The Feeding Interactions Among Species in a Community Are Best Described as Ab

1. Introduction

A long-standing question in ecology is whether resource availability limits biodiversity [1–3]. The 'bottom-up' hypothesis predicts that productivity and species composition of primary producers determine species diversity at all higher trophic levels, e.g. herbivores and their predators. The implications are that changes in species diversity and abundance at higher trophic levels should have no effect on diversity at lower trophic levels. However, for many terrestrial and aquatic ecosystems, the role of bottom-up forcing may be much weaker than typically considered [1–3]. In a recent review, Terborgh [4] argued that 'pervasive bottom-up thinking has blinded us to alternatives', suggesting that top-down forcing (herbivory and predation) should be incorporated into theories of species coexistence [4].

The question of whether top-down forcing regulates biodiversity more strongly than bottom-up forcing can be traced to the 'green world' hypothesis proposed by Hairston et al. [5]. Under this perspective, predators are the key to keeping the world 'green' because they limit herbivore abundance, thus releasing plants from herbivory and allowing them to flourish. Strong top-down forcing by high trophic levels also reduces competition from one or a few dominant species at lower trophic levels, promoting diversification or coexistence among lower trophic taxa [4]. There have been a number of tests of this hypothesis in recent decades, with evidence supporting both top-down and bottom-up controls of biodiversity [6–13]. This suggests that bottom-up and top-down forcing on biodiversity need not be mutually exclusive, and that resources and consumers interact to shape ecosystem composition [6–12]. The strength of these relationships may also vary among ecosystems depending on environmental conditions [14–19]. Generalizing from individual studies within single ecosystems, often at small spatial scales [15], may not therefore be warranted for inferring patterns at broad spatial scales. Large-scale assessments of these trophic relationships is thus important for testing the generality of the results found at regional scales, as well as understanding how these vary across spatial scales, both in terms of magnitude and direction [20]. More comparative examinations across ecosystems are needed, especially with respect to the strengths and interactions between bottom-up and top-down forces on diversity [4,16].

How bottom-up and top-down forces regulate biodiversity at regional and global scales is poorly understood [21,22]. The dynamics of trophic interactions over broad spatial scales can alter species composition and reshape biodiversity patterns by affecting immigration, extinction and speciation processes [23–25]. For example, immigrating consumers provide nutrients for primary producers [26], secondary extinction occurs when specialized consumers lose all of their prey [21,27] and strong trophic interaction favours coevolution and hence changes the optimum phenotype, resulting in faster speciation [25,28]. Understanding variation in trophic interactions across systems provides important insights into questions of biodiversity dynamics, species extinction risk and evolutionary dynamics [13,21,25]. For example, megafauna extinction during the Late Pleistocene and Early Holocene has had a major influence on contemporary ecosystems [29]. The influence of these losses is still evident today with lags in extinction [30] and long-term vegetation shifts [31]. Similar effects occur due to later trophic downgrading [32]. For example, the recent (past 150-year) decline of Australia's largest top predator, the dingo (Canis lupus dingo) has contributed to continent-wide declines of other mammals [33]. In addition, anthropogenic changes to key ecological processes, such as vegetation dynamics and fire, have further altered habitats and thus trophic interactions. Indeed, recent studies of predator–prey interactions have considered humans to be the ultimate superpredator [34–36], while fire is considered a major 'abiotic consumer' of plant biomass having major cascading effects on biodiversity [17,37]. Therefore, when assessing the role of trophic interactions on broad-scale diversity patterns, potentially competing or interacting factors should also be considered.

In this study, we use a comprehensive global dataset on species richness of vascular plants, mammals and breeding birds (figure 1) to evaluate the role of trophic interactions in shaping large-scale diversity patterns. Specifically, we address the following questions. (i) Are relations in large-scale species richness patterns of plants and vertebrate dietary groups consistent with trophic interactions shaping biodiversity patterns? (ii) If so, which direction (top-down or bottom-up) is stronger, or are they complementary? (iii) Are trophic interactions similar among taxa and trophic levels? (iv) How important are trophic interactions relative to extrinsic factors?

2. Material and methods

(a) Plant and vertebrate data

We updated a global dataset on native vascular plant richness originally assembled by Qian and co-workers [38,39]. This dataset was built using a large number of literature sources, including floras, checklists, monographs, atlases and journal articles published during the past several decades, covering 381 geographical units worldwide. Most of these geographical units represent level 4 units referred as 'Basic Recording Units' in the International Working Group on Taxonomic Databases (TDWG; http://www.tdwg.org/). Most geographical units correspond to countries, but with the largest countries (e.g. USA, Canada, China, Brazil and Australia) divided into states or provinces. In a few locations, such as the Far East and Siberian regions of Russia, Mongolia and the Northwest Territories of Canada, data did not match the current political boundaries and thus phytogeographical regions were used for floristic surveys [39].

Species range maps for all terrestrial mammals, excluding bats (Order Chiroptera), were gathered from the IUCN Red List database [40]. For birds, we used breeding distributions of all bird species from the BirdLife International [41]. To match plant richness data, mammal and bird lists were summarized for each geographical unit by overlaying their distribution maps with defined geographical units. In total, we considered 364 geographical units, each of which had at least 10 species in each of the three taxa (i.e. plants, mammals and birds). These geographical units represent nearly 125 × 10⁶ km² or approximately 95% of the total ice-free land of the world (figure 1). Only Antarctica was not represented. The average size of each geographical unit was 343 272 km², with 358 (98%) of 364 geographical units larger than 10 000 km².

