Evolution of plant defenses and herbivore host selection in the neotropical genus of trees inga (fabaceae)

Because plants and their insect enemies are strikingly species-rich groups, understanding their interactions has been a key issue in ecology and evolution. The arms race between plants and herbivores is considered the driver of diversification in both groups. However, we have a poor understanding of how these processes lead to divergence and speciation. This dissertation research tests key theories that relate plant- insect interactions with diversification and coevolution in both groups of organisms. In the first part, I assess the utility and contrast the predictions of two theories aimed to explain the patterns of defense investment across species: The Apparency Theory and the Resource Availability Hypothesis. My results provide strong support for the predictions of the Resource Availability Hypothesis. In particular, the evolution of defenses appears to be related to interspecific differences in inherent growth rate rather than to a species' predictability to herbivores. The theory appears robust across latitude and ontogeny suggesting that it has served as a valid framework for investigating the patterns of plant defenses and that its applicability is quite general. In the second part, I focus on how herbivores may drive the evolution of plant defenses, how plant defenses shape herbivore host choice and how plant-herbivore interactions might influence community composition and diversity focusing on the Neotropical genus of trees Inga (Fabaceae). I characterize the entire suite of anti-herbivore defenses and also quantify the diversity and abundance of leaf-feeders associated with Inga. With the use of phylogenies for both plants and herbivores, I discriminate among possible macroevolutionary hypothesis of host use and plant defense evolution. Contrary to much coevolutionary theory, my results show that closely related Inga species are more divergent in anti-herbivore defenses than in non-defense traits, and that the evolution of host use in herbivorous insects is more conserved with respect to host defenses rather than to host phylogeny. Together, these results suggest that defenses evolve rapidly and that traits related to host choice evolve more slowly. Specifically, although divergence in herbivores might not be driven by their interactions with plants,herbivores may be an important factor driving the divergence among plant species.

EVOLUTION OF PLANT DEFENSES AND HERBIVORE HOST SELECTION IN THE NEOTROPICAL GENUS OF TREES INGA (FABACEAE) by Maria Jose Endara Burbano A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biology The University of Utah August 2015 Copyright © Maria Jose Endara Burbano 2015 All Rights Reserved The Uni v e r s i t y of Utah Graduat e School STATEMENT OF DISSERTATION APPROVAL The dissertation of Maria Jose Endara Burbano has been approved by the following supervisory committee members: Phyllis D. Coley_____________ , Chair _______ 5/5/2015 Date Approved Thomas A. Kursar____________ , Co-chair _______ 5/5/2015 Date Approved Frederick R. Adler___________ , Member _______ 5/5/2015 Date Approved M. Denise Dearing____________ , Member _______ 5/5/2015 Date Approved Eric W. Schmidt_____________ , Member _______ 5/5/2015 Date Approved Michael D. Shapiro___________ , Member _______ 5/5/2015 Date Approved and by __________________ M. Denise Dearing____________ , Chair/Dean of the Department of _______________________Biology and by David B. Kieda, Dean of The Graduate School. ABSTRACT Because plants and their insect enemies are strikingly species-rich groups, understanding their interactions has been a key issue in ecology and evolution. The arms race between plants and herbivores is considered the driver of diversification in both groups. However, we have a poor understanding of how these processes lead to divergence and speciation. This dissertation research tests key theories that relate plant-insect interactions with diversification and coevolution in both groups of organisms. In the first part, I assess the utility and contrast the predictions of two theories aimed to explain the patterns of defense investment across species: The Apparency Theory and the Resource Availability Hypothesis. My results provide strong support for the predictions of the Resource Availability Hypothesis. In particular, the evolution of defenses appears to be related to interspecific differences in inherent growth rate rather than to a species' predictability to herbivores. The theory appears robust across latitude and ontogeny suggesting that it has served as a valid framework for investigating the patterns of plant defenses and that its applicability is quite general. In the second part, I focus on how herbivores may drive the evolution of plant defenses, how plant defenses shape herbivore host choice and how plant-herbivore interactions might influence community composition and diversity focusing on the Neotropical genus of trees Inga (Fabaceae). I characterize the entire suite of anti­herbivore defenses and also quantify the diversity and abundance of leaf-feeders associated with Inga. With the use of phylogenies for both plants and herbivores, I discriminate among possible macroevolutionary hypothesis of host use and plant defense evolution. Contrary to much coevolutionary theory, my results show that closely related Inga species are more divergent in anti-herbivore defenses than in non-defense traits, and that the evolution of host use in herbivorous insects is more conserved with respect to host defenses rather than to host phylogeny. Together, these results suggest that defenses evolve rapidly and that traits related to host choice evolve more slowly. Specifically, although divergence in herbivores might not be driven by their interactions with plants, herbivores may be an important factor driving the divergence among plant species. iv This dissertation is dedicated to my daughter Emilia and my husband David TABLE OF CONTENTS ABSTRACT...........................................................................................................................iii LIST OF TABLES............................................................................................................... viii LIST OF FIGURES................................................................................................................ x ACKNOWLEDGEMENTS..................................................................................................xiii Chapters 1 INTRODUCTION............................................................................................................... 1 Background.................................................................................................................. 1 Chapter summaries.......................................................................................................5 Significance of the study............................................................................................. 9 References...................................................................................................................10 2 THE RESOURCE AVAILABILITY HYPOTHESIS REVISITED: A META-ANALYSIS...............................................................................................................14 Summary.................................................................................................................... 15 Introduction................................................................................................................15 Contrasting plant apparency and RAH.................................................................... 18 Results........................................................................................................................ 18 Discussion.................................................................................................................. 20 Conclusions................................................................................................................21 References.................................................................................................................. 23 Appendix S1: Studies included in the meta-analysis..............................................25 Appendix S2: Materials and methods......................................................................28 Appendix S3: Meta-analysis records....................................................................... 35 3 DIVERGENT EVOLUTION IN ANTI-HERBIVORE DEFENSES WITHIN SPECIES COMPLEXES AT A SINGLE AMAZONIAN SITE........................................49 Abstract.......................................................................................................................49 Introduction................................................................................................................50 Methods.......................................................................................................................52 Results........................................................................................................................ 63 Discussion.................................................................................................................. 68 Conclusions................................................................................................................ 75 References.................................................................................................................. 84 4. TESTING THE COEVOLUTIONARY ARMS RACE: A CASE STUDY WITH THE GENUS OF TREES INGA AND ITS HERBIVORES..............................................90 Abstract.......................................................................................................................90 Introduction................................................................................................................91 Methods.......................................................................................................................94 Results.......................................................................................................................104 Discussion................................................................................................................ 107 Conclusions.............................................................................................................. 115 References................................................................................................................ 127 Appendices A: SUPPORTING INFORMATION FOR CHAPTER 3................................................. 135 B: SUPPORTING INFORMATION FOR CHAPTER 4 ................................................. 162 vii LIST OF TABLES Table S3.1 Prediction (1) meta-analysis records for the effect of resources on plant growth. N= sample size.........................................................................................................36 Table S3.2 Prediction (2) meta-analysis records for the effect of growth on leaf lifespan. N= sample size........................................................................................................................38 Table S3.3 Prediction (3) meta-analysis records for the effect of growth on defense investment. Comparison= Indicates whether the study was used for 1 (within-site) or 2 (among-site) sub analyses; N= sample size......................................................................... 39 Table S3.4 Prediction (4) meta-analysis records for the effect of growth on herbivore damage. Comparison= Indicates whether the study was used for 1 (within-site) or 2 (among-site) sub-analyses; N=sample size.......................................................................... 46 4.1 Pairwise correlations between defense traits among Inga species. Correlation coefficients with f are phylogenetic independent contrasts, the rest are partial mantel r. Significant values (P<0.05) are in bold..............................................................................118 4.2 PCA loadings. Correlations between components and continuous defense traits_119 4.3 Summary statistics for the relationship between herbivore communities and host plant traits. r represents the Mantel correlation between the dissimilarity in host plant traits and their herbivore communities measured by the Bray-Curtis index. Significant values (P<0.05) are in bold..................................................................................................120 4.4 Results of best-fit distance-based redundancy analyses (db-RDA) model for the three most abundant lepidopteran families..................................................................................121 4.5 Results of maximum likelihood analyses for the three most abundant lepidopteran families against host plant traits. AICc: Akaike Information Criteria corrected for small sample sizes. AAICc: difference in AIC scores between the best model (listed first) and each competing model. wi: AIC weight (level of support for a model, max. weight=1). Intercept: null model. Coefficient estimates: raw estimates that indicate a positive or negative effect of each predictor variable on the response variable. Values of coefficients whose 95% credible interval (95% CI) does not include zero are in bold.......................122 A.3 Chromatographic gradient used for the LC-MS analyses 139 A.4 Phenolic content for Inga species as g phenolics per g DW of leaf (± 1 SD) for 5 replicate extractions. Analyses of Variance did not detect differences in phenolic investment among the ESUs within each species complex..............................................140 A.6 Metabolites that distinguish the ESUs (biomarkers) were detected by LC-QToF-MS and were identified by PCA or by inspection. The mass to charge ratio of the ion is indicated as "m/z". "Elemental composition" was obtained from MassLynx (Elemental Composition v 4.0©, Waters Corporation, 2000, Milford, MA). The "expected m/z" equals the monoisotopic mass calculated from "elemental composition" by MassLynx or from isotopic masses in the National Institute of Standards and Technology database (Coursey et al. 2010). The error in the observed m/z is in parts per million (ppm). "GC/EG" is gallocatechin/epigallocatechin. "DOPA" is 3,4-dihydroxyphenylalanine. The columns for capl, cap2, cap3, hetl, and het2 indicate the approximate abundance of each m/z in exponential notation or the approximate abundance as follows: moderately abundant (>105 ion counts), abundant (104 to 105 ion counts), and present (103 to 104 ion counts) indicated as "xx", "x", and "p", respectively. A question mark indicates detected in 1 to 3 samples and missing in 2 to 4 samples................................................................146 A.7 Insect herbivore species found on each I. capitata ESU. Total number of insects observed by ESU are indicated. For the analysis of herbivore association, singletons were not included..........................................................................................................................156 A.9 DIC (Deviance Information Criterion) comparison of unconstrained vs. constrained models estimating the strength of preference of the sawflies for a particular ESU. More negative DIC values mean a better fit.................................................................................161 B .3 PCR protocol................................................................................................................166 ix LIST OF FIGURES 2.1 Effects of defense investment on realized growth. Each curve represents a plant species with a different maximal inherent growth rate. Levels of defense that maximize growth are indicated by arrows. From Colin, Bryant & Chapin (1985)............................ 