Ensemble Forecasting of Large-Scale Fish Species Distribution in Lakes

Species distribution models (SDM) have been routinely used for the purpose of species conservation and biodiversity management, especially in the context of global environment/climate changes. However there is little knowledge about the efficiency/uncertainty source of the SDM for the predictions in aquatic ecosystems, especially in the large-scale research. In this study, a total of 92 fish species were predicted with climatic and geographical variables respectively using machine learning to predict species distribution and composition from more than 100 lakes largely distributed in China. Our results highlight that predictions from single SDM were so variety and unreliable for all species while ensemble approaches could yield more accurate predictions; we emphasized that species characteristics as species prevalence would strongly affect the outcomes of SDMs; our findings finally verified the hypothesis that species distributed with a smaller range size could be more accurately predicted than species with large range size to be plausible in aquatic ecosystems. Our research would provide promising insights into the predicting of fish species in aquatic ecosystems under the impacts of global climate change, especially for the conservation of endemic fish species in China which we inferred could be better predicted.