(b) Trophic groups for mammals and birds

To assess bottom-up and top-down effects among trophic levels, we divided mammals and birds into three trophic groups: carnivores, predominantly feeding on vertebrates; invertivores, predominantly feeding on invertebrates; and herbivores, predominantly feeding on plant materials, such as fruits, nectar, plants and seeds. We used EltonTraits v.1.0 [42] as the primary literature and assigned each species of mammals and birds to one of these three trophic levels. Some mismatched species names in EltonTraits were corrected by checking their synonyms in the IUCN Red List website. When trait information from the EltonTraits database was not available for a species, we used dietary information from MammalDIET v.1.0 [43] to assign species to trophic groups. A small percentage (less than 10%) of species have little trait information in these databases or are difficult to assign to these trophic groups; accordingly, they were excluded from this study. For mammals, 3750 species were assigned to trophic levels, with 232 carnivores, 891 invertivores and 2627 herbivores. For birds, 8278 species were assigned to trophic levels, with 520 carnivores, 4657 invertivores and 3101 herbivores.

(c) Geographical and environmental factors

In addition to accounting for interactions among trophic levels, we evaluated the importance of climate (current and historical) and environmental heterogeneity on patterns of species richness because these factors are widely known to affect broad-scale patterns in species richness. For current climate, we selected two commonly used variables: mean annual temperature (MAT, an indicator of ambient energy) and annual actual evapotranspiration (AET, a measure of water–energy balance). Data for MAT were obtained from the WorldClim Dataset of average climate data for the period 1960–1990 at the 30 arc-second resolution [44]. AET was obtained from the MODIS (Moderate Resolution Imaging Spectrometer) global terrestrial data for the period 2000–2013 at the 30 arc-second resolution [45]. In addition, we assessed the use of primary productivity, measured by the average of NDVI (Normalized Difference Vegetation Index) for the period 2000–2013 in each geographical unit (https://clarklabs.org/additional-products/global-data-archives/) [46,47], but it showed very weak effects on species richness of trophic groups in preliminary analyses and was therefore excluded from the final analysis.

For historical climate, we used velocity at 2.5 arc-minute resolution for mean annual temperature (tempVelo) and annual precipitation (precVelo) between Last Glacial Maximum (LGM; approx. 21 000 years ago) and the present (1950–2000) to represent climate stability [48]. Climate velocity was based on estimates of past mean annual temperature and annual precipitation from the Paleoclimate Modelling Intercomparison Project Phase II [49], using the averages of the CCSM3 and MIROC3.2 simulations.

For environmental heterogeneity, we used topographic relief (TOPO; maximum minus minimum elevation) within each geographical unit [50]. TOPO was calculated from altitude data from WorldClim 30 arc-second data. All other variables were averaged within each geographical unit.

(d) Human and fire as top consumers

Recent studies of predator–prey interactions have considered humans as the ultimate superpredator [34,35,51] and fire as a widespread and key abiotic consumer of vegetation [17,37]. Accordingly, we used a 1 km resolution human influence index (HUMAN) as a surrogate for anthropogenic disturbance over the period 1995–2004 using the Global Human Influence Index Dataset v.2 [52], and fire density (FIRE) at a 0.1° resolution (available at: http://preview.grid.unep.ch) to represent the fire regime. The HUMAN data were created from nine global data layers covering human population density, human land use and infrastructure, and human access. FIRE data were developed by UNEP/GRID-Europe using the World Fire atlas over the period 1997–2010 [53]. As for climate, HUMAN and FIRE variables were averaged for each geographical unit.

(e) Data analysis

We used structural equation models (SEMs) [54,55] to infer relative importance of bottom-up and top-down controls among trophic levels. SEMs allow simultaneous testing of direct and indirect effects between predictor and response variables [54], thus enabling the evaluation of hypothesized causal relationships among interacting trophic groups. We constructed SEMs based on the schematic diagrams (a priori model) in figure 2 using trophic richness and environmental data for each geographical unit. In this a priori SEM model, direct trophic interactions between plants and herbivores, herbivores and carnivores, and invertivores and carnivores were included. The forcing of plant species richness by invertivores was treated as a direct path, because the data on invertebrate diversity were not available. Current climate (MAT and AET), historic climate (tempVelo and precVelo), environmental heterogeneity (TOPO), human influence (HUMAN), fire disturbance (FIRE) and the area of geographical units (AREA) were considered as covariates of species richness for all four trophic groups.

Compared with traditional SEMs commonly used in ecology, our model (figure 2) was a non-recursive SEM model with reciprocal causal relationships that included pathways both for A towards B and for B towards A in the same model. We built non-recursive SEMs that can estimate two unidirectional relationships simultaneously [54] using the R package systemfit [56]. Compared with recursive models, non-recursive SEM can address the problem of endogeneity (i.e. the variable A is hypothesized to be both a cause and a consequence of the variable B) [57], and the system of equations is identified because all the pairs of variables linked by reciprocal effects have different sets of other predictors. The models in figure 2 were fitted using SEM by ordinary least squares, because all regressors are exogenous [55]. A standardized coefficient (β) for each path was calculated and used to compare the relative importance of trophic interactions and environmental effects for species richness at each trophic level. Considering the potential influence of spatial autocorrelation in the modelled results, we adopted the function 'lavSpatialCorrect' by Jarrett Byrnes (https://github.com/jebyrnes/spatial_correction_lavaan). This function can address the problem of correcting sample sizes and standard errors in SEMs with spatial structure in the autocorrelation of endogenous variables. The goodness of fit of the whole model was measured by the McElroy's system R ² value [58], which has been widely used in economics and sociology.

As a control, we also re-ran the analyses excluding 14 oceanic island units because trophic interactions on islands might be influenced by different processes from mainland regions [24,59]. Oceanic units were defined as regions smaller than Australia and surrounded by ocean [60]. To minimize regional and historical effects on trophic interactions, we also assigned the 364 units into six commonly used biogeographical realms [61], including Nearctic, Neotropical, Afrotropic (Ethiopian), Oriental, Palaearctic and Australasian. The Australasian region was excluded for this analysis because it had only eight geographical units and such a small sample size may invalidate SEM analyses. The Palearctic region was divided into eastern and western sections to make geographical extents more consistent among regions [62]. As a result, six biogeographical regions were used in this study. SEMs were performed for each region, and regional differences were assessed by comparing standardized coefficients among paths.

