1. Ecologists are debating the relative role of deterministic and stochastic determinants of community structure. Although the high diversity and strong spatial structure of soil animal assemblages could provide ecologists with an ideal ecological scenario, surprisingly little information is available on these assemblages.
2. We studied species-rich soil oribatid mite assemblages from a Mediterranean beech forest and a grassland. We applied multivariate regression approaches and analysed spatial autocorrelation at multiple spatial scales using Moran’s eigenvectors. Results were used to partition community variance in terms of the amount of variation uniquely accounted for by environmental correlates (e.g. organic matter) and geographical position. Estimated neutral diversity and immigration parameters were also applied to a soil animal group for the first time to simulate patterns of community dissimilarity expected under neutrality, thereby testing neutral predictions.
3. After accounting for spatial autocorrelation, the correlation between community structure and key environmental parameters disappeared: about 40% of community variation consisted of spatial patterns independent of measured environmental variables such as organic matter. Environmentally independent spatial patterns encompassed the entire range of scales accounted for by the sampling design (from tens of cm to 100 m). This spatial variation could be due to either unmeasured but spatially structured variables or stochastic drift mediated by dispersal. Observed levels of community dissimilarity were significantly different from those predicted by neutral models.
4. Oribatid mite assemblages are dominated by processes involving both deterministic and stochastic components and operating at multiple scales. Spatial patterns independent of the measured environmental variables are a prominent feature of the targeted assemblages, but patterns of community dissimilarity do not match neutral predictions. This suggests that either niche-mediated competition or environmental filtering or both are contributing to the core structure of the community. This study indicates new lines of investigation for understanding the mechanisms that determine the signature of the deterministic component of animal community assembly.
Niche-based theories assume that species within a trophic level avoid competition for resources through trade-offs in life traits, which ultimately allow them to coexist locally on limited resources (Chase & Leibold 2003; Tilman 2004). In this framework, demographic stochasticity places upper limits on species richness because, once the system is saturated with tightly packed populations along critical environmental axes, populations become smaller and stochastic events affecting births and deaths are amplified. However, as long as species have relatively narrow niche breadths, deterministic niche-based competitive exclusion is the main process shaping community structure and usually produces species replacement along environmental gradients (Chase & Leibold 2003): local communities sampled at the two extremes of any relevant environmental gradient are expected to diverge in terms of species composition, whereas the closer two local communities are in terms of environmental conditions the more they converge. Two spatially close local communities are generally assembled under fairly similar environmental conditions, and this determines spatial patterns in the distribution of conditions and resources and, consequently, spatial autocorrelation in community dissimilarity (the closer communities are the more similar they are). However, in the niche framework, it does not really matter whether communities are spatially close to each other or not: it is the environment in which they are assembled that determines how similar/dissimilar they are. Two environmentally similar but distant localities are inhabited by communities that converge towards a certain species composition, whereas two close but environmentally dissimilar local communities diverge because their environments diverge, even if at relatively fine spatial scales. This principle, which forms the core of classical theories of community assembly, has been challenged in the last decade by neutral theories, which assume that community dynamics depend only on dispersal limitations and demographic stochasticity (e.g. Bell 2000; Hubbell 2001; Etienne 2007). Although this is clearly a simplistic view of community assembly, the main merit of theories based on this assumption is that they model population dynamics in terms of the individual-based stochastic processes of birth and death. The dispersal-limited sampling theory and neutrality assumption allow the formulation of many testable predictions in terms of patterns of community structure such as beta diversity (Alonso, Etienne & McKane 2006). If the assembly depends on neutral forces alone, local communities should diverge in composition because of dispersal limitations. However, the most important prediction is that divergence increases with increasing geographical distance between local communities. Therefore, as in the case of niche dynamics in spatially structured environments, neutral dynamics also produce spatial autocorrelation in species distribution. The crucial difference between niche and neutral dynamics is that the latter determines spatial patterns independent of environmental change (Hubbell 2001; Chave 2004; Dornelas et al. 2006) and the relative weight of environmentally independent spatial patterns can be detected by multivariate variance partitioning (e.g. Legendre & Legendre 1998; Cottenie 2005). A decade after the publication of the seminal book by Hubbell (2001), much of the current debate on the neutral theory now regards the relative role of stochastic and deterministic drivers of community structure. The question is no longer whether community assembly depends on niche partitioning mediated by environmental filtering or stochastic population drift mediated by limited dispersal, but rather in what proportion these two fundamental mechanisms shape natural communities (e.g. Chave 2004; Tilman 2004). In the framework of the variance partitioning technique, the niche unequivocally contributes to the variation accounted for by spatially independent environmental effects (i.e. after accounting for spatial autocorrelation). A spatially structured environmental effect could actually depend on a spurious correlation between community structure and environmental correlates (Legendre & Legendre 1998). Patterns resulting from limited dispersal are represented by the amount of variation accounted for by spatial autocorrelation independent of the environmental axes (i.e. after accounting for environmental variables).
