces of divergent selection that generates geographic variation and maintains species differences despite gene flow. Using such an approach, Brennan et al. (pp. 702–717) report that both intrinsic selection against hybrids and environmentally imposed divergent selection on several traits actively maintain the integrity of two hybridizing species of Senecio on Mt. Etna.
Genetic architecture of adaptation
New molecular and analytical tools, and new ways of combining them, have made possible a far more empirically rich understanding of the genetic basis of adaptive evolution. At the same time, a contemporary view of genomes as highly interconnected regulatory systems raises critical questions about the genetic architecture of adaptive traits: what kinds of genetic changes underlie novel adaptive phenotypes, and do these same changes arise repeatedly under similar selective pressures? What potential constraints on adaptive evolution are posed by gene interactions and genetic correlations among traits? One key insight to emerge from the recent explosion of genome and transcriptome data is the pivotal role of gene duplication. Flagel & Wendel (pp. 557–564) argue that this is a primary source of evolutionary innovation in plants, whose highly duplicated genomes reflect past and/or recent polyploid events. They describe the diverse array of gene duplication mechanisms known in plants, and explain the distinct ways they can contribute to novel phenotypes. For instance, transposon-mediated duplications can insert a gene into a new regulatory context that alters its expression, while allopolyploidy allows divergent regulatory systems to instantly become ‘co-resident’ genomes, resulting in higher-order gene interaction networks that can produce major phenotypic shifts. More generally, the genetic redundancy created by these various mechanisms can lead to either the evolution of new adaptive functions or to ‘sub-functionalization’ (division of labor among gene copies), which comprises an evolutionary solution to antagonistic pleiotropy.
The evolution of adaptive traits is empirically explored in several strong case studies. Di Stilio et al. (pp. 718–728) explore the evolutionary-developmental basis of floral phenotype in three species of Thalictrum with contrasting pollination syndromes, focusing on a transcription factor in the MYB family that is a candidate gene for floral epidermal features related to petal production. Their study illuminates the active evolution of gene lineages, their developmental impact at the cellular level, and the relationship of gene expression to differently adapted pollination syndromes.
Studies of floral color, defense chemicals, and stress tolerance show that diverse genetic mechanisms can underlie repeated evolution of the same adaptive transition. Cooley & Willis (pp. 729–739) show that red color has evolved repeatedly in a monophyletic group of Mimulus species via different, unique combinations of Mendelian and polygenic factors that influence petal anthocyanins. In another study of adaptive evolutionary convergence, Neiman et al. (pp. 740–750) examine selection on protease inhibitor loci involved in highly specific induced–defense interactions with herbivores. They document surprisingly variable patterns of nucleotide diversity and inferred selective histories within and between Populus species for proteins with putatively similar ecological function. Streisfeld & Rausher (pp. 751–763) examine three independent transitions from blue to red flower color in Ipomoea. Although this change could in theory result from a number of possible types of mutation along the well-characterized anthocyanin pathway, their results implicate the same gene in all three cases, and furthermore suggest that regulatory mutations play the primary role in these repeated evolutionary transitions. Dassanayake et al. (pp. 764–775) explore the genetic basis of physiological adaptation, presenting the first genomic data for two phylogenetically disparate mangrove species, a system of exceptional ecological interest. They characterize transcriptomes obtained by pyrosequencing, functionally annotating them using Arabidopsis and Populus genome data. The results show remarkably similar transcriptome profiles in the mangroves, indicating adaptive convergence at the gene expression level. Lowry et al. (pp. 776–788) examine the evolution of salt tolerance at the population level. In a comparison of coastal and inland Mimulus ecotypes, they identify quantitative trait loci (QTL) involved in salt-spray tolerance and fitness in a coastal site. Interestingly, they found no negative consequences of the alternative QTL's across habitats, suggesting that local adaptation may involve distinct sets of loci that are functionally neutral in other environments.
Kover et al. (pp. 816–825) explicitly focus on the genetic architecture of local adaptation: do genetic changes in response to selection in one environment entail negative fitness effects in other conditions? An experimental evolution approach suggests a complex answer. The authors artificially selected for early flowering in Arabidopsis in two simulated seasonal regimes, producing distinctly adapted lines. They find that in this system, the genetic basis of life-history adaptation includes both alleles with positive effects in both environments, and environment-specific allelic effects that have some negative impact in the alternative treatment. This underlying complexity may act to maintain genetic diversity for fitness-related traits. Muir & Moyle (pp. 789–802) examine a second aspect of genetic architecture that can affect adaptive evolution: the relative phenotypic contributions of additive and epistatic interactions between loci (in this case, target chromosomal regions tested in controlled pairwise combinations in multiple nearly-isogenic lines of Solanum). In this system, not only are epistatic effects a major component of total genetic impact on functional traits, most of these non-additive effects are antagonistic. For instance, chromosomal regions that individually decrease water-use efficiency (WUE) restore drought resistance when they occur in combination. These results point to the important conclusion that evaluating the main effect of individual loci may underestimate the loci that influence a complex adaptive trait, and consequently the genetic constraints on adaptive evolution.
