Relative roles of spatial factors, environmental filtering and biotic interactions in fine-scale structuring of a soil mite community

Soil Biology & Biochemistry 79 (2014) 68e77

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Relative roles of spatial factors, environmental filtering and biotic interactions in fine-scale structuring of a soil mite community Meixiang Gao, Ping He, Xueping Zhang*, Dong Liu, Donghui Wu** a

Key Laboratory of Wetland Ecology and Environment, Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun 130012, China b Key Laboratory of Remote Sensing Monitoring of Geographic Environment, College of Heilongjiang Province, Harbin Normal University, Harbin 150025, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 April 2014 Received in revised form 31 August 2014 Accepted 2 September 2014 Available online 18 September 2014

Community theories suggest that community structuring depends on dispersal limitation, environmental filtering and biotic interactions. However, the relative roles of these factors at fine scale are less well understood. In this study, we attempt to determine the relative roles of spatial factors, environmental filtering and biotic interactions in the fine-scale (5 m) structuring of a soil mite community from a temperate deciduous forest in the Maoershan Ecosystem Research Station in northeastern China. In August 2012, we established three plots and collected 100 samples from each plot in a 5
5 m2 area using a spatially delimited sampling design. To quantify the relative contributions of the spatial and environmental processes, Moran's eigenvector maps (MEMs), variation partitioning analysis and partial Mantel test were used. Null and neutral models were used to disentangle the effects of biotic interactions. Null mode analyses were conducted for non-random patterns of species co-existence and significant species-pairs in the assemblage of soil mites, and to determine whether the observed pattern was the result of biotic interactions. The neutral model was used to identify whether the community structure shows divergence, convergence or neutrality. The results indicated that the relatively large and significant variance was due to spatial factors in all plots. The contribution of environmental filtering was relatively low and non-significant in all plots based on variation partitioning, while it was significant in Plot II based on a partial Mantel test. Soil organic matter content, soil pH, and soil and litter water content explained a significant part of the variance observed in the distribution of the mite community. Furthermore, the null model revealed a non-random co-occurrence pattern in the soil mite community, and the environmental niche overlap indicated a weak contribution of biotic interactions. The observed mean dissimilarity implied significant divergence in communities based on neutral model analysis. Collectively, these results emphasize that both spatial and environmental processes were important drivers in the fine-scale structuring of soil mite communities in a temperate deciduous forest and that biotic interactions were less influential in the observed pattern.

Keywords: Spatial factors Environmental filtering Biotic interactions Soil mite community Fine-scale Temperate deciduous forest

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction Many studies have questioned the manner in which species form assemblages and the rules that govern this process. According to niche and neutral theories in community ecology, the

* Corresponding author. Tel.: þ86 13804576310. ** Corresponding author. Tel.: þ86 15904308293. E-mail addresses: [email protected] (M. Gao), [email protected] (P. He), [email protected] (X. Zhang), [email protected] (D. Liu), [email protected]. cn (D. Wu). http://dx.doi.org/10.1016/j.soilbio.2014.09.003 0038-0717/© 2014 Elsevier Ltd. All rights reserved.

composition of species assemblages can be explained by three processes: dispersal limitation, environmental filtering and biotic interactions (Drake, 1990; Weiher and Keddy, 2001; Gilbert and Lechowicz, 2004; Leibold et al., 2004; Tews et al., 2004). Neutral theories suggest that species are ecologically equivalent and that community structure relies strongly on stochastic processes and dispersal limitation (Hubbell, 2001). In contrast, niche theories emphasize that the appearance of species in a specific habitat is based on biotic interactions and environmental filtering, the latter filtering species from the regional species pool (Diamond, 1975; Chase and Leibold, 2003; Webb et al., 2010). In fact, these

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

theories are not mutually exclusive and evidence for both has been reported for various communities at different spatial scales (Lindo and Winchester, 2009; Dumbrell et al., 2010; Caruso et al., 2012b). However, their relative roles are still not well known (Chase and Myers, 2011; Winegardner et al., 2012). Indeed, the relative roles of spatial factors, environmental filtering and biotic interactions in community structuring are hypothesized to be scale-dependent. Environmental heterogeneity is expected to function at larger scales (Weiher and Keddy, 1995;
n, 2010), Swenson et al., 2007; Cavender-Bares et al., 2009; Sobero because the substantial variation in environmental variables allows species to co-exist. At larger spatial scales, community structuring may be constrained by spatial factors (dispersal limitation) (Declerck et al., 2011; Sokol et al., 2013; Tang et al., 2013) and dispersal limitation might override environmental filtering in certain habitat types (Bello et al., 2013). In addition, the influence of biotic interactions becomes gradually more important as the spatial scale decreases and they might have no obvious effect at scales larger than small (101e103 m) or fine (<101 m) (Hortal et al., 2010). Moreover, in order to describe the underlying mechanisms of species distribution, it is essential to explicitly consider the variables in multiscales (Hortal et al., 2010). However, processes at the small scale often have been neglected when recognizing processes at large scales and it is difficult to infer that variation at the small scale is regulated by stochastic processes (Anderson et al., 2011). Disentangling the mechanisms at larger spatial scales might be challenging, as insufficient considering underlying processes at relative fine-scales (Anderson et al., 2011; Caruso et al., 2012b). Soil harbors a large diversity of organisms, which represent most of the world's terrestrial biodiversity (Wardle, 2002; Bardgett, €ns, 2010). However, 2005; Bardgett and Wardle, 2010; Decae fundamental questions relating to the causes and maintenance of this diversity remain only partially answered (Bardgett, 2002;
ttir et al., 2012; Caruso Lindo and Winchester, 2009; Ingimarsdo et al., 2012b, 2013). Environmental variability is known to allow for the spatial co-existence of competing earthworm species, which
nez emphasizes the importance of environmental filtering (Jime et al., 2012). On the other hand, Caruso et al. (2013) reported that biotic interactions might be a predominant factor in the structuring of soil metacommunity dynamics. Other publications have also demonstrated the important contributions of biotic competition, environmental filtering, or/and spatial factors to the structuring of €ns et al., 2008; Lindo and soil animal communities (Decae
ttir et al., 2012; Caruso et al., Winchester, 2009; Ingimarsdo 2012b). However, the question of the relative roles of these variables at the fine-scale remains little studied. Soil mite communities represent ideal assemblages to test the relative roles of the underlying processes in fine-scale community structuring. According to niche theories, environmental heterogeneity can create varied micro-conditions for co-occurring species, whereas environmental conditions seem to be more homogeneous at the fine scale and thus the environmental heterogeneity may be less important in maintaining diverse species at this level. When considering the highly spatial connectivity and reachability, dispersal might be sufficient to allow species to survive in microhabitats with suitable environmental characteristics, in other words, dispersal might be not limiting (Fuentes, 2002). Thus, it may be that spatial factors are not important in regulating community structure at fine scale. Soil mite communities have large diversity at small spatial scales. Given the highly similar resources at the finescale, the neighboring coexisting species largely share or compete for limited local resources and spaces (Weiher and Keddy, 1995; Kraft and Ackerly, 2010). Consequently, the biotic interactions have the most opportunity to be a major structuring force at the
mez et al., 2010). However, those fine scale (Wardle, 2006; Go

