COS 124-5 - Local tree cover and climate predict mosquito biodiversity and disease vector occurrence in a countryside landscape

Background/Question/Methods

The importance of countryside landscapes to biodiversity and ecosystem services conservation is increasingly recognized, but the features that differentiate patchworks of natural habitat and human infrastructure that retain versus lose biodiversity and ecosystem services remain unclear. Mounting evidence from Latin America shows that percent tree cover at small spatial scales of < 100 m is a reliable predictor of biodiversity for birds, mammals, reptiles, and plants, suggesting local tree cover management as a promising conservation tool in this region. However, the generality of this relationship across taxa and ecosystem services is unknown. Mosquitoes—several species of which transmit human, livestock, and wildlife diseases—are a key taxon where knowledge gaps surrounding community responses to landscape context are directly relevant to both ecosystem and human wellbeing. Therefore, we asked 1) how does tree cover relate to mosquito community assembly in terms of species richness, composition, and presence of key disease vector species across spatial scales? and 2) what factors drive observed relationships between tree cover and mosquito community characteristics? To address these questions, we paired a survey of mosquito communities along a land use gradient in southern Costa Rica with land cover, land use, and climate data.

Results/Conclusions

We found that local land cover and mean annual temperature, but not land use type (residential, agricultural, or forest) predicted mosquito species richness, with more diverse communities occurring at higher tree cover and in cooler climates. Additionally, we found that land cover, land use type, and mean temperature predicted mosquito community composition. In contrast, the occurrence probability of the abundant disease vector Aedes albopictus increased at lower tree cover, warmer temperatures, and with residential land use. Finally, the spatial scales at which tree cover predicted species richness differed compared to Ae. albopictus occurrence: there was a significant effect of tree cover at radii greater than 120 m and greater than 20 m for species richness and Ae. albopictus occurrence, respectively. Overall, our results add to support from other taxa for the value of small areas of both natural and semi-natural habitat in sustaining biodiversity and ecosystem services—here, in the form of protection against an invasive mosquito that is a major vector of human disease.