Climate pressure and protection gaps threaten endemic flora of Brazil’s semi-arid region - Supplementary file

The study evaluates the potential impacts of climate change on the distribution of 157 endemic plant species in the Caatinga biome using species distribution models (SDMs) under contrasting future climate scenarios. Results indicate a severe reduction in climatically suitable areas, particularly under the pessimistic scenario (SSP585). Projected shifts in species richness toward coastal regions suggest that environmental stability in these areas may support the persistence of some species. However, the contraction of suitable habitats could lead to the regional extinction of up to 18 species by 2100. These findings underscore the urgent need for targeted conservation strategies, including the expansion of protected areas and the integration of climate refugia, to mitigate biodiversity loss in the Caatinga.Table S1: Occurrence records of the modeled species, including latitude and longitude coordinates and the data source for each occurrence record.Table S2: Coefficients of the principal components selected from the Principal Component Analysis (PCA) performed with species distribution modeling data under climate change scenarios.Table S3: For each algorithm, the table presents the defined parameters and hyperparameters, including the total number of tested configurations. The column detailing hyperparameter values specifies both the parameter names and the R expressions employed to generate the sequences of values.Table S4: GAP analysis categories and additional conservation indicators for 159 endemic plant species from the Caatinga. For each species, the table presents the climate scenario, future period, GAP analysis category (based on overlap with Protected Areas), total suitable area (km²), preliminary classification according to IUCN Criterion B1 (Extent of Occurrence), and the percentage of change in suitable area relative to the baseline (%), where negative values indicate area loss and positive values indicate area gain.Table S5: Thresholds and performance metrics for species distribution models of 159 endemic plant species from the Caatinga. For each species and algorithm, the table presents the selected threshold value, number of presence and absence records used for evaluation, mean Sørensen similarity index (Sorensen), and standard deviation (SD*) across model replicates. Algorithms include MaxEnt, Random Forest, Support Vector Machine, and an Ensemble consensus model.Figure S1: Proposed biogeographical map of the Caatinga. The entire Caatinga region is considered a Dominion, subdivided into the Caatinga Province and the Chapada Diamantina Province. The Caatinga Province is further divided into Subprovinces and Districts. This map was created from the shapefile available in the supplementary material ofMoro et al. (2024).RasterOutputs_SDM: This folder contains raster files generated from SDMs for 157 plant species endemic to the Caatinga. The models were developed using baseline conditions and future climate change projections under two Shared Socioeconomic Pathways (SSP126 and SSP585) for two time periods: 2041–2070 and 2071–2100. All files are in GeoTIFF format, with a spatial resolution of 5 arc-minutes (approximately 10 km), and use the WGS84 coordinate reference system (EPSG:4326).This dataset provides essential information for understanding the future of Caatinga’s endemic flora under climate change and supports conservation planning for biodiversity protection in the dominion.For further details, please refer to the main article.