69

processes may operate simultaneously in a fine-scale community (Logue et al., 2011; Winegardner et al., 2012) and perhaps different in intensity. In this study we addressed the relative contributions of spatial factors, environmental filtering and biotic interactions on the finescale structuring of soil mite communities in a temperate deciduous forest at the Maoershan Ecosystem Research Station in northeastern China. We tested two hypotheses in this study: 1) spatial factors and environmental filtering should play relative minor roles for these communities and 2) biotic interactions should play an important role in community structuring at the fine scale. 2. Materials and methods 2.1. Study site The study was performed at the Maoershan Ecosystem Research Station (127 300 e340 E, 45 200 e250 N) of the Northeast Forestry University in Heilongjiang Province, China. This area is covered with typical forests of northeastern China. The region lies within a continental temperate monsoon climate. The climate is characterized by an average annual temperature of 2.8  C and an average annual precipitation of 884 mm. The average altitude is approximately 300 m and the average degree of slopes is about 10e15 . The type of parent material is granite bedrock and the type of soil is a Hap-Boric Luvisol (Gong et al., 1999). The annual evaporation is approximately 884 mm. The frost-free days are about 120e140. Soil mite communities were collected in a temperate deciduous forest at the Maoershan Ecosystem Research Station. This location has a 60-yr old secondary forest with an 18 m tall canopy layer. The main tree species include Fraxinus mandshurica Rupr., Ulmus davidiana Planch. var. japonica (Rehd.) Nakai, Betula platyphylla Suk., Populus davidiana Dode, Juglans mandshurica Maxim., Acer mono Maxim., Tilia amurensis Rupr. and Populus ussuriensis Kom., and the main shrub species include Syringa reticulata (Blume) Hara var. amurensis (Rupr.) Pringle, Padus racemosa (Lam.) Gilib., Acer ginnala Maxim. and Corylus mandshurica Maxim. 2.2. Soil mite communities and soil sampling Three experimental plot replicates (5
5 m2) were established at the study site in August 2012. The distance between each replicate (Plot I, Plot II and Plot III) was more than 60 m. Each plot was divided into 100 squares of 0.5
0.5 m2. Soil mite samples were obtained from the left-bottom area of each square. Square soil samples (15
15 cm2 and 10 cm depth) were collected for the extraction of mite communities. These communities were removed from 500 ± 5 g of the square soil samples using the Berlese-Tullgren
rrez-Lo
pez et al., 2010), with mites method (Krantz, 1978; Gutie being preserved in a 95% alcohol solution. Adult soil mites were identified to the species level and then were counted respectively (Balogh and Balogh, 1992; Yin et al., 1998; Walter and Proctor, 2001; Krantz and Walter, 2009). Soil mite juveniles were excluded from
ttir et al., 2012). all analyses (Ingimarsdo A vegetation-free soil core sample (5
5 cm2 and 10 cm depth) was taken directly to the right of each square used to sample the soil mites. These soil samples were air-dried and then sieved to 1 mm. To obtain soil's organic matter content, the colorimetric method after digestion in H2SO4 was used. The soil water content was determined gravimetrically. In deionized water with a soil/ solution ratio of 1:5, the soil pH was obtained (Lao, 1988; Pansu and Gautheyrou, 2003). In addition to the plant species richness (the number of vascular plant species), the DBH (diameter at breast height) and basal diameter of vascular plant, the litter dry weight (10
10 cm2) and the litter water content were measured.

70

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

2.3. Statistical analysis Differences in species richness and abundance of the mite communities between three plots were tested using a one-way ANOVA (analysis of variance). As the variances were not homogeneous in all plots, the one-way ANOVA and post hoc Tamhane-T2 test were used.

2.3.1. Relative roles of spatial factors and environmental filtering When considering the co-varying environmental variables in each plot, a principal component analyses (PCA), grounded on the correlation matrices, was selected to reduce the multivariate variables to uncorrelated variables. As independent environmental variables, the principal components (PCs) were used for subsequent
mez et al., 2010). analyses of each plot (Go The powerful approach of distance-based Moran's eigenvector maps (MEMs, formerly named PCNMeprincipal coordinates of neighbor matrices) (Borcard and Legendre, 2002; Dray et al., 2006; Legendre and Gauthier, 2014) was used to reveal the effects of spatial variables on the community assembly. MEMs analysis produces a range of spatial variables derived from the geographic coordinates of each plot (Dray et al., 2006). The MEMs figure spatial variation across multiple spatial scales and can be served as explanatory spatial factors of community variation (Diniz-Filho and Bini, 2005; Peres-Neto and Legendre, 2010). A Hellinger transformation was applied to the species abundance matrices prior to analysis (Legendre and Gallagher., 2001; Borcard et al., 2011). In Plot I, Plot II and the Mean Plot, significant linear trends were found, so the data of soil mite communities were detrended. The residuals were then used as input data in later spatial analyses of those plots and the non-detrended data were used in Plot III. Next, a forward selection process was performed based on the adjusted R-square to choose the linear combination of vectors that describes the most variation in the mite species matrix with the lowest possible number of vectors (Dray et al., 2006; Borcard et al., 2011). After the forward selection procedure, 9, 10, 3 and 8 MEMs were chosen in Plot I, Plot II, Plot III and the Mean Plot, respectively; these explained 8.91, 13.61, 3.61 and 7.58% of the variation of the soil mite community (P < 0.05), respectively. Variation partitioning was then used to infer the relative contribution of spatial factors and environmental filtering on the soil mite community matrix of each plot (Legendre and Legendre, 1998; Cottenie, 2005; Blanchet et al., 2014). The chosen environmental variables (PCs) and spatial variables (MEMs) were analyzed with the powerful, partial redundancy analysis (pRDA) method. A pRDA allows the total variation of a species matrix in each plot to be partitioned into fractions that represent the contribution of the pure environmental fraction [a], the spatially structured environmental fraction (shared fraction) [b], the pure spatial fraction [c], and the unexplained fraction [d] (Peres-Neto et al., 2006; Borcard et al., 2011). The significance of each source of variation was tested with a Monte Carlo permutation test (999 permutations). In a patchy environmental resource condition, a combination of spatially explicit analysis and partial Mantel test should be helpful for revealing the roles of environmental and spatial variables
nez et al., 2011). A partial Mantel test was calculated to exam (Jime whether the community dissimilarities depended on spatial distances or environmental distances (as an agent for environmental filtering) (Kristiansen et al., 2012; Wang et al., 2013). Although a partial Mantel test might not be suitable for partitioning the variation in community assembly (Legendre et al., 2005, 2008; Legendre and Fortin, 2010), it can be helpful in recognizing whether there is obvious distance decay.

