Anisotropic community turn-over at habitat edges informs on assembly drivers: simulation and empirical test on soil macrofauna

Anisotropic community turn-over at habitat edges informs on assembly drivers: simulation and empirical test on soil macrofauna | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL Oikos This is a preprint and has not been peer reviewed. Data may be preliminary. 7 September 2025 V1 Latest version Share on Anisotropic community turn-over at habitat edges informs on assembly drivers: simulation and empirical test on soil macrofauna Authors : Gwenaelle Auger , Julien Pottier , Laurence Andanson , Pauline Bonnal , Sandrine Revaillot , Marilyn Roncoroni , and Franck Jabot 0000-0002-3113-9510 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.175724475.55678026/v1 Published Oikos Version of record Peer review timeline 272 views 156 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Biodiversity dynamics in heterogeneous landscapes is the result of a complex interplay between movement processes of organisms within and between habitat patches, and niche filtering processes due to spatially varying environmental conditions. Disentangling the relative influences of these different processes on community assembly and dynamics is a central theme of metacommunity ecology. We here propose to take advantage of the anisotropy of environmental variation in the vicinity of habitat edges to understand the drivers of spill-overs between habitats. We develop an analytical framework based on the analysis of the spatial turn-over in community composition at habitat edges. We then test this framework using metacommunity simulations. Finally, we apply this novel approach to an empirical case study on soil macrofauna at a forest-grassland interface. Our analytical framework evidences a very clear habitat-driven filtering of soil macrofauna at this forest-grassland edge, with no detectable influence of movement limitation. Various environmental variables are associated with this abrupt community turn-over, including litter amount and canopy cover, but also soil pH and Mg and Ca micronutrients. ABSTRACT Biodiversity dynamics in heterogeneous landscapes is the result of a complex interplay between movement processes of organisms within and between habitat patches, and niche filtering processes due to spatially varying environmental conditions. Disentangling the relative influences of these different processes on community assembly and dynamics is a central theme of metacommunity ecology. We here propose to take advantage of the anisotropy of environmental variation in the vicinity of habitat edges to understand the drivers of spill-overs between habitats. We develop an analytical framework based on the analysis of the spatial turn-over in community composition at habitat edges. We then test this framework using metacommunity simulations. Finally, we apply this novel approach to an empirical case study on soil macrofauna at a forest-grassland interface. Our analytical framework evidences a very clear habitat-driven filtering of soil macrofauna at this forest-grassland edge, with no detectable influence of movement limitation. Various environmental variables are associated with this abrupt community turn-over, including litter amount and canopy cover, but also soil pH and Mg and Ca micronutrients. Keywords Community assembly, metacommunity, biodiversity, heterogeneous landscape, massive DNA barcoding, soil macrofauna INTRODUCTION Habitat, and especially forest fragmentation is a human-induced process that has accelerated over the past decades in a global context of agriculture and urban expansion (Jaeger et al. , 2016). An important side effect of the size reduction of natural habitat patches is the relative increase in the proportion of habitats that become under edge influence. Hence half of the world’s remaining forest is now situated within 500 m of an edge (Haddad et al. , 2015). In this context, it has become crucial to understand the effects of these increasing edge areas on biodiversity dynamics. Research on habitat edges has primarily focused on documenting spatial variation of population abundances across edges – a feature often called “edge response function”, for a wide range of taxa including plants, invertebrates, birds, herps and mammals (Ries et al. 2004, Ries et al. 2017). These studies found very variable distribution patterns, often presenting one of these three dominant features: i) a monotonous decrease of population abundance from one habitat to the other, ii) a unimodal curve peaking at the edge or iii) the absence of variation in relation to the distance from the edge. Ries and Sisk (2004) proposed that these seemingly idiosyncratic edge response functions could be well explained by the quality, abundance and distribution of the resources of the studied organisms on both sides of the edge. In particular, if resources are more abundant on one side of the edge, their theory predicts that edge response functions should be monotonously decreasing from the high-resource habitat towards the low-resource habitat. In contrast, if the two habitats provide complementary resources, edge response functions should be unimodal and peak at the edge. This resource-based theory received various empirical support (Ries et al. , 2004; Wimp et al. , 2011). Scaling up this population-based theory to the community level is necessary to develop quantitative predictions on the influence of human-induced fragmentation on biodiversity (Ries et al. 2017). However, this is not straightforward due to interspecific variations in resource requirements (Ries et al. 2017) and movement abilities (Ewers & Didham 2006). A way forward is to quantify the respective influences of species movement abilities and of environmental filtering on edge community patterns. The study of ecological metacommunities has given rise to a substantial body of analytical approaches to tease apart the intertwined processes of organisms’ dispersal and environmental filtering (Logue et al. 2011, Jabot et al. 2020, Viana et al. 2022). Many of these approaches rely on the fact that dispersal should lead to a correlation between spatial distance and community turn-over, while environmental filtering should lead to a correlation between environmental distance and community turn-over. Metacommunity studies have typically focused on the spatio-temporal variation of ecological communities in a spatially-distributed habitat (Logue et al. 2011), with relatively few applications to habitat mosaics (but see e.g., Henckel et al. , 2019, Van Schalkwyk et al. 2020). The context of habitat edges brings a number of peculiarities that need to be considered when applying metacommunity analyses. First, while dispersal has so far been considered as the main movement process involved in metacommunity dynamics (Leibold et al., 2004; Vellend 2010), exploratory movements of organisms in their surroundings, also referred to as foraging movements, are more likely to be determinant in driving the spatial turnover of metacommunities in a habitat edge context. Indeed, foraging behaviour in the immediate vicinity of an edge, rather than dispersal, primarily drives the potential spread of individuals across the edge (Harman and Kim, 2024). Second, since environmental properties transition across habitat edges, this leads to a strong correlation between environmental and geographical distances, thereby challenging standard analytical techniques (Gilbert & Bennett, 2010). Third, environmental variation near habitat edges is likely to be anisotropic, with larger variation in the direction perpendicular to the edge than in the direction parallel to it. We here propose to take advantage of this last peculiarity of environmental anisotropy near habitat edges to assess the relative influences of environmental filtering and movement limitation on community turn-over across habitat edges. We develop in Box 1 a hypothesis-driven framework that consists in comparing the slopes of the relationship between geographical distance and community turn-over in transects that are either parallel or perpendicular to the edge. Our core hypothesis is that movement limitation should lead to similar slopes in both types of transects, while environmental filtering should lead to a steeper slope in perpendicular transects than in parallel ones (Box 1). Our aim in this study is twofold. First, we assess this analytical framework using simulated metacommunities. Second, we apply this framework to empirical data on soil macrofauna communities at a forest-grassland interface to assess the importance of environmental filtering and movement limitations on community turn-over. Soil organisms are a group of particular interest to study edge effects, as they represent a major part of the world’s global biodiversity (Anthony et al., 2023) and are recognized to respond strongly to environmental disturbances (Coleman & Hendrix, 2000). Moreover, their fundamental role in the functioning of all terrestrial ecosystems claims for a better understanding of how soil organisms may be affected by newly created edges (Chapin et al., 2011). Edge effects on soil organisms have been mainly documented for specific taxonomic groups separately, such as ants (Dauber & Wolters, 2004) or ground beetles (Magura et al. 2017), so that we lack a comprehensive appraisal of their effects on the soil macrofauna community as a whole. This study should thus contribute to fill this important gap. BOX 1 : Analytical framework to assess the relative influence of movement limitation and environmental filtering on community turn-over at habitat edges Our hypotheses are that 1) environmental conditions are less variable in directions that are parallel to the habitat edge than in perpendicular ones; 2) environmental variations are steeper near the habitat edge. These features should thus lead to increasingly steeper relationships between environmental and geographic distances, as we move the focus among different transect types: from the parallel-to-the-edge transects (least steep relationship), to the perpendicular, within-habitat transects (intermediate slope) and the perpendicular, crossing-habitat transects (steepest relationship, Fig. 1A). If community turn-over is solely driven by environmental filtering, beta-diversity should be low and constant in parallel to the edge-transects (β // , Fig. 1B and 1C). It should linearly increase with geographical distance in perpendicular to the edge-transects if environmental conditions follow a linear gradient (Fig. 1B). In contrast, if environmental conditions abruptly change at the edge (Fig. 1C), beta-diversity should be constant in perpendicular to the edge-transects, being larger for pairs of samples located on both side of the edge (β between ) than for pairs of samples located within a single habitat (β within ). If community turn-over is solely driven by movement limitation, beta-diversity should increase with geographical distance irrespective of the transect type considered (Panel D). Finally, if both movement limitation and environmental filtering are driving community turn-over, and if environmental variation is steeper near the habitat edge, beta-diversity should increase with geographical distance, but with larger slopes and intercepts for β between than β within and β // (Panel E). Hence, by analysing beta-diversity separately in the three transect types, one should be able to disentangle the relative influence of movement limitation and environmental filtering on community turn-over at habitat edges. Figure 1 : Anisotropy-based predictions . Panel A: sampling scheme to take advantage of environmental anisotropy. Each black dot represents a sampling location, along a transect either perpendicular or parallel to the habitat edge. β // and β within represent community turn-over between pairs of samples within the same habitat that respectively belong to a parallel or a perpendicular transect. β between represents community turn-over between pairs of samples located on both sides of the edge along a perpendicular transect. Panels B to E: predicted relationships between community turn-over (betadiversity) and geographical distance in the three transect types under various scenarios. Panel B: environmental-filtering driven community turn-over with a linear environmental gradient from one habitat to the other. Panel C: environmental-filtering driven community turn-over with an abrupt change in environmental conditions at the habitat edge. Panel D: movement limitation-driven community turn-over. Panel E: joint influences of the three previous hypothetical drivers of community turn-over (i.e., movement limitation and environmental filtering with a sigmoid environmental gradient). MATERIAL AND METHODS Metacommunity simulations We used the metacommunity model developed in Jabot et al. (2020) to simulate metacommunity dynamics and adapted it to a habitat edge setting. This model simulates the temporal dynamics (in discrete time) of a set of N communities that are arranged on a rectangular grid of size 2 x n x x n y . The left half of the grid corresponds to the first habitat while the right half corresponds to second habitat. This set of communities is further connected to a regional pool of S species. Each species i has the same regional abundance and a trait value t i equally spaced between 0 and 1. Each community of site j experiences successive steps of reproduction, adult mortality, movement and recruitment of new individuals. Each adult individual produces at each time step a number of juveniles drawn in a Poisson distribution with parameter r . A fraction (1 - m ) of these juveniles stay in the local community while the rest of them equally moves to the eight-neighbouring cells (Moore neighbourhood). Additionally, a fixed number I of juveniles from the regional pool reach each cell at each time step. Individual mortality is modelled as a Bernouilli process with a death probability d ij that depends on the fitness f ij of species i in site j . This fitness depends on the adaptation of species i to the local environmental variable E j and is modelled by a Gaussian-like function: f ij = (1 + A x exp [ - ( t i - E j )² / (2σ²) ]) / (1 + A ) (1) d ij = 1 - (1 - d ) x f ij (2) Finally, recruitment of new adult individuals within a local community is given by a multinomial draw with probabilities equal to juvenile abundances times their local fitness. A more detailed presentation of the model can be found in Jabot et al. (2020) and the R script to run simulations is appended in Supplementary Material S1. Four simulation scenarios corresponding to those presented in Figure 1 were performed with corresponding model parameters given in Table 1. Communities were initialized by independent multinomial draws from the regional pool with probabilities equal to species abundance times their local fitness and a subsequent simulation of 10,000 model time steps was used to reach a dynamical equilibrium. For each environmental scenario, we retained the communities at the end of the simulations, and we applied the same sampling design (perpendicular and parallel transects) as for empirical data (see below) to obtain simulated data comparable to the empirical ones. Table 1 : Model parameters used in the simulations of each scenario . Other parameters had the following fixed values across scenarios: n x = n y = 55; S = 100; m = 0.5; d =0.1. E j was always constant in each habitat column. In scenario 2, E j followed a linear gradient from 0.75 (left habitat interior column) to 0.25 (right habitat interior column), while in scenario 3, E j was constant within each habitat, equal to 0.75 in the left habitat and to 0.25 in the right habitat. In scenario 4, E j followed a logistic function with a rate parameter equal to 20. 