Nutrient Enrichment and Food Web Composition Affect Ecosystem Metabolism in an Experimental Seagrass Habitat

Background

Food web composition and resource levels can influence ecosystem properties such as productivity and elemental cycles. In particular, herbivores occupy a central place in food webs as the species richness and composition of this trophic level may simultaneously influence the transmission of resource and predator effects to higher and lower trophic levels, respectively. Yet, these interactions are poorly understood.

Methodology/Principal Findings

Using an experimental seagrass mesocosm system, we factorially manipulated water column nutrient concentrations, food chain length, and diversity of crustacean grazers to address two questions: (1) Does food web composition modulate the effects of nutrient enrichment on plant and grazer biomasses and stoichiometry? (2) Do ecosystem fluxes of dissolved oxygen and nutrients more closely reflect above-ground biomass and community structure or sediment processes? Nutrient enrichment and grazer presence generally had strong effects on biomass accumulation, stoichiometry, and ecosystem fluxes, whereas predator effects were weaker or absent. Nutrient enrichment had little effect on producer biomass or net ecosystem production but strongly increased seagrass nutrient content, ecosystem flux rates, and grazer secondary production, suggesting that enhanced production was efficiently transferred from producers to herbivores. Gross ecosystem production (oxygen evolution) correlated positively with above-ground plant biomass, whereas inorganic nutrient fluxes were unrelated to plant or grazer biomasses, suggesting dominance by sediment microbial processes. Finally, grazer richness significantly stabilized ecosystem processes, as predators decreased ecosystem production and respiration only in the zero- and one- species grazer treatments.

Conclusions/Significance

Overall, our results indicate that consumer presence and species composition strongly influence ecosystem responses to nutrient enrichment, and that increasing herbivore diversity can stabilize ecosystem flux rates in the face of perturbations.     

Trophic‐level dependent effects on CO2 emissions from experimental stream ecosystems

Concern over accelerating rates of species invasions and losses have initiated investigations into how local and global changes to predator abundance mediate trophic cascades that influence CO₂ fluxes of aquatic ecosystems. However, to date, no studies have investigated how species additions or losses at other consumer trophic levels influence the CO₂ flux of aquatic ecosystems. In this study, we added a large predatory stonefly, detritivorous stonefly, or grazer tadpole to experimental stream food webs and over a 70‐day period quantified their effects on community composition, leaf litter decomposition, chlorophyll‐a concentrations, and stream CO₂ emissions. In general, streams where the large grazer or large detritivore were added showed no change in total invertebrate biomass, leaf litter loss, chlorophyll‐a concentrations, or stream CO₂ emissions compared with controls; although we did observe a spike in CO₂ emissions in the large grazer treatment following a substantial reduction in chlorophyll‐a concentrations on day 28. However, the large grazer and large detritivore altered the community composition of streams by reducing the densities of other grazer and detritivore taxa, respectively, compared with controls. Conversely, the addition of the large predator created trophic cascades that reduced total invertebrate biomass and increased primary producer biomass. The cascading effects of the predator additions on the food web ultimately led to decreased CO₂ emissions from stream channels by up to 95%. Our results suggest that stream ecosystem processes were more influenced by changes in large predator abundance than large grazer or detritivore abundance, because of a lack of functionally similar large predators. Our study demonstrates that the presence/absence of species with unique functional roles may have consequences for the exchange of CO₂ between the ecosystem and the atmosphere.     

Effects of trophic skewing of species richness on ecosystem functioning in a diverse marine community

Widespread overharvesting of top consumers of the world's ecosystems has "skewed" food webs, in terms of biomass and species richness, towards a generally greater domination at lower trophic levels. This skewing is exacerbated in locations where exotic species are predominantly low-trophic level consumers such as benthic macrophytes, detritivores, and filter feeders. However, in some systems where numerous exotic predators have been added, sometimes purposefully as in many freshwater systems, food webs are skewed in the opposite direction toward consumer dominance. Little is known about how such modifications to food web topology, e.g., changes in the ratio of predator to prey species richness, affect ecosystem functioning. We experimentally measured the effects of trophic skew on production in an estuarine food web by manipulating ratios of species richness across three trophic levels in experimental mesocosms. After 24 days, increasing macroalgal richness promoted both plant biomass and grazer abundance, although the positive effect on plant biomass disappeared in the presence of grazers. The strongest trophic cascade on the experimentally stocked macroalgae emerged in communities with a greater ratio of prey to predator richness (bottom-rich food webs), while stronger cascades on the accumulation of naturally colonizing algae (primarily microalgae with some early successional macroalgae that recruited and grew in the mesocosms) generally emerged in communities with greater predator to prey richness (the more top-rich food webs). These results suggest that trophic skewing of species richness and overall changes in food web topology can influence marine community structure and food web dynamics in complex ways, emphasizing the need for multitrophic approaches to understand the consequences of marine extinctions and invasions.     