All statistical analyses were performed using the statistical software r v.3.2.2 [63]. All variables were log-transformed to improve normality. Pairwise correlations were conducted among trophic groups (electronic supplementary material, figure S1), while pairwise correlations among environmental variables were used to assess multicolinearity (electronic supplementary material, figure S2). The R script for non-recursive SEM and data used for the study are provided in the electronic supplementary material.

3. Results

The 364 geographical units include an average of 3487 (±4335 s.d.) species of vascular plants, 86 (±73) species of mammals and 298 (±214) species of breeding birds. The three trophic levels within mammals and birds exhibited strong spatial variation in richness with moderate to high spatial congruence among taxa and trophic levels (figure 1).

At a global scale, plant richness showed positive relations to mammal herbivore richness, with stronger top-down than bottom-up relations (figure 3a). By contrast, plant richness and bird herbivore richness had strong bottom-up, but weak top-down interactions (figure 3b). Relationships between herbivore and carnivore richness were also all positive, with stronger bottom-up than top-down interactions in mammals (figure 3a), and stronger top-down than bottom-up interactions in birds (figure 3b). Relations between invertivore and carnivore richness were moderate and negative for bottom-up interactions and moderate and positive for top-down interactions in mammals (figure 3a), with stronger and positive interactions in both directions in birds (figure 3b). Links between plant and invertivore richness were moderate, with positive bottom-up interactions for both mammals and birds, but weak, top-down interactions for mammals and intermediate, top-down interactions for birds (figure 3). Analyses excluding 14 oceanic island regions yielded similar results (electronic supplementary material, figure S3).

Figure 3. Global-scale non-recursive SEM results for trophic interactions among plant, mammal and bird richness after accounting for environmental variables. Only the central parts of the *a priori* SEM model from figure 2 are shown here. Values next to arrow lines are standardized path coefficients (β). Green arrows represent bottom-up effects, and red-dashed arrows top-down effects. Arrow line thicknesses correspond to the strength of trophic interaction based on standardized coefficients. Significance levels after correcting sample sizes and standard errors in SEMs with spatial structure in the autocorrelation of endogenous variables are denoted by means of asterisks *p < 0.05, **p < 0.01, ***p < 0.001. Adjusted R ² of each path is shown (in blue). The goodness of fit of the whole model is shown as McElroy's R ².

Geographical and environmental factors had relatively consistent effects on species richness across taxa and trophic groups (figure 4). AET, MAT and AREA generally had the strongest effects, with the main difference being weak effects of AET on mammal carnivores and MAT on bird carnivores (figure 4). Topographic heterogeneity and glacial–interglacial velocity in temperature had weaker, but consistently positive or negative effects, respectively. By contrast, fire and human influence had highly variable effects across taxa and trophic groups. Notably, the relative importance of many geographical and environmental variables (figure 4) was much weaker than that of trophic interaction variables (figure 3).

Region-specific analyses demonstrated that trophic interactions were generally consistent among the six biogeographical realms and with that of trophic interactions at the global scale (electronic supplementary material, figure S4). The most inconsistent relationship was top-down interactions between bird invertivore richness and plant richness, which was strongly negative in the Neotropics and Oriental realms, but positive elsewhere. Other major differences included a strong negative top-down forcing between bird herbivore and plant richness in the Afrotropics, and a strong negative bottom-up control between mammal invertivore and carnivore richness in the Oriental realm.

4. Discussion

(a) Bottom-up and top-down controls across trophic levels

Our global-scale analysis of species richness patterns in plant and vertebrate taxa documented patterns of statistical relationships between trophic groups that are consistent with trophic interactions playing a major role in shaping global patterns in biodiversity [4]. The region-specific trophic relations were generally consistent with the global-scale relations, although there were some regional differences in trophic links. After controlling for effects of current and historic climate, topography and disturbances from humans and fire, we detected a strong top-down effect of herbivores on plant richness for mammals. A plausible mechanism would be that by directly reducing the abundance of preferred plant species and indirectly regulating nutrient cycling and soil fertility, mammal herbivores provide opportunities for growth and reproduction of other plant species and therefore increased plant diversity [4,11,64]. Positive top-down effects of invertivore richness on plant richness were identified for birds, consistent with a previous meta-analysis of plants, herbivorous insects and bird predators [65]. Weak top-down effects of bird herbivores and mammal invertivores on plant richness were detected, possibly due to confounding factors that have not been fully accounted for in the model structure.

We also found moderate to strong top-down interactions among trophic groups in mammals and birds. These findings are consistent with a number of prior experimental and field studies at local, regional and continental scales. The early work of Paine [6] on predator removal of the starfish Pisaster in a marine rocky intertidal community showed that predator removal led to sharp reductions in diversity (from 15 species to eight species) of both predator and prey species. In terrestrial systems, Schmitz et al. [11] examined 60 independent tests from 41 studies of top predator removal experiments, finding that top-down controls on plant-eater density were frequent (75%; 45 of 60 tests). At the regional scale, a well-known example of this is the top-down trophic cascade identified following the reintroduction of previously extirpated gray wolves (C. lupus) into Yellowstone National Park. With wolves (apex predator of ungulates) restored over the past two decades, the Yellowstone ecosystem has been restructured, including the recovery of woody plants and the riparian ecosystem [66,67]. At a continental scale, the decline of Australia's top mammal predator, the dingo, has resulted in the decrease in native mammals' diversity through release of invasive mesopredators [33]. In North and South America, Faurby & Svenning [24] used coarse-scale mammal distribution databases to reassess the influence of Great American Biotic Interchange for mammals and concluded that predation has played a critical role in the resulting long-term diversity dynamics. Overall, strong top-down forcing by high trophic levels may affect the processes of species immigration [26], extinction [27,33] and speciation [25,68], and thus reshape biodiversity patterns across space and time even across broad scales.

