Interactions Among Predators and the Cascading Effects of Vertebrate Insectivores on Arthropod Communities and Plants Kailen A

Interactions among predators and the cascading effects of vertebrate insectivores on arthropod communities and plants Kailen A. Mooneya,1, Daniel S. Grunerb, Nicholas A. Barberc, Sunshine A. Van Baeld, Stacy M. Philpotte, and Russell Greenbergf aDepartment of Ecology and Evolutionary Biology, University of California, Irvine, CA 92697-2525; bDepartment of Entomology, University of Maryland, College Park, MD 20742-4454; cDepartment of Biology, University of Missouri, St. Louis, MO 63121; dSmithsonian Tropical Research Institute, Apartado 0843- 03092, Balboa, Ancon, Panama; eDepartment of Environmental Sciences, University of Toledo, Toledo, OH 43606; and fSmithsonian Migratory Bird Center, National Zoological Park, Washington, DC 20008

Communicated by Thomas W. Schoener, University of California, Davis, CA, March 4, 2010 (received for review August 5, 2009) Theory on trophic interactions predicts that predators increase plant arthropod predators and herbivores. Theory predicts that any di- biomass by feeding on herbivores, an indirect interaction called a rect negative effect of vertebrate insectivores on herbivores would trophic cascade. Theory also predicts that predators feeding on be counterbalanced, in part or in whole, by the simultaneous sup- predators, or intraguild predation, will weaken trophic cascades. Al- pression of the arthropod predators of those herbivores (12, 13). though past syntheses have confirmed cascading effects of terrestrial Consequently, predicting the net effects of vertebrate insectivores arthropod predators, we lack a comprehensive analysis for verte- on plants requires knowledge of their relative effects on both brate insectivores—which by virtue of their body size and feeding predatory and herbivorous arthropods (11). Resolving the func- habits are often top predators in these systems—and of how intra- tional role of vertebrate insectivores is increasingly important in guild predation mediates trophic cascade strength. We report here on light of their vulnerability to global change (14). a meta-analysis of 113 experiments documenting the effects of insec- Despite the expectation that intraguild predation dampens the tivorous birds, bats, or lizards on predaceous arthropods, herbivorous strength of trophic cascades, past meta-analyses have not tested this ECOLOGY arthropods, and plants. Although vertebrate insectivores fed as intra- prediction. Meta-analyses of experiments from terrestrial systems guild predators, strongly reducing predaceous arthropods (38%), have focused primarily on the top-down effects of predaceous they nevertheless suppressed herbivores (39%), indirectly reduced arthropods (Table 1) (2, 3). To the extent that these meta-analyses plant damage (40%), and increased plant biomass (14%). Further- included vertebrate insectivores, they emphasized effects on her- more, effects of vertebrate insectivores on predatory and herbivo- bivores, and the role of predatory arthropods as mediators of tro- rous arthropods were positively correlated. Effects were strongest on phic cascade strength has not been studied. Vance-Chalcraft et al. arthropods and plants in communities with abundant predaceous (15) investigated the consequences of intraguild predation for arthropods and strong intraguild predation, but weakincommunities herbivore suppression, but their review included few studies on depauperate in arthropod predators and intraguild predation. The vertebrate insectivores and did not address the indirect effects of naturally occurring ratio of arthropod predators relative to herbi- such dynamics for plants. Consequently, fundamental aspects of vores varied tremendously among the studied communities, and how vertebrate insectivores affect arthropods and plants remain the skew to predators increased with site primary productivity and unclear, despite extensive evidence for the importance of vertebrate in trees relative to shrubs. Although intraguild predation among arthropod predators has been shown to weaken herbivore suppres- insectivores (10, 16, 17). sion, we find this paradigm does not extend to vertebrate insecti- To redress these gaps, we conducted a meta-analysis of 113 vores in these communities. Instead, vertebrate intraguild preda- published studies from 63 publications in which vertebrate insectivores (birds, bats, or lizards) were removed, and their effects tion is associated with strengthened trophic cascades, and insecti- fi vores function as dominant predators in terrestrial plant-arthropod quanti ed not only for herbivores and plants but also for preda- communities. ceous arthropods (for a summary table and list of source publications, see Table S1). With these data, we tested a set of predic- bottom-up and top-down control | intraguild predation | meta-analysis | tions arising from theory on trophic interactions (12, 18) for the trophic cascade | vertebrate predator exclusion effects of vertebrate insectivores on arthropod predators (also referred to as intermediate predators), herbivores, and plants: (P1) generalist predators such as vertebrate insectivores should suppress esearch demonstrates that predators, by feeding on herbi- both predatory and herbivorous arthropods, supporting our asser- vores, can increase plant biomass via the indirect interaction R tion that they feed as intraguild predators; (P2) because of the commonly labeled a trophic cascade (1). In recent years, meta- buffering influence of intermediate predators, the effect strengths analyses have quantified trophic cascades separately in terrestrial of vertebrate insectivores on herbivores and plants should attenu- (2, 3) and aquatic systems (4, 5) and in multiple habitats together ate in communities where intermediate predators are more abun- to compare the strength of trophic cascades among ecosystem dant relative to herbivores; (P3) Because intermediate predators types (6). Although the strengths of trophic cascades vary across should diminish the indirect effects of vertebrate insectivores, the ecosystem types, explanations for the significant residual variation within ecosystems remain enigmatic (6–8). Vertebrate insectivores such as birds, bats, and lizards often Author contributions: K.A.M., D.S.G., N.A.B., S.A.V.B., and S.M.P. designed research; K.A.M., feed as top predators on terrestrial arthropod communities, but D.S.G., N.A.B., S.A.V.B., S.M.P., and R.G. performed research; K.A.M., D.S.G., and N.A.B. based upon current theory it is unclear whether their effects should analyzed data; and K.A.M., D.S.G., N.A.B., S.A.V.B., and S.M.P. wrote the paper. cascade down to affect plant biomass. Because of their large body The authors declare no conflict of interest. size relative to arthropod prey, vertebrate insectivores can con- Freely available online through the PNAS open access option. sume both predatory and herbivorous arthropods (9, 10). As a 1To whom correspondence should be addressed. E-mail: [email protected]. consequence, vertebrate insectivores may feed as so-called intra- This article contains supporting information online at www.pnas.org/cgi/content/full/ guild predators (11), simultaneously consuming intermediate 1001934107/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1001934107 PNAS Early Edition | 1 of 6 Table 1. Summary of results from published trophic cascade meta-analyses involving terrestrial communities Arthropods Plants Source Predators* Herbivores* Damage* Biomass*

(3) — −0.49 −0.95 +0.22 (−0.63 –0.35) (−1.18 –0.72) (+0.11 +0.33) 22% 32% 0% (2) — −0.41 −0.53 +0.12 (−0.49 –0.33) (−0.64 –0.42) (+0.03 +0.20) 37% 48% 10% † (6) — −0.44 — +0.11 (−0.72 –0.16) (−0.06 +0.27) 17% 17% (9) −0.71 −0.60 −0.32 — (−0.95 –0.49) (−0.90 –0.30) (−0.45 –0.20) 100% 100% 100% This study −0.49 −0.47 −0.34 +0.13 (−0.69 –0.29) (−0.66 –0.28) (−0.46 –0.22) (−0.02 +0.28) 100% 100% 100% 100%

*Values reported in each cell: log response ratios, 95% confidence intervals, and % of studies with vertebrate predators. †Based on studies measuring plant biomass of whole communities, limited to grassland and agricultural fields, found stronger effects of vertebrate predators (across terrestrial and aquatic communities). Other meta-analyses use responses of individual plant species and show no effect of predator type. net effects of vertebrate insectivores should be more strongly metric of arthropod community trophic composition under natural negative upon intermediate predators than herbivores, and these conditions. In communities deficient in intermediate predators (low effects should tradeoff such that strong effects on intermediate values of ln[IP:H+]), vertebrate insectivores had relatively weak predators are associated with weak effects on herbivores; (P4) as a effects on intermediate predators, herbivores and plants (Fig. 2 and result of this asymmetry in their net effects on intermediate pred- Table S4). In contrast, in communities with a high relative abun- ators and herbivores, vertebrate insectivores should alter the tro- dance of intermediate predators (high values of ln[IP:H+]) trophic phic composition of the communities on which they prey; and (P5) cascades were strong, with strong negative effects of vertebrate the results of our analysis should show relatively weak indirect effects on plant biomass as compared to past meta-analyses focused on experimental exclusions of arthropod predators that may more often feed directly upon herbivores without intraguild predation. Results and Discussion Tests of Predictions. Consistent with P1, vertebrate insectivores suppressed the abundance of both predaceous and herbivorous arthropods (Fig. 1 and values presented in Table S2). The meta- analysis of Van Bael et al. (9), which was based upon a subset of the data presented here (n = 48 studies), found parallel results for bird effects on predaceous and herbivorous arthropods in trees and understory shrubs. Although negative effects of vertebrate insectivores on arthropod predators may arise in part from competition for a shared prey base of herbivores, diet studies clearly demonstrate that vertebrate insectivores also directly consume predatory arthropods (10, 19, 20). This first result—that vertebrate insectivores act as intraguild predators, suppressing both intermediate predators and herbivores—thus sets the stage for our test of the subsequent predictions. Contrary to our second prediction (P2), trophic cascade strength was correlated with the relative abundance of intermediate predators to herbivores in a community (Fig. 2 andTable S4). Among the studies where authors reported comprehensive arthropod community sampling and assigned all arthropods to a trophic role, the relative abundance of intermediate predators (IP) to herbivores (H) varied by over three orders of magnitude, from herbivore- dominated communities with ratios (IP:H) as low as 1:167 to Fig. 1. Natural log ratio effect sizes (=ln[control / predator exclusion]; ±95% CI) of insectivorous birds, bats, and lizards on arthropods and plants. Hollow predator-dominated communities with values as high as 4:1, with a circles show effect means, dots show values from individual studies. Sample mean ratio of 1.0 ± 0.4 intermediate predators for every herbivore. size provided in parentheses. The bottom axis presents back-transformed We calculated the natural log of the ratio of intermediate predators effect sizes presented as untransformed proportional change. Summary to herbivores in the presence of vertebrate insectivores (exper- statistics provided in Table S2. For intermediate predators: herbivores, a imental control treatments), annotated here as ln(IP:H+), as a single study value of +2.2 is not shown.

