Sampling outdoor, resting Anopheles gambiae and other mosquitoes (Diptera: Culicidae) in western Kenya with clay pots

Abstract

Clay pots were analyzed as devices for sampling the outdoor resting fraction of Anopheles gambiae Giles (Diptera: Culicidae) and other mosquito species in a rural, western Kenya. Clay pots (Anopheles gambiae resting pots, herein AgREPOTs), outdoor pit shelters, indoor pyrethrum spray collections (PSC), and Colombian curtain exit traps were compared in collections done biweekly for nine intervals from April to June 2005 in 20 housing compounds. Of 10,517 mosquitoes sampled, 4,668 An. gambiae s.l. were sampled in total of which 63% were An. gambiae s.s. (46% female) and 37% were An. arabiensis (66% female). The clay pots were useful and practical for sampling both sexes of An. gambiae s.l. Additionally, 617 An. funestus (58% female) and 5,232 Culex spp. (males and females together) were collected. Temporal changes in abundance of An. gambiae s.l. were similarly revealed by all four sampling methods, indicating that the clay pots could be used as devices to quantify variation in mosquito population density. Dispersion patterns of the different species and sexes fit well the negative binomial distribution, indicating that the mosquitoes were aggregated in distribution. Aside from providing a useful sampling tool, the AgREPOT also may be useful as a delivery vehicle for insecticides or pathogens to males and females that enter and rest in them.

Fig. 1

(A) Manipulation of an AgREPOT to retrieve mosquitoes resting in it. (B) Three AgREPOTS deployed outside a human dwelling at the study site. (C) An. gambiae s.l. female resting in an AgREPOT.

Fig. 2

Total number of An. gambiae s.s. and An. arabiensis males and females taken by four sampling methods in western Kenya. The number over each pair of bars is the percentage that were An. gambiae s.s. Pot, clay pot (AgREPOT); PSC, pyrethrum spray catch; CC, Colombian curtain; and Pit, resting pit.

Fig. 3

Percentage of An. gambiae s.s. and An. arabiensis females in three physiological categories (blood fed, gravid, or unfed) taken by four sampling methods in western Kenya. Pot, clay pot (AgREPOT); PSC, pyrethrum spray catch; CC, Colombian curtain; and Pit, resting pit.

Fig. 4

Frequency histograms of distribution of mosquitoes sampled by different methods in villages of western Kenya. Fits to the negative binomial distribution, and estimations of the dispersion parameter, k, are shown.

Fig. 5

Mean number of mosquitoes taken in four different sampling methods per compound, in biweekly sampling intervals from April to July 2005, in rural western Kenya. A-E are species and sexes as indicated, with E as Culex spp. males and females combined. F is total rainfall (millimeters) per day at the study site. Pot, clay pot (AgREPOT); PSC, pyrethrum spray catch; CC, Colombian curtain; and Pit, resting pit.