Bottom-up effects of plants on animals have been reported in a number of other studies [69–73]. However, top-down controls of animals on plants across large spatial scales have been less well explored, despite evidence from controlled experiments [12,18,74] and small-scale field observations [7,75]. Our findings also demonstrate that ascending prey–predator control is stronger than top-down forcing for mammal herbivores and bird invertivores. This is consistent with the findings by Sandom et al. [76] on global-scale mammal predator–prey interactions. By considering both bottom-up and top-down forces in regional and global analyses, our findings suggest that top-down and bottom-up trophic interactions among different trophic levels are not mutually exclusive [77,78]. Both forces should be incorporated in explanatory studies of biodiversity patterns [77,78].

It is worth mentioning that a large number of macroecological studies have documented correlations in species richness among taxa. Many studies have focused only on bottom-up effects of plants on animals [69–73], while others have investigated cross-taxon congruence using pairwise correlations among major taxonomic groups [79–81]. However, the latter approach provides little insight into the magnitude or direction of trophic interactions and a correlation between two trophic levels does not necessarily mean that one drives the other. For example, we observed high pairwise correlations among global patterns in trophic groups (electronic supplementary material, figure S1), with plant and mammal herbivore richness being positively correlated (Pearson's correlation coefficient = 0.692, p < 0.001). This does not, however, provide evidence on the strength of top-down versus bottom-up effects. By disentangling the relative importance of top-down versus bottom-up effects, our analyses provide mechanistic insights into broad-scale trophic interactions.

(b) Importance of trophic versus extrinsic drivers on patterns of diversity

Climatic and topographic variables are often considered major drivers of broad-scale patterns in species richness [79–81]. However, our results suggest that interactions among taxa of different trophic levels may exert additional effects on species richness of particular taxa above and beyond environmental drivers. After considering the effects of trophic interactions, we found that the strength of the examined environmental variables was generally weaker than the influence from trophic levels. Previous studies that have tested plant–animal interactions after accounting for environmental conditions have shown mixed results. For instance, Hawkins & Pausas [70] found that mammal richness at a 100 km² resolution in northeast Spain was better described by climatic variables than by plant richness. Jetz et al. [71] found that environmental variables were better at predicting global patterns in vertebrate diversity than that of plant diversity. By contrast, Qian & Ricklefs [39] analysed cross-taxon correlations between plants and terrestrial vertebrates and found that richness of an animal group was best predicted by richness of other organism groups, rather than environmental variables. Likewise, Greve et al. [64] found that mammalian browsers in Africa were related more to patterns in acacias.

We do not suggest that the role of extrinsic drivers should be ignored when explaining biodiversity patterns for different trophic levels. Of the extrinsic variables examined in our study, AET and MAT were among the most important environmental determinants of richness for most trophic groups. The effects of humans and fire were also important for several trophic groups (e.g. mammal carnivores, and bird herbivores and invertivores), indicating their importance for studies of trophic interactions and biodiversity maintenance [17,51,82]. Additional studies are needed to assess how climate, topography, disturbance (fire and humans) and biotic factors interact to shape biodiversity patterns, and how their importance varies among trophic groups and across spatial and temporal scales.

(c) Limitations and caveats

Considering the ecological complexity of understanding and analysing trophic interactions [21,22], we point out a number of limitations of our work. The first is the broad extent used in sampling biodiversity (364 geographical units) and the large area of the sampling units [50,83]. Some sampling units we used could be larger than the range sizes of some plant and animal species. To reduce the potential influence of variation in geographical unit size, we included sampling area as a predictor in our analyses, which showed weak effects. Second, high-resolution diversity data on plants, vertebrates and invertebrates at a global, or even a regional, scale are scarce, thus limiting the potential for fine-grained, yet broad-scale analyses [50,83,84]. Data deficiency for species distribution, abundance and functional traits limits integrated analyses of trophic ecology and construction of complex food webs [85]. Use of more detailed data on species distribution and traits would further advance our understanding in this field. Third, lack of geographical independence could create spurious results. If some points are clustered in space (e.g. high richness of two trophic groups), it could bias estimates of relationships. However, we accounted for spatial autocorrelation in our analyses and furthermore found generally consistent global and regional patterns. We are therefore confident that our results are robust to spatial autocorrrelation. Finally, we divided mammals and birds into three mutually exclusive trophic levels based on their major dietary preferences [42,43], but interactions among trophic groups are more complex in the real world. While it might be relatively easy to construct complex food webs at small spatial scales, it is much more difficult to do so at broad spatial scales. When these data become available, future studies should evaluate how classification of trophic levels and length of food chains [42,43] affect estimates of trophic interactions at broad spatial scales and test the generality of top-down or bottom-up controls and the validity of these relationships across spatial scales [4].

(d) Conclusion

Our results demonstrate that there are consistent relationships between the large-scale species richness patterns of different trophic groups, consistent with a strong role of trophic interactions in shaping biodiversity patterns. Furthermore, our results suggest a particular importance of herbivores, notably mammalian, for structuring biodiversity patterns across trophic levels, via both top-down and bottom-up forcing [86,87]. These findings underscore the importance of integrating trophic forcing mechanisms into current theoretical studies of biodiversity maintenance [4,32].

Data accessibility

The data and R script are available in the electronic supplementary material.

Authors' contributions

J.Z. and J.-C.S. developed and framed the research questions. J.Z., H.Q., M.G. and V.P. prepared the database. J.Z., M.G. and V.P. analysed data; and all the authors contributed to the writing of the article.

Competing interests

We declare we have no competing interests.

Funding

We gratefully acknowledge financial support from the National Natural Science Foundation of China (grant no. 31670439) and East China Normal University to J.Z., and the European Research Council (ERC-2012-StG-310886-HISTFUNC) to J-C.S.

