Phylogenetic structure and ecological and evolutionary determinants of species richness

Aim: Site-level species richness is thought to result from both local conditions and species’ evolutionary history, but the nature of the evolutionary effect, and how much it underlies the correlation with current environment, are debated. Although tropical conservatism is a widely used explanatory framework along temperature gradients, it is unclear whether cold tolerance is primarily a threshold effect (e.g. freezing tolerance) or represents a more continuous constraint. Nor is it clear whether cold tolerance is the only major axis of conservatism or whether others, such as water-stress tolerance, are additionally important or trade-off against cold tolerance. We address these questions by testing associated predictions for forest plots distributed across 35° latitude.Location: China.Methods: We recorded all trees within 57 0.1-ha plots, generated a phylogeny for the 462 angiosperm species found, and calculated phylogenetic diversity (standardized PD), net relatedness index (NRI) and phylogenetic species variability (PSV) for each plot. We tested the predictions using regression, variance partitioning and structural equation modelling to disentangle potential influences of key climate variables on NRI and PSV, and of all variables on species richness.Results: Species richness correlated very strongly with minimum temperature, nonlinearly overall but linearly where freezing is absent. The phylogenetic variables also correlated strongly with minimum temperature. While NRI and PSV explained little additional variance in species richness, they accounted for part of the species richness–current climate correlation. Water stress added minimal explanatory power. All these variables showed strong latitudinal gradients.Main conclusions: Minimum temperature appeared to primarily control tree species richness, via both a threshold-like freezing effect and a linear relationship in climates without freezing. We found no clear signal of water-stress effects. The modelled contribution of evolutionary history is consistent with cold-tolerance conservatism, but could not account for all the species richness–climate relationship