17 2.2 Mean and 95% confidence intervals for the effect sizes of resources on plant growth measures (weighted standardized mean, Hedges' d): for all studies (n=24) and for studies conducted only in tropical forests (n=17), in temperate forests (n=6), with seedlings (n=9), with saplings (n=5), and with adults (n=8)...............................................................19 2.3 Mean effect sizes (z) and 95% confidence intervals for growth rate effects on investment in plant defenses and herbivory between species within a site. The dependent variables include: leaf lifetime (n=10), all constitutive defenses (n=57), all chemical defenses (n=23), total phenolics (n=6), tannins (hydrolysable and condensed; n=12), terpenes (n=4 records from one study from Fine, Mesones & Coley 2004), mechanical defenses (n=6), nutrient content (n=25) and herbivory (n=16)..........................................19 2.4 Mean effect sizes (z) and 95% confidence intervals for growth rate effects on investment in plant defenses between species in habitats with different nutrient availability. Dependent variables include: all defenses (n=24), chemical defenses (n=18), total phenolics (n=8), tannins (n=10) and mechanical defenses (n=6)..............................20 2.5 Mean effect sizes (z) and 95% confidence intervals for growth rate effects on investment in plant defenses between species in habitats with different light availability. Dependent variables include: all defenses (n=17), chemical defenses (n=9), total phenolics (n=4), tannins (n=5), mechanical defenses (n=8) and nutrient content (n=2)...20 3.1 Heatmap of a hierarchical clustering of I. capitata and I. heterophylla ESUs based on relative abundances of the most important 25 UPLC-MS phenolic metabolites. Each column represents a metabolite with a unique m/z and retention time; analyses are based on 5 individuals per ESU. Each row is one UPLC-MS analysis from one individual plant. Metabolites were identified as "important" based on ANOVA analysis. The color scale for metabolite relative abundance is based on signal intensity (total ion current from the mass spectrometer)................................................................................................................. 76 3.2 Score scatter plots from a PLS-DA model fitted to the relative abundances of peaks obtained by metabolic fingerprinting. (a) and (b) I. capitata species complex. (c) and (d) I. heterophylla species complex. The percentage of the variation explained by each component is indicated on the axes. The ellipses delimited by the dotted lines represent the 95% confidence regions. P values are provided in the Results.................................... 77 3.3 Non-chemical defensive traits of leaves for the ESUs of the I. capitata species complex. In panel (a) circles represent capl (n=24), squares represent cap2 (n=65), and triangles represent cap3 (n=74). Letters denote significant differences between ESUs. Bars are mean ± SE................................................................................................................ 78 3.4 Ordination diagram of 38 I. capitata plants based on the similarities of their insect herbivore faunas (stress value= 0.05). Similarities in herbivore composition were calculated with the Bray-Curtis Index..................................................................................79 3.5 Summary of sawfly host preferences in: (a) choice experiment and, (b) field survey. In both cases, preference is plotted as the median and 95% confidence interval of the posterior probability distribution for population preference, estimated from a hierarchical Bayesian model (Fordyce et al. 2011). Lower-case letters denote posterior probabilities of >0.95 for differences in preferences.................................................................................80 3.6 Non-defensive traits of leaves for the ESUs of the I. capitata species complex. Letters denote significant differences between ESUs. Bars are mean ± SE. LMA=Leaf mass per area, N= Nitrogen...................................................................................................81 3.7 Score scatter plots from a PCA model fitted to GC-MS data of primary metabolites. (a) I. capitata species complex. (b) I. heterophylla species complex. The percentage of the variation explained by each component is indicated on the axes. The ellipses delimited by the dotted lines represent the 95% confidence regions................................. 83 4.1 Comparison between the phylogenetic tree (left) and the defensogram (defense dendrogram, right) for Inga species....................................................................................123 4.2 Relationship between the similarity of lepidopteran communities (based on Bray- Curtis Index) on host plants vs. (a) distance in defenses between Inga hosts and (Partial Mantel r= 0.5, p=0.001) (b) phylogenetic distance between Inga hosts for all pairwise combinations of plants (Mantel r= 0.25, p=0.02)............................................................. 124 4.3 Constrained Analysis of Principal Coordinates of the most parsimonious model for the lepidoptera community similarity measured by the Bray-Curtis index (R2adj= 0.4, p=0.001). Each dot represents an Inga species host color-coded by defense chemistry.125 xi 4.4 Bipartite trophic network of Inga hosts and herbivores. (a) Phylogenies of Inga and Elachistidae plotted in the margins (Parafit test: p=0.52). (b) Phylogeny of Elachistidae and Inga defensogram plotted in the margins (Parafit test: p=0.05). For each network lower bars represent host abundance and upper bars represent herbivores abundance. Linkage width represents frequency of the interaction..................................................... 126 A. 1 Photographs of the Inga species complexes. (a) Inga capitata ESU capl, (b) Inga capitata ESU cap2, (c) Inga capitata ESU cap3, (d) Inga heterophylla ESU hetl and (e) Inga heterophylla ESU het2.................................................................................................136 A.2 Maximum clade credibility tree of the species of the genus Inga in Los Amigos, Peru, from a Bayesian analysis (in BEAST, Drummond & Raumbaut. 2007) of 6000 bp of plastid DNA and the ITS nuclear marker. Branch lengths represent time in millions of years. For the root node, a normally distributed prior was used with a mean of six million years based on divergence times across legumes. Clades containing the I. capitata species complex (ESUs capl, cap2 and cap3) and I. heterophylla species complex (ESUs hetl, het2) are colored in red. This tree shows that the ESUs within each species complex are more closely related to each other than with any other Inga species (Modified from Kursar et al. 2009 by Dr. Kyle Dexter)..............................................................................138 A.5 Total ion chromatograms showing relative intensities of peaks from the LC-QToF-MS for the different ESUs in (a) and (b) positive mode and (c) and (d) negative mode. (a) and (c) I. capitata species complex, (b) and (d) I. heterophylla species complex.......... 141 A.8 . Ordination diagram of 31 Inga capitata plants based on similarities on their herbivore faunas (stress=0.01), with a cutoff number of 3 individuals per herbivore MOTU. Circles represent cap1, triangles represent cap2, and squares represent cap3...160 B.1 Lepidopteran herbivore families associated with Inga in Los Amigos. Percentages represent the relative abundances of the most abundant families. Elachistidae, Erebidae and Riodinidae are the families that were analyzed separately with and without joint-absence information............................................................................................................. 163 B.4 Number of MOTU (Molecular Taxonomic Operational Unit) defined at each cutoff value. The arrow shows the cutoff used for this study (15 base pairs)............................167 B.5 Constrained Analysis of PCO of the most parsimonious model for the lepidoptera community similarity measured by the Bray-Curtis index (R2adj= 0.42, p=0.001). This analysis included only species for which we had data on phenology of leaf production. Each dot represents an Inga species host color-coded by defense chemistry..................168 xii ACKNOWLEDGEMENTS First and foremost, I would like to give heartfelt thanks to Phyllis Coley and Tom Kursar. They were not only my advisors, but also my mentors and friends. Their patience, understanding and genuine caring and concern during graduate school enabled me to have a family while also earning my PhD. They have been inspirational and enlightening. Thank you Lissy and Tom for setting the bar high and demanding a high quality of work in all my endeavors. I would like to thank the members of my committee: Denise Dearing, Fred Adler, Mike Shapiro and Eric Schmidt. They provided useful feedback and insightful comments on my work. I would also like to thank past and present fellow lab members and collaborators for their help and support in the accomplishment of my dissertation work. Special thanks to Alexander Weinhold for his help with metabolomics, to Kyle Dexter and Toby Pennington for providing the phylogeny of Inga, to James Nicholls for his expertise in DNA work, and to Gabby Ghabash for her knack in solving seemingly intractable difficulties with DNA sequencing. Without her efforts the last part of this dissertation would have undoubtedly been much more challenging. My time at Los Amigos Biological Station was made enjoyable in large part due to the exceptional people with whom I interacted. I am grateful for time spent with roommates and friends, for my yoga buddies and for the memorable trips to CM1 and CM2. I thank Jesus and all CICRA employees for obliging me with every request I had, and Don Pascual and Chico for the delicious meals they prepared. I am particularly indebted to Wilder Hidalgo for being an exceptional field assistant and friend. I thank my family and close friends for all their love and encouragement. To my parents for raising me with a love for nature and for supporting me in all my pursuits. To David Donoso for his unconditional friendship, for the science talks and for identifying the ants!!! To Erika Carrera and Xavier Haro for always being there. And most of all, to my loving, encouraging and patient husband David. Please know how much I appreciate your support during my PhD. Thank you. My gratitude is also extended to Shannon Nielsen and Renae Curtz. Thank you for always providing assistance with every question I have ever asked regarding Graduate School, Accounting, you name it. Finally, I would like to acknowledge the funding sources that enabled me to pursue my graduate studies. This work was supported by the Secretaria Nacional de Education Superior, Ciencia, Tecnologia e Innovation del Ecuador (SENESCYT), and grants from CREO (Conservation, Research and Education Opportunities), GCSC (Global Change and Sustainability Center)and ISC (International Student Center) from the University of Utah to M.J.E., and NSF grants (DEB-0640630 and Dimensions 1125733) to P.D.C and T.A.K. Additional funding and logistical support were provided by the University of Utah Biology Department. xiv CHAPTER 1 INTRODUCTION Background A primary goal in ecology and evolution is to understand the origin and maintenance of biodiversity. Major subjects of such inquiries are the plants and their insect herbivores for together they account for more than half of the macroscopic diversity on land. Theory has long predicted that the evolution of plant anti-herbivore defenses and insect counter-adaptations are the drivers of such a great diversity (Ehrlich & Raven 1964). However, we have a poor understanding of how these processes might lead to divergence and speciation. Consequently, for my dissertation research, I focus on a genus of trees and its associated herbivores to test key theories that relate plant-insect interactions with diversification and coevolution in both groups of organisms. Herbivores exert major selective pressures on plants (Agrawal et al. 2012). In response, plants have developed a variety of defensive traits, which are particularly prevalent in young, expanding leaves (Coley & Aide 1991; Kursar & Coley 2003; Brenes-Arguedas et al. 2006). These include physical structures (e.g. hairs, thorns), secondary chemical compounds (e.g., phenolics, saponins, non-protein amino acids), nutritional quality, and phenological escape. These defensive traits are considered direct 2 defenses because they act directly on the herbivore. Plants have also evolved partnerships with species from the third trophic level. These are termed indirect defenses, and involve many different mechanisms. For example, plants may provide food to predators of insect herbivores (e.g., extrafloral nectar) or produce volatile signals to attract parasitoids of herbivores. However, herbivores have also developed a series of physiological, morphological and behavioral counter-adaptations to exploit their host plants. Thus, plants and insects appear to be reciprocally inducing evolutionary change between each other. For many years the study of plant-insect interactions focused primarily on the role of secondary metabolites in plants and on the study of the costs and benefits of defenses. Once the significance of secondary chemicals in defense and in host selection was established (Dethier 1954; Frankel 1959), interest shifted to explaining why the amount and type of defense differed considerable between species. For example, Feeny (1976) proposed that differences in defenses between species are related to a species' predictability to herbivores; Coley et al. (1985) suggested that the evolution of defenses is related to interspecific differences in inherent growth rate. These theories stimulated a multitude of studies and established the basic conceptual framework linking defense and herbivory. However, it was Ehrlich & Raven (1964) who incorporated a macroevolutionary framework into the study of plant-insect interactions. In their seminal paper, they considered the reciprocal nature of the adaptive responses between plants and herbivores and introduced the idea of coevolution between plants and butterflies, a concept that has dominated our understanding of the interactions between plants and insects for the last 50 years. According to Ehrlich & Raven's paradigm there is taxonomic conservatism in the expression of defenses in plants and in the use of hosts in insect herbivores. They suggested that this results from an ongoing coevolutionary arms race between plants and enemies. Specifically, this theory predicts that, after the evolution of a key innovation or "a new defense" in response to herbivore pressure, a plant species would be able to escape most herbivores and radiate into a clade of chemically similar plants. Thus, closely related plants would have similar defenses. Similarly, selection would favor counter-adaptations from herbivores to the novel plant defense and ultimately adaptive radiation of the herbivores onto a set of closely related plants. Hence, related insects should use related plants (Ehrlich & Raven 1964). Because of the defense and counter­defense coevolution pattern, clades of plants and insect herbivores are expected to mirror each other in their patterning of speciation events (e.g., topology of their phylogenies). Ehrlich & Raven suggested that these historical processes of radiation in both organisms may explain a substantial fraction of plant and insect diversities. Ever since its publication, the coevolutionary theory of arms race was rapidly accepted and profoundly shaped the field. Evolutionary conservatism in plant-insect association became the new paradigm. Nevertheless, although explicit tests are few, they have found contradictory results. Examinations of the phylogenetic structure of