Results of the model sensitivity analysis, where one term was dropped from the model at a time.

For each raster cell we calculated the absolute percent difference of the new partial model relative to the full model and determined the mean and standard deviation (SD) of these differences across all cells for each pair. We classed each raster into 5-quantile bins and calculated classification accuracy, which represents the percentage of raster cells in the partial model that were classed in the same bin as in the full model.