Into the woods: Changes in mosquito community composition and presence of key vectors at increasing distances from the urban edge in urban forest parks in Manaus, Brazil

• Published in: Acta tropica 2020-06, Vol.206, p.105441-105441, Article 105441
• Main Authors: Hendy, Adam, Hernandez-Acosta, Eduardo, Chaves, Bárbara Aparecida, Fé, Nelson Ferreira, Valério, Danielle, Mendonça, Claudia, Lacerda, Marcus Vinícius Guimarães de, Buenemann, Michaela, Vasilakis, Nikos, Hanley, Kathryn A.
• Format: Article
• Language: English
• Subjects: Animals; Arbovirus; Brazil; Culicidae - virology; Demography; Forest edge; Forests; Humans; Mosquito; Mosquito Control; Mosquito Vectors - virology; NDVI; Parks, Recreational; Spillback; Zika virus; Zika Virus - isolation & purification; Zika Virus Infection - transmission
• ISSN: 0001-706X; 1873-6254
• DOI: 10.1016/j.actatropica.2020.105441

•Mosquito diversity is similar between urban forest park edge and interior.•Mosquito community composition changes substantially between edge and interior.•Urban arbovirus vectors Ae. aegypti and Ae. albopictus penetrate 100 m into forests.•Urban parks are potential refugia for Ae. albopictus and Ae. aegypti from control.•Both urban and sylvatic mosquitoes may act as arbovirus bridge vectors in parks. Mosquito-borne Zika virus (ZIKV) was recently introduced into the Americas and now has the potential to spill back into a sylvatic cycle in the region, likely involving non-human primates and Aedes, Haemagogus, and Sabethes species mosquitoes. We investigated potential routes of mosquito-borne virus exchange between urban and sylvatic transmission cycles by characterizing mosquito communities in three urban forest parks that receive heavy traffic from both humans and monkeys in Manaus, Brazil. Parks were stratified by both distance from the urban-forest edge (0, 50, 100, and 500 m) and relative Normalized Difference Vegetation Index (NDVI) (low, medium, or high), and mosquitoes were sampled at randomly selected sites within each stratum using BG-Sentinel traps. Additionally, temperature, relative humidity, and other environmental data were collected at each site. A total of 1,172 mosquitoes were collected from 184 sites sampled in 2018, of which 98 sites were resampled in 2019. Using park as the unit of replication (i.e. 3 replicates per sampling stratum), a two-way ANOVA showed no effect of distance or NDVI on the mean number of identified species (P > 0.05 for both comparisons) or on species diversity as measured by the Shannon-Wiener diversity index (P > 0.10 for both comparisons). However, the Morisita overlap index revealed that mosquito communities changed substantially with increasing distance from edge, with communities at 0 m and 500 m being quite distinct. Aedes albopictus and Ae. aegypti penetrated at least 100 m into the forest, while forest specialists including Haemagogus janthinomys, Sabethes glaucodaemon, and Sa. tridentatus were detected in low numbers within 100 m from the forest edge. Trichoprosopon digitatum and Psorophora amazonica were among the most abundant species collected, and both showed distributions extending from the forest edge to its interior. Our results show overlapping distributions of urban and forest mosquitoes at park edges, which highlights the risk of arbovirus exchange via multiple bridge vectors in Brazilian urban fo