Effects of host sex, body mass and infection by avian Plasmodium on the biting rate of two mosquito species with different feeding preferences

The blood-feeding behaviour of mosquitoes is a complex phenomenon that involves different steps. The initial seeking and location of hosts depends on the integration of chemical (e.g. CO₂, odours) and visual cues (e.g. host size and plumage/pelage coloration) emitted by the host [1, 2]. In close proximity between mosquitoes and their hosts, odour, heat and host defensive behaviour may affect the final host choice and blood-feeding success of mosquitoes [3].

Under natural conditions, mosquitoes show different innate feeding preferences, with some species feeding mostly on mammals (mammophilic species, and some of them can be characterized as anthropophilic), while others preferring to bite birds (ornithophilic species), or even amphibians or reptiles, yet other species show a more opportunistic behaviour [4–8]. In addition to this broad tendency for particular host classes, mosquitoes bite certain host species at higher rates than those expected from their abundance [9–11]. For instance, Kilpatrick et al. [9] showed that American robins (Turdus migratorius) were more intensely bitten by Culex pipiens mosquitoes than European starlings (Sturnus vulgaris) in North America. Similarly, in Europe, the feeding preference of Cx. pipiens for blackbirds (Turdus merula) was higher than for European starlings [12]. Within host species, some individuals may receive most mosquito bites and, as a result, they may play a role as superspreaders when infected with vector-borne pathogens [13].

This heterogeneity in vector attraction and host use by mosquitoes could have important impacts on the dynamics of transmission of parasites causing human and animal diseases [14]. Among others, these factors may determine the transmission success of parasites such as protozoans (e.g. Plasmodium spp.) and filarial worms (e.g. Dirofilaria spp.) [2, 15].

Different non-mutually exclusive mechanisms may determine that an individual host receives more mosquito bites, such as, for instance, the use of habitats with higher abundance of mosquitoes, a higher emission of attractive cues, or a less intense or effective anti-mosquito behaviour than other conspecifics [7]. In addition, larger hosts (i.e. with a larger body mass, as a correlate of body size), may receive more bites by mosquitoes [16] probably due to the higher amounts of cues (e.g. CO₂) released by larger individuals [17]. Different studies at the interspecific level have reported a positive relationship between host body mass and the feeding rate of different blood-sucking insects [18, 19]. However, very few studies have experimentally tested the relationship between species variation in body mass and the feeding rate of mosquitoes [20]. In addition, sex-specific morphological, physiological and/or behavioural characteristics could produce differences in the attraction of insect vectors [21]. These differences in vector attraction between host sexes have been argued as a potential explanation for the usually higher prevalence of blood parasites found in male than in female birds [22–24]. However, to the best of our knowledge, only Burkett-Cadena et al. [25] evaluated the effect of bird sex on the variation in mosquito biting preferences. By analysing the blood-meal origin of mosquitoes, authors found that blood meals were biased towards male birds, but only in mammophilic mosquitoes. However, the reasons of these differences remain unclear. Patterns found by Burkett-Cadena et al. [25] could be the result of a differential susceptibility, attraction and/or exposure of bird sexes to mosquito attacks or simply an unbalanced bird sex-ratio in the areas where mosquitoes were captured. Finally, the host infection status by vector-borne parasites may also influence the mosquito biting patterns, potentially determining the pathogen transmission success [26, 27]. For example, humans infected with Plasmodium vivax were more attractive to mosquito vectors [28]. However, studies with avian Plasmodium are less conclusive, because Cx. pipiens, the main vector of avian Plasmodium, was reported to preferentially bite chronically infected birds over uninfected individuals according to Cornet et al. [26, 27], but other studies have reported the opposite pattern [29] or even the absence of significant differences between infected and uninfected birds [30].

In this study, we experimentally assessed the impact of three host-related traits (body mass, sex and infection status by avian Plasmodium) on mosquito feeding patterns, while removing host anti-mosquito behaviour. We performed this study using two mosquito species with different feeding preferences: the ornitophilic Cx. pipiens and the mammophilic Ochlerotatus (Aedes) caspius [8, 31, 32]. We used two bird species as host models, the jackdaw (Corvus monedula) and the house sparrow (Passer domesticus). Both bird species are common hosts of avian malaria parasites [33, 34]. Based on previous evidence [19, 25–27], we predicted (i) a higher biting rate on birds in the ornithophilic Cx. pipiens than in the mammophilic Oc. caspius; (ii) a higher mosquito biting rate on heavier individuals; (iii) a higher biting rate on male birds over females, especially for Oc. caspius; and (iv) a higher biting rate on Plasmodium-infected birds than on uninfected individuals.

Identifying the potential causes underlying the non-random biting patterns of mosquitoes is essential to understand the dynamics of transmission of avian Plasmodium and other vector-borne pathogens [13]. Here, we tested how three important avian traits (i.e. body mass, sex and the infection status by Plasmodium) in two bird species affect the biting rates of two mosquito species potentially involved in the transmission of avian malaria parasites [32]. Nonetheless, other factors not quantified in this study, such as the size of the blood meal, could also be important in the parasite transmission success and should be therefore considered in future studies.

