A Chronosequence Study to Determine Soil Change Due to Long-Term Wastewater Irrigation.

Long-term field experiments are very useful to detect impacts of agricultural practice on soil. However, large efforts are required for their set-up and maintenance, and very few of these experiments have records over more than four decades. Historical records of land use change and agricultural management offer the possibility of analyzing associated soil changes in a “space for time” approach by studying soil chronosequences. We established such chronosequence in the Mezquital Valley, Central Mexico, where soils were irrigated with untreated wastewater from Mexico City for 0, 4, 8, 15, 20, 30, 50, 80 and 100 years. These fields were sampled in 1990 for the first time, and re-sampled in 2009 and analyzed for nutrients and heavy metals, and in 2009 also for emerging contaminants (pharmaceuticals and resistance genes). Models fitted to the data allow calculating rates of change, and to determine the time interval needed to approach new equilibrium conditions.

Total organic C contents increased from less than 20 g/kg to more than 40 g/kg after 50 years, while total Cd, Pb, Cu and Zn contents increased linearly up to 6 fold after 100 years. Total concentrations of ciprofloxacin, sulfamethoxazole, and carbamazepine increased with irrigation duration reaching 95% of their upper limit of 1.4 µg/kg (ciprofloxacin), 4.3 µg/kg (sulfamethoxazole), and 5.4 µg/kg (carbamazepine) in soils irrigated for 19-28 years. Acidic pharmaceuticals (diclofenac, naproxen, bezafibrate) were not retained and thus did not accumulate in soils. Increasing total concentrations of antibiotics in soil are not accompanied by increasing relative abundances of resistance genes. Yet, wastewater irrigation enlarges the absolute concentration of resistance genes in soils due to a long-term increase in total microbial biomass.

The registered changes are soil type specific. Also, the comparison of the nutrient and heavy metal data sets obtained from the same fields at the two sampling dates (1990 and 2009) proves the importance of adequate sampling and of maintaining sample archives.