Evolutionary Trade-Offs in Plants Mediate the Strength of Trophic Cascades

See allHide authors and affiliations

Science  26 Mar 2010:
Vol. 327, Issue 5973, pp. 1642-1644
DOI: 10.1126/science.1184814

Trophic Trade-Offs

There have been many attempts to document and explain the effects of predators on plant biomass in so-called “trophic cascades.” Theory suggests that fast-growing plants are relatively undefended and suffer more from herbivory, which implies a functional trade-off between investment in traits relating to growth and defensive strategies. Mooney et al. (p. 1642; see the Perspective by Hambäck) compared responses to fertilization and aphid predators in 16 milkweed species. As predicted, interspecific variation in the strength of top-down control in terms of a tradeoff with growth was observed.


Predators determine herbivore and plant biomass via so-called trophic cascades, and the strength of such effects is influenced by ecosystem productivity. To determine whether evolutionary trade-offs among plant traits influence patterns of trophic control, we manipulated predators and soil fertility and measured impacts of a major herbivore (the aphid Aphis nerii) on 16 milkweed species (Asclepias spp.) in a phylogenetic field experiment. Herbivore density was determined by variation in predation and trade-offs between herbivore resistance and plant growth strategy. Neither herbivore density nor predator effects on herbivores predicted the cascading effects of predators on plant biomass. Instead, cascade strength was strongly and positively associated with milkweed response to soil fertility. Accordingly, contemporary patterns of trophic control are driven by evolutionary convergent trade-offs faced by plants.

Trophic cascades—the indirect positive effect of predators on plant biomass through herbivore suppression—are the best examples of the importance of indirect interactions as determinants of community structure and ecosystem function. For this reason, there has been great interest in elucidating the sources of variation in trophic cascade strength both within (13) and among ecosystems (4). Much of the research aimed at explaining variation in trophic cascade strength has focused on factors mediating the top-down effects of predators on herbivores, including the influences of intraguild predation (5), synergistic and antagonistic effects of multiple predators (6), trophic subsidies to predators (7), and the nonconsumptive effects of predators on herbivores (8). At the same time, it is also recognized that plant stoichiometry (9), antiherbivore defense traits (1012), and primary productivity (13, 14) can mediate trophic cascade strength from the bottom up. Consequently, a consensus is emerging that multiple, complementary top-down and bottom-up processes determine trophic cascade strength.

Although it is recognized that plant traits can influence interactions with herbivores and herbivore-predator interactions (15, 16), there has been little consideration of how plant growth and defense strategies might result in predictable patterns of trophic cascade strength. There is wide acceptance that plant species evolve in response to fundamental trade-offs that should influence the effects of predators and productivity upon herbivore and plant biomass. For example, plant defense theory predicts that fast-growing species should have relatively low herbivore resistance as compared with slow-growing species (17, 18). Plant resistance to herbivores may in turn influence the indirect effects of predators on plants by altering herbivore susceptibility to predators (19, 20). Similarly, plant growth strategies influence tolerance to herbivory (21, 22), again showing potential to alter the strength of trophic cascades. Although trophic cascades are rightly considered community-level phenomena (23), an understanding of how plant traits influence such dynamics requires first documenting the influence of plant traits on component, species-level cascades.

We conducted a field experiment in which we grew 16 species of milkweeds (Asclepias spp., Apocynaceae) (Fig. 1), factorially manipulated predator access and soil fertility, and monitored plant biomass and populations of the potent herbivore Aphis nerii (Aphididae, Hemiptera), a specialist on the Apocynaceae that occurs naturally on the studied milkweeds (24). It has previously been shown that milkweed species influence this aphids’ population dynamics and interactions with parasitoids (16, 25). We tested whether there are indirect consequences of such variation in trophic dynamics for plant growth as well as whether trade-offs between milkweed growth strategy and herbivore resistance predictably influence the top-down effects of predators. Because all plants were grown in a single environment, any variation in the effects of predators and growth strategy can be attributed to plant species traits. By interpreting these patterns of interspecific variation in trophic structure from a phylogenetic perspective, we link the outcome of fundamental evolutionary trade-offs to contemporary community dynamics.

Fig. 1

Effect sizes for influence of predators and fertilization on herbivore density (calculated per gram of plant dry biomass) and final plant biomass. Black bars for A. nivea and A. candida indicate that soil fertility was not manipulated for these species. Effects are natural log response ratios (LRRs) with 95% confidence limits. Predator effects are calculated across both levels of soil fertility, soil fertility effects are calculated across both levels of predation, and all effects are based on manipulation and control sample sizes of n = 10 plants each. LRRs of 1.0, 2.0, and 3.0 correspond to changes of 2.7-fold, 7.4-fold, and 20-fold, respectively. Predator effects on both herbivore density and plant biomass differ significantly among species, whereas fertilization effects differ for plant biomass but not for herbivore density (table S1). To the left of the effect sizes, the phylogenetic relationship of the studied milkweeds is presented (29).

We first tested for variation among milkweed species in the effects of predators and soil fertility on both plant biomass and herbivore abundance using general linear models. Where species varied in such responses, we then quantified effect sizes for individual species [log response ratios (26)] in order to examine the relationships among species using phylogenetically independent contrasts (27).

