Skip to main content

Hotspots of species loss do not vary across future climate scenarios in a drought-prone river basin

Cite this dataset

Neeson, Thomas; Gill, Ken; Fovargue, Rachel (2020). Hotspots of species loss do not vary across future climate scenarios in a drought-prone river basin [Dataset]. Dryad.


Aim: Climate change is expected to alter the distributions of species around the world, but estimates of species’ outcomes vary widely among competing climate scenarios. Where should conservation resources be directed to maximize expected conservation benefits given future climate uncertainty? Here, we explore this question by quantifying variation in fish species’ distributions across future climate scenarios.

Location: Red River basin, south-central United States.

Methods: We modeled historical and future stream fish distributions using a suite of environmental covariates derived from high-resolution hydrologic and climatic modeling of the basin. We quantified variation in outcomes for individual species across climate scenarios and across space, and identified hotspots of species loss by summing changes in probability of occurrence across species.

Results: Under all climate scenarios, we find that the distribution of most fish species in the Red River Basin will contract by 2050. However, the variability across climate scenarios was more than 10 times higher for some species than for others. Despite this uncertainty in outcomes for individual species, hotspots of species loss tended to occur in the same portions of the basin across all climate scenarios. We also find that the most common species are projected to experience the greatest range contractions, underscoring the need for directing conservation resources towards both common and rare species

Main conclusions: Our results suggest that while it may be difficult to predict which species will be most impacted by climate change, it may nevertheless be possible to identify spatial priorities for climate mitigation actions that are robust to future climate uncertainty. These findings are likely to be generalizable to other ecosystems around the world where future climate conditions follow prevailing historical patterns of key environmental covariates.


This dataset includes Maxent model outputs of the distributions of 31 fish species in the Red River of Oklahoma and Texas. Data are probability of occurrence maps for the 31 species for historical and future (year 2050) time periods for each of nine climate scenarios.

Usage notes

This text explains the data files that accompany:

Gill, K. et al. (2020) Hotspots of species loss do not vary across future climate scenarios in a drought-prone river basin. Ecology and Evolution, in press.

Data are organized in 18 sub-directories, each of which contains Maxent model outputs from a particular climate scenario and time period (historical, and for the year 2050). The nine climate scenarios represent all combinations of three general circulation models (GCMs; CCSM4, MIROC5, and MPI_ESM_LR) and three representative concentration pathways (RCPs; 2.6, 4.5, and 8.5 W/m2). For example, the sub-directory “RCP26-CCSM4-Future2050” contains Maxent model outputs for the year 2050 under the RCP 2.6 scenario and the CCSM4 general circulation model.

Each sub-directory (i.e., climate scenario and time period combination) contains 31 raster files, one per each of the 31 fish species in this study. Each raster file is a Maxent-generated probability of occurrence map for a species in a given climate scenario and time period.


South Central Climate Adaptation Science Center