deletions | additions       diff --git a/figures/Fig_7/caption.tex b/figures/Fig_7/caption.tex  index f92e182..0d468d2 100644  --- a/figures/Fig_7/caption.tex  +++ b/figures/Fig_7/caption.tex 
...
Fig \textbf{Fig  8: Classification of target regions based on the effect of predictors (local local  regression coefficient of the given predictor). predictor}.  The coefficients of slope and average minimum  temperature ofthe  coldest month (bio6) were (bio6), slope, and precipitation of  the best criteria to group warmest quarter (bio18) were  the localities. most important factors in this model.  Example of interpretation: the habitat suitability values of Southern Italy (ID=35) can be explained by a coefficient of the slope lower than -0.006, that implies slope values slightly higher than the optimum, and a coefficient of the temperature of the coldest month lower than 0.004, what means that bio6 is around its optimum.