Spatio-ecological determinants of freshwater snail intermediate hosts and schistosome infection in the Lango subregion, Northern Uganda: a geostatistical approach to targeted disease control

Background Freshwater snails of the genera Biomphalaria , Bulinus , and Lymnaea serve as intermediate hosts for trematodes causing schistosomiasis and fascioliasis, diseases of major public health concern in sub-Saharan Africa. In Uganda's Lango subregion, schistosomiasis remains endemic despite control efforts, yet comprehensive spatial and ecological analyses of snail intermediate host distributions are lacking. This study employed geostatistical approaches to identify high-risk snail habitats and sites where schistosome cercarial shedding was detected to inform targeted control strategies. Methods A cross-sectional study was conducted during the dry and rainy seasons of 2023 across 26 georeferenced sites in Lira and Kole districts. Freshwater snails were collected using standardized methods and identified morphologically. Cercarial shedding tests determined infection status but were limited to detecting patent infections. Physicochemical parameters (pH, salinity, total dissolved solids, dissolved oxygen, temperature, and conductivity) and ecological variables were measured. Spatial analysis included Moran’s I for autocorrelation, Getis-Ord Gi* for hotspot detection, and inverse distance weighting for interpolation. Generalized linear mixed models with spatial random effects were used to assess predictors of snail prevalence and density, compared with non-spatial models using the Akaike Information Criterion. Results A total of 4,802 snails from 13 species were collected, with Biomphalaria choanomphala (25.8%) being most abundant. Significant spatial clustering was detected for B. choanomphala (Moran’s I = 0.32, p = 0.004) and B. sudanica (Moran’s I = 0.24, p = 0.018). Three density hotspots and one site where infected snails clustered were identified, primarily in rice paddies and swamps near human settlements. Overall infection rate was 0.15% (5/3404 tested snails), with B. choanomphala showing the highest infection 0.06% (2/1241). Spatial GLMMs outperformed non-spatial models (ΔAIC = 12.7–15.3), revealing significant effects of salinity (odds ratio = 0.21, p < 0.001), total dissolved solids (β = -0.03, p = 0.002), dissolved oxygen (β = 0.54, p = 0.003), and anthropogenic activities. Spatial random effects accounted for 18–24% of residual variation. Conclusions This study demonstrates the significant added value of geostatistical methods in identifying snail intermediate host clusters and sites with detected infections. The integration of spatial analysis with ecological modeling provides a robust framework for potential targeted snail control. Our findings suggest that focused interventions in identified high-density areas, integration of spatial risk maps into district health planning, and community engagement in modifying high-risk water contact sites could help reduce schistosomiasis transmission in the Lango subregion.