However, variance partitioning does not always disentangle different assembly processes (Legendre & Legendre 1998; Smith & Lundholm 2010): first, one assumes that all the environmental variables explaining community structure were included in the analysis, which is arguably often not the case; secondly, a recent study based on simulations (Smith & Lundholm 2010) strongly suggests that the interaction between dispersal limitation and environmental selection actually affects the relative proportions of variance explained by environmental and spatial components.
Following another approach, the proponents of neutral models assert that all affirmations about niche-driven local communities converging in terms of species composition should be based on the comparison between the observed local communities and their hypothetical neutral counterparts (e.g. Etienne 2007). A quantitative neutral expectation in terms of community dissimilarity (Dornelas et al. 2006; Dornelas 2010) can therefore be used to detect the signature of the deterministic drivers of communities (e.g. niche partitioning), although the mechanisms responsible for this signature may remain unknown.
Within this context, soil animal communities provide a very interesting case for investigating the relative role of neutral and niche processes: their distribution is highly structured at multiple spatial scales, and the small body size of these animals allows the study of key community assembly processes at distances that are easy to manage during field studies (e.g. a few hundred metres). Over such short distances, taxonomic assemblages such as those of oribatids or collembolans have very high, even astonishing diversity that is often, though not always, coupled with limited dispersal capability (Ettema & Wardle 2002; Coulson, Hodkinson & Webb 2003; Lindo & Winchester 2009). Studies assessing the potential role of neutral dynamics (e.g. Caruso & Migliorini 2006; Lindo & Winchester 2009) surprisingly have not addressed the topic by combining methods proposed in the last 10 years (e.g. Cottenie 2005; Dornelas et al. 2006; Etienne 2007, 2009). There is still lively debate on the relevance of niche differentiation in explaining the so-called enigma of soil animal diversity (Anderson 1975). Although soil organisms may offer an interesting perspective on the general validity of ecological theories especially when applied to animal assemblages, relevant hypotheses based on neutral models have not been tested by soil animal ecologists.
We studied very diverse oribatid mite (Acari, Oribatida) assemblages from a Mediterranean beech forest and a grassland. Using a spatially explicit sampling design, we analysed community dissimilarities to assess the significance of habitat effects and spatial autocorrelation and quantify their relative contribution by means of multivariate variance partitioning (Legendre & Legendre 1998). Given the limits of variance partitioning, we then applied neutral models to a soil animal group for the first time to test whether observed levels of community dissimilarity were consistent with neutral expectations (Dornelas et al. 2006; Etienne 2007, 2009; Dornelas 2010; Caruso et al. 2011).
