A risk index for fecal contamination of irrigation water during drought

Abstract

Groundwater is increasingly relied upon for irrigation water in semi-arid regions, where recurrent droughts and intensive abstraction have intensified pressure on both groundwater quantity and quality. This study developed a risk index for the contamination of groundwater used for irrigation during drought in the Berrechid Plain, Morocco, where irrigation depends almost entirely on groundwater. Groundwater fecal contamination was assessed using in situ tryptophan-like fluorescence (TLF) measurements collected at 84 sites during drought. To identify the dominant factors controlling TLF variability, 12 environmental variables describing contaminant propagation conditions and 5 source-related variables representing anthropogenic pressures were analyzed using principal component analysis, correlation analysis, and multiple linear regression. Source-related variables structured TLF variability more clearly than environmental factors, with the first two source components explaining 66% of the variance, compared with 31% for the first two environmental components. Correlation and regression analyses identified PC_src1 as the main significant driver of TLF variability, indicating that TLF concentrations increased with proximity to industrial areas, livestock farms, and wastewater treatment plants. The regression model was significant (F = 4.20, p = 0.00398), with PC_src1 as the only significant predictor (β = −2.205, p = 0.001). The selected contamination-pressure gradient was combined with long-term irrigation intensity derived from remote sensing to generate the spatial risk index. Low, moderate, and high-risk areas covered 43%, 54%, and 4% of the aquifer, respectively. The continuous risk index was significantly correlated with measured TLF concentrations (Spearman ρ = 0.443; Pearson r = 0.469; p < 0.05). Unlike conventional vulnerability approaches, this TLF-based risk index integrates anthropogenic pressure and irrigation intensity, providing an operational ecological indicator for assessing irrigation-water contamination risk in drought-affected semi-arid aquifers.