Abstract: Aim: Despite their often high-trophic position and their contribution to many ecosystem functions, little is known about the factors affecting assemblage structure of medium- and large-sized neotropical vertebrates. We examined the relative roles played by the physical and biological environment, and by purely spatial processes, in shaping the composition and diversities of these vertebrate assemblages. Then, based on the theory that the Guianan forest cover shrank to isolated pockets during the late Pleistocene–Holocene, we tested if the past forest refugia may have shaped current vertebrate assemblages. Location: French Guiana, Guiana Shield, South America. Methods: Abundances of 19 medium- and large-sized vertebrates were estimated at 21 locations in undisturbed Guianan rain forests. Using taxonomic, functional and phylogenetic metrics, we partitioned the effects of a range of physical and biological environmental conditions and purely spatial predictors in shaping both assemblage composition and (alpha and beta) diversities. Results: We identified a significant, but weak relationship between taxonomic, functional and phylogenetic assemblage composition and environmental conditions. Assemblage diversity patterns were mainly explained by spatial predictors irrespective of the metrics. Current assemblage diversities are correlated with Pleistocene–Holocene forest history, with the highest alpha diversities outside of putative forest refugia, and the highest beta diversities inside these areas. Main conclusions: Current vertebrate assemblage composition is not strongly marked by common environmental factors. Our main conclusion is that assemblage composition results from individual species responses to the environment. Our findings also suggest that dispersal-related processes or more probably historical processes shape (alpha and beta) diversity patterns. In fact, forest fragmentation during Pleistocene–Holocene climate changes could have led to isolated vertebrate assemblages evolving into unique species assemblages creating the current high beta diversity inside refugia, whereas the lower habitat stability outside of refugia could have led to mixed assemblages in areas recolonized by forest vertebrates (current high alpha diversity outside of refugia).

Abstract: Both niche and neutral theories have been suggested as potential frameworks for modelling biodiversity. Niche models assume that biological traits represent evolutionary adaptations and define individuals in terms of functional trade-offs. Neutral models assume that all individuals at a single trophic level are functionally equivalent on a per capita basis with respect to their birth, death, dispersal and speciation. The opinion of many researchers is that neutral and niche processes operate simultaneously to generate diversity without knowing how the unification of both models can be achieved. Recently, several theoretical papers have reported evidence on the evolutionary emergence of niche structures shaping the emergence of groups of similar species. In this way, an Emergent Group is defined as a set of species that have a similar functional niche owing to a convergent ecological strategy. Central to the Emergent Group concept are the assumptions of functional equivalence within and of functional divergence between Emergent Groups. Within an Emergent Group, species richness is subject to a zero-sum rule set by the balance between the rate of individual loss and of immigration. Between Emergent Groups, tradeoffs such as seed size/seedling competitivity, investment in reproductive system/investment in vegetative systems or competitive ability/predator invulnerability are cornerstones of the evolutionary divergence. Delineating Emergent Groups amounts to reaching a compromise between maximizing niche differentiation (i.e. maximizing differences in functional tradeoffs) between Emergent Groups and maximizing neutrality within Emergent Groups. Up to now, the Emergent Group concept has been mostly proposed by theoretical scientists but it should be tested by empirical ecologists. The way in which niche and neutral models could be combined provides a profitable opportunity for theoretical and empirical scientists to collaborate fruitfully. (c) 2007 Rubel Foundation, ETH Zurich. Published by Elsevier GmbH. All rights reserved.

Abstract: entropart is a package for R designed to estimate diversity based on HCDT entropy or similarity-based entropy. It allows calculating species-neutral, phylogenetic and functional entropy and diversity, partitioning them and correcting them for estimation bias. © 2015, American Statistical Association. All rights reserved.

Abstract: An identification key is provided for 21 larval types of Elmidae (riffle beetles) known to occur in French Guiana. Not all elmid genera known to occur in French Guiana are known in the larval stage. Nor are all the known larval types assigned to known elmid genera. © 2020, Universidade de Sao Paulo. All rights reserved.

Abstract: In the context of an ongoing monitoring study on the impacts of gold-mining activities on critical ecosystem processes, we explored the use of soil Denitrifying Enzyme Activity (DEA) and Substrate Induced Respiration (SIR) as ecosystem indicators in tropical rainforests of French Guiana. We also propose DEA/SIR ratio as ecosystem attribute able to describe the state of an ecosystem and to reflect changes in ecological processes. With this purpose, we measured SIR, DEA and DEA/SIR ratio in five gold-mining areas and five surrounding natural reference rainforests. We also measured indicators in two conditions of spontaneous regeneration of vegetation (stratified or not) and two conditions of soil rehabilitation (prior preparation of soils or not). We showed a high variability of DEA, SIR and DEA/SIR ratio in the natural reference forests. This pointed out the necessity to identify relevant reference systems – i.e. proving a close match in all relevant ecological dimensions – to compare with closed perturbed systems in order to assess the levels of alterations after disturbances. Results showed a high impact of gold mine on microbial processes with a strong decrease of DEA (10-fold lower), SIR (2-fold lower) and DEA/SIR ratio (8-fold lower) in perturbed areas in comparison with natural reference forests. The type of spontaneous vegetation (stratified or not) influenced the values of indicators as well as prior rehabilitation of soils, demonstrating the capacity of DEA, SIR and DEA/SIR ratio to respond in proportion to the perturbation (robustness) and to the different levels of restoration (sensitivity). The systematic decrease of the ratio DEA/SIR observed in the studied perturbed situations demonstrates clearly that the structure of microbial communities has been also modified. The ratio DEA/SIR proved to be robust and sensitive, and able to describe in fairly fine way changes of soil microbial communities in terms of structure and function in gold mine areas and during processes of restoration. We propose to use DEA, SIR and DEA/SIR ratio as bioindicators of both structural and functional aspects of C and N cycling in soils. Together with others bioindicators based on key supporting functions in soils, these indicators should accurately evaluate the ecological potential of natural ecosystems and the levels of degradation in case of land-use changes. © 2012 Elsevier Ltd.