plant defenses have revealed that close relatives tend to be dissimilar in defenses (Agrawal & Fishbein 2006; Bursera 1997; Becerra et al. 2009; Kursar et al. 2009; Sedio 2013). Furthermore, host shifts among closely related herbivores are more strongly correlated 3 4 with the chemistry of the new hosts than with their phylogenetic relationships (Becerra 1997; Becerra & Venable 1999; Berenbaum 2001; Wahlberg 2001), and phylogenies of insects rarely match those of their hosts (Futuyma & Agrawal 2009; Thompson 1994). Thus, the wide acceptance of a strong phylogenetic conservatism for defenses in plants and, hence, for host use in herbivores may not be warranted, suggesting that these ideas need refining and more rigorous testing. Here, I examine macroevolutionary patterns of defense evolution and the contribution of plant-insect interactions to divergence in defenses among species and to community assembly. First, by means of a meta-analysis, I assess the utility of two of the most influential theories aimed to explain the patterns of defense investment across species. Second, I perform a complete characterization of plant functional traits and evolutionary relationships for plants and insect herbivores and determine how they fit into the arms race scenario. For this, I use Inga, a genus of trees and its associated herbivores in the Peruvian Amazon, a habitat where the diversity of plants and invertebrates is among the highest in the terrestrial world and where the arms race may be particularly pronounced. Study system: The genus of trees Inga and its herbivores. I focus my study on the speciose (>300 species), ecologically important, and abundant Neotropical tree genus, Inga (Fabaceae). This high diversity appears to be the result of a recent, explosive radiation within the last 4-10 million years (Richardson et al. 2001). At my study site in Peru, Los Amigos, and elsewhere, the genus Inga constitutes one of the most abundant and diverse genera (N. Pitman unpublished data), with more 5 than 40 species occurring in 25 ha and 6% of the stems (Valencia et al. 2004). If not the highest, Inga certainly is among those genera with the greatest number of congeneric species at a given site. The study of defenses in Inga is exceptional in that in a single genus we can find multiple resistance traits. Inga produces a variety of secondary metabolites including non-protein amino acids, flavonoids, flavan-3-ols, saponins and amines (Lokvam et al. 2004; Lokvam & Kursar 2005; Brenes-Arguedas et al. 2006; Lokvam et al. 2007). The content of nitrogen in the leaf, a key nutritional trait for herbivores, varies twofold during the development of the young leaf (Kursar & Coley 2003). Also, Inga is characterized by the presence of nectaries on leaves that produce nectar and attract ant defenders (Koptur 1984). Some species have other physical (trichomes) and phenological (synchrony in leaf production) defenses as well (Coley et al. 2005). Inga is associated with several groups of herbivores including Coleoptera, Orthoptera, phloem-feeding Coreidae, Diptera, sawflies, Phasmida and Lepidoptera. However, the group causing the most damage to leaves of Inga is Lepidoptera (Kursar et al. 2006). For this study I focused only on lepidopteran herbivores due to their importance. Chapter summaries Revisiting the Resource Availability Hypothesis (Chapter 2) Among theories of plant defense, the Apparency Theory (Feeny 1976) and the Resource Availability Hypothesis (Coley, Bryant & Chapin 1985) stand out as two of the most influential plant defense theories in the last few decades. These theories have aimed 6 to explain patterns of defense investment and the selective pressures that have led to the variety of defensive strategies across species. The Resource Availability Hypothesis relates the evolution of defenses to interspecific differences in inherent growth rate, whereas Apparency Theory assumes that defenses are related to a species' predictability to herbivores. Although the theories have different assumptions regarding the reasons leading to defense differences, some of the predictions are similar. For example, both theories agree that long-lived, slow-growing species (apparent species) should invest more in defenses than short-lived, fast-growing species (unapparent species). However, a fundamental difference between the theories is their contrasting predictions for the amount of herbivory. The Resource Availability Hypothesis predicts that fast-growing species should suffer greater herbivore damage, while Apparency Theory predicts similar losses for apparent and unapparent species. Predictions from the Apparency Theory and the Resource Availability Hypothesis have been repeatedly tested. Overall, the evidence provides mixed support: some of the results are consistent with the predictions of the hypotheses whereas others not. In an early review (Stamp 2003), the assumptions and evidence for all theories of plant defense were summarized. However, to our knowledge, no attempt to revise quantitatively the utility of these two theories has been considered. For this manuscript, we provide a historical review of both theories, present evidence that shaped their development and, by means of a meta-analysis, contrast their predictions with results from many studies. Our results provide strong support for the predictions of the Resource Availability Hypothesis. In particular, species adapted to resource-poor environments 1) grow inherently more slowly, 2) invest more in constitutive defenses, and 3) support lower herbivory than species from more productive habitats. We also 7 found that the application of this theory appears robust across latitude and ontogeny, as the magnitude of the effect sizes for most of the predictions did not vary significantly between ecosystems nor across ontogenetic stages. Our data also showed that variation in growth rate among species better explains the differences in herbivory than variation in apparency, suggesting that the evolution of different defensive strategies across species is resource, rather than herbivore-driven. Therefore, we suggest that the Resource Availability Hypothesis is very broadly applicable for understanding the evolution of anti-herbivore defenses on timescales that are long and that probably correspond to the origin of clades. Defense evolution, herbivore host selection and plant-insect evolutionary relationships in the species-rich Neotropical genus of trees Inga (Chapters 3 & 4) The remainder of my dissertation addresses the evolution of anti-herbivore defenses on timescales that correspond to the evolution of species. The arms race between plants and insect herbivores has been invoked as one of the main mechanisms driving trait divergence and coevolution for both groups (Becerra 1997; Thompson 1998; Becerra 2009; Futuyma & Agrawal 2009; Kursar et al. 2009; Agrawal 2012). Specifically, this theory proposes that adaptations between both organisms are reciprocal and that their interactions might have driven the diversification in both groups (Ehrlich & Raven 1964). Key interpretations from this theory are that closely related species of plants share similar defenses and support a community of closely related species of insect 8 herbivores. However, the few studies that have tested these critical assumptions have found contradictory results. In Chapter 3, we present evidence for enemy-related differentiation among closely related species within two clades in the genus of trees Inga. We hypothesize that herbivores may exert divergent selection on defenses, such that closely related plant species will be more different in defensive than in non-defensive traits. Contrary to much coevolutionary theory, we found that sister Inga species are more divergent in anti­herbivore defenses than in non-defense traits (e.g., for habitat use and resource acquisition). Furthermore, the assemblages of insect herbivore communities are dissimilar between the populations of coexisting Inga species. Taken together, our results suggest that herbivores may have played a significant role in selecting for their phenotypic divergence and potentially in their diversification and coexistence. In Chapter 4 we provide a more rigorous scrutiny of the coevolutionary hypothesis proposed by Ehrlich & Raven which has dominated our understanding of plant-insect interactions during the last 50 years. A particularly significant contribution of my research is the inclusion of extensive anti-herbivore traits. These are ignored in most other studies and, in their absence, alternative hypotheses cannot be rigorously compared. Specifically, for 38 Inga occurring at our study site, we analyze if closely related species of Inga are similar in defenses and determine if the herbivore community structure on Inga is based on host plant relatedness or on host defensive traits. We also compare the topologies of the phylogenies for both groups. Our results show that close relatives in Inga are highly dissimilar in defenses, that shared host plant defensive traits are more important than phylogenetic association in the community structure of the herbivores, and 9 that host use is conserved. But host use is not conserved in the classical sense, because closely related herbivores feed on plants that are similar in defensive traits but are not necessarily closely related. In other words, the topologies of their phylogenies are not congruent, suggesting that divergence in the traits of herbivore species might not be driven by their interactions with their Inga host plants and that closely related species of herbivores may diverge slowly in the traits that determine host choice. When taken together, these results show that, as expected, plant defenses determine host choice. But, they also strongly suggest that plant-antiherbivore defenses evolve rapidly and that herbivore traits involved in host choice evolve more slowly and depend more on existing host-choice traits. Hence, there is an apparent asymmetry in the interaction between Inga and its herbivores. Specifically, although divergence in herbivores might not be driven by their interactions with plants, herbivores may be an important factor driving the divergence among plant species. Significance of the study Results from my project provide mechanistic explanations about how plant defenses determine phytophagous insect host associations. They also shed light on the process of evolution, and lead to a reevaluation of the classic expectations from the coevolutionary theory of the arms race. On an ecological scale, my study provides evidence that herbivore pressure may favor neighbors that are dissimilar in their defenses. These insights will address fundamental questions about how so many species can coexist and what processes may have facilitated the evolution of the great diversity of plants and herbivores in tropical rainforests. 10 In addition to contributions to fundamental questions in ecology and evolution, a better understanding of Inga may provide practical applications in agroforestry. Many species of Inga have multiple uses on small-scale farms. Their fruits are edible and the wood is used domestically or marketed for fuel in many Latin American countries. More than 30 species are used as shade trees for perennial crops in agroforestry, such as cacao and cafe plantations. Many other benefits from Inga have been also reported, such as the fertilizing effect of leaf litter and mulch, nitrogen fixation by their symbiotic rhyzobia, maintenance of soil moisture, erosion control and weed suppression. Therefore, Inga is one of the most promising tree genera for agroforestry. By identifying Inga-specific herbivores and understanding how Inga defends against them, data from my dissertation could help to identify species combinations that differ in their defenses and do not share herbivore species. This may reduce herbivory under agroforestry situations. Additionally, Inga provides nectar on the leaves to attract predatory ants that protect the leaves by eating herbivores. My data could identify Inga species that are particularly effective at attracting ants. Ants attracted to Inga nectar may also reduce herbivory on associated shade crops. Thus my basic research should help inform and improve agroforestry use of this popular tree species. References Agrawal, A.A. & Fishbein, M. (2006) Plant defense syndromes. Ecology, 87, S132-S149. Agrawal, A.A., Hastings, A.P., Johnson, M.T., Maron, J.L. & Salminen, J.P. (2012) Insect herbivores drive real-time ecological and evolutionary change in plant populations. Science, 338, 113-117. Becerra, J.X. (1997) Insects on plants: macroevolutionary chemical trends in host use. Science, 276, 253-256. 11 Becerra, J.X. & Venable, D.L. (1999) Macroevolution of insect-plant associations: the relevance of host biogeography to host affiliation. Proceedings o f the National Academy o f Sciences (USA), 96, 12626-12631. Becerra, J., Noge, K. & Venable, D. (2009) Macroevolutionary chemical escalation in an ancient plant-herbivore arms race. Proceedings o f the National Academy o f Sciences (USA), 106, 18062-18066. Berenbaum, M.R. (2001) Chemical mediation of coevolution: phylogenetic evidence for Apiaceae and associates. Annals o f the Missouri Botanical Garden, 88, 45-59. Brenes-Arguedas, T., Horton, M.W., Coley, P.D., Lokvam, J., Waddell, R.A., Meizoso- O'Meara B.E. & Kursar, T.A. (2006) Contrasting mechanisms of secondary metabolite accumulation during leaf development in two tropical tree species with different leaf expansion strategies. Oecologia, 149, 91-100. Coley, P.D., Bryant, J.P. & Chapin, F.S. (1985) Resource availability and plant antiherbivore defense. Science, 230, 895-899. Coley, P.D. & Aide, T.M. (1991) Comparison of herbivory and plant defenses in temperate and tropical broad-leaved forests. Plant-Animal interactions: evolutionary ecology in tropical and temperate regions. (eds. P.W. Price, T.M. Lewinsohn, W.W. Fernandes & W.W. Benson) John Wiley & Sons, New York. pp. 25-49. Coley, P.D., Lokvam, J., Rudolph, K., Bromberg, K., Sackett, T.E., Wright, L., Brenes- Arguedas, T., Dvorett, D., Ring, S., Clark, A., Baptiste, C., Pennington, R.T. & Kursar, T.A. (2005) Divergent defensive strategies of young leaves in two Neotropical species of Inga. Ecology, 86, 2633-2643. Dethier, V.G. (1954) Evolution of feeding preferences in phytophagous insects. Evolution, 8, 32-54. Ehrlich, P. & Raven, P. (1964) Butterflies and plants: a study in plant coevolution. Evolution, 18, 586-608. Fraenkel, G. (1959) The raison d'etre of secondary plant substances. Science, 129, 1466­1470. Feeny, P.P. (1976) Plant apparency and chemical defense. Biochemical interaction between plants and insects. (eds. Wallace, J.W. & Mansell, R.L.). Plenum, New York. pp 1-40. Futuyma, D.J. & Agrawal, A.A. (2009) Macroevolution and the biological diversity of plants and herbivores. Proceedings o f the National Academy o f Sciences (USA), 106, 18054-18061. 