Materials and methods
Study Area and Sampling Strategy
We collected soil fauna samples from two main habitats located in the ‘Pietraporciana Nature Reserve’ (Italy): an ancient beech (Fagus sylvatica L.) forest and a natural grassland dominated by species such as Bromus erectus Huds. Soil fauna species, especially small-bodied microarthropods (e.g. oribatid mites that are about 0·5–0·25 mm long), have low dispersal capabilities (Berthet 1964; Ojala & Huhta 2001) and generally present highly aggregated spatial distributions at scales ranging from centimetres to tens and hundreds of metres. Unfortunately, the processes underlying these patchy distributions are poorly understood (but see for example Ettema & Wardle 2002; Keitt et al. 2002; Coulson, Hodkinson & Webb 2003). In the classical niche theory, species distribution and co-occurrence patterns at the microscale of the local community are ascribed to interspecific interactions and/or the environment, whereas in the neutral theory, these are considered the result of migration/dispersal dynamics only. To resolve this controversy, we applied a nested sampling design (Fig. 1) for discerning patterns acting at different scales, from the overall meta-community (each habitat or both together) to the local community (soil samples; Hubbell 2001; Etienne 2007). In each habitat, we defined a study area (500 × 500 m) in which we randomly sampled four plots (10 × 10 m), and within each plot, we randomly selected three subplots (1 × 1 m). Three soil samples for arthropod extraction were collected from each subplot using a box corer (8 × 8 × 8 cm). We thus obtained 36 community samples from each habitat. For each community, we took a supplementary soil sample for quantifying environmental factors: organic matter, texture and moisture. Distances between the beech forest and the grassland (a few hundred metres) represented the broad scale at which macroscopic patterns of the composition and abundance of soil animal communities should be obvious and easily interpreted in terms of basic ecological concepts (e.g. Bardgett 2005), as available data confirm (Migliorini, Petrioli & Bernini 2002). Within each habitat, the distance between plots (not less than 20 m) matched the scale of processes regulating intraspecific competition between individual trees within the forest, since 15 m is the average distance between beeches in this forest. Within each plot, distances between subplots match the scale of processes that can regulate the community structure of soil animals through changes in resources, e.g. those provided by microbial communities. At this scale, other important drivers are the architecture of the environment (e.g. soil aggregates), which may affect critical features such as moisture (Bardgett 2005). Soil samples from each subplot represent local communities whose microscale animal distribution and co-occurrence patterns are explained by the classical niche theory or neutral theory.
By employing a nested sampling strategy, we accounted for multiple spatial scales (Lindo & Winchester 2009). We also included several different ‘intervals’ (average distance separating sampling units) for the overall scale of the study (sensuLegendre & Legendre 1998); these ranged from tens of cm (subplot) to tens (plot) and hundreds (habitat) or) of m. This approach allowed us to detect patterns generated by fine- to broad-scale ecological processes that are presumed to play important roles (e.g. Underwood 1997) in soil fauna communities (e.g. Lindo & Winchester 2009).
Sample Processing and Analysis
We extracted Oribatid mites from each sample using a modified Berlese–Tullgren apparatus (animals preserved in 75% ethyl alcohol) and counted/identified adult individuals at the species level (see Appendix S1 in Supporting Information for a list of references used for taxonomic identification). Water content, texture, organic matter and biomass are considered the most important variables affecting oribatid assemblages in the study area (Migliorini, Petrioli & Bernini 2002). Soil organic matter content was estimated using sieved material (mesh size 2 mm) by loss on ignition at 550 °C for 4 h. All visible debris and coarse fragments were removed before sieving. Water content was determined by drying soil samples under the hood to a constant weight at room temperature. Dry subsamples were sieved through a 2-mm mesh to assess the relative proportion of coarse and fine material. All measures were expressed as % of soil weight. Lastly, the overall biomass of oribatids was measured on a sample basis using the methods reported in Caruso & Migliorini (2009). Oribatid biomass (sum of the biomass of all individuals in the sample) was intended as a proxy for the amount of energy available to the local community, which does not necessarily coincide with the amount of organic matter (see Fig. S1 in Appendix S1, Supporting Information).
To analyse patterns in community structure, we created a presence/absence and abundance matrix for species recorded in each local community: 72 soil samples (rows) × 156 species (columns). From these matrices, we created sample dissimilarity matrices (72 × 72) using the indices recommended by Anderson, Ellingsen & McArdle (2006). The choice of index did not alter our main conclusions (see Appendix S1, Supporting Information). We here report results obtained using the Jaccard distance, which accounts for change in species composition only (presence/absence) and is a very well-known, widely employed dissimilarity index (Anderson, Ellingsen & McArdle 2006; Appendix S1, Supporting Information).
Environmental factors were summarized in a matrix presenting water content, texture, organic matter and biomass values (four columns) in each local community (72 rows). Using this matrix, we calculated the correlation matrix of environmental factors (4 × 4), which we visualized and inspected through principal component analysis (PCA; see Fig. S1 in Appendix S1, Supporting Information). Water was highly correlated with organic matter and was excluded in posterior analysis to avoid the inclusion of collinear predictors. Only organic matter, oribatid biomass and coarse material were therefore used for posterior analyses.
