Leaching of N, P and glyphosate from two soils after herbicide treatment and incorporation of a ryegrass catch crop

During 2005-2007, studies were carried out in two field experiments in southwest Sweden with separately tile-drained plots on a sandy soil (three replicates) and on a clay soil (two replicates). The overall aim was to determine the effects of different cropping systems with catch crops on losses of N, P and glyphosate. Different times of glyphosate treatment of undersown ryegrass catch crops were examined in combination with soil tillage in November or spring. Drainage water was sampled continuously in proportion to water flow and analysed for N, P and glyphosate. Catch crops were sampled in late autumn and spring and soil was analysed for mineral N content. The yields of following cereal crops were determined. The importance of keeping the catch crop growing as long as possible in the autumn is demonstrated to decrease the risk of N leaching. During a year with high drainage on the sandy soil, annual N leaching was 26 kg/ha higher for plots with a catch crop killed with glyphosate in late September than for plots with a catch crop, while the difference was very small during 1 yr with less drainage. Having the catch crop in place during October was the most important factor, whereas the time of incorporation of a dead catch crop did not influence N leaching from either of the two soils. However, incorporation of a growing catch crop in spring resulted in decreased crop yields, especially on the clay soil. Soil type affected glyphosate leaching to a larger extent than the experimental treatments. Glyphosate was not leached from the sand at all, while it was found at average concentrations of 0.25 mu g/L in drainage water from the clay soil on all sampling occasions. Phosphorus leaching also varied (on average 0.2 and 0.5 kg/ha/yr from the sand and clay, respectively), but was not significantly affected by the different catch crop treatments.