Human malaria, caused by an infection with
Plasmodium parasites and transmitted by
Anopheles mosquitoes, is one of the greatest causes of mortality in the world. Nearly half-a-million people die from malaria each year, with the greatest burden (> 90%) of morbidity and mortality occurring in sub-Saharan Africa [
1]. Malaria control efforts have increased in recent years, with a resurgence in investment in research and practical prevention efforts such as targeted vector control strategies. These targeted vector control strategies use insecticide-treated bed nets (ITNs), as well as indoor residual spraying (IRS) to reduce vector populations and to protect humans from malaria transmission via mosquito bites [
2]. The overwhelming focus on insecticides, in particular with ITNs and IRS, for malaria prevention over the past several decades has led to shifts in mosquito susceptibility to these insecticides [
3‐
7], and may lead to facultative or long-term shifts in the behaviour of vector species [
8‐
12]. Most populations of the primary malaria vector species in Africa,
Anopheles gambiae sensu stricto (
s.s.) and
Anopheles funestus s.s., are now at least partially resistant to most pyrethroid insecticides [
11,
12], while development of resistance to other commonly used classes of insecticides, such as carbamates and organophosphates, is ongoing [
11,
13]. Certain populations of
An. gambiae sensu lato (
s.l.) and
An. funestus s.l. also seem to have undergone behavioural changes following implementation of ITN campaigns, switching to feeding at times when people are less likely to be under bed nets [
4,
14,
15] as well as displaying increased levels of exophagy, avoiding contact with insecticide while seeking a host [
4,
6,
7]. In regions where
Anopheles vectors are becoming increasingly exophagic, they also appear to be becoming increasingly zoophagic [
3,
6,
7,
16,
17], supporting research which indicates that extrinsic factors such as host availability influence host blood meal preference, although methodological variables, such as indoor/outdoor trapping location may influence the type of blood meal detected [
18,
19]. While measures such as ITNs are effective in reducing the bulk of human malaria cases caused by endophagic mosquitoes, residual transmission, facilitated by exophagic mosquitoes, remains an issue of concern [
6,
16,
20‐
23]. In order to achieve long-term malaria control and to fight residual transmission, it is critical to consider mosquito ecology and to integrate alternative vector control strategies with the methods currently used [
20,
24,
25].
The island nation of Madagascar represents an area of ongoing residual malaria transmission where alternative strategies may need to be integrated with current approaches to combat transmission. One hundred percent of Madagascar’s population lives in areas where malaria is endemic [
26]. Despite successful control strategies in Madagascar in recent decades, the impact of ongoing efforts to distribute ITNs, as well as widespread application of IRS, has not been enough to halt the transmission of malaria-causing parasites. This is evident in the extensive outbreaks seen in recent years [
27]. The continued malaria burden and impact on mortality in Madagascar suggests current vector control strategies need to be revised and improved. The high ratio of cattle to humans could create potential for a zooprophylactic approach, one in which livestock are used to divert malaria vectors from human populations. The success of such an approach depends on the importance of various vector species and their host feeding preferences.
The goals of this study were to: (1) determine community compositions of host-seeking malaria vector species in southeastern Madagascar, (2) determine host preference of
Anopheles spp. collected in this area, where habitat alteration is common and livestock are prevalent, and (3) determine the proportion of
Anopheles spp. in this area harbouring infective sporozoite stages of
Plasmodium parasites
Plasmodium falciparum,
Plasmodium vivax VK210 (
P. vivax 210) and
P. vivax VK247 (
P. vivax 247). Due to the high level of zoophagy demonstrated by
Anopheles in other parts of sub-Saharan Africa where livestock are common, it was hypothesized that in this study area, mosquito species which are typically considered to be anthropophagic, such as
An. gambiae s.l. and
An. funestus s.l., would exhibit feeding preference toward livestock, such as cattle. It was further hypothesized that typically zoophagic species, such as
Anopheles coustani and
Anopheles squamosus/cydippis, would maintain preference for feeding on livestock. Additionally, it was hypothesized that these zoophagic vectors would exhibit
Plasmodium sporozoite rates similar to those of the primary vectors,
An. gambiae s.s. and
An. funestus s.s. Finally, based on reported case data, it was hypothesized that
P. vivax would be present, but its prevalence in the vector population would be lower than that of
P. falciparum [
26].