In the current study, An. coustani complex showed susceptibility for the development of P. falciparum and P. vivax parasites in experimental infection. However, An. gambiae s.l. and An. pharoensis had a significantly higher infection rate than An. coustani complex which is considered as a suspected malaria vector in Ethiopia and all three species showed high mean oocysts density, with the highest oocyst density found in An. gambiae s.l., which is one of the main malaria vector species in sub-Saharan Africa. Studies in some countries under natural conditions indicate that An. coustani complex supports the development of both P. falciparum and P. vivax [33, 34]. In contrast to what we are reporting here, similar experimental studies conducted in Jimma town of southwest Ethiopia indicated that An. coustani complex is not susceptible to P. vivax development . This discrepancy might be due to larval breading site variability and the presence of different member species of An. coustani complex i.e., there might be different sibling species present in different parts of the country.
Determining the susceptibility of Anopheline mosquito’ species to malaria parasites in time and space is important in vector control programs [31–33]. In this study, it is documented that there is a difference in susceptibility among the three species. The rate of mosquito infection is affected by various factors such as environmental, biological as well as behavioral [19, 35, 36]. An. coustani complex was known as suspected vector in Ethiopia [33, 34]. Immunological studies using wild catch mosquitoes of An. coustani in different countries in sub-Sharan Africa have shown its susceptibility to P. falciparum and P. vivax infection [33–40].
The susceptibility of An. gambiae s.l., and An. pharoensis to malaria parasites P. vivax and P. falciparum infection has been long established  and growing evidence including this study confirm the susceptibility of An. coustani complex to both P. vivax and P. falciparum from Ethiopia, Kenya, Zambia, DRC, and Cameroon [18, 33, 36, 38, 42], field populations were susceptible, 33, 34, 43].
Vector competency varies from species to species and usually there is high proportion of uninfected mosquitoes when the infection process is controlled. This is confirmed in studies conducted in Ethiopia and elsewhere in world [32, 44, 45]. On the other hand, infection rates were much higher in experimental studies than those reported in natural infection for the three species examined. As determined by ELISA technique, based on the use of species-specific anti-sporozoite monoclonal antibodies, mosquito populations had different infection rates in sub-Saharan countries. For instance, the infection rate of An. gambiae s.l ranged from 0.3 to 9.3%; An. pharoensis from 0.4 to 5.2%; and An. coustani complex from 0.3 to 1.81% [18, 33, 46].
In this study, gametocyte density was found to be significantly positively correlated with infection rates. Irrespective of the differences in mosquito and parasite species, in our findings also, Plasmodium gametocyte density was significantly correlated to infection rates and mean oocysts. This is also true in other studies conducted in Manaus, in the western Brazilian Amazon and Bengbu, Anhui Province, central China [44, 45].
In this study, infected blood with equivalent gametocyte density was provided to mosquitoes’ species, but variable susceptibility was observed. This discrepancy in infection rate might be due to different factors, such as, gametocyte maturity, gametocyte sex ratio, different Plasmodium genotypes, rhythms in the density and infectivity of transmission forms (gametocytes), immune factors in patient sera, mosquitoes’ innate immunity, all of which could alter gametocyte infectivity [47–50]. In addition, recent studies indicate that sub-microscopic gametocyte density is capable of infecting mosquitoes; this shows that rather than gametocyte density, the above contributing factors play a role in the variability of infection rate and mean oocyst. Studies from different countries show that sub-microscopic infections might be the major contributor to malaria transmission [51–54]. This shows that gametocyte density is not the only factor in mosquitoes’ infection.
Furthermore, secondary and suspected vectors such as An. pharoensis and An. coustani complex have relatively higher abundance during the dry season . This might have importance in local malaria transmission, as they may help to increase or prolong the malaria transmission period . Mostly secondary vectors are often outdoor biting and outdoor resting [46, 56]. The role of outdoor-resting anopheles’ mosquitoes in malaria transmission is important, secondary vectors are vectors which contribute to malaria transmission in Africa, and their role in transmission is not negligible [56, 57]. Most secondary vectors have a short survival rate with 50–60% natural mortality rates per gonotrophic cycle [23, 24]. This might explain the reason, why population density of the secondary vector has never been high in many settings.
Most anopheles vectors found in nature have only a few oocysts and oocysts have little importance in malaria epidemiology; rather, it is the outdoor biting and resting behavior in nature that contribute to residual malaria transmission in many parts of sub-Saharan Africa and pose new challenges as they cannot be reliably monitored or controlled using conventional tools and outdoor biting proportion increased by 10% [58–60]. In more recent studies in sub-Saharan Africa, An. pharoensis and An. coustani complex likely became primary role in malaria transmission; this might be as a result of existing interventions targeting primary vectors to achieve complete malaria control .