| 作者： |
Guo, Chuanbo; Chen, Yushun; Gozlan, Rodolphe E.; Li, Zhongjie; Mehner, Thomas; Lek, Sovan; Paukert, Craig P. |
| 论文摘要： |
Understanding drivers of freshwater fish assemblages is critically important for biodiversity conservation strategies, especially in rapidly developing countries, which often have environmental protections lagging behind economic development. The influences of natural and human factors in structuring fish assemblages and their relative contributions are likely to change given the increasing magnitude of human activities. To discriminate natural and human drivers of fish diversity and assemblage patterns in developing countries with rapid socio-economic development, a dataset of 908 freshwater fish species and 13 metrics including three categories of both natural (i.e., biogeographic) and human drivers (i.e., economic growth, inland fisheries) in China were analysed with machine learning algorithms (i.e., self-organizing map, random forest). Here, we found that biogeographic drivers explained 21.8% of the observed fish assemblage patterns in China and remained stronger predictors when compared to human drivers (i.e., 15.6%, respectively). Freshwater fish species richness was positively correlated to rainfall, air temperature, surface water area and inland fisheries production but negatively correlated with urbanization. In addition, the strong structuring effects of climatic variables on Chinese fish richness patterns suggested that the fish assemblages could be particularly vulnerable to climate change. Our results showed that natural biogeographic factors still dominate in driving freshwater fish assemblage patterns despite increased human disturbances on aquatic ecosystems in a rapidly developing country. These findings consequently suggested that we should consider both natural (e.g., climate) and human (e.g., urbanization, inland fisheries) factors when establishing aquatic conservation strategies and priorities for developing countries that are experiencing rapid socio-economic changes. |
