Uncertainty Assessment of Species Distribution Prediction Using Multiple Global Climate Models on the Tibetan Plateau: A Case Study of Gentiana yunnanensis and Gentiana siphonantha

Abstract

Species distribution models (SDMs) have been widely used to project how species respond to future climate changes as forecasted by global climate models (GCMs). While uncertainties in GCMs specific to the Tibetan Plateau have been acknowledged, their impacts on species distribution modeling needs to be explored. Here, we employed ten algorithms to evaluate the uncertainties of SDMs across four GCMs (ACCESS-CM2, CMCC-ESM2, MPI-ESM1-2-HR, and UKESM1-0-LL) under two shared socioeconomic pathways (SSP2-4.5 and SSP5-8.5) at two time stages. We selected two endemic species of the Tibetan Plateau, Gentiana yunnanensis and G. siphonantha, distributed in the Hengduan Mountain regions of the southeast plateau and northeast plateau regions, respectively, as case studies. Under the two SSPs and two time periods, there are significant differences in the distribution areas of G. yunnanensis predicted by different GCMs, with some showing increases and others showing decreases. In contrast, the distribution range trends for G. siphonantha predicted by different GCMs are consistent, initially increasing and then decreasing. The CMCC-ESM2 model predicted the largest increase in the distribution range of G. yunnanensis, while the UKESM1-0-LL model predicted the greatest decrease in the distribution range of G. siphonantha. Our findings highlight that the four selected GCMs still lead to some variations in the final outcome despite the existence of similar trends. We recommend employing the average values from the four selected GCMs to simulate species potential distribution under future climate change scenarios to mitigate uncertainties among GCMs.