2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1001934107 Mooney et al. strongest in systems assumed to have strong intraguild predation and weak trophic cascades. According to P3, we predicted that top predator effects should be more strongly negative upon intermediate predators than herbivores and that these effects should tradeoff and thus be neg- atively correlated. Contrary to P3, we found that the mean effects of vertebrate insectivores upon intermediate predators and herbivores were nearly identical in strength (Fig. 1). For the subset of studies that measured effects on both intermediate predators and herbivores, these effects were positively correlated and similar in magnitude (r = 0.41, n = 52, P = 0.0023; Fig. S1). Across all studies, vertebrate insectivores reduced intermediate predator and herbivore abundance by 38% [log response ratios (LRR) mean ± 95% CI:= −0.47 ± 0.19] and 39% (LRR = −0.49 ± 0.19), respectively (Fig. 1). Certainly, trophic levels are artiﬁcial con- structions for many arthropods, and there is an increasing recog- nition that many taxa feed as facultative or obligate omnivores (23). These trophic “tangles,” characterized by increasing trophic indeterminacy at higher positions, are prevalent across marine, freshwater, and terrestrial systems (24). Our assignment of all predatory arthropods to a single trophic role almost certainly mischaracterizes some of these omnivores. The equality of vertebrate insectivore effects on predatory and herbivorous arthropods may result, in part, from a gradient of variable feeding roles within each trophic construction. Yet vertebrate insectivores had strong effects of nearly identical magnitude on spiders (Araneae; 45%

reduction, LRR = −0.60 ± 0.21) and lepidopteran larvae (49% ECOLOGY reduction, LRR = −0.67 ± 0.26), two taxonomic groups that are abundant and unambiguously predaceous and herbivorous, respectively (Fig. 1). Consequently, a degree of imprecision in the assignment of trophic roles is unlikely to account for the symmetrical effects of vertebrate insectivores on predaceous and herbivorous arthropods. Because vertebrate insectivores caused effects of equivalent magnitude upon predaceous and herbivorous arthropods, they had little effect on the trophic composition of arthropod communities despite a strong overall effect on arthropod abundance (Fig. 1 and Fig. S1). This result contradicts P4, which states that vertebrate insectivores should alter community trophic composition because of the expected asymmetry of their effects on intermediate predators and herbivores. We used two complementary analyses to test whether vertebrate predators altered the trophic composition of arthropod communities. First, we compared the intermediate predator to herbivore ratio in the presence (IP:H+) and absence (IP:H−) of vertebrate predators in a log response ratio (ln[IP:H+/IP:H−]). This analysis did not reveal any effect of vertebrate predators on arthropod community trophic composition (Fig. 1). In our second approach, we inspected the correlation between the natural log of IP:H between treatments with and without vertebrate insectivores (ln[IP:H+] and ln[IP:H−], respectively). The relationship was positive (r = 0.85, n = 49, P < 0.0001) (Fig. 3) and the slope did not differ from 1.0 ( = 0.87 ± Fig. 2. Relationship between arthropod community composition (the log β ratio of intermediate predators to herbivores in the presence of vertebrate 0.15). Taken together, this evidence suggests that vertebrate insectivores = ln[IP:H+]) and the strength of vertebrate insectivore effects (log predators have symmetrical negative effects upon predaceous and response ratios, LRRs) on predaceous and herbivorous arthropods and plants. herbivorous arthropods (Figs. 1 and 3) despite the great under- Effects on plants are both effects on plant biomass and effects on plant dam- lying variation in trophic composition of the communities upon age. LRRs for plant damage are multiplied by −1 to make the effect directions which they fed (Fig. 3). parallel with those of plant biomass. Dashed vertical lines show arthropod Contrary to P5, our results for the effect magnitudes of top communities with an equal abundance of intermediate predators and herbi- vertebrate predators were highly consistent with those from pre- vores. Dashed horizontal lines show vertebrate insectivore effect sizes of 0. vious meta-analyses of terrestrial systems based largely upon the exclusion of predaceous arthropods (Table 1). Despite their strong negative effects on intermediate predators, vertebrate insectivores insectivores on both intermediate predators and herbivores and also suppressed herbivores (see above; Fig. 1) and exerted an strong positive effects on plants. Although food chain theory pre- overall beneﬁcial effect on plants by reducing plant damage by 40% dicts that the effects of vertebrate insectivores upon herbivores and (LRR = −0.34 ± 0.13) and there was a trend towards an increase in plants should attenuate or reverse with the increased abundance of plant biomass of 14% (LRR = 0.13 ± 0.15; Fig. 1). As noted in intermediate predators (21, 22), our results suggest the opposite: previously published meta-analyses (2, 3), plant damage responses the effects of vertebrate insectivores on herbivores and plants were were stronger than biomass responses (F1, 67 = 3.77, P = 0.0563),

Mooney et al. PNAS Early Edition | 3 of 6 of a system should result in progressive lengthening of food chains and increasing abundance at higher trophic levels. Empirical evidence also supports the hypothesis that bottom-up and top-down factors can interactively determine trophic structure (29, 30). To test this possibility, we investigated whether variation in primary productivity was associated with the relative abundance of intermediate predators and the strength of top predator effects on arthropods and plants. We used mean annual precipitation and the Normalized Difference Vegetation Index (NDVI) as indica- tors of primary productivity and also examined latitude (absolute value) with which mean annual precipitation and NDVI are neg- atively correlated. Although we found no evidence for an influence of precipitation or latitude, our analyses reveal that ln(IP:H+) increased significantly with NDVI (Fig. 4 and Table S3). We also found that ln(IP:H+) was significantly higher in trees than in shrubs (herbaceous plants yielded insufficient replication for analysis) (Fig. 4 and Table S3). If trees represent a higher-quality resource for arthropods than shrubs (e.g., because of higher pro- Fig. 3. Effect of vertebrate insectivores on arthropod community trophic ductivity in the canopy vs. understory) a similar productivity effect composition. Each datum shows the natural log of the intermediate predator: may also explain these differences. In combination, host plant herbivore ratio (= ln[IP:H]) in treatments with and without vertebrate insec- growth form and NDVI explained one-third of the overall varia- tivores. Symbols with error bars show mean community composition for shrubs tion in the relative abundance of intermediate predators to her- fi fi and trees ± 95% con dence intervals. Trees have a signi cantly higher ln(IP:H) bivores (Table S3). However, we failed to detect the predicted than shrubs (Table S3). There were only two studies on herbs. Data are presented with study as the replicate and multiple studies per field site. Dashed positive association between NDVI (or precipitation or latitude) vertical and horizontal lines show arthropod communities with an equal and the effects of vertebrate insectivores on arthropods and plants abundance of intermediate predators and herbivores. (Table S3). Consequently, our analyses provide mixed support for the hypothesis that variation in primary productivity drives the positive association between the abundance of intermediate but we found little evidence for a correlation between damage and predators and the strength of top predator effects on arthropods biomass responses among the 16 studies that measured both (r = and plants. 0.38, P = 0.16; Fig. S2). As with previous meta-analyses, effects on Second, an influx of allochthonous arthropods into structurally herbivores were stronger than effects on plants, suggesting a pro- complex plant canopies might simultaneously subsidize preda- gressive attenuation of top-down effects (25). ceous arthropods and provoke functional and numerical responses All of the above conclusions were robust with respect to the of vertebrate insectivores (31). As mobility and foraging behavior type of vertebrate insectivore studied, plant growth form, and tend to scale with trophic position (32), trophic subsidies might several aspects of experimental design. Due to sample size simultaneously increase the abundance of predaceous arthropods restrictions, we were only able to compare the strength of bird and the foraging intensity of vertebrate insectivores that in turn and bat effects (combined) with those of lizards, and only upon reduce herbivore abundance and benefit plants (13, 33). Preda- arthropods. We found no detectable effect of top predator type ceous arthropods were frequently present at densities that may not upon predatory or herbivorous arthropods, or for all arthropods be sustained on herbivore prey alone (Fig. 3), suggesting such combined (Table S3). Similarly, we found no difference in plant subsidies are likely important. Although some studies in our an- or arthropod responses based upon whether the studied plants alyses reported abundances of detritivores (Fig. 1), there were were trees or shrubs (Table S3), although insufficient data pre- insufficient data to rigorously test whether the relative abundance vented a comparison of these woody plants to herbs. Finally, of detritivores was correlated with the abundance of intermedi- there were no indications that arthropod or plant effects varied ate predators and the strength of vertebrate insectivores effects. as a function of the scale of predator exclusions, experimental Third, the effects of intermediate and vertebrate insectivores duration, or sample size used in the examined studies (Table S3). may operate synergistically to suppress herbivore abundance to Summarizing our results, we show that vertebrate insectivores the benefit of plants (34, 35). Under this hypothesis, inter- suppressed both intermediate predators and herbivores to an mediate predators may enhance the effects of vertebrate insec- equal extent and that the effects of these vertebrate insectivores tivores if each predator group forces herbivores to engage in on herbivores and plants strengthened with increasing relative behaviors that increase their vulnerability to attack by the other. abundance of intermediate predators. This result is especially Several studies have tested for interaction between the effects of noteworthy given the overwhelming evidence that arthropod vertebrate and arthropod predators, with synergistic effects predators and parasitoids themselves are capable of suppressing demonstrated in some (36) but not all systems (37, 38). Such herbivore populations (2, 3, 26). Theory on intraguild predation explanations would similarly explain the stronger effects of ver- predicts that intermediate predators should buffer and weaken the tebrate insectivores on intermediate predators if interactions net effects of top predators upon herbivores (21), and empirical among those intermediate predators increase their vulnerability evidence from arthropod food webs supports this prediction (15). to vertebrate insectivores. Indeed, functional diversity among Our contrary findings thus suggest that current models of trophic vertebrate insectivores within communities, in terms of body size, interactions do not adequately describe dynamics involving foraging strategy, foraging strata, and dietary preferences, can vertebrate insectivores. enhance prey suppression (9, 39). Fourth, the parallel effects of vertebrate insectivores on preda- Why Does Theory Fail for Vertebrate Insectivores? First, resource tory and herbivorous arthropods could arise through competition availability and primary productivity may cascade up food webs for a shared prey base of herbivores. None of the experiments (27), leading to a parallel increase in the abundance of inter- included in this meta-analysis documented the extent to which mediate predators and to an increase in the strength of top vertebrate insectivores reduced predatory arthropods directly via predator effects. In the exploitation ecosystems hypothesis, consumption versus indirectly by consuming shared herbivorous Oksanen et al. (28) proposed that increasing primary productivity prey. On Bahamian islands, experimental studies in our database

