Disentangling the effect of local and global spatial variation on a mosquito-borne infection in a neotropical heterogeneous environment.

Mosquito-borne pathogen transmission exhibits spatial-temporal variability caused by ecological interactions acting at different scales. We used local spatial statistics and geographically weighted regression (GWR) to determine the spatial pattern of malaria incidence and persistence in northeastern Venezuela. Seven to 11 hot spots of malaria transmission were detected by using local spatial statistics, although disease persistence was explained only for four of those hot spots. The GWR models greatly improved predictions of malaria risk compared with ordinary least squares (OLS) regression models. Malaria incidence was largely explained by the proximity to and number of Anopheles aquasalis habitats nearby (1-3 km), and low-elevation terrains. Disease persistence was associated with greater human population density, lower elevations, and proximity to aquatic habitats. However, there was significant local spatial variation in the relationship between malaria and environmental variables. Spatial modeling improves the understanding of the causal factors operating at several scales in the transmission of malaria.