01 January 1970 0 3K Report

Is it reasonable to use predictions outputs (probability rasters 0-100) as variable in niche/distribution models? For example SoilGrids (https://soilgrids.org/) layer of individual soil type. I am especially interested in case when the probabilities are low in the study area, but species response is high. I will be thankful for any references, suggestions and comments

Similar questions and discussions