No abstract

Aim Snake faunal dissimilarity within tropical forests is not well characterized, nor are the factors underlying these patterns. Our aim was to disentangle the ecological and historical factors driving biogeographical subregions (BSR) for snakes.Location Brazilian Atlantic Forest (BAF). MethodsWe compiled 274 snake inventories to build a species-by-site matrix and used unconstrained ordination and clustering techniques to identify the number of snake BSR. We applied an interpolation method to map axes of compositional variation over the whole extent of the BAF, and then classified the compositional dissimilarity according to the number of snake BSR identified a priori. We used multinomial logistic regression models and deviance partitioning techniques to investigate the influence of contemporary climatic stability, productivity, topographic complexity, and historical climate shifts in explaining the BSR. ResultsWe identified 198 snake species organized into six BSR, three of them located along the BAF coast and the other three predominantly inland BSR. Climatic stability made the largest contribution to explaining the variability in snake BSR, followed by productivity and historical variation in climate. Topography was important only if historical variation in climate was excluded from the analysis.Main conclusions The highest rates of snake endemism within BAF were in the coastal BSR, as compared to the inland BSR that are mostly composed of open habitat specialists. Our findings suggest that the topographic complexity of the BAF acts on snake distributions not as a physical barrier, but rather as a climatic barrier, providing historical climate refuges for species living along altitudinal gradients. Overall, the predominance of climatic stability and historic variation in climate in explaining snake BSR reinforces the importance of thermoregulatory constraints in shaping the distribution of tropical ectotherm species.Dev. expl (%) = percentage of deviance explained, AICc = Akaike's information criterion corrected for small sample sizes, wAICc = AICc weight. Cstab = contemporary climatic stability, Prod = productivity, Topo = topographic complexity, Cclim = contemporary climate (Cstab + Prod), Hclim = historical variation in climate. See Methods for individual predictor abbreviation.

Factors driving the spatial configuration of centres of endemism have long been a topic of broad interest and debate. Due to different eco-evolutionary processes, these highly biodiverse areas may harbour different amounts of ancient and recently diverged organisms (paleo-and neo-endemism, respectively). Patterns of endemism still need to be measured at distinct phylogenetic levels for most clades and, consequently, little is known about the distribution, the age and the causes of such patterns. Here we tested for the presence of centres with high phylogenetic endemism (PE) in the highly diverse Neotropical snakes, testing the age of these patterns (paleo-or neoendemism), and the presence of PE centres with distinct phylogenetic composition. We then tested whether PE is predicted by topography, by climate (seasonality, stability, buffering and relictualness), or biome size. We found that most areas of high PE for Neotropical snakes present a combination of both ancient and recently diverged diversity, which is distributed mostly in the Caribbean region, Central America, the Andes, the Atlantic Forest and on scattered highlands in central Brazil. Turnover of lineages is higher across Central America, resulting in more phylogenetically distinct PE centres compared to South America, which presents a more phylogenetically uniform snake Research * equally contributed. 329fauna. Finally, we found that elevational range (topographic roughness) is the main predictor of PE, especially for paleoendemism, whereas low paleo-endemism levels coincide with areas of high climatic seasonality. Our study highlights the importance of mountain systems to both ancient and recent narrowly distributed diversity. Mountains are both museums and cradles of snake diversity in the Neotropics, which has important implications for conservation in this region.

No abstract

The ongoing biodiversity crisis increases the importance and urgency of studies addressing the role of environmental variation on the composition and evolutionary history of species assemblages, but especially the tropics and ectotherms remain understudied. In regions with rainy summers, coexistence of tropical ectothermic species may be determined by the partitioning of the climatic niche, as ectotherms can rely on water availability and thermoregulatory behaviour to buffer constraints along their climatic niche. Conversely, tropical ectotherms facing dry summers would have fewer opportunities to climatic niche partitioning and other processes rather than environmental filtering would mediate species coexistence. We used 218 snake assemblages to quantify the compositional (CBD) and phylogenetic (PBD) beta-diversity of snakes in the Atlantic Forest (AF) hotspot, South America. We identify two AF regions with distinct climatological regimes: dry summers in the northern-AF and rainy summers in the southern-AF. While accounting for the influence of multiscale spatial processes, we disentangle the relative contribution of thermal, water-related and topographic conditions in structuring the CBD and PBD of snake assemblages, and determine the extent in which snake assemblages under distinct climatological regimes are affected by environmental filtering. Thermal conditions best explain CBD and PBD of snakes for the whole AF, whereas water-related factors best explain the structure of snake assemblages within a same climatological regime. CBD and PBD patterns are similarly explained by spatial factors but snake assemblages facing dry summers are more affected by spatial processes operating at fine to intermediate spatial scale, whereas those assemblages in regions with rainy summers have a stronger signature of coarse-scale processes. As expected, environmental filtering plays a stronger role in southern-AF than northern-AF, and the synergism between thermal and water-related conditions is the key cause behind this difference. Differences in climatological regimes within the tropics affect processes mediating species coexistence. The influence of broad-scale gradients (e.g. temperature and precipitation) in structuring tropical ectothermic assemblages is greater in regions with rainy summers where climatic niche partitioning is more likely. Our findings highlight the potential stronger role of biotic interactions and neutral processes in structuring ectothermic assemblages facing changes towards warmer and dryer climates.