Mapping human risk of infection with Borrelia burgdorferi sensu lato, the agent of Lyme borreliosis, in a periurban forest in France.

Lyme borreliosis is a major zoonosis in Europe, with estimates of over 26,000 cases per year in France alone. The etiological agents are spirochete bacteria that belong to the Borrelia burgdorferi sensu lato (s. l.) complex and are transmitted by hard ticks among a large range of vertebrate hosts. In Europe, the tick Ixodes ricinus is the main vector. In the absence of a vaccine and given the current difficulties to diagnose and treat chronic Lyme syndromes, there is urgent need for prevention. In this context, accurate information on the spatial patterns of risk of exposure to ticks is of prime importance for public health. The objective of our study was to provide a snapshot map of the risk of human infection with B. burgdorferi s. l. pathogens in a periurban forest at a high resolution, and to analyze the factors that contribute to variation in this risk. Field monitoring took place over three weeks in May 2011 in the suburban Sénart forest (3,200ha; southeast of Paris), which receives over 3 million people annually. We sampled ticks over the entire forest area (from 220 forest stands with a total area of 35,200m(2)) and quantified the density of questing nymphs (DON), the prevalence of infection among nymphs (NIP), and the density of infected nymphs (DIN), which is the most important predictor of the human risk of Lyme borreliosis. For each of these response variables, we explored the relative roles of weather (saturation deficit), hosts (abundance indices of ungulates and Tamias sibiricus, an introduced rodent species), vegetation and forest cover, superficial soil composition, and the distance to forest roads. In total, 19,546 questing nymphs were collected and the presence of B. burgdorferi s. l. was tested in 3,903 nymphs by qPCR. The mean DON was 5.6 nymphs per 10m(2) (standard deviation=10.4) with an average NIP of 10.1% (standard deviation=0.11). The highest DIN was 8.9 infected nymphs per 10m(2), with a mean of 0.59 (standard deviation=0.6). Our mapping and modeling revealed a strong heterogeneity of risk within the forest. The highest risk was found in the eastern part of the forest and localized patches in the northwestern part. Lyme borreliosis risk was positively associated with stands of deciduous trees (mainly oaks) and roe deer abundance. Contrary to expectations, DIN actually increased with distance from the point of introduction of T. sibiricus (i.e., DIN was higher in areas with potentially lower abundances of T. sibiricus). Thus, despite the fact that T. sibiricus is an important reservoir host for B. burgdorferi s. l., our study found that other explanatory factors played a more important role in determining the density of infected ticks. Precise mapping of the risk of exposure to Lyme borreliosis in a highly visited forest represents an important tool for targeting prevention and control measures, as well as making the general public and local health officials aware of the risks.