Résumé
Aim: Examining tree species-environment association can offer insight into the drivers of vegetation patterns and key information of practical relevance to forest management. Here, we aim to quantify the contribution of climate and soil gradients to variation in Central African tree species composition (abundance and occurrence). Location: Tropical rainforests of southern and eastern Cameroon. Methods: We established 82 1-ha permanent plots across seven localities and censused all trees ≥ 10 cm in diameter, representing a total of 37,733 trees and 455 species. In 60 of those plots, we measured ten soil variables describing texture and nutrients levels and extracted ten bioclimatic variables from global-gridded climate databases. We synthesized the main environmental gradients by conducting principal component analyses on climate and soil data respectively. We performed unconstrained and constrained non-symmetric correspondence analyses to account for the individual and joint contributions of climate and soil on species abundance and occurrence. Results: Climate and soil contributed similarly to variances of species abundance and occurrence (12–15 % variance for climate vs. 11–12 % variance for soil). Climate influence mostly concerns some abundant species, while some of the less abundant species were mainly driven by soil. Fractions of species variances accounted for by climate and soil show strong correlation when assessed from species occurrence and abundance data. Conclusion: Variation in occurrence and abundance of tropical forest trees can be partly shaped by both climate and soil gradients in Cameroon, which emphasizes the importance to jointly consider soil and climate in species distribution modeling. Less abundant species may express environmental influence differently than abundant species and convey complementary information about community assemblage. Though showing congruent patterns here, species abundance and occurrence reflect different interacting community processes and both should be examined to better understand vegetation patterns.