The present study evaluated the attractiveness of the ASB Sarabi v1.2.1 developed by Westham Co. for mosquito vectors in a region of South-Central Tanzania with both malaria and dengue transmission using a camera station. This is the first study to demonstrate that such camera stations offer a simple solution for assessing and comparing the attractiveness of ASBs and other potential mosquito attractants in semi-field and field settings. In the semi-field system, comparisons between Westham ASB Sarabi v1.2.1 and sucrose solution showed that mosquitoes overall were similarly attracted to both. This is in line with a previous study conducted in Coastal Tanzania which found that locally made ASBs were equally attractive to sucrose solution [22]. Here, however, in the semi-field system, An. arabiensis visited the ASBs more than the sucrose solution whereas An. funestus did not. Comparisons between ASBs and ASB blanks lacking the attractive odour blend also showed that the ASB attractant was attractive to the mosquitoes in the semi-field system. The low overall number of mosquito visits observed in relation to the number of mosquitoes released in these experiments may be attributed to the fact that they were carried out during the dry season which made for low overall mosquito activity. Fewer visits also meant lower statistical power, thus seasonality and patterns of mosquito activity play a major role in the success and need for replication in such experiments.
In contrast to that the field studies were conducted during the rainy season and showed more definitively that the ASB attractant which was used in the Westham ASB stations v1.2.1 is very effective in attracting mosquitoes when compared to the ASB blank in the control arm. This confirms that the attraction of the Westham ASB station is associated with its odour bait and therefore that olfactory attraction appears more important than any visual attraction of the bait station. It is noteworthy, that despite the clear attraction of An. arabiensis and C. quinquefasciatus to ASBs, the results obtained from the human landing catch conducted simultaneously with the camera recordings suggest a relatively low overall attractiveness of these ASB v1.2.1 stations compared to humans. HLCs indicated that mosquito densities in Lupiro village were high during the month of May, which aligns with the rainy season in the region. On average, 378 mosquitoes landed on capturers per night compared to an average of 2 landings on baits per night. Our findings suggest two hypotheses regarding the observed low visitation to the baits when compared to the HLC. Firstly, it may be possible that sugar feeding on the ASB bait stations is limited during the rainy season as there are a lot of flowers and fruits that mosquitoes can rely on for sugar at that time. In such circumstances, mosquitoes may feed mainly on natural sugar sources because of their abundance compared to the limited number of ASBs. This first hypothesis, therefore, focuses on sugar usage as a dietary complement for body maintenance and energy for flight and mating. A slightly different explanation would be that the high availability of water sources reduces the reliance on nectar, which can be sought by mosquitoes both for its water and sugar content. During the rainy period, mosquitoes might reduce their reliance on nectar or sugar solution because they find water droplets or puddles very easily and thus exhibit an overall reduced attraction to natural sugar sources and artificial baits. Thus, in future, it will be important to understand the dynamics of sugar feeding, in relation to its dual role as a source of water and or energy, and its changes in availability throughout different seasons. Interestingly, the data collected from the two stations baited with sucrose suggest that the ASB attractant performed better compared to this comparator in the field, which contrasts with the finding in the semi-field system. However, there were only two sucrose stations available for the field comparison and so further work is required to confirm that finding. Further work using larger sample sizes should also demonstrate whether ASB competes well in terms of long-range attraction with natural sources of sugar and nectar and under what conditions. Further studies should, therefore, formally test the efficacy of ASB attractant across different seasons and geographical sites to highlight the relationship between natural sugar availability and ASB efficacy.
With regard to the specificity of the ASB attractant, this study confirmed that the Westham ASB attractant was most attractive to An. arabiensis compared to An. funestus in the Kilombero Valley. However this has to be tested in the semi-field to see if it is related to relative survival or ASB attractiveness. An. arabiensis made longer visits to the ASB station with attractant and sucrose bait system than on ASB blank, that is a good indication that feeding activities took place.Although An. arabiensis were more frequently found on ASBs in the field, it's essential to note that this species was generally more abundant than An. funestus according to HLCs. Therefore, the higher presence of An. arabiensis on ASBs isn't necessarily indicative of greater attraction to these baits. Additionally, in the[35] Zambia field trial, a higher proportion of An. funestus were observed to have fed on ASBs compared to An. arabiensis. This finding underscores the importance of considering species-specific feeding behaviors and abundances in evaluating the efficacy and attractiveness of control measures such as ASBs.