Effects of food diversity on diatom selection by harpacticoid copepods

The diversity of species assemblages which occupy a basal position in the trophic pyramid (typically unicellular algae in aquatic environments) is known to influence the interaction with organisms of higher trophic levels. A laboratory feeding experiment was conducted with cultures of three benthic diatom species (Navicula phyllepta, Grammatophora marina and Cylindrotheca closterium) as primary producers and three harpacticoid copepod species (Harpacticus obscurus, Paramphiascella fulvofasciata and Tigriopus brevicornis) as grazers to evaluate the effects of food diversity (and concomitant food availability) on grazing selectivity. This kind of selectivity experiments is singular for benthic harpacticoid copepods as so far, information on food selection of harpacticoids is scarce.Uptake of a unispecific food source by a single copepod species decreased as food diversity (and concomitant overall food concentration) increased. All three consumers reacted similarly to changing food diversity, but exhibited strong species-specific responses to food identity i.e. which diatom was added was crucial. Irrespective of level of food diversity, H. obscurus took up high amounts of G. marina, whereas both P. fulvofasciata and T. brevicornis preferred C. closterium when given the choice between different diatoms. As for zooplanktonic taxa, this experiment showed that in lower benthic marine food webs both prey organisms (primary producers) and grazers play a very specific role. Diversity of food and its identity are of critical importance at the base of the trophic pyramid, influencing trophic transfer from primary producers over grazers to higher trophic levels.     

Ecological stoichiometry of indirect grazer effects on periphyton nutrient content

Ecological stoichiometry has been successful in enhancing our understanding of trophic interactions between consumer and prey species. Consumer and prey dynamics have been shown to depend on the nutrient composition of the prey relative to the nutrient demand of the consumer. Since most experiments on this topic used a single consumer species, little is known about the validity of stoichiometric constraints on trophic interactions across consumers and ecosystems. We conducted a quantitative meta-analysis on grazer–periphyton experiments to test (1) if benthic grazers have consistent effects on the nutrient composition of their prey, and (2) whether these effects can be aligned to the nutrient stoichiometry of grazer and periphyton, other environmental factors, or experimental constraints. Grazers significantly lowered periphyton C:N and C:P ratios, indicating higher N- and P-content of grazed periphyton across studies. Grazer presence on average increased periphyton N:P ratios, but across studies the effect size did not differ significantly from zero. The sign and strength of grazer effects on periphyton nutrient ratios was strongly dependent on the nutrient content of grazers and their food, but also on grazer biomass, the amount of biomass removal and water column nutrients. Grazer with low P-content tended to reduce periphyton P-content, whereas grazers with high P-content increased periphyton P-content. This result suggests that low grazer P-content can be an indication of physiological P-limitation rather than a result of having relatively low and fixed P-requirements. At the across-system scale of this meta-analysis, predictions from stoichiometric theory are corroborated, but the plasticity of the consumer nutrient composition has to be acknowledged. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

