Abstract
Background Attractive Toxic Sugar Baits (ATSBs) successfully reduced Anopheles mosquito vector populations and malaria parasite transmission in Mali, but application methods need to be improved for wide-scale use, and effects on non-target organisms (NTOs) must be assessed. The goals of this study were to determine on a village level the effect of different outdoor configurations of ATSB bait stations to 1) achieve > 25% Anopheles mosquito vector daily feeding rate for both males and females and 2) minimize the effect on non-target organisms. Methods Dye was added to Attractive Sugar Bait Stations (ASB – without toxin) to mark mosquitoes feeding on the sugar baits, and CDC UV light traps were used to monitor mosquitoes for the presence of the dye. Yellow plates, pitfall traps, Malaise traps, UV light traps, UV tray traps, and sweep nets were used to trap and sample non-target organisms (NTOs) for dye, indicating feeding on the ASB. ASB stations were hung on outer walls of village homes to determine the impact of different densities of ASBs (1,2, or 3 per home) as well as the impact of ASB height (1 m or 1.8 m above the ground on sugar feeding by anophelines. These experiments were carried out separately, on consecutive nights for mosquito and NTO monitoring. Eight villages in the Koulikoro province were chosen as the experimental locations. Results The use of one ASB station per house marked 23.11% of female and 7.11% of male An. gambiae s.l. While two and three ASB stations per house gave feeding rates above the 25% goal, there was no statistical difference in the percentage of marked mosquitoes (p=0.3141 females; p=0.9336 males). There was no difference in sugar feeding on ASB stations when hung at 1.0 and 1.8 m and (p=0.5170 females; p=0.9934 males); however, ASBs at 1.8 m had less accidental damage from village residents and animals, and subsequent invasion of non-targets through rips or holes produced. ASB stations at 1.8 m above ground were fed on by three of seven monitored insect orders. Feeding rates were less than 0.015% of total trap catches and as low as 0.0001%. The monitored orders were: Hymenoptera [ants (Formicidae), bees (Apidae), and wasps (Vespidae)], Lepidoptera (Rhopalocera, Bombyces, Geometroidea, Noctuoidea, Sphingidae, Pyraloidea), Coleoptera (Carabidae, Tenebrionidae, Scarabaeidae, Cerambycidae, and Chrysomelidae), Diptera (Brachycera, Chironomidae), Hemiptera (Cicadomorpha and Heteroptera), Neuroptera (Myrmeleontiformia) and Orthoptera (Caelifera and Ensifera). Using one or two stations limited evidence of NTO feeding to ants (Hymenoptera), Brachycera, Heteroptera, Noctuiodea, Rhopalocera, wasps (Vespidae) and wild bees (Apidae) (both Hymenoptera) and had a significantly reduced percentage of stained individuals compared to three stations which had the highest feeding rates amongst NTOs. The percentages of stained individuals were as follows: 6.84 ± 2.03% Brachycera were stained followed by wasps (Hymenoptera: Vespidae) 5.32 ± 2.27%, and Rhopalocera 2.22 ± 1.79%. Hanging the optimal number of stations per house for catching mosquitoes (two) 1.8 m above ground, limited the groups of non-targets to Brachycera, Chironomidae, Noctuoidea, Rhopalocera, parasitic wasps and wasps (both Hymenoptera: Vespidae). The three most commonly stained non-target insect groups at this height were wasps (Vespidae) (1.65 ± 0.75%), Chironomidae (0.99 ± 0.37), and Brachycera (1.55 ± 0.69%). Feeding at this height only occurred when stations were damaged.Conclusions The goal of marking one quarter of the total Anopheles mosquito vector population per day was obtained using 2 bait stations at 1.8 m height above the ground on the outer walls of houses. This configuration of ATSB stations also had minimal effects on non-target insects: only 0.0001% to 0.013% of specimens (in three orders) were marked. Stations hung 1.8 m above the ground had less accidental damage from passing people and livestock. The minimal marking of non-target insects may be attributed to visual orientation of non-mosquito insects while mosquitoes, are mostly guided by olfactory cues. Furthermore, the bait stations have a membrane cover, which if intact, is impenetrable to most sugar feeding non-target insects but is pierced by the stylets of the mosquito proboscis. Thus, most non-target insects are not exposed to the toxin even if they approach the bait stations.
Physical and behavioural abilities of commensal rodents related to the design of selective rodenticide bait stations