4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1001934107 Mooney et al. Fig. 4. Effects of NDVI (A) and host plant growth form (B) on the relative abundance of intermediate predators to herbivores in the presence of vertebrate insectivores (=ln [IP:H+]). (A) Filled circles indicate trees, hollow circles shrubs, x herbaceous plants. Data are presented with field site as the replicate, compiling multiple studies in some cases. (B) Hollow circles show effect means (± 95% CI), and dots show values from individual studies. Sample sizes are provided in parentheses. Data are presented with study as the replicate, and multiple studies per field site. that documented the negative effects of lizards on spiders (38, 40) will be critical to document not only the net effects of vertebrate were insufficient to disentangle these mechanisms, but parallel insectivores on predatory arthropods but also the extent to which mechanistic studies demonstrated these effects occur via both such effects are due to direct consumption versus competition for predation and competition (19). If vertebrate insectivores fed a shared prey base (40). To separate intraguild predation into its

exclusively upon herbivores, a parallel indirect effect on predatory competitive and predatory components, studies must determine ECOLOGY arthropods would be expected (Figs. 1 and 3). However, such an the overlap in diets of vertebrate insectivores and predatory explanation would not explain the strengthening of top-down arthropods in terms of prey taxonomic identity and size and also controls by vertebrate insectivores with the increasing relative document per capita herbivore consumption when vertebrate and abundance of predatory arthropods (Fig. 2). Most importantly, arthropod predators act alone and in combination (19, 43). there is extensive evidence that vertebrate insectivores do consume Second, studies must be conducted over significantly larger predatory arthropods (10, 19), and there is little reason to assume scales than those published to date. Although we observed no vertebrate insectivores would consume arthropods selectively with significant effects of the size of experimental units or the dura- respect to their trophic level. tion of studies (Table S3), we note that there are no published Although the above hypotheses seek to explain the positive studies carried out at a scale relevant to the home ranges of the association between intermediate predators and trophic cas- vertebrate insectivores or that ran for sufficient duration to cades strength, the pattern that intermediate predators did not encompass more than a single generation of the host plants. Our dampen herbivore suppression also requires exploration. We ability to “scale-up” results from relatively small-scale manipu- propose that vertebrate insectivores may, in fact, release herbi- lations remains unclear for most systems. vores from consumption by intermediate predators, but these Current theory successfully emulates the trophic dynamics relatively large and highly mobile vertebrates in turn switch operating within arthropod communities (15), but ourfindings show among trophic groups opportunistically and dynamically. Pre- they are not extensible to predict effects of vertebrate insectivores in vious studies suggest that many avian and mammalian insecti- these terrestrial systems. Vertebrate insectivores such as birds, bats, vores show frequency-dependent pressure on prey groups to the and lizards reduce the densities of both predaceous and herbivorous extent that those prey are abundant relative to other potential arthropods according to their proportional representation in the prey (41). If vertebrate insectivores dynamically switch to the arthropod community. Because terrestrial vertebrate insectivores most abundant trophic groups (42), this frequency-dependent are large and mobile, they may rapidly compensate for herbivore predation may equalize their effects on predatory and herbivo- release, thus leading to an emergent equality in their top-down rous prey, resulting in little net effect on the trophic composition effects (41, 42). However, other factors may lead to a positive of arthropod communities. Accordingly, the failure of intraguild association between intermediate predators and top-down control predation theory to predict vertebrate insectivore effects may from vertebrate insectivores, perhaps including—but not limited arise from aspects of their biology that facilitate dynamic to—the bottom-up effects of resource availability, trophic subsidies, switching between consumption of predaceous and herbivorous multiplicative predator effects, and competition for shared prey. As trophic groups, namely their large size relative to both arthropod a result of these dynamics, vertebrate insectivores play a key role in groups, high mobility, and behaviorally sophisticated foraging suppressing arthropods and increasing plant biomass that was not strategies that vary within and among individuals and species. anticipated by theory.

Future Directions. Although our analysis of the published literature Methods clearly documents patterns in the top-down cascading effects of We searched the literature for manipulativefield studies where (i) birds, bats, or vertebrate insectivores, limitations of the available data prevent lizards were experimentally excluded or enclosed withfield cages or netting, (ii) definitive conclusions on the precise mechanisms at work. We matched with appropriate open-access controls, and (iii) effects were measured outline multiple, nonmutually exclusive hypotheses that can only on terrestrial plants and with naturally occurring arthropod populations or communities. In some instances, multiple studies from a single publication were be tested with additional experimentation. In this regard, two treated as independent experiments (seeSI Appendix 1 for details). issues are of paramount importance. First, more studies are Tests for all hypotheses were based upon unweighted LRR (44), calculated needed that factorially manipulate both vertebrate insectivores from mean responses in the presence YI and absence Y I − of vertebrate fi ð þÞ ð Þ and arthropod predators to de nitively determine the role of insectivores ln Y I =YI − . We tested for effects of vertebrate insectivores on all ð ½ þ Þ intermediate predators as mediators of top-down effects. Here, it arthropods, arthropod trophic groups, abundant arthropod orders, arthropod

Mooney et al. PNAS Early Edition | 5 of 6 diversity, arthropod community trophic composition (defined as the relative arthropod community trophic composition as the natural log of the ratio of abundance of predaceous and herbivorous arthropods [IP:H]), plant damage intermediate arthropod predators to herbivores (ln[IP:H]), taking the value of by herbivores, and plant biomass. If a study included multiple responses of dif- this ratio in the presence of vertebrate insectivores (ln[IP:H+]) to be a descriptor fering units, we averaged the LRRs within that study to generate a single LRR. of each studied community. For 33 of the 49 studies the authors reported When a study reported effects over time, we used the final time point. comprehensive sampling of the arthropod community. The value of ln(IP:H+) For responses with sufficient sample sizes, we used general linear models did not differ based upon complete versus incomplete sampling (F1,48 = 2.56, P = to test for the effects of biological and methodological covariates (Table S3). 0.12). In 26 of these 49 studies, there were arthropods not assigned to a trophic We investigated whether LRRs were dependent upon methodological cova- group by the authors and that were trophically ambiguous based solely upon riates (exclusion area, sample size, experimental duration), host plant growth their taxonomy. In these studies, trophically ambiguous taxa constituted 45% form (herb, shrub, or tree), and insectivore taxon (birds and bats vs. lizards) of arthropod abundance or biomass. Here we based our analyses only upon with each study serving as the unit of replication (study-based covariates). In trophically assigned arthropods only. We did not detect a relationship be- separate analyses, we tested for an influence of latitude (absolute value), tween ln(IP:H+) and the proportion of the community consisting of unassigned mean annual precipitation, and site productivity (NDVI) using study sites as arthropods (r = 0.11, n = 26, P = 0.58). the replicate by averaging the responses of multiple studies as needed (site- − based covariates). Latitude and longitude were taken directly from papers or We investigated both the determinants of ln(IP:H+) and the association extracted using the software application Google Earth (Google Inc.). Data on between ln(IP:H+) and trophic cascade strength (Table S3 and S4, respectively). total annual precipitation were obtained from the WorldClim database (45) Tests for the determinants of ln(IP:H+) proceeded as described above, with and mean annual NDVI from the National Oceanic and Atmospheric separate tests for the influence of site- and study-based covariates on ln(IP:H+) Administration Advanced Very High Resolution Radiometer satellite (46). (Table S3). When testing for the effects of arthropod community upon trophic When investigating the influence of covariates on plant responses, we con- cascade strength we used linear regression with ln(IP:H+) as a predictor and ducted a single analysis in which biomass responses were preferentially taken LRRs for top predator effects as the dependent variables (Table S4). Results when available, but otherwise used plant damage responses multiplied by −1 were qualitatively identical when these analyses were repeated using the ratio to convert the sign of these effects to parallel those of plant biomass of intermediate arthropod predators to herbivores in the absence of verte- responses. In these models, we included an indicator term for response metric brate insectivores (ln[IP:H-]) or averaged across treatments. Finally, we com- (damage, biomass) to account for any difference between these measures. In pared the intermediate predator to herbivore ratio between treatments with all analyses we modeled LRRs as a linear function of all predictors and then (IP:H+) and without (IP:H−) vertebrate insectivores using the log response fi discarded any terms where P > 0.15 to arrive at a nal reduced model. Because ratio ln(IP:H+/IP:H−). of small sample sizes and difficulties in interpretation, we did not test for interactions among covariates in effects on LRRs. ACKNOWLEDGMENTS. We thank A. Hurlbert for providing NDVI data and Our analyses of arthropod community trophic composition werebased upon A. Agrawal, E. Borer, D. Spiller, O. Schmitz, T. Schoener, L. Yang, and two 49 studies where the abundance of intermediate predators and herbivores anonymous reviewers for critical comments and suggestions on the were reported in the same units and were thus comparable. We quantified manuscript.