In situ warming strengthens trophic cascades in a coastal food web

Global warming may affect most organisms and their interactions. Theory and simple mesocosm experiments suggest that consumer top–down control over primary producer biomass should strengthen with warming, since consumer respiration increases faster with warming than plant photosynthesis. However, these predictions have so far not been tested on natural communities that have experienced warming over many generations. Natural systems display a higher diversity, heterogeneity and complexity than mesocosms, which could alter predicted effects of warming. Here we used an artificially heated part of the northern Baltic Sea (the Forsmark Biotest basin) to test how warming influences trophic interactions in a shallow coastal food web with four trophic levels: omnivorous fish, invertivorous fish, herbivorous invertebrates, and filamentous macroalgae. Monitoring of fish assemblages over six years showed that small invertivorous fish (gobiids, sticklebacks and minnows) were much less abundant in the heated basin than in unheated references areas. Stomach content analyses of the dominating omnivorous fish – Eurasian perch Perca fluviatilis – revealed a strikingly different diet within and outside the Biotest basin; gammarid crustaceans were the dominating prey at heated sites, whereas invertivorous fish (e.g. gobiids) dominated at unheated sites. A 45‐day cage experiment showed that fish exclusion did not affect the biomass of algal herbivores (gastropods and gammarids), but reduced algal biomass in heated sites (but not unheated). This suggests that warming induced a trophic cascade from fish to algae, and that this effect was mediated by predator‐induced changes in herbivore behavior, rather than number. Overall, our study suggests that warming has effectively compressed the food chain from four to three trophic levels (algae, gammarids and perch), which have benefitted the primary producers by reducing grazing pressure. Consequently, warming appears to have restructured this coastal food web through a combination of direct (physiological) and indirect (species interactions) effects.     

Influence of consumer mutual interference on the stabilization of a three-level trophic system

In this paper, we present a three-level (food–prey–predator) trophic food chain which includes consumer mutual interference (MIF). In contrast with other analyses, we consider the effect of both prey and predator MIF on the dynamics of a three-level trophic system. MIF is generally considered to exert a stabilizing effect on population dynamics based on the predator–prey model. However, results from analytical and numerical simulations utilizing a simple three-species food chain model suggest that while the addition of prey MIF to the model provides a stabilizing influence, as the chaotic dynamics collapse to a stable steady state, adding only predator MIF to the model can only stabilize the system at intermediate MIF values. The three-species trophic food chain is also stabilized when combination of both prey and predator MIF is added to the model. Our work serves to provide insight into the effects of MIF in the real world.     

Predators and Resources Influence Phosphorus Transfer along an Invertebrate Food Web through Changes in Prey Behaviour

Predators play a fundamental role in prey trophic behaviour, with indirect consequences for species coexistence and ecosystem functioning. Resource quality and availability also influence prey trophic behaviour, with potential effects on predator-prey dynamics. Although many studies have addressed these topics, little attention has been paid to the combined effects of predators and resources on prey species coexistence and nutrient transfer along food chains, especially in detritus-based systems. To determine the influence of predators and resource quality on the movement and P uptake of detritivores, we carried out a field experiment on the River Kelvin (Scotland) using ³²P to test the hypothesis of reduced prey vagility among resource patches as a strategy to avoid predation. Thirty leaf sacks containing alder leaves and two detritivore prey populations (Asellus aquaticus and Lymnaea peregra) were placed in cages, half of them with two predator species (Dendrocoelum lacteum and Erpobdella octoculata) and the other half without predators. Five alder leaf bags, each individually inoculated with a different fungus strain to simulate a patchy habitat, were placed inside each leaf sack. One bag in each sack was labelled with ³²P, in order to assess the proportion of detritivores using it as food and thus their movement among the five resource patches. Three replicates for each labelled fungus and each predation treatment (i.e. with and without predators) were left on the riverbed for 7 days. The presence of predators had negligible effects on the number of detritivores in the leaf bags, but it did reduce the proportion of ³²P-labelled detritivores and their P uptake. The most strongly affected species was A. aquaticus, whose vagility, trophic overlap with L. peregra and P uptake were all reduced. The results confirm the importance of bottom-up and top-down forces acting simultaneously to regulate nutrient transfer along food chains in patchy habitats.     

Effects of Grazer Presence on Genetic Structure of a Phenotypically Diverse Diatom Population

Studies of predator–prey systems in both aquatic and terrestrial environments have shown that grazers structure the intraspecific diversity of prey species, given that the prey populations are phenotypically variable. Populations of phytoplankton have traditionally considered comprising only low intraspecific variation, hence selective grazing as a potentially structuring factor of both genetic and phenotypic diversity has not been comprehensively studied. In this study, we compared strain specific growth rates, production of polyunsaturated aldehydes, and chain length of the marine diatom Skeletonema marinoi in both grazer and non-grazer conditions by conducting monoclonal experiments. Additionally, a mesocosm experiment was performed with multiclonal experimental S. marinoi populations exposed to grazers at different levels of copepod concentration to test effects of grazer presence on diatom diversity in close to natural conditions. Our results show that distinct genotypes of a geographically restricted population exhibit variable phenotypic traits relevant to grazing interactions such as chain length and growth rates. Grazer presence affected clonal richness and evenness of multiclonal Skeletonema populations in the mesocosms, likely in conjunction with intrinsic interactions among the diatom strains. Only the production of polyunsaturated aldehydes was not affected by grazer presence. Our findings suggest that grazing can be an important factor structuring diatom population diversity in the sea and emphasize the importance of considering clonal differences when characterizing species and their role in nature.     

