Inﬂuence of Anthropogenic Load in River Basins on River Water Status: A Case Study in Lithuania

: Twenty-four rivers in different parts of Lithuania were selected for the study. The aim of the research was to evaluate the impact of anthropogenic load on the ecological status of rivers. Anthropogenic loads were assessed according to the pollution sources in individual river catchment basins. The total nitrogen (TN) values did not correspond to the “good” and “very good” ecological status classes in 51% of the tested water bodies; 19% had a “bad” to “moderate” BOD 7 , 50% had “bad” to “moderate” NH 4 -N, 37% had “bad” to “moderate” NO 3 -N, and 4% had “bad” to “moderate” PO 4 -P. The total phosphorus (TP) values did not correspond to the “good” and “very good” ecological status classes in 4% of the tested water bodies. The largest amounts of pollution in river basins were generated from the following sources: transit pollution, with 87,599 t/year of total nitrogen and 5020 t/year of total phosphorus; agricultural pollution, with 56,031 t/year of total nitrogen and 2474 t/year of total phosphorus. The highest total nitrogen load in river basins per year, on average, was from transit pollution, accounting for 53.89%, and agricultural pollution, accounting for 34.47%. The highest total phosphorus load was also from transit pollution, totaling 58.78%, and agricultural pollution, totaling 28.97%. Multiple regression analysis showed the agricultural activity had the biggest negative inﬂuence on the ecological status of rivers according to all studied indicators.


Introduction
Lithuania is committed to achieving the objectives of the EU Water Framework Directive by 2027 and to achieving good water status in inland waters. There are approximately 30 thousand rivers and creeks in Lithuania, with a longer than 200 m, reaching an overall sum of 63,700 km. Although the ecological condition of Lithuanian rivers has been mostly improving over the last few years, it has been determined that only 49% of them correspond to a good ecological state [1], in the period of 2010-2013.
Sources of pollution are divided as follows: background pollution (forests), diffuse (non-point) pollution from agricultural lands, surface sewage that is not treated in wastewater treatment plants (WWTPs), and concentrated (point) pollution caused by households, urban, municipal, industrial wastewater (wastewater treatment plants), and others. Human activities affect the condition of water bodies differently in rural areas (agricultural activities, livestock) and urban areas (industrial, municipal, domestic wastewater discharges). Changes of land use also have a negative effect on water condition of rivers [15][16][17][18][19][20][21]. Landscape changes caused by anthropogenic activities and land cover make a significant

Study Area
To determine the risk of water bodies that do not comply with the water quality standards, the physicochemical quality indicators at 94 locations of 24 rivers were studied. Water samples were taken between January and March, April and June, July and September, and October and December in 2014-2020. The investigated river's water sampling areas and their hydrological data are presented in Figure 1 and Table 1.

Presentation of Pollution Sources
The assessment of contamination sources considered the nature of land use, the nature of cities and settlements, the location of potential sources of concentrated source pollution, the nature and intensity of economic activities in the basin and their potential impact on water bodies, recreational activities, and other economic activities that may not be in good condition according to condition requirements, and so forth.
Diffuse agricultural pollution, consisting of manure and mineral fertilizer loads resulting from agricultural activities and from the load on the population whose households are not connected to sewage collection systems.
The main sources of concentrated pollution are wastewater from cities, settlements, industrial enterprises and rain and surface water, wastewater from urban areas.
High potential for concentrated pollution to enter water bodies directly or through river tributaries.
For the quantification of pollution indicators, the following factors have been assessed: • Domestic and industrial wastewater disposal facilities in the study areas, the extent of their pollution loads, the impact on the status of the water body and the average TN and TP value of wastewater in the period 2015-2020. Data from the Environmental Protection Agency (EPA) on wastewater dischargers, identified pollutant concentrations, and annual wastewater volumes were estimated by dividing their statistical values by water body feeding basins; • The number of people connected to the sewage collection systems and sewage management (i.e., central, individual, or no management (statistics)). The contamination loads in the environment released by the residents whose wastewater was not collected were assessed according to the HELCOM recommendations, which specify that one resident generates 25.6 kg of waste according to the BOD 7 , 4.4 kg of TN, and 0.9 kg of TP; • To determine the nutrient loads from residential and commercial areas, data from the SWAT (small watershed to river basin-scale model used to simulate the quality and quantity of surface and ground water) model were used to calculate and evaluate pollution loads. SWAT model is a basin-scale continuous-time model that operates on a continuous basis and assesses the impact of management practices on water, sediment, and agrochemicals in non-monitored basins [61]. SWAT is widely used in assessing soil erosion prevention and control, diffuse source pollution control and regional management in watersheds; • To assess the impact of the transformation of biogenic nutrients in soil and water body pollution, a SWAT model was used to calculate the average of total nitrogen (TN) and total phosphorus (TP) leaching.