1 - Movement limitation 1 10 0 1 2 - Environmental filtering – linear gradient 0 1000 1 0.05 3 - Environmental filtering – habitat contrast 0 1000 1 0.05 4 - Movement limitation and environmental filtering – sigmoid gradient 1 10 1 1 Study area and sampling design We conducted our empirical study in Laqueuille, Central Massif, France (core location: 45.6236 N, 2.7393 E, altitude: 1182m, Fig. 2A). The landscape was characterized by a grassland matrix mainly used for cattle grazing, and fragmented forest patches dominated by beech trees. Our study focused on a South-West facing interface between a grazed grassland and an 8-ha beech forest patch, whose area has remained constant for at least 75 years. An electric fence delimited the edge (distance ‘0m’) that was characterized by a sharp transitional area between the forest and the grassland (Fig. 2B). We chose an irregularly-spaced sampling grid (Fig. 2C) for two main reasons. First, invertebrate organisms present large interspecific variations in movement limitation, even within taxonomic groups (Auger et al. 2024), and environmental variation is multidimensional and multiscale (Dray et al. 2006). This challenges the use of a single spatial scale of sampling (Laroche 2024). Second, we aimed at intensifying the sampling effort near the edge, in order to best capture its associated ecological changes at fine spatial scales, as classically done in habitat edge studies (Boetzl et al. 2023). Transects were separated by increasing distances of 1, 2, 7 and 40m. Sampling was carried out along each transect at respective distances of 1, 3, 10, 25 and 50m from the edge in both habitats (Fig. 2C). To prevent an effect of edge interactions (Porensky & Young, 2013), we centred our sampling grid in the middle of the edge of interest that was 320 meters long, so that every sampling point from the forest patch was distant from any other edge by at least 100m. Figure 2 : Field sampling design . Panel A: Location of the study site. Panel B: Aerial view of the study site (© IGN, 2024). Panel C: Sampling design. Each dot represents a sampling point. Environmental variables and macrofauna sampling Sampling took place in June 2022, during the peak vegetation period. We measured 14 environmental variables at each of the 50 sampling points (10 points per crossing transect × 5 transects). First, vegetation cover and mean vegetation height were estimated within quadrats of 1m² around the sampling points. Litter was collected separately when extracting the macrofauna soil blocks and later weighted in the lab. Canopy cover was obtained from photographs taken from the soil surface, then binarized using the R function reclassify from the raster package (Hijmans et al. , 2024) with threshold values specific to each image, to distinguish the relative proportions of canopy versus sky areas. Finally, additional soil cores (5 cm diameter, 10 cm depth) were extracted and analysed for ten soil characteristics. Total organic carbon and nitrogen contents were assessed by a dry combustion method according to the NF ISO 10694 and NF ISO 13878 norms. They were then used to calculate the C:N ratio. Soil pH was measured in a soil water suspension according to NF ISO 10390. Available inorganic phosphorus content (phosphate P 2 O 5 ) was determined by the Olsen method (NF ISO 11263). The cation exchange capacity (CEC) and exchangeable cation concentrations (K⁺, Mg²⁺, Ca²⁺, Na⁺) were obtained using a solution of cobalt-hexamine trichloride, [Co(NH 3 ) 6 ]Cl 3 , (NF X 31-130 and NF EN ISO 23470). At each sampling point, an undisturbed soil block (18 × 18 cm, 10 cm depth) was extracted and hand-sorted to collect all macrofauna individuals ( i.e. body size \(\geq\) 3mm). Earthworms and gastropods on one side, and macro-arthropods on the other, were preserved separately in 96% ethanol. At one sampling point located at 1m to the edge on the grassland side, no adult or larval macro-individuals were recovered. The plot was kept in the data analysis as corresponding to an empty macrofauna community. Macrofauna identification All individuals were first visually identified under a stereomicroscope. Individuals that could not be unambiguously assigned to a species or a morphospecies were identified by DNA barcoding. We used the protocol described in Srivathsan et al. (2024). This consists in extracting DNA from entire specimens (or legs when the individual is too large) using a HotShot solution (Truett et al. 2000), in amplifying DNA with a set of tagged primers and in sequencing this amplified DNA using an ONT Flongle sequencer. We used both the BF3-BR2 and the LCO1490-HC02198 primer sets (Folmer et al. 1994, Elbrecht and Leese 2017). Consensus barcodes were obtained with ONTbarcoder 2.0 (Srivathsan et al. 2024) and compared to the BOLD database using BOLDigger3 (Buchner and Leese 2020). Barcodes that could not be assigned to a species in this database were further clustered into operational taxonomic units (OTUs) using an UPGMA algorithm with a 99% similarity threshold. Using this DNA barcoding identification tool, we were further able to identify larvae individuals (Jabot et al. 2025). Data analysis We first applied our proposed analytical framework (Box 1) on both simulated and empirical data. This consisted in computing beta-diversity between pairs of community samples. We used the abundance-based Bray-Curtis dissimilarity index encoded in the vegdist function of the vegan R package (Oksanen et al. 2024). We then performed linear regressions to assess the relationships between beta-diversity and the log-transformed geographical distance between pairs of community samples in the three types of transects considered: transects perpendicular to the habitat edge within and between habitats, and transects parallel to the habitat edge. We finally compared the support for the alternative scenarios depicted in Fig. 1 through Akaike Information Criterion (AIC) computation. We report in the main text the analyses performed on the empirical communities as a whole including both adult and larvae individuals. Qualitatively similar results were obtained when analysing adult and larvae individuals separately (Fig. S3 and S4). From the empirical data, we further assessed the environmental drivers of community turn-over by performing a redundancy analysis (RDA) on the environmental matrix (14 variables). Then, for each environmental variable separately and for the first RDA axis, we fitted a logistic response curve to assess the distance of edge influence of these environmental variables, following the methodology of Ewers and Didham (2006). All data analyses were performed using