The multivariate analyses and partial Mantel test were implemented using the vegan (Oksanen et al., 2013) and packages in the R software, version 3.0.1. 2.3.2. Null model and neutral model analysis To determine whether patterns of species co-occurrence in the overall matrix showed signs of non-random processes, we performed a formal null model analysis (Gotelli, 2000; Gotelli and Ulrich, 2012). The commonly used indices of C-score and V-ratio were used (Gotelli, 2000; Gotelli and McCabe, 2002). The cooccurrence procedures are very sensitive to variation in species occurrence frequencies, hence row totals were saved as a constraint in the null model (Gotelli, 2000). Additionally, the V-ratio is determined by the row and column sums of the matrix (Gotelli, 2000), therefore it is not valid for the FF (fixedefixed) algorithm. Thus, the FF, FE (fixedeequiprobable) and FP (fixedeprobability) algorithms were used to calculate the C-score, and the FE and FP algorithms were used to calculate the V-ratio (Gotelli, 2000; Gotelli and Ulrich, 2010, 2012). The C-score is an aggregate index that describes the average behavior of a metric that is calculated on a species-pair basis. Consequently, important information can be detected by testing for non-random patterns in species associations on a species-pair basis. Nevertheless, this approach creates the statistical problem that the amount of possible pairs is high. This problem considerably increases the risk of Type I errors (Gotelli and Ulrich, 2010). Thus, this analysis was performed by calculating the C-score for each species pair and then identifying its significance using the four methods suggested by Gotelli and Ulrich (2010). Of these, the CL (confidence limit criterion) is the most liberal and simplest method. For the CL, when the number of species pairs in a matrix is large, then 5% of them will fall outside the 95% confidence limits simply by chance (Gotelli and Ulrich, 2010; Krasnov et al., 2011). To address this problem, the conservative criteria of BY (after sequential Bonferroni correction) (Benjamini and Yekutieli, 2001), BM (empirical Bayes mean-based criterion) and BCL (empirical Bayes confidence limits-based criterion) were introduced (Gotelli and Ulrich, 2010). The steps undertaken to fulfill the four criteria can be found in Gotelli and Ulrich (2010). Next, the standardized effect size (SES) was calculated to evaluate the direction and extent of deviation from the null model. The SES evaluates the degree of standard deviations by which the investigated index is higher or lower than the mean index of the simulated assemblages (Gotelli and McCabe, 2002). Supposing the SES values a normal distribution, a 95% confidence interval of the SES values is supposed to occur between 2.0 and 2.0. The C-score and V-ratio indices were obtained using the Ecosim 7.0 software (Gotelli and Entsminger, 2009). Significant species-pairs were obtained based on the PAIRS software (Ulrich, 2008). The neutral model has suggested as a specific form of more general null model in community ecology (Gotelli and McGill, 2006). Moreover, the neutral model is a valuable tool for community ecologists with dynamic quantitative expectation in the light of community dissimilarity (Dornelas, 2010). Thus, a neutral model was also applied to identify whether the community showed significant convergence, divergence or just neutrality. Based on the species abundance distribution (SAD) in each plot and the Mean plot, the neutral diversity (q) and immigration parameters (Hubbell's m or Etienne's I) (Hubbell, 2001; Etienne, 2007) were evaluated according to a neutral sampling formula for multiple samples, as introduced by Etienne (Etienne, 2007, 2009). The formula allows one to figure an exam of q and m (Etienne, 2007, 2009). Fundamental diversity (q) represents the product of the size of the metacommunity and the speciation rate (Etienne, 2007), and the average immigration parameter (m) calculates the amount of

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

dispersal limitation of the local community and can be explained as the number of potential immigrants competing with local individuals for vacant sites (Etienne and Olff, 2004; Etienne, 2009). The PARI/GP codes written by Etienne (Etienne, 2007, 2009) were used to calculate q and m, from which expected SADs were produced. Jaccard indices were then used to identify community dissimilarities based on species abundance for both the observed and expected communities, which were compared with a bootstrapped t test (Efron and Tibshirani, 1993; Caruso et al., 2012a, 2012b). Then, we performed a meta-analysis of the results of calculated and expected community dissimilarities to explore the effects of moderators (i.e. predicted effect). Notionally, the moderators are partitioned into three models: neutral, divergent and convergent. Based on a random-effects model, a continuous meta-analysis was performed. Then a Kendall's rank correlation (Begg and Mazumdar, 1994; Viechtbauer, 2010) was selected to evaluate the funnel plot asymmetry. All analyses were performed in R platform with “vegan”, “bootstrape” and “metafor” packages (Viechtbauer, 2010). 2.3.3. Pinka Ojk niche overlap index To identify whether the pattern obtained was due to interspecific competition or environmental filtering that structured the soil mite community, Pianka's Ojk (Pianka, 1973), a community-level niche overlap index, was performed using the mean niche overlap of the total possible pairwise species. The range of the Pianka Ojk metrices is between 0 and 1. We also figured the values of the SES €ns et al., 2009). A value smaller than expected by chance (Decae represents a competitively assemblage, which perhaps resulted from biotic interactions, especially interspecific competition. A value above that expected by chance indicates that all species have similar patterns of resource utilization, which may be caused by environmental constraints (Albrecht and Gotelli, 2001). The niche partitioning for environmental variables was calculated from the individual matrices, in which the rows represented individual species and the columns represented environmental variables (PCs). Each entry represented the number of individuals caught in
nez each plot for a given range of the environmental variables (Jime €ns et al., 2012). For a detailed description of the method, see Decae
nez et al. (2006, 2012) and Pianka (1973). et al. (2009), Jime Calculation and tests of Pinka Ojk niche overlap were carried out using the “niche overlap” module of the Ecosim 7.0 software (Gotelli and Entsminger, 2009).

71

Table 1 Species richness (number of mite species in each plot) and abundance (total number of individual mites in each plot) of the soil mite communities in a deciduous forest at Maoershan Mountain Station (northeastern China). Plot I Macrocheles sp. Pachyseius sp. Epicriidae sp. Gamasolaelaps sp. Nanhermannia sp. Eulohmannia sp. Belba sp. 1 Scheloribates sp. Suctobelbella sp. Geholaspis sp. Protoribates sp. Oribatida sp. Acrotritia ardua (Koch, 1841) Prostigmata sp. Ceratozetes sp. Holaspulus sp. Belba sp. 2 Hypochthonius sp. Trombidiidae sp. Species richness (J)a Total number of mites (S)a

d

603 (D) 3302 (E) 530 (D) NFb 67 (B) 1331 (E) 739 (E) 1 (A) 2671 (E) 184 (C) 182 (C) 1184 (E) 602 (D) 745 (D) 1284 (E) 126 (B) 14 (A) 1321 (E) 2 (A) 18H 14 888H

Plot II

Plot III

Mean plotc

598 (E) 3874 (E) 232 (B) 1 (A) 11 (A) 1701 (E) 126 (C) 2 (A) 2149 (E) 832 (B) 919 (D) 276 (D) 243 (D) 14 (A) 227 (D) 275 (D) 1 (A) 361 (D) 2 (A) 19I 11 844I

575 (E) 3412 (E) 110 (B) 48 (A) 47 (16) 1657 (E) 210 (C) 63 (A) 1965 (E) 397 (70) 1149 (E) 386 (E) 230 (D) 102 (A) 314 (D) 229 (D) 39 (A) 483 (D) 1 (A) 19H 11 417I

592 (E) 3529 (E) 291 (E) 16 (A) 42 (C) 1563 (E) 358 (E) 22 (A) 2262 (E) 471 (E) 750 (E) 615 (E) 358 (E) 287 (D) 608 (E) 210 (E) 18 (A) 722 (E) 2 (A) 19 12 716

a Different letters indicate significant differences. J represents community size and S represents observed richness when estimating the neutral model parameters for each plot and the Mean plot. b NF indicates not found. c Data in the Mean plot were the averages of Plots I, II and III combined. d Raunkiaer's frequency class. A: 1e20%; B: 21%e40%; C: 41e60%; D: 61e80%; E: 81e100%.

3.3. Relative contributions of spatial factors and environmental filtering The amount of variation accounted for by the pure spatial fraction [c] was relatively large and significant in all plots. The variation explained by the pure environmental fraction [a] and the spatially structured environmental fraction [b] were relatively lower and non-obvious in each plot (Fig. 2). Based on the results of the redundancy analysis, the effects of the specific PCs were obvious in each plot (Table 2). In plot II, the community dissimilarity exhibited obvious correlation coefficients with the environmental variables after accounting for the effect of spatial variables (Table 3).

3. Results 3.1. Community assembly In total, 38 149 soil mites individuals were detected. The most frequent and abundant soil mite species were Pachyseius sp., Eulohmannia sp. and Suctobelbella sp. The species richness of the soil mite community was obviously different between Plots I and II, and between Plots II and III. Additionally, the abundance of the soil mite community was significantly different between Plots I and II, and between Plots I and III (Table 1). Overall, the individual-based rarefaction curves (Gotelli and Colwell, 2001) showed that the sampling efforts were sufficient to describe the overall richness of the soil mite communities (Supplementary Material Appendix 1). 3.2. Results of environmental principal components Eigenvalues of the first four axes of the PC explained 80.82, 78.84, 79.15 and 73.56% of the total inertia for Plot I, Plot II, Plot III and the Mean plot, respectively. Four PCs were selected for each plot (Fig. 1). A summary of the statistics for the environmental variables analyzed is given in the Supplementary Material Appendix 2.