1. Paine RT (1980) Food webs: Linkage, interaction strength and community infrastructure. 24. Thompson RM, Hemberg M, Starzomski BM, Shurin JB (2007) Trophic levels and J Anim Ecol 49:667–685. trophic tangles: The prevalence of omnivory in real food webs. Ecology 88:612–617. 2. Halaj J, Wise DH (2001) Terrestrial trophic cascades: How much do they trickle? Am 25. Schmitz OJ (2008) Herbivory from individuals to ecosystems. Annu Rev Ecol Evol Syst Nat 157:262–281. 39:133–152. 3. Schmitz OJ, Hambäck PA, Beckerman AP (2000) Trophic cascades in terrestrial systems: 26. Matsumoto T, Itioka T, Nishida T (2003) Cascading effects of a specialist parasitoid on A review of the effects of carnivore removals on plants. Am Nat 155:141–153. plant biomass in a Citrus agroecosystem. Ecol Res 18:651–659. 4. Micheli F (1999) Eutrophication, fisheries, and consumer-resource dynamics in marine 27. Kagata H, Ohgushi T (2006) Bottom-up trophic cascades and material transfer in pelagic ecosystems. Science 285:1396–1398. terrestrial food webs. Ecol Res 21:26–34. 5. Brett MT, Goldman CR (1996) A meta-analysis of the freshwater trophic cascade. Proc 28. Oksanen L, Fretwell SD, Arruda J, Niemela P (1981) Exploitation ecosystems in Natl Acad Sci USA 93:7723–7726. gradients of primary productivity. Am Nat 118:240–261. 6. Shurin JB, et al. (2002) A cross-ecosystem comparison of the strength of trophic 29. Gutierrez AP, Mills NJ, Schreiber SJ, Ellis CK (1994) A physiologically based tritrophic cascades. Ecol Lett 5:785–791. perspective on bottom-up-top-down regulation of populations. Ecology 75:2227–2242. 7. Shurin JB, Gruner DS, Hillebrand H (2006) All wet or dried up? Real differences 30. Gruner DS (2004) Attenuation of top-down and bottom-up forces in a complex – between aquatic and terrestrial food webs. Proc Biol Sci 273:1 9. terrestrial community. Ecology 85:3010–3022. 8. Borer ET, et al. (2005) What determines the strength of a trophic cascade? Ecology 86: 31. Nakano S, Murakami M (2001) Reciprocal subsidies: Dynamic interdependence between – 528 537. terrestrial and aquatic food webs. Proc Natl Acad Sci USA 98:166–170. 9. Van Bael SA, et al. (2008) Birds as predators in tropical agroforestry systems. Ecology 32. Eveleigh ES, et al. (2007) Fluctuations in density of an outbreak species drive diversity 89:928–934. cascades in food webs. Proc Natl Acad Sci USA 104:16976–16981. 10. Gunnarsson B (2007) Bird predation on spiders: Ecological mechanisms and evolutionary 33. Leroux SJ, Loreau M (2008) Subsidy hypothesis and strength of trophic cascades across consequences. J Arachnol 35:509–529. ecosystems. Ecol Lett 11:1147–1156. 11. Polis GA, Myers CA, Holt RD (1989) The ecology and evolution of intraguild predation: 34. Sih A, Englund G, Wooster D (1998) Emergent impacts of multiple predators on prey. Potential competitors that eat each other. Annu Rev Ecol Syst 20:297–330. Trends Ecol Evol 13:350–355. 12. Polis GA, Holt RD (1992) Intraguild predation: The dynamics of complex trophic 35. Schmitz OJ (2007) Predator diversity and trophic interactions. Ecology 88:2415–2426. interactions. Trends Ecol Evol 7:151–154. 36. Swisher BJ, Soluk DA, Wahl DH (1998) Non-additive predation in littoral habitats: 13. Polis GA, Strong DR (1996) Food web complexity and community dynamics. Am Nat Influences of habitat complexity. Oikos 81:30–37. 147:813–846. 37. Mooney KA (2007) Tritrophic effects of birds and ants on a canopy food web, tree 14. Şekercioğlu ÇH, et al. (2002) Disappearance of insectivorous birds from tropical forest growth, and phytochemistry. Ecology 88:2005–2014. fragments. Proc Natl Acad Sci USA 99:263–267. 38. Spiller DA, Schoener TW (1994) Effects of top and intermediate predators in a 15. Vance-Chalcraft HD, et al. (2007) The influence of intraguild predation on prey – suppression and prey release: A meta-analysis. Ecology 88:2689–2696. terrestrial food web. Ecology 75:182 196. 16. Sekercioglu CH (2006) Increasing awareness of avian ecological function. Trends Ecol 39. Philpott SM, et al. (2009) Functional richness and ecosystem services: Bird predation – Evol 21:464–471. on arthropods in tropical agroecosystems. Ecol Appl 19:1858 1867. 17. Whelan CJ, Wenny DG, Marquis RJ (2008) Ecosystem services provided by birds. Ann N 40. Spiller DA, Schoener TW (1990) A terrestrial experiment showing the impact of Y Acad Sci 1134:25–60. eliminating top predators on foliage damage. Nature 347:469–472. fl 18. Rosenheim JA (1998) Higher-order predators and the regulation of insect herbivore 41. Allen JA, Raison HE, Weale ME (1998) The in uence of density on frequency-dependent fi populations. Annu Rev Entomol 43:421–447. selection by wild birds feeding on arti cial prey. Proc Biol Sci 265:1031–1035. 19. Spiller DA, Schoener TW (1990) Lizards reduce food consumption by spiders: Mech- 42. Chesson PL (1984) Variable predators and switching behavior. Theor Popul Biol 26: anisms and consequences. Oecologia 83:150–161. 1–26. 20. Federico P, et al. (2008) Brazilian free-tailed bats as insect pest regulators in 43. Chase JM, et al. (2002) The interaction between predation and competition: A review transgenic and conventional cotton crops. Ecol Appl 18:826–837. and synthesis. Ecol Lett 5:302–315. 21. Holt RD, Polis GA (1997) A theoretical framework for intraguild predation. Am Nat 44. Hedges LV, Gurevitch J, Curtis PS (1999) The meta-analysis of response ratios in 149:745–764. experimental ecology. Ecology 80:1150–1156. 22. Daugherty MP, Harmon JP, Briggs CJ (2007) Trophic supplements to intraguild predation. 45. Hijmans RJ, et al. (2005) Very high resolution interpolated climate surfaces for global Oikos 116:662–677. land areas. Int J Climatol 25:1965–1978. 23. Coll M, Guershon M (2002) Omnivory in terrestrial arthropods: Mixing plant and prey 46. Hurlbert AH, Haskell JP (2003) The effect of energy and seasonality on avian species diets. Annu Rev Entomol 47:267–297. richness and community composition. Am Nat 161:83–97.

6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1001934107 Mooney et al. Supporting Information

Mooney et al. 10.1073/pnas.1001934107

Fig. S1. Relationships between the strength of vertebrate insectivore effects (log response ratio effect sizes) on predaceous and herbivorous arthropods. plant growth (LRR) Vertebrate insectivore effect on insectivore effect Vertebrate

Vertebrate insectivores effect on plant damage (LRR)

Fig. S2. Absence of a correlative relationship between log response ratios for plant damage and plant biomass from studies that measured both variables (n = 16, r = 0.38, P = 0.16).

Mooney et al. www.pnas.org/cgi/content/short/1001934107 1 of 2 Other Supporting Information Files

SI Appendix (DOC ).

Table S1. Analyzed studies, methodological and biological covariates, and dependent variables

Table S1.

