Urbanization and Landscape Heterogeneity Influence Culex Species Ecology and Genetics in Eastern North America

Abstract

Vector-borne disease is an important facet of public health, as they account for nearly 20% of global disease burden. Multiple species, including at least one vector, at least one host, and a pathogen, must interact in vector-borne disease transmission cycles, and thus understanding human risk of vector-borne disease and public health outcomes requires a community ecology framework. Members of the Culex genus, including Cx. pipiens, Cx. quinquefasciatus, and Cx. restuans are sympatric in eastern North America and are vectors of West Nile virus. This dissertation explores the roles of habitat use, community ecology, phenology, and landscape heterogeneity on Culex spatiotemporal dynamics and genetics along urban to rural gradients in eastern North America. Through surveillance of belowground structures in Washington, D.C. over two years, I found that mosquito species of public health importance, including Aedes aegypti, Aedes albopictus, and members of the Culex pipiens assemblage, use these structures for breeding and development. Belowground structures may serve as refugia against extreme climatic conditions and facilitate overwintering survival for non-diapausing taxa and/or taxa with thermal tolerance limitations, potentially expanding their suitable ranges. On an urban to rural gradient in greater Washington, D.C. and Maryland, a complex of cryptic Culex WNv vectors co-occur on the landscape. Using molecular techniques and constrained ordination, I found that these cryptic Culex species were differently distributed at fine spatial scales, likely due to the impacts of urbanization on vector habitat and subsequent niche segregation. Culex pipiens were cosmopolitan and dominant across sites in greater Washington D.C. and Maryland. However, individuals with Cx. quinquefasciatus ancestry were limited to urban and peri-urban sites closest to the city center, and Cx. restuans were most abundant in rural and suburban sites furthest from the city center with dense and heterogeneous canopy cover. Previous work suggested that phenology has a considerable impact on Culex species dynamics; Cx. restuans was thought to be an early season species that cedes to Cx. pipiens over the course of the season. Initially, I did not detect an effect of season on Culex spatiotemporal dynamics when collections were undertaken from June through October, but when I expanded the collection season to include the months of April and May, the influence of season was evident. Therefore, the hallmark “crossing-over” point that is common in the Culex literature happens prior to the local mosquito abatement season in Washington and D.C. and Maryland. During the active surveillance and management period, season has little impact on Culex species abundance as compared to environmental factors measured along our urban to rural gradient. A replicated comparison of the abundance and relative frequency of Cx. pipiens and Cx. restuans along urbanization gradients in Washington D.C., greater Philadelphia, PA and greater Chicago, IL, using gradient forests demonstrated that phenology was consistently the most important predictor of the shift between a Cx. restuans-dominant community and a Cx. pipiens-dominant community. This crossing-over point trended later in the season with increasing latitude. Turnover in species abundance tended to occur at intermediate points along environmental gradients associated with urbanization, like percent impervious surface, percent tree cover, distance to city center, and vegetation index. Results of two analytical approaches (ordination and regression trees) and from three metropolitan areas support Cx. restuans as an early season species that is otherwise associated with sites with cooler temperatures, less impervious surface, more tree cover, a shallower water table, and increased distance from city center. Conversely, Cx. pipiens is more abundant than Cx. restuans in sites that are more characteristic of urbanization. Culex pipiens is globally ubiquitous and was common across site classes in the three localities in this study. This species comprises two bioforms, pipiens and molestus, which are characterized by divergent ecological, physiological, and behavioral traits. These bioforms can interbreed in the field and the lab. However, at all sites analyzed across three northeastern metropolitan areas, analysis of genotypes at a single neutral locus violated assumptions of Hardy Weinberg Equilibrium, suggesting that there is not unrestricted geneflow between bioforms across the landscape. The proportion of molestus alleles increased with increasing percent impervious surface and decreased vegetation, two environmental correlates of urbanization. Molestus alleles may confer an advantage in urban environments because they can leverage human infrastructure to overcome thermal limitations and persist in isolated belowground populations via autogeny and use of mammalian hosts. Overall, Culex WNv vectors are differentially distributed across urban to rural gradients in the northeastern United States. These aspects are influenced by a heterogeneous land use and landscape-level changes associated with urbanization. A clear understanding of vector life history, genetics, interspecies interactions, and distribution across the landscape can improve practitioners’ power and precision in predicting and managing vector borne disease transmission. While some patterns in species distribution and composition were universal across metropolitan areas, there was variation between localities that could significantly contribute to WNv transmission and human disease risk. Therefore, I conclude that modeling, as well as development of surveillance and management strategies for WNv vectors should be implemented locally to have the greatest impact on public health outcomes.