Pathogen-induced host phenotypic changes are widespread phenomena that can dramatically influence host-vector interactions. Enhanced vector attraction to infected hosts has been reported in a variety of host-pathogen systems, and has given rise to the parasite manipulation hypothesis whereby pathogens may adaptively modify host phenotypes to increase transmission from host to host. However, host phenotypic changes do not always favour the transmission of pathogens, as random host choice, reduced host attractiveness and even host avoidance after infection have also been reported. Thus, the effects of hosts' parasitic infections on vector feeding behaviour and on the likelihood of parasite transmission remain unclear. Here, we experimentally tested how host infection status and infection intensity with avian Plasmodium affect mosquito feeding patterns in house sparrows (Passer domesticus). In separate experiments, mosquitoes were allowed to bite pairs containing (i) one infected and one uninfected bird and (ii) two infected birds, one of which treated with the antimalarial drug, primaquine. We found that mosquitoes fed randomly when exposed to both infected and uninfected birds. However, when mosquitoes were exposed only to infected individuals, they preferred to bite the non-treated birds. These results suggest that the malarial parasite load rather than the infection itself plays a key role in mosquito attraction. Our findings partially support the parasite manipulation hypothesis, which probably operates via a reduction in defensive behaviour, and highlights the importance of considering parasite load in studies on host-vector-pathogen interactions.

Mosquito host utilization is a key factor in the transmission of vector‐borne pathogens given that it greatly influences host–vector contact rates. Blood‐feeding patterns of mosquitoes are not random, as some mosquitoes feed on particular species and/or individuals more than expected by chance. Mosquitoes use a number of cues including visual, olfactory, acoustic, and thermal stimuli emitted by vertebrate hosts to locate and identify their blood meal sources. Thus, differences in the quality/intensity of the released cues may drive host selection by mosquitoes at both inter‐ and intra‐specific levels. Such patterns of host selection by mosquitoes in space and time can be structured by factors related to mosquitoes (e.g. innate host preference, behavioural plasticity), to hosts (e.g. emission of host‐seeking cues, host availability) or to both (e.g. pathogen infection). In this study, we review current evidence, from phenomena to mechanisms, of how these factors influence host utilization by mosquitoes. We also review the methodologies commonly used in this research field and identify the major challenges for future studies. To bridge the knowledge gaps, we propose improvements to strengthen traditional approaches and the use of a functional trait‐based approach to infer mosquito host utilization in natural communities.

BackgroundThe life-cycle of many vector-borne pathogens includes an asexual replication phase in the vertebrate host and sexual reproduction in the insect vector. However, as only a small array of parasites can successfully develop infective phases inside an insect, few insect species are competent vectors for these pathogens. Molecular approaches have identified the potential insect vectors of blood parasites under natural conditions. However, the effectiveness of this methodology for verifying mosquito competence in the transmission of avian malaria parasites and related haemosporidians is still under debate. This is mainly because positive amplifications of parasite DNA in mosquitoes can be obtained not only from sporozoites, the infective phase of the malaria parasites that migrate to salivary glands, but also from different non-infective parasite forms in the body of the vector. Here, we assessed the vectorial capacity of the common mosquito Culex pipiens in the transmission of two parasite genera.MethodsA total of 1,560 mosquitoes were allowed to feed on five house sparrows Passer domesticus naturally infected by Haemoproteus or co-infected by Haemoproteus/Plasmodium. A saliva sample of the mosquitoes that survived after 13 days post-exposure was taken to determine the presence of parasite DNA by PCR.ResultsOverall, 31.2% mosquito’s head-thorax and 5.8% saliva samples analysed showed positive amplifications for avian malaria parasites. In contrast to Haemoproteus DNA, which was not found in either the body parts or the saliva, Plasmodium DNA was detected in both the head-thorax and the saliva of mosquitoes. Parasites isolated from mosquitoes feeding on the same bird corresponded to the same Plasmodium lineage.ConclusionsOur experiment provides good evidence for the competence of Cx. pipiens in the transmission of Plasmodium but not of Haemoproteus. Molecular analyses of saliva are an effective method for testing the vector competence of mosquitoes and other insects in the transmission of vector-borne pathogens.