12 Koptur, S. (1984) Experimental evidence for defense of Inga (Mimosoideae) saplings by ants. Ecology, 65, 1787-1793. Kursar, T.A. & Coley, P.D. (2003) Convergence in defense syndromes of young leaves in tropical rainforests. Biochemical Systematics and Ecology, 21, 929-949. Kursar, T.A., Dexter, K.G., Lokvam, J., Pennington, R.T., Richardson, J.E., Weber, M.G., Murakami, E.T., Drake, C., McGregor, R. & Coley, P. D. (2009) The evolution of anti-herbivore defenses and their contribution to species coexistence in the tropical tree genus Inga. Proceedings o f the National Academy o f Sciences (USA), 106, 18073-18078. Lokvam, J., Coley, P.D. & Kursar, T.A. (2004) Cinnamoyl glucosides of catechin and dimeric procyanidins from young leaves of Inga umbellifera (Fabaceae). Phytochemistry, 65, 351-358. Lokvam, J. & Kursar, T.A. (2005) Divergence in structure and function of young leaf chemical defenses in two co-occurring Inga species. Journal o f Chemical Ecology, 31, 2563-2580. Kursar, T.A., Wolfe, B.T., Epps, M.J. & Coley, P.D. (2006) Food quality, competition, and parasitism influence feeding preference in a Neotropical lepidopteran. Ecology, 87, 3058-3069. Lokvam, J., Clausen, T.P., Grapov, D., Coley, P.D. & Kursar, T.A. (2007) Galloyl depsides of tyrosine from young leaves of Inga laurina. Journal o f Natural Products, 70, 134-136. Richardson, J.E., Pennington, R.T., Pennington, T.D. & Hollingsworth, P.M. (2001) Rapid diversification of a species-rich genus of neotropical rain forest trees. Science, 293, 2242-2245. Sedio, B. (2013) Trait evolution and species coexistence in the hyperdiverse tropical forest tree genus Psychotria. PhD thesis, University of Michigan, USA. Thompson, J.N. (1994) The coevolutionaryprocess. University of Chicago Press. Chicago IL. Thompson, J.N. (1988) Coevolution and alternative hypothesis on insect/plant interactions. Ecology, 69, 893-895. Valencia, R., Condit, R., Foster, R.B., Romoleroux, K., Villa Munoz, G., Svenning, J-C., Magard, E., Bass, M., Losos, E.C. & Balslev, H. (2004) Yasuni Forest Dynamics Plot, Ecuador. Tropical forest diversity and dynamism: findings from a large-scale plot network. (eds. E.C. Losos & E.G. Leigh Jr.). University of Chicago Press, Chicago. pp. 609-628. 13 Walbergh, N. (2001) The phylogenetics and biochemistry of host-plant specialization in Melitaeine butterflies (Lepidoptera: Nymphalidae). Evolution, 55, 522-537. CHAPTER 2 THE RESOURCE AVAILABILITY HYPOTHESIS REVISITED: A META-ANALYSIS Reprinted from Functional Ecology, Vol 25, Issue 2, M.-J. Endara, and P. D. Coley "The resource availability hypothesis revisited: a meta-analysis," copyright © 2011, with permission from John Wiley and Sons. 15 & i Ecological Society Functional Ecology 2010 doi: 10.1111/j. 1365-2435,2010.01803.x EVOLUTIONARY ECOLOGY OF PLANT DEFENCES The resource availability hypothesis revisited: a meta-analysis Mari'a-Jose Endara*1 2 and Phyllis D. Coley1,3 1 Department o f Biology, University of Utah, Salt Lake City, Utah 84112, USA; 2Departamento de Ciencias Bioldgicas, Herbario QCA, Pontificia Universidad Catdlica del Ecuador, Av. 12 de Octubre y Roca, Aptdo. 17-01-2184, Quito, Ecuador; and 3Smithsonian Tropical Research Institute, Box 0843-03092, Balboa, Panama Summary 1. S e v e ra l th e o r ie s h a v e p ro v id e d a f r am ew o rk f o r u n d e r s ta n d in g v a r ia tio n in p la n t d e fe n c e a g a in s t h e rb iv o re s . A m o n g th em , th e p l a n t a p p a r e n c y th e o ry a n d th e r e s o u r c e a v a ila b ility h y p o th e s is (R A H ) h a v e a im e d to e x p la in th e p a t t e r n s o f d e fe n c e in v e s tm e n t a n d th e se le c tiv e p r e s s u re s t h a t h a v e led to th e v a r ie ty o f d e fe n s iv e s tr a te g ie s a c ro s s sp e c ie s. H e r e we p ro v id e a h is ­to r ic a l re v iew o f b o th th e o r ie s , p r e s e n t e v id e n c e t h a t s h a p e d th e ir d e v e lo pm e n t a n d c o n t r a s t th e ir p r e d ic tio n s . 2. W e p r e s e n t th e r e s u lts o f a m e ta - a n a ly s is o f th e u tility o f th e R A H 25 y e a r s a f te r i t w a s p r o ­p o s e d a n d c om p a r e it to a p p a r e n c y th e o ry . W e p e r fo rm e d a m e ta - a n a ly s is o f 50 s tu d ie s t h a t h a v e e x am in e d p l a n t g row th , d e fe n c e s a n d h e rb iv o ry in r e la tio n to r e s o u r c e a v a ila b ility a c ro s s la titu d e a n d o n to g e n y . S p e c ific a lly , we te s te d f o u r p re d ic tio n s t h a t fo llow th e R A H : (i) sp e c ie s a d a p te d to re s o u r c e - r ic h e n v iro nm e n ts h a v e in trin s ic a lly f a s te r g r o w th r a te s th a n sp e c ie s a d a p t e d to r e s o u rc e - p o o r e n v iro nm e n ts ; (ii) f a s t-g row in g sp e c ie s h a v e s h o r te r le a f life tim e s th a n s low -g row in g spe cies; (iii) f a s t-g row in g sp e c ie s h a v e low e r am o u n ts o f c o n s titu tiv e d e fe n c e s th a n s low -g row in g spe cies; a n d (iv ) f a s t-g row in g sp e c ie s s u p p o r t h ig h e r h e rb iv o ry r a te s th a n s low -g row in g spe cies. 3. O u r r e s u lts c o n f irm th e p r e d ic tio n s t h a t sp e c ie s a d a p te d to r e s o u r c e - p o o r e n v iro nm e n ts g row in h e r e n tly m o r e slow ly , in v e s t m o r e in c o n s titu tiv e d e fe n c e s a n d s u p p o r t low e r h e rb iv o ry th a n sp e c ie s f rom m o r e p r o d u c tiv e h a b ita ts . O u r d a t a a lso sh ow e d t h a t v a r i a t io n in g ro w th r a te am o n g sp e c ie s b e tte r e x p la in s th e d if fe re n c e s in h e r b iv o ry th a n v a r ia tio n in a p p a r e n c y , su g g e s t­in g th a t th e e v o lu tio n o f d if fe r e n t d e fe n s iv e s tr a te g ie s a c ro s s sp e c ie s is r e s o u r c e , r a th e r th a n h e r ­b iv o r e d riv e n . W e a ls o f o u n d t h a t th e a p p lic a tio n o f th is th e o ry a p p e a r s r o b u s t a c ro s s la titu d e a n d o n to g e n y , a s th e m a g n itu d e o f th e e ffe c t sizes fo r m o s t o f th e p re d ic tio n s d id n o t v a ry s ig n ifi­c a n tly b e tw e e n e c o s y s tem s o r a c ro s s o n to g e n ic s ta g e s. 4. W e c o n c lu d e th a t th e R A H h a s se rv ed a s a v a lid f ram ew o rk f o r in v e s tig a tin g th e p a t t e r n s o f p l a n t d e fe n c e s a n d th a t its a p p lic a b ility is q u ite g e n e ra l. Key-words: h a b i t a t r e s o u rc e s , h e rb iv o ry , m e ta - a n a ly s is , p l a n t a p p a r e n c y , p l a n t d e fe n c e s , p la n t d e fe n c e th e o ry , p l a n t g row th , r e s o u r c e a v a ila b ility h y p o th e s is Connell 1971). Additionally, the evolutionary trajectory o f b o th plan t an d herbivore tra its is driven by the 'a rm s race' where plants are und e r continual selection to optimize defence investments an d herbivores respond with counter ad ap ta tio n s to detoxify o r avoid the defences (Ehrlich & Raven 1964; T homp so n 1988). In this paper, we provide some historical context for the development o f theories that have aimed to explain the p a tte rn s o f defence investment an d the selective pressures th a t have led to the variety of defensive strategies across species. We focus on two main © 2010 The Authors. Journal compilation © 2010 British Ecological Society Introduction Because p lants an d herbivores constitute over h a lf o f the macroscopic diversity o n earth, their in teractions play a fundamental role in biodiversity a n d ecosystem function. F o r example, the diversity o f plan t species coexisting a t a single site may frequently be shaped by the negative density-an d distance-dependent effects o f herbivores (Janzen 1970; C o rrespondence author. E-mail: [e-mail] Functional Ecology 16 2 M.-J. Endara & P. D. Coley theories, plan t apparency theory (Feeny 1976) an d the resource availability hypothesis (Coley, Bryant & Chapin 1985), present evidence th a t shaped their development and co n tra st their predictions. We then present results from a meta-analysis o f studies examining interspecific v a riation in defence to assess th e utility o f these theories. Although plants were credited with having effective anti­herbivore defences as early as 1888 (Stahl), it was not until Dethier's (1954) and F ra enkel's (1959) papers th a t the signif­icance o f p lan t secondary metabolites was widely appreci­ated. Since then, the details o f myriad defensive traits and the concept o f b ottom-up control o f herbivores have per­meated the literature (Ehrlich & Raven 1964; Whittak e r & Feeny 1971; Levin 1976; Haukioja 1980; L in d ro th & Batzli 1984; Power 1992). While the concept o f p lan t defences was embraced, the puzzle remained as to why the am o u n t and type o f defence differed so much among species. Fun d amen ­tal to explaining interspecific variation in defences is, under­standing the costs and benefits o f defensive traits. The costs o f defence have been extensively studied, and a lthough they have occasionally been difficult to quantify, many examples o f direct, indirect and ccological costs have been docu­mented (Simms 1992; Koricheva 2002; Strauss et at. 2002). The benefits o f reduced herbivory, while n o t universal, have also been shown (Marqiiis 1984; Belsky 1986). Most o f the synthetic theories addressing interspecific differences in defence assume th a t selection has optimized investments, such th a t the benefits outweigh the costs (Feeny 1976; Grime 1977; McKey 1979; Rhoades 1979; Coley, Bryant & Chapin 1985; Crawley 1985). In the next section we focus on two of these theories, apparency theory and the resource availabil­ity hypothesis th a t sought explanations for why species differed in their investment in defences. We sta rt with a historical review o f the theories and the evidence that shaped their development. Apparency theory The first major a ttem p t to identify interspecific p a tte rn s of plant defences and to infer the processes responsible was apparency theory (Feeny 1976). A similar idea was simulta­neously presented by Rhoades & Cates (1976). This theory revolutionized the field, as it shifted the focus from catalogu­ing the array o f defensive traits, to asking why species differed in defences. The theory not only identified patterns of defences b u t suggested th a t the apparency o f species to herbi­vores was the cause. Feeny posited th a t species th a t were long-lived would be apparent o r ‘bound to be found' by both generalist an d specialist herbivores and therefore would be under strong selection for effective defences against both. The high investments in secondary m etabolites, such as tannins in oaks, were consistent with this. T annins were th o u g h t to reduce digestibility o f leaves by binding with proteins, a mechanism o f action th a t would be difficult for herbivores to circumvent. Feeny referred to these types o f defences as 'quantitative' because the greater the investment, the more effective they would be. Furthe rmore, he posited th a t they would present an effective defence against all herbivores, both specialists and generalists. In contrast, he suggested th a t u n ap p a ren t species were short-lived and ephemeral in time and space, an d because o f this unpredictability, it would be difficult for herbivores to specialize on them. Thus, u n ap p a r­ent species could evade specialists and would only need defences th a t were effective against generalists. Using herba­ceous crucifers as an example, he called these ‘qualitative' defences. Qualitative defences o f apparent plants were typi­cally present in low concentrations and were low molecular weight molecules such as sinigrins and alkaloids. They were thought to act on specific animal targets an d present signifi­cant barriers to generalists and non-adapted insects. Although Feeny hypothesized that it would be possible for herbivores to evolve counter ad aptations to qualitative defences, the o p p ortunity for specialization would n o t arise because u n ap p a ren t plants were ephemeral and unreliable food sources. U n apparent plants would therefore escape from specialists and have qualitative defences against general­ists. Apparent plants would have quantitative defences th a t would be effective against both generalists an d specialists. Thus, the apparency o f plants would determine if they were attacked by specialist herbivores o r n o t, and the herbivores in tu rn would determine which defences, quantitative o r qualita­tive., were optimal. Because o f the elegance o f apparency the­ory and the plausible fit with nature, the theory was rapidly accepted an d profoundly shaped the field. It has been cited 1400 times an d established the paradigm against which subse­quent theoretical and empirical work has been judged. Resource availability hypothesis The resource availability hypothesis (RAH), also called the growth ra te hypothesis (Coley 1987; Stamp 2003), accepted Feeny's premise th a t long-lived species (apparent) invested heavily in defences an d short-lived species (unapparent) did not, b u t presented an alternative explanation o f the mecha­nism. Coley, Bryant & Chapin (1985) proposed th a t the observed range o f defence investment was n o t due to differ­ences among species in apparency, but to differences among species in the cost/benefit ra tio o f defences. T hey argued th a t the costs and benefits o f investing in defence depended o n the inherent growth rate o f the species. In a fast-growing species, the o p portunity cost o f investing in defence would be high, as reallocating resources from photosynthetic leaves w ould have a much bigger negative impact on a fast grower compared to a slow grower. However, for fast growers, the negative impact o f losing leaf area would be low. as they could more quickly replace lost leaves and a given am o u n t o f damage would rep­resent a smaller percentage o f their annual growth. F u rth e r­more, the RAH postulated that herbivore pressure was a characteristic o f the envi ronment, rather th an o f a species' ap ­parency, and th a t even if the risk o f herbivory were uniform across species, selection could favour different levels o f defence in species with different inherent growth rates. This is because the inherent growth ra te determines the o p portunity cost o f defence and the impact o f herbivory. © 2010 The Authors. Journal compilation © 2010 British Ecological Society, Functional Ecology 17 18 4 M .