Acknowledgements

We thank the Editor Prof. Diego Vázquez, Dr James Grace and three anonymous reviewers for constructive and helpful comments. We thank Kate Broadley at Fuse Consulting, Yuchu Qin and Shuqing Teng for their assistance on data analysis. J.-C.S. also considers this work a contribution to his VILLUM Investigator project (VILLUM FONDEN, grant 16549).

Footnotes

Electronic supplementary material is available online at https://dx.doi.org/10.6084/m9.figshare.c.4157570.

References

1
Hutchinson GE
. 1959 Homage to Santa Rosalia or why are there so many kinds of animals? Am. Nat. 93 , 145–159. (doi:10.1086/282070) Crossref, ISI, Google Scholar
2
MacArthur R, Levins R
. 1967 The limiting similarity, convergence and divergence of coexisting species. Am. Nat. 101 , 377–385. (doi:10.1086/282505) Crossref, ISI, Google Scholar
3
Tilman D
. 1982 Resource competition and community structure . Princeton, NJ: Princeton University Press. Google Scholar
4
Terborgh JW
. 2015 Toward a trophic theory of species diversity. Proc. Natl Acad. Sci. USA 112 , 11 415–11 422. (doi:10.1073/pnas.1501070112) Crossref, ISI, Google Scholar
5
Hairston NG, Smith FE, Slobodkin LB
. 1960 Community structure, population control, and competition. Am. Nat. 94 , 421–425. (doi:10.1086/282146) Crossref, ISI, Google Scholar
6
Paine RT
. 1966 Food web complexity and species diversity. Am. Nat. 100 , 65–75. (doi:10.1086/282400) Crossref, ISI, Google Scholar
7
Terborgh Jet al.
2001 Ecological meltdown in predator-free forest fragments. Science 294 , 1923–1926. (doi:10.1126/science.1064397) Crossref, PubMed, ISI, Google Scholar
8
Murdoch WW
. 1966 Community structure, population control, and competition—a critique. Am. Nat. 100 , 219–226. (doi:10.1086/282415) Crossref, ISI, Google Scholar
9
Power ME
. 1992 Top-down and bottom-up forces in food webs: do plants have primacy? Ecology 73 , 733–746. (doi:10.2307/1940153) Crossref, ISI, Google Scholar
10
Polis GA
. 1999 Why are parts of the world green? Multiple factors control productivity and the distribution of biomass. Oikos 86 , 3–15. (doi:10.2307/3546565) Crossref, ISI, Google Scholar
11
Schmitz OJ, Hamback PA, Beckerman AP
. 2000 Trophic cascades in terrestrial systems: a review of the effects of carnivore removals on plants. Am. Nat. 155 , 141–153. (doi:10.1086/303311) Crossref, PubMed, ISI, Google Scholar
12
Duffy JEet al.
2015 Biodiversity mediates top-down control in eelgrass ecosystems: a global comparative-experimental approach. Ecol. Lett. 18 , 696–705. (doi:10.1111/ele.12448) Crossref, PubMed, ISI, Google Scholar
13
Edgar GJ, Alexander TJ, Lefcheck JS, Bates AE, Kininmonth SJ, Thomson RJ, Duffy JE, Costello MJ, Stuart-Smith RD
. 2017 Abundance and local-scale processes contribute to multi-phyla gradients in global marine diversity. Sci. Adv. 3 , e1700419. (doi:10.1126/sciadv.1700419) Crossref, PubMed, ISI, Google Scholar
14
Oksanen L, Fretwell SD, Arruda J, Niemela P
. 1981 Exploitation ecosystems in gradients of primary productivity. Am. Nat. 118 , 240–261. (doi:10.1086/283817) Crossref, ISI, Google Scholar
15
Gripenberg S, Roslin T
. 2007 Up or down in space? Uniting the bottom-up versus top-down paradigm and spatial ecology. Oikos 116 , 181–188. (doi:10.1111/j.2006.0030-1299.15266.x) Crossref, ISI, Google Scholar
16
Leroux SJ, Loreau M
. 2015 Theoretical perspectives on bottom-up and top-down interactions across ecosystems. In Trophic ecology: bottom-up and top-down interactions across aquatic and terrestrial systems (eds
Hanley T, St Pierre K
), pp. 3–27. Cambridge, UK: Cambridge University Press. Crossref, Google Scholar
17
Bond WJ, Keeley JE
. 2005 Fire as a global 'herbivore': the ecology and evolution of flammable ecosystems. Trends Ecol. Evol. 20 , 387–394. (doi:10.1016/j.tree.2005.04.025) Crossref, PubMed, ISI, Google Scholar
18
Hillebrand Het al.
2007 Consumer versus resource control of producer diversity depends on ecosystem type and producer community structure. Proc. Natl Acad. Sci. USA 104 , 10 904–10 909. (doi:10.1073/pnas.0701918104) Crossref, ISI, Google Scholar
19
Hopcraft JG, Olff H, Sinclair AR
. 2010 Herbivores, resources and risks: alternating regulation along primary environmental gradients in savannas. Trends Ecol. Evol. 25 , 119–128. (doi:10.1016/j.tree.2009.08.001) Crossref, PubMed, ISI, Google Scholar
20
Belmaker J, Zarnetske P, Tuanmu MN, Zonneveld S, Record S, Strecker A, Beaudrot L
. 2015 Empirical evidence for the scale dependence of biotic interactions. Global Ecol. Biogeogr. 24 , 750–761. (doi:10.1111/geb.12311) Crossref, Google Scholar
21
Boyce DG, Frank KT, Worm B, Leggett WC
. 2015 Spatial patterns and predictors of trophic control in marine ecosystems. Ecol. Lett. 18 , 1001–1011. (doi:10.1111/ele.12481) Crossref, PubMed, ISI, Google Scholar