Statistical Analyses
To statistically assess the significant impacts on quality factors related to the ecological status of water bodies, the impacts of anthropogenic load indicators TP and TN from municipal wastewater, surface wastewater, households not connected to sewage networks, agricultural land, background, and transit pollution (t/year); agricultural land, forests, wetlands, meadows, arable, infertile land, and green land (ha) on water quality indicators (Y) for the water in rivers were determined. A multiple linear regression model was applied: The coefficient b j shows how much the value of Y increases (or decreases) by one unit, as x j increases when the remaining x k are fixed. t is Student's criterion, according to which we determined whether the b j coefficients differed statistically significantly from zero, and according to this, we decided whether the predicted values depended upon x j . The standardized coefficient beta was used to determine the relative influence of independent variables on the predicted Y. In absolute terms, a higher beta coefficient indicates greater dependence of Y on x j .
The regression model is appropriate due to the following: • The Levene test was applied as an endogeneity test; the R code was applied to generate the analyses in this area, R 2 ≥ 0.20; • ANOVA was performed with a p-value of <0.05; • t-tests were performed, showing significance at p < 0.05; • All SWFs (Dispersion reduction factor) were ≤4 (no diversity problems); • All Cook measure values were ≤1.

Ecological Status Classes of the Stretches of Rivers According to the Physicochemical Values of Elemental Indicators
Studies on the physicochemical quality of element indicators were performed for NO 3 -N (mg/L), NH 4 -N (mg/L), TN (mg/L), PO 4 -P (mg/L), TP (mg/L), and the BOD 7 (mg/L). The results are shown in Figure 2. The results presented in Figure 2 show that, according to the TN, 51% of the studied rivers did not meet the requirements of the "good" ecological class; 19% of the rivers had a "bad" to "moderate" BOD7, 50% had ''bad'' to ''moderate'' NH4-N, 37% had ''bad'' to ''moderate'' NO3-N, 4% had ''bad'' to ''moderate'' TP, and 4% had ''bad'' to ''moderate'' PO4-P.

Assessment of Nutrient Loads in River Basins
Nutrient loads in the river basins were calculated by collected the SWAT model data. Calculations were performed in tons per year for the inflows into the rivers for the total nitrogen and total phosphorus. The TN and TP loads in river basins (t/year) are presented in Figure 3.   The results presented in Figure 2 show that, according to the TN, 51% of the studied rivers did not meet the requirements of the "good" ecological class; 19% of the rivers had a "bad" to "moderate" BOD 7, 50% had ''bad" to ''moderate" NH 4 -N, 37% had ''bad" to ''moderate" NO 3 -N, 4% had ''bad" to ''moderate" TP, and 4% had ''bad" to ''moderate" PO 4 -P.

Assessment of Nutrient Loads in River Basins
Nutrient loads in the river basins were calculated by collected the SWAT model data. Calculations were performed in tons per year for the inflows into the rivers for the total nitrogen and total phosphorus. The TN and TP loads in river basins (t/year) are presented in Figure 3. The results presented in Figure 2 show that, according to the TN, 51% of the studied rivers did not meet the requirements of the "good" ecological class; 19% of the rivers had a "bad" to "moderate" BOD7, 50% had ''bad'' to ''moderate'' NH4-N, 37% had ''bad'' to ''moderate'' NO3-N, 4% had ''bad'' to ''moderate'' TP, and 4% had ''bad'' to ''moderate'' PO4-P.