Table S2. Data for insectivore effect sizes, sample sizes, and 95% conﬁdence intervals and P values depicted in Fig. 1

Table S2.

Table S3. Statistical tables for analysis of association between methodological and biological covariates with vertebrate insectivores effects (log response ratios) on arthropods and plants

Table S3.

Table S4. Statistical tables for analysis of the association between the ratio of intermediate predators to herbivores in the presence of vertebrate insectivores and the strength of effects upon arthropods and plants

Table S4.

Mooney et al. www.pnas.org/cgi/content/short/1001934107 2 of 2 1 APPENDIX 1: STUDY SELECTION CRITERIA 2 We searched the literature for manipulative field studies where (i) birds, bats or lizards 3 were experimentally excluded or enclosed with field cages or netting, (ii) matched with 4 appropriate open-access controls, and (iii) effects were measured on terrestrial plants and with 5 naturally occurring arthropod populations or communities. Examples of the basis for exclusion 6 of a study included the following: studies were based upon ‘natural experiments’ (or were 7 correlative in nature) (e.g., 1); manipulations did not use matched physical structures in the same 8 habitat, as with introductions of insectivores to islands compared to islands lacking insectivores 9 (e.g., 2); studies used acceptable manipulations but measured predation with experimentally 10 deployed sentinel prey rather than natural communities (e.g., 3); studies measured effects on 11 pollinators (e.g., 4); data from the same experimental study were used in more than one 12 publication (e.g., 5); and effects of birds were measured on aquatic (as opposed to terrestrial) 13 plants (e.g., 6). 14 Concerning methodological aspects of studies beyond these strict criteria above, we 15 adopted the view of Englund et al. (7) who argue for inclusive, rather than exclusive, study 16 selection criteria. Rather than eliminate studies from the database by imposing a priori 17 methodological standards, we included all studies meeting the above criteria, while noting salient 18 aspects of the design such as experimental duration (days), sample size (number of replicates) 19 and the size of experimental exclosures (area in m2). We thus quantified these covariates and 20 analyzed their importance relative to other predictors in models for all response variables. In 21 most cases, authors studied the effects of vertebrate insectivore exclusion upon a single plant 22 species. 23 Studies from a single publication were treated as independent when one or more of the 24 following conditions were met: (i) the effects of more than one top predator were measured 25 independently; (ii) the experiments were replicated across multiple field sites; or (iii) when at a 26 single site the effects of a top predator were measured across more than one ecological context or 27 crossed experimental factor (e.g., [8] where birds were excluded on trees with and without ants) Table S1. Analyzed studies, methodological and biological covariates, and dependent variables.

Factor(s) for splitting single publication into multiple annual studies with the specific level(s) of those factors given in duration exclusion precip. Citation* parentheses latitude longitude (months) area (m2) (mm) NDVI

1 -- 57.68 12.65 5 0.25 919 191.4

2 -- 63.55 20.22 2.5 2 616 171.6

3 -- 38.05 -90.55 5 4 1045 196.7

4 -- 47.32 -114.23 75 100 394 181.9

5 -- 36.60 -91.20 15 464 530 162.5

6 strata / fertilization (canopy, no fertilization) 19.30 -105.03 5 3 790 195.5

6 strata / fertilization (understory, no fertilization) 19.30 -105.03 5 1 790 195.5

6 strata / fertilization (canopy, fertilization) 19.30 -105.03 5 3 790 195.5

6 strata / fertilization (understory, fertilization) 19.30 -105.03 5 1 790 195.5

7 predator excluded (birds) 18.30 -66.55 4 50 1967 206.3

7 predator excluded (lizards) 18.30 -66.55 4 50 1967 206.3

8 field site (Whited) 47.47 -104.07 27 100 477 162.2

8 field site (Miles) 46.37 -105.08 27 83 440 165.0

9 -- 47.19 -113.51 1 1 522 185.0

10 field site (McCall) 44.89 -116.08 1 1 668 180.4 10 field site (Okanogan Highlands) 48.36 -119.58 1 1 308 177.2

11 -- 39.73 -123.65 2 2 1547 212.2

12 -- 18.32 -65.75 6 1 2335 205.4

13 -- -32.38 115.85 12 1 878 192.9

14 field site (urban) 33.44 -112.07 12 1 218 159.7

14 field site (desert remnant) 33.44 -112.07 12 1 218 159.7

14 field site (desert) 33.44 -112.07 12 1 218 159.7

15 -- 32.48 -106.77 15 2 259 155.2

16 host species / fertilization (Qercus prinus, unfertilized) 33.89 -83.27 6 1 1241 200.7

16 host species / fertilization (Quercus rubra, unfertilized) 33.89 -83.27 6 1 1241 200.7

16 host species / fertilization (Quercus prinus, fertilized) 33.89 -83.27 6 1 1241 200.7

16 host species / fertilization (Quercus rubra, fertilized) 33.89 -83.27 6 1 1241 200.7

17 field site (year 1) 47.54 -103.86 2 225 369 162.3

17 field site (year 2) 47.54 -103.86 2 225 369 162.3

18 field site (year 1) 31.53 -111.50 3 25 453 162.9

18 field site (year 2) 31.53 -111.50 3 25 453 162.9

19 -- 9.18 -79.83 1.3 3.2 2630 197.8

20 -- 16.12 -95.90 33 0.5 1165 198.9

21 field site (Inga shade) 15.28 -90.11 2 0.5 2695 206.0 21 field site (no shade) 15.28 -90.11 2 10 2695 206.0

21 field site (non-Inga shade) 15.28 -90.11 2 10 2695 206.0

22,23 fertilization (unfertilized) 19.66 -155.28 33 40 2484 198.6

22,23 fertilization (fertilized) 19.66 -155.28 33 40 2484 198.6

24 experiment / season (Year 1, Fall) 57.68 12.65 5 0.25 919 191.4

24 experiment / season (Year 1, Spring) 57.68 12.65 11 0.25 919 191.4

24 experiment / season (Year 2, Fall) 57.68 12.65 17 0.25 919 191.4

24 experiment / season (Year 2, Spring) 57.68 12.65 23 0.25 919 191.4

25 field site / host species (DES, Betula) 57.68 12.02 4 2 809 189.8

25 field site / host species (SSK, Betula) 57.68 11.93 4 2 759 189.8

25 field site/ host species (RUD, Betula) 57.68 11.91 4 2 753 189.8

25 field site / host species (DES, Quercus) 57.68 12.02 4 2 809 189.8

25 field site / host species (SSK, Quercus) 57.68 11.93 4 2 759 189.8

25 field site/ host species (RUD, Betula) 57.68 11.95 4 2 778 189.8

26 -- 43.93 -71.75 2 20 1232 198.1

27 -- 21.58 -158.13 1 0.64 859 201.9

28 -- 41.57 -101.69 1.3 100 653 166.5

29 field site (year 1 slope) 41.57 -101.69 1.3 100 653 166.5

29 field site (year 2 slope) 41.57 -101.69 1.3 100 653 166.5 29 field stie (year 1 valley) 41.57 -101.69 1.3 100 653 166.5

29 field site (year 2 valley) 41.57 -101.69 1.3 100 653 166.5

30 predator excluded (birds) 9.10 -79.51 2.5 2 2635 195.2

30 predator excluded (bats) 9.10 -79.51 2.5 2 2635 195.2

31 -- 4.47 118.23 1 1 2099 210.9

32 canopy cover (open) 36.29 -93.73 16 1 1152 203.5

32 canopy cover (closed) 36.29 -93.73 16 1 1152 203.5

33 -- 39.69 -120.46 16 5 787 189.2

34 -- 38.05 -90.55 18 1 1045 196.7

35 field site (xeric) -40.60 -71.08 15 0.8 672 171.3

35 field site (hydric) -40.62 -71.83 15 0.8 1509 209.3

36 -- 39.04 -76.79 0 1.44 1072 188.6

37 -- 51.70 5.31 3 3.14 783 194.0

38 ants (present) 39.10 -105.80 3 2 337 169.4

38 ants (absent) 39.10 -105.80 3 2 337 169.4

39 -- 39.10 -105.80 36 4 346 166.8

40 ants (present) 39.10 -105.80 27 6 350 166.8

40 ants (absent) 39.10 -105.80 27 6 350 166.8

41 -- 42.67 141.60 1 1 1146 182.9 42 -- 42.70 141.60 2 225 1123 188.2

43 field site (riparian) 42.72 141.60 1 3.14 1117 188.2

43 field site (upland) 42.72 141.60 1 3.14 1117 188.2

44 -- 17.49 -62.98 6 144 1153 173.5

45 agricultural practice (traditional polyculture) 15.25 -92.39 18 80 3209 187.1

45 agricultural practice (commercial polyculture 1) 15.18 -92.37 18 80 3718 187.1

45 agricultural practice (commercial polyculture 2) 15.26 -92.39 18 80 3013 187.1

45 agricultural practice (shade monoculture) 15.18 -92.32 18 80 3645 190.9

46 experiment (dry season) 15.18 -92.37 2 1 3718 187.1

46 experiment (wet season) 15.18 -92.37 2.5 1 3718 187.1

47 agricultural practice (traditional polyculture) 15.25 -92.39 25 80 3209 187.1

47 agricultural practice (commercial polyculture 1) 15.18 -92.37 25 80 3718 187.1

47 agricultural practice (commercial polyculture 2) 15.26 -92.39 25 80 3013 187.1

