Projected future bioclimate-envelope suitability for amphibian species in South Central USA

This dataset contains the result of the bioclimatic-envelope modeling of the three amphibian species -- the Sacramento Mountain Salamander (Aneides hardii), the Jemez Mountains Salamander (Plethodon neomexicanus), and the Chiricahua Leopard Frog (Lithobates chiricahuensis) -- in the South Central US using the downscaled data provided by WorldClim. We used five species distribution models (SDM) including Generalized Linear Model, Random Forest, Boosted Regression Tree, Maxent, and Multivariate Adaptive Regression Splines (MARS) and ensembles to develop the present day distributions of the species based on climate-driven models alone. We then projected future distributions of the species using data from four climate models: Community Climate System Model version 4 (CCSM4), Hadley Centre Global Environment Model version 2-Earth System (HadGEM2-ES), Model for Interdisciplinary Research on Climate version 5 (MIROC5), and Max Planck Institute Earth System Model, low resolution (MPI-ESM-LR). We ran the climate models according to two greenhouse gas concentration pathways (RCP2.6 and RCP8.5). Datasets in this file are the results for models RCP2.6 and RCP8.5 for the years 2050 and 2070. It shows a comparison of ensembles of suitable bioclimatic conditions between present day and future day. The dataset shows areas where ensembles agree and suitable conditions are stable (stable represented in green), future ensemble projects new suitable conditions (gain represented in yellow), present ensemble may be converted to unsuitable in the future (loss represented in red), and areas where conditions are unsuitable in the future (non represented in gray).

本数据集包含美国中南部地区3种两栖动物的生物气候包络建模结果，分别为萨克拉门托山钝口螈（Sacramento Mountain Salamander, Aneides hardii）、赫梅斯山蝾螈（Jemez Mountains Salamander, Plethodon neomexicanus）与奇里卡瓦豹蛙（Chiricahua Leopard Frog, Lithobates chiricahuensis），建模所用数据源自WorldClim提供的降尺度数据。本研究采用5种物种分布模型（Species Distribution Model, SDM），包括广义线性模型、随机森林、提升回归树、Maxent及多元自适应回归样条（MARS），并通过模型集成方法，仅基于气候驱动模型构建了上述物种的当前适宜分布区。随后，研究团队借助4种气候模型对物种的未来适宜分布区进行投影，这4种气候模型分别为：社区气候系统模型第4版（CCSM4）、哈德利中心全球环境模型第2版-地球系统模式（HadGEM2-ES）、跨学科气候研究模型第5版（MIROC5）以及马克斯·普朗克研究所地球系统模式低分辨率版本（MPI-ESM-LR）。上述气候模型的运行依据两种温室气体浓度情景（RCP2.6与RCP8.5）。本文件收录的数据集为两种情景下2050年与2070年的模型运算结果。该数据集展示了当前与未来生物气候适宜区的集成模型对比结果，具体涵盖：集成模型一致判定且适宜条件保持稳定的区域（以绿色标记稳定区域）、未来集成模型新增的适宜区域（以黄色标记增益区域）、当前适宜区域在未来转为不适宜的区域（以红色标记流失区域），以及未来始终不适宜的区域（以灰色标记非适宜区域）。