TEMPERATURE AND PRECIPITATION AS PREDICTORS OF SPECIES RICHNESS IN NORTHERN ANDEAN AMPHIBIANS FROM COLOMBIA

Our objective was to explore the spatial distribution patterns of amphibian species

richness in Antioquia, as model for the tropical Andes, and determine how annual

mean temperature, annual precipitation, and elevation range influence it. We also

briefly compare local and global regression models for estimating the relation

between environmental variables and species richness. Distribution maps for 223

amphibian species and environmental variables were generalized onto grid maps

of 752 blocks each covering the entire Department of Antioquia. We explored the

relationship between species richness and environment using two global regression

models (the Ordinary Least Squares “OLS” and Generalized Linear Squares

“GLS” models) and one local model (the Geographically Weighted Regression

“GWR” model). We found a significant relationship between species richness and

environmental variables (GLS r2: 0.869; GRW r2: 0.929). The GLS model efficiently

incorporated the spatial autocorrelation effect and handled spatial dependence

in the regression error terms while the GWR model showed the best fit (r2) and

balance between number of parameters and fit (AICc). GWR parameters show wide

variation within the study area, indicating that relationship between species richness

and climate is spatially complex. Temperature was the most important variable

in the GLS and GWR models, and altitude range the least significant. The strong

relationship between environment and amphibian richness is possibly due to life

history traits of amphibians, such as ectothermy and water dependency to complete

the life cycle