COMBINING PARTICIPATORY MAPPING WITH FINE-SCALE SPATIAL DATA TO UNDERSTAND LIVESTOCK-PREDATOR CONFLICT AT MULTIPLE SCALES.
Alex McInturff; UC Berkeley; 1923 9th St, APT B, Berkeley, CA, 94710; (205) 447-3741; amcinturff@gmail.com; Jennifer R. B. Miller, Kaitlyn M. Gaynor, Justin S. Brashares
Fear of livestock predation drives carnivore declines in pastoral landscapes throughout the world. Spatial risk models have made important strides in identifying patterns of conflict at broad geographic scales, but a dearth of georeferenced data has limited inference at the finer scales at which husbandry practices occur. Here we present a novel combination of participatory mapping of rancher risk perceptions with a unique 10-year data set of GPS-located coyote attacks on sheep on a California ranch to show conflict patterns at unprecedentedly fine scales. Analysis of GPS data shows that while vegetation patterns predictably shape conflict at large scales, NDVI and water availability are unexpectedly pronounced drivers of conflict within individual pastures. This finding surprised participating ranchers and has profound implications for future drought conditions. Ranchers typically summarized their risk perceptions for whole pastures, and thus did not identify the intra-pasture drivers of conflict produced by the data analysis. Nevertheless, because ranchers decide which pastures are grazed, their perceptions are always partially determinative of inter-pasture patterns of risk exposure, making their knowledge a key consideration in understanding risk. Further data collection and exchange between scientists and ranchers has broad potential to better predict livestock predation and reduce conflict.
Ecology and Conservation of Mammals II   Student Paper