Climate change and land-use changes are key drivers of global biodiversity loss. Many species are shifting to higher elevations or latitudes in response to global warming, often encountering unfavorable land-use conditions during the shift. This leads to reduced range size and increased extinction risks, particularly for mountain species, often confined to narrow, high-altitude habitats. Predicting future distributions of mountain species requires an accounting for their bioclimatic responses, detailed topographical distribution, land-use preferences, and ability to colonise new areas via dispersal mechanisms. These elements are rarely considered together over large scales. Here, we projected the future distribution of 32 mountain mammal and 344 non-migratory mountain bird species by 2050 under different emission scenarios (SSP-RCP 1-2.6 and SSP-RCP 5-8.5). Using Species Distribution Models (SDMs) that incorporated topography, climate, and land-use data, we assessed the impacts of global change on species’ ranges across mountain regions worldwide, accounting for realistic dispersal scenarios. Under the high-emissions scenario, species were projected to experience significantly greater range loss compared to the low-emissions scenario, with a difference of 17% of loss for birds and 16% for mammals. The number of species that shift their range also increased, passing from 73 to to 84. The most severe range losses were projected for species located in tropical mountain ranges, while European and North American mountains showed lower losses, highlighting substantial regional differences in vulnerability. When land-use changes were included in the models, projected range losses increased further, particularly under the low-emissions scenario (+2%). Our findings emphasize the importance of considering both climate and land-use changes when assessing biodiversity risks in mountain regions. Our results highlight the urgency of mitigating climate change and managing land use to preserve the unique biodiversity of these areas.

Inferring assembly processes from empirical community diversity patterns has always been a major goal in Ecology. Many empirical studies rely on the "filtering framework", which characterizes community assembly as a sequence of abiotic and biotic filters. The success of the ecological filtering framework lies in its theoretical foundation, linking environmental filtering to niche theory, and competitive interactions to coexistence theory. Empirical studies have provided evidence of environmental filtering in a wide range of environments. However, while competitive interactions are omnipresent, few applications of the filtering framework found significant evidence of competition in real-life settings. Consequently, the framework has been criticised for being overly simplistic. We argue that this unbalanced picture is likely due to specific conceptual challenges. First, many traits are commonly used in empirical work without a clear distinction between traits that capture species responses' to the environment vs. traits that capture the competitive interactions between species, and without consideration of how these two sets of traits may co-vary. Second, it neglects that environmental filter and competition can produce the same traits patterns. Third, the spatial scale at which the community is observed strongly impacts the resulting patterns. Here, we explore these three conceptual challenges and test how trait patterns vary depending on different assembly processes, traits and scales vary. Using a theoretical simulation model, we demonstrate that the trait patterns resulting from environmental filtering and competition respond differently to variations in traits' correlation structure and observation scales. We then identify the actual conditions in which it is possible to distinguish signals of distinct assembly processes from patterns, given the correlation and relevance of traits and the inherent constraints of the observational scale.

Biotic interactions drive multitrophic species community assembly. Yet, explicitly incorporating this process in species distribution models (SDMs) is particularly challenging, even when biotic interactions are known. Here, we propose a framework that combines knowledge of trophic interactions with Bayesian structural equation models to model each species as a function of its prey or predators and environmental conditions. We tested and validated our framework on realistic simulated communities spanning different theoretical models and ecological setups. We showed that our framework improves the inference of both species’ potential and realized niches compared to single SDMs (mean performances increased by 8% and 6% respectively), especially for species with strong biotic control, thus increasing model predictive performance. Our framework can easily integrate various SDM extensions (e.g., occupancy models) and algorithms, and stands out as a novel solution for modeling multitrophic community distributions when trophic interactions are known or assumed.

A document by Pierre Gaüzère. Click on the document to view its contents.

Soil trophic networks are key to biogeochemical cycles, in particular decomposition. However, few studies have yet quantified how microbial decomposition activity along environmental gradients is jointly driven by bacteria, fungi, and their respective consumers. Here, we quantified these direct and indirect effects on decomposition and contrasted them between forests and open habitats using multiple elevational gradients in the French Alps. While environmental control on microbial decomposition activity was comparable in the two habitats, the pathways and strengths of biotic predictors strongly differed. The fungal channel composition played a moderate role in forests, while the bacterial channel composition was critical in open habitats. Importantly, we found trophic regulation by consumers to be a key modulator of the direct environmental effects on decomposition in open habitats. These results highlight the need to integrate trophic regulation when predicting future ecosystem functioning.

Land use intensification favours particular trophic groups which can induce architectural changes in food-webs. These changes can deeply impact ecosystem functioning, stability and robustness to extinctions. However, the imprint of land management intensity on food-web architecture has rarely been characterised across large spatial extent and various land uses. We investigated the influence of land management intensity on six facets of food-web architecture for 67,051 European tetrapod communities, and its dependency on land use and climate. We found that intensification promoted lower proportions of both apex and basal species, with more connected and less compartmentalized food-webs, and unexpectedly, favoured longer trophic chains in cities and decreased omnivory in mediterranean climates. By favouring mesopredators and undermining basal tetrapods, intensification might lead to new forestry and agricultural pest outbreaks. Our results support mesopredator regulation and apex predator protection where possible, but urban and mediterranean contexts might need alternative strategies.

