Nitrogen leakage from different land use types : a comparison between the watersheds of Graisupis and Vardas, Lithuania

Abstract: Eutrophication problems in the Baltic Sea have drawn attention to the contribution of nutrients from surrounding countries. The problems have attracted considerable attention to non-point source nitrogen pollution of rivers and lakes. The ability to predict nitrogen export from inland sources is essential in order to evaluate the effects of anthropogenic activities, of which the leakage from agriculture is one. In this study the nitrogen leakage from different kinds of land use is investigated in Lithuania. The overall aim is to develop an easy, fast and cheap methodology, in order to measure nitrogen leakage. Thus, a method used in the Genevad drainage basin, Sweden, was applied (Wickberg, 2000). The latter uses an established relationship between nitrate/nitrite and conductivity in order to count for the nitrogen leakage. Another part of the aim is to present a new method of land use mapping at the Lithuanian Institute of Water Management. The study includes a comparision between the watersheds of river Graisupis, an agricultural region, and river Vardas, in a more hilly part of the country. In order to carry out the study, land use maps, based on aerial photos from 1995, was created and water samples were collected and analysed. The data material was then analysed and presented with statistical- and GIS- methods. A Digital Elevation Model (DEM) was used together with an application written in Visual Basic 6.0, in order to isolate sub-drainage basins, which were used in the analyse of leakage. The results show that there is no significant relationship between conductivity and nitrite/nitrate in either of the two watersheds. Thus, the regression equation for these two variables could not be used to count for the nitrogen leakage from different land use classes using the collected conductivity values. This shows that the method used in the Genevad drainage basin is not valid in the study area, and thus the overall aim of the study (easy, fast, cheap) could not be achieved. When instead using the measured conductivity values, in order to observe the leakage of nutritive salts from different land use classes, they show that there is a high leakage from forest in both watersheds. Also, the correlations between the distribution of a specific land use class and the amount of nitrogen in the river, show a small significance only in forests in Graisupis. Thus, forest seems to affect the leakage of nutrients. However, riverine transport of nitrogen cannot only be based exclusively on land use. Other catchment characteristics must also be considered in order to understand the influences from different processes concerning nitrogen leakage.