47 agricultural practice (shade monoculture) 15.18 -92.32 25 80 3645 190.9

48 -- -32.75 116.92 19 517 192.0

49 -- 45.43 -93.21 90 81 749 175.5

50 host species / fertliization (Salix phylicifolia, unfertilized) 61.55 29.55 3 4 643 176.8

50 host species / fertliization (Salix myrsinifolia, unfertilized) 61.55 29.55 3 4 643 176.8

50 host species / fertliization (Salix phylicifolia, fertilized) 61.55 29.55 3 4 643 176.8 50 host species / fertliization (Salix myrsinifolia, fertilized) 61.55 29.55 3 4 643 176.8

51 field site (1971) 51.43 -2.66 4 1 853 197.1

51 field site (1972) 51.43 -2.66 6 1 853 197.1

52 -- 24.18 -76.44 45 83.6 1254 132.7

53,54 -- 24.18 -76.44 37 83.6 1254 132.7

55 -- 24.18 -76.44 20 83.6 1254 132.7

56 fertilization (unfertilized) 64.23 19.77 37 4 581 183.6

56 fertilization (fertilized) 64.23 19.77 37 4 581 183.6

57 -- 43.94 -71.75 25 10 1223 198.1

58 field site (Robinson Creek) 48.35 -120.12 1 0.75 383 175.0

58 field site (Twisp River) 48.35 -120.12 1 0.75 383 175.0

58 field site (Bill Smith Place) 48.35 -120.12 1 0.75 383 175.0

59 -- 45.24 -73.56 14.5 42 933 173.7

60 field site / strata (San Lorenzo, canopy) 9.17 -79.58 9 1 3188 192.9

60 field site / strata (Panama City, canopy) 8.59 -79.33 9 1 1866 137.9

60 field site / strata (San Lorenzo, understory) 9.17 -79.58 9 1 3188 192.9

60 field site / strata (Panama City, understory) 8.59 -79.33 9 1 1866 137.9

61 -- 9.20 -82.28 13 36 2689 208.2

62 experiment (1982) 43.41 -121.08 1 4 282 159.4 62 experiment (1983) 43.41 -121.08 13 4 282 159.4

63 predator excluded (birds) 15.18 -92.37 1.75 1 3209 187.1

63 predator excluded (bats) 15.18 -92.37 1.75 1 3209 187.1

"#$%&!'(!)*+,-+.&/!

host plant LRR LRR excluded arthropod growth LRR pred. herbiv. LRR total arthropod LRR plant LRR plant Citation* predator sampling form arthropods arthropods arthropods diversity ln(IP:H+) damage biomass

1 bird partial tree -1.10153 -1.10153

2 bird partial shrub -0.89517 -0.89517 1.03799

3 bird complete tree 0.27975 0.02408 0.09223 -0.88563 0.26125

4 bird partial herb 0.36445 0.14901 -0.41358

5 bird partial herb -0.94764 0.06701 0.47199 0.03337

6 bird partial tree -0.92792 -0.92792 0.26977 0.69622 -0.33958

6 bird partial tree -0.22051 -0.22051 0.0854 0.52991 -0.39749

6 bird partial tree -0.4634 0.14046 0.34741

6 bird partial tree -0.33052 -0.13661 0.38872

7 bird partial shrub -0.03977 -1.45225 -0.6356 0.4034 7 lizard partial shrub -0.16004 -0.45564 -0.39278 -0.71349

8 bird partial herb 0.31475 0.17563 -0.16314

8 bird partial herb -0.22314 -0.0714 0.02028

9 bird partial tree -0.27202 -0.27202

10 bird partial tree -0.09121 -0.09121

10 bird partial tree 0.06006 0.06006

11 lizard partial herb -1.22671 -1.22671 0.77428

12 lizard complete tree -0.48143 -0.3959 -0.01479 1.44186 0.57054

13 bird complete tree -0.48643 -2.69366 -0.39526 -0.29424 2.47293 1.09861

14 bird partial shrub 0.34657 -0.36205 -2.89202

14 bird partial shrub 0.28798 -0.69848 -2.74613

14 bird partial shrub -0.07688 0 -1.99586

15 bird complete shrub -0.4284 -0.70439 -0.70054

16 bird complete tree -0.02303 -0.20459 -0.06908 -2.2083 -0.24856

16 bird complete tree 1.05919 1.18452 1.45063 -1.43631 -0.13043

16 bird complete tree 1.26642 0.75045 0.85196 -2.54314 -0.29207

16 bird complete tree -1.61181 -2.24752 -2.44074 -1.30815 -0.30191

17 bird partial herb -0.36397 -0.0267

17 bird partial herb -0.32769 -0.00173 18 bird partial herb -0.78338 -0.78338

18 bird partial herb -1.06108 -1.06108

19 bird complete herb -0.50035

20 bird complete tree -0.63325 -1.64728 -1.36214 1.12214

21 bird complete shrub -0.67156 -0.11468 -1.41454 -1.60516 0.07696

21 bird complete shrub -0.46852 -2.37928 -1.34791 -0.89597 0.24583

21 bird complete shrub -0.21825 -0.25611 -1.37364 -0.81364 0.16616

22,23 bird complete tree 1.17449 0.85925 0.32311 0.49674 0.63291 0.7135

22,23 bird complete tree -1.03483 -0.07677 -0.44291 1.14079 -0.61018 -0.22201

24 bird complete tree -0.5988 -0.2393

24 bird complete tree -1.32797 -1.33634

24 bird complete tree -0.77367 -0.76228

24 bird complete tree -0.84852 -0.88692

25 bird complete tree -1.8092 -2.62995 -2.13423 -0.05892

25 bird complete tree -2.15211 1.45395 -1.66426 0.22127

25 bird complete tree -1.15896 0.8679 -0.53114 -0.81887

25 bird complete tree -0.45923 -0.3269 -0.67663 2.78336

25 bird complete tree -1.7642 -3.19867 -1.87262 4.61677

25 bird complete tree -2.06199 -1.68287 -2.38465 2.42985 26 bird partial tree 0.03704 -0.26663 -0.05884 0.42381

27 bird partial herb -0.03175 -0.76072 -0.97641 0.75984 0.78846 0.17728

28 bird partial herb -0.33273 -0.33273 -0.15459

29 bird partial herb 0.0019 0.0019 -0.07663

29 bird partial herb -0.17365 -0.17365 -0.16525

29 bird partial herb -0.32653 -0.32653 -0.11194

29 bird partial herb -0.1186 -0.1186 -0.03435

30 bat complete shrub -0.92846 1.12915

30 bird complete shrub -0.50263 0.51547

31 bird none tree 1.08463

32 bird partial tree -1.60944 0.95551 0.33647 -2.56495

32 bird partial tree -0.64185 0.24116 -2.53828 -1.02962

33 bird partial tree 0.21883 0.23224 -0.01754

34 bird partial tree -0.52609 -0.57596 0.41253 0.18555

35 bird partial tree -0.78897 0.66932

35 bird partial tree -0.66732 0.79511

36 bird partial tree 0.44992

37 bird none tree 0.20875 0.51212

38 bird complete tree -0.69477 -0.70821 -0.69759 -0.0899 38 bird complete tree -0.6026 -0.77225 -0.72075 -0.71948

39 bird complete tree -1.07212 -2.80215 -1.72108 1.4061 0.22314 0.13149

40 bird complete tree -0.54537 -1.08896 -1.49172 -1.11521 0.09022 0.39335

40 bird complete tree -0.41409 -0.5058 -0.47753 -0.76301 0.04228 0.48347

41 bird partial tree -0.26606

42 bird partial tree -0.25465 -0.55144 -0.40305 1.01419

43 bird partial shrub -0.3043 -0.3043

43 bird partial shrub -0.06871 -0.06871

44 lizard partial shrub -0.7427

45 bird complete shrub -0.11297 -0.39249 -0.14272 0.24207 -0.16487

45 bird complete shrub -0.49923 0.14983 0.05714 -0.70486 0.62933

45 bird complete shrub -0.13393 -0.60957 -0.10037 0.5852 0.17829

45 bird complete shrub -0.12427 -0.31439 0.20874 -0.23768 0.3406

46 bird complete tree -1.23661 -0.59989 -0.467 0.3394

46 bird complete tree -0.68534 -1.20127 -0.89088 0.26548

47 bird partial shrub 0.04445

47 bird partial shrub 0.07637

47 bird partial shrub -0.12014

47 bird partial shrub 0.15123 48 bird complete tree -0.47156 -0.8473 -0.41443 1.78557

49 bird partial herb -0.47957 -0.30142 -4.0955

50 bird partial tree 1.93516 1.93516 0.6888 0.44709

50 bird partial tree -0.09196 -0.09196 -0.06565 0.03306

50 bird partial tree 0.06131 -0.15619 0.32159 0.1148

50 bird partial tree -0.25465 -0.25465 0.55016 0.04011

51 bird partial herb -2.97593

51 bird partial herb -3.01821

52 lizard partial shrub 0.93288 0.13469

53,54 lizard partial shrub -0.83292 -0.72246 -0.46597 -0.29151 1.37685

55 lizard partial shrub -1.30731 -0.0873 -1.07389 -0.26399

56 bird none shrub 0.33715

56 bird none shrub 0.5866

57 bird partial tree -0.09685 0.30099 0.10598

58 bird partial tree -1.08011

58 bird partial tree -1.28048

58 bird partial tree -0.09081

59 bird partial herb -0.00768 0.06016

60 bird + bat partial tree -0.14601 0.15123 -0.04508 -0.40109 0.11341 60 bird + bat partial tree -0.59168 -0.01439 -0.16959 -1.44382 0.6213

60 bird + bat partial tree -0.11635 0.05804 -0.00455 0.71361 0.17917

60 bird + bat partial tree 0.24621 0.16645 0.13184 -0.3824 -0.13134

61 bird + bat partial tree -0.07197 -0.54459 -0.11204 0.84663 0.44306 0.05033

62 bird complete shrub 0.71138 0.08669 0.09714 0.10461 -5.12253

62 bird complete shrub -0.12173 0.15446 -0.10744 -0.0981 -2.96377

63 bird partial shrub -0.33153

63 bat partial shrub -0.27554

"#$%&'(%)'&!*+)',&*!

