Heat map of variable importance across models.

Variable importance, defined as the proportion of each predictor’s contribution to the model, is shown relative to the predictor with the highest contribution (p/p_max). The color-shading within each row reflects the underlying numbers from high (red) to low (blue) and is included to make the differences in the values more apparent. For the coastwide model,”NA” indicates that Fetch was not used as a predictor.