The biting rate of the mammophilic Oc. caspius on jackdaws was higher than the biting rate of the ornithophilic Cx. pipiens, while non-significant differences were found when mosquitoes faced house sparrows. Although most of the blood meals of Oc. caspius analysed in different studies was derived from mammals, birds including chickens and house sparrows represented between 9.1–19.9% of the blood meals in this species [6, 49]. This pattern clearly contrasts with that of Cx. pipiens, for which birds represent between 85.1–91.67% of the blood meals [6]. In this study, the Cx. pipiens biotypes were not considered although potential differences in the feeding preferences have been reported [50]. However, mosquitoes in this study were captured in an area where the two biotypes (Cx. pipiens pipiens and Cx. pipiens molestus) and their hybrids coexist. A previous study conducted in southern Spain did not find significant differences in the proportion of blood meals from birds between Culex pipiens biotypes [45]. Our results clearly support the ability of Oc. caspius to feed on birds, at least when they are not allowed to choose between other host classes (i.e. mammals). This fact is especially relevant as our study focuses on the biting rate of mosquitoes and not on the feeding preferences of these species. Contrary to Cx. pipiens, Oc. caspius is traditionally considered as an aggressive mosquito and an important nuisance to human populations [51], but experimental studies supporting this assumption are scarce. Differences in the biting rate between mosquito species could be associated with their life history traits and breeding requirements, especially those related to the availability of water sources. While Oc. caspius depends on tidal cycles and uses temporal flooded areas for larval development [52], Cx. pipiens uses more permanent water sources [53] and, consequently, their life-cycle may be less time-constrained. In addition, it is possible that a differential activity pattern between mosquito species could have affected our results, with Oc. caspius showing a strong peak of activity during the day while Cx. pipiens peaks its activity at night and sunset [49]. Although this possibility could potentially explain the differences found between species, the biting rates of Cx. pipiens found in this study are similar to those found in a previous experiment developed during the night [30]. The variation in climatic conditions throughout the year is expected to affect the phenology of both the host and the vector and thus may potentially influence the host-vector interactions. We found that mosquito biting rates increased over time (i.e. from spring to autumn), but only when exposed to house sparrows. In this regard, Edman [54] reported differences in the feeding patterns of Culex nigripalpus depending on the season. In our case, this effect seems to be associated to particular host-mosquito assemblages, since it was apparent in only one of the two bird species studied. Nevertheless, it is important to highlight that the relevance of this variable should be considered low, as the model including the significant effect of the date explained only 7% of the variance.

The fact that differences in biting patterns were only detected when mosquitoes faced jackdaws, the larger host species, suggests these may be related to differences in the amounts of cues emitted by each bird species. In close proximity to their hosts, the relevance of visual and thermo-sensory stimulation of mosquitoes increases with respect to larger distances. Moreover, the use of multiple sensory cues may increase the likelihood of mosquito feeding success [3]. Due to their larger size, jackdaws may emit a higher amount of attractants, including CO₂, heat and odours, than house sparrows, potentially leading to the differences found here.

Previous studies have reported a positive relationship between host body mass and biting rates of blood-sucking insects [19, 55], although this pattern usually corresponds to studies comparing different host species. As expected from its larger size, jackdaws were bitten at a higher rate than house sparrows by Oc. caspius, although we did not find significant differences for Cx. pipiens. The amount of bare skin exposed, usually positively correlated with body mass, may affect mosquito feeding success [56]. In fact, body mass was a key variable explaining the prevalence of infection by the mosquito-borne pathogen West Nile virus in birds [57]. Further support for this possibility can be derived from the study by Yan et al. [58], who found that Cx. pipiens fed more often on birds with longer tarsus, suggesting that larger areas of exposed skin are important for determining biting patterns. In this respect, Burkett-Cadena et al. [39] found that the proportion of bites on nestlings with respect to the adult female in the nest increased as they grow, suggesting that size in addition to the exposed skin may be key factors influencing mosquito feeding patterns. However, we did not find any significant relationship between the biting rate of mosquitoes and bird body mass at the intraspecific level. In this regard, Lalubin et al. [29] found that the attraction of Cx. pipiens to house sparrows was not significantly associated with their body mass. This suggests that, at short distances, the slight intraspecific differences in body mass are probably less important than other cues determining mosquito bites, like heat, humidity or odour [3].

As predicted, host sex influenced the biting rates of Oc. caspius mosquitoes when facing jackdaws. However, and in contrast to our prediction, this mosquito species preferred to bite females than males, but these differences were not found when mosquitoes were exposed to house sparrows. The biting rates of Cx. pipiens were not significantly related to host sex. Burkett-Cadena et al. [25] found male-biased blood meals in mosquitoes (64% of the blood meals analysed derived from male birds), although they suggest this could be due to skewed sex ratios in wild birds. However, there are not significant sex-biased differences in the feeding patterns of bird-biting mosquitoes, including Culex species [25]. Moreover, Simpson et al. [20] concluded that bird sex has no effect on the probability of Cx. pipiens choosing an individual over its partner. The preference of mammophilic mosquitoes for a particular sex of bird could be associated with the sexual differences in the composition of odour profiles. Among other factors, the volatile and non-volatile substances of the secretions of the preen gland may affect the feeding preferences of blood-sucking insects [59, 60] and their composition differ between bird sexes [61, 62]. Differences in the response of Oc. caspius and Cx. pipiens to secretions of the preen gland could explain, at least in part, discrepancies found between mosquito species [63].

This study highlights that the magnitude and direction of the effects of host traits such as body mass, sex or the infection status by the mosquito-borne avian Plasmodium on the feeding patterns of mosquitoes are far from being generalizable. Only sex was associated to differences in mosquito biting rates, and this effect was only detected for one of the mosquito species studied here. Consequently, the biting patterns of mosquitoes may differ according to vector and host species characteristics. The reasons underlying the preference of mammophilic mosquitoes for individuals of a particular sex are unclear and need detailed analyses with regard, for instance, to the olfactory cues released by male and female birds.