**6.1 River flow**

122 Studies on Water Management Issues

In the steep vineyards scenario (STV35), all vineyards on the slopes greater than 35 % were changed into forest, to verify the environmental impact of abandonment of vineyards on steep slopes (Fig. 7). In the STV35 scenario, 17 % (Reka) and 1.4 % (Dragonja) of grassland is changed into forest, which is equivalent to 3.93% (River) and 0.06% (Dragonja) of the total

than 35 % for the Reka and Dragonja catchment

the Reka and Dragonja catchment

**6. Results and discussion** 

1 and 3,489 ± 11,742 kg TP year-1.

Fig. 7. Hydrological response units with the vineyard land use (VINE) and slopes greater

the Reka (8 %) and Dragonja (18%) catchments area was turned into a forest (Fig. 8).

Fig. 8. Hydrological response units with the grassland land use (TRAV) and slope classes for

The base scenario indicates a high average annual variability in the transport of the sediment, total nitrogen (TN) and total phosphorus (TP) in the river flow (Table 7). The standard deviations for the Reka subcatchment 8 reveal that the sediment, TN and TP 2/3 of transported quantities are expected in the interval 1,844 ± 1,075 t sediment year-1, 88,728 ± 63,255 kg TN year-1 and 3,489 ± 2,993 kg TP year-1 and for the Dragonja subcatchment 14 in the interval 4,804 ± 1,576 t sediment year t-1, 163,763 ± 98,949 kg TN year-

Extensive grassland scenario (ETA) objective was to determine what would be the impact on water quantity and quality, if the whole grassland would be overgrown with forest. Extensive grassland use with one cutting is widespread in both areas. Whole grassland in

catchments.

Changes in average annual flow between base and agri-environmental scenarios are minimal for both catchments for the research period. Maximum changes on an annual basis are less than 0.5 % (Table 8) and on a monthly basis close to 1% (Reka) and 5% (Dragonja) (Fig. 9). Student t-statistics for average annual flows reveal that the results of the agrienvironmental scenarios are not statistically different from the base scenario (Table 9).


Table 8. Impacts (change in %) of agri-environmental scenarios on the river flow, sediment load, total nitrogen and total phosphorus load in the watercourse; compared to the baseline scenario

Modelling of Surface Water Quality by Catchment Model SWAT 125

Fig. 10. Change in average monthly river loads of sediment (%) between the base (Base = 0)

The effect of agri-environmental scenarios on the annual TP transport in the river flow has proved to be negligibly small, due to the small proportion of land on which the scenarios were set up (Table 8). The results of the agri-environmental scenarios for the TN transport in both catchments are not statistically significantly different from the base scenario (Table 9). Large monthly variations in the loads of TP transported were typical for the scenarios with higher levels of organic matter (EKO20, EKO100, ETA) (Fig. 11). The decomposition of the organic matter is difficult to control, monitor and predict. However, on an annual basis, the

The effects of agri-environmental scenarios on the TP transport in the stream are low (Table 8) and may be observed in scenarios EKO 100 and EVP (Reka) and ETA (Dragonja) (Fig. 12). Student t-statistics for average annual TP load in both catchments are not statistically significantly different (Table 9). In case of Rivers, maximum difference between the scenarios resulting in cooler and wetter period of the year, and in the Dragonja catchment,

Fig. 11. Change in average monthly river loads of total nitrogen (%) between the base (Base = 0) and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja

**-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55**

**Percentage change in TN (%)**

and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja

**-75 -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40**

**Percentage change in sediment (%)**

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

**EVP EKO20 EKO100 S35 S50 STV35 ETA DragonjaBase = 0**

> **EVP EKO20 EKO100 S35 S50 STV35 ETA DragonjaBase = 0**

**EVP EKO20 EKO100 S35 S50 STV35 ETA RekaBase = 0**

**-75 -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40**

**-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55**

**Percentage change in TN (%)**

**Percentage change in sediment (%)**

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

subcatchment 14 (1994−2008)

variation between months are equalized.

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

subcatchment 14 (1994−2008)

in the warmer and more stormy period.

**6.4 Total phosphorus** 

**6.3 Total nitrogen** 

**EVP EKO20 EKO100 S35 S50 STV35 ETA RekaBase = 0**

Fig. 9. Change in average monthly flow (%) between the base (Base = 0) and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja subcatchment 14 (1994−2008)


Note: The results of the scenarios are statistically significantly different from the base scenario, if the value of Student t-test exceeds tα = 2.145. If the value is negative, scenario is reducing the quantities in the river flow, and vice versa.

