Biotic filtering by species' interactions constrains food-web variability across spatial and abiotic gradients.

Despite intensive research on species dissimilarity patterns across communities (i.e. ß-diversity), we still know little about their implications for variation in food-web structures. Our analyses of 50 lake and 48 forest soil communities show that, while species dissimilarity depends on environmental and spatial gradients, these effects are only weakly propagated to the networks. Moreover, our results show that species and food-web dissimilarities are consistently correlated, but that much of the variation in food-web structure across spatial, environmental, and species gradients remains unexplained. Novel food-web assembly models demonstrate the importance of biotic filtering during community assembly by (1) the availability of resources and (2) limiting similarity in species' interactions to avoid strong niche overlap and thus competitive exclusion. This reveals a strong signature of biotic filtering processes during local community assembly, which constrains the variability in structural food-web patterns across local communities despite substantial turnover in species composition.

Diversity and functional structure of soil animal communities suggest soil animal food webs to be buffered against changes in forest land use.

Forest soil and litter is inhabited by a diverse community of animals, which directly and indirectly rely on dead organic matter as habitat and food resource. However, community composition may be driven by biotic or abiotic forces, and these vary with changes in habitat structure and resource supply associated with forest land use. To evaluate these changes, we compiled comprehensive data on the species composition of soil animal communities and environmental factors in forest types varying in land-use intensity in each of three regions in Germany, i.e., coniferous, young managed, old managed, and unmanaged beech forests. Coniferous forests featured high amounts of leaf litter and low microbial biomass concentrations contrasting in particular unmanaged beech forests. However, soil animal diversity and functional community composition differed little between forest types, indicating resilience against disturbance and forest land use. Structural equation modelling suggested that despite a significant influence of forest management on resource abundance and quality, the biomass of most soil fauna functional groups was not directly affected by forest management or resource abundance/quality, potentially because microorganisms hamper the propagation of nutrients to higher trophic levels. Instead, detritivore biomass depended heavily on soil pH. Macrofauna decomposers thrived at high pH, whereas mesofauna decomposers benefitted from low soil pH, but also from low biomass of macrofauna decomposers, potentially due to habitat modification by macrofauna decomposers. The strong influence of soil pH shows that decomposer communities are structured predominantly by regional abiotic factors exceeding the role of local biotic factors such as forest type.

Predator traits determine food-web architecture across ecosystems.

Predator-prey interactions in natural ecosystems generate complex food webs that have a simple universal body-size architecture where predators are systematically larger than their prey. Food-web theory shows that the highest predator-prey body-mass ratios found in natural food webs may be especially important because they create weak interactions with slow dynamics that stabilize communities against perturbations and maintain ecosystem functioning. Identifying these vital interactions in real communities typically requires arduous identification of interactions in complex food webs. Here, we overcome this obstacle by developing predator-trait models to predict average body-mass ratios based on a database comprising 290 food webs from freshwater, marine and terrestrial ecosystems across all continents. We analysed how species traits constrain body-size architecture by changing the slope of the predator-prey body-mass scaling. Across ecosystems, we found high body-mass ratios for predator groups with specific trait combinations including (1) small vertebrates and (2) large swimming or flying predators. Including the metabolic and movement types of predators increased the accuracy of predicting which species are engaged in high body-mass ratio interactions. We demonstrate that species traits explain striking patterns in the body-size architecture of natural food webs that underpin the stability and functioning of ecosystems, paving the way for community-level management of the most complex natural ecosystems.

Roots rather than shoot residues drive soil arthropod communities of arable fields.

