Effects of Climate and Built Environments on Aedes aegypti distribution in the Western United States

Abstract

Since its establishment in California in 2013, the dengue and Zika vector Aedes aegypti (L.) has rapidly spread across the state, and as of 2025, has been found in 64% of counties with active vector surveillance. Previous studies modeling this vector’s distribution in the United States typically neglect human-mediated dispersal effects that may be driving rapid, long distance movement of vectors that lead to introductions into new locations. To address this gap, we incorporated connectivity measures between trap locations into a shape-constrained additive model (SCAM). The model was trained on Ae. aegypti occurrence data from 2014–2025 from the California Vector-borne Disease Surveillance System (CalSurv) with connectivity measured by traffic flow from sites that collected Ae. aegypti in the prior year. We found that this connectivity measure, alongside fractional impervious surface and annual average minimum temperatures, had the strongest partial effect on Ae. aegypti occurrence throughout California. Model predictions for the current risk of Ae. aegypti invasion throughout California also indicated that there is significant risk within unsampled regions of southeast and northern California, while Oregon and Washington state were predicted to be homogeneously low-risk. We document the significant increase in the range of Ae. aegypti throughout California over the past 12 years, and highlight the relevance of these results in guiding proactive and targeted mosquito trap placement and protecting the public from future Aedes-borne disease transmission in the western United States.