Blood-meal host selection, malaria infection, and genotyping to delineate human to mosquito plasmodium transmission : implication for malaria parasite infections in malawi

Malaria is endemic in Malawi with continuing transmission and high rates of infection in the human population, despite ongoing interventions by the national malaria control program. The two main malaria control interventions in Malawi are long-lasting insecticidal treated nets (LLINs) and indoor residual spraying (IRS). The research described in this dissertation investigated Anopheles vector blood host selection among vertebrate hosts and within the human population, and from these investigations inferred pathways of malaria transmission from human to mosquito populations. The primary hypothesis was that biases in mosquito human feeding could drive malaria infection transmission. The study setting was two districts in southeastern Malawi. One district, Balaka, had conventional LLINs, those containing pyrethroid insecticide only; and the other, Machinga, had LLINs containing pyrethroids and the synergist piperonyl butoxide (PBO). Adult female Anopheles mosquitoes were sampled indoors in rural villages by CDC light traps, Pyrethrum spray catches and aspiration. Human demographic data and blood spots on filter paper by finger prick were collected from consented participants. Anopheles mosquito species were identified by morphological and molecular methods. Mosquito blood meal sources and Plasmodium falciparum infection in abdomen or head-thorax of mosquitoes were determined by qPCR with host-specific oligonucleotide probes and parasite specific probes, respectively. Blood meals from humans and human blood spots were analyzed by genotyping 24 microsatellite loci, to generate genetic profiles that were matched one to the other, and between blood meals and blood spots, using an algorithm executed in RStudio (Version 1.1.456). The results showed similar mosquito species community composition of the three main malaria vectors (An. arabiensis, An. gambiae, An. funestus), but different in abundance, with substantial malaria infection rates (ranging from 14-36%) in the mosquitoes at both sites. Potential secondary malaria vector mosquito species were present. Most mosquitoes fed upon humans (81%) and relatively less upon dogs (2%) and goats (5.2%). There were strong biases in human host selection, with a highly nonrandom and aggregated distribution of blood meals, with few individuals contributing most of the blood meals and most individuals contributing few. In human blood meals, males of the age group 6-15 were the most malaria infected group compared to younger and older age groups, and individuals of this group contributed most of the blood meals. These results incriminate males of age 6-15 as drivers of malaria infection from human to mosquitoes. The district with LLINs containing PBO had a lower Human Blood Index (HBI) and Entomological Inoculation Rate (EIR) compared to the district with conventional LLINs. Mixed blood meals of human and non-human hosts suggest interrupted feeding and the adaptability of the mosquitoes to switch amongst host species, possibly circumventing malaria control interventions. Malaria transmission was well sustained despite implementation of LLINs containing synergist PBO or not. Likely, other factors such as net durability and efficacy may be reducing effectiveness of the LLIN program and need further investigation. Continued efforts in malaria control and prevention should include deliberate efforts focused on key demographic population groups, such males of 6-15 years old, which while not the most vulnerable population to malaria, are reservoirs of infection.