Figure 1.
Major Atlantic Forest eco-regions, modified from the World Wildlife Fund designations (A): AMF = Araucaria Moist Forests; APF = Alto Paraná (semideciduous/deciduous) Forests; BCF = Bahia Coastal (moist) Forests; BIF = Bahia Interior (semidecidual/decidual) Forests; PCF = Pernambuco Coastal (moist) Forests; PIF = Pernambuco Interior (semideciduous) Forests; and SMCF = Serra do Mar Coastal (moist) Forests.
Biogeographic regions based on the anuran fauna generated through k-means clustering with v-fold cross-validation (B).
Table 1.
Generalized Linear Models of amphibians' k-means group in the Atlantic Forest.
Table 2.
Regression coefficients of determination of the full multiple-variable Generalized Linear Model of amphibians' k-means group in the Atlantic Forest (eco-regional variables are omitted due to the lack of statistical significance with any cluster).
Figure 2.
Principal component analysis (PCA) based on abiotic (annual precipitation, precipitation seasonality, mean annual temperature, annual actual evapotranspiration, and standard deviation of elevaton) and biotic (species richness and range size) variables of all 469 grid cells in the Atlantic Forest.
Different symbols represent biogeographic regions based on the anuran fauna generated through k-means clustering with v-fold cross-validation.
Figure 3.
Deviance partitioning analysis representing the deviance in cluster configurations explained by climate (annual precipitation, precipitation seasonality, mean annual temperature, and annual actual evapotranspiration), vegetation structure of the AF (eco-regions considered in Figure 1A), and topography (range in elevation).
The light-dark gradient of the figure represents the low-high deviance explained by the predictors.