Table 9. Review of statistically significant results of Student t-statistics for average annual flow and average annual load of sediment, total nitrogen and total phosphorus

#### **6.2 Sediment**

Impacts of agri-environmental scenarios EVP, EKO20, EKO100, S35, S50, STV35, ETA on an average annual load of sediment transported with the flow are evident for certain scenarios (Table 8). Statistically significant changes in the Reka catchment have been calculated for the EKO100 scenario, while the EVP scenario result is slightly lower to be statistically significantly different (Table 9). The river Dragonja results show that changes in the scenarios EVP, EKO20, EKO100 and ETA are statistically significantly different from the base scenario (Table 9). The biggest differences between scenarios in transported sediment load are in autumn and winter months, when the loads for scenarios EKO100 (Reka) and ETA (Dragonja) get considerably reduced (Fig. 10).

Fig. 10. Change in average monthly river loads of sediment (%) between the base (Base = 0) and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja subcatchment 14 (1994−2008)

#### **6.3 Total nitrogen**

124 Studies on Water Management Issues

**Percentage change in river flow (%)**

 **-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10**

agri-environmental scenarios for the Reka subcatchment 8 and Dragonja subcatchment 14

Student distribution of the sample with n-1 degrees of freedom

Table 9. Review of statistically significant results of Student t-statistics for average annual

Impacts of agri-environmental scenarios EVP, EKO20, EKO100, S35, S50, STV35, ETA on an average annual load of sediment transported with the flow are evident for certain scenarios (Table 8). Statistically significant changes in the Reka catchment have been calculated for the EKO100 scenario, while the EVP scenario result is slightly lower to be statistically significantly different (Table 9). The river Dragonja results show that changes in the scenarios EVP, EKO20, EKO100 and ETA are statistically significantly different from the base scenario (Table 9). The biggest differences between scenarios in transported sediment load are in autumn and winter months, when the loads for scenarios EKO100 (Reka) and

flow and average annual load of sediment, total nitrogen and total phosphorus

**Reka – subcatchment 8 Dragonja – subcatchment 14 Scenario Flow Sediment TN TP Flow Sediment TN TP**  EVP 0.000 –1.214 –0.148 –0.712 0.000 **–5.630** –0.122 –0.215 EKO20 0.009 –0.348 0.448 0.389 –0.047 **–3.056** 0.750 0.603 EKO100 0.005 **–2.435** –0.105 –1.439 0.053 **–3.023** 0.209 0.080 S35 0.002 –0.057 –0.023 –0.027 0.000 –0.274 –0.014 –0.019 S50 0.000 –0.004 –0.001 –0.002 0.000 –0.006 –0.001 0.000 STV35 0.018 –0.157 –0.281 –0.112 0.000 –0.001 0.000 0.000 ETA 0.018 –0.216 –0.127 –0.159 0.013 **–14.386** –0.450 –0.594 Note: The results of the scenarios are statistically significantly different from the base scenario, if the value of Student t-test exceeds tα = 2.145. If the value is negative, scenario is reducing the quantities in

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

**EVP EKO20 EKO100 S35 S50 STV35 ETA DragonjaBase = 0**

**EVP EKO20 EKO100 S35 S50 STV35 ETA RekaBase = 0**

Fig. 9. Change in average monthly flow (%) between the base (Base = 0) and

**Student t-test** (Significance level 0.05)

**-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10**

(1994−2008)

the river flow, and vice versa.

**6.2 Sediment** 

**Percentage change in river flow (%)**

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

α=0.025, SP=14, tα =2.145

ETA (Dragonja) get considerably reduced (Fig. 10).

The effect of agri-environmental scenarios on the annual TP transport in the river flow has proved to be negligibly small, due to the small proportion of land on which the scenarios were set up (Table 8). The results of the agri-environmental scenarios for the TN transport in both catchments are not statistically significantly different from the base scenario (Table 9). Large monthly variations in the loads of TP transported were typical for the scenarios with higher levels of organic matter (EKO20, EKO100, ETA) (Fig. 11). The decomposition of the organic matter is difficult to control, monitor and predict. However, on an annual basis, the variation between months are equalized.

