SEARCH

SEARCH BY CITATION

The relationship between climate/productivity and historical/regional contingency and their relative influence on geographical patterns of species richness (GPSR) are still unresolved. Based on field data from 1494 plots from forests on 63 mountains across China, we document the GPSR for forest communities. Regression tree and generalized linear models were used to explore the discreteness and gradient of the distribution of tree species richness (α-diversity), and to estimate the correlations of climate, historical floristic region, and local habitat with species richness. The collinearity between climatic variables and region were further disentangled; and the spatial autocorrelation in the patterns of α-diversity and the residuals of alternative predictive models were compared. Overall, 75% of variation in plot-based α-diversity of trees was accounted for by all variables included, and about 66.5%, 64.5% and 27.9% by climate, region, and local habitat respectively. Importantly, the explanatory power of these variables differed in particular for coniferous, deciduous broadleaved and evergreen broadleaved species. Ambient temperature was more important for α-diversity of trees than were the other climatic variables across China. Spatial autocorrelation in the pattern of α-diversity could be accounted for mainly by spatial variation climate. The concordance between tree α-diversity, historical flora, contemporary climate, and Quaternary climate change mode suggests the climate/productivity and historical/regional contingency both contribute to the GPSR in a complimentary manner. Taken together, our results provide unique evidence to link of the effects of contemporary climate and historical climate change on species richness across scales.