• Citer

Drivers of tree diversity in tropical rain forests: New insights from a comparison between littoral and hilly landscapes of Central Africa

par Dauby, Gilles ;Hardy, Olivier J. ;Leal, Miguel;Breteler, F.J.;Stevart, Tariq
Référence Journal of Biogeography, 41, 3, page (574-586)
Publication Publié, 2014-03

Article révisé par les pairs

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Aim: Our aim was to document African rain forest tree diversity patterns to test whether they support one of the following hypotheses: the structure of tree species diversity is at equilibrium and mostly driven by neutral processes (H0); the structure of tree species diversity results from Pleistocene forest cover dynamics (H1); or the structure of tree species diversity results from ecological sorting along environmental gradients (H2). Location: West Central Africa: the hilly region of Gabon, a putative Pleistocene forest refuge, and the adjacent flat littoral region, putatively colonized by forests from the hilly region after the last glaciation. Methods: We used 103 standardized 0.1-ha tree plots distributed among five landscapes in the two regions. We partitioned the tree diversity within and among plots at a local and regional scale using the Hurlbert effective number of species framework. The explanatory power of geographical and elevational distances on species turnover was tested. At the landscape scale, we partitioned the beta diversity into nestedness and turnover components. Results: Mean alpha and gamma diversities were much higher in the hilly region, where differences in elevation explained a significant part of species turnover. Substantial species turnover correlated with geographical distance also occurred in the littoral region despite its low elevational heterogeneity. Floristic turnover between plots from different regions remained stable whatever the elevational/geographical distances. At the regional scale, species replacement remained the main contributor to beta diversity (low nestedness values). Main conclusions: The heterogeneous distribution of alpha and gamma diversity, similar levels of species turnover among plots and the low contribution of the nestedness component to beta diversity between hilly and littoral landscapes do not support hypotheses H0 and H1. The combined impacts of limited dispersal and ecological sorting along environmental gradients (H2) can explain patterns within regions. The marked floristic discontinuity between regions could reflect contrasting histories (two distinct refugia) although the role of contrasting environments cannot be discounted. © 2013 John Wiley & Sons Ltd.