UC Davis

Aedes aegypti, the primary vector for dengue virus (DENV), as well as other viruses such as chikungunya, Zika, and yellow fever, is well-adapted to urban environments and has become established across tropical, subtropical, and some temperate regions, including parts of the continental United States. With no effective treatments available for dengue, prevention relies primarily on vector control. Early warning systems that would allow vector control interventions to be targeted and timely would mitigate the disease burden, reduce the duration of outbreaks, and lower associated costs.Entomological surveillance has the potential to provide timely and spatially accurate risk estimates, yet metrics for Ae. aegypti abundance and DENV prevalence in Ae. aegypti have thus far not been consistent and effective predictors of dengue risk. The high spatial and temporal variability of dengue transmission complicates the use of entomological data for risk prediction. Additionally, the spatial and temporal dimensions of vector control coverage necessary for effective interventions remain poorly understood. This dissertation aims to define spatial and temporal scales and effort required for both surveillance and control of Ae. aegypti informed by realistic scenarios, including a major DENV-2 outbreak in Iquitos, routine entomological surveillance in Miami-Dade County, Florida, and a large-scale study of sequential ultra-low volume (ULV) indoor pyrethroid spray applications in Iquitos. Chapter 1 characterizes the major DENV-2 outbreak in Iquitos, including variation in abundance and DENV prevalence in Ae. aegypti and their lagged associations with dengue cases in humans. We identified Ae. aegypti DENV prevalence as the entomological metric that best predicted dengue risk, with the strongest association at a one-week lag. Chapter 2 focused on determining the surveillance effort required for early detection of dengue transmission in both endemic and non-endemic settings through simulations informed by real data from entomological surveillance during the dengue outbreak in Iquitos. Results showed that sample sizes of 320 and 1,600 Ae. aegypti females could provide an 80% DENV detection probability in endemic and non-endemic settings, respectively. The surveillance effort necessary to achieve optimal sample sizes varied markedly by season, collection method, and the epidemiological context. Chapter 3 assessed the temporal and spatial effects of ultra-low volume (ULV) indoor insecticide sprays on Ae. aegypti populations. We disentangled the impact of sprays within a household from those in neighboring households and compared the effect of the most recent spray with the cumulative effect of multiple sprays on the number of Ae. aegypti in a household. Our findings suggest that the reduction of Ae. aegypti in a household is primarily determined by time elapsed since the last spray intervention within that household, with no significant additional reduction provided by neighboring houses being sprayed or by multiple past sprays in the same household. Taken together, my research provides evidence to optimize entomological surveillance, early warning systems, and vector control strategies for more effective dengue prevention.