Environment and dispersal influence changes in species composition at different scales in woody plants of the Western Ghats, India

Abstract

AbstractQuestionWhat are the relative roles of environmental and spatial factors in influencing variation in species composition of tropical woody plants at different spatial scales?LocationTropical evergreen forests, Western Ghats, South India.MethodsUsing a plot‐based species inventory spanning the entire latitudinal extent (1,200 km) of the Western Ghats’ wet evergreen forests, we collected primary data on spatial variation in species composition of woody plants. Each plot was characterized by a set of environmental descriptors consisting of topographic, edaphic and climatic variables, while eigenvector‐based spatial variables and plot coordinates were used as spatial descriptors. We used ordination‐based as well as distance‐based variation partitioning techniques to partition the variation in species composition into components uniquely and jointly explained by environmental and spatial factors.ResultsThe compositional similarity of woody plants largely showed a linear decline with log‐geographic distance. However, this relationship was spatially structured. After controlling for the differences in environment, compositional similarity was found to be strongly associated with geographic distance only at the smallest spatial scale. Variation partitioning analysis revealed that environmental variables explained a much larger proportion of variation in species composition overall compared to spatial variables. Among environmental variables, climatic variables emerged as the most important predictors of variation in species composition at regional and landscape scales.ConclusionsStrong association between compositional similarity and geographic distance at local scales indicates the influence of dispersal limitation, while niche differentiation seems to be a more important driver of variation in species composition at larger spatial scales. Overall, our results provide evidence for scale‐dependent shifts in the relative importance of factors that are responsible for variation in species composition.