Trophic cascades alter eco-evolutionary dynamics and body size evolution

Trait evolution in predator–prey systems can feed back to the dynamics of interacting species as well as cascade to impact the dynamics of indirectly linked species (eco-evolutionary trophic cascades; EETCs). A key mediator of trophic cascades is body mass, as it both strongly influences and evolves in response to predator–prey interactions. Here, we use Gillespie eco-evolutionary models to explore EETCs resulting from top predator loss and mediated by body mass evolution. Our four-trophic-level food chain model uses allometric scaling to link body mass to different functions (ecological pleiotropy) and is realistically parameterized from the FORAGE database to mimic the parameter space of a typical freshwater system. To track real-time changes in selective pressures, we also calculated fitness gradients for each trophic level. As predicted, top predator loss generated alternating shifts in abundance across trophic levels, and, depending on the nature and strength in changes to fitness gradients, also altered trajectories of body mass evolution. Although more distantly linked, changes in the abundance of top predators still affected the eco-evolutionary dynamics of the basal producers, in part because of their relatively short generation times. Overall, our results suggest that impacts on top predators can set off transient EETCs with the potential for widespread indirect impacts on food webs.     

Trophic cascades and trophic trickles in pelagic food webs

Prey-dependent models, with the predation rate (per predator) a function of prey numbers alone, predict the existence of a trophic cascade. In a trophic cascade, the addition of a top predator to a two-level food chain to make a three-level food chain will lead to increases in the population size of the primary producers, and the addition of nutrients to three-level chains will lead to increases in the population numbers at only the first and third trophic levels. In contrast, ratio-dependent models, with the predation rate (per predator) dependent on the ratio of predator numbers to prey, predict that additions of top predators will not increase the population sizes of the primary producers, and that the addition of nutrients to a three-level food chain will lead to increases in population numbers at all trophic levels. Surprisingly, recent meta-analyses show that freshwater pelagic food web patterns match neither prey-dependent models (in pelagic webs, ''prey'' are phytoplankton, and ''predators'' are zooplankton), nor ratio-dependent models. In this paper we use a modification of the prey-dependent model, incorporating strong interference within the zooplankton trophic level, that does yield patterns matching those found in nature. This zooplankton interference model corresponds to a more reticulate food web than in the linear, prey-dependent model, which lacks zooplankton interference. We thus reconcile data with a new model, and make the testable prediction that the strength of trophic cascades will depend on the degree of heterogeneity in the zooplankton level of the food chain.     

Temperature and prey density jointly influence trophic and non‐trophic interactions in multiple predator communities

相似文献   

Restricting Prey Dispersal Can Overestimate the Importance of Predation in Trophic Cascades

Predators can affect prey populations and, via trophic cascades, predators can indirectly impact resource populations (2 trophic levels below the predator) through consumption of prey (density-mediated indirect effects; DMIEs) and by inducing predator-avoidance behavior in prey (trait-mediated indirect effects; TMIEs). Prey often employ multiple predator-avoidance behaviors, such as dispersal or reduced foraging activity, but estimates of TMIEs are usually on individual behaviors. We assessed direct and indirect predator effects in a mesocosm experiment using a marine food chain consisting of a predator (toadfish – Opsanus tau), prey (mud crab - Panopeus herbstii) and resource (ribbed mussel – Geukensia demissa). We measured dispersal and foraging activity of prey separately by manipulating both the presence and absence of the predator, and whether prey could or could not disperse into a predator-free area. Consumption of prey was 9 times greater when prey could not disperse, probably because mesocosm boundaries increased predator capture success. Although predator presence did not significantly affect the number of crabs that emigrated, the presence of a predator decreased resource consumption by prey, which resulted in fewer resources consumed for each prey that emigrated in the presence of a predator, and reduced the overall TMIE. When prey were unable to disperse, TMIEs on mussel survival were 3 times higher than the DMIEs. When prey were allowed to disperse, the TMIEs on resource survival increased to 11-times the DMIEs. We found that restricting the ability of prey to disperse, or focusing on only one predator-avoidance behavior, may be underestimating TMIEs. Our results indicate that the relative contribution of behavior and consumption in food chain dynamics will depend on which predator-avoidance behaviors are allowed to occur and measured.     