R, version 4.4.0 (R Core Team). R scripts to replicate metacommunity simulations and data analyses are provided in Supplementary Text S1 and S2 respectively. All data (simulated and empirical) are also available in Supplementary Files S7-9 and at . RESULTS Our metacommunity simulations confirm the operationality of our analytical framework. In the first scenario, in which community turn-over is solely driven by movement limitation, we observe that community composition is spatially auto-correlated (inserted image of Fig. 3A) but without anisotropy related to the habitat edge. As a consequence, beta-diversity increases with geographical distance similarly among the three transect types (regression lines superimposed in Fig. 1A, see text S5 for the full statistical results). In the second scenario, community turn-over is solely driven by environmental filtering; environmental conditions are constant in the direction parallel to the habitat edge and follow a linear gradient in the direction perpendicular to the edge. As a consequence, the turn-over in community composition is very low parallel to the edge (due to the stochasticity of the model) and progressive in directions perpendicular to the edge (inserted image of Fig. 3B). In this case, we find, as expected, that beta-diversity is not influenced by geographical distance in transects parallel to the habitat edge (green line in Fig. 3B) and increase with geographical distance in transects perpendicular to the edge irrespective of whether they cross or not the habitat edge (superimposed blue lines in Fig. 3B). In the third scenario, community turn-over is solely driven by environmental filtering; environmental conditions are constant within each habitat and distinct between the two habitats (inserted image of Fig. 3C). In this case, we find that beta-diversity is not influenced by geographical distance in the three types of transects and is larger in pairs of samples belonging to different habitats (dark blue line in Fig. 3C) than in within-habitat pairs (superimposed light blue and green lines in Fig. 3C). Finally, in the fourth scenario, community turn-over is driven by both movement limitation and environmental filtering; environmental conditions are constant in the direction parallel to the edge and follow a sigmoid gradient in the direction perpendicular to the edge. In this case, we find, as expected, that beta-diversity is increasing with geographical distance in the three transects types and with a slope that is larger in perpendicular-to-the-edge transects (blue lines in Fig. 3D) than in parallel ones (green line in Fig. 3D), and larger in transects crossing the habitat edge (dark blue line in Fig. 3D) than in within-habitat ones (light blue line in Fig. 3D). Figure 3 : Application of the framework to simulated metacommunities . Each panel corresponds to a distinct scenario as in Figure 1 and depicts the relationship between beta-diversity and geographical distance between pairs of simulated community samples. The inserted image depicts the community-weighted mean of the trait value t in each simulation pixel. Panel A: movement limitation and absence of environmental filtering. Panels B, C, D: environmental filtering and absence of movement limitation. Panel B: linear environmental gradient. Panel C: habitat-based environmental contrast. Panel D: abrupt (sigmoid) change of environmental conditions at the habitat edge. Green (resp. light blue, dark blue) dots correspond to pairs of samples in transects parallel to the habitat edge (resp. perpendicular within habitat, perpendicular between habitat). Regression lines adopt the same colour coding. They are horizontal (panels B, C) when the slope coefficient is not significantly different from zero and superimposed (panels A, B, C) when intercept and/or slope values are not significantly different between transect types. Macrofauna sampling gathered 2098 individuals (1254 and 844 in the forest and the grassland side, respectively) belonging to 213 species (OTUs). A large proportion (22 %) of these individuals were larvae, which represented 27% of the individuals encountered in the grassland plots, and 18% of the individuals in the forest plots. Dominant taxonomic classes included Insecta and Arachnida in both forest and grassland plots, and Chilopoda and Diplopoda in forest plots. Among insects, Coleoptera individuals were abundant in both forest and grassland plots, Hymenoptera (ants) were abundant in grassland plots while Dermaptera and Diptera (mainly larvae) dominated forest plots (Supplementary File S7). When applying our framework to these soil macrofauna communities, we demonstrate a clear habitat-driven environmental filtering at the forest-grassland edge (Fig. 4). Indeed, the influence of geographical distance on community turn-over is very low: beta-diversity is not significantly influenced by geographical distance in parallel-to-the-egde transects (green line, p = 0.32), nor in within habitat perpendicular-to-the-edge transects (light blue line, p = 0.97). In transects crossing the edge, the positive influence of geographical distance on beta-diversity is significant, although with a low effect size (dark blue line, t = 2.75, p = 0.008). Besides, the difference in beta-diversity between parallel (green line) and within habitat (light blue) transects is not significant ( p = 0.06) although a model with different intercepts has a slightly lower AIC (-241.4) than a model with a single intercept (-239.9). In contrast, the difference in beta-diversity between crossing-the-edge transects (dark blue line) and the other ones is clearly significant with a large effect size ( t = 12.3, p < 0.001). Finally, our analyses also reveal a large intra-habitat beta-diversity with a Bray-Curtis dissimilarity index averaging at 0.79. Very similar results were obtained when analysing separately adult (Fig. S3) and larval individuals (Fig. S4). Figure 4 : Relationship between beta-diversity and geographical distance at the forest-grassland edge . Green (resp. light blue, dark blue) dots correspond to pairs of samples in transects parallel to the habitat edge (resp. perpendicular within habitat, perpendicular between habitat). Regression lines adopt the same colour coding. The redundancy analysis revealed that the environmental variables explaining community turn-over were mainly structured around the first RDA axis that explained 45 % of total variation. This first RDA axis was mostly influenced by litter quantity, followed by soil pH, soil magnesium content, canopy cover and soil calcium content (Fig. 5B). Environmental variables had variable patterns of spatial variation in relation to the habitat edge, with very different distance of edge influences (DEI, Fig. 5B). Some environmental variables, like pH (Fig. 5A) or canopy cover, presented smooth changes from the forest to the grassland habitat (large DEI), while other environmental variables, like litter quantity, magnesium and calcium content, presented sharp changes at the habitat edge (small DEI, Fig. S6). The first RDA axis also presented a sharp change at the edge with a DEI of 2 m (Fig. 5C). Figure 5 : Environmental variables contributing to habitat filtering . Panel A: example of logistic fit to compute the distance of edge influence (DEI) on pH variation. Each dot represents a soil pH value at a sampling point as a function of its distance to the habitat edge. Negative (resp. positive) distance values correspond to forest (resp. grassland) samples. The