Records from virtual databases were downloaded using the function occ() from the R package spocc. All occurrences were thoroughly assessed for their completeness and reliability. Occurrence records located exactly over centroids of municipalities and political polygons were removed from the dataset, in addition to duplicates, incomplete coordinates, and those from museums, using the suite of clean_coordinates functions from R package CoordinateCleaner. Occurrences with spatial autocorrelation structures widely divergent from the rest of the dataset coupled with coordinates outside the known extent of occurrence of the species, taken from the International Union for the Conservation of Nature, were also removed. Maintaining the contemporaneity with the climatic dataset, we also removed records dated before the year 1970.