SYMP 19-5
Mammal species ranges shift in response to changes in vegetation and climate

Thursday, August 14, 2014: 3:40 PM
Camellia, Sheraton Hotel
Maria João Santos, The Bill Lane Center for the American West, Stanford University, Stanford, CA

Several environmental factors have been associated with range shifts, and cumulative effects of interacting drivers are likely to occur. We assessed whether the observed changes in the elevation ranges of small mammals in the Sierra Nevada can be explained either directly by changes in climate or indirectly via vegetation changes. We used occupancy modelling and species distribution modelling (SDM: MAXENT) to estimate the changes in elevation range of small mammals from 1910-1930 to 2000-2010 using vegetation, climate, and vegetation+climate as predictor variables. We compared and contrasted the consistency of the results across modelling results. 


Only 6 out of 19 species showed a consistency between occupancy modelling range shifts and those predicted by SDMs based on either climate or vegetation.  Seven species showed consistency between occupancy and climate-based SDMs; and nine others showed consistency between occupancy and vegetation-based SDMs. Predictions from the two SDMs were more consistent than those between SDMs and occupancy models. The results were consistent for the upper and lower range limits. All the six species whose elevation ranges significantly contracted according to the occupancy modelling showed an agreement with the SDMs: three agreed with climate and vegetation predictions, two with vegetation alone and one with climate alone. However, the predictions were inconsistent for species that either expanded their elevation range or showed no change. These results suggest that predictions based on climate or vegetation may predict very different directions of range shift; nonetheless, SDMs based on either climate or vegetation seem to agree more in predictions of range contraction. They also suggest that species shifting range in the same direction may be responding to different factors.  They also indicate that multiple directions of change can be predicted depending on which predictor variable is used (e.g., climate or vegetation), and that they may act cumulatively or cancel one another out.