green line is a logistic fit that enables to compute DEI as the distance on the x axis between inflexion points of the curve. Panel B: DEI values of each environmental variable, as well as its contribution to the first RDA axis. L: litter quantity, pH: soil pH, Mg: soil magnesium content, CC: canopy cover, Ca: soil calcium content, N: soil nitrogen content, VH: herbaceous vegetation height, CEC: soil cation exchange capacity, VC: herbaceous vegetation cover, CN: soil C/N ratio, C: soil carbon content, K: soil potassium content, P: soil phosphorus content, Na: soil sodium content. Panel C: logistic fit of the spatial variation of the first RDA axis. Taking advantage of anisotropy at habitat edges We proposed an analytical workflow to disentangle the influences of environmental filtering and movement limitation on the spatial turn-over in community composition in the vicinity of habitat edges. The main idea of this workflow is to contrast the dependence of beta-diversity on geographical distance in transects parallel and perpendicular to the habitat edge. We assessed this workflow on simulated metacommunities. This demonstrated that it was indeed possible to disentangle the influences of environmental filtering and movement limitation on community assembly by comparing the relationship between beta-diversity and geographical distance in the different transect types (Fig. 3). Many methodologies have been developed for two decades to tease apart the influences of environmental and movement filtering on community assembly. A first class of methods analyse the spatial turn-over in community composition as a function of environmental and spatial variables using a variety of statistical methods (Graco-Roza et al. 2022, Viana et al. 2022). An ongoing challenge for these methods is to reliably partition the effects of spatial and environmental predictors when they covary (Gilbert and Bennett 2010, Clappe et al. 2018, Viana et al. 2022). A second class of methods aim at fitting dynamical models of community assembly encapsulating various assembly processes (Jabot et al. 2008, Munoz et al. 2018, Laroche and Ehlers 2025). These methods are less flexibly adaptable to empirical case studies, although they present the advantage of offering a mechanistic understanding of ecological processes and enable dynamical projections (Sokol et al. 2015). Both classes of methods have developed extensions to incorporate additional information sources such as functional traits (Pouget et al. 2021, Jeliazkov and Chase 2024), phylogenetic information (Mouquet et al. 2012) or spatio-temporal dynamics (Legendre and Gauthier 2014, Jabot et al. 2020, Guzman et al. 2022). However, few developments of metacommunity methodologies have been targeted on the peculiarity of habitat mosaics (Pandit et al. 2009, Ryberg and Fitzgerald 2016). Our present contribution on this topic is thus twofold. First, we push forward the idea to structure the sampling design to take advantage of the environmental anisotropy in the vicinity of habitat edge by sampling communities in directions parallel and perpendicular to the edge. Second, we propose to contrast the relationship between beta-diversity and geographical distance in these different transect types. Our approach assumes that environmental variation is anisotropic in the vicinity of habitat edges, being larger in the direction perpendicular to the edge than in the direction parallel to the edge. This assumption is backed up by numerous edge studies that have documented environmental gradients across habitat edges (Murcia 1995). Such environmental gradients can extend over large distances (Berges et al. 2013) and therefore influence community assembly on a large proportion of our fragmented world (Haddad et al. 2015). Our approach should thus be widely applicable for the analysis of community assembly in heterogeneous landscapes. We presented this workflow with the case of a straight habitat edge to make it easier to understand. However, it could be easily applied to irregular edges. In this case, subtler geomatical techniques should be used to delineate non-straight transect lines being at a constant distance from the habitat edge. \received DD MMMM YYYY \acceptedDD MMMM YYYY A sharp habitat-driven turn-over in soil macrofauna community composition Applying this novel methodology to a forest-grassland interface, we detected a very clear signature of a habitat-driven turn-over in soil macrofauna community composition. We found beta-diversity to be larger in pairs of samples belonging to different habitats than in pairs of samples within the same habitat (Fig. 4). The influence of movement limitation on community turn-over was not apparent with no significant relationship between beta-diversity and geographical distance in transects parallel to the habitat edge (Fig. 4). Finally, we detected a slightly significant increase in beta-diversity with geographical distance in pairs of samples from different habitats (Fig. 4). This last result indicates a sharp environmental transition at the habitat edge. This result is further supported by the analysis of the spatial variation of the first axis of environmental variation obtained by redundancy analysis (Fig. 5C) that sharply varies at the edge with a very low distance of edge influence of only 2 meters. This turn-over in community composition has also been reported in previous studies on forest-grassland edges, but generally with less sharp transitions. The recent meta-analysis of Magura and Lovei (2024) centred on the influence of forest edge on ground beetles reported that distance of edge influence on beetle abundance in the forest side was below 10 m for herbivorous and omnivorous species and below 20 m for predatory ones. Besides, edge effects at forest-grassland interfaces have often been reported to be asymmetric with larger distances of edge influence in the grassland side than in the forest (Roume et al. 2011, Boetzl et al. 2016). Roume et al. (2011) reported a DEI of 11.7 m on ground beetle composition and Boetzl et al. (2016) reported that spill-over of forest specialists beetles in the grassland habitat extended up to 20 m. Similar orders of magnitude of DEI were reported for other taxa such as spiders and centipedes (Lacasella et al. 2015) or slightly lower (3-6 m) for staphylinids and diplopods (Hänggi and Baur 1998). Many of these previous studies have used pitfall traps to document variation in community composition, whereas we used monolith extraction in this study (as standard soil macrofauna monitoring studies, Potapov et al. 2022). This difference in sampling methodology may contribute to explain our finding of a sharper turn-over in community composition at the edge than previously reported. Indeed, pitfall traps are likely to sample foraging individuals performing foray loops out of their main resting habitat, while this is relatively less likely to happen with our monolith extraction sampling method, especially for nocturnal species. This last component of foraging may explain why Magura and Lovei (2024) found larger DEI for predatory than herbivorous species in their meta-analysis. Hence pitfall sampling is likely to integrate the effects of both spill-over across habitats and foray loops, while monolith extraction is likely to be comparatively less influenced by foraging movements and to essentially measure spill-over (sensu Harman and King 2024). Several environmental variables have a predominant role in the