The Aedes aegypti mosquito first invaded the Americas about 500 years ago and today is a widely distributed invasive species and the primary vector for viruses causing dengue, chikungunya, Zika, and yellow fever. Here we test the hypothesis that the North American colonization by Ae. aegypti occurred via a series of founder events. We present findings on genetic diversity, structure, and demographic history using data from 70 Ae. aegypti populations in North America genotyped at 12 microsatellite loci and/or ~20,000 single nucleotide polymorphisms (SNPs), the largest genetic study of the region to date. We find evidence consistent with a colonization driven by serial founder effect (SFE), with Florida as the putative source for a series of westward invasions. This scenario was supported by 1) a decrease in the genetic diversity of Ae. aegypti populations moving west, 2) a correlation between pairwise genetic and geographic distances, and 3) demographic analysis based on allele frequencies. A few Ae. aegypti populations on the west coast do not follow the general trend, likely due to a recent and distinct invasion history. We argue that SFE provides a helpful albeit simplified model for the movement of Ae. aegypti across North America, with outlier populations warranting further investigation.

To investigate whether serial founder effect explains the westward spread of Ae. aegypti across southern North America, we analyzed 12 microsatellite loci and SNP array data from 70 North American Ae. aegypti populations, the largest study of the region to date. The microsatellite dataset includes 2,132 individuals from 63 populations genotyped at 12 loci (Slotman et al. 2007; Brown et al. 2011). The SNP dataset includes 359 individuals from 36 populations genotyped at ~15,000 SNPs with the Axiom_aegypti array (Evans et al. 2015). Microsatellite and SNP genotypes for most populations in this study have been reported in Gloria-Soria et al. 2014, Evans et al. 2015, Kotsakiozi et al. 2017, Pless et al. 2017, Saarman et al. 2017, Gloria-Soria et al. 2017, Pless et al. 2020, and Pless et al. 2021. New data presented here include microsatellite genotypes from 1) La Altagracia, Dominican Republic, 2) San Jose de Ocoa, Dominican Republic, 3) St. Croix, USVI, and 4) St. Thomas, USVI and genome-wide SNP data for 1) St. Thomas, USVI, 2) Alamagordo, NM, USA, 3) Las Cruces, NM, USA, 4) Lubbock, TX, USA, 5) Bexar, TX, USA.

 The mean sample size is 33.8 for microsatellites and 10.4 for SNPs. The year range of sampling collections is 2006-2018 (with >75% of the samples collected in 2014 or later). All mosquito samples were collected as adults or eggs from traps and were shipped as adults to Yale University for analysis. No more than six individuals were used from a single ovitrap to minimize sampling relatives. Additional information is available in the associated manuscript. 

The microsatellite data are in Genepop format, and "0" is the missing value.

The SNP array data are in Plink format (bed/bim/fam).