Application of taxonomic distinctness indices of fish assemblages for assessing effects of river-lake disconnection and eutrophication in floodplain lakes

The taxonomic distinctness (TD) indices have been recommended as robust measures for detecting human disturbances, due to their appealing properties such as independence of sampling settings and statistical framework for discriminating disturbed habitats by randomization test. Here we test the application of two widely-used TD indices, namely average taxonomic distinctness (Delta(+)) and variation in taxonomic distinctness (Lambda(+)), for assessing the effects of anthropogenic impacts (loss of river-lake connectivity and deterioration of water quality) based on the presence/absence data of fish assemblages from the floodplain lakes in the Yangtze River Basin, China. The river-disconnected lakes showed significantly lower Delta(+) values than connected lakes, and the Delta(+) values slightly decreased with increasing water eutrophication. Whereas no significant difference was detected for Lambda(+) values of fish assemblages among lake groups. The Delta(+) funnel plot can relatively well discriminate anthropogenically disturbed lakes, identifying about half of the disconnected lakes as disturbed, and the percentage of lakes identified as disturbed increased with increasing deterioration of water quality. Furthermore, we found species richness was highly related to lake area but not to nutrient levels. On the contrary, the Delta(+) was independent of lake area and responded significantly to water quality gradients, implying that the index is comparable among lakes with different surface area. In summary, the Delta(+) index of fish assemblages showed advantages over species richness for assessing anthropogenic effects in the Yangtze floodplain lakes. Therefore, the Delta(+) of fish assemblages could be considered as a potential metric for lake bioassessment and management in the studied region in particular and for other river floodplains in general.