-J . Endara & P. D. Coley inducing production o f defences (Karban & Myers 1989; Ka rb an 2011). Another strategy is to tolerate herbivore d am ­age by storing sufficient resources to allow regrowth (Strauss & Agrawal 1999). We d o n o t discuss these ideas, as the goal o f this paper is to review theories whose m ain objective was to u nderstand interspecific differences in constitutive defences. However, it is worth noting th a t there is much confusion in the literature regarding the predictions o f some o f the above mentioned theories an d the circumstances und e r which they are applicable (Stamp 2003). In o u r literature review, it was common to find studies claiming they supported the RAH when they did not, an d others refuting the theory when their results were in agreement. The RAH was also frequently invoked when comparing phenotypic responses o f plants to different environments, even though the RAH explicitly refers to optimal levels o f defence th a t have evolved in species adapted to different environments. Although phenotypic plasticity theoretically could m irro r adapta tio n s seen across species, they often do not. but instead seem to reflect imbal­ances in allocation. The c arbon-nutrient balance hypothesis (CNB), which was designed to explain these phenotypic shifts in defences, does not assume optimality and therefore makes different predictions th an the RAH (Bryant, Chapin & Klein 1983). Conversely, the growth-differentiation balance hypothesis (GDBH), as elaborated by Herms & Ma ttson (1992), does assume th a t phenotypic variation in secondary metabolism represents adaptive plasticity consistent with p re ­dictions o f optimal defence theory (see also Glynn et at. 2007). F urthe rmore, we found th a t the CNB hypothesis was often misused to explain interspecific differences, as did Stamp (2003) for GDBH. Bryant, Chapin & Klein (1983) and Herms & Ma ttson (1992) addressed bo th phenotypic and evo­lutionary responses o f plants to resource availability, which no d o u b t has contributed to this confusion. Contrasting plant apparency and RAH Because the theories o f resource availability and apparency are not mutually exclusive and in some cases make similar predictions, in the next section o f the paper we examine the generality and utility o f the RAH 25 years after it was p ro ­posed and, compare it to apparency theory. We examined interspecific patterns o f growth, defence and herbivory by means o f meta-analyses based on 50 studies published between 1985 and 2010 and conducted o n > 600 different plant species (see references o f studies included in Appen­dix S I, Supporting information). Specifically, we performed separate meta-analyses for each o f the four predictions from the RAH: (i) species adapted to resource-rich environments have intrinsically faster growth rates than species adapted to resource-poor environments; (ii) fast-growing species have shorter leaf lifetimes than slow-growing species; (iii) fast-growing species have lower amounts o f constitutive defences than slow-growing species; and (iv) fast-growing species su p ­po rt higher herbivory rates th an slow-growing species. We examine these predictions across latitude an d ontogeny. We selected for relevant studies using Web o f Knowledge, Google Scholar an d Web o f Science, searching for the terms ‘p lan t' and ‘herbiv*' and ‘defens*' and ‘resource*' (or Might' or ‘nutrient*') and ‘growth'. Other relevant studies were found by searching the reference section in the articles retrieved from the te rm searches. We restricted o u r analyses to studies th a t examined interspecific differences in plant species within a site o r between sites differing in their degree o f resource availability. Thus, studies that compared growth, defences or herbivory in the same p lan t species in different resource envi­ronments were not considered. F o r a complete description o f our inclusion criteria see methods in Appendix S2 (Support­ing information). Fo r the last two meta-analyses (predictions 3 and 4 from the RAH), the articles were grouped into two types: studies th a t compared investment in p lan t defences, herbivory and growth between two o r more different species within a site, and those th a t compared two o r more different species growing in sites with divergent resource levels (light an d nutrients, see Appendix S3, Supporting information for fu rth e r categorization o f studies included in these meta­analyses). In the original articles, the h abitats in which the studies were conducted were usually classified as either resource-poor environments o r resource-rich environments based on the levels o f nutrient availability o r o f light availabil­ity. All the meta-analyses were conducted with the program Me ta Win version 2,1,5 (Rosenberg, Adams & Gurevitch 2000), and using th e mixed effects model (Gurevitch & Hedges 1993; see Appendix S2, Supporting information for a complete description o f materials an d methods and Appen­dix S3, Supporting information for effect sizes). The results th a t follow are based primarily on studies o f woody terrestrial species. Although we did not specifically exclude studies o f herbaceous plants, o r o f marine an d aq u a ­tic plants, mo st o f these studies evaluated intraspecific differ­ences in growth, defences and herbivory, and as such, did n o t meet o u r inclusion criteria (see Appendix S2, Supporting information). P RE DICTION 1: SPECI ES ADAPTED TO RESOURCE-RICH ENVIRONMENTS HAVE INTRINSICALLY FASTER GROWTH RATES THAN SPECIES ADAPTED TO RESOURCE-POOR ENVIRONMENTS Overall, we found th a t species from resource-rich environ­ments grew faster th an those from resource-poor environ­ments (d = 2-75, 95% C l = 1 01-4 85, n = 24, nfs = 232; Fig. 2; Table 1 in Appendix S3, Supporting information). We did n o t find significant varia tio n among studies conducted in tropical forests vs. temperate forests (Qg = 0-6, d.f. = 1, P = 0-59), n o r among ontogenic stages (£?B = 707, d.f. = 2, P = 0'33). However, we found th a t the magnitude o f the effect was significantly different among the different growth traits ( g B - 16-35, P - 0-04). The lower variance was found among those studies th a t reported growth ra te and height. When only these studies were analysed, the results Results © 2010 The Authors, journal compilation © 2010 British Ecological Society, Functional Ecology 19 Revisiting the resource availability hypothesis 5 f ig. 2. Mean and 95% confidence intervals for the effect sizes o f resources on plant growth measures (weighted standardized mean, Hedges' d): for all studies (n = 24) and for studies conducted only in tropical for­ests (n = 17), in temperate forests (n = 6), with seedlings (n = 9), with saplings in = 5), and with adults (n = 8). it LU 7 j 6 ­5 - ­4 - 3 - ­2 ­1 - ­0 - See' ,W3S were similar to those obtained from the whole d a ta set (d = 1-22, 95% C l = 0-17-2-14, n = 17,n fs = 36-6). PREDI CT ION 2: FAST-GROWING SPECI ES HAVE SHORTER LEAF LI F ET IM ES TH AN SLOW-GROWING SPECIES We found a strong and negative effect o f growth rate on leaf lifetime (z - -1-78, 95% Cl - -2-55 to -1 0 6 , n - 10, KfS = 110; Fig. 3; Table 2 in Appendix S3, Supporting infor­mation), confirming the prediction th a t slow-growing species have longer leaf lifetimes than fast-growing species. We were unable to compare the magnitude o f the effect between habitats with different resources, different ecosystems or ontogenic stages because most studies included in o u r meta-analysis were conducted with species in the same site, in tropical forests and with adult individuals (Appendix S2, Supporting information). PREDI CT ION 3: FAST-G ROWI NG SPECI ES HAVE LOWER INVESTMENTS IN CONST ITUTI VE DEFENCES THAN SLOW-GROWING SPECI ES To test this prediction we conducted two analyses. We analysed growth effects on defences for fast- and slow-growing species in the same habitat to control, a t least in part, for differences in the expression o f defences across species caused by varied environments. We also compared growth effects between two o r more different species in different habitats, which would include plasticity as well as evolved differences for constitutive defences (see me th ­ods in Appendix S2, Supporting information). The studies only reported quantitative results for phenolic compounds and terpenes, but not for other classes o f chemical defences (see Table 3 in Appendix S3, Supporting infor­mation). Effect of plant growth on investment in plant defences between species within a site As predicted by the RAH, when all types o f defences were considered together (chemical and mechanical), fast-growing species invested less in constitutive defences th an slow-grow­ing species (z = -0-52, 95% C l = -0-66 to -0-38, n = 57, fifs = 1824-6; Fig, 3), This effect was more pronounced for seedlings and saplings than for adults as differences between seedlings and saplings (-0 61 vs. -0*59) were virtually n o n ­existent (-0-18; Qb - 9-36, d.f. = 2, P = 0 04), and also for studies conducted in tropical (-0-62) vs. temperate (-0-3) for­ests (Qb - 6-44, d.f. = 2, P = 0-04). The result was also sig­nificant when considering only chemical defences (z = -0-3, 95% Cl = -0-45 to -0 1 2 , n - 23,n ts = 139 6; Fig. 3). The same p a tte rn was maintained for the effect o f growth on investment in terpenes (z = -0-43, 95% C l = - 0 7 to -0-23, n = 4, Hfs = 105-7; Fig. 3) an d to ta l phenolics and tannins, although it was no t significant for the last two. We also found th a t fast-growing species invested less in mechanical defences (z = -0-85, 95% C l = -1-2 to -0-59, n = 25, nk = 584-2; Fig. 3), an d the magnitude o f the effect was significantly higher in seedlings (-1-67) vs. saplings (-0-64) an d adults (-0-33; Qb = 26-29, d.f. = 2, P = 0 004). O u r meta-analy­sis also confirmed th a t fast-growing species h ad higher leaf Fiy. 3. Mean effect sizes (z) and 95% confi­dence intervals for growth rate cffccts on investment in plant dclfcnccs and herbivory between species within a site. The dependent variables include: leaf lifetime (n - 10), all constitutive defences (n = 57), all chemical defences (n = 23), total phenolics (it ~ 6), tannins (hydrolysable and condensed; n - 12), terpenes (n - 4 records from one study by Fine, Mesones & Coley 2004), mechanical defences (n = 6), nutrient con­t e n t ^ = 25) and herbivory (n = 16). N .C 1 -- % 0 5-- O 0 - - CD -0-5-- o -1 -- (A O -1 -5 - N - 2 - t> -2 5-- £LU S.® A©' nr 2010 The Authors. Journal compilation © 2010 British Ecological Socicty, Functional Ecology 20 6 M.-J. Endara & P. D. Coley nutrient content (z = 0-51, 95% C l = 0-33-0-72, n = 6, Hfs = 42-5; Fig. 3). Effect of plant growth on investment in defences between species in habitats with different nutrient availability Our meta-analyses showed th a t studies comparing invest­ment in defences between two o r more different species from habitats with different levels o f nutrients had contradictory results compared to within site comparisons. In these studies, there were no differences between fast-growing an d slow-growing species in defence investment when all defences were combined (chemical an d mechanical). We found the same result when all chemical defences were combined in to one response variable (total phenolics, hydrolysable tannins and condensed tannins), o r individually for tannins an d leaf toughness (Fig. 4). There were n o t enough studies to compare trichomes. However, fast-growing species invested more in total phenolics I: = 0-85, 95% Cl = 0-04-1-41, n = 8, nfs = 0; Fig. 4). We did not find differences in the effect of growth on defences between studies conducted in tropical vs. temperate forests, n o r among studies performed with seed­lings o r saplings or adult individuals. Since only one o f our selected studies compared nutrient content in leaves between species from habitats with different nutrient levels we did not conduct a m eta-analysis for this tra it (Appendix S I , S upport­ing information). Effect of plant growth on investment in defences between species in habitats with different light availability Although in general slow-growing species invested more in defences than fast-growing species, this difference was not sig­nificant when comparing different specics from h abitats with different levels o f light (95% C l = -0-35 to 0-64; Fig. 5). Individually, we found a significant negative effect o f growth on mechanical defences (only leaf toughness, as there were not enough studies comparing p roduction o f trichomes) (z = -0-82, 95% C l = -1-42 to -0-13, /; = 8, nh = 18-8) b u t n o t on other defences. There were no differences in stud­ies conducted in different ecosystems an d w ith different on to ­genetic stages. PREDICTION 4: FAST-GROWING SPECI ES SUPPORT HIGHER HERB IVORY RATES THAN SLOW-GROWING S PEC IE S When comparing different species within the same site, we found th a t fast-growing species suffered higher herbivory compared to slow-growing species (z = 0-35, 95% Cl = 0-15-0-55, n = 16, nfs = 27-4; Fig. 3; Table 4 in Appendix S3, Supporting information). In contrast, the com­parison across sites includes n o t only differences among spe­cies in their growth rate, b u t also differences among sites in overall herbivore pressure. In this comparison, herbivory for fast growers in resource-rich sites was higher th an for slow growers at resource-poor sites, but the effect was not signifi­cantly different from zero (z = 0-29, 95% C l = -0-18 to 0*68, n = 2 9 ,« fs = 0). Discussion The goal o f both apparency theory and the RAH has been to provide a theoretical framework th a t adequately explains the interspecific variation in plan t defensive strategies. The RAH relates the evolution o f defences to interspecific differences in 2 j sz% 1-5-- o CD 1 -- H-o (fl 05 N i --------------------I- iO o o t/5 -0-5-- it -1 -- LU -1-5 -L <^evv Fig. 4. Mean effect sizes (z) and 95% confi­dence intervals for growth rate effects on investment in plant defences between species in habitats with different nutrient availability. Dependent variables include: all defences (n = 24), chemical defences (n = 18), total phenolics (n = 8), tannins (n = 10) and mechanical defences (n = 6). 1-5 1 0-5 0 -0-5 - ­- 1 -1-5 -2 -2-5 d ‘°eS c ^ ' ca >\\cs ,j\eo^'ca' Fig. 5. Mean effect sizes (z) and 95% confi­dence intervals for growth rate effects on investment in plant defences between species in habitats with different light availability. Dependent variables include: all defences (n = 17), chemical defences (n = 9), total phenolics (n = 4), tannins (n = 5), mechani­cal defences (/? = 8) and nutrient content (n = 2). © 2010 The Authors. Journal compilation © 2010 British Ecological Socicty, Functional Ecology 21 Revisiting the resource availability hypothesis 7 inherent growth rate, w hereas apparency theory assumes th a t defences are related to a species' predictability to herbivores. Although the theories have different assumptions regarding the reasons leading to defence differences, some o f the predic­tions are similar. For example, both theories agree th a t long-lived, slow-growing species (apparent species) should invest more in defences th an short-lived, fast-growing species (unap-parent species). However, a fundamental difference between the theories is their contrasting predictions for the amount of herbivory. The RAH predicts th a t fast-growing species should suffer greater herbivore damage, while apparency the­ory predicts similar losses for apparent and unapp a ren t spe­cies, In the discussion th a t follows, vvc examine results for defence and herbivory, as these apply to both theories. We also examine two predictions th a t apply only to the RAH, th a t resources affect growth and th a t growth affects leaf life­times. PREDI CT ION 1: RESOURCE EFFECTS ON PLANT GROWTH RATE Our meta-analysis suggests th a t, in agreement with the RAH, plan t species from resource-rich environments had higher growth rates th an species from resource-poor environments (Fig. 2). These patterns hold across different ecosystems and ontogenic stages, as we did n o t find significant differences between studies conducted in tropical forests vs. temperate forests an d in seedlings, saplings and adults. It is less certain whether these patterns will also hold for herbaceous species since all the studies included in our meta-analysis were based on woody species. However, a similar association between resources an d inherent growth was found in a m eta-analytical study performed with temperate herbs (Taub 2007), Our results are consistent with the well-established fact th a t spe­cies growth rates vary with fertility levels (Grime 2001) and light requirements (Swaine & Whitmore 1988). High rates of growth are hallmark characteristics o f p lan t species adapted to high-resource environments (Grime 1979; Chapin 1980; Lambers & P o o rte r 1992). In contrast, species adapted to low-resource environments grow slowly and retain their growth habit even under high-resource conditions (Grime 2001). PREDI CT ION 2: GROWTH RATE EFFECTS ON LEAF LIFETIME As predicted, slow-growing species have leaves with signifi­cantly longer leaf lifetimes th an fast-growing species. Long-lived leaves minimize nutrient losses (Aerts 1995) and constitute an essential ad aptation o f slow-growing species to habitats with low-resource availability (Grime 1977). The relationship between growth rate and leaf life span was the foundation for suggesting th a t qualitative defences, because o f a higher maintenance cost, would be favoured in leaves with short life spans, and quantitative defences, with high ini­tial costs but low maintenance costs, would be favoured in leaves with long life spans (Coley 1987). PREDI CT ION 3: GROWTH RATE EFFECTS ON DEFENCES Both theories predicted greater investment in defence for slow-growing species, b u t for different reasons. The RAH predicts th a t for slow-growing species the opportunity cost o f defence will be low and the negative impact o f herbivory high. Therefore, slow growers should exhibit higher investments in constitutive defences (Coley 1987). Apparency theory pre­dicted th a t apparent plants would need effective defences against bo th specialists and generalists. The results from our