Consumer effects decline with prey diversity

While consumer species diversity is known to influence the capture of limited resources, little is known about how prey diversity impacts the transfer of energy and matter among trophic levels. Here, we perform a meta‐analysis of experiments that have examined the impact of grazers on the biomass of periphytic algae to test the hypothesis that the magnitude of consumer (grazer) effects on prey (algae) depends on the species diversity of the prey assemblage. The analysis reveals that consumer effects tend to decrease as the diversity of a prey assemblage increases. This trend is robust for several different, yet complementary indices of grazer effect size and algal diversity. The trend also remains significant after statistically controlling for a variety of factors that can covary with prey diversity among studies. We discuss several possible mechanisms for the documented pattern, such as diversity enhancing the probability of inedibility and of positive interactions.     

Grazer identity changes the spatial distribution of cascading trophic effects in stream pools

Non-lethal effects of predators on prey behavior can mediate trophic cascades, but the extent of effects depends on habitat characteristics and risk sensitivity of prey. Furthermore, predation risk for stream organisms varies along the depth gradient and strongly influences their behavior. Grazing minnows (Campostoma anomalum) and crayfish (Orconectes virilis) are both prey for largemouth bass (Micropterus salmoides) in streams, but differ in their predator-avoidance behavior. This study contrasts the effects and mechanisms of non-lethal trophic cascades on the spatial distribution of filamentous green algae among stream pools and along a depth gradient within pools. Presence/absence of a largemouth bass was crossed with four combinations of the two grazer species (0 grazers, 30 minnows, 30 crayfish, and 15 each) in outdoor, experimental streams. Grazer densities were maintained by restocking. I used geostatistics to quantify spatial patterns of predator and grazer habitat use, height of filamentous algae in the water column, and spatial covariation of water depth with algal height, and depth with grazer habitat use. In streams with only minnows, bass were sedentary, and hid within tall algae in a single "bass pool". In pools with grazed algae, bass actively pursued prey within and among pools and used deeper water. This set up a hierarchy of risk to grazers along the depth gradient from bass in deep water to potential risk from terrestrial predators in shallow water. Thus, minnows were more sensitive than crayfish to predation risk from bass, but less sensitive than crayfish to risk from terrestrial predators. Minnows mediated cascades at the scale of whole pools by avoiding "bass pools", but only if crayfish were absent. Crayfish avoided potential interactions both with terrestrial predators and bass by grazing and burrowing in deeper water at night (when bass were inactive), and by hiding in burrows during daytime. Crayfish without burrows avoided bass and crayfish defending burrows by using shallow edges of pools as corridors, but did not graze there. Thus, crayfish-mediated cascades were limited to pool edges. Effects of grazer identity may extend to other consumers via modification of risk for biota that use filamentous algae as either foraging or refuge habitat.     

Seasonal Trophic Niche Shift and Cascading Effect of a Generalist Predator Fish

Few studies have examined how foraging niche shift of a predator over time cascade down to local prey communities. Here we examine patterns of temporal foraging niche shifts of a generalist predator (yellow catfish, Pelteobagrus fulvidraco) and the abundance of prey communities in a subtropical lake. We predicted that the nature of these interactions would have implications for patterns in diet shifts and growth of the predator. Our results show significant decreases in planktivory and benthivory from late spring to summer and autumn, whereas piscivory increased significantly from mid-summer until late autumn and also increased steadily with predator body length. The temporal dynamics in predator/prey ratios indicate that the predation pressure on zooplankton and zoobenthos decreased when the predation pressure on the prey fish and shrimps was high. Yellow catfish adjusted their foraging strategies to temporal changes in food availability, which is in agreement with optimal foraging theory. Meanwhile the decrease in planktivory and benthivory of yellow catfish enabled primary consumers, such as zooplankton and benthic invertebrates, to develop under low grazing pressure via trophic cascading effects in the local food web. Thus, yellow catfish shifts its foraging niche to intermediate consumers in the food web to benefit the energetic demand on growth and reproduction during summer, which in turn indirectly facilitate the primary consumers. In complex food webs, trophic interactions are usually expected to reduce the strength and penetrance of trophic cascades. However, our study demonstrates strong associations between foraging niche of piscivorous fish and abundance of prey. This relationship appeared to be an important factor in producing top-down effects on both benthic and planktonic food webs.     

