Spatial–temporal distribution of Anopheles larval habitats in Uganda using GIS/remote sensing technologies

• Published in: Malaria Journal
• Main Authors: Ryan Tokarz, Robert J. Novak
• Format: Article in Journal/Newspaper
• Language: English
• Published: BMC 2018
• Subjects: Mosquito control; Mosquito surveillance; Autocorrelation; Moran’s I; Remote sensing; Arctic medicine. Tropical medicine; RC955-962; Infectious and parasitic diseases; RC109-216; Arctic; Human health
• Online Access: https://doi.org/10.1186/s12936-018-2567-z; https://doaj.org/article/642b8dcdf9524d849385c519fa3609b3

Abstract Background Anopheles mosquitoes impose an immense burden on the African population in terms of both human health and comfort. Uganda, in particular, boasts one of the highest malaria transmission rates in the world and its entire population is at risk for infection. Despite the immense burden these mosquitoes pose on the country, very few programmes exist that directly combat the issue at the vector control level and even fewer programmes focus on the vector in its most vulnerable juvenile stages. This study utilizes remote sensing techniques and spatial autocorrelation models to identify and prioritize the most prolific Anopheline larval habitats for control purposes in a rural community in Uganda. Methods A community-based mosquito surveillance programme was developed and implemented in Papoli Parish in Eastern Uganda over a 4-month period. Each day, a trained field team sampled the larval habitats of Anopheles mosquitoes within the population-dense areas of the community. Habitats and their productivity were identified and plotted spatially on a daily basis. Daily output was combined and displayed as a weekly habitat time-series. Additional spatial analysis was conducted using the Global and Anselin’s Local Moran’s I statistic to assess habitat spatial autocorrelation. Results Spatial models were developed to identify highly significant habitats and dictated the priority of these habitats for larval control purposes. Weekly time-series models identified the locations and productivity of each habitat, while Local Moran’s I cluster maps identified statistically significant clusters (Cluster: High) and outliers (High Outlier) that were then interpreted for control priority. Models were stitched together in a temporal format to visually demonstrate the spatial shift of statically significant, high priority habitats over the entire study period. Discussion The findings show that the spatial outcomes of productive habitats can be made starkly apparent through initial habitat modelling and resulting ...