Anopheles arabiensis (Diptera: Culicidae), second only to An. gambiae s.s. in its malarial vectorial capacity, is generally described as the 'less anthropophilic' or 'more opportunistic' and more 'exophagic' of the two species, particularly in eastern Africa [1–3]. This suggests that the presence of cattle within human settlements could divert malaria vectors away from humans, thereby reducing malaria transmission by passive zooprophylaxis. As tempting and logical as this proposition sounds, there are two important points to bear in mind.
First, if there is a high underlying entomological inoculation rate (EIR) of, say, 1 infective bite/person/night, then halving this to, say, 0.5 bites/person/night will not provide any material reduction in the incidence of malaria .
Second, terms such as anthropophily and exophagy do not describe specific behaviours per se, but, rather, are generally descriptive behavioural types inferred from the observed origin of bloodmeals. Drawing inferences about 'host preference' from information about bloodmeals alone can be confounded by the locations where mosquitoes are sampled. For example, in a study of An. arabiensis in southern Ethiopia, Tirados et al  found that in a village where the ratio of cattle:humans was ~ 1:1, the proportions of bloodmeals from humans was 51%. In a nearby cattle camp, however, where the ratio of cattle:humans was ~17:1, the percentage of human bloodmeals was very similar (46%). Thus, An. arabiensis seems to be 'opportunistic' in the village, but 'anthropophilic' in the cattle camp. The high proportion of human bloodmeals in the cattle camp begs the question: what is the behavioural mechanism that leads An. arabiensis to obtain such a biased feeding pattern?
Attempts to assess the inherent strength of response of An. arabiensis to human and cattle odours have generally concluded that this species is significantly more strongly attracted to human than cattle odours. In the study of Tirados et al , the numbers of An. arabiensis females caught in human odour-baited entry traps (OBETs ) were ~7.8 times greater than in cattle-baited OBETs, and in response to whole baits, human-landing catches caught about six times more than cattle-baited traps. Tirados et al  suggested that the paradoxical evidence that a so-called 'opportunistic' species appears to be highly anthropophilic in a cattle camp but not in a village could be explained as an interaction between a preference for feeding on humans and a preference for feeding outdoors. Accordingly, in the village context, the overall cattle:human ratio might be ~ 1:1, but with most humans indoors at night, the effective cattle:human ratio would be much higher. In the cattle camp, humans were always outdoors and hence available to mosquitoes, leading to a higher than expected percentage of meals from them.
A separate study, conducted on An. arabiensis in Zimbabwe, investigated possible interactions between behaviours associated with bloodfeeding: (1) the attraction to odour and (2) the 'entry' response associated with pursuit of host indoors . The study focussed explicitly on overcoming the problems of biases in trapping systems by using an arrangement of electrocuting nets to quantify the numbers of mosquitoes attracted to odours of cattle and/or humans dispensed outdoors or indoors. Outdoors, odour from a single human and a single ox attracted similar numbers of An. arabiensis. However, if these odours were dispensed indoors, then human odour caught significantly more An. arabiensis than ox odour. These results suggest that outdoors, odour from a single human or ox are equally attractive to An. arabiensis, but human odour elicits a stronger entry response than does cattle odour. The upshot of these findings is that proportion of bloodmeals from humans will depend on the type (cattle or human) and location (indoors or outdoors) of hosts.
Across the various spatial permutations of cattle and humans, two scenarios might be regarded as being most likely to result in a bias towards feeding from cattle: (1) a large herd of cattle that surrounds a single human outdoors, or (2) the same scenario but with the human indoors. Accordingly, the present study evaluated the impact of cattle barriers on the collection of two malaria vectors, An. arabiensis and An. pharoensis, from human hosts, indoors and outdoors. The main aim was to discover the degree to which herds might, under ideal circumstances, protect their herders given the underlying blood-feeding behaviour of the mosquito species.