Biotic diversity of ecological communities can be driven by a mixture of climatic, soil and biotic factors from local to regional scales. Patterns of diversity change were often examined along latitudinal or elevational gradients, which were mainly driven by climatic factors. However, few studies have assessed biodiversity patterns along both abiotic and biotic gradients simultaneously. Here, we established 309 forest dynamics plots of typical forest vegetation types (tropical rainforest, subtropical evergreen deciduous broad-leaf mixed forest, warm temperate conifer broad-leaf mixed forest, and temperate conifer forest) in seven biogeographic regions across four climatic regions (tropical, subtropical, warm temperate, and temperate regions) in China. A total of 46,280 tree individuals of 801 species in these plots were tagged, investigated and mapped, and six functional traits and seven soil factors were sampled and measured, in addition, data on three climatic factors were extracted from Worldclim. Principal component analysis (PCA) was used to build the compound habitat gradient (CHG) combining the biotic and abiotic factors. Patterns of changes in species and functional diversity along the CHG were analyzed. Results showed that species richness, Shannon-Wiener index and functional richness (FRic) increased but functional divergence (FDiv) decreased along the first axis of the CHG. The models using the first four PCA axes of the CHG and models with the individual variables had more power to explain species diversity than functional diversity. PC1 was the most important predictor in explaining patterns of diversity variation. The null models of FRic and FDiv were significantly negatively correlated with PC1 in the compound habitat gradient, and not with other axes. Our study demonstrates that the compound habitat gradient analysis is an effective approach of exploring patterns of biodiversity change and understanding both the abiotic and biotic factors driving community assembly across different climatic regions.

生态群落的生物多样性可以由地方到区域尺度的气候，土壤和生物因素混合驱动。经常沿着纬度或海拔梯度检查多样性变化的模式，这主要是由气候因素驱动的。但是，很少有研究同时评估非生物和生物梯度上的生物多样性模式。在这里，我们在四个气候区域（热带）的七个生物地理区域中建立了309个典型森林植被类型（热带雨林，亚热带常绿落叶阔叶混交林，温带温带针叶树阔叶混交林和温带针叶林）的森林动态图。 ，亚热带，温带和温带地区）。在这些地块中，总共对801个物种的46,280个树个体进行了标记，进行了调查和制图，对六个功能性状和七个土壤因子进行了采样和测量，此外，还从Worldclim中提取了三个气候因子的数据。主成分分析（PCA）用于建立结合生物和非生物因素的复合栖息地梯度（CHG）。分析了沿CHG物种和功能多样性的变化模式。结果表明，沿CHG的第一个轴，物种丰富度，Shannon-Wiener指数和功能丰富度（FRic）增大，但功能差异（FDiv）减小。使用CHG的前四个PCA轴的模型以及具有单个变量的模型具有比功能多样性更大的解释物种多样性的能力。PC1是解释多样性变化模式的最重要预测指标。FRic和FDiv的无效模型在复合生境梯度中与PC1显着负相关，而与其他轴则不相关。我们的研究表明，复合栖息地梯度分析是探索生物多样性变化模式并了解驱动不同气候区域的社区聚集的非生物和生物因素的有效方法。