22
Kissling WD, Schleuning M
. 2015 Multispecies interactions across trophic levels at macroscales: retrospective and future directions. Ecography 38 , 346–357. (doi:10.1111/ecog.00819) Crossref, ISI, Google Scholar
23
Gravel D, Massol F, Canard E, Mouillot D, Mouquet N
. 2011 Trophic theory of island biogeography. Ecol. Lett. 14 , 1010–1016. (doi:10.1111/j.1461-0248.2011.01667.x) Crossref, PubMed, ISI, Google Scholar
24
Faurby S, Svenning J
. 2016 The asymmetry in the Great American Biotic Interchange in mammals is consistent with differential susceptibility to mammalian predation. Glob. Ecol. Biogeogr. 25 , 1443–1453. (doi:10.1111/geb.12504) Crossref, Google Scholar
25
Schemske DW, Mittelbach GG, Cornell HV, Sobel JM, Roy K
. 2009 Is there a latitudinal gradient in the importance of biotic interactions? Annu. Rev. Ecol. Evol. Syst. 40 , 245–269 (doi:10.1146/annurev.ecolsys.39.110707.173430) Crossref, ISI, Google Scholar
26
Loreau M, Holt RD
. 2004 Spatial flows and the regulation of ecosystems. Am. Nat. 163 , 606–615. (doi:10.1086/382600) Crossref, PubMed, ISI, Google Scholar
27
Sanders D, Thebault E, Kehoe R, van Veen FJF
. 2018 Trophic redundancy reduces vulnerability to extinction cascades. Proc. Natl Acad. Sci. USA 115 , 2419–2424. (doi:10.1073/pnas.1716825115) Crossref, PubMed, ISI, Google Scholar
28
Dobzhansky T
. 1950. Evolution in the tropics. Am. Sci. 38 , 209–221. Google Scholar
29
Barnosky AD, Koch PL, Feranec RS, Wing SL, Shabel AB
. 2004 Assessing the causes of Late Pleistocene extinctions on the continents. Science 306 , 70–75. (doi:10.1126/science.1101476) Crossref, PubMed, ISI, Google Scholar
30
Doughty CEet al.
2016 Megafauna extinction, tree species range reduction, and carbon storage in Amazonian forests. Ecography 39 , 194–203. (doi:10.1111/ecog.01587) Crossref, ISI, Google Scholar
31
Doughty CE, Faurby S, Svenning J
. 2016 The impact of the megafauna extinctions on savanna woody cover in South America. Ecography 39 , 213–222. (doi:10.1111/ecog.01593) Crossref, ISI, Google Scholar
32
Estes JAet al.
2011 Trophic downgrading of planet Earth. Science 333 , 301–306. (doi:10.1126/science.1205106) Crossref, PubMed, ISI, Google Scholar
33
Johnson CN, Isaac JL, Fisher DO
. 2007 Rarity of a top predator triggers continent-wide collapse of mammal prey: dingoes and marsupials in Australia. Proc. R. Soc. B 274 , 341–346. (doi:10.1098/rspb.2006.3711) Link, ISI, Google Scholar
34
Darimont CT, Fox CH, Bryan HM, Reimchen TE
. 2015 The unique ecology of human predators. Science 349 , 858–860. (doi:10.1126/science.aac4249) Crossref, PubMed, ISI, Google Scholar
35
Dorresteijn I, Schultner J, Nimmo DG, Fischer J, Hanspach J, Kuemmerle T, Kehoe L, Ritchie EG
. 2015 Incorporating anthropogenic effects into trophic ecology: predator–prey interactions in a human-dominated landscape. Proc. R. Soc. B 282 , 20151602. (doi:10.1098/rspb.2015.1602) Link, ISI, Google Scholar
36
Barnes ADet al.
2017 Direct and cascading impacts of tropical land-use change on multi-trophic biodiversity. Nature Ecol. Evol. 1 , 1511. (doi:10.1038/s41559-017-0275-7) Crossref, PubMed, ISI, Google Scholar
37
Bond WJ
. 2005 Large parts of the world are brown or black: a different view on the 'Green World' hypothesis. J. Veg. Sci. 16 , 261–266. (doi:10.1658/1100-9233(2005)016[0261:Lpotwa]2.0.Co;2) ISI, Google Scholar
38
Qian H, Song JS, Krestov P, Guo QF, Wu ZM, Shen XS, Guo XS
. 2003 Large-scale phytogeographical patterns in East Asia in relation to latitudinal and climatic gradients. J. Biogeogr. 30 , 129–141. (doi:10.1046/j.1365-2699.2003.00807.x) Crossref, ISI, Google Scholar
39
Qian H, Ricklefs RE
. 2008 Global concordance in diversity patterns of vascular plants and terrestrial vertebrates. Ecol. Lett. 11 , 547–553. (doi:10.1111/j.1461-0248.2008.01168.x) Crossref, PubMed, ISI, Google Scholar
40
IUCN (International Union for Conservation of Nature). 2014The IUCN Red List of Threatened Species. V. 2014.2. See http://www.iucnredlist.org. Downloaded 5 July 2014. Google Scholar
41
BirdLifeInternational N
. 2014 Bird species distribution maps of the world . Cambridge, UK: BirdLife International. Google Scholar
42
Wilman H, Belmaker J, Simpson J, de la Rosa C, Rivadeneira MM, Jetz W
. 2014 EltonTraits 1.0: species-level foraging attributes of the world's birds and mammals. Ecology 95 , 2027. (doi:10.1890/13-1917.1) Crossref, Google Scholar
43
Kissling WD, Dalby L, Flojgaard C, Lenoir J, Sandel B, Sandom C, Trojelsgaard K, Svenning JC
. 2014 Establishing macroecological trait datasets: digitalization, extrapolation, and validation of diet preferences in terrestrial mammals worldwide. Ecol. Evol. 4 , 2913–2930. (doi:10.1002/ece3.1136) Crossref, PubMed, ISI, Google Scholar