meta-analysis found th a t, when considering only the studies th a t compared defence investment across species in the same habitat, there was a significant negative effect o f growth rate on overall defence investment. This result was also main­tained when considering chemical an d mechanical defences independently (Fig. 3). Moreover, this p a tte rn appears robust, as the direction o f the growth effect o n defences was the same when comparing different latitudes and ontogenetic stages. Although defences were universally higher in slow growers, o u r mcta-analysis showed th a t defence differences between fast and slow growers were significantly greater in tropical ecosystems. Possible explanations for this pattern might lie in the fact th a t, in the tropics, there is a higher absolute invest­ment in defences (Coley & Aide 1991), a higher variance in defensive compounds (Gauld & Gaston 1994), and a greater range o f plan t growth rates (Van Z an d t 2007). Greater amounts and ranges could facilitate detection o f differences. Similarly, there was a negative effect o f growth on overall defences for all ontogenetic stages, but the magnitude o f this effect was significantly higher for seedlings. The reason for this is unclear, however, again, it may be easier to detect dif­ferences in defences if seedlings invest more than other age classes because o f the potentially devastating effects o f her­bivory (Barton & Koricheva 2010; b u t see Boege & Marquis 2005). However, when analysing the studies comparing two or more different species from different sites, we did n o t find a significant elTect o f growth rate on overall defences. This was consistent whether h abitats differed with respect to nutrients or light. We interpret this as resulting from a combination o f phenotypic responses o f plants to short-term changes in resources with selection for different defence strategies in dif­ferent habitats. Thus, these results can be better explained by integrating both the RAH and the carbon-nutrient balance hypothesis (CNB; Bryant, Chapin & Klein 1983; Dyer & Co­ley 2002; Stamp 2003). The CNB hypothesis suggests that when resources are in excess o f wh a t can be used for growth, they will be invested in defences. Accordingly, under high light where carbon is in excess relative to nutrients, this theory predicts higher amounts o f carbon-based defences, whereas the RAH predicts lower defences for species adapted to this low-resource condition. Because o f these counterbalancing influences, we would expect no significant effect o f plant growth on defences, and this is what we found in our meta­analysis for studies comparing species from sites with differ­ent levels o f light. In another study, Baldwin & Schultz (1988) © 2010 The Authors. Journal compilation © 2010 British Ecological Society, Functional Ecology 22 8 M .-J . Endara & P. D. Coley also found n o significant differences in phenol investment when comparing species o f the genus Piper from gaps and un­derstorey. F o r mechanical defences, the CNB theory does n o t have a prediction, while the RAH predicts lower mechanical defences for species adapted to high-light levels. Again, our results were consistent with this, as leaves o f slow-growing species were significantly tougher. In contrast to the defence comparisons across light gradi­ents, which were consistent with the combined effecls o f RAH and CNB theories, results from habitats with different nutrient levels were confusing. Under high nutrient levels, b o th the CNB hypothesis and the RAH predict lower carbon-based defences, however, we found a non-significant opposite trend. An o th e r meta-analysis (Koricheva 1998) also found a weak but negative effect o f fertilization o n carbon-based defences. Although Herms & Ma ttso n (1992) proposed a model that integrates genetic and phenotypic plasticity, the predictions are nonlinear an d complex, making it difficult or impossible to capture secondary metabolic responses to varia­tion in resource availability (Stamp 2003). Thus, when com­parisons are made within a site, there is a clear negative relationship between plant growth and defence following the RAH, however, when confounding effects o f environmental plasticity are included (Figs 4 an d 5), particularly those asso­ciated with nutrient gradients, it is obvious th a t o u r under­standing is incomplete. The RAH also predicts higher inducible defences in fast-growing species. This is because the opportunity cost o f defence is higher for fast growers, and because fast growers may more often occur under conditions th a t favour induc­tion, such as predictable, but periodic herbivore atta ck (ICar-ban 2010). Although we did n o t analyse this prediction, supporting evidence has been found. In a literature review of 68 studies, Nykanen & Koricheva (2004) found th a t the p ro ­duction o f phenolics and protein-precipitation capacity o f tannins increased in fast-growing species after herbivore dam­age more than in slow-growing species. Van Z an d t (2007) found a similar result in an experimental study with nine spe­cies o f temperate herbaceous plants. A p attern first identified by Feeny was th a t un ap p a ren t plants invested in qualitative defences and ap p a ren t plants in quanti­tative defences. Although this observation has been fairly well supported, the reasons why are still unclear. A quantitative review o f defensive classes in different plant guilds co rrobo­rated this idea by finding that fasl-growing plants (apparent plants) are most often defended with quantitative, dose-dependent defences and slow-growing plants (unapparent plants) with qualitative defences (T. Massad & L. Dyer, pers. comm.). Feeny suggested it was because quantitative defences worked against all herbivores, while qualitative defences worked only against generalists and non-adapted specialists. However, the fact th a t the herbivores attacking apparent and u n apparent plants are a similar mix o f specialists and general­ists (Futuyma & G ould 1979), an d th a t quantitative an d qual­itative defences do n o t have different effects on generalists vs. specialists herbivores (Smilanich 2008; Carmona, Lajeunesse & Johnson 2011; T. Massad & L. Dyer, pers. comm.) brings this explanation into question. Another criticism o f apparen­cy theory, though one th a t we do not view as a fatal attack, is th a t the primary function o f tannins may n o t be to bind p ro ­teins and reduce digestion (Bernays 1981). Instead, more recent evidence shows th a t oxidation o f hydrolysable tannins forms reactive oxygen species, which can overwhelm the an ti­oxidant defences o f herbivorous insects and damage midgut tissues (Martin, Ma rtin & Bernays 1987; Appel 1993; Summers & Felton 1994; J. Salminen & M. Karonen, u n p u b ­lished). Nonetheless, this could be considered a quantitative defence as higher concentrations o f hydrolysable tannins will lead to greater levels o f oxidative stress. The RA H proposed th a t leaf lifetime, which is related to plant growth rates, is the key factor directing selection for the type o f defence. They argued th a t qualitative defences, in addition to being present in low concentrations, are low molecular weight molecules with high turnover or mainte­nance rates. In contrast, quantitative defences such as con­densed tannins, would require a considerable initial investment since they are present a t high concentrations, but because they do not turnover, there would be no subsequent maintenance costs. Thus, for species with short-leaf lifetimes, it would be more cost effective to invest in qualitative com­pounds, whereas for long-lived leaves, the cumulative cost would be lower for quantitative compounds. However, this argument rests o n differences in turnover rates for qualitative and quantitative compounds, an assumption th a t also has been challenged (Mihaliak, Gershenzon & Croteau 1991; Baldwin & Ohnmeiss 1994; van Dam et al. 1995; Salminen & Karonen 2011). Thus, the underlying factors favouring com­pounds along the quantitative/qualitative continuum remain to be determined. PREDI CT ION 4: GROWTH RATE EFFECTS ON HERBIVORY One o f the key differences between the RAH and apparency theory is related to the predicted herbivore damage. Apparen­cy theory (Feeny 1976; Rhoades & Cates 1976) predicts similar rates o f damage. Un ap p a ren t plants escape from specialists and have secondary metabolites th a t are effective against generalists, whereas, apparent plants have m e tab o ­lites th a t are effective against bo th specialists and generalists. In contrast, RAH predicts th a t fast-growing species will sup­p o rt higher levels o f herbivory than slow-growing species because they are less defended. Our results support the last prediction, since we found a negative and significant effect size o f growth rate on herbivory when analysing studies com­paring species with different growth rates within the same h abitat. Thus, u napparent plant species (fast-growing species according to the RAH) did n o t escape from herbivory, but had significantly higher levels th an app a ren t species (slow-growing species according to the RAH). Therefore, variation in growth rate among species explains better the differences in PLANT DEFENCES: UNANSWERED QUESTIONS © 2010 The Authors, Journal compilation © 2010 British Ecological Socicty, Functional Ecology 23 Revisiting the resource availability hypothesis 9 herbivory th an variation in apparency. We found similar trends in the meta-analyses for studies comparing herbivory and growth ra te between species growing in sites with differ­ent level o f resources. This comparison n o t only takes into account differences in growth rates, b u t also differences between sites in overall herbivore pressure. A negative effect size o f growth suggested th a t fast-growing species from resource-rich habitats suffered higher herbivory th an slow-growing species from resource-poor habitats. However, the greater variance and absence o f significance is consistent with herbivore pressure varying among habitats. In addition to high herbivory on unapparent, fast-growing species, there is no evidence th a t they are attacked more by specialists than ephemeral species (Futuyma & Gould 1979; Cates 1980; Basset 1992), a key element o f apparency theory posited to drive selection for different defence strategies. The host-finding abilities o f insect herbivores are sufficiently good th a t escape from discovery does n o t appear to occur, except perhaps for extremely ephemeral species o r tissues. Thus the patterns o f defence first described by Feeny may not be ade­quately explained by a plan t's apparency, as this does n o t lead to differential atta ck by specialist vs. generalist herbivores. Conclusions Both apparency theory and the RA H have provided testable hypotheses for investigating interspecific v ariation in patterns o f p lan t defences and have stimulated a multitude o f studies. Both have been extremely influential and are widely cited (1400 and 1600 citations respectively). Our evaluation o f the generality o f the RAH 25 years afte r its first publication shows strong support for the basic tenets linking resources, p lan t growth, defence and herbivory. I t has been suggested th a t the predictive power o f the RAH is mostly supported in tropical forests, with mixed support in temperate forests (Van Z andt 2007). Although wc found a higher mean effect size for all our predictions in the tropics, this difference was signifi­can t for only one o f the predictions. Therefore, we suggest th a t the applicability o f the RAH is general. In addition, because o f its simplicity an d wide application, the RAH has provided a coherent framework for the generation o f new ideas about plan t - insect interactions, For example, it has been proposed th a t resource availability and enemy release may interact in plan t invasions (Blumenthal 2006). More recent approaches in understanding the origin and maintenance o f plant defences are often framed in an explicit phylogenetic context. Other approaches o f promise ask mech­anistic questions regarding the macroevolutionary trends in plant defences, and how selection by herbivores could influ­ence bo th the speed and direction o f selection. Furthermore, how could these interactions be shaped across species ranges and depend on the mosaic o f oth e r interacting species? And finally, can plant - herbivore interactions promote plant diversity by promoting rates o f speciation or slowing extinc­tion? New phylogenetic and molecular techniques as well as new theoretical approaches in studying plant - herbivore interactions should further enhance o u r understanding o f these fundamentally imp o rtan t interactions across evolution­ary and ecological time-scales. We are grateful to Thomas Kursar and Fred Adler for suggestions and com­ments during the preparation o f this manuscript and to Daniel Herms and Julia Koricheva for thoughtful reviews. P.D.C. acknowledges NSF DEB-0640630. Aerts. R. (1995) The advantages o f being evergreen. Trends in Ecology and Evo­lution, 10,302-407. Agrawal, A.A. & Fishbein, M. (2008) Phylogenetic escalation and decline of plant defense strategies. Proceedings o f the National Academy o f Sciences, USA, 105.10057-10060. Appel, H.M, (1993) Phenolics in ecological interactions: the importance o f oxi­dation. Journal o f Chemical Ecology, 19, 1521-1552. Baldwin, IfT. & Ohnmeiss, T.E. (1994) Swords into plowshares? Nicotiana syl-vestris does not use nicotine as a nitrogen source under nitrogen-limited growth. Oecologia, 98, 385-392. Baldwin, FT. & Schultz, J.C. (1988) Phylogeny and the patterns of leaf pheno­lics in gap-adapted and forest-adapted Piper and Miconia understory shrubs. Oecologia,1 5 .105-109. Barton, K.E, & Koricheva, J. (2010) The ontogeny o f plant defense and herbiv­ory: characterizing general patterns using meta-analysis. The American Nat­uralist, 174.481-493. Basset, Y. (1992) Host specificity o f arboreal and free-living insect herbivores in rain forests. Biological Journal o f the Linnean Society, 47, 115-133. Belsky, A.J. (1986) Does herbivory benefit plants? A review of the evidence. American Naturalist, 127,870-892. Bernays. E.A. (1981) Plant tannins and insect herbivores: an appraisal. Ecologi­cal Entomology, 6, 353-360. Blumenthal, D.M. (2006) Interactions between resource availability and enemy release in plant invasion. Ecology Letters, 9 ,887-895. Boege, K. & Marquis, R.J. (2005) Facing herbivory as you grow up: the ontogeny o f resistance in plants. Trends in Ecology and Evolution, 20, 441-448. Bryant, J.P., Chapin, F.S. Ill & Klein, D.R. (1983) Carbon/nutrient balance of boreal plants in relation to vertebrate herbivory. Oikos, 40,357-368. Bryant, J.P. & Kuropat, P.J. (1980) Selection of winter forage by subartic browsing vertebrates: the role o f plant chemistry. Annual Review o f Ecology and Systematics, 11,261-285. Carmona, D., Lajeunesse, M. & Johnson, M. (2011) Plant traits that predict resistance to herbivores. Functional Ecology, in press. 001:10.111 l/j.1365- 2435.2010.01794.x. Cates, R.G. (1980) Feeding patterns of monophagous, oligophagous, and polyphagous insect herbivores: the effect of resource abundance and plant chemistry. Oecologia, 46,22-31. Chapin, F.S. Ill (1980) The mineral nutrition of wild plants. Annual Review o f Ecology and Systematics, 11, 233-260. Coley, P.D. (1983) Herbivory and defensive characteristics o f tree species in a lowland tropical forest. Ecological Monographs, 53,209-233. Coley, P.D. (1987) Interspecific variation in plant anti-herbivore properties: the role of habitat quality and rate of disturbance. New Phytologisi, 106. 