Fig. 11. Change in average monthly river loads of total nitrogen (%) between the base (Base = 0) and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja subcatchment 14 (1994−2008)

#### **6.4 Total phosphorus**

The effects of agri-environmental scenarios on the TP transport in the stream are low (Table 8) and may be observed in scenarios EKO 100 and EVP (Reka) and ETA (Dragonja) (Fig. 12). Student t-statistics for average annual TP load in both catchments are not statistically significantly different (Table 9). In case of Rivers, maximum difference between the scenarios resulting in cooler and wetter period of the year, and in the Dragonja catchment, in the warmer and more stormy period.

Modelling of Surface Water Quality by Catchment Model SWAT 127

sufficiently enriched with organic matter and nutrients to supply plants for a several decades (Mihelič et al., 2009). In the Dragonja catchment, which is subject to a high degree of afforestation, the scenario EVP reflected in the significant concentration reduction below the recommended value. We used 3 meters wide vegetation bands that have reflected a 14 % (Reka) and 31 % (Dragonja) reduction of sediment in the watercourse, but with broader bands, an even greater impact could be achieved. For the effectiveness of the bands, the identification of critical points is important (Garen & Moore, 2005; Wolfe, 2000). A small proportion of the area can have a significant impact on the sediment, N and

> DRAGONJA – subcatchment 14 (Podkaštel 9300) - cyprinid river

**Average annual concentration (mg l-1)** 

**Sediment Nitrate TP Sediment Nitrate TP** 

**Measured 29.3 2.7 0.043 32.6 2.7 0.109**  EVP 19.9 2.6 0.042 **28.8** 2.7 0.100 EKO20 23.2 2.7 0.045 **31.1** 2.5 0.121 EKO100 23.2 2.6 0.044 **27.6** 2.9 0.092 S35 **28.6** 2.7 0.043 **32.2** 2.7 0.108 S50 **29.3** 2.7 0.043 **32.6** 2.7 0.109 STV35 **29.3** 2.7 0.043 **32.1** 2.6 0.108

ETA 13.8 2.6 0.039 **30.6** 2.5 0.104 Limit and guide concentrations (mg l-1) set by EU directives and Slovenian regulations: **Sediment** (river) **25 mg l-1**; **Nitrate (NO3-)** in drinking water **50 mg l-1** and in surface water **14,08 - 30,8** (very good state)

Table 10. Impacts of the alternative scenarios on the average annual concentration (mg l-1) of

Following the trend of afforestation of agricultural land, the ETA scenario could become practicable, under which all grassland (18 %) would be overgrown by forest. However, such a scenario is not viable, since larger farmers round up their vineyards and olive groves and reduce overgrowth. However, this process is considerably slower than natural afforestation, which has affected the water cycle and erosion processes in the last decade (Globevnik, 2001). Sediment reductions in the catchment are expected with progressive land abandoned with afforestation and with parallel establishment of buffer zones on larger agriculturally rounded areas. The negative effect of erosion buffer zones is an exclusion of a certain percentage of agricultural land from agricultural production. At 3 m wide buffer zones on 1 ha of land (10,000 m2) the loss of the land in production would be 12 % (1,200 m2). An important element, which partially contributes to increased sediment loads in the river Dragonja are cliffs and steep eroded slopes without vegetation, which are eroded at the

**1** (good state); **Total phosphorus (TP)** for salmonid waters **0,2 mg l-1** and for

REKA – subcatchment 5 (Neblo 8700) - salmonid river

P loads in the watercourses.

Scenarios

and **28,6 - 41,8 mg l-**

cyprinid waters **0,4 mg l-1**.

the sediment, nitrate and total phosphorus

foothills by the river and torrential tributaries.

Fig. 12. Change in average monthly river loads of total phosphorus (%) between the base (Base = 0) and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja subcatchment 14 (1994−2008)

#### **6.5 Scenario evaluation**

The evaluation of impacts of the agri-environmental scenarios on the sediment and nutrients transport processes on the catchment level was performed in the light of the EU Water Framework Directive (WFD 2000/60/ES) and Republic of Slovenia legislation. Both set guide concentrations with the purpose of limiting impacts of excessive levels on flora and fauna in the rivers. When interpreting the concentrations we need to have in mind the geological and pedological characteristics of the catchment. There is also the question of whether to consider set guide levels for the rivers that do not represent an economic interest (Lohse, 2008), however rivers are not only economic asset. When recommending possible agri-environmental mitigation measures to deliver water quality improvements, careful evaluation and prioritization of each measure has to be performed according to its positive and negative issues on the environment, agriculture, social life and economy (Bockstaller et al., 2009; Everard, 2004; Glavan et al., 2011).