3.4. Species co-occurrence, significant species-pairs, niche overlap and neutral model According to the C-score with the FF algorithm for Plot II and Plot III, species co-occurrence showed significant spatial segregation. The co-occurrence patterns of the other assemblies, which showed non-randomness based on the C-score or V-ratio with FF or FE algorithms, all showed significantly spatial aggregations (Table 4). In terms of the pair-based co-occurrence analysis with the FE algorithm based on the four criteria, there were more significantly aggregated species-pairs than segregated species-pairs in all plots. According to the same analysis with the FF algorithm, for those plots in which significant species pairs were identified, there were more significantly segregated species-pairs than aggregated species-pairs, except for the Mean plot based on the CL criterion, and Plot III based on the BM criterion (Supplementary Material Appendix 3). The average values of community Ojk for the environmental factors were 0.86, 0.81, 0.83 and 0.89 in Plot I, Plot II, Plot III and in

72

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

Fig. 1. The first four axes derived from principal component analysis (PCA) of the environmental variables in each plot. The factor coordinates (arrows) are built from the PC1, PC2, PC3 and PC4 eigenvector coefficients. LDW-litter dry weight (g), LWC-litter water content (w/w%), SWC-soil water content (w/w%), PR-plant species richness (the number of vascular plant species), GD-basal diameter of vascular plant (mm), DBH-diameter at breast height (mm), SOM-soil organic matter content (g kg1), pH-soil pH. (a), (b), (c) and (d) represent Plot I, Plot II, Plot III and the Mean plot.

the Mean plot, respectively. For all environmental variables, the observed values of the Ojk niche overlap were larger than the simulated values. Except for PC2 in Plot II, the values of the SES were obviously larger than 2, which indicates that soil mite communities were not competitively structured (Table 5). According to the evaluation of the neutral model parameters and the corresponding test based on observed and simulated mean dissimilarities, neutrality can be rejected. Obvious funnel plot asymmetry was not founded (P > 0.05) and significant heterogeneity was detected (P < 0.001) (Fig. 3).

4. Discussion 4.1. Relative roles of spatial factors and environmental filtering Our analyses suggest that pure spatial pattern had a significant influence on the fine-scale structuring of each mite community. The analysis based on MEMs allowed us to identify spatial factors (dispersal limitation and demographic stochasticity) in mite communities. The spatial pattern might result from dispersal-related processes, such as physical barriers or dispersal ability (Cottenie,

Fig. 2. Variation partitioning for soil mite community in each plot tested by partial redundancy analysis (pRDA). Pure environmental [a], pure spatial [c] and shared fractions [b] are provided. Negative values are not shown. (a), (b), (c) and (d) represent Plot I, Plot II, Plot III and the Mean plot. **P < 0.01.

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77 Table 2 The effect of environmental factors on the soil mite community structures analyzed by redundancy analysis and Monte Carlo permutation test (999 permutations). Factor

Plot I

PC1a PC2 PC3 PC4

P P P P

Plot II

¼ 0.11 ¼ 0.07 < 0.001 ¼ 0.03

P P P P

Plot III

¼ 0.06 < 0.001 ¼ 0.12 ¼ 0.03

P P P P

Mean plot

< 0.01 ¼ 0.11 ¼ 0.03 ¼ 0.02

P P P P

¼ 0.96 < 0.001 < 0.001 ¼ 0.36

a PC indicates each of the factors that were obtained from the PCA for each of the data sets.

Table 3 Partial Mantel test of soil mite community dissimilarity against spatial distance and environmental distance for each plot (999 permutations). Plot I

EnvironmentjSpacea SpacejEnvironmentb

Plot II

Plot III

Mean plot

R

P

R

P

R

P

R

P

0.01 0.06

0.38 0.90

0.10 0.01

0.01 0.40

0.07 0.01

0.11 0.57

0.04 0.03

0.18 0.76

a Soil mite community dissimilarity with environmental distance, controlling for spatial distance. b Soil mite community dissimilarity with spatial distance, controlling for environmental distance.

€nroos et al., 2013), 2005; Soininen et al., 2007; Soininen, 2012; Gro and it is more important when considering the lower active dis€nroos et al., persers (Cottenie, 2005; Soininen et al., 2007; Gro 2013). Active dispersal is the primary means of reaching different locations for most soil mite species. Terrestrial oribatid mite assemblages are restricted by low dispersal activities at the local scale € nroos et al., 2013). Fine-scale (Lindo and Winchester, 2009; Gro dispersers show significant spatial limitation, perhaps because their limited dispersal capabilities prevent them from tracking € nroos et al., 2013). Unenvironmental micro-heterogeneity (Gro fortunately, information on the dispersal abilities of the soil mite species found in the study area are still lacking and this aspect needs further investigation.

73

According to neutral theory, community structure can be explained by spatial factors independent of environmental variables. However, observed community dissimilarity of each plot and the Mean plot were obviously greater than those predicted using a neutral model. A quantitative neutral model based on community dissimilarity (Dornelas et al., 2006; Etienne, 2007) can be used to recognize the sign of the deterministic regulator of assembly (Caruso et al., 2012b), nevertheless the underlying mechanisms accounting for this sign still indistinct (Gotelli and McGill, 2006; Caruso et al., 2012b). It can be inferred that community structuring can be controlled not only by spatial factors but also by other processes, such as environmental filtering. The significant heterogeneity and divergence of each community based on neutral model analysis suggest that environmental variables might filter species into local communities (Dornelas et al., 2006). Although a much smaller and non-significant fraction of variation was accounted for by the spatially structured environmental factors and the pure environmental factors, we cannot simply exclude the effect of environmental filtering on soil mite assemblages. The soil environmental factors (soil organic matter content, soil pH value, soil water content and litter water content) were important for the soil mite communities in Plots I and II. The above-ground vegetation factors (basal diameter, plant species richness and DBH) were important in Plot III. It is difficult to conclude that above-ground plant variables govern soil mite community assemblages. The food resources of soil mite communities include litter ingredient and soil fungi (Siepel and Ruiter-Dijkman, 1993; Schneider and Maraun, 2005). Therefore, the driver of soil mite community assemblies is most likely not the plants themselves. Instead, their influence is probably indirect: plant litter affects soil organic matter and soil water content, which in turn affect the soil microorganisms
ttir et al., 2012). In on which mite communities feed (Ingimarsdo
nez et al., this research, an environmental regulating structure (Jime 2012) was detected in each plot according to the analysis of the environmental niche partitioning. Moreover, the results of the partial Mantel test indicated a significant contribution of environmental fractions in Plot II. In addition, according to Raunkiaer's

Table 4 Species co-occurrence analysis for testing whether species distributions can be approximated by random distributions generated by statistical models (null model analysis). Plot

Index

Null model

Observed indexa

Mean of simulated indexb

SESc

Initial P valued

Corrected Pe

Plot I

C-score

FE FF FP FE FP FE FF FP FE FP FE FF FP FE FP FE FF FP FE FP

110 110 110 1.45 1.45 130 130 130 1.93 1.93 112 112 112 2.46 2.46 18.7 18.7 18.7 1.33 1.33

122 108 118 1.00 1.15 161 127 151 1.00 1.28 148 104 137 1.00 1.39 21.3 17.9 21.1 1.00 1.02