The 16 milkweed species varied in herbivore resistance (quantified here as –1 × aphid density in the absence of predators) (fig. S1 and table S1) and the top-down effects of predators on herbivore density and plant biomass (Fig. 1 and table S1). This variation in top-down control among species is equivalent in magnitude to that previously observed among ecosystem types (4). That such variation occurs among closely related species in a single abiotic environment, with the same herbivore species and with a common guild of predators, underscores the powerful influence of plant traits upon trophic structure.

Milkweed species also differed strongly with respect to two aspects of the plants’ growth strategy, growth rate [which is defined as mean species biomass at the conclusion of the experiment (fig. S1 and table S1)], and the growth response to increased soil fertility (Fig. 1 and table S1). The strength of soil fertility effects on plant biomass was stronger when predators were absent than when present (fig. S2 and table S1), but these dynamics were consistent among milkweed species (table S1). Despite the species-specific effects of soil fertility on milkweed growth, the indirect positive effect of soil fertility on aphid density was indistinguishable among milkweed species (Fig. 1 and table S1). Accordingly, there was asymmetry in how milkweed species influenced top-down and bottom-up trophic dynamics: Predator effects on both herbivore density and plant biomass were species-specific, whereas species variation in soil fertility effects were limited to the direct influence on plant biomass.

Having shown milkweed species differences in growth strategies, resistance, and the top-down effects of predators, we examined the relationships among these species traits while controlling for phylogenetic history (27). Variation in herbivore density among milkweed species was determined by means of a combination of top-down and bottom-up processes. Although theory and data predict that herbivore resistance in plants should influence predator-herbivore interactions (19, 20), predator effects on herbivores were unrelated to milkweed resistance (fig. S3). In addition, the strength of predator effects on herbivores did not vary as a function of milkweed growth or growth response to soil fertility (table S2). However, both components of milkweed growth strategy convergently traded off against resistance so that species that were fast growing and responsive to soil fertility had low resistance (meaning, higher herbivore densities) (Fig. 2 and table S2), which is consistent with predictions from plant defense theory (17, 18). Thus, herbivore density was jointly and independently determined by means of interspecific variation in the top-down effect of predators and the bottom-up effect of milkweed growth strategies.

Fig. 2

(A to D) Relationship between milkweed species’ growth rate, growth response to soil fertility, resistance to herbivores, and predator effects on plant biomass. For ease of interpretation, raw data are depicted with both raw and phylogenetically corrected R2 values indicated in each panel. Where the phylogenetically independent correlations were significant [conducted by using generalized least-squares methods (27)], a linear best fit is shown through the raw data. Growth rate and response to soil fertility themselves are uncorrelated (table S2).

Surprisingly, predator effects on plant biomass were unrelated to either predator effects on herbivores or herbivore density (resistance) (table S2). Instead, plant growth response to soil fertility predicted more than half of the variation in the top-down effects of predators on plant biomass (Fig. 2). At the same time, milkweed growth rate was not predictive of predator effects on plant biomass [and growth rate and response to soil fertility themselves are uncorrelated (table S2)]. Because the impact of predators on plant biomass was not related to the strength of herbivore suppression, variation in the indirect effects of predators on plants is probably attributable to variation in tolerance of milkweed species to herbivory. Consequently, an evolutionary trade-off leads to an association between high growth in response to soil fertility, low tolerance to herbivory, and an increase in predator effects on plant biomass.

Of several plant traits assayed, we found evidence suggestive of one mechanism behind the observed variation in the top-down effects of predators on plants. Plant emissions of sesquiterpene volatile organic compounds (VOCs) were significantly positively correlated with the top-down effects of predators on plant biomass (fig. S4 and table S3). Sesquiterpenes are a group of VOCs that can play a key role in indirect defense through recruitment of predators to plants (28). Therefore, variation among species in trophic cascade strength may be driven at least in part by interspecific variation in this ecologically important group of volatile compounds.

We have documented wide variation in top-down regulation of plant and herbivore biomass among a group of closely related species; this variation corresponded with a fundamental evolutionary trade-off faced by plants. Interspecific variation in herbivore density was determined jointly and independently through variable effects of predators and two components of plant growth. In contrast, the cascading effects of predators on plant biomass were not tied to predator-herbivore interactions but instead fell along a probable trade-off between tolerance to herbivores and the bottom-up effects of soil fertility. Whereas food web models have predicted the correspondence of top-down and bottom-up effects on the basis of thermodynamic principles of energy flow (13, 14), we show here that such dynamics can similarly arise from convergent trade-offs faced by plants over evolutionary time. Our results underscore the importance of considering the plant-herbivore linkage as a determinant of trophic cascade strength, and the dynamic interplay between past evolutionary processes and contemporary ecological dynamics.

Supporting Online Material

Materials and Methods

Figs. S1 to S4

Tables S1 to S4


References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. We thank R. A. Smith and J. Goldstein for field assistance; S. Rasmann for contributing cardenolide data; M. Fishbein for providing phylogenetic information; and the University of California Berkeley Botanical Garden, S. Malcolm, A. Rapini, and M. Fishbein for providing plant material. The manuscript was improved by comments from A. Flecker, S. Rasmann, D. Raboski, and D. Gruner and supported by NSF, the Cornell Center for a Sustainable Future, and the University of California Irvine School of Biological Sciences.
View Abstract

Navigate This Article