EARLY HEARING DETECTION AND INTERVENTION VIRTUAL CONFERENCE
MARCH 2-5, 2021
(Virtually the same conference, without elevators, airplane tickets, or hotel room keys)
5/25/2021 | 2:00 PM - 3:30 PM | CHOICE OF CLIMATE DATA AFFECTS THE PERFORMANCE AND INTERPRETATON OF SPECIES DISTRIBUTION MODELS | Virtual Platform
CHOICE OF CLIMATE DATA AFFECTS THE PERFORMANCE AND INTERPRETATON OF SPECIES DISTRIBUTION MODELS
The proliferation of readily-accessible climate data is stimulating increasing use of species distribution models (SDMs) in conservation planning and management. However, it is uncertain how sensitive models are to the choice of climate data used. We compared the performance of maximum-entropy SDMs developed for six imperiled California species (4 amphibians, 2 reptiles) based on seven different datasets (WorldClim, Chelsa, TerraClimate, Climate Research Unit Time-Series [CRU-TS], PRISM, StreamCat, and EarthEnv). For each species, we used a standardized, objective procedure to select climate predictors from each dataset; assessed performance with standard metrics; and compared variable importance scores, partial dependence plots, and predicted distributions. SDM performance was sensitive to the climate dataset used. For some species, use of similar climate predictors from different datasets produced opposite responses. Models based on the spatially coarsest data (CRU-TS) performed less well than models based on the other data sources. For some species, freshwater-focused datasets produced better models than terrestrial-focused datasets. Some models overpredicted and others underpredicted distributions compared with historical range maps. Care should be taken when developing, applying, and interpreting climate-based SDMs, and ensembles of models may be needed to estimate uncertainties in predictions.
- Models
- Big data
- Aquatic–terrestrial biodiversity
Presentation:
This presentation has not yet been uploaded.
Handouts:
Handout is not Available
Transcripts:
CART transcripts are NOT YET available, but will be posted shortly after the conference
Presenters/Authors
Umarfarooq Abdulwahab
(), Utah State University, farooqadavudi@gmail.com;
ASHA DISCLOSURE:
Financial -
Nonfinancial -
Edward Hammill
(), Utah State University, edd.hammill@usu.edu;
ASHA DISCLOSURE:
Financial -
Nonfinancial -
Charles Hawkins
(), Utah State University, chuck.hawkins@usu.edu;
ASHA DISCLOSURE:
Financial -
Nonfinancial -