Wednesday, August 4, 2010

PS 64-158: Habitat and demographic factors that influence urban fox and gray squirrel populations

Andrew L. Cave, Nathan A. Weber, and Charles H. Nilon. University of Missouri

Background/Question/Methods   We examined how urban habitats are shaped, at the neighborhood scale, in areas with different demographics and socioeconomic conditions and how those habitats influence fox and gray squirrel populations. Our study, in Columbia, MO, used eight randomly selected census track block groups which we paired to represent the four types of neighborhoods found in the city. These four neighborhoods differed in percentage of black residents, percentage of those who owned their home, age of residence, education level, and mean household income. Thirty randomly selected 100m transects were established across the four neighborhoods. Each transect's habitat was described by the number and size(s) of tree species, canopy coverage, and the presence or absence of supplemental food sources. Each season, between June of 2009 and April of 2010, a pair of fifteen minute time-counts were performed on each transect. ANOVA was implemented to analyze differences in the mean number of squirrels counted between the neighborhoods and seasonal variations. Tukey's test was performed to assess pairwise differences among the four neighborhoods. We used Poisson Regression to produce models for both gray and fox squirrels to identity if canopy coverage, hard-mast production, and potential den abundance influenced their densities.

Results/Conclusions   There was no statistically significant difference in seasonal variation for fox squirrels (F = 0.060, df = 2/177, p = 0.942). We found that the mean number of fox squirrels per transect (M = 0.356, SD = 0.802) differed significantly among the neighborhoods (F = 18.710, df = 3/176, p < 0.001). Fox squirrels were more abundant in neighborhood type 1, which is primarily low-income black residents. We found that our model did not explain a significant amount of variation seen for fox squirrels, adjusted R2 = 0.032, F(5,174) = 2.199, p = 0.057. We also found that there was no significant difference in seasonal variation for gray squirrels (F = 0.462, df = 2/177, p = 0.631). The mean number of gray squirrels (M = 1.35, SD = 1.996) differed significantly among the neighborhoods (F = 22.788, df = 3/176, p < 0.001). Gray squirrels were more abundant in neighborhood type 4, which is primarily a middle-class, white neighborhood with high homeownership. We found that the number of hard mast producing trees, canopy coverage and potential den trees explained 38% of the variation seen for gray squirrels, adjusted R2 = 0.378, F(5,174) = 23.586, p < 0.001.