Regionalisation of nitrate leaching on pasture land in Southern Manitoba

Abstract

Nitrogen is a key agricultural input which is crucial for crop growth, development, and yield. However, an excess application of nitrogen may result in the nitrate contamination of groundwater. This research focused on the estimation of nitrate leaching fluxes upon the application of pig slurry on pasture land in Southern Manitoba using physical based modeling and to further spatially disperse the point estimates of nitrate leaching fluxes at the field scale. The field site was located in Southeastern Manitoba. Grazing livestock tended to concentrate near areas such as mineral feeders, water troughs and shelters causing higher deposition density of urine and feces, resulting in six bare earth areas in the study site. HYDRUS-1D was used to determine continuous recharge and nutrient leaching estimates. The simulated leaching estimates were spatially dispersed using Cokriging. The difference in simulated and observed nitrate concentrations in groundwater was expressed in terms of RMSE between 0.023 and 5.12 mg NO3−-N L-1, NSE between 0.66 and 0.96 and the ME between −1.03 mg NO3−-N L-1 and 1.05 mg NO3−-N L-1. Simulation results suggested that two years of pig slurry application to the hay pasture on a coarse sandy soil did not cause significant accumulation of nitrate in the shallow groundwater since the maximum nitrate leaching flux were below 11 kg ha-1 resulting in NO3− concentration in groundwater below 1.2 mg L-1. In contrast, nitrate concentrations five times above the drinking water threshold were simulated in bare earth areas. Leaching of large amounts of nitrate occurred in these areas. Overall, HYDRUS-1D can be considered a useful tool in quantifying nitrate leaching estimates for pasture in a cold climate such as southeastern Manitoba, and Cokriging can be considered a reliable method for a study site where cross-correlations between variables are important to consider for carrying out interpolation.