Soil food webs are driven by plant-derived carbon (C) entering the soil belowground as rhizodeposits or aboveground via leaf litter, with recent research pointing to a higher importance of the former for driving forest soil food webs. Using natural abundance stable isotopes of wheat (C3 plant) and maize (C4 plant), we followed and quantified the incorporation of shoot residue- and root-derived maize C into the soil animal food web of an arable field for 1 year, thereby disentangling the importance of shoot residue- versus root-derived resources for arable soil food webs. On average, shoot residue-derived resources only contributed less than 12% to soil arthropod body C, while incorporation of root-derived resources averaged 26% after 2 months of maize crop and increased to 32% after 1 year. However, incorporation of root-derived maize C did not consistently increase with time: rather, it increased, decreased or remained constant depending on species. Further, preference of shoot residue- or root-derived resources was also species-specific with about half the species incorporating mainly root-derived C, while only a few species preferentially incorporated shoot residue-derived C, and about 40% incorporated both shoot residue- as well as root-derived C. The results highlight the predominant importance of root-derived resources for arable soil food webs and suggest that shoot residues only form an additional resource of minor importance. Variation in the use of plant-derived C between soil arthropod species suggests that the flux of C through soil food webs of arable systems can only be disentangled by adopting a species-specific approach.

Unifying elemental stoichiometry and metabolic theory in predicting species abundances.

While metabolic theory predicts variance in population density within communities depending on population average body masses, the ecological stoichiometry concept relates density variation across communities to varying resource stoichiometry. Using a data set including biomass densities of 4959 populations of soil invertebrates across 48 forest sites we combined these two frameworks. We analyzed how the scaling of biomass densities with population-averaged body masses systematically interacts with stoichiometric variables. Simplified analyses employing either only body masses or only resource stoichiometry are highly context sensitive and yield variable and often misleading results. Our findings provide strong evidence that analyses of ecological state variables should integrate allometric and stoichiometric variables to explain deviations from predicted allometric scaling and avoid erroneous conclusions. In consequence, our study provides an important step towards unifying two prominent ecological theories, metabolic theory and ecological stoichiometry.

Lack of energetic equivalence in forest soil invertebrates.

Ecological communities consist of small abundant and large non-abundant species. The energetic equivalence rule is an often-observed pattern that could be explained by equal energy usage among abundant small organisms and non-abundant large organisms. To generate this pattern, metabolism (as an indicator of individual energy use) and abundance have to scale inversely with body mass, and cancel each other out. In contrast, the pattern referred to as biomass equivalence states that the biomass of all species in an area should be constant across the body-mass range. In this study, we investigated forest soil communities with respect to metabolism, abundance, population energy use, and biomass. We focused on four land-use types in three different landscape blocks (Biodiversity Exploratories). The soil samples contained 870 species across 12 phylogenetic groups. Our results indicated positive sublinear metabolic scaling and negative sublinear abundance scaling with species body mass. The relationships varied mainly due to differences among phylogenetic groups or feeding types, and only marginally due to land-use type. However, these scaling relationships were not exactly inverse to each other, resulting in increasing population energy use and biomass with increasing body mass for most combinations of phylogenetic group or feeding type with land-use type. Thus, our results are mostly inconsistent with the classic perception of energetic equivalence, and reject the biomass equivalence hypothesis while documenting a specific and nonrandom pattern of how abundance, energy use, and biomass are distributed across size classes. However, these patterns are consistent with two alternative predictions: the resource-thinning hypothesis, which states that abundance decreases with trophic level, and the allometric degree hypothesis, which states that population energy use should increase with population average body mass, due to correlations with the number of links of consumers and resources. Overall, our results suggest that a synthesis of food web structures with metabolic theory may be most promising for predicting natural patterns of abundance, biomass, and energy use.

Body masses, functional responses and predator-prey stability.

The stability of ecological communities depends strongly on quantitative characteristics of population interactions (type-II vs. type-III functional responses) and the distribution of body masses across species. Until now, these two aspects have almost exclusively been treated separately leaving a substantial gap in our general understanding of food webs. We analysed a large data set of arthropod feeding rates and found that all functional-response parameters depend on the body masses of predator and prey. Thus, we propose generalised functional responses which predict gradual shifts from type-II predation of small predators on equally sized prey to type-III functional-responses of large predators on small prey. Models including these generalised functional responses predict population dynamics and persistence only depending on predator and prey body masses, and we show that these predictions are strongly supported by empirical data on forest soil food webs. These results help unravelling systematic relationships between quantitative population interactions and large-scale community patterns.