The results of the scenarios demonstrate that in the Reka and Dragonja catchments major problems with the concentrations of NO3- and TP are excluded, as both are lower than the limit values (Table 10). Nevertheless, the results reveal the difficult path to achieve the recommended value for sediment in both catchments, especially in the case of the river Reka catchment. With the realization of agri-environmental scenarios for the Dragonja catchment, particularly the EVP and ETA, we could expect reduction of the sediment concentration below the recommended level and consequently water quality improvements. In the Dragonja catchment, the guide concentration of 25 mg l-1 was reached with the scenarios EVP, EKO20, EKO10 and ETA. However, in the Reka catchment, scenarios sediment reductions are not sufficient to reduce the concentration below the guide level. This leads us to thinking, that catchment is dominated by certain land use (vineyard) and soils, which have a negative impact on the river concentrations (Komac & Zorn, 2007; Petek, 2007; Volk et al., 2009).

The EKO100 scenario is considering the low proportion of land involved in organic production in research areas almost impracticably, since it would require too much labour-intensive work, which results in a higher final price of the crop. Organic production is advised in the areas with long-term organic fertilization where soils were

The evaluation of impacts of the agri-environmental scenarios on the sediment and nutrients transport processes on the catchment level was performed in the light of the EU Water Framework Directive (WFD 2000/60/ES) and Republic of Slovenia legislation. Both set guide concentrations with the purpose of limiting impacts of excessive levels on flora and fauna in the rivers. When interpreting the concentrations we need to have in mind the geological and pedological characteristics of the catchment. There is also the question of whether to consider set guide levels for the rivers that do not represent an economic interest (Lohse, 2008), however rivers are not only economic asset. When recommending possible agri-environmental mitigation measures to deliver water quality improvements, careful evaluation and prioritization of each measure has to be performed according to its positive and negative issues on the environment, agriculture, social life and economy (Bockstaller et

The results of the scenarios demonstrate that in the Reka and Dragonja catchments major

limit values (Table 10). Nevertheless, the results reveal the difficult path to achieve the recommended value for sediment in both catchments, especially in the case of the river Reka catchment. With the realization of agri-environmental scenarios for the Dragonja catchment, particularly the EVP and ETA, we could expect reduction of the sediment concentration below the recommended level and consequently water quality improvements. In the Dragonja catchment, the guide concentration of 25 mg l-1 was reached with the scenarios EVP, EKO20, EKO10 and ETA. However, in the Reka catchment, scenarios sediment reductions are not sufficient to reduce the concentration below the guide level. This leads us to thinking, that catchment is dominated by certain land use (vineyard) and soils, which have a negative impact on the river concentrations (Komac & Zorn, 2007; Petek, 2007; Volk

The EKO100 scenario is considering the low proportion of land involved in organic production in research areas almost impracticably, since it would require too much labour-intensive work, which results in a higher final price of the crop. Organic production is advised in the areas with long-term organic fertilization where soils were

Fig. 12. Change in average monthly river loads of total phosphorus (%) between the base (Base = 0) and agri-environmental scenarios for the Reka subcatchment 8 and Dragonja

**-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50**

**Percentage change in TP (%)**

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

and TP are excluded, as both are lower than the

**EVP EKO20 EKO100 S35 S50 STV35 ETA DragonjaBase = 0**

**-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50**

**Percentage change in TP (%)**

**EVP EKO20 EKO100 S35 S50 STV35 ETA RekaBase = 0**

**1 2 3 4 5 6 7 8 9 10 11 12 Month**

al., 2009; Everard, 2004; Glavan et al., 2011).

problems with the concentrations of NO3-

et al., 2009).

subcatchment 14 (1994−2008)

**6.5 Scenario evaluation** 

sufficiently enriched with organic matter and nutrients to supply plants for a several decades (Mihelič et al., 2009). In the Dragonja catchment, which is subject to a high degree of afforestation, the scenario EVP reflected in the significant concentration reduction below the recommended value. We used 3 meters wide vegetation bands that have reflected a 14 % (Reka) and 31 % (Dragonja) reduction of sediment in the watercourse, but with broader bands, an even greater impact could be achieved. For the effectiveness of the bands, the identification of critical points is important (Garen & Moore, 2005; Wolfe, 2000). A small proportion of the area can have a significant impact on the sediment, N and P loads in the watercourses.