Whilst the baits tested here did not include a killing agent, it remains important that mosquitoes not only explore the baits but effectively feed on them to pick up a sufficient dose of killing agent [22]. C. quinquefasciatus also fed on ASBs, but their numbers on baits were much less than expected given their great abundance in HLC catches. Ae. aegypti was never observed on the baits in the field despite being present in HLC catches, although at very low numbers. Under semi-field settings, this species was only very rarely observed on the baits. This would suggest that Ae. aegypti might require a different blend of attractants or may be even more prone to feeding on natural sugar sources than Anopheles or Culex. Additionally, this study highlighted other important mosquito behavioural factors, such as sex-specific differences in attraction to sugar feeding In our study, female mosquitoes visited more the ASB than male mosquitoes, which may imply that the ASB attractant or bait format is more attractive to female mosquitoes but less so for males, though semi-field sex comparisons may have been affected by relative survival rate. In contrast to our finding, a field study that used the ASB station v1.1.1 developed by Westham Co. in Zambia, observed a higher proportion of uranine positive male mosquitoes than females, implying that male mosquitoes were feeding more on the bait stations [35]. Our finding of female attraction to ATSBs in South-Central Tanzania supports the potential use of ASBs for malaria control programmes in that region as female mosquitoes are the ones responsible for blood feeding on hosts, hence transmitting pathogens [36]. Interestingly, the present study documented for the first time and through direct observation the timing of landing on the baits in the field by different species. In the field, landing on the ASB started from 17:00 h and continued till 20:00 h in the evening, with a second peak of landings taking place around 5:00 h to 7:00 h in the morning. Therefore the start and end time of sugar feeding activity were comparable to those observed in host seeking female An. arabiensis and C. quinquefasciatus from the HLC sampling performed in this study which has also in line with studies describing the natural host seeking behaviour of An. gambiae s.l. An.arabiensis and An. funestus (Degefa et al., 2021; Kabbale et al., 2013). However, unlike for host-seeking, our results and those of other studies show that sugar feeding starts early in the evening and continues for 3–4 h, followed by a clear drop in activity later at night and another distinct peak of activity early in the morning (Degefa et al. 2021; Kabbale et al. 2013 Müller et al. 2010)
In order to assess the proportion of male and female sugar feeding ahead of ATSB trials, other studies have used the cold anthrone method to detect sugar uptake in anopheline species [39]. Another approach used consisted in collecting the content of light-traps baited with flowers at 1h time intervals to infer the timing of feeding on natural sugar by male and female An. gambiae s.l. [21]. It is noteworthy that such indirect methods cannot possibly generate data on the relative proportion and timing of visits to baits by vector species as accurately as those measured from direct visual recording on ASB as implemented in this study.
The camera stations deployed in the field also generated important data on visits by non-target organisms (NTOs) to the baits. These visits were rare and the taxa involved included Araneae (spiders), Hymenoptera (ants, wasps), and Lepidoptera (moths). Such observations are in line with findings from previous studies [40–44]. The latter studies relied on identifying NTOs that fed from ATSB through detecting food dye or staining in all insects collected by Malaise traps, plate traps, UV traps, sweep nets, and pitfall traps [40–44]. The varying efficacy of the trapping and marking methods used and complexity of detection in these studies make them susceptible to various biases. Camera stations are a much more direct method for recording the attraction to baits of any NTO taxa, including visits by vertebrates such as that of a gecko which our study recorded. We also found variation in local NTO abundance between ASBs with Camera number 817 recording more NTO visits than other camera traps. This may be attributed to the location of that camera trap at the fringe of the village in a wooded area.