1. Askenmo C, von Bromssen A, Ekman J, Jansson C (1977) Impact of some wintering birds on spider abundance in spruce. Oikos 28:90-94.

2. Atlegrim O (1989) Exclusion of birds from bilberry stands: impact on insect larval density and damage to the bilberry. Oecologia 79:136- 139.

3. Barber NA, Marquis RJ (2009) Spatial variation in top-down direct and indirect effects on white oak (Quercus alba L.). Am Midl Nat 162:169-179.

4. Belovsky GE, Slade JB (1993) The role of vertebrate and invertebrate predators in a grasshopper community. Oikos 68:193-201.

5. Bock CE, Bock JH, Grant MC (1992) Effects of bird predation on grasshopper densities in an Arizona grassland. Ecology 73:1706-1717.

6. Boege K, Marquis RJ (2006) Plant quality and predation risk mediated by plant ontogeny: consequences for herbivores and plants. Oikos 115:559-572.

7. Borkhataria RRC, J A, Groom MJ (2006) Additive effects of vertebrate predators on insects in a Puerto Rican coffee plantation. Ecol Appl 16:696-703.

8. Branson DH (2005) Direct and indirect effects of avian predation on grasshopper communities in northern mixed-grass prairie. Environ Entomol 34:1114-1121.

9. Campbell RW, Torgersen TR (1983) Effect of branch height on predation of western spruce budworm (Lepidoptera: Tortricidae) pupae by birds and ants. Environ Entomol 12:697-699.

10. Campbell RW, Torgersen TR, Srivastava N (1983) A suggested role for predaceous birds and ants in the population dynamics of the western spruce budworm. For Sci 29:779-790.

11. Chase JM (1998) Central-place forager effects on food web dynamics and spatial pattern in northern California meadow. Ecology 79:1236-1245.

12. Dial R, Roughgarden J (1995) Experimental removal of insectivores from the rain forest canopy: direct and indirect effects. Ecology 76:1821-1834.

13. Evelegh NCP, Majer JD, Recher HF (2001) The effects of reducing bird predation on canopy arthropods of marri (Eucalyptus calophylla) saplings on the Swan Coastal Plain, Western Australia. J Roy Soc W Aust 84:13-21.

14. Faeth SH, Warren PS, Shochat E, Marussich WA (2005) Trophic dynamics in urban communities. BioScience 55:399-407.

15. Floyd T (1996) Top-down impacts on creosotebush herbivores in a spatially and temporally complex environment. Ecology 77:1544- 1555.

16. Forkner RE, Hunter MD (2000) What goes up must come down? Nutrient addition and predation pressure on oak herbivores. Ecology 81:1588-1600.

17. Fowler AC, Knight RL, George TL, McEwen LC (1991) Effects of avian predation on grasshopper populations in North Dakota grasslands. Ecology 72:1775-1781.

18. Gardner KT, Thompson DC (1998) Influence of avian predation on a grasshopper (Orthoptera: Acrididae) assemblage that feeds on threadleaf snakeweed. Environ Entomol 27:110-116.

19. Gradwohl J, Greenberg R (1982) The effect of a single species of avian predator on the arthropods of aerial leaf litter. Ecology 63:581- 583.

20. Greenberg R, Ortiz JS (1994) Interspecific defense of pasture trees by wintering yellow warblers. Auk 111:672-682.

21. Greenberg R, et al. (2000) The impact of avian insectivory on arthropods and leaf damage in some Guatemalan coffee plantations. Ecology 81:1750-1755.

22. Gruner DS (2004) Attenuation of top-down and bottom-up forces in a complex terrestrial community. Ecology 85:3010-3022.

23. Gruner DS, Taylor AD (2006) Richness and species composition of arboreal arthropods affected by nutrients and predators: a press experiment. Oecologia 147:714-724.

24. Gunnarsson B (1996) Bird predation and vegetation structure affecting spruce-living arthropods in a temperate forest. J Anim Ecol 65:389-397.

25. Gunnarsson B, Hake M (1999) Bird predation affects canopy-living arthropods in city parks. Can J Zool 77:1419-1428.

26. Holmes RT, Schultz JC, Nothnagle P (1979) Bird predation on forest insects: an exclosure experiment. Science 206:462-463.

27. Hooks CRR, Pandey RR, Johnson MW (2003) Impact of avian and arthropod predation on lepidopteran caterpillar densities and plant productivity in an ephemeral agroecosystem. Ecol Entomol 28:522-532.

28. Joern A (1986) Experimental study of avian predation on coexisting grasshopper populations (Orthoptera: Acrididae) in a sandhills grassland. Oikos 46:243-249.

29. Joern A (1992) Variable impact of avian predation on grasshopper assemblies in sandhills grassland. Oikos 64:458-463.

30. Kalka MB, Smith AR, Kalko EKV (2008) Bats limit arthropods and herbivory in a tropical forest. Science 320:71.

31. Koh LP (2008) Birds defend oil palms from herbivorous insects. Ecol Appl 18:821-825.

32. Lichtenberg JS, Lichtenberg DA (2002) Weak trophic interactions among birds, insects and white oak saplings (Quercus alba). Am Midl Nat 148:338-349.

33. Low C, Connor EF (2003) Birds have no impact on folivorous insect guilds on a montane willow. Oikos 103:579-589.

34. Marquis RJ, Whelan CJ (1994) Insectivorous birds increase growth of white oak through consumption of leaf-chewing insects. Ecology 75:2007-2014.

35. Mazia CN, Kitzberger T, Chaneton J (2004) Interannual changes in folivory and bird insectivory along a natural productivity gradient in northern Patagonia forests. Ecography 27:29-40.

36. Medina RF, Barbosa P (2002) Predation of small and large Orgyia leucostigma (J. E. Smith) (Lepidoptera: Lymatriidae) larvae by vertebrate and invertebrate predators. Environ Entomol 31:1097-1102.

37. Mols CMM, Visser ME (2002) Great tits can reduce caterpillar damage in apple orchards. J Appl Ecol 39:888-899.

38. Mooney KA (2006) The disruption of an ant-aphid mutualism by birds increases effects on pine herbivores. Ecology 87:1805-1815.

39. Mooney KA, Linhart YB (2006) Contrasting cascades: insectivorous birds increase pine but not parasitic mistletoe growth. J Anim Ecol 75:350-357.

40. Mooney KA (2007) Tritrophic effects of birds and ants on a canopy food web, tree growth, and phytochemistry. Ecology 88:2005-2014.

41. Murakami M (1999) Effect of avian predation on survival of leaf-rolling lepidopterous larvae. Res Popul Ecol 41:135-138.

42. Murakami M, Nakano S (2000) Species-specific bird functions in a forest-canopy food web. Proc R Soc Lond Ser B-Biol Sci 267:1597- 1601.

43. Murakami M, Nakano S (2002) Indirect effect of aquatic insect emergence on a terrestrial insect population through by birds predation. Ecol Lett 5:333-337.

44. Pacala S, Roughgarden J (1984) Control of arthropod abundance by Anolis lizards on St. Eustatius (Neth. Antilles). Oecologia 64:160- 162.

45. Perfecto I (unpublished data).

46. Philpott SM, Greenberg R, Bichier P, Perfecto I (2004) Impacts of major predators on tropical agroforest arthropods: comparisons within and across taxa. Oecologia 140:140-149.

47. Philpott SM (unpublished data).

48. Recher HF, Majer JD (2006) Effects of bird predation on canopy arthropods in wandoo Eucalyptus wandoo woodland. Austral Ecol 31:349-360.

49. Ritchie ME (2000) Nitrogen limitation and trophic vs. abiotic influences on insect herbivores in a temperate grassland. Ecology 81:1601- 1612.

50. Sipura M (1999) Tritrophic interactions: willows, herbivorous insects and insectivorous birds. Oecologia 121:537-545.

51. Solomon ME, Glen DM, Kendall DA, Milsom NF (1976) Predation of overwintering larvae of codling moth (Cydia pomonella (L.)) by birds. J Appl Ecol 13:341-351.

52. Spiller DA, Schoener TW (1990) A terrestrial experiment showing the impact of eliminating top predators on foliage damage. Nature 347:469-472.

53. Spiller DA, Schoener TW (1994) Effects of top and intermediate predators in a terrestrial food web. Ecology 75:182-196.

54. Spiller DA, Schoener TW (1998) Lizards reduce spider species richness by excluding rare species. Ecology 79:503-516.

55. Spiller DA, Schoener TW (1988) An experimental study of the effect of lizards on web-spider communities. Ecol Mono 58:57-77.

