Figure 2. Methods flowchart. Occurrence data forCryptotis mexicanus and other small non-volant mammal localities
for the target-group (TG) background are obtained (yellow). Monthly
temperature and precipitation are used to create bioclimatic variables
(ENVS) at several temporal resolutions (blue). Darker colours correspond
to the standard 30-year average approach (STA), while lighter colours
correspond to time-matched approaches (T01, T05, T10). Bioclimatic
values are extracted to create a sample with data (SWD) file to run a
MaxEnt model to select the best model for each approach (green).
Occurrence data are partitioned using repeated cross-validation to
obtain 1000 model iterations for each approach (purple). Model
comparisons of each temporal resolution are performed against the
standard 30-year average models (dashed lines).