Must epidemiologically impactful vector control interventions disrupt mosquito population structure? A case study of a cluster-randomised controlled trial.

Abstract

Large epidemiological impacts resulting from disease vector control interventions are typically associated with significant disruption of vector populations. While vector density is a frequently measured response, impacts on demography and connectivity are suspected but rarely quantified. We analysed low-coverage whole-genome sequence data of 893 Anopheles gambiae mosquitoes collected between 2014-2015 during a cluster-randomized control trial (cRCT) in Burkina Faso to compare a pyrethroid-only net (ITN) with a pyrethroid-pyriproxyfen (ITN-PPF) net. Despite reductions of clinical malaria by 12%, and vector density by 22% in the ITN-PPF arm, we found no significant changes in An. gambiae population genetic structure or diversity. We found remarkably low population differentiation, and a lack of discernible clustering by treatment, time, or space. Nucleotide diversity and inbreeding coefficient remained stable between treatments, and genome-wide scans showed no putative signatures of selection between trial arms. These results show that ITN-PPF did not alter An. gambiae genetic structure, possibly due to large, vagile populations in West Africa. More widely, this is first evidence that epidemiologically meaningful reductions in vector density may not impact genetic diversity or connectivity, and challenges what constitutes adequate vector control in large populations.