56. Strengbom J, Witzell J, Nordin A, Ericson L (2005) Do multitrophic interactions override N fertilization effects on Operophtera larvae? Oecologia 143:241-250.

57. Strong AM, Sherry TW, Holmes RT (2000) Bird predation on herbivorous insects: indirect effects on sugar maple seedlings. Oecologia 125:370-379.

58. Torgerson TR, Campbell RW (1982) Some effects of avian predators on the western spruce budworm in North Central Washington. Environ Entomol 11:429-431.

59. Tremblay A, Mineau P, Stewart RK (2001) Effects of bird predation on some pest insect populations in corn. Agr Ecosys Environ 83:143-152.

60. Van Bael SA, Brawn JD (2005) The direct and indirect effects of insectivory by birds in two contrasting Neotropical forests. Oecologia 143:106-116.

61. Van Bael SA (unpublished data).

62. Wiens JA, et al. (1991) Arthropod dynamics on sagebrush (Artemisia tridentata): effects of plant chemistry and avian predation. Ecol Mono 61:299-321.

63. Williams-Guillen K, Perfecto I, Vandermeer J (2008) Bats limit insects in a Neotropical agroforestry system. Science 320:70.

! Table S3. Statistical tables for analysis of association between methodological and biological covariates with vertebrate insectivores effects (log response ratios) on arthropods and plants.

Note: For each dependent variable, separate analyses are presented for study-based and field site-based covariates, and a reduced model is analyzed for those factors with P < 0.15 in the initial model.

S3-A. Dependent variable = LRR predatory arthropods

Study-based covariates

Source DF Sum Squares Mean Square F P

Model 6 3.9776 0.6629 1.28 0.2861

Error 49 25.4753 0.5199

Corrected total 55 29.4529

Experimental duration 1 0.0745 0.0745 0.14 0.7067

Exclusion area 1 0.7079 0.7079 1.36 0.2489

Sample size 1 1.3173 1.3173 2.53 0.1179

Bird & bat vs. lizard 1 0.0627 0.0627 0.12 0.7299

Host plant growth form 2 1.2689 0.6344 1.22 0.304

Reduced model based on P<0.15

Source DF Sum Squares Mean Square F P

Sample size 1 1.1257 1.1257 2.19 0.145

Error 55 28.3275 0.5150

Corrected total 56 29.4532

Field site-based covariates

Source DF Sum Squares Mean Square F P

Model 3 1.0153 0.3384 1.61 0.2124

Error 25 5.2589 0.2104

Corrected total 28 6.2742

Precipitation 1 0.0917 0.0917 0.44 0.5152

NDVI 1 0.1133 0.1133 0.54 0.4698

Latitude (absolute value) 1 0.8125 0.8125 3.86 0.0606

Reduced model based on P<0.15

Source DF Sum Squares Mean Square F P

Latitude (absolute value) 1 0.7312 0.7312 3.56 0.0699

Error 27 5.5430 0.2053

Corrected total 28 6.2742

Note, P=0.87 if all studies by Gunnarsson and colleagues are averaged to single datum.

S3-B. Dependent variable = LRR herbivorous arthropods

Study site as unit of replication

Source DF Sum Squares Mean Square F P

Model 6 1.3911 0.2319 0.3 0.9347

Error 86 66.2126 0.7699

Corrected total 92 67.6038

Experimental duration 1 1.1567 1.1567 1.5 0.2237

Exclusion area 1 0.0001 0.0001 0 0.9894

Sample size 1 0.0339 0.0339 0.04 0.8343 Bird & bat vs. lizard 1 0.0004 0.0004 0 0.983

Host plant growth form 2 0.0050 0.0025 0 0.9967

Field site-based covariates

Source DF Sum Squares Mean Square F P

Model 3 0.9158 0.3053 0.81 0.4966

Error 49 18.5583 0.3787

Corrected total 52 19.4740

Precipitation 1 0.5922 0.5922 1.56 0.2171

NDVI 1 0.3830 0.3830 1.01 0.3195

Latitude (absolute value) 1 0.4623 0.4623 1.22 0.2746

S3-C. Dependent variable = LRR total arthropods

Study-based covariates

Source DF Sum Squares Mean Square F P

Model 6 1.3760 0.2293 0.37 0.8931

Error 71 43.5123 0.6128

Corrected total 77 44.8882

Experimental duration 1 0.0117 0.0117 0.02 0.8904

Exclusion area 1 1.1107 1.1107 1.81 0.1825

Sample size 1 0.1064 0.1064 0.17 0.6782

Bird & bat vs. lizard 1 0.1826 0.1826 0.3 0.5869

Host plant growth form 2 0.1715 0.0857 0.14 0.8697

Field site-based covariates

Source DF Sum Squares Mean Square F P

Model 3 1.0547 0.3516 1.65 0.1954

Error 35 7.4547 0.2130

Corrected total 38 8.5094

Precipitation 1 0.7902 0.7902 3.71 0.0622

NDVI 1 0.4441 0.4441 2.09 0.1576

Latitude (absolute value) 1 0.3002 0.3002 1.41 0.2431

Reduced model based on P<0.15

Source DF Sum Squares Mean Square F P

Precipitation 1 0.1611 0.1611 0.71 0.4035

Error 37 8.3483 0.2256

Corrected total 38 8.5094

S3-D. Dependent variable =Ln(predatory arthropods: herbivorous arthropods)

Study-based covariates

Source DF Sum Squares Mean Square F P

Model 5 43.3866 8.6773 4.1 0.0038

Error 44 93.1654 2.1174

Corrected total 49 136.5520

Experimental duration 1 0.1973 0.1973 0.09 0.7616

Exclusion area 1 3.8027 3.8027 1.8 0.1871 Sample size 1 5.0851 5.0851 2.4 0.1284

Bird & bat vs. lizard 1 0.0848 0.0848 0.04 0.8423

Host plant growth form 1 21.1692 21.1692 10 0.0028

Note, host plant growth form limited to shrub and tree because of insufficient sample size to include herbs.

Reduced model based on P<0.15

Source DF Sum Squares Mean Square F P

Model 2 41.5358 20.7679 10.01 0.0002

Error 48 99.5673 2.0743

Corrected total 50 141.1031

Sample size

Host plant growth form 1 18.9175 18.9175 9.12 0.004

Field site-based covariates

Source DF Sum Squares Mean Square F P

Model 3 12.7849 4.2616 1.67 0.1975

Error 27 69.0129 2.5560

Corrected total 30 81.7977

Precipitation 1 0.0179 0.0179 0.01 0.9338

NDVI 1 11.0378 11.0378 4.32 0.0473

Latitude (absolute value) 1 0.7302 0.7302 0.29 0.5974

Reduced model based on P<0.15

Source DF Sum Squares Mean Square F P

NDVI 1 11.6213 11.6213 4.8 0.0366 Error 29 70.1764 2.4199

Corrected total 30 81.7977

S3-E. Dependent variable = LRR plants

Biomass and damage responses included in single analysis, with biomass responses taken preferentially over damage responses when available. Each analysis includes a term to account for differences in effect size due to the response metric used.

Study-based covariates

Source DF Sum Squares Mean Square F P

Model 6 1.4826 0.2471 1.4 0.2327

Error 47 8.2663 0.1759

Corrected total 53 9.7489

Experimental duration 1 0.0848 0.0848 0.48 0.4908

Exclusion area 1 0.0593 0.0593 0.34 0.5642

Host plant growth form 2 0.1682 0.0841 0.48 0.6229

Sample size 1 0.1602 0.1602 0.91 0.3447 biomass vs. damage 1 0.1824 0.1824 1.04 0.3138

Field site-based covariates

Source DF Sum Squares Mean Square F P

Model 4 1.5060 0.3765 2.54 0.0587

Error 32 4.7391 0.1481

Corrected total 36 6.2451

Precipitation 1 0.0013 0.0013 0.01 0.9269

NDVI 1 0.2965 0.2965 2 0.1667 Latitude (absolute value) 1 0.1025 0.1025 0.69 0.4117

biomass vs. damage 1 1.1096 1.1096 7.49 0.01

! Table S4. Statistical tables for analysis of the association between the ratio of intermediate predators to herbivores in the presence of vertebrate insectivores and the strength of effects upon arthropods and plants.

S4-A. Dependent variable = LRR predatory arthropods

Source DF Sum Squares Mean Square F P ln(IP:H+) 1 3.4796 3.4796 7.44 0.0087

Error 52 24.3297 0.4679

Corrected Total 53 27.8093

S4-B. Dependent variable = LRR herbivorous arthropods

Source DF Sum Squares Mean Square F P ln(IP:H+) 1 13.1664 13.1664 15.59 0.0002

Error 52 43.9196 0.8446

Corrected Total 53 57.0861

S4-C. Dependent variable = LRR plants

Source DF Sum Squares Mean Square F P

Model 2 0.7911 0.3955 3.96 0.0341

Error 22 2.2002 0.1000

Corrected Total 24 2.9913 ln(IP:H+) 1 0.7676 0.7676 7.68 0.0112 biomass vs. damage 1 0.0064 0.0064 0.06 0.802

Reduced model based on P<0.15

Source DF Sum Squares Mean Square F P

ln(IP:H+) 1 0.7847 0.7847 8.18 0.0089

Error 23 2.2066 0.0959

Corrected Total 24 2.9913