Examining the Ecological Drivers of Bluetongue Virus Transmission

Vector-borne pathogens cause many diseases of medical and veterinary concern, and are largely expanding with changing environments. Bluetongue virus (BTV), a Culicoides-borne pathogen of wild and domestic ruminants, has been spreading in recent decades. BTV has a complex transmission ecology, infecting several species of both hosts and vectors and circulating in highly variable environments. In this dissertation, I focus on factors influencing BTV transmission, specifically host communities, development habitats, and temperature variability. First, I study the effect of host abundance on vector abundance, trapping midges across livestock operations of varying sizes, assuming that if vector abundance increases with hosts, there is support for density-dependent transmission. I use hierarchical Bayesian models, estimating unobserved trapping efficacy, to separate this from the ecologically relevant midge abundance. Results indicate a positive effect of host abundance on vector-to-host ratio, suggesting that transmission is amplified in high-density environments. Next, I conducted a survey of Culicoides development habitats, to classify what landscapes and microhabitats support development. I collected moist substrate from variable landscapes, and observed whether midges emerged, to characterize differential emergence across land use and substrate types. Diverse site types supported midge emergence, with natural spaces producing as many midges as more classically implicated livestock operations, though the latter showed higher proportions of Culicoides sonorensis, which are thought to be the most competent vectors of BTV in this region. Lastly, I modeled the effect of seasonal temperature variability on Culicoides populations and BTV transmission, simulating across annual mean temperatures, seasonal variation, and pathogen introduction relative to seasonality. Epidemiological outcomes depended jointly on annual mean temperatures and seasonal variation, with larger outbreaks occurring when conditions were suitable for longer periods. This result yields different predictions of than traditional temperature-derived models, because our model incorporates cumulative effects of temperature, demonstrating the importance of considering seasonality when predicting transmission.<p></p>