2.76 1.00 1.63 3.33 1.92 6.03 3.04 3.44 6.86 3.93 7.94 6.95 4.39 10.7 5.95 1.33 1.23 1.22 2.50 2.35

<0.01 0.84 0.06 <0.01 0.04 <0.001 <0.01 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.09 0.11 0.11 0.01 0.02

0.01 0.84 0.06 <0.01 0.06 <0.001 <0.01 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.27 0.17 0.11 0.02 0.02

V-ratio Plot II

C-score

V-ratio Plot III

C-score

V-ratio Mean plot

C-score

V-ratio a

The observed indices are those values that were calculated from the observed assemblages. Mean of simulated indices were values after 50 000 randomizations. c A SES (standardized effect size) is calculated as [(observed value-mean of simulated value)/standard deviation of simulated value]. For the C-score, a SES >2 indicates significant species segregation, and a SES <2 indicates significant species aggregation. For the V-ratio, a SES >2 indicates significant species aggregation, and a SES <2 indicates significant species segregation. d Initial tail probability of observed indices that were obtained by random permutations. e The corrected P-associated tailed probability (P < 0.05) is indicated after the false discovery rate (FDR) procedure (Benjamini and Yekutieli, 2001). b

74

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

Table 5 Community niche analysis for selected environmental principal components. Plot

Environmental variableb

Observed index

Mean of simulated index

SES

Initial Pc

Corrected Pd

Plot Ia

PC1 PC2 PC3 PC4 PC1 PC2 PC3 PC4 PC1 PC2 PC3 PC4 PC1 PC2 PC3 PC4

0.96 0.83 0.80 0.85 0.96 0.70 0.82 0.74 0.95 0.79 0.81 0.84 0.95 0.89 0.86 0.88

0.48 0.77 0.75 0.80 0.44 0.68 0.59 0.72 0.16 0.71 0.63 0.74 0.73 0.72 0.80 0.79

20.4 6.63 4.68 5.96 20.0 1.49 12.5 2.20 22.5 7.09 11.6 9.54 19.7 14.3 6.96 11.4

<0.001 <0.001 <0.01 <0.001 <0.001 0.92 <0.001 0.04 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

<0.001 <0.001 <0.01 <0.001 <0.001 0.92 <0.001 0.05 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

Plot IIa

Plot IIIa

Mean plota

a Average niche overlap for each plot was calculated for multidimensional environmental principal components by averaging the single Ojk values for each resource that was
nez et al., 2012). exploited in the environmental principal components by the community and compared with a null model (30 000 simulations) (Jime b PC indicates each of the factors that was obtained from the PCA for each of the data sets. c The initial P value indicates the probability that the standardized effect size (SES) differed from zero. d The corrected P value indicates the probability at P < 0.05, after the FDR procedure correction.

frequency class distribution in all plots, more than 50% of the species belong to classes D and E (Table 1), meaning that over half the species are distributed broadly in each plot. Therefore, we can infer that the mite species were widely distributed in different micro-habitats, indicating the important governing abilities of the
nez et al., 2012). micro-environmental spatial heterogeneity (Jime 4.2. Relative role of biotic interactions The co-occurrence patterns of the soil mite communities showed significantly non-random for all plots. According to the results of our null model analysis on the overall species matrix, a significantly aggregated spatial co-occurrence pattern was found in all plots taking into account the C-score and V-ratio with the FE and FP algorithms, whereas a significantly segregated spatial cooccurrence pattern was only detected in Plots II and III according to the C-score with the FF algorithm. Based on the pairwise null model analysis, when the soil mite assemblage was segregated according to the C-score with the FF algorithm (Plots II and III), there were more segregated species pairs than aggregated species pairs. Additionally, when the soil mite assemblages were aggregated according to the C-score with the FE algorithm, there were more aggregated species pairs than segregated species pairs. The co-occurring distribution of the observed significant species pairs was in line with the overall matrix based on the same null model

algorithm. However, few significant negative species pairs were detected based on the pairwise null model analysis using the conservative criterion. The contribution of these negative species pairs to the non-random community structure may be insignificant due to dilution effects (Gotelli and Ulrich, 2010; Krasnov et al., 2011). Moreover, the analysis of neutral model also implies a little possible function of biotic interactions in community structuring. What is more, the results of the environmental niche partitioning suggested a common environmental structuring constraint rather
ttir et al., than interspecific competition (Leibold, 1998; Ingimarsdo
nez et al., 2012). Greater micro-variations in the envi2012; Jime ronment allow for species co-existence, hence environmental process might predominate over the contribution of biotic interactions. Actually, oribatid mite species are belong to three or four trophic levels (Schneider et al., 2004; Maraun et al., 2011). Within the same mite assemblages, true decomposers and predators have been demonstrated and others occupy intermediate positions (Schneider et al., 2004; Maraun et al., 2011; Caruso et al., 2012b). Additionally, predators also control community structures (Wardle, 2002, 2006). Thus we suggest that the stable-isotope method should be performed in the future (Schneider et al., 2004; Maraun et al., 2011). We also suggest that a functional trait-based approach (Bello et al., 2013; Liu et al., 2013; Algarte et al., 2014; Michel and Knouft, 2014) and a phylogenetic method (Cavender-Bares et al., 2009) should also be carried out.

Fig. 3. Results based on the meta-analysis evaluating deviations from predictions of the neutral model. The null hypothesis is that mean observed dissimilarities (Jaccard dissimilarity index) between any two samples are the same as those expected under ecological neutrality. Thus, a positive effect size (observed mean dissimilarity e expected mean dissimilarity) indicates that dissimilarity is higher than predicted (divergence), while a negative effect size indicates that dissimilarity is lower than predicted (convergence). The null hypothesis corresponds to the vertical line at zero (neutrality). Error bars are 95% confidence limits. q is neutral diversity (Hubbell, 2001). m is the average immigration parameters in terms of migration rate (calculated by averaging the parameter estimate obtained for each local community, after Etienne, 2009). For the two other neutral model parameters (J-community size; S-observed richness), please see Table 1. ***P < 0.001. (a) e Plots I, II, III and the Mean plot were considered during meta-analysis process. (b) e Plots I, II and III were considered during meta-analysis process.