44
Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A
. 2005 Very high resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25 , 1965–1978. (doi:10.1002/Joc.1276) Crossref, ISI, Google Scholar
45
Mu QZ, Zhao MS, Running SW
. 2011 Improvements to a MODIS global terrestrial evapotranspiration algorithm. Remote Sens. Environ. 115 , 1781–1800. (doi:10.1016/j.rse.2011.02.019) Crossref, ISI, Google Scholar
46
Phillips LB, Hansen AJ, Flather CH
. 2008 Evaluating the species energy relationship with the newest measures of ecosystem energy: NDVI versus MODIS primary production. Remote Sens. Environ. 112 , 4381–4392. (doi:10.1016/j.rse.2008.08.002) Crossref, ISI, Google Scholar
47
Pettorelli N, Gaillard JM, Mysterud A, Duncan P, Stenseth NC, Delorme D, Van Laere G, Toigo C, Klein F
. 2006 Using a proxy of plant productivity (NDVI) to find key periods for animal performance: the case of roe deer. Oikos 112 , 565–572. (doi:10.1111/j.0030-1299.2006.14447.x) Crossref, ISI, Google Scholar
48
Sandel B, Arge L, Dalsgaard B, Davies RG, Gaston KJ, Sutherland WJ, Svenning JC
. 2011 The influence of late quaternary climate-change velocity on species endemism. Science 334 , 660–664. (doi:10.1126/science.1210173) Crossref, PubMed, ISI, Google Scholar
49
Braconnot Pet al.
2007 Results of PMIP2 coupled simulations of the Mid-Holocene and Last Glacial Maximum—Part 1: experiments and large-scale features. Clim. Past 3 , 261–277. (doi:10.5194/cp-3-261-2007) Crossref, ISI, Google Scholar
50
Rahbek C, Graves GR
. 2001 Multiscale assessment of patterns of avian species richness. Proc. Natl Acad. Sci. USA 98 , 4534–4539. (doi:10.1073/pnas.071034898) Crossref, PubMed, ISI, Google Scholar
51
Muhly TB, Hebblewhite M, Paton D, Pitt JA, Boyce MS, Musiani M
. 2013 Humans strengthen bottom-up effects and weaken trophic cascades in a terrestrial food web. PLoS ONE 8 , e64311. (doi:10.1371/journal.pone.0064311) Crossref, PubMed, ISI, Google Scholar
52
CIESIN WCS. 2005 Last of the wild project, version 2, 2005 (LWP-2): global human influence index (HII) dataset (geographic) . Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). Google Scholar
53
Mota BW, Pereira JMC, Oom D, Vasconcelos MJP, Schultz M
. 2006 Screening the ESA ATSR-2 World Fire Atlas (1997–2002). Atmos. Chem. Phys. 6 , 1409–1424. (doi: 10.5194/acp-6-1409-2006) Crossref, ISI, Google Scholar
54
Grace JB
. 2006 Structural equation modeling and natural systems . Cambridge, UK: Cambridge University Press. Crossref, Google Scholar
55
Shipley B
. 2009 Confirmatory path analysis in a generalized multilevel context. Ecology 90 , 363–368. (doi:10.1890/08-1034.1) Crossref, PubMed, ISI, Google Scholar
56
Henningsen A, Hamann JD
. 2007 systemfit: a package for estimating systems of simultaneous equations in R. J. Stat. Softw. 23 , 1–40. (doi:10.18637/jss.v023.i04) Crossref, ISI, Google Scholar
57
Paxton P, Hipp JP, Marquart-Pyatt S
. 2011 Nonrecursive models: endogeneity, reciprocal relationships, and feedback loops . Beverley Hills, CA: SAGE Publishing. Crossref, Google Scholar
58
McElroy MB
. 1977 Goodness of fit for seemingly unrelated regressions: Glahn's and Hooper's r̄ ² . J. Econ. 6 , 381–387. (doi:10.1016/0304-4076(77)90008-2) Crossref, Google Scholar
59
Whittaker RJ
. 1998 Island biogeography: ecology, evolution, and conservation . Oxford, UK: Oxford University Press. Google Scholar
60
Kreft H, Jetz W, Mutke J, Kier G, Barthlott W
. 2008 Global diversity of island floras from a macroecological perspective. Ecol. Lett. 11 , 116–127. (doi:10.1111/j.1461-0248.2007.01129.x) PubMed, ISI, Google Scholar
61
Pielou EC
. 1992 Biogeography . Malabar, FL: Krieger Publishing Company. Google Scholar
62
Qian H
. 2008 Effects of historical and contemporary factors on global patterns in avian species richness. J. Biogeogr. 35 , 1362–1373. (doi:10.1111/j.1365-2699.2008.01901.x) Crossref, ISI, Google Scholar
63
R Core Team. 2015 R: A language and environment for statistical computing . Vienna, Austria: R Foundation for Statistical Computing. Google Scholar
64
Greve Met al.
2012 Continental-scale variability in browser diversity is a major driver of diversity patterns in acacias across Africa. J. Ecol. 100 , 1093–1104. (doi:10.1111/j.1365-2745.2012.01994.x) Crossref, ISI, Google Scholar
65
Mäntylä E, Klemola T, Laaksonen T
. 2011 Birds help plants: a meta-analysis of top-down trophic cascades caused by avian predators. Oecologia 165 , 143–151. (doi:10.1007/s00442-010-1774-2) Crossref, PubMed, ISI, Google Scholar
66
Beschta RL, Ripple WJ
. 2009 Large predators and trophic cascades in terrestrial ecosystems of the western United States. Biol. Conserv. 142 , 2401–2414. (doi:10.1016/j.biocon.2009.06.015) Crossref, ISI, Google Scholar
67
Ripple WJ, Beschta RL
