Nitrate nitrogen (NO₃N) and, to a lesser extent, ammonium nitrogen (NH₄N) are important nutrients that can influence the nutrient enrichment (eutrophication) of surface waters and, to meet EU Directives, are subject to legislative controls. Nutrient export coefficients are widely used in water quality modelling assessments as they offer a means to assess the impacts of land use change on the eutrophication of both marine and freshwater systems. Queen's University Belfast and the Agri-Food and Biosciences Institute are analysing how river nitrate levels in Northern Ireland are changing and this PhD research is entitled, ‘spatial analysis approaches to modelling water quality in Northern Ireland'. This research involves deriving nitrogen export rates measured in rivers in Northern Ireland for CORINE landcover classes, with the main aim of updating CORINE-based annual NO₃N riverine export coefficients for 2003-2008 so as to compare them with coefficients determined in the 1990s. A secondary aim is to estimate CORINE-based NH₄N export coefficients as these have not been previously derived.
In-stream NO₃N and NH₄N concentrations from 534 river monitoring stations for 2003-2008 were supplied by the Northern Ireland Environment Agency (NIEA). Mean concentrations per catchment were calculated for two time periods: 2003-2005 and 2006-2008. Using ESRI's ArcGIS software, river catchment boundaries were overlaid with UK Met Office long-term annual mean rainfall and potential transpiration grids for the period 1971-2000 to estimate mean annual flow rates per catchment. Mean annual nitrate loadings for 2003-2005 and 2006-2008 were calculated by multiplying annual mean NIEA nitrate concentrations with estimated annual river flows on a catchment-by-catchment basis. By allowing for the human N contribution from sewers and septic tanks, we were able to determine the agricultural nitrate load for each catchment. CORINE landcover data were extracted by overlaying the CORINE landcover map with the river catchment boundaries to give a breakdown of CORINE landcover areas for each catchment.
Independent, small agricultural catchments distributed across Northern Ireland were selected for regression analysis. Using the CORINE landcover classification areas of these catchments as independent (x-axis) variables and mean annual agricultural nitrate loadings as dependent (y-axis) variables, mean annual nitrate export coefficients were derived for each landcover class using stepwise, backward linear regression. Results obtained for NO₃N show a statistically significant decrease in newly-derived export coefficients of 3.64 kgN/ha/yr for improved grassland and 2.57 kgN/ha/yr for non-improved grassland, for the years 2003-2005, from pre-2000 levels. Similarly, decreases from pre-2000 levels were also observed for the period 2006-08. To validate these results, a further regression of predicted (newly derived export coefficient) NO₃N loadings and observed NO₃N loadings (based on the NIEA data) was undertaken for the selected independent catchments, for each time period. The slope of each regression was close to 1.0, with 99% significance, indicating that CORINE landcover-based nutrient export coefficients are a good indicator of riverine NO₃N loadings.
In contrast, NH₄N export coefficients showed little variation between the dominant landcover classes of Northern Ireland viz. good pasture, poor pasture, coniferous forest and peat bogs. For the period 2006-2008, coefficients for NH₄N ranged from 0.54-0.67 kgN/ha/yr. This range was similar for 2003-2005. It was concluded that NH₄N is independent of landcover class and therefore riverine NH₄N loadings may be more closely related to other variables. Future work will investigate the relationship of NO₃N and NH₄N catchment loadings to soil type and also to manure N from farm livestock.
The results of this research show how NO₃N exports vary between landcover types. This is of potential use to water and land use managers as the updated export coefficient model can be used to assess both changes in loss of nutrients from specific land cover classes (which, in this study, were found to be declining) and to predict the effects of future land use change on riverine NO₃N and NH4N levels.
Judith Watson, Queen's University Belfast.