habitat-driven turn-over in soil macrofauna that we evidenced (Fig. 5B). Litter amount is larger in the forest habitat. It constitutes both a food resource and a microhabitat for a variety of soil organisms (Curry 1987). Its demonstrated influence on community composition is consistent with previous findings in both grassland (Curry 1987) and forest habitats (Ott et al. 2014). Soil pH has been shown to influence soil invertebrate community composition in both grasslands (Sanderson et al. 1995, Hoeffner et al. 2021) and temperate forests (Kaneko and Kofuji 2000, Mueller et al. 2016). Canopy cover is a proxy of light intensity whose influence on soil fauna assemblages has been demonstrated in forests (Antvogel and Bonn 2001). Finally, magnesium and calcium are known drivers of plant digestibility (Mladkova et al. 2018), of their consumption by insect herbivores (Tesitel et al. 2021) and of litter decomposition (Garcia-Palacios et al. 2016). They have contrasted concentrations in dicotyledon versus monocotyledon plants (Mladkova et al. 2018, Gross et al. 2024). Furthermore, multiple studies have documented the influence of calcium on forest soil organisms such as earthworms (Reich et al. 2005), fungi, nematodes and enchytraeids (Persson et al. 1989) and forest macroinvertebrates (Ohta et al. 2014, Mueller et al. 2016), as well as in grassland arthropods (Reihart et al. 2021). Our finding that Ca and Mg are more influential variables than other, more often studied macronutrient variables (such as C, N and P) further supports the growing recognition of the widespread importance of micronutrients for plant and animal community assembly and ecosystem functioning (Kaspari 2021). A large within-habitat beta-diversity of soil macrofauna communities Another finding of our study is the large within-habitat turn-over between sampled points, with beta-diversity values (abundance-based Bray-Curtis dissimilarities) around 0.79 in both the forest and grassland habitats. Furthermore, beta-diversity values were almost always above 0.5 even between pairs of samples that were distant of less than a few meters (Fig. 4). This indicates a strong patchiness of soil macrofauna as previously stressed (Decaëns 2010). In grasslands, grazing has been shown to promote environmental heterogeneity at the paddock scale (Bloor and Pottier 2014). This large within-habitat environmental heterogeneity was recovered in our study site for all measured variables in the grassland habitat but also in the forest habitat (Fig. S6). This environmental heterogeneity is a likely explanation for the large within-habitat beta-diversity values that we measured. A second potential explanation is that movement limitation might be extreme, even at the fine spatial scale that we explored, inducing large community turn-over at sub-meter scales. If this is the case, this would explain why we did not detect any effect of movement limitation at the range of spatial scales (above one meter) that we studied. Current syntheses on invertebrate movements indicate that macrofauna movement capacities are frequently above one meter (Auger et al. 2024). However, many taxa are still poorly studied on this topic (especially centipedes and millipedes), so that we cannot completely rule out this possibility. The potential of DNA megabarcoding for invertebrate metacommunity ecology Another feature of our study is the use of massive DNA barcoding (also referred as “megabarcoding”) to identify soil macrofauna. Massive and low-cost DNA barcoding may enable to speed up research on taxa with lacks of suitable taxonomic expertise (Eisenhauer et al. 2017, Srivathsan et al. 2023). It will also enable to consider larval stages that currently challenge standard morphology-based identifications (Yeo et al. 2018) and can make up a substantial part of invertebrate communities (Jabot et al. 2025). In our case study, larvae individuals totalled 22 % of the individuals (up to 27% in the grassland side) and presented spatial community patterns that were very similar to those of adult individuals (Fig. S3 and S4). A growing number of recent studies make use of massive DNA barcoding to analyse the spatial structure of biodiversity (e.g., Hartop et al. 2024, Lewthwaite et al. 2024). This offers very promising perspectives to extend the understanding of the drivers of biodiversity dynamics to the neglected taxa majority (Srivathsan et al. 2023). \received DD MMMM YYYY \acceptedDD MMMM YYYY REFERENCES Anthony, M.A., Bender, S.F. & van der Heijden, M.G.A. (2023) Enumerating soil biodiversity. Proceedings of the National Academy of Sciences, 120, e2304663120. Antvogel, H., & Bonn, A. (2001). Environmental parameters and microspatial distribution of insects: a case study of carabids in an alluvial forest. Ecography, 24(4), 470-482. Auger, G., Pottier, J., Mathieu, J., & Jabot, F. (2024). Space use of invertebrates in terrestrial habitats: phylogenetic, functional and environmental drivers of interspecific variations. Global Ecology and Biogeography, 33(4), e13811.Berges et al. 2013 Bloor, J. M., & Pottier, J. (2014). Grazing and spatial heterogeneity: implications for grassland structure and function. Grassland biodiversity and conservation in a changing world, 135-162. Boetzl, F.A., Schneider, G. & Krauss, J. (2016) Asymmetric carabid beetle spillover between calcareous grasslands and coniferous forests. Journal of Insect Conservation, 20, 49–57. Boetzl, F. A., Sponsler, D., Albrecht, M., Batáry, P., Birkhofer, K., Knapp, M., … & Schirmel, J. (2024). Distance functions of carabids in crop fields depend on functional traits, crop type and adjacent habitat: a synthesis. Proceedings of the Royal Society B, 291, 20232383. Buchner, D., & Leese, F. (2020). BOLDigger–a Python package to identify and organise sequences with the Barcode of Life Data systems. Metabarcoding and Metagenomics, 4, e53535. Chapin III, F. S., Matson, P. A., & Vitousek, P. M. (2011). Landscape heterogeneity and ecosystem dynamics. In Principles of terrestrial ecosystem ecology (pp. 369-397). New York, NY: Springer New York. Clappe, S., Dray, S. & Peres-Neto, P.R. (2018) Beyond neutrality: disentangling the effects of species sorting and spurious correlations in community analysis. Ecology, 99, 1737–1747. Coleman, D.C. & Hendrix, P.F. eds. (2000) Invertebrates as webmasters in ecosystems, CABI Pub, Wallingford, Oxon, UK ; New York. Curry, J. P. (1987). The invertebrate fauna of grassland and its influence on productivity. II. Factors affecting the abundance and composition of the fauna. Grass and Forage Science, 42(3), 197-212. Dauber, J. & Wolters, V. (2004) Edge effects on ant community structure and species richness in an agricultural landscape. Biodiversity & Conservation, 13, 901–915. Decaëns, T. (2010). Macroecological patterns in soil communities. Global Ecology and Biogeography, 19(3), 287-302. 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, 483–493. Eisenhauer, N., Antunes, P. M., Bennett, A. E., Birkhofer, K., Bissett, A., Bowker, M. A., … & Powell, J. R. (2017). Priorities for research in soil ecology. Pedobiologia, 63, 1-7. Elbrecht, V., & Leese, F. (2017). Validation and development of COI metabarcoding primers for freshwater macroinvertebrate bioassessment. Frontiers in Environmental Science, 5, 11. Ewers, R. M., & Didham, R. K. (2006). Continuous response functions for quantifying the strength of edge effects. Journal of applied ecology, 43(3), 527-536. Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol. 1994 Oct;3(5):294-9. PMID: 7881515. García‐Palacios, P., McKie, B. G., Handa, I. T., Frainer, A., & Hättenschwiler, S. (2016). The importance of litter traits and decomposers for litter decomposition: a comparison of aquatic and terrestrial ecosystems within and across biomes. Functional Ecology, 30(5), 819-829. Gilbert, B., & Bennett, J. R. (2010). Partitioning variation in ecological communities: do the numbers add up?. Journal of applied ecology, 47(5), 1071-1082. Graco‐Roza, C., Aarnio, S., Abrego, N., Acosta, A. T., Alahuhta, J., Altman, J., … & Soininen, J. (2022). Distance decay 2.0–a global synthesis of taxonomic and functional turnover in ecological communities. Global Ecology and Biogeography, 31(7), 1399-1421. Gross, N., Maestre, F. T., Liancourt, P., Berdugo, M., Martin, R., Gozalo, B., … & Le Bagousse-Pinguet, Y. (2024). Unforeseen plant phenotypic diversity in a dry and grazed world. Nature, 632(8026), 808-814. Guzman, L. M., Thompson, P. L., Viana, D. S., Vanschoenwinkel, B., Horvath, Z., Ptacnik, R., … & Chase, J. M. (2022). Accounting for temporal change in multiple biodiversity patterns improves the inference of metacommunity processes. Ecology, 103(6), e3683. Haddad, N.M., Brudvig, L.A., Clobert, J., Davies, K.F., Gonzalez, A., Holt, R.D., Lovejoy, T.E., Sexton, J.O., Austin, M.P., Collins, C.D., Cook, W.M., Damschen, E.I., Ewers, R.M., Foster, B.L., Jenkins, C.N., King, A.J., Laurance, W.F., Levey, D.J., Margules, C.R., Melbourne, B.A., Nicholls, A.O., Orrock, J.L., Song, D.-X. & Townshend, J.R. (2015) Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science Advances, 1, e1500052. Hänggi, A. & Baur, B. (1998) The effect of forest edge on ground-living arthropods in a remnant unfertilized calcareous grassland in the Swiss Jura mountains. Mitteilungen der Schweizerischen Entomologischen Gesellschaft, 343–354. Harman, R. R., & Kim, T. N. (2024). Differentiating spillover: an examination of cross-habitat movement in ecology. Proceedings of the Royal Society B, 291, 20232707. Hartop, E., Lee, L., Srivathsan, A., Jones, M., Peña-Aguilera, P., Ovaskainen, O., … & Meier, R. (2024). Resolving biology’s dark matter: species richness, spatiotemporal distribution, and community composition of a dark taxon. BMC biology, 22(1), 215. Henckel, L., Meynard, C.N., Devictor, V., Mouquet, N. & Bretagnolle, V. (2019) On the relative importance of space and environment in farmland bird community assembly. PLOS ONE, 14, e0213360. Hijmans, R.J., Etten, J. van, Sumner, M., Cheng, J., Baston, D., Bevan, A., Bivand, R., Busetto, L., Canty, M., Fasoli, B., Forrest, D., Ghosh, A., Golicher, D., Gray, J., Greenberg, J.A., Hiemstra, P., Hingee, K., Ilich, A., Geosciences, I. for M.A., Karney, C., Mattiuzzi, M., Mosher, S., Naimi, B., Nowosad, J., Pebesma, E., Lamigueiro, O.P., Racine, E.B., Rowlingson, B., Shortridge, A., Venables, B. & Wueest, R. (2024) raster: Geographic Data Analysis and Modeling. Hoeffner, K., Santonja, M., Monard, C., Barbe, L., Le Moing, M., & Cluzeau, D. (2021). Soil properties, grassland management, and landscape diversity drive the assembly of earthworm communities in temperate grasslands. Pedosphere, 31(3), 375-383. Jabot, F., Etienne, R. S., & Chave, J. (2008). Reconciling neutral community models and environmental filtering: theory and an empirical test. Oikos, 117(9), 1308-1320. Jabot, F., Laroche, F., Massol, F., Arthaud, F., Crabot, J., Dubart, M., … & Datry, T. (2020). Assessing metacommunity processes through signatures in spatiotemporal turnover of community composition. Ecology Letters, 23(9), 1330-1339. Jabot, F., Auger, G., Bonnal, P., Pizaine, M., Roncoroni, M., Revaillot, S., & Pottier, J. (2025). Use of massive DNA barcoding to monitor biodiversity: A test on forest soil macrofauna. Forest Ecology and Management, 595, 123004. Jaeger, J. A., Soukup, T., Schwick, C., Madriñán, L. F., & Kienast, F. (2016). Landscape fragmentation in Europe. In European landscape dynamics (pp. 187-228). CRC Press. Jeliazkov, A., & Chase, J. M. (2024). When do traits tell more than species about a metacommunity? A synthesis across ecosystems and scales. The American Naturalist, 203(1), E1-E18. Kaneko, N., & Kofuji, R. I. (2000). Effects of soil pH gradient caused by stemflow acidification on soil microarthropod community structure in a Japanese red cedar plantation: an evaluation of ecological risk on decomposition. Journal of forest research, 5(3), 157-162. Kaspari, M. (2021). The invisible hand of the periodic table: how micronutrients shape ecology. Annual Review of Ecology, Evolution, and Systematics, 52(1), 199-219. Lacasella, F., Gratton, C., De Felici, S., Isaia, M., Zapparoli, M., Marta, S. & Sbordoni, V. (2015) Asymmetrical responses of forest and “beyond edge” arthropod communities across a forest–grassland ecotone. Biodiversity and Conservation, 24, 447–465. Laroche, F. (2024). Efficient sampling designs to assess biodiversity spatial autocorrelation: should we go fractal?. Peer Community Journal, 4. Laroche, F., & Ehlers, B. (2025). Detecting non-neutral modules in species co-occurrence data: principles and application to plant communities. Royal Society Open Science, 12(7), 241375. 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(1778), 20132728. Leibold, M. A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J. M., Hoopes, M. F., … & Gonzalez, A. (2004). The metacommunity concept: a framework for multi‐scale community ecology. Ecology letters, 7(7), 601-613. Lewthwaite, J. M., Baiotto, T. M., Brown, B. V., Cheung, Y. Y., Baker, A. J., Lehnen, C., … & Guzman, L. M. (2024). Drivers of arthropod biodiversity in an urban ecosystem. Scientific Reports, 14(1), 390. 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, 482–491. Magura, T., & Lövei, G. L. (2024). Edge history modulates the depth of edge influence: evidence from ground beetles with different feeding traits. Forest Ecology and Management, 561, 121874. Magura, T., Lövei, G.L. & Tóthmérész, B. (2017) Edge responses are different in edges under natural versus anthropogenic influence: a meta-analysis using ground beetles. Ecology and Evolution, 7, 1009–1017. Mladkova, P., Mladek, J., Hejduk, S., Hejcman, M., & Pakeman, R. J. (2018). Calcium plus magnesium indicates digestibility: the significance of the second major axis of plant chemical variation for ecological processes. Ecology letters, 21(6), 885-895. Mouquet, N., Devictor, V., Meynard, C. N., Munoz, F., Bersier, L. F., Chave, J., … & Thuiller, W. (2012). Ecophylogenetics: advances and perspectives. Biological reviews, 87(4), 769-785. Mueller, K. E., Eisenhauer, N., Reich, P. B., Hobbie, S. E., Chadwick, O. A., Chorover, J., … & Oleksyn, J. (2016). Light, earthworms, and soil resources as predictors of diversity of 10 soil invertebrate groups across monocultures of 14 tree species. Soil Biology and Biochemistry, 92, 184-198. Munoz, F., Grenié, M., Denelle, P., Taudière, A., Laroche, F., Tucker, C., & Violle, C. (2018). ecolottery: Simulating and assessing community assembly with environmental filtering and neutral dynamics in R. Methods in Ecology and Evolution, 9(3), 693-703. Murcia, C. (1995) Edge effects in fragmented forests: implications for conservation. Trends in Ecology & Evolution, 10, 58–62. Ohta, T., Niwa, S., Agetsuma, N., & Hiura, T. (2014). Calcium concentration in leaf litter alters the community composition of soil invertebrates in warm-temperate forests. Pedobiologia, 57(4-6), 257-262. Oksanen, J., Simpson, G.L., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., Solymos, P., Stevens, M.H.H., Szoecs, E., Wagner, H., Barbour, M., Bedward, M., Bolker, B., Borcard, D., Carvalho, G., Chirico, M., Caceres, M.D., Durand, S., Evangelista, H.B.A., FitzJohn, R., Friendly, M., Furneaux, B., Hannigan, G., Hill, M.O., Lahti, L., McGlinn, D., Ouellette, M.