Climate Envelope Modeling for Evaluating Anticipated Effects of Climate Change on Threatened and Endangered Species
Modeling both potential changes in climate and responses of species and habitats can increase certainty in management decisions by helping managers to understand the range of possible species and habitat responses under different alternative futures. Climate envelope modeling is one type of modeling that can be useful in understanding species and habitat responses to climate change because it identifies key links between drivers of change (e.g., climate) and relevant responses.
PROJECT COMPLETED
The Science Issue and Relevance: Climate change will accelerate threats that challenge our ability to restore, preserve, and protect natural ecosystems and the species that depend on them. Successful conservation strategies will require an understanding of climate change and the ability to predict how it will affect species and habitats at multiple scales. Modeling both potential changes in climate and responses of species and habitats can increase certainty in management decisions by helping managers to understand the range of possible species and habitat responses under different alternative futures. Climate envelope modeling is one type of modeling that can be useful in understanding species and habitat responses to climate change because it identifies key links between drivers of change (e.g., climate) and relevant responses. Climate envelope models describe relationships between species occurrences and bioclimate variables (temperature and precipitation) to define a species climate niche (envelope). Relationships derived from contemporary data can be projected to the future using estimates of anticipated climate change.
Methodology for Addressing the Issue: Climate envelope models are a subset of the more general family of species distribution models that correlate species occurrence or abundance with climate variables to make spatially-explicit predictions of potential distribution. The general approach involves five steps: 1) acquiring species occurrence data and subsequent partitioning into ‘training’ and ‘validation’ subsets; 2) testing for statistical associations between occurrence and climate in the training data set; 3) applying associations between occurrence and climate revealed in the training dataset to predict species distributions; 4) evaluating performance of model predictions using occurrences in the validation dataset; and 5) using the associations between occurrence and contemporary climate conditions to forecast the occurrence of species under future climate projections.
Future Steps: We plan to take the next step in model refinement by adding data on land cover to species models. This additional layer of information will increase the accuracy of our models, and allow users to evaluate the relative strength of climate versus non-climate factors on species distributions. By including both climate and land cover predictors of species distributions, our models come closer to modeling the true geographic range of species rather than a more general climate envelope. Our objectives are to: 1) Improve our existing climate envelope models for T&E species to include data describing contemporary land cover associations and produce revised contemporary distribution output; 2) Create models that forecast future shifts in natural land cover under two emissions scenarios and three global circulation models; and 3) Using output from objectives 1 and 2, forecast potential species responses to direct effects of climate change (altered precipitation and temperature) as well as indirect climate change effects (land cover shifts).
Modeling both potential changes in climate and responses of species and habitats can increase certainty in management decisions by helping managers to understand the range of possible species and habitat responses under different alternative futures. Climate envelope modeling is one type of modeling that can be useful in understanding species and habitat responses to climate change because it identifies key links between drivers of change (e.g., climate) and relevant responses.
PROJECT COMPLETED
The Science Issue and Relevance: Climate change will accelerate threats that challenge our ability to restore, preserve, and protect natural ecosystems and the species that depend on them. Successful conservation strategies will require an understanding of climate change and the ability to predict how it will affect species and habitats at multiple scales. Modeling both potential changes in climate and responses of species and habitats can increase certainty in management decisions by helping managers to understand the range of possible species and habitat responses under different alternative futures. Climate envelope modeling is one type of modeling that can be useful in understanding species and habitat responses to climate change because it identifies key links between drivers of change (e.g., climate) and relevant responses. Climate envelope models describe relationships between species occurrences and bioclimate variables (temperature and precipitation) to define a species climate niche (envelope). Relationships derived from contemporary data can be projected to the future using estimates of anticipated climate change.
Methodology for Addressing the Issue: Climate envelope models are a subset of the more general family of species distribution models that correlate species occurrence or abundance with climate variables to make spatially-explicit predictions of potential distribution. The general approach involves five steps: 1) acquiring species occurrence data and subsequent partitioning into ‘training’ and ‘validation’ subsets; 2) testing for statistical associations between occurrence and climate in the training data set; 3) applying associations between occurrence and climate revealed in the training dataset to predict species distributions; 4) evaluating performance of model predictions using occurrences in the validation dataset; and 5) using the associations between occurrence and contemporary climate conditions to forecast the occurrence of species under future climate projections.
Future Steps: We plan to take the next step in model refinement by adding data on land cover to species models. This additional layer of information will increase the accuracy of our models, and allow users to evaluate the relative strength of climate versus non-climate factors on species distributions. By including both climate and land cover predictors of species distributions, our models come closer to modeling the true geographic range of species rather than a more general climate envelope. Our objectives are to: 1) Improve our existing climate envelope models for T&E species to include data describing contemporary land cover associations and produce revised contemporary distribution output; 2) Create models that forecast future shifts in natural land cover under two emissions scenarios and three global circulation models; and 3) Using output from objectives 1 and 2, forecast potential species responses to direct effects of climate change (altered precipitation and temperature) as well as indirect climate change effects (land cover shifts).