Nitrate contamination of groundwater: measurement and prediction

Agriculture makes a significant contribution to the diffuse source contamination of surface and groundwater resources, particularly contributing to the NO3- contamination of groundwater. Two approaches were adopted to evaluate management practices (within the context of the whole farming system) for their impacts on the environment: (1) measurement of the quality of groundwater under different farming systems, and (2) comparison of predictions of the impact of farming systems on water quality, obtained using whole farm N budgets, with measured values. The Ontario Farm Groundwater Quality Survey evaluated the rural groundwater quality in Ontario, with respect to common contaminants including NO3-. Approximately 1300 domestic farm wells were sampled, and wells were drilled in some fields of farms involved in the study. NO3- was present at concentrations above the maximum acceptable for drinking water (10 mg N l -1) in 14% of wells, including 7% of wells that also had unacceptable concentrations of coliform bacteria. Significant levels of NO3- contamination were observed under most agricultural land use practices investigated. Calculation of N budgets was simplified by assuming that there was no net change in the N content of farm assets. The N inputs to agricultural systems considered were: purchases from off-farm suppliers, N2 fixation and atmospheric deposition. Symbiotic N2 fixation was estimated from empirical relationships between crop yield and N2 fixed. The N outputs were in sales of plant and animal produce, gaseous and leaching losses. Gaseous loss was assumed to result only from volatilization of ammonia, estimated to be 39% of total manure N. We have identified one cash crop farming system where there was a true balance. The rotation included corn soybeans and wheat, with two years of soybean always being grown before corn. Many livestock farms, including two organic farms, gave large imbalances of N which might indicate that these operations were not in equilibrium. The relationship between measured and predicted values of NO3- -N expected in the groundwater under the different management systems showed that the simplified N budget overestimated the NO3- -N concentration by about one third. However, the budget approach appeared to identify farms where contamination was likely even if the actual amount was over estimated. Simplified budgets could therefore be used to compare the potential of different farming systems for causing environmental contamination.