Study areas
The study was carried out in two ecological settings, Vallée du Kou and Soumousso, two villages close to Bobo-Dioulasso, Burkina Faso:
Vallée du Kou, a village located in the north-west of Bobo-Dioulasso. The site is characterized by wooded Savannah and covers 1,200 ha. Vallée du Kou (11˚23' N, 4˚24' W) is a rice area with high mosquito densities all the year. A high insecticide resistance level is then occurring in this area due to the large use of insecticide in rice culture and the cotton around. The village is composed of seven quarters and the trial has been conducted in one of these quarters, Vallée du Kou 3 (VK3). Given the presence of water all year round, mosquitoes are found accordingly with a peak density observed in August-September during the rainy season. An. gambiae and An. coluzzii are alternated during the year with a predominance of An. coluzzii throughout the year. Both species in this area are highly resistant to pyrethroids and DDT (kdr based mechanism, (0.8–0.95). A raise in ace-1 resistance frequency is also observed.
Soumousso, located in the southern of Bobo-Dioulasso, 11˚04' N, 4˚03' W), is a drier setting than VK3 with less water and where the most dominant species are An. gambiae and a mixture of An. funestus, An. arabiensis and other sibling species of An. gambiae s.l.
In this area, the mosquito density is lower compared to that of Vallée du Kou and the dynamics of the mosquitoes follow the rainy season scale.
Description Of The Traps
The original LFET prototype (Prototype 1) (P1) was described by Diabaté and collaborators in 2013 [18] and was made of a metal frame (Long = 69 cm, Large = 51 cm, Height = 165 cm) fitted from the bottom to the top with a regular mosquito net to prevent mosquitoes and other insects entering the trap to escape (Fig. 1 ). A funnel made of metal frame was inserted at the top of the trap in such a way that mosquitoes approaching the window go first through the big opening of the funnel and enter the trap passing through the small opening.
The big opening of the funnel is long of 70 cm and the diagonal is long of 54 cm while the small opening is long of 13.3 cm and large of 11.2 cm. The small opening of the funnel is distant of 10 cm from the backside of the trap. The funnel, the way and the place where it is inserted in the cage allow mosquitoes to enter easily the trap but prevent them from escaping. Once mosquitoes enter the trap, they have a big volume beneath the funnel where they can find refuge. For a mosquito to escape, it will have to fly all the way up towards the small opening of the funnel and be able to navigate through the 10 cm space separating the small opening of the funnel and the back side of the cage. Ultimately it will fly to exhaustion before it can find the way out. The principle of LFET is to lose the mosquito inside the volume of the trap until dehydration and finally death without finding the way out to exit. The trap is fitted with three sleeves on the side (one below, one in the middle and one on the top) through which mosquitoes were aspirated from the cage. The trap was secured to window using nails.
The Prototype 2 (P2) was built on the same model of the original LFET prototype but half reduced, with the same dimensions except the funnel height of 82.5 cm (Fig. 2). The whole trap is covered by a net, fitted with three sleeves for mosquito collection.
The Prototype 3 (P3) is made of a metal funnel frame measuring (Length = 81 cm, Height = 80 cm). Here the small funnel opening used as entry is a circular metal funnel which opening size is 16.5 cm of diameter, instead of small rectangle in the original prototype. Two circular openings on right and left (distance between both circular openings is 12.5 cm) were planned to fit the window position and the place where it should be fixed to the wall in the house. When one is used, the second is closed by the net. The volume of trap was made of a rectangular metal form (dimensions are, Length = 110 cm, width = 16 cm, height = 56 cm) covered by a net. A net is covering the trap on backside of the funnel. This trap fitted five sleeves allowing mouth aspiration of mosquito within the cage (Fig. 3).
The prototype 4 (P4) is similar to the prototype 3 with the same dimensions of funnel and volume, but the difference between both is a little circular funnel frame extending the funnel. This small circular funnel is 9 cm-long from the beginning to the end, like a shrinking canal giving access to the beneath volume. This last circular funnel (10.5 cm of diameter) is the funnel whereby mosquitoes must pass to access the trap volume. This last circular funnel is far from the back side of the trap of 10 cm. In addition, this trap was equipped with a mirror for internal uses within the house. This trap was also outfitted by five sleeves enabling the mosquito collection. Other difference between both prototypes 3 and 4, is the mirror added. Finding a useful trap for the occupants coupled to reducing mosquito density may help increasing the acceptability and friendly use of the trap (Fig. 4).
Study Design And Mosquito Collection
Three LFET prototypes (P 2, 3, 4) were tested in two ecological settings in comparison to the original LFET prototype (P1). The performance of the traps was assessed in terms of trapping and suppressing the indoor malaria vectors as well as other vectors entering the house through window. A total of 12 houses were chosen in each testing site. Only houses with single room, single window and single door were included in the study. To ensure that mosquitoes had no other alternative except the windows to enter in the house, all the small holes of the ceiling and the wall were closed with sponges or clothes and a curtain was placed at the entrance of each house (Fig. 5). The inhabitants were informed on the aim of the study and therefore free to use their doors as they will. The window with LFET was left continuously opened to allow the mosquitoes to access the inside of house. Trap installation and mosquito sampling were performed over 8 consecutive days per month, from September to November in VK3 and Soumousso.
The daily mosquito collection started at 07:00 am using a mouth aspirator and a pyrethrum spray catch [19] was performed in the trap and the corresponding house. All the mosquitoes caught were kept in single cup, then killed with chloroform and morphologically identified on site. The daily work on site consisted in mosquito separation, sorting, identification, sexing according to Gillies key [20], counting and scoring per genus, species, physiological status and the numbers recorded on a spreadsheet.
All traps set up each month were removed at the end of the trial. Along this study, the villagers involved in the trial, were asked to feel free to give their opinion on any inconvenience related to the new LFET prototypes.
Statistical analysis
Microsoft Excel 2007 (Microsoft®, New York, USA) was used to record data, and R-3.6.2 (package dplyr, questionr, and coin) to produce the graphs and the statistical analyses.
In the study, four main variables were analyzed:
1) Total number of An. gambiae s.l. caught calculated as (number in trap) + (number in corresponding room);
2) Proportion of mosquitoes caught in trap: calculated as (number caught in trap)/ (total number)
3) Proportion of entry reduction: calculated as the proportion of mosquitoes caught in trap/total of mosquitoes caught (trap + room) × 100.
4) A daily removal of mosquitoes was calculated as the number of all mosquitoes collected per site (trap + house) /24 days (8 consecutive days over 3 months)
All collected mosquitoes in the traps were sex-sorted. Female mosquitoes were sorted according to the gonotrophic status and the proportion of each status (gravid, blood fed and unfed) was estimated by dividing the number of females of the specific status by the total number of females caught in the traps.
The numbers of mosquitoes in the trap and the matching house did not follow a normal distribution. Therefore, a non-parametric test, Kruskal-wallis sum test, was used to assess the overall performance of the traps. The effects of monthly collection in mosquito density reduction were evaluated using a Levene’s test for homogeneity of variance (center = median) with a Wilcoxon rank sum test with a p-value bonferroni adjustment method. A Tukey multiple comparison of means with 95% family-wise confident level test, was used for pair-wise comparison between LFET prototypes.
To assess whether the traps caught more mosquitoes than those entered in the matching houses, a comparison using Wilcoxon rank sum test with a holm p-value adjustment method was used.