Comparative evaluation of different versions of exposure-free mosquito electrocuting traps and barrier screen trap for monitoring outdoor densities and biting time phenotypes by malaria and filariasis vectors in Tanzania.

BACKGROUND: Estimating human exposure to mosquito vectors is crucial for the prediction of malaria transmission and intervention impact. The human landing catch method is frequently used to directly measure estimate exposure rates; however, there has been an increasing shift from this method to exposure-free alternatives, such as the mosquito electrocuting traps (MET) and other approaches. While these latter methods can provide robust and representative values of human exposure and mosquito density, they often still require a human volunteer, which poses logistical challenges. Additionally, in the case of the MET, the early MET prototype (METe) required human volunteers to wear protective clothing that could be uncomfortable. We investigated two alternative trapping approaches to address these challenges by comparing the performance of the METe prototype to: (i) a modified caged MET prototype that offers full protection to users (METc) and (ii) a barrier screen trap (BST) designed to passively sample (host-seeking and blood-fed) mosquitoes outdoors without requiring a human participant. METHODS: The relative performance of the METe, METc and BST were evaluated in a 3 × 3 Latin square field experiment design conducted in south-eastern Tanzania over 12 nights of sampling. The outcomes of interest were the nightly catch of mosquitoes and biting time estimates. RESULTS: The METc and BST caught similar numbers of An. arabiensis as the METe (relative ratio [RR] = 0.76, 95% confidence interval [CI]: 0.42-1.39, P = 0.38 and RR = 1.13, 95% CI: 0.63-2.04, P = 0.69, respectively). Similarly, the METc and BST caught similar numbers of Culex spp. as the METe (RR = 0.87, 95% CI: 0.62-1.22, P = 0.42 and RR = 0.80, 95% CI: 0.57-1.12, P = 0.199, respectively). All three trapping methods indicated a similar pattern of biting activity by An. arabiensis and Culex spp., characterized by biting starting in the early evening (18:00-22:00), peaking when people are typically sleeping (22:00-05:00) and dropping off drastically toward the morning (05:00-07:00). CONCLUSIONS: The modifications made to the METe design to improve user comfort and remove the need for protective clothing did not result in an underestimation of mosquito vector abundance nor misrepresentation of their biting time pattern. We recommend the METc for use over the METe design. Similarly, the BST demonstrated potential for monitoring malaria and filariasis vector densities in Tanzania.

Modified veranda-trap hut for improved evaluation of vector control interventions.

Experimental huts with veranda traps have been used in Tanzania since 1963 for the study of residual insecticides for use with insecticide-treated nets and indoor residual spraying. Mosquitoes are allowed unrestricted entry through the eaves to facilitate the collection of an estimable proportion of mosquitoes that attempt to exit through the eave gaps, which are left open on two sides of the hut. This study was designed to validate the use of eave baffles to funnel entry and to prevent mosquito escape, and to determine biting times of Anopheles arabiensis (Patton) (Diptera: Culicidae). Anopheles arabiensis and Culex quinquefasciatus (Say) (Diptera: Culicidae) were released into the room at 20.30 hours and collected the following morning from veranda traps, window traps and the room. Centers for Disease Control light traps hung overnight next to volunteers were emptied every 2 h to determine peak biting times. A total of 55% of An. arabiensis were trapped before 22.30 hours and the highest peak in 'biting' was recorded during 18.30-20.30 hours. Of the released An. arabiensis that exited into veranda traps, 7% were captured in veranda traps entered through baffles and 93% were captured in traps entered through unmodified eaves. When veranda screens were left open to allow for escape outdoors, recapture rates were 68% for huts with eave baffles and 39% for huts with unmodified eaves. The comparison of open eaves with baffled eaves validated the assumption that in huts of the traditional non-baffled design, 50% of mosquitoes escape through open eaves. Eave baffles succeeded in reducing the potential for mosquito exit and produced more precise estimates of effect.