PS 64-70
Predicting climate change impacts on the potential distribution of 12 endemic Skinks (Genus:Plestiodon) of Mexico

Friday, August 15, 2014
Exhibit Hall, Sacramento Convention Center
Mayra Oyervides, Biology, University of Texas-Pan American, Edinburg, TX
T. Patricia Feria, Department of Biology, The University of Texas-Pan American, Edinburg, TX
Manuel Feria, Museo de Zoología, Facultad de Estudios Superiores Zaragoza, Universidad Nacional Autónoma de México, Distrito Federal, Mexico
Background/Question/Methods

Anthropogenic global climate change is recognized as a large scale threat to biodiversity worldwide.  Species distribution models (SDMs) were created to assess the possible effects of climate change on the distribution of 12 endemic Plestiodon lizards of Mexico.  We gathered a database of presence of species information using different sources such as online databases (GBIF, HerpNET, REMIB), literature, and field data. Databases per species were exhaustively reviewed by a Mexican expert on the group, as well as using ArcGIS v 10.1 to eliminate or correct geographic locations (e.g. longitude should be negative in Mexico).  Species occurrences were correlated with 19 climatic variables (temperature, precipitation) from www.worldclim.org to model species' current suitable habitat.  These models were then projected to future (2050) conditions for which bioclimatic information was downloaded from the Climate Change, Agriculture and Food Security (CCAFS) website (http://www.ccafs-climate.org/).  Forecasts were made using 4 General Circulatory Models (CGCM2, CSIRO, HadCM3, and NIES99), 2 Intergovernmental Panel on Climate Change Scenarios (A2 and B2), and software MaxEnt (v3.3.3k).  Models were run 10 times using the cross-validation option in MaxEnt, and evaluated using the AUC on a ROC plot.  The resulting maps and AUC are the average of the 10 replicates.

Results/Conclusions

Findings showed AUC scores >0.9, demonstrating a robust level of accuracy for the models obtained from MaxEnt. With the exception of 4 species (P. colimensis, P. lynxe, P. parviauriculatus, and P. parvulus), suitable habitat for most species will contract significantly.  These findings could help establish sound conservation programs for these species.  Additional studies to understand species population trends and other threats for these species, such as habitat fragmentation are encouraged.