AUTHOR=Holeva-Eklund Whitney M. , Young Steven J. , Will James , Busser Nicole , Townsend John , Hepp Crystal M.
TITLE=Species distribution modeling of Aedes aegypti in Maricopa County, Arizona from 2014 to 2020
JOURNAL=Frontiers in Environmental Science
VOLUME=10
YEAR=2022
URL=https://www.frontiersin.org/journals/environmental-science/articles/10.3389/fenvs.2022.1001190
DOI=10.3389/fenvs.2022.1001190
ISSN=2296-665X
ABSTRACT=Background:Aedes aegypti mosquitoes transmit dengue, yellow fever, Zika, and chikungunya viruses. Their range has recently been expanding throughout the world, including into desert regions such as Arizona in the southwestern United States. Little is understood about how these mosquitoes are surviving and behaving in arid environments, habitat that was previously considered inhospitable for the vector. The goal of this study is to create quarterly species distribution models based on satellite imagery and socioeconomic indicators for Ae. aegypti in Maricopa County, Arizona from 2014 to 2020.
Methods: Trapping records for Ae. aegypti in Maricopa County, Arizona from 2014 to 2020 were split into 25 quarterly time periods. Quarterly species distribution models (Maxent) were created using satellite imagery-derived vegetation and moisture indices, elevation, and socioeconomic factors (population density, median income) as predictors. Maps of predicted habitat suitability were converted to binary presence/absence maps, and consensus maps were created that represent “core” habitat for the mosquito over 6 years of time. Results were summarized over census-defined zip code tabulation areas with the goal of producing more actionable maps for vector control.
Results: Population density was generally the most important predictor in the models while median income and elevation were the least important. All of the 25 quarterly models had high test area under the curve values (>0.90) indicating good model performance. Multiple suburban areas surrounding the Phoenix metropolitan core area were identified as consistent highly suitable habitat.
Conclusion: We identified long term “core” habitat for adult female Ae. aegypti over the course of 6 years, as well as “hotspot” locations with greater than average suitability. Binary maps of habitat suitability may be useful for vector control and public health purposes. Future studies should examine the movement of the mosquito in this region over time which would provide another clue as to how the mosquito is surviving and behaving in a desert region.