From the geographical distance matrix of soil samples, spatial patterns were summarized in the Moran eigenvector mapping matrix (MEM) that best accounted for autocorrelation (Dray, Legendre & Peres-Neto 2006). Spatial autocorrelation accounts for how the response value observed in a locality is predictable in terms of response values observed in neighbouring localities. The eigenvector matrix (MEM) is the entrywise product of (i) a connectivity binary matrix representing a graph in which samples are the vertices to be connected in a network and (ii) a weighting matrix that assigns a weight to each link of the network using functions that quantify sample dissimilarity. Multiple spatial patterns can be tested by modifying connectivity and/or weighting matrices (Dray, Legendre & Peres-Neto 2006), and model selection procedures (Akaike information criterion, AIC; Burnham & Anderson 2002) can be applied to select the best linear combination of eigenvectors that maximizes their correlation with the data and minimizes their number (Dray, Legendre & Peres-Neto 2006). Quantitative details are given in Appendix S1 (Supporting Information) and the cited references (Borcard & Legendre 2002; Dray, Legendre & Peres-Neto 2006). Once extracted, each eigenvector accounts for spatial patterns at a certain scale, so that the overall matrix of eigenvectors accounts for multiple spatial scales. The matrix of spatial eigenvectors (below MEM) can then be used as a predictor in multivariate regression approaches (Dray, Legendre & Peres-Neto 2006). To disentangle the relative role of environmental and spatial factors, we applied redundancy analysis (RDA) and variance partitioning (Legendre & Legendre 1998). The response matrix was the community matrix, and the variables used as explanatory factors were environmental vectors such as organic matter, biomass and coarse material and spatial autocorrelation vectors (MEM matrix). Variance partitioning can be used to quantify the unique contribution of these two components plus the confounding effect of spatially structured environmental variation, which makes it difficult to disentangle the effects of neutral and niche forces (e.g. Smith & Lundholm 2010).
Given the limitations of variance partitioning, we completed the analysis with the novel aid of neutral models. The neutral diversity (θ) and immigration (Hubbell’s m or Etienne’s I: Hubbell 2001; Etienne 2007) parameters were estimated using a recent version of the neutral sampling formula for multiple samples by Etienne (2009). Estimation of θ and I using a formula that accounts for multiple samples allows the creation of artificial local communities of the same size as the real communities (Etienne 2007, 2009). We then used Jaccard’s index to calculate the ecological distance (below we use the expression ‘community dissimilarity’) separating the simulated communities. We thus obtained a null but neutral (i.e. dynamic) expectation that can be compared with that of the real community (Dornelas et al. 2006).
We collected a total of 8875 oribatid mites belonging to 156 species (see Appendix S1 in Supporting Information for a complete checklist and the species distribution). The total size of the beech forest community Jbeech= 5744, and the total number of species Sbeech= 124; there were 31 singletons, and the most abundant (771 ind.) species was Medioppia media, which is typical of forest habitats. For the grassland, Jgrass= 3130 and Sgrass= 72; there were 16 singletons, and the most abundant (500 ind.) species was Zygoribatula exarata, which is very typical of Mediterranean open habitats.
