Little is known about the Anopheles species of the coastal areas of French Guiana, or their spatiotemporal distribution or environmental determinants. The present study aimed to (1) document the distribution of Anopheles fauna in the coastal area around Cayenne, and (2) investigate the use of remotely sensed land cover data as proxies of Anopheles presence. To characterise the Anopheles fauna, we combined the findings of two entomological surveys that were conducted during the period 2007-2009 and in 2014 at 37 sites. Satellite imagery data were processed to extract land cover variables potentially related to Anopheles ecology. Based on these data, a methodology was formed to estimate a statistical predictive model of the spatial-seasonal variations in the presence of Anopheles in the Cayenne region. Two Anopheles species, known as main malaria vectors in South America, were identified, including the more dominant An. aquasalis near town and rural sites, and An. darlingi only found in inland sites. Furthermore, a cross-validated model of An. aquasalis presence that integrated marsh and forest surface area was extrapolated to generate predictive maps. The present study supports the use of satellite imagery by health authorities for the surveillance of malaria vectors and planning of control strategies.

In French Guiana, malaria vector control and prevention relies on indoor residual spraying and distribution of long lasting insecticidal nets. These measures are based on solid epidemiological evidence but reveal a poor understanding of the vector. The current study investigated the behaviour of both vectors and humans in relation to the ongoing prevention strategies. In 2012 and 2013, Anopheles mosquitoes were sampled outdoors at different seasons and in various time slots. The collected mosquitoes were identified and screened for Plasmodium infection. Data on human behaviour and malaria episodes were obtained from an interview. A total of 3,135 Anopheles mosquitoes were collected, of which Anopheles darlingi was the predominant species (96.2%). For the December 2012-February 2013 period, the Plasmodium vivax infection rate for An. darlingi was 7.8%, and the entomological inoculation rate was 35.7 infective bites per person per three-month span. In spite of high bednet usage (95.7%) in 2012 and 2013, 52.2% and 37.0% of the participants, respectively, had at least one malaria episode. An. darlingi displayed heterogeneous biting behaviour that peaked between 20:30 and 22:30; however, 27.6% of the inhabitants were not yet protected by bednets by 21:30. The use of additional individual and collective protective measures is required to limit exposure to infective mosquito bites and reduce vector densities.