251­263. Coley, P.D. & Aide, T.M. (1991) Comparison of herbivory and plant defenses in temperate and tropical broad-leaved forests. Plant-Animal Interactions: Evolutionary Ecology in Tropical and Temperate Regions {eds P.W. Price, T.M. Lewinsohn, G.W. Fernandes & W.W. Benson), pp. 25-49. Wiley & Sons, NY. Coley, P.D., Bryant, J.P. & Chapin, F.S. Ill (1985) Resource availability and plant antiherbivore defense. Science, 230,895-899. Connell, J.H. (1971) On the role of natural enemies in preventing competitive exclusion in some marine animals and rain forest trees. Dynamics o f Numbers in Populations (eds P.J. den Boer & G.R. Gradwell), pp. 298-312. Proceed­ings of the advanced Study Institute, Osterbeek, 1970. Centre for Agricul­tural Publication and Documentation, Wageningen. Crawley, M.J. (1985) Reduction o f oak fecundity by low-density herbivore pop­ulations. Nature, 314, 163-164. van Dam, N.M., Witte, L„ Theuring, C. & Hartmann, T. (1995) Distribution, biosynthesis and turnover of pyrrolizidine alkaloids in Cynoglossum offici­nale. Phytochemistry, 39,287-292. Acknowledgements References © 2010 The Authors. Journal compilation © 2010 British Ecological Society, Functional Ecology 24 Dethier, V. (1954) Evolution o f feeding preferences in phytophagous insects. Evolution, 8 ,32-54. Dyer, L.A. & Coley, P.D. (2002) Tritrophic interactions in tropical and temper­ate communities. Multitrophic Level Interactions (eds T. Tscharntke & B. Hawkins), pp, 67-88. Cambridge University Press, Cambridge. Ehrlich, P. & Raven, P. (1964) Butterflies and plants: a study in plant coevolu­tion. Evolution, 18, 586-608. Farrell, B.D., Dussourd, D.E. & Mitter, C. (1991) Escalation o f plant defense: do latex and resin canals spur plant diversification? American Naturalist, 128,881-900. Feeny. P. (1976) Plant apparency and chemical defense. Recent Advances in Phytochemistry (eds J.W. Wallace & R.L. Mansell), pp. 1-40. Plenum Press, New York. Fine, P.V.A., Mesones, I. & Coley, P.D. (2004) Herbivores promote habitat specialization by trees in Amazonian forests. Science, 305,663-665. Fraenkel, G.S. (1959) The raison d'etre o f secondary plant substances. Science, 129,1466-1470. Futuyma, D.J. & Gould, F. (1979) Associations of plants and insects in a decid­uous forest. Ecological Monographs, 49. 33-50. Gauld, ID . & Gaston, J. (1994) The taste of enemy-free space: parasitoids and nasty hosts. Parasitoid Community Ecology (eds B.A. Hawkings & W. Shee­han), pp. 279-299. Oxford University press, New York. Glynn, C., Herms, D.A., Orians, C.M., Hansen, R.C. & Larsson, S. (2007) Testing the growth-differentiation balance hypothesis: dynamic responses of willows to nutrient availability. New Phyiologist, 176,623-634. Grime, J.P. (1977) Evidence for the cxistencc of three primary strategics in plants and its relevance to ecological and evolutionary theory. American Naturalist, 111,1169-1194. Grime, J.P. (1979) Plant Strategies and Vegetation Processes. John Wiley & Sons, New York. Grime, J.P. (2001) Plant Strategies, Vegetation Processes, and Ecosystem Prop­erties, 2nd edn. John Wiley & Sons, New York. Gurevitch, J. & Hedges, L.V. (1993) Meta-analysis: combining the results of independent experiments. Design and Analysis o f Ecological Experiments (eds S.M. Schemer & J. Gurevitch), pp. 378-425. Chapman & Hall, New York. Haukioja, E. (1980) On the role o f plant defenses in the fluctuation of herbivore populations. Oikos, 35.202-213. Herms, D.A. & Mattson, W.J. (1992) The dilemma of plants: to grow or defend. Quarterly Review o f Biology, 67,283-335. Janzen, D.H. (1970) Herbivores and the number of tree species in tropical for­ests. American Naturalist, 104, 501-528. Janzen, D.H. (1974) Tropical blackwater rivers, animals and mast fruiting by the Dipterocarpaceae. Biotropica, 6,69-103. Johnson, M.T., Smith, S.D. & Rausher, M.D. (2009) Plant sex and the evolu­tion of plant defenses against herbivores. Proceedings o f the National Acad­emy o f Sciences, USA, 105,18079-18084. Karban, R. (2011) The ecology and evolution of induced resistance against herbivores. Functional Ecology, in press. DOI: 10.1111/j. 1365-2435.2010. 01789.x. Karban, R. & Myers, J.H. (1989) Induced plant responses to herbivory. Annual Review o f Ecology andSystematics, 20,331-348. Korichcva, J. (1998) Regulation of woody plant secondary metabolism by resource availability: hypothesis testing by means o f meta-analysis. Oikos, 83,212-226. Koricheva, J. (2002) Meta-analysis o f sources of variation in fitness costs of plant antiherbivore defenses. Ecology, 83, 176-190. Lambers, H. & Poorter, H. (1992) Inherent variation in growth rate between plants: a search for physiological causes and ecological consequences. Advances in Ecological Research, 2 3 ,187-261. Levin, D.A. (1975) Pest pressure and recombination systems in plants. Ameri­can Naturalist, 109,437-451. Levin, D, (1976) The chemical defenses o f plants to pathogens and herbivores. Annual review o f Ecology and Systematics, 7 ,1 21-159. Lindroth, R. & Batzli, G. (1984) Plant phenolics as chcmical defenses: cffects of natural phenolics on survival and growth of prairie voles (Microtus ochro-gaster). Journal o f Chemical Ecology, 10,229-244. Marquis, R. (1984) Leaf herbivores decrease fitness of a tropical plant. Science, 226, 537-539. Martin, J.S., Martin, M.M. & Bemays, E.A. (1987) Failure o f tannin acid to inhibit digestion or reduce digestibility on plant protein in gut fluids o f insect herbivores. Journal o f Chemical Ecology, 13,605-621. Mattson, W.J. (1980) Herbivory in relation to plant, nitrogen content. Annual Review o f Ecology and Systematics, 11,119-161. McKey, D, (1979) The distribution o f secondary compounds within plants. Herbivores: Their Interactions with Secondary Plant Metabolites (eds 10 M.-J. Endara & P. D. Coley G.A. Rosenthal & D.H. Janzen), pp. 55-133. Academic Press, New York. Mihaliak, C.A., Gcrshcnzon, J. & Croteau, R. (1991) Lack of spatial monoter-pene turnover in root plants: implications for theories of plant-chemical defense. Oecologia, 87,373-376. Nykanen, H. & Koricheva, J. (2004) Damage-induced changes in woody plants and their effects on insect herbivore performance: a meta-analysis. Oikos, 104, 247-268. Power, M. (1992) Top-down and bottom-up forces in food webs: do plants have primacy? Ecology, 7 3 ,733-746. Rhoades, D.F. (1979) Evolution o f plant chemical defense against herbi­vores. Herbivores: Their Interaction with Secondary Plant Metabolites (eds G.A. Rosenthal & D.H. Janzen), pp. 1-55. Academic Press, New York. Rhoades, D.F. & Cates, R.G. (1976) Toward a general theory of plant antiher­bivore chemistry. Recent Advances in Phytochemistry (eds J.W. Wallace & R.L. Mansell), pp. 168-213. Plenum Press, New York. Rosenberg, M.S., Adams, D.C. & Gurevitch, J. (2000) MetaWin: Statistical Software for Meta-Analysis. Sinauer Associates, Sunderland, Massachu­setts, USA. Salminen, J. & Karonen, M. (2011) Chemical ecology o f tannins and other phenolics: do we need a change in approach? Functional Ecology, in press. Simms, E.L. (1992) Costs of plant resistance to herbivory. Plant Resis­tance to Herbivores and Pathogens: Ecology, Evolution, and Genetics (eds R.S. Fritz & E.L. Simms), pp. 392^425. University o f Chicago Press, Chicago. Smilanich, A.M. (2008) Variation in plant chemical defense and the physiological response o f specialist and generalist herbivores. PhD dissertation, Tulane Uni­versity, New Orleans. Stahl, E. (1888) Pflanzen und Schneden. Jenaische Zeitschrift F. Naturwissens-chaften, 22, 557-684. Stamp, N. (2003) Out of the quagmire of plant defense hypotheses. The Quar­terly Review o f Biology, 78,23-54. Strauss, S.Y. & Agrawal, A.A, (1999) Theecology and evolution o f plant toler­ance to herbivory. Trends in Ecology and Evolution, 14. 179-185. Strauss, S.Y.. Rudgers. J.A., Lau, J.A. & Irwin, R.E. (2002) Direct and ecologi­cal costs o f resistance to herbivory. Trends in Ecology and Evolution, 17,278­285. Summers, C.B. & Felton, G.W. (1994) Prooxidant effects o f phenolic acids on the generalist herbivore Helicoverpa zea (Lepidoptera: Noctui-dae): potential mode of action for phenolic compounds in plant anti­herbivore chemistry. Insect Biochemistry and Molecular Biology, 24, 943-953. Swaine, M.D. & Whitmore, T.C. (1988) On the definition o f ecological species groups in tropical rain forests. Vegetation, 7 5 ,81-86. Taub. D. (2007) A Meta-analysis o f Studies on Plant Growth Rate and Allocation to Roots and Shoots. Available from Nature Proceedings: http://dx.doi.org/ 10.1038/npre.2007,185. (Accessed on April 29,2010), Thompson, J. (1988) Coevolution and alternative hypotheses on insect/plant interactions. Ecology, 69,893-895. Van Zandt, P.A. (2007) Plant defense, growth, and habitat: a comparative assessment o f constitutive and induced resistance. Ecology, 88, 1984­1993. Whittaker, R.H. & Feeny. P.P. (1971) Allelochemics: chemical interactions between species. Science, 171,757-770. Received8 June 2010; accepted 28 September 2010 Handling Editor: Marc Johnson Supporting Information Ad d itio n a l Su p p o rtin g In fo rm a tio n m a y be fo u n d in the online ver­sion o f this article. Appendix S I . Studies included in th e m eta-analyses. Appendix S2. M ate rials an d methods. Appendix S3. Meta-analysis records. As a service to o u r a u th o rs a n d readers, this jo u rn a l p ro v id e s sup­p o rtin g in fo rm a tio n supplied by th e au th o rs. Such ma terials may be re-organized for online delivery, b u t a re n o t copy-edited o r typeset. Technical su p p o rt issues arising from su p p o rtin g in fo rm a tio n (o the r th a n missing files) sh o u ld be addressed to th e au th o rs. © 2010 The Authors. Journal compilation © 2010 British Ecological Society, Functional Ecology 25 APPENDIX S1: Studies included in the meta-analyses Agrawal, A. A. & Fishbein, M. (2008). Phylogenetic escalation and decline of plant defense strategies. Proceedings o f the National Academy o f Science o f the United States o f America 105: 10057-10060. Almeida-Cortez, J. S., Shipley, B. & Arnason, J. T. (1999). Do plant species with high relative growth rates have poorer chemical defences? Functional Ecology 13: 819-827. Baldwin, I. T. & Schultz, J. C. (1988). Phylogeny and the patterns of leaf phenolics in gap-adapted and forest-adapted Piper and Miconia understory shrubs. Oecologia 75:105-109. Bryant, J. P., Kuropat, P. J., Cooper, S. M., Frisby K. & Owen-Smith, N. (1989). Resource availability hypothesis of plant antiherbivore defense tested in a South African savanna ecosystem. Nature 340: 227-229. Cebrian, J. & Duarte, C. M. (1994). The dependence of herbivory on growth rate in natural plant communities. Functional Ecology 8: 518-525. Coley, P.D. (1987). Interspecific variation in plant anti-herbivore properties: the role of habitat quality and rate of disturbance. New Phytologist 106: 251-263. Coley, P. (1988). Effects of plant growth rate and leaf lifetime on the amount and type of anti-herbivore defense. Oecologia 74: 531-536. Cunningham, S. A., Summerhayes, B. & Westoby, M. (1999). Evolution divergences in leaf structure and chemistry, comparing rainfall and soil gradients. Ecological Monographs 69: 569-588. Dalling, J. W., Pearson, T. R.H., Ballesteros, J., Sanchez, E. & Burslem, D.F.R.P. (2009). Habitat partitioning among neotropical pioneers: a consequence of differential susceptibility to browsing herbivores? Oecologia 161: 361-370. De la Cruz, M. & Dirzo, R. (1987). A survey of the standing levels of herbivory in seedlings from a Mexican rain forest. Biotropica 19: 98-106. 25 Dominy, N. J., Grubb, P. J., Jackson, R.V., Lucas, P.W., Lucas, W., Metcalfe, D.J., Svenning, J.C. & Turner, I. M. (2008). In tropical lowland rain forests monocots have tougher leaves than dicots, and include a new kind of tough leaf. Annals of Botany 101: 1363-1377. Downum, K., Lee, D., Halle, F., Quirke, M. & Towers, N. (2001). Plant secondary compounds in the canopy and understory of a tropical rain forest in Gabon. Journal o f Tropical Ecology 17: 477-481. Dudt, J. F. & Shure, D. J. (1994). The influence of light and nutrients on foliar phenolics and insect herbivory. Ecology 75: 86-98. Eichhorn, M. P., Fagan, K. C. & Compton, S. G. (2007). Explaining leaf herbivory rates on tree seedlings in a Malaysian rain forest. Biotropica 39: 416-421. Farji-Brener, A. G. (2001). Why are leaf-cutting ants more common in early secondary forests than in old-growth tropical forests? An evaluation of the palatable forage hypothesis. Oikos 92: 169-177. Fine, P.V.A., Mesones, I. & Coley, P.D. (2004). Herbivores promote habitat specialization by trees in Amazonian forests. Science 305:663-665. Fine, P.V.A., Miller, Z. J., Mesones, I., Irazuzta, S., Appel, H. M., Stevens, H. H., Saaksjarvi, I., Schultz, J. C. & Coley, P. D. (2006). The growth-defense trade-off and habitat specialization by plants in Amazonian. Ecology 87: S150-S162. Folgarait, P. J. & Davidson, D.W. (1994). Antiherbivore defenses on myrmecophytic Cecropia under different light regimes. Oikos 71:305-320. Folgarait, P. J. & Davidson, D.W. (1995). Myrmecophytic Cecropia: Antiherbivore defenses under different nutrient treatments. Oecologia 104: 189-206. Fraser, L. H. & Grime, P. (1999). Interacting effects of herbivory and fertility on a synthesized plant community. Journal o f Ecology 87: 514-525. Hallam, A. & Read, J. (2006). Do tropical plant species invest more in anti-herbivore defense than temperate species? A test in Eucryphia (Cunoniaceae) in eastern Australia. Journal o f Tropical Ecology 22:41-51. Hendriks, R. J. J., de Boer, N. J. & van Groenendael, J. M. (1999). Comparing the preferences of three herbivore species with resistance traits of 15 perennial dicots: the effects of phylogenetic constraints. Plant Ecology 143:141-152. Holdo, R. M. (2003). Woody plant damage by African elephants in relation to leaf nutrients in Western Zimbabwe. Journal o f Tropical Ecology 19: 189-196. 26 Howlett, B. E. & Davidson, D.W. (2001). Herbivory on planted dipterocarp seedlings in secondary logged forests and primary forests of Sabah, Malaysia. Journal o f Tropical Ecology 17: 285-302. Lowman, M. D. (1987). Relationships between leaf growth and holes caused by herbivores. Australian Journal o f Ecology 12: 189-191. Lowman, M. D. (1992). Leaf growth dynamics and herbivory in five species of Australian rain-forest canopy trees. Journal o f Ecology 80: 433-447. Massey, F. P., Roland, E. & Hartley, S. E. (2007). Grasses and the resource availability hypothesis: the importance of silica-based defenses. Journal o f Ecology 95: 414­424. Matsuki, S. & Takayoshi, K. (2006). Comparison of leaf life span, photosynthesis and defensive traits across seven species of deciduous broad-leaf tree seedlings. Annals o f Botany 97: 813-817. McCanny, S. J., Keddy, P. A., Arnason, T. J., Gaudet, C. L., Moore, D. R. J. & Shipley, B. (1990). Fertility and the food quality of wetland plants: a test of the resource availability hypothesis. Oikos 59: 373-381. Miller, R. E. & Woodrow, I. E. (2008). Resource availability and the abundance of an N-based defense in Australian tropical rain forests. Ecology 89: 1503-1509. Newbery, D. M. & de Foresta, H. (1985). Herbivory and defense in pioneer, gap and understory trees of tropical rain forest in French Guiana. Biotropica 17: 238-244. Read, J., Sanson, G. D., de Garine-Wichatitsky, M. & Jaffre, T. (2006). Sclerophylly in two contrasting tropical environments: low nutrients vs. low rainfall. American Journal o f Botany 93:1601-1614. Reich, P. B., Uhl, C., Walters, M. B. & Ellsworth, D. S. (1991). Leaf lifespan as a determinant of leaf structure and function among 23 Amazonian tree species. Oecologia 86: 16-24. Reich, P. B., Walters, M. B. & Ellsworth, D. S. (1992). Leaf life-span in relation to leaf, plant, and stand characteristics among diverse ecosystems. Ecological Monographs 62: 365-392. Reich, P. B., Walters, M. B. & Ellsworth, D. S. (1997). From tropics to tundra: global convergence in plant functioning. Proceedings o f the National Academic o f Science o f the United States o f America 94: 13730-13734. 