物种丰富度分布格局的成因机制一直是宏观生态学研究的热点问题之一。中国西南地区喀斯特地貌区(以广西、云南和贵州为主)是世界上面积最大的喀斯特地貌区, 也是全球范围内34个生物多样性热点地区之一。为了解该区域两栖动物物种丰富度分布格局及其与环境因子之间的关系, 本研究根据中国科学院成都生物研究所标本馆、中国科学院昆明动物研究所标本馆、广西壮族自治区自然博物馆和中南林业科技大学动物标本室收藏的标本数据, 以及公开发表的文献数据, 共获得18,246条两栖动物记录(219个物种), 然后运用生态位模型估测每个物种的潜在分布区, 并把每个物种的潜在分布区叠加起来, 最终得到该区域在10 km ´10 km生态位模型空间尺度上的两栖物种丰富度地理分布格局图, 最后进行多元回归和模型选择分析。结果表明: 有12种两栖动物仅在喀斯特地貌区分布, 占物种总数的5.48%; 有104种两栖动物仅在非喀斯特地貌区分布, 占物种总数的47.49%; 有103种两栖动物在喀斯特地貌区和非喀斯特地貌区均有分布, 占物种总数的47.03%; 两栖动物物种丰富度随纬度的增高而降低; 地貌类型(喀斯特地貌和非喀斯特地貌)对两栖动物物种丰富度的分布格局有显著影响(χ² = 36.47, P < 0.0001), 但模型拟合效果差(McFadden’s Rho square = 0.0037)。影响该区域两栖动物物种丰富度分布格局最大的环境因子是年均降雨量(R² = 0.232, P < 0.001), 其次是最干月平均降雨量(R² = 0.221, P < 0.001)。该区域两栖动物物种丰富度的格局主要是由地貌和不同的环境因子共同相互作用的结果, 不过仍有相当一部分物种丰富度的分布格局未被解释。因此, 要更全面地认识该区域两栖动物物种丰富度格局的形成机制, 有必要加强干扰、捕食、竞争等其他生物因子的影响研究。

Patterns in the distribution of species richness have always been a central theme in macroecology. The karst landforms in Southwest China (mainly Guangxi, Yunnan and Guizhou provinces) are among the largest of the global biodiversity hotspots. In this study, we sought to understand spatial patterns of amphibian species richness and its relationship with environmental factors. We compiled a large dataset of 18,246 records of point location data for 219 amphibian species occurring in China. We retrieved this data from published literature, Herpetology museums of Chengdu Institute of Biology and Kunming Institute of Zoology, Chinese Academy of Sciences, Guangxi Zhuang Autonomous Region Museum of Nature and the Central South University of Forestry and Technology, and published sources. We used this data to generate the potential distributions of each species using ecological niche modeling. We combined the potential distributions maps of all species into a composite map to describe species richness patterns on the grid cell of 10 km × 10 km, and then conducted multivariate regression and model selection. Our results showed that 12 species were distributed only in karst area, accounting for 5.48% of the total species pool, 104 species were found in non-karst area (47.49% of total species), and 103 species were found in both karst area and non-karst area (47.03% of total species). Based on the raw data of museum collections data and MaxEnt species distribution modeling, we found that amphibian species richness in the study area decreased at higher latitudes. Karst landforms and non-karst landforms differed in their distribution patterns of amphibian species richness (χ² = 36.47, P < 0.0001), but the model was a poor fit to the data (McFadden’s Rho square = 0.0037). The most significant environmental predictors of species richness were mean annual rainfall (R² = 0.232, P < 0.001) and precipitation of driest Month (R² = 0.221, P < 0.001). The results based on model selection showed that underlying mechanisms related to landforms and different ecological hypotheses might simultaneously explain patterns of amphibian species richness in the study area. Future research should examine other biological factors such as interference, predation, and competition to understand the mechanisms controlling patterns of amphibian species richness.