The Effect of Forest Fragmentation on Lyme Disease Risk

Unknown to the United States three decades ago, Lyme disease is now the most prevalent vectorborne disease in the country. Caused by the bacterium Borrelia burgdorferi, Lyme disease is typically transmitted to humans by the bite of a blacklegged tick (Ixodes scapularis) in the nymph stage. For this reason the density of infected nymphs is considered the most precise ecological indicator of Lyme disease risk. The white-footed mouse (Peromyscus leucopus) is the most competent reservoir for the transmission of Lyme disease to blacklegged ticks, and has been shown to reach extraordinary densities in small woodlots (<1-2 ha) created by forest fragmentation. Thus we hypothesized that Lyme disease risk would be inversely related to the area of forest patches. Forest fragments, ranging from 0.7 ha – 7.6 ha, were chosen using GIS maps of the forested cover of Dutchess County, New York, as were three continuously forested areas. Density estimates were taken by drag sampling, and infection prevalence was assessed by dissecting ticks and determining the presence or absence of Borrelia burgdorferi spirochetes. Analysis of forest fragments demonstrated a strong, inverse relationship with patch size for blacklegged nymphal tick density (R 2 = 0.51, P = 0.02, N = 14), infection prevalence (R 2 = 0.43, P = 0.011, N = 14), and their product, the density of infected nymphs (R 2 = 0.65, P = 0.01, N = 14). Larval density, on the other hand, showed no relationship with patch area (R 2 = 0.01, P = 0.78, N = 14). Multiple linear regression of the combined influence of patch area and density of larvae, for fragments only, found the density of infected nymphs to be a negative linear function of patch area (P = 0.018), but not significantly related to the density of larvae (P = 0.099) (model R 2 = 0.52). When forest fragments were combined with three continuous sites, the effect of patch area was dampened, seemingly because high larval densities in the continuous sites appeared to drive very high nymph densities. Nevertheless, infection prevalence of those nymphs was quite low compared to the small fragments, and nymph infection prevalence was, in fact, a negative exponential function of forest area for all sites (R 2 = 0.57, P < 0.0005, N = 17). These results suggest that Lyme disease risk is inversely related to forest patch area, with small patches (<1-2 ha) having higher Lyme disease risk due to higher densities of infected nymphs. Efforts to manage the spread of Lyme disease should be directed towards the prevention of the fragmentation of forests into patches of less than 1-2 ha, given that these patches appear particularly prone to high densities of infected nymphal-blacklegged ticks.