Over a period of 28months, 309 female Anopheles, from 10 species, were collected in 383 traps placed in the habitat of two chimpanzee communities (Kanyawara and Kanyanchu) in Western Uganda. The number of species caught was consistent with other studies conducted in tropical areas: 10 species were collected in Malaysia in eight nights . The number of anopheline mosquitoes caught might be considered to be quite low as compared to the numbers obtained in some studies conducted in human and livestock environments. For example, in a 1.5-year study conducted with light traps in Kenya, about 15,000 and 60,000 female Anopheles from only four species were collected inside and outside houses, respectively . On the other hand, when the mosquito surveys were conducted in an environment similar to that where the chimpanzees live, namely cool highlands, the number of Anopheles collected is often very low: for instance using the same methodology and sampling protocols, 107 Anopheles were collected at Kyenjojo (about 20km to Kibale NP) as compared to the 10,127 Anopheles collected in Tororo, a dry savannah grassland area .
Ultimately the ecology and host preference of the Anopheles species identified in the home range of chimpanzees remain poorly documented. However, the meager observations do not exclude a role in malaria transmission. Only one of these species, Anopheles paludis, has been incriminated as an important vector of human malaria , though this role appeared to vary with geographical location. The forest species A. implexus has been described as an anthropophilic taxon, and in 1960, Lips found one female infected with malaria out of 1,200 dissected. Breeding places were usually described to be swamps, stagnant shady drains, water containing dead leaves and elephant tracks , but Lambrecht  failed to find any larva or egg in the forest gallery where adults where collected despite repeated efforts. Anopheles obscurus, commonly found in forests, does not bite humans but was found carrying malaria oocysts [68, 69]. Anopheles squamosus, abundant from the peak to the end of the rainy season , is close to Anopheles pharaoensis but its role in human malaria transmission does not seem to be major. Anopheles ziemani is mostly described as a zoophilic species but has nevertheless been implicated in human malaria transmission . Anopheles demeilloni lives in altitude (700 to1,800m), as does Anopheles harperi a rather zoophilic and exophilic species. Anopheles vinckei is poorly described and its distribution is limited to Oriental Kivu in Democratic Republic of Congo and Western Uganda.
Three Anopheles species (A. implexus, A. vinckei and A. demeilloni) represented more than 94% of the catch but only one, A. implexus, was dominant in the vicinity of the chimpanzees nests where it was consistently captured (21months over 24 of study). In a previous study , blood samples were obtained serendipitously from three Kanyawara community chimpanzees: one was injured in the course of a fight in September 2006, one as it was released from a poachers snare in October 2006, and the third in the course of a post-mortem in January 2007. Molecular analysis of the genomic DNA prepared from these samples revealed that all these three chimpanzees sampled randomly, within the study-period of the present survey, were infected with multiple species of Plasmodium. Given that the Kanyawara community comprises only 44 individual chimpanzees, the presence of infections in three strongly suggests a very high prevalence of malaria in these apes, and this can only be due to high transmission rates, chronic persistence of parasites over long durations, or both. Observations of mixed infections in wild-caught chimpanzees are common as are sub-patent infections that persist for many years [15, 16]. The prevalence of infection was high in the human blood samples analyzed, with Plasmodium parasites detected using the same PCR assays in 44 of the 74 (60%) persons sampled.
Plasmodium parasites could not be detected by PCR amplification in any of the Anopheles mosquitoes caught. Given the very low numbers of mosquitoes that were actually caught in the vicinity of the chimpanzee nests, this was to be expected since the infection rates observed in mosquitoes are usually low (< 1%) as are the parasite burdens (an average of one oocyst) even in mosquitoes caught in areas of high malaria endemicity. Ideally, one can optimize the chances to uncover the mosquitoes that transmit the parasite by collecting specimens that have fed on the chimpanzees, but this is impossible to achieve for individuals in the wild. One can also minimize the possibility of degradation by processing the mosquito material for DNA extraction immediately after collection. In this manner, it will be possible to target studies of chimpanzee behaviour in relation to mosquito avoidance. However, such studies are technically and practically challenging to conduct.
