Epidemiology of suburban deer

Urban and suburban development has expanded beyond the traditional core of major cities. Preservation of open spaces in metropolitan landscapes provides aesthetic and recreational values for humans. However, white-tailed deer (Odocoileus virginianus) populations in North America are overabundant in many suburban areas. This trend is directly related to changes to habitat as a result of landscape alteration and fragmentation caused by humans. This close relationship between humans and wildlife increases the risk for zoonotic disease transmission from many vectors, such as mosquitoes and ticks. There is limited research for suburban deer examining the relationship between zoonotic prevalence and ecological factors, such as habitat. Additionally, an extensive search of deer ecology and wildlife disease literature revealed no peer-reviewed publications of zoonoses from suburban white-tailed deer. From 1995-99, I conducted a zoonotic serological survey of live-captured (tagged and radio-collared) and culled deer from Chicago, Illinois forest preserves for Jamestown Canyon (JC) and LaCrosse (LAC) encephalitis viruses, toxoplasmosis, and leptospirosis. The prevalence of sera samples collected from suburban white-tailed deer (live-captured and culled) was 55.9% for Toxoplasma gondii (n=443), 16.2% for Leptospira spp. (n=444), 38.1% for JC virus (n=404), and 6.7% for LAC virus (n=404). Prevalence was higher for adult deer for all pathogens sampled except Leptospira spp. Prevalence also was less likely in males for JC virus. A higher prevalence for T. gondii and JC virus was detected at Des Plaines (DP) and prevalence for Leptospira spp. was less likely at DP. Differences in year the sample was collected were present for T. gondii, Leptospira spp., and JC virus. Variation in month the samples were collected was evident for T. gondii, Leptospira spp., and LAC virus. Main effects variables (age, sex, site, year, and month) for multivariate logistic regression (LR) models were screened using backward stepwise elimination (p<0.20). The final LR model for T. gondii contained all main effects variables. Leptospira spp. and JC viral models both contained age, site, and year, while the final LAC viral LR model contained age, sex, and month. A subsample of deer (n=169 from DP and Palos; radio-collared and culled) analyzed for JC and LAC viral antibodies was used to evaluate habitat characteristics. Radio-collared does selected for forest, savanna, and grassland habitats (available habitat=p<0.0001; home range=p