In this study, reductions in the PCV of human blood of a magnitude consistent with severe anaemia were found to significantly reduce the amount of protein that malaria vectors acquire from blood feeding, and their resultant long-term survival. However mosquitoes feeding on blood of low PCV experienced no reduction in their oviposition rate, and contrary to initial prediction, exhibited increased fecundity of around 15% relative to those fed on blood of normal PCV. Although low PCV blood was associated with a reduction in the long-term survival of mosquitoes, this difference was not evident within the first 14 days after blood feeding, which approximates the minimal time period required for malaria parasites to complete their extrinsic development in mosquitoes. These results suggest that contact with human hosts whose PCV has been depleted due to severe anaemia may not significantly reduce the fitness or transmission potential of malaria vectors.
The observed reduction in blood meal size in mosquitoes feeding on blood of low PCV is consistent with previous studies conducted on rodent models [2, 17]. Specifically, anaemia in mice due to infection with Plasmodium chabaudi (which reduced their PCV by 20% in comparison to uninfected controls) was associated with a 25% reduction in the blood meal size of An. stephensi. A similar study of the rodent malaria parasite P. yoelii nigeriensis found that the blood meal size of An. stephensi fed on infected mice with low PCV (15–35%) and high parasitaemia was lower than those fed on infected mice with normal PCV (42–45%) . However, as reductions in host PCV in both these studies occurred as a consequence of malaria infection, the impacts of PCV and parasites themselves could not be disentangled. The results on An. gambiae s.s presented here demonstrate that PCV variation on its own can account for reductions in mosquito blood meal size that have been associated with feeding on malaria parasite-infected hosts.
The results presented here are in agreement with theoretical predictions that mosquito blood meal size may increase with more moderate levels of PCV reduction [15, 16], because reductions in red cell density reduce blood viscosity and thus the rate mosquitoes can imbibe blood. This advantage is predicted to be outweighed by the cost of the lower protein content (red cell density) of blood as its PCV falls [15, 16]. Reductions in PCV investigated here were substantial (PCV <15%, consistent with severe anaemia in humans), and resulted in smaller blood meal sizes, confirming the prediction that the advantages of increased blood flow rate with low PCV blood are outweighed by the cost of reduced red cell density at more extreme levels of anaemia (e g, PCV <15%).
In contrast to its influence on mosquito blood meal size, blood PCV had no impact on the oviposition rate of An. gambiae s.s., even after controlling for variation in mosquito wing length which was positively correlated with oviposition rate here and elsewhere [40, 41]. It is unclear why extreme reductions in blood PCV that significantly reduced mosquito blood meal size did not reduce their oviposition rate, as in previous studies these factors have been shown to be positively correlated [8, 40–42]. Possible reasons could be that the oviposition rate of mosquitoes fed on low PCV blood was enhanced by other chemical substrates in the blood serum such as free amino acids, lipids or other chemical elements. However, currently there is no evidence that blood serum factors might be important in triggering oviposition. Regardless of the mechanism, these results suggest that blood meal size may be an unreliable indicator of the mosquito oviposition rate when females feed from hosts with blood of variable PCV.
Also contrary to expectation on the basis of its impact on blood meal size, blood of low PCV was associated with higher mosquito fecundity. Of mosquitoes that laid at least one egg, those fed on low PCV blood laid approximately 15% more eggs than mosquitoes fed on blood of normal PCV. The finding of enhanced fecundity with low PCV blood is in contrast with previous studies, which have shown that as host PCV decreases due to malaria infection, both mosquito blood meal size and fecundity are reduced [42, 43]. The enhanced fecundity of mosquitoes fed on low PCV blood here cannot be due to its higher quality as indexed by protein intake from haemoglobin, as the haematin assay indicated that these blood meals had reduced protein levels. An alternative possibility is that mosquitoes fed on low PCV blood compensated for the lower resource quality of their blood meal by reallocating resources acquired during larval development towards enhanced short-term reproduction, at a cost to long-term survival. This hypothesis is drawn from life-history theory, which predicts that in maximizing their fitness, organisms face a trade-off between the allocation of resources to short-term versus long-term reproduction [44–46]. Trade-offs between survival and reproduction are predicted to be most extreme when resource availability is low . Studies on insects such as Drosophila, Ceratitis capitata, as well as Aedes and Anopheles mosquitoes, indicate that they can shift resource allocation to short-term reproductive increases, at the expense of longer term survival, when resources are low [47–50]. Further experimentation requires measurement of the total lifetime reproductive success of mosquitoes that are repeatedly fed on low PCV blood meals to confirm whether this apparent enhancement of fecundity is a short-term phenomenon due to resource re-allocation, or a consistent fitness advantage associated with low PCV blood. Such a re-allocation of resources may not completely compensate for the impact of poor blood meal quality (low PCV); eggs laid by mosquitoes fed on low PCV blood may have been provisioned with fewer maternal reserves and consequently have lower hatch rates than those produced from normal PCV hosts (not tested here).
Although variation in blood PCV had a significant impact on mosquito long-term survival, it did not affect An. gambiae s.s survival throughout the minimum period required for malaria parasite development (up to 14 days post feeding). This suggests that even extreme variation in host blood PCV may have limited impact on the vectorial capacity of An. gambiae s.s mosquitoes. However, although mosquito survival through the parasite’s extrinsic development period is the most crucial determinant of malaria transmission potential, any effect of PCV on mosquito longer term survival beyond this point could have some effect on transmission by reducing the number of biting opportunities mosquitoes have once they become infectious. Consequently, variation in host blood PCV may be expected to have a minor impact on malaria transmission in terms of mosquito long-term survival.
The relatively modest, although mostly statistically significant, differences in mosquito long-term survival associated with blood PCV reported here may be an underestimate of what occurs in the wild, where environmental conditions are harsher and more heterogeneous. Under the standardized insectary conditions used here, mosquitoes were given access to glucose ad libitum after blood feeding. Previous studies have shown that sugar feeding enhances mosquito survival [36, 47–50]. Thus the provision of glucose to mosquitoes in these experiments may have minimized the fitness costs of low PCV in blood by providing additional energetic resources to fuel survival. Furthermore, under the standardized laboratory conditions used here mosquitoes were exposed to few of the environmental stresses that would be encountered in nature (e.g., temperature and humidity fluctuation, high energetic demand of finding an oviposition site, predators), and thus the cost of resource limitation as experienced in nature may have been underestimated. Furthermore, although not considered in this work, PCV may also influence the frequency of interrupted feeding [16, 51]. Field-based experiments considering the impact of human host PCV under natural conditions will thus be required to confirm the role of host haematological factors on malaria vector fitness.
The impact of blood PCV on the energetic reserves of uninfected mosquitoes at death was found to be relatively minor. Lipid abundance at death was higher in mosquitoes fed on normal rather than low PCV blood, but glycogen and glucose levels were similar. Lipid levels are known to be the primary determinant of long-term survival , so the greater longevity of mosquitoes fed on normal PCV blood may be explained by this finding. The similarity of glycogen and glucose levels in mosquitoes fed on low and normal PCV blood is perhaps not surprising as these were measured at the time of natural death, when all of these resources may be similarly exhausted as result of general activity, or would have been replenished by ad libitum glucose. To more conclusively determine the impact of blood PCV on mosquito energetic reserves, experiments where resource levels are assayed at several points throughout the mosquito life time are needed. However, as reductions in blood PCV had few detrimental impacts on the mosquito fitness traits measured it may also be unlikely to have large impacts on mosquito energetic reserves.