The hypothesis tested in this study was that Anopheles abundance would be higher in areas of the chimpanzees home range that have lower altitude and are more humid. Furthermore, if diseases transmitted by anophelines were to impact on chimpanzee health, then the apes would be more likely to select nesting sites away from the wetter areas and at levels with the lowest Anopheles abundance. Although the data presented here could be interpreted to support this hypothesis, the numbers of mosquitoes caught were low. Several recent studies emphasized the strong and synergistic effects of microclimate and altitude on malaria risk in human population, especially in highland sites [35, 53, 72]. In such areas, the valleys and basin-like depressions were recognized as less desirable areas to live, people living in the valleys receive more infective bites under such ambient conditions and the human density in these foci was relatively lower . In the present study, the number of Anopheles females caught varied with the altitude, temperature and hygrometry of the various sites where the traps were placed and the same patterns of choice for sleeping sites are observed in chimpanzees. Differences related to captures in the two close sites sampled (separated by less than 20km but characterized by different microclimates) are not surprising. Ernst et al. conducted a study in a 16km2 area where elevation ranged from 1,829 to 2,132m and showed striking magnitude of the differences even within this small area (up to 39-fold differences in incidence between the sub-unit areas of highest and lowest incidence). In our survey, the study in Kanyawara, the site of highest altitude, wetter atmosphere and cooler temperature, monitored during both rainy and dry seasons underlined the effects of climatic and spatial factors on the trap yields.
The low abundance of adult mosquitoes caught in the chimpanzees night environment might be in part explained by aspects of chimpanzee behaviour that lead to a reduction of exposure to malaria vectors. Chimpanzees are mobile and every evening they build a night nest in a new location within a large home range, about 20km2 in the present study-sites. In addition, even if chimpanzees are social, their population density is low (2.4 individuals/km2) and their system of fission-fusion prevents a high concentration of chimpanzees in a same nesting site at a same time. In Kibale NP, wet areas are usually found in the valley where swamps are frequent and footprints of elephants are very abundant. In human communities within a small area of 16km2, proximity to forest and swamp have both been associated with significant increased vector density: vector density has been shown to cluster in low-lying swampy areas . The present results indicated that chimpanzees build nests at a higher altitude than the sites where they feed, suggesting that they prefer nesting site above the wet valley. A topographic preference of chimpanzees for nesting on ridges and shoulders was also noted by Furuichi and Hashimoto . The negative correlation observed between nest number and mosquito abundance is consistent with these findings, although it cannot be exclude that mosquitoes might be more attracted by chimpanzees than by the traps. Nonetheless, the higher the nest site, the less diverse were the species of Anopheles encountered. There are many factors, such as predation pressure, body size or comfort, that have been proposed to influence the construction and selection of sleeping shelters and their height by great apes [74, 75]. It was even suggested that nest construction could explain the cognitive evolution of hominoids through long-term memory consolidation related to higher quality sleep . The present survey suggests that mosquito avoidance should be also considered as a factor in the selection of the nest location and height. This is especially relevant in Kibale NP where no predators exist and where chimpanzees are not hunted. Provided that the negative correlation between nest number/height and mosquito abundance does not reflect a higher attraction of mosquitoes to chimpanzees than to the traps, it would appear that chimpanzees select a nesting site where relatively few biting mosquitoes occurred. Moreover, mosquitoes are vectors of other diseases, including arboviral infections, that affect great apes and thus that can have significant impact on mortality and morbidity of wild primate populations .
From a conservation point of view, it would be of great importance to collect mosquitoes in the different sites where chimpanzees live. This is of particular importance with respect to degraded forest and at the edge of the park sites, where the vector species and their infections might differ from those collected in the middle of the park. Data records extending back to 1903 indicate that the Kibale region has become moister , which would likely lead to an increase in mosquito abundance. It is possible that chimpanzees, which had been adapted over thousands of years to forest vectors and parasites, might face novel threats to their health as changing climate and land conversion force then to live in fragmented forests and to use more frequently the forest edge.