Mosquito and arbovirus surveillance is essential to the protection of public health. A majority of surveys are undertaken at ground level. However, mosquitoes shelter, breed, and quest for hosts across vertical strata, thus limiting our ability to fully describe mosquito and arboviral communities. To elucidate patterns of mosquito vertical stratification, canopy traps were constructed to sample mosquitoes at heights of 1.5, 5.0, and 8.7 m across three different landscape types in a Florida coastal conservation area. We assessed trapping efforts using individual-based rarefaction and extrapolation. The effects of height, landscape, site location, and sampling date on mosquito community composition were parsed out using permutational ANOVA on a Hellinger-transformed Bray-Curtis dissimilarity abundance matrix. Lastly, a generalized linear mixed effects model (GLMM) was used to explore species-specific vertical patterns. We observed differences in sampling effort and community composition structure across various heights and landscapes. Our GLMM revealed significant effects of trap height for Aedes taeniorhynchus, Anopheles crucians, Anopheles quadrimaculatus, and Culex coronator, but not for Culex nigripalpus, the ultra-dominant species present in this area. Together these data provide evidence that height and landscape significantly affect mosquito community structures and highlight a need to develop sampling regimes to target specific vector and nuisance species at their preferred height and across different landscape types.
Keywords: GLMM; canopy; mosquito; stratification; vector.