. 2012 Trophic cascades in yellowstone: the first 15 years after wolf reintroduction. Biol. Conserv. 145 , 205–213. (doi:10.1016/j.biocon.2011.11.005) Crossref, ISI, Google Scholar
68
Broeckhoven C, Diedericks G, Hui C, Makhubo BG, Mouton PLFN
. 2016 Enemy at the gates: rapid defensive trait diversification in an adaptive radiation of lizards. Evolution 70 , 2647–2656. (doi:10.1111/evo.13062) Crossref, PubMed, ISI, Google Scholar
69
Hawkins BA, Porter EE
. 2003 Does herbivore diversity depend on plant diversity? The case of California butterflies. Am. Nat. 161 , 40–49. (doi:10.1086/345479) Crossref, PubMed, ISI, Google Scholar
70
Hawkins BA, Pausas JG
. 2004 Does plant richness influence animal richness? The mammals of Catalonia (NE Spain). Divers. Distrib. 10 , 247–252. (doi:10.1111/j.1366-9516.2004.00085.x) Crossref, ISI, Google Scholar
71
Jetz W, Kreft H, Ceballos G, Mutke J
. 2009 Global associations between terrestrial producer and vertebrate consumer diversity. Proc. R. Soc. B 276 , 269–278. (doi:10.1098/rspb.2008.1005) Link, ISI, Google Scholar
72
Qian H, Kissling WD
. 2010 Spatial scale and cross-taxon congruence of terrestrial vertebrate and vascular plant species richness in China. Ecology 91 , 1172–1183. (doi:10.1890/09-0620.1) Crossref, PubMed, ISI, Google Scholar
73
Zhang J, Kissling WD, He FL
. 2013 Local forest structure, climate and human disturbance determine regional distribution of boreal bird species richness in Alberta, Canada. J. Biogeogr. 40 , 1131–1142. (doi:10.1111/Jbi.12063) Crossref, ISI, Google Scholar
74
Souza L, Zelikova TJ, Sanders NJ
. 2016 Bottom-up and top-down effects on plant communities: nutrients limit productivity, but insects determine diversity and composition. Oikos 125 , 566–575. (doi:10.1111/oik.02579) Crossref, ISI, Google Scholar
75
Mclaren BE, Peterson RO
. 1994 Wolves, moose, and tree rings on Isle Royale. Science 266 , 1555–1558. (doi:10.1126/science.266.5190.1555) Crossref, PubMed, ISI, Google Scholar
76
Sandom C, Dalby L, Flojgaard C, Kissling WD, Lenoir J, Sandel B, Trojelsgaard K, Ejrnaes R, Svenning JC
. 2013 Mammal predator and prey species richness are strongly linked at macroscales. Ecology 94 , 1112–1122. (doi:10.1890/12-1342.1) Crossref, PubMed, ISI, Google Scholar
77
Chesson P, Kuang JJ
. 2008 The interaction between predation and competition. Nature 456 , 235–238. (doi:10.1038/nature07248) Crossref, PubMed, ISI, Google Scholar
78
Pace ML, Cole JJ, Carpenter SR, Kitchell JF
. 1999 Trophic cascades revealed in diverse ecosystems. Trends Ecol. Evol. 14 , 483–488. (doi:10.1016/S0169-5347(99)01723-1) Crossref, PubMed, ISI, Google Scholar
79
Wolters V, Bengtsson J, Zaitsev AS
. 2006 Relationship among the species richness of different taxa. Ecology 87 , 1886–1895. (doi:10.1890/0012-9658(2006)87[1886:RATSRO]2.0.CO;2) Crossref, PubMed, ISI, Google Scholar
80
Castagneyrol B, Jactel H
. 2012 Unraveling plant–animal diversity relationships: a meta-regression analysis. Ecology 93 , 2115–2124. (doi:10.1890/11-1300.1) Crossref, PubMed, ISI, Google Scholar
81
Westgate MJ, Barton PS, Lane PW, Lindenmayer DB
. 2014 Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nat. Commun. 5 , 3899. (doi:10.1038/ncomms4899) Crossref, PubMed, ISI, Google Scholar
82
Ellis EC, Ramankutty N
. 2008 Putting people in the map: anthropogenic biomes of the world. Front. Ecol. Environ. 6 , 439–447. (doi:10.1890/070062) Crossref, ISI, Google Scholar
83
Jetz W, Rahbek C
. 2002 Geographic range size and determinants of avian species richness. Science 297 , 1548–1551. (doi:10.1126/science.1072779) Crossref, PubMed, ISI, Google Scholar
84
Kreft H, Jetz W
. 2007 Global patterns and determinants of vascular plant diversity. Proc. Natl Acad. Sci. USA 104 , 5925–5930. (doi:10.1073/pnas.0608361104) Crossref, PubMed, ISI, Google Scholar
85
Hortal J, de Bello F, Diniz-Filho JAF, Lewinsohn TM, Lobo JM, Ladle RJ
. 2015 Seven shortfalls that beset large-scale knowledge of biodiversity. Annu. Rev. Ecol. Evol. Syst. 46 , 523–549. (doi:10.1146/annurev-ecolsys-120213-054400) Crossref, ISI, Google Scholar
86
Bakker ES, Gill JL, Johnson CN, Vera FW, Sandom CJ, Asner GP, Svenning JC
. 2016 Combining paleo-data and modern exclosure experiments to assess the impact of megafauna extinctions on woody vegetation. Proc. Natl Acad. Sci. USA 113 , 847–855. (doi:10.1073/pnas.1502545112) Crossref, PubMed, ISI, Google Scholar
87
van der Plas F, Howison RA, Mpanza N, Cromsigt JPGM, Olff H
. 2016 Different-sized grazers have distinctive effects on plant functional composition of an African savannah. J. Ecol. 104 , 864–875. (doi:10.1111/1365-2745.12549) Crossref, ISI, Google Scholar

scotttherhang.blogspot.com

Source: https://royalsocietypublishing.org/doi/10.1098/rspb.2018.0949