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

The relative contributions of spatial factors, environmental filtering and biotic interactions in soil animal community structuring vary depending upon the spatial scales and study objects. At the landscale scale (104  2
105 m) (Hortal et al., 2010),
ttir et al. (2012) found collembolan and oribatid species Ingimarsdo avoid each other, but emphasized the contribution of spatial and environmental filtering. At the local scale (103e104 m) (Hortal et al., 2010), spatial patterns and environmental factors (Caruso et al., 2012b) and biotic interactions (especially interspecific competition) (Caruso et al., 2013) are important drivers for micro-soil animal communities. At the small scale (101e103 m) (Hortal et al.,
nez et al., 2011, 2012) and 2010), soil environmental filtering (Jime biotic interactions (Nachman and Borregaard, 2010) permit species to co-exist. At the fine scale (<101 m) (Ettema and Wardle, 2002; Hortal et al., 2010), our study showed that both spatial and environmental processes govern co-existence assemblages, with little evidence being found for the contribution of biotic interactions. According to our results, the underlying processes at fine scale seem complicated and significant. During the processes of identifying the rules of community structuring at regional or local scales, the underlying processes at the fine scales should not be ignored. Despite a density sampling strategy and an above/below-ground environmental factors collecting design, most (about 90% on average) of the community composition remained unexplained. Performing comparisons and interpreting results should therefore be made with care, since pure spatial and environmental variables can be confused by different sources of errors (Smith and Lundholm, 2010). For example, pure spatial patterns explained by dispersal limitation might simply result from unmeasured environmental variables (Smith and Lundholm, 2010; Anderson et al., 2011). The effects of pure environmental patterns are also related to niche-mediated competition (Caruso et al., 2012b) and the extent of measured variables (Jones et al., 2008). Although there are critiques about the ability of variation partitioning to identify the relative roles of environmental and spatial variables (Tuomisto and Ruokolainen, 2008; Gilbert and Bennett, 2010), it is helpful for uncovering community pattern (Anderson et al., 2011). Some publications have shown that it can be efficiently applied (DinizFilho et al., 2012) and it has been widely performed for this purpose (Caruso et al., 2013; Sokol et al., 2013; Michel and Knouft, 2014). At the fine-scale, the unexplained variation might result from those unmeasured and important environmental factors (Borcard et al., 2004), such as vertical distributions (soil depth and structure) (Anderson and Hall, 1977), through its threedimensional structure variability, which may provide high niche differentiation and the opportunity to segment resources (Wardle,
nez et al., 2006). The unexplained variation might also 2002; Jime arise from unconsidered temporal variations, which could also provide important information for identifying underlying mechanisms. Opportunities for species co-existence in a community may be increased by temporal segregation (Chase and Leibold, 2003; Dumbrell et al., 2011). We therefore suggest that future studies of community structuring should better consider both spatial and temporal variations in environmental factors. Collectively, these results suggest that both spatial factors and environmental filtering were important drivers in the fine-scale (5 m) structuring of a soil mite community in the temperate deciduous forest in northeastern China, whereas biotic interactions were less influential in the observed pattern. Acknowledgments We thank Pierre Legendre, Jiangshan Lai and Xiangcheng Mi for their assistance in the MEMs and variation partitioning analysis.
Jime
nez for his assistance with the We thank Juan-Jose

75

environmental niche overlap analysis. We thank Xiaoguang Du for his assistance with neutral model analysis. We thank Judson Mark for his assistance in English polishing. We thank managers of the Maoershan Ecosystem Research Station for the convenience they provided. We also thank the anonymous reviewers. This study was supported by the National Natural Sciences Foundation of China (grant nos. 41101049, 41471037, 40601047, 41371072, 41430857 and 31200331) and by the China Postdoctoral Science Foundation (grant no. 2012M511361).

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.soilbio.2014.09.003.

References Albrecht, M., Gotelli, N.J., 2001. Spatial and temporal niche partitioning in grassland ants. Oecologia 126, 134e141. Algarte, V.M., Rodrigues, L., Landeiro, V.L., Siqueira, T., Bini, L.M., 2014. Variance partitioning of deconstructed periphyton communities: does the use of biological traits matter? Hydrobiologia 722, 279e290. Anderson, J.M., Hall, H., 1977. Cryptostigmata species diversity and soil habitat structure. Ecological Bulletins 25, 473e475. Anderson, M.J., Crist, T.O., Chase, J.M., Vellend, M., Inouye, B.D., Freestone, A.L., Sanders, N.J., Cornell, H.V., Comita, L.S., Davies, K.F., Harrison, S.P., Kraft, N.J.B., Stegen, J.C., Swenson, N.G., 2011. Navigating the multiple meanings of b diversity: a roadmap for the practicing ecologist. Ecology Letters 14, 19e28. Balogh, J., Balogh, P., 1992. The Oribatid Mites Genera of the World (Vol. 1 and 2). The Hungarian National Museum Press, Budapest, Hungary. Bardgett, R., 2005. The Biology of Soil: a Community and Ecosystem Approach. Oxford University Press, Oxford. Bardgett, R.D., 2002. Causes and consequences of biological diversity in soil. Zoology 105, 367e375. Bardgett, R.D., Wardle, D.A., 2010. Aboveground-belowground Linkages: Biotic Interactions, Ecosystem Processes, and Global Change. Oxford University Press, Oxford. Begg, C.B., Mazumdar, M., 1994. Operating characteristics of a rank correlation test for publication bias. Biometrics 50, 1088e1101. Bello, F.d., Vandewalle, M., Reitalu, T., Leps, J., Prentice, H.C., Lavorel, S., Sykes, M.T., 2013. Evidence for scale- and disturbance-dependent trait assembly patterns in dry semi-natural grasslands. Journal of Ecology 101, 1237e1244. Benjamini, Y., Yekutieli, D., 2001. The control of the false discovery rate in multiple testing under dependency. The Annals of Statistics 29, 1165e1188. Blanchet, F.G., Legendre, P., Bergeron, J.A.C., He, F., 2014. Consensus RDA across dissimilarity coefficients for canonical ordination of community composition data. Ecological Monographs 84, 491e511. http://dx.doi.org/10.1890/18130648.1891. Borcard, D., Gillet, F., Legendre, P., 2011. Numerical Ecology with R. Springer, New York. Borcard, D., Legendre, P., 2002. All-scale spatial analysis of ecological data by means of principal coordinates of neighbour matrices. Ecological Modelling 153, 51e68. Borcard, D., Legendre, P., Avois-Jacquet, C., Tuomisto, H., 2004. Dissecting the spatial structure of ecological data at multiple scales. Ecology 85, 1826e1832. Caruso, T., Hempel, S., Powell, J.R., Barto, E.K., Rillig, M.C., 2012a. Compositional divergence and convergence in arbuscular mycorrhizal fungal communities. Ecology 93, 1115e1124. Caruso, T., Taormina, M., Migliorini, M., 2012b. Relative role of deterministic and stochastic determinants of soil animal community: a spatially explicit analysis of oribatid mites. Journal of Animal Ecology 81, 214e221. Caruso, T., Trokhymets, V., Bargagli, R., Convey, P., 2013. Biotic interactions as a structuring force in soil communities: evidence from the micro-arthropods of an Antarctic moss model system. Oecologia 172, 495e503. Cavender-Bares, J., Kozak, K.H., Fine, P.V.A., Kembel, S.W., 2009. The merging of community ecology and phylogenetic biology. Ecology Letters 12, 693e715. Chase, J.M., Leibold, M.A., 2003. Ecological Niches: Linking Classical and Contemporary Approaches. The University of Chicago Press, Chicago. Chase, J.M., Myers, J.A., 2011. Disentangling the importance of ecological niches from stochastic processes across scales. Philosophical Transactions of the Royal Society B: Biological Sciences 366, 2351e2363. Cottenie, K., 2005. Integrating environmental and spatial processes in ecological community dynamics. Ecology Letters 8, 1175e1182. €ns, T., 2010. Macroecological patterns in soil communities. Global Ecology and Decae Biogeography 19, 287e302. €ns, T., Jime
nez, J.J., Rossi, J.P., 2009. A null-model analysis of the spatioDecae temporal distribution of earthworm species assemblages in Colombian grasslands. Journal of Tropical Ecology 25, 415e427.

