This is a bit of a beginner question. I am attempting to model the species distribution of the little fire ant but I would like to use global presence data to output a map of the small island of Moorea as well as the chain of islands of French Polynesia. I have already downloaded the 30s bioclim data that appears to be a good resolution for the islands.
My question is: does clipping environmental layers restrict the samples the model is trained on? For example, if I clip the layers to Moorea in order to output the asc heat map of moorea, will the model still be trained on global data to create these predictions, or will it only be able to use the presence data from Moorea? If so, what is the best way to output the map that I desire?