MEMs selected as the best vectors for accounting for spatial autocorrelation in the data revealed that the oribatid community is spatially structured at multiple spatial scales: broad (500–100 m) to intermediate (100–10 m) spatial patterns were observed at the habitat (beech forest vs. grassland) and plot scale (e.g. see MEM1 and MEM3 in Fig. S2 in Appendix S1 in Supporting Information), whereas intermediate to fine-scale patterns were observed within plots (10 m, e.g. MEM 10) and subplots (1 m, e.g. MEM 69). These spatial patterns may depend on either an underlying spatially structured environmental variation or processes such as limited dispersal. RDA analysis and the Monte Carlo permutation test showed that the effects of organic matter, biomass and coarse material were not significant (P ≥ 0·05 for all variables and the entire model) after removing the conditional effect of the MEM matrix (i.e. of spatial autocorrelation), whereas the same test performed on the MEM matrix after removing the conditional effect of organic matter, biomass and coarse material showed that all eigenvectors but MEM1 were significant (in all case P ≪ 0·05). The reason why MEM1 was not significant after removing the effects of organic matter, biomass and coarse material is that most of the variation in these variables depended on the spatial segregation of the beech forest from the grassland, and MEM1 mostly accounted for spatial patterns in the community that were dependent on this segregation (e.g. Fig. S3 in Appendix S1, Supporting Information). Variance partitioning showed that the overall relative contribution of pure spatial patterns, spatially structured environmental changes and pure environmental effects was, respectively, 40%, 2% and 2%. Therefore, the total contribution of environmental vectors plus spatially structured environmental effects only accounted for 4% of community structure. The test of homogeneity of multivariate dispersion (β) indicated that βgrass > βbeech (P < 0·001). This is clearly seen in a PCoA representation of Jaccard distances (Fig. 2a). Interestingly, the environmental variables were much more variable in the beech forest than in the grassland (Fig. 2b and Fig. S1). The estimates of neutral model parameters θ and I and the results of the neutrality test based on predicted and observed levels of dissimilarity revealed (Table 1; Fig. 3) that neutrality is always rejected, irrespective of whether it was assumed that the two habitats belonged to the same metacommunity or that (more likely) each habitat had its own metacommunity (i.e. θ and I were estimated for each individual habitat). Local communities always diverged (Fig. 3) with respect to levels of dissimilarity predicted by the neutral model (i.e. they were less similar than predicted by the neutral model). Interestingly, the size effect was greater for the beech forest than for the grassland.
Table 1. Estimates of neutral model parameters (Etienne 2009) for the beech forest (beech), the grassland (grass) and the hypothetical metacommunity obtained by merging samples from the two habitats into one data set
J, community size; S, observed richness; θ, neutral diversity (Hubbell 2001); m, average immigration parameter in terms of migration rate (calculated by averaging the parameter estimate obtained for each local community according to Etienne 2009); P, significance level of a bootstrapped t-test to test the hypothesis that mean levels of observed (Obs) Jaccard distances were significantly shifted with respect to the mean of neutral communities simulated (Sim) using the estimated θ and m values. See also Fig. 3.
Beech + Grass
Obs > Sim (P <0·001)
Obs > Sim (P <0·001)
Obs > Sim (P <0·001)
The most important ecological feature of the study area is that the beech forest and the grassland differ markedly in terms of species composition and diversity of oribatid mites. For example, more than 50% of the 156 detected species were not present in both habitats. Furthermore, differences observed in terms of community structure are consistent with the main environmental differences that generally distinguish a mature forest from a grassland (i.e. soil moisture, structure and organic matter). Soil animals such as oribatid mites feed on litter and soil micro-organisms such as fungi (e.g. Schneider & Maraun 2005): by grazing significantly, they therefore affect the detritus food web (Bardgett 2005). The clear differences between the grassland and beech forest communities can be simply explained by the niche theory, because the observed variance in community structure seems to reflect the average environmental variance between the two habitats.
The adopted sampling design and statistical modelling (MEMs: Dray, Legendre & Peres-Neto 2006) allowed us to detect several spatial structures in the community from the habitat (500 m) to the subplot (1 m) scale. The most obvious of these spatial patterns is the broadest one: most of the variation observed in the community corresponds to the mere spatial separation of the beech forest from the grassland (see MEM1 in Figs S2 and S3 in Appendix S1, Supporting Information). Therefore, community patterns at this scale do not actually provide definitive statistical proof that the community is structured by the observed environmental differences between the two habitats (Legendre & Legendre 1998). In fact, variance partitioning results indicate that once we accounted for the fact that the correlation between community patterns and environmental variables is spatially structured at a scale corresponding to the geographical distance between the two habitats, a multitude of spatial patterns remained significant and were uncorrelated with the measured environmental variables. Overall, these spatial patterns accounted for 40% of community variation. This large amount of variation is consistent with that observed not only in other soil assemblages but also in different types of communities (see the meta-analysis by Cottenie 2005).