Upwelling-like bottom intrusion enhances the pelagic–benthic coupling by a fish predator in a coastal food web

Upwelling regions where nutrients are transported from deep to surface waters are among the most productive in the oceans. Although it is well known that the upwelling affects fishery production through bottom-up trophic cascading, it remains unexplored how temporal variation in its intensity alters overall trophic energy flows within a focal food web. In the present study, we demonstrate that inter-annual variation in the intensity of upwelling-like bottom intrusion alters food web properties in coastal waters of the Uwa Sea by focusing on the levels of δ¹³C and δ¹⁵N for a demersal fish predator, Acropoma japonicum. This approach integrates information on prey–predator interactions. In the season following a stratification period when pelagic productivity is limited by nutrient availability, A. japonicum showed lower levels of δ¹³C in years with high bottom intrusion intensity than in those with low intensity. One possible cause for this isotopic depletion is that the bottom intrusion-induced nutrient supply enhances pelagic productivity and consequently facilitates a foraging shift by A. japonicum from ordinary benthic prey to supplementary pelagic prey with a lower δ¹³C. In conclusion, the increased intensity of bottom intrusion results in coupling of two major trophic energy flows, pelagic and benthic food chains, through the demersal predator’s foraging shift.     

Light and nutrients regulate energy transfer through benthic and pelagic food chains

The amount of energy flowing to top trophic levels depends on primary production and the efficiency at which it is converted to production at each trophic level. In aquatic systems, algal production is often limited by light and nutrients, and the nutritional quality of algae depends on the relative balance of these two resources. In this study, we used a mesocosm experiment to examine how light and nutrient variation affected food chain efficiency (FCE, defined as the proportion of primary production converted to top trophic level production), using a food web with benthic and pelagic food chains. We also related variation in benthic and pelagic efficiencies to the nutritional quality of primary producers, i.e. carbon:nitrogen:phosphorus stoichiometry. As predicted, pelagic and benthic FCEs were highest under low light/high nutrient conditions, the treatment with the best algal food quality, i.e. the lowest C:nutrient ratios. Pelagic FCE and pelagic herbivore efficiency (HE_P) were more responsive than benthic FCE to variation in light and nutrients. Furthermore, pelagic FCE and HE_P were highly correlated with algal C:P, suggesting ‘carryover effects’ of algal food quality on carnivores (larval fish) via effects on herbivore (zooplankton) quality. Benthic (tadpole) production was primarily explained by primary production rate, suggesting food quantity rather than quality drives their production. However, benthic FCE was also highest at low light/high nutrients and was significantly correlated with food quality. The stronger effect of food quality in mediating pelagic compared to benthic efficiencies, is consistent with differences in the stoichiometric mismatches between algae and consumers. Pelagic FCE and HE_P were more likely to be P‐limited, whereas benthic FCE was more likely N‐limited. This study is the first to examine both pelagic and benthic FCE within the same system, and highlights the importance of differential consumer needs in determining how food quality affects energy transfer efficiency.     

Ecosystem consequences of diversity depend on food chain length in estuarine vegetation

Biodiversity and food chain length each can strongly influence ecosystem functioning, yet their interactions rarely have been tested. We manipulated grazer diversity in seagrass mesocosms with and without a generalist predator and monitored community development. Changing food chain length altered biodiversity effects: higher grazer diversity enhanced secondary production, epiphyte grazing, and seagrass biomass only with predators present. Conversely, changing diversity altered top‐down control: predator impacts on grazer and seagrass biomass were weaker in mixed‐grazer assemblages. These interactions resulted in part from among‐species trade‐offs between predation resistance and competitive ability. Despite weak impact on grazer abundance at high diversity, predators nevertheless enhanced algal biomass through a behaviourally mediated trophic cascade. Moreover, predators influenced every measured variable except total plant biomass, suggesting that the latter is an insensitive metric of ecosystem functioning. Thus, biodiversity and trophic structure interactively influence ecosystem functioning, and neither factor's impact is predictable in isolation.