Secondary malaria vectors in western Kenya include novel species with unexpectedly high densities and parasite infection rates.

Background: Malaria vector control has been implemented chiefly through indoor interventions targeting primary vectors resulting in population declines-pointing to a possible greater proportional contribution to transmission by secondary malaria vectors with their predominant exophagic and exophilic traits. With a historical focus on primary vectors, there is paucity of data on secondary malaria vectors in many countries in Africa. This study sought to determine the species compositions and bionomic traits, including proportions infected with Plasmodium falciparum and phenotypic insecticide resistance, of secondary vectors in three sites with high malaria transmission in Kisumu County, western Kenya.

Methods: Cross-sectional sampling of adult Anopheles was conducted using indoor and outdoor CDC light traps (CDC-LT) and animal-baited traps (ABTs) in Kakola-Ombaka and Kisian, while larvae were sampled in Ahero. Secondary vectors captured were exposed to permethrin using WHO bioassays and then analyzed by ELISA to test for proportions infected with P. falciparum sporozoites. All Anopheles were identified to species using morphological keys with a subset being molecularly identified using ITS2 and CO1 sequencing for species identification.

Results: Two morphologically identified secondary vectors captured-An. coustani and An. pharoensis-were determined to consist of four species molecularly. These included An. christyi, An. sp. 15 BSL-2014, an unidentified member of the An. coustani complex (An. cf. coustani) and a species similar to that of An. pharoensis and An. squamosus (An. cf. pharoensis). Standardized (Anopheles per trap per night) capture rates demonstrate higher proportions of secondary vectors across most trapping methods-with overall indoor and outdoor CDC-LTs and ABT captures composed of 52.2% (n = 93), 78.9% (n = 221) and 58.1% (n = 573) secondary vectors respectively. Secondary vectors were primarily caught outdoors. The overall proportion of secondary vectors with P. falciparum sporozoite was 0.63% (n = 5), with the unidentified species An. cf. pharoensis, determined to carry Plasmodium. Overall secondary vectors were susceptible to permethrin with a > 99% mortality rate.

Conclusions: Given their high densities, endophily equivalent to primary vectors, higher exophily and Plasmodium-positive proportions, secondary vectors may contribute substantially to malaria transmission. Unidentified species demonstrate the need for further morphological and molecular identification studies towards further characterization. Continued monitoring is essential for understanding their temporal contributions to transmission, the possible elevation of some to primary vectors and the development of insecticide resistance.