No serious issues with the camera stations were observed. The camera traps produced adequate image quality with no major difficulties in recognizing species and determining the sex of mosquitoes. No major data collection difficulties were encountered, except that identifying images positive for mosquitoes or NTOs from all generated images takes time. It is important to note that mosquitoes, being cold-blooded organisms, cannot trigger the camera's built-in passive infrared trigger when landing on the ATSB or entering the camera's field of view. Therefore, we recommend using a 24 h time-lapse approach with images taken at 1 min intervals for tracking mosquito landings in ATSB studies. The downside of that approach is that one camera trap will produce approximately 4,320 images per 72 h. To save time in viewing all those images, we converted 24 h stacks of images into mp4 video files using the Mac iMovie software. This enabled fast scrolling through the video created to quickly detect those frames with mosquitoes and recording their landing and departure times. The possibility of analyzing videos using more elaborate image analyses software assisted by machine learning is currently being evaluated [45, 46].
One important finding from our dry season semi-field experiments comparing ASBs to ASB blanks was that when conditions in enclosures were very dry (above 35°C), mosquitoes did not engage at all in our experiments and hid and died. Considering that 198 mosquitoes were released in the experimental chamber and observations were made for 72 h, we collected very few positive images of mosquitoes landing on the ASB. Thus, mosquitoes seemed to have hidden in the clay pots provided to mitigate desiccation but did not visit the ASB attractant under those conditions. Considering these observations, predicting what level of ASB visitation will be detected under field conditions during very dry weather is difficult. Whilst increased sugar feeding is expected as long as mosquitoes are active, very harsh conditions are also likely to induce further protective behaviour such as actively avoiding open and arid locations, thus enabling them to conserve moisture and survive. This might include seeking shelter, resting in cool and shaded areas, and even estivating [47, 48].
In addition to threshold levels of drought tolerance, the availability and quality of sugar sources to mosquitoes may also play a role. In the dry season, natural sugar sources such as nectar-producing plants might be scarce, further increasing the attractiveness of sugar-based attractants like those used in ASBs to mosquitoes. This is supported by a study in central Tanzania in semi-field enclosures where dense-, sparse- and no-vegetation settings were simulated during the dry and wet seasons; ATSBs were more visited under the bare site without any vegetation settings than the one with dense or sparse vegetation [49]. Though that study was performed in the semi-field with a controlled environment in both dry and rainy seasons, it further emphasizes the need for field studies comparing ATSB attractiveness in settings or regions with dense vegetation and rich in flowering plants and in regions with sparse or semi-arid areas to understand the full complexity of sugar feeding behaviour of mosquitoes in the field across the season.
The camera traps used in this study are well suited for this purpose because they can generate direct data for comparisons of ATSB attractiveness in different settings. In the field setting, it may be difficult to rely on detecting food dye[11] or the uranine marker (the Westham ASB used in this study had a uranine marker while the ASB blank had no marker) to estimate the percentage of mosquitoes that have fed on the ASB under investigation [35]. The recapture rate in the field is normally very low, due to the harsh conditions of the environment [50]. It is, therefore, difficult to trace or capture a mosquito that visited and fed on the ASB bait station without mass ASB deployment. Additionally, the position of traps seems to be critical in assessing feeding rate if using a bait dye in the field (uranine or food dye). Studies conducted in Mali, West Africa successfully used a glue trap method to evaluate the relative attractiveness of ATSBs in the field, however the smell of glue itself can have a repellent effect [21]. Here again, the camera trap method can be used to assess the attraction of ATSBs without the need to deploy large numbers of ATSBs and positioning glue or other traps to indirectly evaluate ATSB attractiveness in the field. The camera stations allow a simple measure of attraction whereas trap-based feeding rate assessments, though important, may further depend on: 1) the short-range stimulus needed to feed once mosquitoes have landed, 2) the accessibility of the bait, and 3) its palatability. So the use of the camera traps in the field may arguably generate the best measure of ATSB attractiveness independent of any other factors. Further studies may be necessary to understand whether all the landings observed led to effective feeding. This could simply be established through comparing landing rates estimated by camera stations to feeding rates estimated via dye detection in semi-field studies.