Neutral models predict that spatial patterns in community structure can be explained by limited dispersal only, independently of environmental variables. Given that only 4% of community variation was explained by environmental variables and 40% was explained by spatial autocorrelation independent of environmental variation, one might conclude that patterns of community structure in the studied assemblages suggest the prevalence of neutral forces. However, variance partitioning is limited by the fact that it is hardly possible to take into account all environmental variables relevant to niche dynamics. A recent study (Smith & Lundholm 2010) has shown that the relative proportions of variance explained by environmental and spatial components partially depend on the degree of interaction between the two. Several authors have instead proposed that an appropriate null hypothesis for testing the niche theory is paradoxically provided by the neutral models developed in recent years (Bell 2000; Hubbell 2001; Alonso, Etienne & McKane 2006; Etienne 2007, 2009) and explicitly tested in this study. Results clearly reject models based on neutral theories. Note that neutral models predicted that local communities in the beech forest were experiencing a higher rate of immigration than those in the grassland. According to the neutral theory (Hubbell 2001), this is consistent with the fact that the beech forest has much lower beta diversity (Fig. 2a). However, observed levels of dissimilarity were always significantly greater than those predicted by the neutral model: both the beech forest and the grassland have higher levels of dissimilarity, and their local communities diverge with respect to their neutral counterparts (Fig. 3). This important finding implies that patterns of beta diversity in Fig. 2a cannot be interpreted in the light of demographic stochasticity and limited dispersal alone, as models based on these mechanisms failed to predict patterns of ecological dissimilarity or beta diversity (Dornelas et al. 2006; Etienne 2007). Given that the data refuted the neutral hypothesis and despite the fact that variance partitioning indicated that environmentally independent spatial effects are prominent features of the communities, we accept the hypothesis that environmental filtering and/or niche-mediated competition are operating and, in particular, that demographic stochasticity and limited dispersal are not the only forces driving community structure. The divergence of communities suggests that high environmental heterogeneity is sorting species into relatively heterogeneous local assemblages (Dornelas et al. 2006). We exclude a disturbance regime, given that the study area is in a nature reserve showing no sign of physical alteration. Some environmental axes should account for the observed community patterns; however, the ones we measured had limited explanatory power. Despite the range of scales we could resolve with our sampling design and modelling strategy, and the apparently obvious ecological differences between a grassland and a beech forest, most (56%) of the community structure remained unexplained and only a small fraction (2%) of community variation was accounted for by the spatial structure in the environmental variables or a spatially independent environmental effect (2%). We found a total of 156 species, most of which were not shared by the two habitats. Our data therefore raise the question of the unsolved ‘enigma of soil diversity’ (Anderson 1975): how is it possible that such a large number of species coexist in a relatively homogenous environment such as soil? Wardle (2002) stated that soil actually provides both high niche dimensionality (owing to its three-dimensional structure) and the opportunity to partition resources considerably. Furthermore, differential rates of activity can enhance coexistence by avoiding competition (Chase & Leibold 2003). Differential rates of activity and local patchy variations in resource availability, soil structure or features affecting life cycles (e.g. temperature) may increase local species richness by favouring local niche partitioning and increasing beta diversity (Wardle 2002). The local variations in our environmental variables are not consistent with this hypothesis because, for example, organic matter and water content were more variable in the beech forest (Fig. 3b), which has lower beta diversity (Fig. 3a) but higher species richness (Table 1, estimates of θ). Nevertheless, the pattern of heterogeneity observed in the environmental variables (Fig. 2b) matches that observed in the neutral analysis (Fig. 3, compare the more divergent beech with the less divergent grass), which offers a novel key for interpreting results. Indeed, there thus remain three not mutually exclusive possibilities: some environmental variables driving local communities at relatively fine scales were not measured, although we detected the strong spatial structures they determine (environmental filtering); local species interaction determines a substantial amount of community structure (niche), such that the neutral model was rejected and beta diversity was highly variable (Figs 2b and 3); temporal variations in the spatial patterns of the environment not accounted for in this study explain spatiotemporal variations in the assemblages. We cannot exhaustively address this latter potentially critical point. Nevertheless, we have good temporal information on the seasonal variability of the arthropod community in the study area (Migliorini, Petrioli & Bernini 2002), confirming that the analysed data set is representative of the community in terms of diversity and species composition. The large number of singletons might suggest that several species have not been sampled. Estimates of