Into the woods: Insertion of oil palm plantations in Borneo narrows mosquito diversity but favors a key vector of sylvatic dengue virus

Background/Question/Methods : Mosquito-borne viruses, including dengue, Zika, and chikungunya, cause global pandemics that sicken or kill millions of people. Each of these viruses emerged into the human population via spillover from sylvatic transmission cycles maintained between wildlife and Aedes mosquitoes in tropical forests. These ancestral sylvatic strains continue to circulate and to spill over into the human population in present day, and it is crucial to identify environmental factors that shape the likelihood of spillover. We hypothesized that land cover change, specifically transition from forest to oil palm plantation, alters the distribution, abundance, and diversity of mosquitoes and thereby influences rates of spillover. We tested four predictions: (i) mosquito diversity and abundance would decline in plantations; (ii) mosquito diversity would be greatest at the forest edge iii) Aedes albopictus, an opportunistic vector of multiple viruses, would occur in equal abundance in plantations and forest, (iv) Ae. niveus, a canopy-living vector of sylvatic dengue virus, would be most abundant in forests. Adult mosquitoes were collected at 340 sites at 4 distances from forest edge into oil palm plantations (10, 20, 50, 100 m), and 6 distances from forest edge into forests (0, 10, 20, 50, 100, 500 m) in Sarawak, Malaysian Borneo.

Results/Conclusions : In total 894 mosquitoes comprising 24 genera were collected. Genus-level richness and gamma diversity increased with distance into the forest; interior forest sites were significantly more similar to each other than to oil palm sites. Alpha diversity increased with distance into the forest; however, contrary to our prediction, sites 100 m into forest were the most diverse. Genus level composition was different between distances, but remained more similar from 10 m up to 500 m. There was a relationship between elevation and mosquito community similarity, further explaining the spatial patterns observed. Mean mosquito abundance declined significantly in interior plantation; however, as predicted there was no difference in the mean number of Ae. albopictus sampled at any distance. Ae. niveus were rarely collected; however, all specimens were collected at 10m or deeper into forests, supporting our fourth prediction. These data indicate that spillover of sylvatic viruses may occur in oil palm where they could be transmitted via Ae. albopictus but that likelihood of spillover increases with distance into the forest, where more types of potential vector species occur and where mosquito abundance is, on average, higher.