Diversity, composition, altitude, and seasonality of high-altitude windborne migrating mosquitoes in the Sahel: Implications for disease transmission

Abstract

Recent studies have reported Anopheles mosquitoes captured at high-altitude (40-290 m above ground) in the Sahel. Here, we describe this migration modality across genera and species of African Culicidae and examine its implications for disease transmission and control. As well as Anopheles, six other genera-Culex, Aedes, Mansonia, Mimomyia, Lutzia, and Eretmapodites comprised 90% of the 2,340 mosquitoes captured at altitude. Of the 50 molecularly confirmed species (N = 2,107), 33 species represented by multiple specimens were conservatively considered high-altitude windborne migrants, suggesting it is a common migration modality in mosquitoes (31-47% of the known species in Mali), and especially in Culex (45-59%). Overall species abundance varied between 2 and 710 specimens/species (in Ae. vittatus and Cx. perexiguus, respectively). At altitude, females outnumbered males 6:1, and 93% of the females have taken at least one blood meal on a vertebrate host prior to their departure. Most taxa were more common at higher sampling altitudes, indicating that total abundance and diversity are underestimated. High-altitude flight activity was concentrated between June and November coinciding with availability of surface waters and peak disease transmission by mosquitoes. These hallmarks of windborne mosquito migration bolster their role as carriers of mosquito-borne pathogens (MBPs). Screening 921 mosquitoes using pan-Plasmodium assays revealed that thoracic infection rate in these high-altitude migrants was 2.4%, providing a proof of concept that vertebrate pathogens are transported by windborne mosquitoes at altitude. Fourteen of the 33 windborne mosquito species had been reported as vectors to 25 MBPs in West Africa, which represent 32% of the MBPs known in that region and include those that inflict the heaviest burden on human and animal health, such as malaria, yellow fever, dengue, and Rift Valley fever. We highlight five arboviruses that are most likely affected by windborne mosquitoes in West Africa: Rift Valley fever, O'nyong'nyong, Ngari, Pangola, and Ndumu. We conclude that the study of windborne spread of diseases by migrating insects and the development of surveillance to map the sources, routes, and destinations of vectors and pathogens is key to understand, predict, and mitigate existing and new threats of public health.

Keywords: Africa; arbovirus; disease-spread; dispersal; malaria; mosquito-borne pathogen; one health; surveillance.

Figure 1

Mosquito composition at altitude (40–290 m above ground) in the Mali Sahel (2013–2015) by genera (color) for species represented by two or more specimens in the collection (text). The number of specimens identified using mtCOI DNA barcode similarity is shown above bars (note: logarithmic scale of the Y-axis). The fraction of species found at altitude per genus from the corresponding total known in Mali (see text) are given for each genus above bars and tested against the overall fraction of 28.3% (30/106) using a binomial test (ns and * denote P > 0.05 and P < 0.05, respectively). The pie chart shows composition of specimens by genera including specimens that were identified by molecular methods (see text).

Figure 2

Sex imbalance in mosquitoes at altitude and composition of female's gonotrophic states by species (note: taxa with < 4 females of known gonotrophic state are not shown). The expected equal proportion of females and males is denoted by the broken blue line. Note that except unfed females (gray fill color), all other gonotrophic states must have taken at least one blood meal from a vertebrate host prior to their high-altitude flight and hence could be harboring pathogens from that exposure (see text). Numbers on top (black) and bottom (white) of the bars denote available sample sizes to calculate sex and gonotrophic state composition, respectively.

Figure 3

Relative flight altitude (A) and seasonality (B) of high-altitude flying mosquitoes over three years (2013–2015) across the Sahelian aerial stations. Note: species order varies between panels. (A) The relative mean flight height of mosquito species (including only those whose N ≥ 5) weighted by the ratio of the proportion of mosquitoes collected on each panel height divided by the proportion of nights this panel height was available and their 95% confidence interval (CI). Overall mean panel altitude is 150 m (denoted by broken blue line). Sample size of each species is given above the mean. (B) Seasonality of high-altitude flying Sahelian mosquitoes, showing the period of flight of each species (blue frames), and aerial abundance (filled color intensity), measured as the logarithm of mean panel density * 1,000 (see text). Note: Sampling was not performed in January–February.

Figure 4

Risk of disease spread by windborne mosquitoes. A heatmap showing pathogen (Y-axis, pathogen acronym, see Supplementary Table S3) and vector combinations with weight given by the product of vector importance (primary vs. secondary) and aerial abundance (see text). Vertical bars show the number of diseases reportedly transmitted by each vector, either as primary or secondary vector with the total risk weight (color intensity, see text). Horizontal bars show the number of reported vectors of each pathogen, either as primary or secondary vector with the total risk weight (color intensity, see text) for each pathogen. Pathogens circulating between humans (H), mammals (M), birds (B), or unknown (U) natural hosts are listed above horizonal bars (M > B denotes mostly mammals but some indication of birds; B/M indicates evidence for both without clear ranking; MH indicates that wild mammals are reservoir hosts but human to human transmission by vector can sustain human infection over several years, M(H) denotes that human-to-human transmission occurs rarely). Red stars indicate MBPs of highest risk for being primarily affected by windborne mosquitoes (see text).