ABSTRACT In multilayer aquifer systems beneath urbanising coastal plains, vertical connectivity between hydrostratigraphic units controls groundwater flow, recharge transmission and vulnerability to contamination, yet conventional piezometric and hydrochemical approaches often yield ambiguous evidence in low‐relief settings. Stable isotopes of water (δ 18 O, δ 2 H) offer a complementary tracer because they integrate flow pathways at basin scales and can formally test whether superposed aquifers share equivalent meteoric signatures. This framework was applied to the Pampeano–Puelche Aquifer System (PPAS) beneath Gran La Plata, Argentina, where an unconfined unit overlies a semi‐confined aquifer through a laterally variable aquitard. Winter 2025 sampling ( n = 102) encompassed groundwater from both aquifers, surface water and the dual‐source urban supply (tap) network. Statistical comparisons revealed isotopic indistinguishability between aquifers (Mann–Whitney p > 0.05; Cohen's d ≤ 0.33), providing quantitative evidence for effective hydraulic connectivity despite hydrostratigraphic differentiation. Multivariate analysis identified three clusters, with coastal salinisation driven by mixing with marine‐influenced waters rather than evaporation‐driven isotopic fractionation (systematic increases in δ 18 O and δ 2 H and decreases in d ‐excess), as supported by d ‐excess behaviour together with preserved alignment to the local meteoric water line. Monte Carlo mixing models demonstrated sfubstantial spatial heterogeneity in network source contributions (0%–100% groundwater), including undocumented blending in operationally classified zones. Regression‐kriged isoscapes revealed coherent south‐to‐north isotopic gradients transcending aquifer boundaries. These results provide the first regional groundwater isoscapes for a major Argentine urban centre, demonstrate a transferable conditional modelling framework for multilayer systems, and establish isotopic tracing as a diagnostic tool for infrastructure performance invisible to conventional hydraulic monitoring.
Niño et al. (Wed,) studied this question.