Vector composition, abundance, biting patterns and malaria transmission intensity in Madang, Papua New Guinea: assessment after 7 years of an LLIN-based malaria control programme

BACKGROUND: A malaria control programme based on distribution of long-lasting insecticidal bed nets (LLINs) and artemisinin combination therapy began in Papua New Guinea in 2009. After implementation of the programme, substantial reductions in vector abundance and malaria transmission intensity occurred. The research reported here investigated whether these reductions remained after seven years of sustained effort. METHODS: All-night (18:00 to 06:00) mosquito collections were conducted using human landing catches and barrier screen methods in four villages of Madang Province between September 2016 and March 2017. Anopheles species identification and sporozoite infection with Plasmodium vivax and Plasmodium falciparum were determined with molecular methods. Vector composition was expressed as the relative proportion of different species in villages, and vector abundance was quantified as the number of mosquitoes per barrier screen-night and per person-night. Transmission intensity was quantified as the number of sporozoite-infective vector bites per person-night. RESULTS: Five Anopheles species were present, but vector composition varied greatly among villages. Anopheles koliensis, a strongly anthropophilic species was the most prevalent in Bulal, Matukar and Wasab villages, constituting 63.7-73.8% of all Anopheles, but in Megiar Anopheles farauti was the most prevalent species (97.6%). Vector abundance varied among villages (ranging from 2.8 to 72.3 Anopheles per screen-night and 2.2-31.1 Anopheles per person-night), and spatially within villages. Malaria transmission intensity varied among the villages, with values ranging from 0.03 to 0.5 infective Anopheles bites per person-night. Most (54.1-75.1%) of the Anopheles bites occurred outdoors, with a substantial proportion (25.5-50.8%) occurring before 22:00. CONCLUSION: The estimates of vector abundance and transmission intensity in the current study were comparable to or higher than estimates in the same villages in 2010-2012, indicating impeded programme effectiveness. Outdoor and early biting behaviours of vectors are some of the likely explanatory factors. Heterogeneity in vector composition, abundance and distribution among and within villages challenge malaria control programmes and must be considered when planning them.