76

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77

€ns, T., Margerie, P., Aubert, M., Hedde, M., Bureau, F., 2008. Assembly rules Decae within earthworm communities in north-western France e a regional analysis. Applied Soil Ecology 39, 321e335. Declerck, S.A.J., Coronel, J.S., Legendre, P., Brendonck, L., 2011. Scale dependency of processes structuring metacommunities of cladocerans in temporary pools of High-Andes wetlands. Ecography 34, 296e305. Diamond, J.M., 1975. Assembly of Species Communities. Harvard University Press, Cambridge. Diniz-Filho, J.A.F., Bini, L.M., 2005. Modelling geographical patterns in species richness using eigenvector-based spatial filters. Global Ecology and Biogeography 14, 177e185. Diniz-Filho, J.A.F., Siqueira, T., Padial, A.A., Rangel, T.F., Landeiro, V.L., Bini, L.M., 2012. Spatial autocorrelation analysis allows disentangling the balance between neutral and niche processes in metacommunities. Oikos 121, 201e210. Dornelas, M., 2010. Disturbance and change in biodiversity. Philosophical Transactions of the Royal of Society B 365, 3719e3727. Dornelas, M., Connolly, S.R., Hughes, T.P., 2006. Coral reef diversity refutes the neutral theory of biodiversity. Nature 440, 80e82. Drake, J.A., 1990. Communities as assembled structures: do rules govern pattern? Trends in Ecology & Evolution 5, 159e164. Dray, S., Legendre, P., Peres-Neto, P.R., 2006. Spatial modelling: a comprehensive framework for principal coordinate analysis of neighbour matrices (PCNM). Ecological Modelling 196, 483e493. Dumbrell, A.J., Ashton, P.D., Aziz, N., Feng, G., Nelson, M., Dytham, C., Fitter, A.H., Helgason, T., 2011. Distinct seasonal assemblages of arbuscular mycorrhizal fungi revealed by massively parallel pyrosequencing. New Phytologist 190, 794e804. Dumbrell, A.J., Nelson, M., Helgason, T., Dytham, C., Fitter, A.H., 2010. Relative roles of niche and neutral processes in structuring a soil microbial community. The ISME Journal 4, 337e345. Efron, B., Tibshirani, R.J., 1993. An Introduction to the Bootstrap. Chapman and Hall, New York. Etienne, R.S., 2007. A neutral sampling formula for multiple samples and an ‘exact’ test of neutrality. Ecology Letters 10, 608e618. Etienne, R.S., 2009. Improved estimation of neutral model parameters for multiple samples with different degrees of dispersal limitation. Ecology 90, 847e852. Etienne, R.S., Olff, H., 2004. How dispersal limitation shapes speciesebody size distributions in local communities. The American Naturalist 163, 69e83. Ettema, C.H., Wardle, D.A., 2002. Spatial soil ecology. Trends in Ecology & Evolution 17, 177e183. Fuentes, M., 2002. Seed dispersal and tree species diversity. Trends in Ecology & Evolution 17, 550.
mez, J.P., Bravo, G.A., Brumfield, R.T., Tello, J.G., Cadena, C.D., 2010. A phylogenetic Go approach to disentangling the role of competition and habitat filtering in community assembly of Neotropical forest birds. Journal of Animal Ecology 79, 1181e1192. Gilbert, B., Bennett, J.R., 2010. Partitioning variation in ecological communities: do the numbers add up? Journal of Applied Ecology 47, 1071e1082. Gilbert, B., Lechowicz, M.J., 2004. Neutrality, niches, and dispersal in a temperate forest understory. Proceedings of the National Academy of Sciences of the United States of America 101, 7651e7656. Gong, Z., Chen, Z., Luo, G., Zhang, G.L., Zhao, W., 1999. Soil reference with Chinese soil taxonomy. Soils 31, 57e63 (in Chinese). Gotelli, N.J., 2000. Null model analysis of species co-occurrence patterns. Ecology 81, 2606e2621. Gotelli, N.J., Colwell, R.K., 2001. Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecology Letters 4, 379e391. Gotelli, N.J., Entsminger, G.L., 2009. Ecosim: Null Models Software for Ecology, Version 7. Acquired Intelligence Inc. and Kesey-Bear: Jericho, VT, USA. http:// garyentsminger.com/ecosim.htm. Gotelli, N.J., McCabe, D.J., 2002. Species co-occurrence: a meta-analysis of J.M. Diamond's assembly rules model. Ecology 83, 2091e2096. Gotelli, N.J., McGill, B.J., 2006. Null versus neutral models: what's the difference? Ecography 29, 793e800. Gotelli, N.J., Ulrich, W., 2010. The empirical Bayes approach as a tool to identify nonrandom species associations. Oecologia 162, 463e477. Gotelli, N.J., Ulrich, W., 2012. Statistical challenges in null model analysis. Oikos 121, 171e180. €nroos, M., Heino, J., Siqueira, T., Landeiro, V.L., Kotanen, J., Bini, L.M., 2013. Gro Metacommunity structuring in stream networks: roles of dispersal mode, distance type, and regional environmental context. Ecology and Evolution 3, 4473e4487.
rrez-Lo
pez, M., Jesús, J.B., Trigo, D., Fern
Gutie andez, R., Novo, M., Díaz-Cosín, D.J., 2010. Relationships among spatial distribution of soil microarthropods, earthworm species and soil properties. Pedobiologia 53, 381e389. Hortal, J., Roura-Pascual, N., Sanders, N.J., Rahbek, C., 2010. Understanding (insect) species distributions across spatial scales. Ecography 33, 51e53. Hubbell, S.P., 2001. The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press, Princeton.

ttir, M., Caruso, T., Ripa, J., Magnúsdo
ttir, O.B., Ingimarsdo Migliorini, M., Hedlund, K., 2012. Primary assembly of soil communities: disentangling the effect of dispersal and local environment. Oecologia 170, 745e754.
nez, J.J., Decae €ns, T., Ame
zquita, E., Rao, I., Thomas, R.J., Lavelle, P., 2011. ShortJime range spatial variability of soil physico-chemical variables related to earthworm

