Particulate-facilitated leaching of glyphosate and phosphorus from a marine clay soil via tile drains

Losses of commonly used chemical pesticides from agricultural land may cause serious problems in recipient waters in a similar way to phosphorus (P). Due to analytical challenges concerning determination of glyphosate (Gly), transport behaviour of this widely used herbicide is still not well known. The objective of the present study was to quantify and evaluate leaching of Gly in parallel with P. Leaching losses of autumn-applied Gly (1.06 kg ha(-1)) via drainage water were examined by flow-proportional sampling of discharge from 20 drained plots in a field experiment in eastern Sweden. Samples were analysed for Gly in particulate-bound (PGly) and dissolved (DGly) form. The first 10 mm water discharge contained no detectable Gly, but the following 70 mm had total Gly (TotGly) concentrations of up to 6 mu g L-1, with 62% occurring as PGly. On average, 0.7 g TotGly ha(-1) was leached from conventionally ploughed plots, compared with 1.7 g TotGly ha(-1) from shallow-tilled plots (cultivator to 12 cm working depth). Higher Gly losses occurred in snowmelt periods in spring, but then with the majority (60%) as DGly. All autumn concentrations of PGly in drainage water were significantly correlated (p<0.001) to the concentrations of particulate-bound phosphorus (PP) lost from the different plots (Pearson correlation coefficient 0.84), while PP concentrations were in turn significantly correlated to water turbidity (Pearson correlation coefficient 0.81). Leaching losses of TotGly were significantly lower (by 1.3 g ha(-1); p<0.01) from plots that had been structure-limed three years previously and ploughed thereafter than from shallow-tilled plots. Turbidity and PP concentration also tended to be lowest in discharge from structure-limed plots and highest from shallow-tilled plots. This difference in TotGly leaching between soil management regimes could not be explained by differences in measured pH in drainage water or amount of discharge. However, previously structure-limed plots had significantly better aggregate stability, measured as readily dispersed clay (RDC), than unlimited plots. The effects of building up good soil structure, with strong soil aggregates and an appropriate pore system in the topsoil, on mitigating Gly and P losses in particulate and dissolved form should be further investigated.