Assessment of Nutrient Loads in River Basins
Nutrient loads in the river basins were calculated by collected the SWAT model data. Calculations were performed in tons per year for the inflows into the rivers for the total nitrogen and total phosphorus. The TN and TP loads in river basins (t/year) are presented in Figure 3.    The river basins get the largest amounts of pollution from transit loads located above the research locations, with the total nitrogen equaling 87,599.32 t/year and total phosphorus amounting to 5019.51 t/year. From agricultural activities, the total nitrogen reached 56,030.53 t/year and the total phosphorus was 2474.14 t/year. The amounts from background pollution (urban areas and forests) were 17,941.31 t/year of total nitrogen and 619.09 t/year of phosphorus. The amounts from municipal sewage were 368.66 t/year of total nitrogen and 342.12 t/year of phosphorus. The amounts from surface sewage were 321.64 t/year of total nitrogen and 29.02 t/year of phosphorus. Residents whose sewage was not discharged into sewage treatment systems generated 281.64 t/year of total nitrogen and 57.06 t/year of phosphorus. Figure 4 shows the percentage distribution of total nitrogen and total phosphorus loads in the studied river basins.  The river basins get the largest amounts of pollution from transit loads located above the research locations, with the total nitrogen equaling 87,599.32 t/year and total phosphorus amounting to 5019.51 t/year. From agricultural activities, the total nitrogen reached 56,030.53 t/year and the total phosphorus was 2474.14 t/year. The amounts from background pollution (urban areas and forests) were 17,941.31 t/year of total nitrogen and 619.09 t/year of phosphorus. The amounts from municipal sewage were 368.66 t/year of total nitrogen and 342.12 t/year of phosphorus. The amounts from surface sewage were 321.64 t/year of total nitrogen and 29.02 t/year of phosphorus. Residents whose sewage was not discharged into sewage treatment systems generated 281.64 t/year of total nitrogen and 57.06 t/year of phosphorus. Figure 4 shows the percentage distribution of total nitrogen and total phosphorus loads in the studied river basins.
The highest annual total nitrogen load for river basins per year, on average, came from transit pollution, accounting for 53.89%. A total of 34.47% came from agricultural pollution, 11.04% came from background pollution (urban areas and forests), 0.17% came from pollution from residents who were not connected to sewage systems, 0.20% came from surface sewage, and 0.23% came from municipal wastewater.
The highest annual load of total phosphorus in river basins was from transit pollution, accounting for 58.78%. A total of 28.97% came from agricultural pollution, 7.25% came from background pollution (urban areas and forests), 0.67% came from pollution from inhabitants who were not connected to sewage systems, 0.34% came from surface sewage, and 4.01% came from municipal wastewater. The highest annual total nitrogen load for river basins per year, on average, came from transit pollution, accounting for 53.89%. A total of 34.47% came from agricultural pollution, 11.04% came from background pollution (urban areas and forests), 0.17% came from pollution from residents who were not connected to sewage systems, 0.20% came from surface sewage, and 0.23% came from municipal wastewater.
The highest annual load of total phosphorus in river basins was from transit pollution, accounting for 58.78%. A total of 28.97% came from agricultural pollution, 7.25% came from background pollution (urban areas and forests), 0.67% came from pollution from inhabitants who were not connected to sewage systems, 0.34% came from surface sewage, and 4.01% came from municipal wastewater.

Influence of Anthropogenic Loading on Total Nitrogen, Ammonium Nitrogen, Nitrate Nitrogen, Total Phosphorus, and Phosphate Phosphorus
The influence of anthropogenic loading on total nitrogen concentration (TN is dependent variable Y) was calculated by multiple regression analysis and results are presented in Table 2.

