Mosquito species (Diptera: Culicidae) and associated entomofauna breeding in tree holes and artificial containers in three adjacent forests in Western Kenya

Abstract

Tree hole breeding mosquitoes and other associating arthropods were collected from three adjacent forests in western Kenya to determine distribution and habitat preferences. Immature stages were collected from natural tree holes and artificial traps set along horizontal and vertical transects. The horizontal transect passed through ecotone (forest edge), closed canopy, open grass land and river-rhine zones while the vertical transect extended from ground along a tall tree to a height of 15 m in the forest. Seventeen mosquito species were recorded, of which 13 occurred in Kakamega Forest and 11 each in Kaimosi and South Nandi Forests. These species represented five genera; Aedes (9 spp.), Culex (3 spp.), Culiseta (1 sp.), Eretmopodites (2 spp.) and Toxorohynchites (2 spp.). Five non-culicid arthropod species larvae were identified comprising a dragonfly, a tipulid (crane fly), a helodid (marsh beetle), a chironomid (midge) and a psychodid (moth fly). The ecotone, closed canopy and river-rhine zones yielded highest numbers of mosquito species in the tree forests studied. Eretmopodites semisimplicipes (Edwards), Aedes apicoargenieus (Theobald) and Aedes capensis (Edwards) were most abundant specis in the three forests. Ten species occurred in all of the three forests whereas seven species were found only in particular forests. Aedes ajricanus (Theobald) occurred in all the zones in Kaimosi Forest but preferred closed canopy zone. Toxorhynchites barbipes (Edwards) was recorded only at the ecotone zone in each of the forests. The mean number of species per trap did not vary significantly among vegetational zones in Kakamega and Kaimosi Forests but the ecotone zone in South Nandi Forest significantly differed with other zones within the forest. With the exception of Culiseta fraseri (Edwards) and Culex tigripes (Granpre and Charmoy) the 11 mosquito species collected from 65 natural tree holes in 43 trees (mainly Ficus spp.), were also obtained from black tin cans. However, black tin can and open top bamboo pot breeding habitats differed significantly in KakamegaForest but not in South Nandi Forest. Tree holes collections were not significantly different from either black tin cans or open bamboo pots collections. Closed ends bamboo pots significantly differed from open top bamboo pot, black tin can and tree hole breeding habitats. Ae. capensis and Culex nebulosus (Theobald) were significantly associated in black tin cans but were not significantly associated in bamboo pots and in tree holes. Among traps placed 1, 7 and 15 m above the ground, Ae. capensis and Ae. africanus showed a significant preference for the 1 m level. No species significantly preferred either 7 or 15 m height habitat. The least number of species was caught at the 15 m height level and the highest number of species was recorded at the 1 m level in all of the three forests. Seasonalityof the most abundant specie.s in tree holes and black tin cans were not directly correlated. Ae. africanus and Toxory'!chites brevipalpis (Theobald) population peaks were in the relatively dry period. The predatory mosquito, T. brevipalpis tracked or coincided with population patterns of the five most abundant mosquito species in tree holes but not in black tin cans. Since two confirmed vectors of yellow fever, Ae. africanus and Ae. aegypti were collected and non-human primates densities are very high, these forests must be considered areas of potential risk for yellow fever transmission