27 Richards, L. A. & Coley, P. D. (2007). Seasonal and habitat differences affect the impact of food and predation on herbivores: a comparison between gaps and understory of a tropical forest. Oikos 116: 31-40. Rogers, W. E. & Siemann, E. (2002). Effects of simulated herbivory and resource availability on native and invasive exotic tree seedlings. Basic and Applied Ecology 3: 297-307. Shure, D. & Wilson, L. (1993). Patch size effects on plant phenolics in successional openings of the southern Appalachians. Ecology 74: 55-67. Skarpe, C., Bergstrom, R., Braten, A. L. & Danell, K. (2000). Browsing in a heterogeneous savanna. Ecography 23: 632-640. Smith, D. M. & Nufio, C. R. (2004). Levels of herbivory in two Costa Rican rain forests: implications for studies of fossil herbivory. Biotropica 36: 318-326. Southwood, T. R. E., Brown, V. K. & Reader, P. M. (1986). Leaf palatability, life expectancy and herbivore damage. Oecologia 70: 544-548. Turner, I. M., Choong, M. F., Tan, H. T. W. & Lucas, P.W. (1993). How tough are sclerophylls? Annals o f Botany 71: 343-345. Turner, I. M. (1995). Foliar defenses and habitat adversity of three woody plant communities in Singapore. Functional Ecology 9: 279-284. Turner, I. M., Lucas, P.W, Becker, P, Wong, S.C. , Yong, J. W. H., Choong, M. F. & Tyree M.T. (2000). Tree leaf form in Brunei: a heath and mixed dipterocarp forest compared. Biotropica 32: 53-61. Van Zandt, P. A. (2007). Plant defense, growth, and habitat: a comparative assessment of constitutive and induced resistance. Ecology 88: 1984-1993. Vasconcelos, H. (1999). Levels of leaf herbivory in Amazonian trees from different stages in forest regeneration. Acta Amazonica 29: 615-623. Wardle, D. A., Barker, G. M., Bonner, K. I. & Nicholson, K. S. (1998). Can comparative approaches based on plant ecophysiological traits predict the nature of biotic interactions and individual plant species effects in ecosystems? Journal o f Ecology 86:405-420. Wilf, P., Labandeira, C.C., Johnson, K.R., Coley, P.D. & Cutter, A.D. (2001). Insect herbivory, plant defense and early Cenozoic climate change. Proceedings o f the National Academy o f Science 98:6221-6226. 28 APPENDIX S2: Materials and Methods We reviewed published studies comparing growth, defenses and herbivory among different plant species and under different resource environments. We selected suitable studies using three electronic databases: Web of Knowledge and Google Scholar for 1985-2010 and Web of Science for 1997-2009. We searched for the terms "plant" and "herbiv*" and "defens*" and "resource*" (or "light" or "nutrient*") and "growth". Other relevant studies were also found by searching the reference section in the articles retrieved from the term searches. These searches led to a large number of articles (~600) that were examined but only a subset met our criteria (described below) and were, therefore, included in our meta-analyses. Thus our analysis is based on 50 studies published between 1985-2010 and conducted on > 600 different plant species (see references of studies included in Appendix 1). We set a specific inclusion criterion to choose the studies for our meta-analyses. We restricted our analyses to studies that examined interspecific differences in plant species within a site or between sites differing in their degree of resource availability. Thus, studies that compared growth, defenses or herbivory in the same plant species in different resource environments were not considered. In order to be included in our review, a study had to provide data for growth estimates or growth categorization of the plant species and information related to the resource availability of the habitat at which the study was conducted. 29 In the original articles, the habitats at which the studies were conducted were usually classified as either resource-poor environments or resource-rich environments based on the levels of nutrient availability (low or high, infertile or fertile) or of light availability (shade or gap, understory or canopy, primary forest or secondary forest). Plant growth estimates included growth measurements (biomass, height, leaf area and leaf production) and growth categorizations (slow growing or fast growing as classified by the authors). Leaf lifetime measurements were expressed as the number of days or months a leaf was monitored until senescence. Antiherbivore defenses included plant traits proven to decrease herbivore consumption and/or herbivore growth and survival. Defenses were classified as chemical (concentrations or percentages of secondary compounds per unit weight of tissue), mechanical (trichome density and leaf toughness) or nutritional (concentrations or percentages of water and/or nitrogen per unit weight of tissue). Herbivory measurements were usually estimated as the percentage of leaf area eaten or biomass consumed. We performed separate meta-analyses for each of the four predictions from the RAH. For the prediction (1) meta-analysis we used 24 records from 13 studies that compared inherent growth measurements between species from resource-poor environments vs. species from resource-rich environments. For meta-analysis of the prediction (2) we included 10 records from 4 studies that examined leaf lifespan in relation to growth rate. For prediction (3) meta-analysis we used 103 records from 30 studies that contrasted investment of constitutive defenses between fast-growing and slow-growing species. And, for the meta-analysis of the prediction (4), we summarized the results for 42 records from 25 studies that related herbivore damage with plant growth rate. For the last two meta-analyses, the articles were grouped into two types: studies that compared investment in plant defenses, herbivory and growth between two or more different species within a site, and those that compared two or more different species growing in sites with different resource levels (light and nutrients). In addition, for all the meta-analyses, we further distinguished between studies that were conducted in different ecosystems (tropical vs. temperate forests) and with different ontogenetic stages (adults, saplings and seedlings). We performed these further categorizations in order to analyze for other sources of variation in the expression of plant growth, defenses and herbivory in relation to resource availability. From most studies, more than one record suitable for our meta-analysis could be recovered. Some studies reported results from more than one response variable (growth, leaf lifespan, defense and herbivory). In these cases, we created four separate data sets corresponding to the four response variables and the results for each response variable were included in different meta-analyses. In addition, if a study reported data for several plant species, we included each species separately in the meta-analysis to avoid statistical problems related to non-independent comparisons. STATISTICAL ANALYSES Studies included in the meta-analyses of the first prediction addressed plant growth differences in relation to resource availability by comparing growth between species from resource-rich environments vs. species from resource-poor environments. The outcomes from these studies were usually in the form of mean and standard 30 deviations or standard errors. As a first step in our meta-analysis, we converted all these estimates into a common measure of effect size, the standardized mean difference statistic, Hedges' d (Gurevitch & Hedges 2001), which is a statistic commonly used in meta-analyses (Hawkes & Sullivan 2001; Gomez-Aparicio et al. 2004; Maestre, Valladares & Reynolds 2005). For the reported means and variances of growth measures for species from resource-rich and resource-poor environments we directly calculated d: ( X r - X p ) d =--------------- J S Y v where R and P are the means, S is the pooled standard deviation and J is a correction term for small sample sizes (Rosenberg, Adams & Gurevitch 2000). positive values of Hedges' d indicate that plant species in high-resource environments grow faster than those in low-resource environments and vice versa. Studies used for the meta-analyses of predictions (2), (3) and (4) examined the relationships between continuous variation in leaf lifetime, defenses or herbivory and growth. The outcomes from these studies were usually correlation coefficients or regression equations. However, some studies reported t, x2, and F values from statistical tests comparing such variation in relation to growth. For these studies we selected the Pearson product-moment correlation coefficient (r) as the common measure of association between leaf lifetime, defenses or herbivory and growth and calculated effect sizes from these coefficients. The correlation coefficient (r) is considered an advantageous effect size statistic because it is easy to interpret (Koricheva 2002, 31 32 Koricheva, Nykanen & Gianoli 2004), and principally, because most of the commonly used test statistics can be translated into an r value (Rosenberg, Adams & Gurevitch 2000). When the results of the studies were reported in the form of correlation coefficients, they were directly included into the data sets. When the associations were examined as regression analyses, we took the square root from the coefficient of determination (R2). F-statistics, ^-statistics and x2-statistics were transformed into r following Rosenberg , Adams & Gurevitch (2000). Individual r coefficients were z-transformed and weighted by their respective sample sizes. The sign of r reflects the patterns of plant defenses or herbivory in relation to growth. If growth decreases plant investment in defenses, r will be negative (and vice versa). If growth increases nutritional content and plant herbivory, r will be positive (and vice versa). All the meta-analyses were conducted with the program MetaWin version 2.1.5 (Rosenberg , Adams & Gurevitch 2000), and using the mixed effects model (Gurevitch & Hedges 1993), which assumes that variation observed among studies is due to sampling error and random variation (Koricheva 2002). As we mentioned above, we conducted separate meta-analyses for each prediction following the RAH. The magnitude of the effect size was considered to be statistically significant when the bias-corrected 95% confidence interval of the z-transformed effect size, generated from 9,999 iterations, did not include zero (Gurevitch & Hedges 1993). In addition, following Rosenberg, Adams & Gurevitch (2000), we analyzed if the size of the effect varied across different measurements of growth rate for the meta-analysis of prediction (1), and between studies that were conducted in tropical vs. temperate forests and between different ontogenic stages (seedlings, saplings and adults) for all meta-analyses. For this, we used the statistic Qb, which is a weighted sum of squares that can be tested against an x2 distribution with n -1 degrees of freedom (Gurevitch & Hedges 2001). Significant values of Qb imply that there are differences in effect sizes between groups. In order to test publication bias, e.g., the tendency of journals to publish only studies with significant results, we calculated a fail-safe number (nfs) by means of the weighted method proposed by Rosenthal (1979). A fail-safe number is the number of non-significant or unpublished studies needed to change a significant effect to a non­significant effect in a meta-analysis (Rosenberg, Adams & Gurevitch 2000). If this number is larger than 5n + 10, where n is the number of observed studies in the meta-analysis, we can be confident of the robustness of our analyses against publication bias (Rosenberg, Adams & Gurevitch 2000). Another source of bias in meta-analytical methods constitutes the phylogenetic nonindependence. Given that the goal of meta-analysis is to summarize research from multiple taxa, the absence of phylogenetic independence, resulting from shared phylogenetic history among closely related taxa, violates statistical assumptions of independence (Lajeunesse 2009). Thus, the integration of phylogenetic information into ecological meta-analysis is becoming a new and exciting area. However, because of the novelty of its application, the use of phylogenetic information in quantitative reviews still has several theoretical and practical issues, such as the accessibility to information and difficulty of analysis needed to construct evolutionary relationships (Lajeunesse 2009), especially for unbalanced designs like our study. For these reasons, we decided to not include phylogenetic information in our analyses. Moreover, we worked with a broad diversity of plant species from many different families and from many different 33 34 geographic localities. Therefore, we consider that the absence of phylogeny will not threaten our conclusions because, by inspection, there is no a phylogenetic bias in our results. REFERENCES Gomez-Aparicio, L., Zamora, R., Gomez, J. M., Hodar, J. A., Castro, J. & Baraza, E. (2004). Applying plant facilitation to forest restoration: a meta-analysis of the use of shrubs to nurse plants. Ecological Applications 14: 1128-1138. Gurevitch, J. & Hedges, L. V. (1993). Meta-analysis: combining the results of independent experiments. Design and analysis of ecological experiments (eds. S. M. Scheiner & J. Gurevitch), pp: 378-425. Chapman & Hall, New York. Gurevitch, J. & Hedges, L. V. (2001). Meta-analysis: combining the results of independent experiments. Design and analysis of ecological experiments. (eds. S. M. Scheiner & J. Gurevitch), pp: 347-369. Oxford University Press, Oxford. Hawkes, C. V. & Sullivan, J. J. (2001). The impact of herbivory on plants in different resource conditions: a meta-analysis. Ecology 82: 2045-2058. Koricheva, J. (2002). Meta-analysis of sources of variation in fitness costs of plant antiherbivore defenses. Ecology 83:176-190. Koricheva, J., Nykanen, H. & Gianoli, E. (2004). Meta-analysis of trade-offs among plant antiherbivore defenses: Are plants jacks-of-all-trades, masters of all? The American Naturalist 163: E64-E75. Lajeunesse, M. J. (2009). Meta-analysis and the comparative phylogenetic method. The American Naturalist 174: 369-381. Maestre, F. T., Valladares, F. & Reynolds, J. F. (2005). Is the change of plant-plant interactions with abiotic stress predictable? A meta-analysis of field results in arid environments. Journal of Ecology 93: 748-757. Rosenberg, M. S., Adams, D. C. & Gurevitch, J. (2000). MetaWin: statistical software for meta-analysis. Sinauer Associates, Sunderland, Massachusetts, USA. Rosenthal, R. (1979). The "file drawer problem" and tolerance for new results. Psychological Bulletin 86: 638- 641. APPENDIX S3 Meta-analysis records Table S3.1. Prediction (1) meta-analysis records for the effect of resources on plant growth. N= sample size. Reference Ecosystem Habitat Resource Ontogeny Functional Group Effect size (d) N Bryant et al. 1989 Tropical Nutrients Seedling Woody 6.21 30 Bryant et al. 1989 Tropical Nutrients Seedling Woody 19.54 30 Coley, 1987 Tropical Light Saplings Woody 0.58 47 Coley, 1987 Tropical Light Saplings Woody 0.29 47 Coley, 1987 Tropical Light Saplings Woody 0.43 47 Coley, 1988 Tropical Light Saplings Woody 0.76 41 Coley, 1988 Tropical Light Saplings Woody 0.52 41 Cunningham, et al. 1999 Temperate Nutrients Adults Woody 1.68 9 Cunningham, et al. 1999 Temperate Nutrients Adults Woody 1.68 9 Dalling et al. 2009 Tropical Light Seedlings Woody -5.88 30 Dalling et al. 2009 Tropical Light Seedlings Woody 0.32 30 Dalling et al. 2009 Tropical Light Seedlings Woody -4.28 30 Dudt & Shure, 1994 Temperate Light, Nutrients Seedlings Woody 0.33 240 Fine et al. 2004 Tropical Nutrients Seedlings Woody 0.13 880 Fine et al. 2006 Tropical Nutrients Seedlings Woody 3.8 880 Fine et al. 2006 Tropical Nutrients Seedlings Woody 1.96 880 Folgarait & Davidson 1994 Tropical Light Seedlings Woody 1.52 24 36 Table S3.1. (continued) Reference Ecosystem Habitat Resource Ontogeny F