Figure 7. Maps showing
the summation of thresholded suitability maps based on 1000 iterations
(4 folds x 250 repeats) of Cryptotis mexicanus for each of the
four approaches: standard 30-year average (STA), and three temporal
resolutions (T01, one-year; T05, five-year; T10, ten-year). Areas with
no suitable predictions are indicated in blue, while those where most
iterations predict suitability are shown in red. Occurrences that were
fully withheld during model training and cross-validation are indicated
with a white cross mark, while the number of times an occurrence was
predicted as suitable is represented the size of a semi-transparent
black circle, with larger circles indicating a greater number of
suitable predictions.