Frontiers in chemical ecology and coevolution


(tel +1 607-254-4255; email

7th New Phytologist Workshop in Ithaca, NY, USA, September 2013

There has been tremendous growth in the area of chemical ecology and the study of coevolution. Over the past decade, the pages of New Phytologist have been publishing increasingly important reviews (Strauss et al., 2005; Keeling & Bohlmann, 2006; Heil, 2008; Dudareva et al., 2013) and original studies (Leitner et al., 2005; Agrawal et al., 2009; Huang et al., 2012; Zhang et al., 2013) that have advanced our ecological and evolutionary understanding of chemically-mediated interactions. In particular, classic questions on the interactions of plants and herbivores, pollinators, and microbes have been infused with modern chemical and molecular methods, which has enhanced progress. In addition, there continues to be a healthy tension between the use of model and nonmodel study systems in chemical ecology, two ways to make progress on interdisciplinary issues. The good news is that new methods (e.g. genome-wide association mapping) can turn nearly any organism into a ‘model’. As always, issues of chemical ecology stand at the border between advancing knowledge of biodiversity and basic plant biology with that of pest management and maximizing plant production in the face of enemy attack (Fig. 1).

Figure 1.

Chemical ecology and coevolution. (a, b) A monarch butterfly caterpillar (Danaus plexippus) taking bites of common milkweed (Asclepias syriaca). After the exudation of sticky toxic latex, the caterpillar wipes its mouth clean. (c) A honey bee (Apis melifera) approaching common milkweed flowers. (d) Predation of a juvenile Colorado potato beetle (Leptinotarsa decemlineata) by a stink bug (Podisus maculiventris). All photographs courtesy of Ellen Woods (University of Connecticut, USA).

It was in this context that the 7th New Phytologist Workshop was held in Ithaca, NY, USA early in the autumn of 2013. The workshop, made up of over 30 participants (Fig. 2), was convened by an organizing committee based at Cornell University, Ithaca, NY, USA (Anurag Agrawal, André Kessler, Georg Jander, Robert Raguso, Jennifer Thaler), spanning four departments (Ecology & Evolutionary Biology, Plant Biology, Neurobiology and Behavior, and Entomology). A committee of graduate students and postdoctorates was also central to the planning and implementation of the workshop: Geoffrey Broadhead, Clare Casteel, Mônica Kersch-Becker, Maya Lim, and Marjorie Weber. In addition to funding from the New Phytologist Trust, further support came from Cornell's University Lectures Committee. In this paper I have highlighted some of the meeting's major emergent themes and discussions.

Figure 2.

Meeting participants at the 7th New Phytologist Workshop (Taughannock Falls, Ulysses, NY, USA).

Community and evolutionary ecology

Although it has long been recognized that the interaction ‘modules’ we study are embedded in a larger biotic and abiotic context, the field has been slow to embrace that larger context, especially in an evolutionary sense. Yes, we have the classic papers and concepts related to vegetational diversity impacting plant–herbivore interactions (Root, 1973), the concepts of diffuse coevolution (Strauss et al., 2005), and advances in our understanding of geographic variation in interactions within and between continents (Zangerl & Berenbaum, 2003). However, admittedly, it is a challenge to incorporate community complexity into our models. At the workshop, Anurag Agrawal and André Kessler, and invited speakers Erik Poelman (Wageningen University, the Netherlands) and Julia Koricheva (Royal Holloway, University of London, UK) addressed these issues in a variety of natural systems. Predicting the extent of context dependence will always be a sticky issue, and this group highlighted the importance of novel interactions between defense and competition, plant biodiversity, and specificity in response to the diverse assemblages of insects on plants (Poelman et al., 2011; Milligan & Koricheva, 2013; Uesugi & Kessler, 2013). Perhaps the biggest open gap is still identifying the most important (read: fitness impacting) players in any given system; only after identifying these important species can we make progress on evolutionary issues in a community context.

‘Most species interactions involve at least three trophic levels … and it behoves us to not only recognize this, but to incorporate such interactions into our conceptual models.’