clustering in a neotropical gallery forest. Soil Biology and Biochemistry 43, 1071e1080.
nez, J.J., Decae €ns, T., Rossi, J.P., 2006. Stability of the spatio-temporal distribuJime tion and niche overlap in Neotropical earthworm assemblages. Acta Oecologica 30, 299e311.
nez, J.J., Decae €ns, T., Rossi, J.P., 2012. Soil environmental heterogeneity allows Jime spatial co-occurrence of competitor earthworm species in a gallery forest of the Colombian ‘Llanos’. Oikos 121, 915e926. Jones, M.M., Tuomisto, H., Borcard, D., Legendre, P., Clark, D.B., Olivas, P.C., 2008. Explaining variation in tropical plant community composition: influence of environmental and spatial data quality. Oecologia 155, 593e604. Kraft, N.J.B., Ackerly, D.D., 2010. Functional trait and phylogenetic tests of community assembly across spatial scales in an Amazonian forest. Ecological Monographs 80, 401e422. Krantz, G.W., 1978. A Manual of Acarology. Oregon State University Book Stores Inc., Corvallis, Oregon. Krantz, G.W., Walter, D.E., 2009. A Manual of Acarology, third ed. Texas Tech University Press, Lubbock, Texas. Krasnov, B.R., Shenbrot, G.I., Khokhlova, I.S., 2011. Aggregative structure is the rule in communities of fleas: null model analysis. Ecography 34, 751e761. Kristiansen, T., Svenning, J.C., Eiserhardt, W.L., Pedersen, D., Brix, H.,
ndez, C., Balslev1, H., 2012. Environment Kristiansen, S.M., Knadel, M., Gra versus dispersal in the assembly of western Amazonian palm communities. Journal of Biogeography 39, 1318e1332. Lao, J.C., 1988. Handbook of Soil Agrochemistry Analysis. Agriculture Publishing House, Beijing. Legendre, P., Borcard, D., Peres-Neto, P.R., 2005. Analyzing beta diversity: partitioning the spatial variation of community composition data. Ecological Monographs 75, 435e450. Legendre, P., Borcard, D., Peres-Neto, P.R., 2008. Analyzing or explaining beta diversity? Comment. Ecology 89, 3238e3244. Legendre, P., Fortin, M.J., 2010. Comparison of the Mantel test and alternative approaches for detecting complex multivariate relationships in the spatial analysis of genetic data. Molecular Ecology Resources 10, 831e844. Legendre, P., Gallagher, E., 2001. Ecologically meaningful transformations for ordination of species data. Oecologia 129, 271e280. Legendre, P., Gauthier, O., 2014. Statistical methods for temporal and space-time analysis of community composition data. Proceedings of the Royal Society B: Biological Sciences 281, 1e9. Legendre, P., Legendre, L., 1998. Numerical Ecology. Developments in Environmental Modelling. Elsevier, Amsterdam. Leibold, M.A., 1998. Similarity and local co-existence of species in regional biotas. Evolutionary Ecology 12, 95e110. Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J.M., Hoopes, M.F., Holt, R.D., Shurin, J.B., Law, R., Tilman, D., Loreau, M., Gonzalez, A., 2004. The metacommunity concept: a framework for multi-scale community ecology. Ecology Letters 7, 601e613. Lindo, Z., Winchester, N.N., 2009. Spatial and environmental factors contributing to patterns in arboreal and terrestrial oribatid mite diversity across spatial scales. Oecologia 160, 817e825. Liu, J., Soininen, J., Han, B.P., Declerck, S.A.J., 2013. Effects of connectivity, dispersal directionality and functional traits on the metacommunity structure of river benthic diatoms. Journal of Biogeography 40, 2238e2248. Logue, J.B., Mouquet, N., Peter, H., Hillebrand, H., 2011. Empirical approaches to metacommunities: a review and comparison with theory. Trends in Ecology & Evolution 26, 482e491. Maraun, M., Erdmann, G., Fischer, B.M., Pollierer, M.M., Norton, R.A., Schneider, K., Scheu, S., 2011. Stable isotopes revisited: their use and limits for oribatid mite trophic ecology. Soil Biology and Biochemistry 43, 877e882. Michel, M.J., Knouft, J.H., 2014. The effects of environmental change on the spatial and environmental determinants of community-level traits. Landscape Ecology 29, 467e477. Nachman, G., Borregaard, M.K., 2010. From complex spatial dynamics to simple Markov chain models: do predators and prey leave footprints? Ecography 33, 137e147. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, Pierre, Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H., 2013. Vegan: Community Ecology Package. R package version 2.0-8. Pansu, M., Gautheyrou, J., 2003. Handbook of Soil Analysis: Mineralogical, Organic and Inorganic Methods. Springer, Berlin Heidelberg. Peres-Neto, P.R., Legendre, P., 2010. Estimating and controlling for spatial structure in the study of ecological communities. Global Ecology and Biogeography 19, 174e184. Peres-Neto, P.R., Legendre, P., Dray, S., Borcard, D., 2006. Variation partitioning of species data matrices: estimation and comparison of fractions. Ecology 87, 2614e2625. Pianka, E.R., 1973. The structure of lizard communities. Annual Review of Ecology and Systematics 4, 53e74. Schneider, K., Maraun, M., 2005. Feeding preferences among dark pigmented fungal taxa (“Dematiacea”) indicate limited trophic niche differentiation of oribatid mites (Oribatida, Acari). Pedobiologia 49, 61e67. Schneider, K., Migge, S., Norton, R.A., Scheu, S., Langel, R., Reineking, A., Maraun, M., 2004. Trophic niche differentiation in soil microarthropods (Oribatida, Acari): evidence from stable isotope ratios (15N/14N. Soil Biology & Biochemistry 36, 1769e1774.

M. Gao et al. / Soil Biology & Biochemistry 79 (2014) 68e77 Siepel, H., Ruiter-Dijkman, E.M.d, 1993. Feeding guilds of oribatid mites based on their carbohydrase activities. Soil Biology and Biochemistry 25, 1491e1497. Smith, T.W., Lundholm, J.T., 2010. Variation partitioning as a tool to distinguish between niche and neutral processes. Ecography 33, 648e655.
n, J.M., 2010. Niche and area of distribution modeling: a population ecology Sobero perspective. Ecology 33, 159e167. Soininen, J., 2012. Macroecology of unicellular organisms e patterns and processes. Environmental Microbiology Reports 4, 10e22. Soininen, J., McDonald, R., Hillebrand, H., 2007. The distance decay of similarity in ecological communities. Ecography 30, 3e12. Sokol, E.R., Herbold, C.W., Lee, C.K., Cary, S.C., Barrett, J.E., 2013. Local and regional influences over soil microbial metacommunities in the Transantarctic Mountains. Ecosphere 4, 1e24. Swenson, N.G., Enquist, B.J., Thompson, J., Zimmerman, J.K., 2007. The influence of spatial and size scale on phylogenetic relatedness in tropical forest communities. Ecology 88, 1770e1780. Tang, T., Wu, N.C., Li, F.Q., Fu, X.C., Cai, Q.H., 2013. Disentangling the roles of spatial and environmental variables in shaping benthic algal assemblages in rivers of central and northern China. Aquatic Ecology 47, 453e466. €rger, K., Wichmann, M.C., Schwager, M., Tews, J., Brose, U., Grimm, V., Tielbo Jeltsch, F., 2004. Animal species diversity driven by habitat heterogeneity/diversity: the importance of keystone structures. Journal of Biogeography 31, 79e92. Tuomisto, H., Ruokolainen, K., 2008. Analyzing or explaining beta diversity? Reply. Ecology 89, 3244e3256. Ulrich, W., 2008. Pairs e a FORTRAN Program for Studying Pair-wise Species Associations in Ecological Matrices, Version 1.0. http://www.uni.torun.pl/-ulrichw.

77

Viechtbauer, W., 2010. Conducting meta-analyses in R with the metafor package. Journal of Statistical Software 36, 1e48. Walter, D.E., Proctor, H.C., 2001. Mites in Soil. CD-ROM. CSIRO Publishing, Collingswood, Australia. Wang, S.X., Wang, X.A., Guo, H., Fan, W.Y., Lv, H.Y., Duan, R.Y., 2013. Distinguishing the importance between habitat specialization and dispersal limitation on species turnover. Ecology and Evolution 3, 3545e3553. Wardle, D.A., 2002. Communities and Ecosystems: Linking the Aboveground and Belowground Components. Princeton University Press, New Jersey. Wardle, D.A., 2006. The influence of biotic interactions on soil biodiversity. Ecology Letters 9, 870e886. Webb, C.T., Hoeting, J.A., Ames, G.M., Pyne, M.I., Poff, N.L., 2010. A structured and dynamic framework to advance traits-based theory and prediction in ecology. Ecology Letters 13, 267e283. Weiher, E., Keddy, P., 2001. Ecological Assembly Rules: Perspectives, Advances, Retreats. Cambridge University Press, Cambridge. Weiher, E., Keddy, P.A., 1995. The assembly of experimental wetland plant communities. Oikos 73, 323e335. Winegardner, A.K., Jones, B.K., Ng, I.S.Y., Siqueira, T., Cottenie, K., 2012. The terminology of metacommunity ecology. Trends in Ecology & Evolution 27, 253e254. Yin, W.Y., Hu, S.H., Shen, Y.F., Ning, Y.Z., Sun, X.D., Wu, J.H., Zhuge, Y., Zhang, Y.M., Wang, M., Chen, J.Y., Xu, C.G., Liang, Y.L., Wang, H.Z., Yang, T., Chen, D.N., Zhang, G.Q., Song, D.X., Chen, J., Liang, L.R., Hu, C.Y., Wang, H.F., Zhang, C.Z., Kuang, B.R., Chen, G.X., Zhao, L.J., Xie, R.D., Zhang, J., Liu, X.W., Han, M.Z., Bi, D.Y., Xiao, N.N., Yang, D.R., 1998. Pictorical Keys to Soil Animals of China. Science Press, Beijing, China.