Influence of Anthropogenic Loading on Total Nitrogen, Ammonium Nitrogen, Nitrate Nitrogen, Total Phosphorus, and Phosphate Phosphorus
The influence of anthropogenic loading on total nitrogen concentration (TN is dependent variable Y) was calculated by multiple regression analysis and results are presented in Table 2.
Multiple regression analysis of the influence of anthropogenic loads on the total nitrogen concentration in the water showed that the total nitrogen value was affected by N from agricultural land, and the total nitrogen amount was generated from agricultural land and arable land (p < 0.05). The higher the concentrations of TN were from arable land and agricultural land, the higher the value of TN was in the water (positive function).
The effect of anthropogenic loads on the ammonium nitrogen concentration (NH 4 -N is dependent variable Y) was calculated by multiple regression analysis. The results are presented in Table 3.  Multiple regression analysis of the influence of anthropogenic loads on the ammonium nitrogen concentration in the water showed that the total NH 4 -N value was affected by the TN from households not connected to sewage networks, the TN from agricultural land and transit pollution, and the NH 4 -N amount generated from forests, wetlands, meadows, arable land, infertile land, and green land (p < 0.05). The higher the concentrations of NH 4 -N were from households not connected to the sewage networks, agricultural land, transit pollution, wetlands, meadows, and arable land, the higher the value of the NH 4 -N was in the water (positive function). The higher the NH 4 -N concentrations were from forests, infertile land, and green land, the lower the NH 4 -N concentration was in the water (negative function).
The effect of anthropogenic loads on the nitrate nitrogen concentration (NO 3 -N is dependent variable Y) was calculated by multiple regression analysis. The results are presented in Table 4. Multiple regression analysis of the influence of anthropogenic loads on the nitrate nitrogen concentration in the water showed that the NO 3 -N value was affected only by arable land (p < 0.05). The higher the concentration of NO 3 -N was from arable land, the higher the value of NO 3 -N was in the water (positive function).
The effect of anthropogenic loads on the total phosphorus concentration (TP is dependent variable Y) was calculated by multiple regression analysis. The results are presented in Table 5.
Multiple regression analysis of the influence of anthropogenic loads in basins on the concentration of total phosphorus in the water showed that the total phosphorus value was influenced by the discharge of surface wastewater from households not connected to sewage networks, agricultural land, arable land, infertile land, and green land (p < 0.05). The higher the TP concentration was in the surface wastewater from households not connected to sewage networks, agricultural land, arable land, the higher the TP value was in the water (positive function). The larger the infertile and green areas were in the river basins, the lower the total phosphorus concentration was in the water (negative function).
The effect of anthropogenic loads on the phosphate phosphorus concentration, (PO 4 -P is dependent variable Y), was calculated by multiple regression analysis. The results are presented in Table 6.  Multiple regression analysis of the influence of anthropogenic loads in basins on the concentration of phosphate phosphorus in the water showed that the PO 4 -P value was influenced by the discharge of municipal wastewater from background and transit pollution, agricultural land, forests, meadows, arable land, infertile land, and green land (p < 0.05). The higher the PO 4 -P concentration was in the municipal wastewater from background pollution, agricultural land, meadows, arable land, the higher the PO 4 -P value was in the water (positive function). The higher the transit pollution, and the larger the forests, infertile, and green areas were in the river basins, the lower the PO 4 -P concentration was in the water (negative function).

