COS 178-10 - Ecosystem and microbial drivers of Lyme disease risk in southern Vermont

Friday, August 11, 2017: 11:10 AM
D137, Oregon Convention Center
William J. Landesman, Biology, Green Mountain College, Poultney, VT, Brian F. Allan, Department of Entomology, University of Illinois at Urbana-Champaign, Urbana-Champaign, IL and Kenneth Mulder, Mathematics, Green Mountain College, Poultney, VT
Background/Question/Methods

For the past three years (2013 - 2015), Vermont has reported the highest per capita incidence of Lyme disease in the U.S. In the northeastern U.S. Lyme disease is caused by Borrelia burgdorferi, and transmitted by the bite of an infected Ixodes scapularis. The objective of this study was to characterize spatial patterns in tick density and infection rate in relation to forest fragmentation, which has been found previously to be predictive of Lyme disease risk. Additionally, we used quantitative real-time PCR to quantify variation in the B. burgdorferi pathogen load of ticks, another important risk factor in transmission. We tested 406 nymphs from eight different forests (four patch sizes of <2 ha and four of 2 - 5 ha) in southern Vermont for infection with B. burgdorferi. On a separate set of tick samples we characterized bacterial community composition of infected and uninfected I. scapularis using amplicon sequencing of the 16S gene (V4 region). We used non-metric multidimensional scaling (NMDS) of the unweighted Unifrac metric to test the hypothesis that I. scapularis infection (presence/absence) is affected by the tick microbiome. Bioinformatics analysis was performed using Qiime 1.9.

Results/Conclusions

Average density of infected nymphs/m2 ranged from 0.01 - 0.10 among sites. Bootstrap ANOVA revealed a trend that large patches had an average nymph density that was 0.015m2 higher than for small patches (P = 0.059). The nymphal infection rate among sites ranged from 22% to 50% and average loading from 50 B. burgdorferi per nymph to > 900,000. Bootstrap ANOVA was suggestive of potential differences in mean loading between sites (P = 0.064). NMDS revealed no patterns between B. burgdorferi infection status and bacterial community composition. Thus, we found significant differences in nymph density, infection prevalence and pathogen load among sites. In contrast to previous studies, patch size had a small but positive effect on tick density and no effect on infection status. These differences may be attributable to a relatively high level of connectivity for the forest patches in this study to larger nearby forested sites. We show that bacterial community composition was not associated with the infection status (presence/absence) of adult I. scapularis. However, the currently available microbiome testing was performed on adult ticks from a separate population. Future testing will be performed on the nymphal stage ticks from Vermont, for which B. burgdorferi loading has been determined.