Comparison of Different Trapping Methods to Collect Malaria Vectors Indoors and Outdoors in Western Kenya

Abstract

Abstract Background Enhanced vector surveillance, is one of the 4 pillars of the WHO’s global vector control response (2017–2030). Human landing catches are the gold standard for entomological surveys but are difficult to implement and expose collectors to mosquito bites and potentially to malaria infection. Other surveillance tools such as light traps, pyrethrum spray catches and aspiration are less expensive and do not expose collectors to potentially infectious mosquitoes, but they are difficult to implement outdoors and/or to assess duration of collection/standardize collection effort. This study evaluated four mosquito trapping methods that may be cheaper, easier, and less risky to implement compared to human landing catch. Methods Three mosquito sampling methods (UV light traps, CDC light traps and Prokopack aspiration) were evaluated against human landing catches in two villages of Rarieda sub-county, in Siaya County, western Kenya. UV light traps, CDC light traps and human landing catches were conducted in three locations: inside houses, 10 meters from the house and 10 meters from the compound boundary. These were done every hour from 17:00 until 07:00. Prokopack aspiration was done indoors and outdoors of houses adjacent to the light trap and HLC houses from 07:00 until 11:00. Analyses of mosquito densities, species abundance and sporozoite infection prevalence were performed across all sampling methods. Species within the An. gambiae and An. funestus species complexes were identified using PCR. ELISAs were used to determine mosquito sporozoite infection prevalence. Data analysis was done in R statistical software. Results A total of 5,370 male and female Anopheles mosquitoes were sampled from 608 trapping efforts. An. funestus constituted 70.3% (n = 3,877) of the sampled Anopheles mosquitoes while An. coustani was 19.7% and An. gambiae s.l. was much lower at 7.2%. 93.8% of An. funestus s.l. samples processed through PCR were An. funestus s.s. and 97.8% of An. gambiae s.l. were confirmed to be An. arabiensis. Only An. funestus samples were positive for sporozoites, with a species specific sporozoite infection prevalence of 3.1%. Indoor aspiration captured the highest number of An. funestus (mean = 6.74; RR = 7.49 compared to indoor HLC, 95% CI 3.95–14.22, P < 0.001) followed by indoor UV-LT, (mean = 3.7; RR = 3.6, 95% CI 2.02–6.42, P < 0.001) and indoor CDC-LT (mean = 1.74; RR = 1.85, 95% CI 1.02–3.33, P = 0.042). In pairwise comparisons, significantly different numbers of An. funestus were collected by all indoor methods with the most collected by aspiration and the fewest by HLC. For An. arabiensis, indoor UV-LT and indoor CDC-LT each captured an average of 0.18 per trap-night which were significantly higher than HLC indoors. Outdoors, UV-LT collected significantly higher numbers of Anopheles mosquitoes across all species analyzed (An. funestus: mean = 1.69, RR = 4.27 compared to outdoor HLC, 95% CI 2.20–8.31, P < 0.001; An. arabiensis: mean = 0.22, RR = 15.64, 95% CI 1.97-124.36, P = 0.009; An. coustani: mean = 3.74, RR = 10.48, 95% CI 4.37–25.14, P < 0.001) when compared to outdoor HLC. Hourly biting in UV-LT and CDC-LT indicated different peaks compared to HLC for An. funestus collected indoors. Conclusions Anopheles funestus remains the predominant malaria vector in the region and was primarily caught indoors. Anopheles arabiensis were trapped in similar both indoors and outdoors while and An. coustani were mostly collected outdoors with UV-LTs. UV-LT and CDC-LT collected higher numbers of the primary Anopheles mosquitoes indoors and outdoors except for An. funestus indoors where aspiration was the most efficient method. The UV-LT generally collected more mosquitoes than the CDC-LT indicating UV-LTs may be an efficient tool for monitoring populations of Anopheles mosquitoes. Differences in hourly biting by different collection methods indicate the need to further investigate the behaviour of An. funestus.