Mosquito blood-feeding patterns and nesting behavior of American crows, an amplifying host of West Nile virus

Abstract

Abstract Background Although American crows are a key indicator species for West Nile virus (WNV) and mount among the highest viremias reported for any host, the importance of crows in the WNV transmission cycle has been called into question because of their consistent underrepresentation in studies of Culex blood meal sources. Here, we test the hypothesis that this apparent underrepresentation could be due, in part, to underrepresentation of crow nesting habitat from mosquito sampling designs. Specifically, we examine how the likelihood of a crow blood meal changes with distance to and timing of active crow nests in a Davis, California, population. Methods Sixty artificial mosquito resting sites were deployed from May to September 2014 in varying proximity to known crow nesting sites, and Culex blood meal hosts were identified by DNA barcoding. Genotypes from crow blood meals and local crows (72 nestlings from 30 broods and 389 local breeders and helpers) were used to match mosquito blood meals to specific local crows. Results Among the 297 identified Culex blood meals, 20 (6.7%) were attributable to crows. The mean percentage of blood meals of crow origin was 19% in the nesting period (1 May–18 June 2014), but 0% in the weeks after fledging (19 June–1 September 2014), and the likelihood of a crow blood meal increased with proximity to an active nest: the odds that crows hosted a Culex blood meal were 38.07 times greater within 10 m of an active nest than > 10 m from an active nest. Nine of ten crow blood meals that could be matched to a genotype of a specific crow belonged to either nestlings in these nests or their mothers. Six of the seven genotypes that could not be attributed to sampled birds belonged to females, a sex bias likely due to mosquitoes targeting incubating or brooding females. Conclusion Data herein indicate that breeding crows serve as hosts for Culex in the initial stages of the WNV spring enzootic cycle. Given their high viremia, infected crows could thereby contribute to the re-initiation and early amplification of the virus, increasing its availability as mosquitoes shift to other moderately competent later-breeding avian hosts. Graphical Abstract