Underestimation of foraging behaviour by standard field methods in malaria vector mosquitoes in southern Africa

Smita Das, Tyler C Henning, Limonty Simubali, Harry Hamapumbu, Lukwa Nzira, Edmore Mamini, Aramu Makuwaza, Mbanga Muleba, Douglas E Norris, Jennifer C Stevenson

Defining the anopheline mosquito vectors and their foraging behaviour in malaria endemic areas is crucial for disease control and surveillance. The standard protocol for molecular identification of host blood meals in mosquitoes is to morphologically identify fed mosquitoes and then perform polymerase chain reaction (PCR), precipitin tests, or ELISA assays. The purpose of this study was to determine the extent to which the feeding rate and human blood indices (HBIs) of malaria vectors were underestimated when molecular confirmation by PCR was performed on both visually fed and unfed mosquitoes. In association with the Southern Africa International Centers of Excellence in Malaria Research (ICEMR), mosquito collections were performed at three sites: Choma district in southern Zambia, Nchelenge district in northern Zambia, and Mutasa district in eastern Zimbabwe. All anophelines were classified visually as fed or unfed, and tested for blood meal species using PCR methods. The HBIs of visually fed mosquitoes were compared to the HBIs of overall PCR confirmed fed mosquitoes by Pearson's Chi-Square Test of Independence. The mosquito collections consisted of Anopheles arabiensis from Choma, Anopheles funestus s.s., Anopheles gambiae s.s. and Anopheles leesoni from Nchelenge, and An. funestus s.s. and An. leesoni from Mutasa. The malaria vectors at all three sites had large human blood indices (HBI) suggesting high anthropophily. When only visually fed mosquitoes tested by PCR for blood meal species were compared to testing those classified as both visually fed and unfed mosquitoes, it was found that the proportion blooded was underestimated by up to 18.7%. For most Anopheles species at each site, there was a statistically significant relationship (P < 0.05) between the HBIs of visually fed mosquitoes and that of the overall PCR confirmed fed mosquitoes. The impact on HBI of analysing both visually fed and unfed mosquitoes varied from site to site. This discrepancy may be due to partial blood feeding behaviour by mosquitoes, digestion of blood meals, sample condition, and/or expertise of entomology field staff. It is important to perform molecular testing on all mosquitoes to accurately characterize vector feeding behaviour and develop interventions in malaria endemic areas.