Multivariate Assessment of Waterborne Bacterial Contamination across Boreholes, Wells and Streams in Kuje Agro-Ecological Area, Nigeria

Abstract

Waterborne bacterial contamination remains a major public health concern in agroecological communities that rely on multiple water sources for domestic and agricultural use. This study assessed the distribution and interrelationships of selected waterborne bacteria across boreholes, hand-dug wells and streams in the Kuje agroecological area of the Federal Capital Territory, Nigeria, using multivariate statistical techniques. Water samples were collected from boreholes, wells, streams and a treated water control source and analyzed for Escherichia coli, Salmonella spp., Shigella spp. and Pseudomonas spp. using standard microbiological methods. Multivariate analysis of variance (MANOVA) was applied to evaluate differences in bacterial distributions across water sources, supported by Box’s M test, Dunnett’s post-hoc comparisons and Pearson correlation analysis. MANOVA results showed no statistically significant differences using Wilks’ Lambda and Pillai’s Trace (p > 0.05); however, Roy’s Largest Root revealed a significant multivariate effect (p < 0.05), indicating subtle but meaningful variations in bacterial contamination patterns. Correlation analysis demonstrated significant positive associations among key bacterial indicators, suggesting shared contamination pathways linked to fecal pollution and environmental exposure. The findings highlight the value of multivariate statistical tools in water quality assessment and underscore the need for routine microbiological monitoring, improved sanitation practices and integrated water safety planning in agro-ecological communities.