Abstract
Abstract
Context
Climate change and anthropogenic stressors have contributed to rapid declines in biodiversity worldwide, particularly for amphibians. Amphibians play important ecological roles, yet little is known about how distribution hotspots may change or how the environmental factors influence distribution patterns in the North American Great Plains.
Objectives
Ecological niche models improve understanding of biotic and abiotic factors associated with species' distributions and can highlight potential threats to species conservation. Here, we identify important predictors of amphibian distributions and predict how land use and climate change may alter amphibian distributions in the Upper Missouri River Basin.
Methods
We used publicly available occurrence data, 16 environmental and climatic predictors, and the machine-learning algorithm, Random Forests, to create spatially explicit distribution models for eight amphibian species. Models were scored to baseline conditions (2005) and two future climate-change/land-use scenarios to predict changes in amphibian distributions for 2060.
Results
Models were highly accurate and revealed more pronounced distribution changes under the intensive RCP8.5/CONUS A2 scenario compared to the moderate RCP6.0/CONUS B2 scenario. Both scenarios predicted gains for most eastern species (i.e., Blanchard’s cricket frogs, Plains leopard frogs, Woodhouse’s toads, and Great Plains toads) and declines for all western montane species. Overall, distribution changes were most influenced by climatic and geographic predictors, (e.g., mean temperature in the warmest quarter, precipitation, and elevation), and geography, versus anthropogenic land-use variables.
Conclusions
Changes in occurrence area varied by species and geography, however, high-elevation western species were more negatively impacted. Our distribution models provide a framework for conservation efforts to aid the persistence of amphibian species across a warming, agriculturally dominated landscape.
Funder
National Science Foundation ESPCoR WAFERx
Publisher
Springer Science and Business Media LLC