Masters Thesis

Hotspot analysis of roadkill in Southern California: a GIS approach

Areas of high probability for road kill hotspots were identified and land cover patterns that best distinguished where wildlife crossed road networks were measured to examine relationships between road kill and landscape/road variables in eight counties in Southern California from 1994-2012. The spatial association of road kill hotspots with specific land cover types was assessed using several geospatial analysis techniques. The Point Density analysis determined that there were eight hotspots for road kill in three counties for Southern California. Four hotspots were in San Diego County, two were in Los Angeles County, and two were in Ventura County. The habitat characteristics for the three counties indicated that road kill occurred in highly urbanized locations with 61% in urban land cover, while the remaining 39% of road kills were unevenly distributed between eleven other land cover types. For this study, there were a variety of species affected by road kills, from very small to very large animals including birds and reptiles. Rabbits were the most frequently killed animals followed by snakes, birds, squirrels, and coyotes. The landscape metrics were measured with FRAGSTATS (version 4.0). The FRAGSTATS results computed a multiplicity of landscape metrics for the categorical map patterns by quantifying the spatial configuration of patches near the roads. The analysis identified a difference in the recognized hotspot areas for the patches of land cover. While road kill incidents, the majority of the time, occurred in urban areas in seven out of the eight hotspots, urban land cover was the dominant type for only three of the eight hotspot locations; two in central San Diego on the coast and the other in southern Los Angeles County. Spearman's rho analysis revealed that the density of road kill is strongly correlated with the number of land cover patches and patch richness. As the number of patches (NP) and the number of patch types in each area (PR) increase, road kill events also increase. This may be due to a fragmented landscape in which wildlife have a variety of patch types they must navigate, thus increasing the need for them to move across the landscape and the probability of being hit on a road. Although road mortality may not affect large and fecund populations, it can have a significant impact on small populations and threatened or endangered species. The outcome for the hotspot analyses clearly showed urban land cover type as the highest among road kill sites. The results suggest that wildlife are crossing roads at distinct locations in the landscape. Measures can be taken to improve the chances for survival of many animals in the future such as, constructing wildlife crossings like overpasses and underpasses for future road constructions in identified hotspot areas. As the population in Southern California continues to increase so does the need to identify and protect critical habitats for wildlife.

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.