actual richness based on rarefaction curves (Chao 1, Jackknife 1, and Bootstrap: Magurran 2004; : Table S2 in Appendix S1, Supporting Information) showed that observed richness values did indeed underestimate total richness, especially in the beech forest. This result may be due not only to the very high diversity of the studied system but also to the use of the Berlese–Tullgren apparatus, which exposes soil arthropods to high thermal stress. However, under the dry conditions of Mediterranean areas (including relatively moister habitats such as beech forests), the Berlese–Tullgren apparatus has generally proved to be very efficient for estimating the composition of taxa such as oribatid mites (see Migliorini, Petrioli & Bernini 2002). Note that the main result of the analysis based on neutral models did not depend on whether we calculated community dissimilarity using presence/absence data or abundance data (for the latter, see Fig S4 in Appendix S1, Supporting Information). Based on these pieces of information, we can state that the most important results of the study (rejection of neutral models, high beta diversity and strong spatial structure from fine to broad scales) are fairly robust to temporal changes and sampling bias; nevertheless, temporal replication of spatial data could provide important information for a more mechanistic interpretation of the main patterns presented in this study. For example, well-replicated temporal data would allow us to test the hypothesis that the various species have demographic peaks in different seasons, which can greatly increase the spatiotemporal region of coexistence by avoiding competition through temporal segregation (Chase & Leibold 2003). Previous data on the study area (Migliorini, Petrioli & Bernini 2002) are fairly consistent with this hypothesis. As for the unmeasured fine-scale variations in environmental variables, note that changes in soil structure have been proposed as crucial determinants for soil animals, which in turn also positively contribute to soil structure (Wardle 2002; Bardgett 2005). We broadly quantified soil structure by the percentage of coarse material (>2 mm) in the soil: this was sufficient to distinguish the soils of the two habitats, but was later found to have a low discriminatory power at finer scales. We also roughly quantified organic matter in a broad sense by measuring its overall amount in the soil by ignition; however, stoichiometric ratios of key elements such as C, N and P may have a much greater explanatory potential (as demonstrated for the most disparate soil organisms), considering that these ratios drive the microbiological community on which most microarthropod species feed (Bardgett 2005). As for the latter point, note that the feeding habits of oribatids and other microarthropods such as springtails have been investigated in detail over the last years using different approaches ranging from food choice experiments to isotopic analysis (e.g. Schneider et al. 2004; Schneider & Maraun 2005). Results indicate that these taxa have different degrees of specialization and highly variable feeding habits. If one of the life traits accounting for niche differences was scale dependent or highly unpredictable, neutral-like and niche processes could interplay over a range of spatial scales, possibly resulting in complex spatial patterns such as the multiple patterns observed in our study. To address this point in the future, we propose that studies such as those conducted by Schneider et al. (2004) and Pollierer et al. (2009), who investigated trophic levels and habits through a careful and standardized use of stable isotope ratios, should be coupled with an analysis of community structure such as the one presented herein. For example, a critical issue is that stable isotopes revealed a continuity of feeding spectra within taxonomic assemblages such as oribatid assemblages, which are classically believed to be relatively homogeneous in terms of feeding ecology. There are a few species that show isotopic signatures close to the litter layer instead; these can thus be considered true decomposers. Within the same assemblage, there are even a few species that show isotopic signatures typical of predators in the ground above. As most species are in an intermediate position, the classical concept of discrete and well-defined trophic levels simply does not apply to soil food webs. Strictly speaking, the neutral theory should be applied to a community of species belonging to the same trophic level and potentially competing for a relatively homogenous set of resources (Hubbell 2001), although this definition is rather broad. Soil animal assemblages such as oribatid mite assemblages exemplify what animal ecologists have recently started to observe in the soil: sets of communities that are intermediate between the classical definitions of the community analysed by neutral models and real systems with several more or less discrete trophic levels. A new generation of theories and observations that take into account multiple spatiotemporal scales and different environments are thus required to unravel the causes of the complex community patterns arising in animal communities because of interacting stochastic and deterministic processes.
This study was partly supported by Freie Universität Berlin, MIUR and University of Siena grants. We thank Robin Hankin and Jeff Powell for their helpful comments on a previous version of the manuscript. T.C. was supported by the Alexander von Humboldt Foundation. This work is a contribution to a DFG standard grant to T.C. and Stefan Hempel (Freie Universität Berlin). We are grateful to two anonymous reviewers for very constructive criticisms that substantially improved earlier versions of the manuscript.