Our knowledge of the factors influencing the distribution of soil organisms is limited to specific taxonomic groups. Consequently, our understanding of the drivers shaping the entire soil food web is constrained. To address this gap, we conducted an extensive soil biodiversity monitoring program in the French Alps, using environmental DNA to obtain multi-taxon data from 418 soil samples. The spatial structure of resulting soil food webs varied significantly between and within habitats. From forests to grasslands, we observed a shift in the abundance of trophic groups from fungal to bacterial feeding channels, reflecting different ecosystem functioning. Furthermore, forest food webs were more strongly spatially structured which could only partly be explained by abiotic conditions. Grassland food webs were more strongly driven by plant community composition and soil characteristics. Our findings provide valuable insights into how climate and land use changes may differentially affect soil food webs in mountains.

Protists are major actors of soil communities and play key roles in shaping food webs, community assembly, and ecosystem processes, yet their functional diversity is understudied. High-throughput sequencing data have revealed their ubiquity and diversity, but lack of standardized traits has hampered the integration of functional information, limiting our understanding of soil ecosystems. Here we propose a framework for soil protists, identify a set of common traits to characterize their functional diversity, and apply the framework on a broad-scale, real-world dataset. We reviewed studies on soil protists to identify the traits used in the literature, and define a framework based on 10 key traits that satisfy two criteria: availability of information, and applicability to most taxa. The framework was tested on a dataset of environmental DNA metabarcoding data from 1123 soil samples collected in 48 glacier forelands worldwide. Traits were assigned to all the 570 Molecular Operational Taxonomic Units (MOTUs) detected in our dataset, leading to the production of a global trait-based dataset from glacier forelands. We estimated the functional space of protist communities and evaluated if the selected traits were effective in describing protist diversity. The functional space of protist communities showed that the MOTUs are clustered in three regions, mainly reflecting different nutritional and habitat preferences. The proposed framework is appropriate for multiple applications, including estimation of functional diversity and food web analyses, and provides a basis for ecological studies on soil protists, enabling the functional characterization of this essential but often neglected component of soil biodiversity.

Although how rare species persist in communities is a major ecological question, the critical phenotypic dimension of rarity is broadly overlooked. Recent work has shown that evaluating functional distinctiveness, the average trait distance of a species to other species in a community, offers essential insights into biodiversity dynamics, ecosystem functioning, and biological conservation. However, the ecological mechanisms underlying the persistence of functionally distinct species are poorly understood. Here we propose a heterogeneous fitness landscape framework, whereby functional dimensions encompass peaks representing trait combinations that yield positive intrinsic growth rates in a community. We identify four fundamental causes leading to the persistence of functionally distinct species in a community. First, environmental heterogeneity or alternative phenotypic designs can drive positive population growth of functionally distinct species. Second, sink populations with negative growth can deviate from local fitness peaks and be functionally distinct. Third, species found at the margin of the fitness landscape can persist but be functionally distinct. Fourth, biotic interactions (either positive or negative) can dynamically alter the fitness landscape. We offer examples of these four cases and some guidelines to distinguish among them. In addition to these deterministic processes, we also explore how stochastic dispersal limitation can yield functional distinctiveness.

Re-analyzing data from our study, Bruun & Ejrnaes (2022) show that key species to productivity are more abundant than species threatened by extinction. They therefore conclude that biodiversity loss hardly hampers ecosystem processes. Acknowledging the validity of the findings, we clarify why we believe their conclusions are drawn too far.

Ice-free areas are increasing worldwide due to the dramatic glacier shrinkage and are undergoing rapid colonization by multiple lifeforms, thus representing key environments to study ecosystem development. Soils have a complex vertical structure. However, we know little about how microbial and animal communities differ across soil depths and development stages during the colonization of deglaciated terrains, how these differences evolve through time, and whether patterns are consistent among different taxonomic groups. Here, we used environmental DNA metabarcoding to describe how community diversity and composition of six groups (Eukaryota, Bacteria, Mycota, Collembola, Insecta, Oligochaeta) differ between surface (0-5 cm) and relatively deep (7.5-20 cm) soils at different stages of development across five Alpine glaciers. Taxonomic diversity increased with time since glacier retreat and with soil evolution; the pattern was consistent across different groups and soil depths. For Eukaryota, and particularly Mycota, alpha-diversity was generally the highest in soils close to the surface. Time since glacier retreat was a more important driver of community composition compared to soil depth; for nearly all the taxa, differences in community composition between surface and deep soils decreased with time since glacier retreat, suggesting that the development of soil and/or of vegetation tends to homogenize the first 20 cm of soil through time. Within both Bacteria and Mycota, several molecular operational taxonomic units were significant indicators of specific depths and/or soil development stages, confirming the strong functional variation of microbial communities through time and depth. The complexity of community patterns highlights the importance of integrating information from multiple taxonomic groups to unravel community variation in response to ongoing global changes.

While species interactions are fundamental for linking biodiversity to ecosystem functioning and for conservation, large-scale empirical data are lacking for most species and ecosystems. Accumulating evidence suggests that trophic interactions are predictable from available functional trait information, but we have yet to understand how well we can predict interactions across large spatial scales and food webs. Here, we built a model predicting predator-prey interactions based on functional traits for European vertebrates. We found that even models calibrated with very few known interactions (100 out of 71k) estimated the entire food web reasonably well. However, predators were easier to predict than prey, with prey in some clades being particularly difficult to predict (e.g., fowls and storks). Local food web connectance was also consistently over-estimated. Our results demonstrate the potential for filling gaps in sparse food webs, an important step towards a better description of biodiversity with strong implications for conservation planning.