Scoville et al. (pp. 803–815) focus on the pleiotropic impact of allelic variation as a potential evolutionary constraint. Their study integrates QTL mapping and G-matrix estimation to quantify evolutionary potential of a suite of floral and life-history traits in a natural population of Mimulus guttatus. The results of this innovative work exemplify how allelic changes at even one locus can realign the pattern of genetic covariances for fitness-related traits, altering both constraints to adaptation and levels of standing genetic variation. Galloway et al. (pp. 826–838) examine another critical aspect of potential constraints on adaptive evolution: maternal effects. Because maternal and offspring traits are expressed at different times, idiosyncratic trajectories of selective change can occur for traits influenced by these inter-generational effects. Building on an impressive body of work on environmental maternal effects in the forest herb Campanulastrum americanum, Galloway and colleagues identify strong effects of maternal genotype on the potential for selective response in offspring traits.
Evolution of functional traits
Papers in this area include both creative interdisciplinary work on the evolutionary history of functional innovations, and studies of the process of selection on functional traits. To begin the section, a provocative review by Sadras & Denison (pp. 565–574) re-evaluates the traditional view among physiologists and crop scientists that growing plant organs compete with one another as alternative ‘sinks’ for finite resources. They instead propose to consider resource allocation mechanisms in plants within an explicitly evolutionary framework. Such a framework has two components: first, the recognition that any conflict for resources within a genetic individual is selectively constrained by its effect on the fitness of the entire organism; and second, a view of genetically distinct plant structures (such as outcrossed progeny developing on a maternal plant) in the context of selective models for parent-offspring fitness trade-offs and sibling rivalry.
Brodribb and colleagues (pp. 839–847) consider a critical macroevolutionary aspect of ecophysiological adaptation. Their comparative study assesses stomatal control sensitivity to a broad range of CO2 levels in taxa representing major land plant lineages. The results are striking: in contrast to the sampled ferns, lycopods and gymnosperms, angiosperms have evolved the unique ability to close in response to high CO2 levels, a response that maximizes water-use efficiency (but reduces carbon fixation) under such conditions. In their ‘phylogenetic ecology’ study, Agrawal et al. (pp. 848–867) examine the evolutionary gain and loss of leaf surface traits important to both ecophysiology and herbivore resistance in the monophyletic, ecologically diverse genus Asclepias. Gibson & Waller (pp. 575–587) follow a passionate interest of Darwin's to evaluate the selective pressures that likely promoted the transition from sticky traps to highly modified snap-traps in carnivorous plants, a remarkable functional innovation that evolved just once.
At the population level, Donovan et al. (pp. 868–879) evaluate evidence for resource-based natural selection on ecophysiological traits in two distinct dune habitats in desert Helianthus taxa. Although they find evidence for direct selection on several functional leaf traits, both the strength and the direction of selection are highly variable, creating changing temporal and spatial patterns rather than consistent, habitat-specific selective pressures. McGoey & Stinchcombe (pp. 880–891) show empirically that selection on adaptive shade avoidance plasticity in Impatiens capensis depends on whether competitors are conspecific or heterospecific, refining our understanding of this important aspect of selection in natural communities. Two papers examine the adaptive relevance of plant morphological structures. Mao & Huang (pp. 892–899) demonstrate that in plants in which pollen is protected from rain by floral structures, pollen is less resistant to water damage than pollen from plants in which pollen is not protected. Wise (pp. 900–907) describes a growth-form polymorphism in Solidago altissima in which some individuals, instead of producing erect flowering stems, produce stems that recurve toward the ground. He shows that this alternate growth form reduces attack by apex-galling herbivores, suggesting that it may be maintained in populations by selection imposed by herbivores.
Finally, two papers focus on phenotypic plasticity in physiological traits. Maherali et al. (pp. 908–918) consider the question of whether changes in physiological traits over the life cycle are adaptive. They demonstrate that in Avena barbata, ontogenic change in photosynthesis is adaptive in wet-soil environments, though apparently neutral in dry soils. Lev-Yadun & Holopainen (pp. 506–512) address a long-standing question in plant biology: why trees in North American and East Asia produce red leaves in autumn. Using a comparative geographical analysis, they present evidence supporting a new hypothesis: that red autumn leaves are a relict of adaptation to different climates and herbivores of the Tertiary.
Emerging themes and future directions
Adaptation and selection will remain central foci of research activity for as long as organisms interact with the physical and living components of their environments. Building on Darwin's foundation, evolutionary studies now aim to reveal the molecular and genomic properties of these interactions at the individual, population, community and phylogenetic levels. Recent work provides keen insights into the environmental context dependence of gene expression, the complex interplay of biotic and abiotic selection pressures in real habitats, the pathways and constraints to adaptive transitions, and the nature of evolutionary innovations in functional and reproductive traits. Evolutionary perspectives are generating new ways of thinking about plant development, physiology, geographic distribution, and community ecology. These interdisciplinary research efforts will also continue to inform invasion biology and biodiversity conservation in powerful ways.
This special issue of New Phytologist, which is devoted to plant adaptation, celebrates the bicentennial of Charles Darwin's birth and the sesquicentennial of the publication of On the origin of species. This issue was edited by Richard Abbott, David Ackerly, Laura Galloway, Mark Rausher, Ruth Shaw and Sonia Sultan. In addition, Laura Galloway, Mark Rausher, Ruth Shaw and Holly Slater provided comments and suggestions on the introductory Editorial. We thank the many Authors and Reviewers who have also contributed to this special issue and we hope that you, our Readers, will enjoy and take inspiration from it.