-H., Cunha, E.R., Smith, T., Stier, A., Braak, C.J.F.T. & Weedon, J. (2024) vegan: Community Ecology Package. Ott, D., Digel, C., Klarner, B., Maraun, M., Pollierer, M., Rall, B. C., … & Brose, U. (2014). Litter elemental stoichiometry and biomass densities of forest soil invertebrates. Oikos, 123(10), 1212-1223. Pandit, S.N., Kolasa, J. & Cottenie, K. (2009) Contrasts between habitat generalists and specialists: an empirical extension to the basic metacommunity framework. Ecology, 90, 2253–2262. Persson, T., Lundkvist, H., Wirén, A., Hyvönen, R., & Wessén, B. (1989). Effects of acidification and liming on carbon and nitrogen mineralization and soil organisms in mor humus. Water, Air, and Soil Pollution, 45(1), 77-96. Porensky, L.M. & Young, T.P. (2013) Edge-Effect Interactions in Fragmented and Patchy Landscapes. Conservation Biology, 27, 509–519. Potapov, A.M., Sun, X., Barnes, A.D., Briones, M.J., Brown, G.G., Cameron, E.K., Chang, C.-H., Cortet, J., Eisenhauer, N., Franco, A.L.C., Fujii, S., Geisen, S., Guerra, C., Gongalsky, K., Haimi, J., Handa, I.T., Janion-Sheepers, C., Karaban, K., Lindo, Z., Mathieu, J., Moreno, M.L., Murvanidze, M., Nielsen, U., Scheu, S., Schmidt, O., Schneider, C., Seeber, J., Tsiafouli, M., Tuma, J., Tiunov, A., Zaytsev, A.S., Ashwood, F., Callaham, M. & Wall, D. (2022) Global monitoring of soil animal communities using a common methodology. Soil Organisms, 94, 55–68. Pouget, C., Pottier, J., & Jabot, F. (2021). Fine‐scale functional metacommunity dynamics: Analysing the role of disturbance‐driven environmental variability in grasslands. Journal of Vegetation Science, 32(4), e13068. Reich, P. B., Oleksyn, J., Modrzynski, J., Mrozinski, P., Hobbie, S. E., Eissenstat, D. M., … & Tjoelker, M. G. (2005). Linking litter calcium, earthworms and soil properties: a common garden test with 14 tree species. Ecology letters, 8(8), 811-818. Reihart, R. W., Angelos, K. P., Gawkins, K. M., Hurst, S. E., Montelongo, D. C., Laws, A. N., … & Prather, C. M. (2021). Crazy ants craving calcium: macronutrients and micronutrients can limit and stress an invaded grassland brown food web. Ries, L. & Sisk, T.D. (2004) A Predictive Model of Edge Effects. Ecology, 85, 2917–2926. Ries, L., Jr, R.J.F., Battin, J. & Sisk, T.D. (2004) Ecological Responses to Habitat Edges: Mechanisms, Models, and Variability Explained. Annual Review of Ecology, Evolution, and Systematics, 35, 491–522. Ries, L., Murphy, S.M., Wimp, G.M. & Fletcher, R.J. (2017) Closing Persistent Gaps in Knowledge About Edge Ecology. Current Landscape Ecology Reports, 2, 30–41. Roume, A., Deconchat, M., Raison, L., Balent, G., & Ouin, A. (2011). Edge effects on ground beetles at the woodlot–field interface are short‐range and asymmetrical. Agricultural and Forest Entomology, 13(4), 395-403. Ryberg, W. A., & Fitzgerald, L. A. (2016). Landscape composition, not connectivity, determines metacommunity structure across multiple scales. Ecography, 39(10), 932-941. Sanderson, R. A., Rushton, S. P., Cherrill, A. J., & Byrne, J. P. (1995). Soil, vegetation and space: an analysis of their effects on the invertebrate communities of a moorland in north-east England. Journal of Applied Ecology, 506-518. Sokol, E. R., Brown, B. L., Carey, C. C., Tornwall, B. M., Swan, C. M., & Barrett, J. E. (2015). Linking management to biodiversity in built ponds using metacommunity simulations. Ecological Modelling, 296, 36-45. Srivathsan, A., Ang, Y., Heraty, J.M., Hwang, W.S., Jusoh, W.F.A., Kutty, S.N., Puniamoorthy, J., Yeo, D., Roslin, T. & Meier, R. (2023) Convergence of dominance and neglect in flying insect diversity. Nature Ecology & Evolution, 7, 1012–1021. Srivathsan, A., Feng, V., Suárez, D., Emerson, B., & Meier, R. (2024). ONTbarcoder 2.0: rapid species discovery and identification with real‐time barcoding facilitated by Oxford Nanopore R10. 4. Cladistics, 40(2), 192-203. Těšitel, J., Tahadlová, M., Lepš, J., & Hölzel, N. (2021). Linking insect herbivory with plant traits: phylogenetically structured trait syndromes matter. Journal of Vegetation Science, 32(4), e13061. Truett, G.E., Heeger, P., Mynatt, R.L., Truett, A.A., Walker, J.A., Warman, M.L., 2000. Preparation of PCR-quality mouse genomic DNA with hot sodium hydroxide and tris (HotSHOT). Biotechniques 29 (1), 52–54. Van Schalkwyk, J., Pryke, J.S., Samways, M.J. & Gaigher, R. (2020) Environmental filtering and spillover explain multi-species edge responses across agricultural boundaries in a biosphere reserve. Scientific Reports, 10, 14800. Vellend, M. (2010) Conceptual Synthesis in Community Ecology. The Quarterly Review of Biology, 85, 183–206. Viana, D. S., Keil, P., & Jeliazkov, A. (2022). Disentangling spatial and environmental effects: Flexible methods for community ecology and macroecology. Ecosphere, 13(4), e4028. Wimp, G.M., Murphy, S.M., Lewis, D. & Ries, L. (2011) Do edge responses cascade up or down a multi-trophic food web? Ecology Letters, 14, 863–870. Yeo, D., Puniamoorthy, J., Ngiam, R. W. J., & Meier, R. (2018). Towards holomorphology in entomology: rapid and cost‐effective adult–larva matching using NGS barcodes. Systematic entomology, 43(4), 678-691. SUPPLEMENTARY INFORMATION Supplementary Text S1: R script to simulate metacommunity dynamics at edges. Supplementary Text S2: R script to replicate our data analyses. Fig. S3: adult-only results. Fig. S4: larvae results. Supplementary Text S5: full statistical results of the analyses of metacommunity simulations. Fig. S6: Distances of edge influence (DEI) for the 14 measured environmental variables. Supplementary File S7: Rdata file of the community matrix. Supplementary File S8: csv file containing the environmental variables at the sampled points. Supplementary File S9: Rdata file containing the simulated metacommunities. Information & Authors Information Version history V1 Version 1 07 September 2025 Peer review timeline Published Oikos Version of Record 5 Jan 2026 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection Oikos Keywords biodiversity community assembly heterogeneous landscape massive dna barcoding metacommunity soil macrofauna Authors Affiliations Gwenaelle Auger INRAE View all articles by this author Julien Pottier INRAE View all articles by this author Laurence Andanson VetAgro Sup - Campus Agronomique de Clermont View all articles by this author Pauline Bonnal INRAE View all articles by this author Sandrine Revaillot INRAE View all articles by this author Marilyn Roncoroni INRAE View all articles by this author Franck Jabot 0000-0002-3113-9510 [email protected] INRAE View all articles by this author Metrics & Citations Metrics Article Usage 272 views 156 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Gwenaelle Auger, Julien Pottier, Laurence Andanson, et al. Anisotropic community turn-over at habitat edges informs on assembly drivers: simulation and empirical test on soil macrofauna. Authorea . 07 September 2025. DOI: https://doi.org/10.22541/au.175724475.55678026/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175724475.55678026/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a00e80ea1d3806f3',t:'MTc3OTY0ODgxMA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();