Modelling of human and climatic impact on nitrogen load in a Swedish river 1885-1994

Changes in environmental conditions within a river basin in South Central Sweden (1400 km(2)) and impacts on riverine nitrogen (N) transport were evaluated. A historical database was compiled and the process-based HBV-N model used to estimate flow normalised N loads in 1885, 1905, 1927, 1956, 1976, and 1994, using a standard climatological record (1985-1994). The study shows the value of process-based modelling in environmental impact assessment, by making it possible to assess and integrate the effect of a number of factors, both with regard to human impact and natural climatic variability. Factors taken into account include: the effects of land use, agricultural practices, atmospheric deposition, human dietary intake, use of flush toilets, lowering of lakes, building of dams, and climatic variability. For all years studied, agriculture was the overriding source of N, and changes in riverine N over time mainly reflected changes in land use and agricultural practices. In spite of decreasing N-leaching from agriculture, the net load remained fairly constant between 1885 and 1927, due to reduced N retention. Drainage of agricultural land had a dominating impact on reducing N retention, which increased the N loads, while the effects of the lowering of lake levels and dam building were less pronounced. Household N emission per capita was higher in 1994 than in 1927, as the increased consumption of meat and dairy products alone resulted in a higher increase of the emission than was compensated for with wastewater treatment improvement. In addition, introduction of flush toilets increased the emission from households. In total, the net load in 1976 was twofold higher than that in 1885, 1905 and 1927, due to increased leaching from agriculture, wastewater emission, and atmospheric deposition on lake surfaces. Finally, the impact of climatological variability was assessed, using a 110-yr climatological record. The choice of 10-yr period of climatological data was the factor that had the largest impact on calculated N load.