The chemistry of attraction and defense

Following these lines, Scott Armbruster (Portsmouth University, UK) presented a historical perspective on evolutionary transitions between pollinator rewards and defense against herbivores (Armbruster et al., 2009). Indeed, it was his classic work, and through the continuing contributions of speakers Lynn Adler (University of Massachusetts, Amherst, MA, USA) and Robert Raguso that additional emphasis is being placed on the role of chemistry in pollen, nectar, and floral volatile emissions (Raguso, 2009; Adler et al., 2012). Although we clearly have much to learn and explore, it became apparent from the presentations that floral chemistry is perhaps the most underappreciated aspect of pollination biology. Furthermore, not only are there important implications for pollinators, but also for the community of other plant visitors. Given the complexity of interactions, we cannot assume that floral traits are targeted at pollinators or herbivores; they may have multiple roles, some of which are unexpected, and likely to be context dependent (Huang et al., 2012).

Mechanisms of defense and counter-defense

The remarkable diversity of secondary compounds that have been isolated from even a single species, for example, the diverse glucosinolates from the model plant Arabidopsis thaliana, present tremendous challenges for the study and interpretation of secondary metabolism. Does this compound diversity represent specific adaptations targeted at different attackers, synergistic combinations of toxins, or some nonadaptive outcome (Müller et al., 2010; Bekaert et al., 2012)? Genetically modified plants, network analysis, and the dissection of hormonal pathways are all contributing to unravelling the physiological basis and evolutionary underpinnings of secondary metabolism (Kliebenstein, 2012) (with contributions from Georg Jander and invited speakers Dan Kliebenstein, University of California, Davis, CA, USA and Martin Heil, Cinvestav, Irapuato, Mexico). For defensive responses, there are general mechanisms through which plants respond, and many additional pathways for fine-tuning the specific plant response (Heil et al., 2012). Finally, Noah Whiteman (University of Arizona, Tucson, AZ, USA) spoke about the evolution of herbivory in flies, and the role of adaptations in metabolism and detoxification that have facilitated shifts of drosophilids as leaf mining herbivores of the Brassicaceae (Whiteman et al., 2012). Indeed, the transition to herbivory (from saprophagy) appears to be dependent both on the evolution of novel genes and the modification of existing stress and toxin-detoxification genes. Although it is still unclear how general these results are, studying the genetic basis and evolution of plant feeding (and specialization to various plant parts) could well be revolutionized in the coming decades.

Scaling up and out

If what happens in our Petri dishes, laboratories, or even local field sites cannot be generalized to larger scales, then our science is in trouble. Four speakers focused on different aspects of scaling in plant–animal chemical ecology and coevolution. Thinking of plants as dynamic organisms has been the rule within circles of plant biologists for decades. Nonetheless, this is often not widely acknowledged. Karina Boege (Universidad Nacional Autónoma de México, Mexico City, Mexico) addressed the generality of plant phenotypes by considering changes that occur during plant ontogeny. Indeed, ontogenetic switches in plant phenotypes bear on many issues from simply the methods of measuring plant traits, to constraints on the expression of phenotypes, to the evolution of multiple integrated traits (Boege & Marquis, 2005). The intersection of developmental changes through a season and the order of pest arrival (Viswanathan et al., 2007) generated interest and discussion in terms of understanding how plants and insects can dynamically shape the assemblage of insect communities. At the other end of the spectrum, Katja Poveda (Cornell University, Ithaca, NY, USA) focused on landscape ecology, and whether environmental heterogeneity at the landscape scale translated into predictable local trophic interactions that can impact plant productivity (Poveda et al., 2012). Higher trophic level interactions, focusing on predation risk, and its cascading effects on prey behavior and physiology was the focus of presentations by Jennifer Thaler and invited speaker Oswald Schmitz (Yale University, New Haven, CT, USA). That predation risk alone can have major consequences not only for prey behavior and plant damage, but also nutrient cycling, highlights the dramatic extent to which pieces of ecological webs are connected (Hawlena et al., 2012; Thaler et al., 2012). Most species interactions involve at least three trophic levels, occur in a landscape context, and change over ontogeny, and it behoves us to not only recognize this, but to incorporate such interactions into our conceptual models. Although tri-trophic theory is well-established, we are still developing strong predictions for these other axes.

Conclusion and speculation

These are exciting times to be studying chemical ecology and coevolution. Advances and availability of tools, techniques, and information (i.e. genomes and phylogenies) are enhancing our ability to address classic questions. The most influential emerging work is integrative in some way, crossing boundaries or approaches (i.e. looking at roots, including microbes, incorporating insect physiology, etc.). If plant–animal interactions are truly a model for how species interact more generally, then we can expect to see substantial progress in understanding both mechanisms and outcomes of such interactions in the coming decades.