Heavy metal concentrations in a soil irrigated with treated municipal wastewater: use of mixed models to analyze the effect over time

In arid and semi-arid regions of the world, non-conventional waters may represent an important complementary irrigation source. To investigate potential soil contaminations risks, repeated measurement experiments and rigorous methodological approaches may be adopted to assess variations of contaminant concentrations in the soil, eliminating possible confounding effects due to correlations over time. A field trial was carried out in Southern Italy to assess variation of heavy metals concentrations in a soil irrigated with: (i) freshwater (FW), (ii) a secondary-treated (SW) and (iii) a tertiary-treated municipal effluent (TW). Concentrations of cadmium, chromium, copper, nickel, lead and zinc were quantified by ICP-OES on soil samples collected over three sampling dates. Mixed effects models accounting for residual autocorrelation were used to evaluate differences over time and treatments and the following variance-covariance structures were compared: compound symmetry (CS); autoregressive of first order (AR(1)); autoregressive of first order with heterogeneous variances (ARH(1)). ARH(1) models were significant in most of the cases investigated, confirming heterogeneity of variances observed over time. At the end of each irrigation season, soil heavy metal contents did not show significant differences among treatments. Significant interactions between water source and sampling time were observed mainly under SW and TW supply; in any case no increase of metals over time was recorded. Our results confirmed low risk of soil metals contamination after short and medium term irrigation with treated municipal wastewater. However, when using non-conventional water resources, proper irrigation management and continuous monitoring of water and soil are needed to avoid soil degradation.