The potential for water, nutrient and faecal bacteria transport in border-dyke run-off from dairy pasture was monitored within the South Canterbury catchment of Waikakahi during the 2002/2003 milking season. The Waikakahi stream runs lengthways through the catchment and characteristically has mean summer flows approximately four-times that for winter (2002; 1850 l/s vs 450 l/s, respectively). This extra flow is assumed to be fed largely from irrigation run-off and drainage. Three borders of a flood-irrigated dairy paddock in the upper part of the catchment, located on Temuka gley soils, were directed off to a collection weir over which seven irrigation events were recorded for runoff volume, Escherichia coli (E. coli) and phosphorus (P) and nitrogen (N) concentrations. Irrigation volume loss from the Waikakahi field site was, on average, 50% of total inflow and considerably higher than the accepted recommended maximum in Australia of 10%. However, it was accepted that this represented a possible "worst-case" scenario. Run-off totalled 2600 m3/ha over the first six events with the large volume at least partly attributed to insufficient infiltration into the soil due to the low hydraulic conductivity of the Temuka soils. Irrigation volumes entering the catchment were generally sufficient to supply 90-100 mm depth of water across the area but border gradients appear too steep to allow sufficient infiltration before the water ran to the end of the border. Re-grading borders to allow for the slower infiltration rates of heavy texture soils is suggested. Concentrations of P, N and E. coli in irrigation run-off were consistently higher than the acceptable critical limits for water quality and even with in-stream dilution, would continue to exceed water standards. Total-P and dissolved reactive phosphorus (DRP) concentrations in nonfertiliser affected run-off (first six events) were high at ~0.8 mg and 0.6 mg P/l, respectively. Total phosphorus and nitrogen losses over the seven events totalled 3.4 and 2.0 kg/ha, respectively. The source of most P appears to be from high soil P levels (soil Olsen P levels >45 ìg/ ml), indicating the importance of ensuring soil Olsen P values remain within the agronomic target range (20-30 ìg/ml). The presence of fertiliser and timing of application, and days between irrigation and last grazing, were also important determinants of nutrient concentrations inirrigation run-off. Flow obstructions within the headrace channel were linked to differences in volumes entering individual borders and also caused unintended inflow from irrigation of the adjacent set of borders. Improving the smoothness of the headrace channel is required to achieve more even watering. A number of other contributing factors that may lead to excessive irrigation run-off were also noted. In this particular set in the first instance, inflow times needed shortening to reduce water loss. Reducing the volume of irrigation run-off overall entering the stream remains the chief means of reducing nutrient and faecal bacteria contaminant loadings and improving water quality. Keywords: border-dyke irrigation, dairy pasture, faecal bacteria, nitrogen, phosphorus
Degradation of water quality in Lough Neagh, Northern Ireland, by diffuse nitrogen flux from a phosphorus-rich catchment