Discussion
Agricultural activity has strict negative impact on condition of surface water bodies, their ecosystems, the degradation of vegetation, and the quantitative and qualitative changes in fish populations in the Mediterranean basin [62]. The main factor affecting the Baltic Sea region environment is the increased amount of nutrients in rivers, mainly from diffuse agricultural sources [63]. The diffused nitrogen of anthropogenic origin account for about 70% of the total load deposited into rivers and lakes of the Baltic Sea basin area. Of the total diffuse load of nitrogen deposited into the Baltic Sea, 80% is from agriculture [64,65]. In Estonia, Latvia, and Lithuania, agriculture was intensified, and the amount of nitrogen fertilizers was increased after the 1990s [66].
Ikauniece and Lagzdinš assessed the status of two rivers, the Slocene and the Age, in Latvia. It was found that the ecological and chemical status of these rivers depended on the following factors: climatic conditions, types of soil and land-use, and human activities. The impact of land-use types and concentrations of total nitrogen, NO 3 -N, NH 4 -N, total phosphorus, and PO 4 − -P on the water of rivers was established. The highest concentrations of these substances were determined in the spring. It can be stated that snow melt during the spring period increases losses of biogenic compounds from concentrated sources [36].
An increase in sensitivity was found in basins with more agricultural land and more fertilizer. A change in the use of chemical fertilizers by ±20% affected the NO 3 -N loads in the water body between zero effect and an increase of ±13%, while a change in manure use by ±20% affected the NO 3 -N loads in the water body from zero effect to a change of −6% to +7% [63]. Ferrier [67] pointed out that nitrate concentrations in water of the rivers depend on the area of arable land, and there is a relationship between orthophosphate-P, suspended solids concentrations and meadow cover. Studies conducted in the Liswarta basin (Poland) have showed that high concentrations of nutrients in the Liswarta River and its tributaries are closely linked to the agriculture activities in this basin. However, urban wastewater effluents effected the highest concentrations of nutrients set in the Biała Oksza River [34]. Other Polish researchers assessed the influence of land use on the condition of the Dunajec, Czarny Dunajec, Biały Dunajec, and Białka rivers in the Podhale region (southern Poland). The results of their study showed that the concentrated pollution sources, such as effluents from WWTPs or untreated sewage from households, were more important than diffuse sources but agricultural activities significantly affect water quality of rivers [39].
Watershed modelling was used to discover the critical areas of water quality of rivers, as well as to define impacts and identify the most significant pollution sources in the river basins in Lithuania. Regional diffuse pollution leaching patterns were estimated using this model. The largest leaching rates total phosphorus were assessed in the southeastern and western parts of Lithuania. The largest leaching of total nitrogen was determined to occur in the center of the country. It can be seen from the modelling results that agriculture is the dominant pollution source in all Lithuanian river basins. The organic loads from diffuse pollution sources accounted for 60-90% of the annual loads in all of the river basins, excluding the urban catchments of the Neris and Nemunas rivers. The total phosphorus loads from agricultural sources accounted for 50-93% of the annual TP load. The pollution from concentrated sources and non-sewered households had almost no influence on the nitrate loads, and agriculture was the only dominant source of pollution, contributing 90-99% of the annual nitrate load [68]. It was determined that, satisfactorily, 90% of all nitrogen entered the Mūša sub-catchment from the diffuse pollution sources, including 87% from the arable land and just a little more than 3% from the forest territory and pastures. A total of 10% of all nitrogen in the basin came from the concentrated pollution sources. The largest amounts of total phosphorus in the Mūša sub-catchment entered the basin from the concentrated pollution sources (about 49%), arable land (36%) and about 15% from the forest area and pastures [69].
Various sources indicate measures for protection against diffuse pollution. In Poland, the recommendations for the protection of river valleys from biogenic pollution include the activities such as preserving natural vegetation on the banks of rivers, reducing of intensive agriculture activities and others [34]. Scholz [70] introduced diffuse pollution control strategies involving draining the natural wetlands by ditches in Germany.
In Lithuania, the main measures that should be applied to reduce the input of pollution from agricultural activities into rivers and other inland waters are as follows [71]: The application of fertilization plans and targeted/precision farming. Balanced fertilization reduces the need for fertilizers and pesticides and saves water resources. This results in less nitrogen and phosphorus leaching and less eutrophication in surface water bodies; Additional protection strips for surface water bodies. The protective strips of natural vegetation left along the water bodies help to absorb excess nutrients and control water pollution; Stubble fields left during the winter help conserve water resources and prevent nutrient leaching; The installation of controlled drainage. An intelligent drainage system increases yields by reducing the need for fertilizer and stopping the leaching of nutrients into surface water bodies.

1.
The total nitrogen values did not comply with the requirements of to the ''good" and ''very good" ecological status classes in 51% of the tested water bodies; 19% had a ''bad" to ''moderate" BOD 7 , 50% had ''bad" to ''moderate" NH 4 -N, 37% had ''bad" to ''moderate" NO 3 -N, 4 % had ''bad" to ''moderate" PO 4 -P, and the total phosphorus values did not correspond to the ''good" or ''very good" ecological status classes in 4% of the tested water bodies; 2.
River basins accumulate the biggest quantities from the following sources: transit pollution, contributing 87,599 t/year of total nitrogen and 5020 t/year of phosphorus; agricultural pollution, contributing 56,031 t/year of total nitrogen and 2474 t/year of total phosphorus; 3.
The biggest total nitrogen load in river basins per year is from transit pollution, accounting for 53.89%; agricultural pollution accounts for 34.47%. The highest total phosphorus load is also from transit pollution, accounting for 58.78%; agricultural pollution accounts for 28.97%; 4.
The multiple regression analysis showed that the agricultural activity had the biggest negative influence on the ecological status of rivers according to all studied indicators. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data are available in a publicly accessible repository.

Conflicts of Interest:
The authors declare no conflict of interest.