Spatial pattern of woody plants and their environmental interpretation in the karst forest of southwest China

We investigated the spatial patterns of woody species and their relationships with environmental factors in the karst forest of southwest China. Data for 30 woody plant species with an importance value >= 5 and 15 environmental factors in 50 quadrats in a forest plot of 200mx100m were used in this study. Two-way indicator species analysis revealed that the forest communities could be divided into 11 vegetation groups and classified into four ecotypes at the third divisional level. Detrended canonical correspondence analysis demonstrated that the variables significantly related to plant patterns were slope position, slope aspect, percentage of outcropping rocks, organic matter, total phosphorus, available nitrogen, and available phosphorus. Variation partitioning revealed that soil properties accounted for 21.0%, topographical features for 13.2%, and unmeasured variables and spatial processes for 47.7% of the variation in the species pattern, which suggest that both the deterministic factors (i.e., related to niche differentiation) and stochastic processes (i.e., related to dispersal limitation) are substantial determinants of the distribution of woody species. Our analysis suggests that picking the right late-succession species for a particular subregion or increasing soil nutrient content in poor habitats would promote species replacement in karst regions.