Limit and guide concentrations (mg l-1) set by EU directives and Slovenian regulations: **Sediment** (river) **25 mg l-1**; **Nitrate (NO3-)** in drinking water **50 mg l-1** and in surface water **14,08 - 30,8** (very good state) and **28,6 - 41,8 mg l-1** (good state); **Total phosphorus (TP)** for salmonid waters **0,2 mg l-1** and for cyprinid waters **0,4 mg l-1**.

Table 10. Impacts of the alternative scenarios on the average annual concentration (mg l-1) of the sediment, nitrate and total phosphorus

Following the trend of afforestation of agricultural land, the ETA scenario could become practicable, under which all grassland (18 %) would be overgrown by forest. However, such a scenario is not viable, since larger farmers round up their vineyards and olive groves and reduce overgrowth. However, this process is considerably slower than natural afforestation, which has affected the water cycle and erosion processes in the last decade (Globevnik, 2001). Sediment reductions in the catchment are expected with progressive land abandoned with afforestation and with parallel establishment of buffer zones on larger agriculturally rounded areas. The negative effect of erosion buffer zones is an exclusion of a certain percentage of agricultural land from agricultural production. At 3 m wide buffer zones on 1 ha of land (10,000 m2) the loss of the land in production would be 12 % (1,200 m2). An important element, which partially contributes to increased sediment loads in the river Dragonja are cliffs and steep eroded slopes without vegetation, which are eroded at the foothills by the river and torrential tributaries.

Modelling of Surface Water Quality by Catchment Model SWAT 129

7. Physical landscape spatial variability within catchments (topography, soils, land use, land management etc.) have important influence on the model results. This means that pollutant sources and loads are not evenly distributed in space. Rather than impose blanket agri-environmental measures in the model, it is better to target key source areas

8. The scenarios assume that all farmers in the catchment take up the structural measures or the changes in land use and management uniformly. However, field work shows that this is not the case. A close cooperation with all key stakeholders on local, regional, national and transnational level and financial support, like EU Common Agriculture

At the end of this chapter we would like to increase awareness that model results and their interpretation by the modeller must lead to constructive discussion, which aims to achieve and maintain good water quality in research catchments, which is the objective of the Water

Financial support for this study was provided by the Slovenian Research Agency founded

Abbaspour, K.C.; Yang, J.; Maximov, I.; Siber, R.; Bogner, K.; Mieleitner, J.; Zobrist, J. & R.

Bockstaller, C.; Guichard, L.; Makowski, D.; Aveline, A.; Girardin, P. & Plantureux, S. (2009).

Bowatte, S.; Tillman, R.; Carran, A. & Gillingham, A. (2006). Can phosphorus fertilisers

Bracmort, K.S.; Arabi, M.; Frankenberger, J.R.; Engel, B.A. & Arnold, J.G. (2006). Modelling

Buda, A.R.; Kleinman, P.J.A; Srinivasan, M.S.; Bryant, R.B. & Feyereisen, G.W. (2009). Effects

Čarman, M.; Mikoš, M. & Pintar, M. (2007). Različni vidiki erozije tal v Sloveniji = Different

Thur watershed using SWAT. *Journal of Hydrology,* Vol.333, pp. 413-430 Arnold, J.G.; Srinivasan, R.S.; Muttiah, R.S. & Williams, J.R. (1998). Large area hydrological

Srinivasan (2007). Modelling hydrology and water quality in the pre-alpine/alpine

modelling and assessment Part I: Model development. *Journal of the American Water* 

Agri-environmental indicators to assess cropping and farming systems – A review.

alone increase levels of soil nitrogen in New Zeland hill country pastures? *Nutrient* 

long-term water quality impact of structural BMPs. *Agricultural Society of* 

of Hydrology and Field Management on Phosphorus Transport in Surface Runoff.

aspects of soil erosion in Slovenia. In: *Strategija varovanja tal v Sloveniji*: *Zbornik referatov,* M. Knapič, (Ed.), 39–50, Pedološko društvo Slovenije = Slovenian Soil

Policy, which enable areas to develop in a sustainable way, is necessary.

by the Government of the Republic of Slovenia. Contract number: 1000-06-310163.

account for the nutrient lag times in the groundwater.

or HRU combinations that deliver excessive loads.

Framework Directive and other legislation related to water.

*Resources Association*, Vol.34, No.1, pp. 73–89

*Cycling in Agroecosystems*, Vol.75, pp. 57–66

science Society, Ljubljana

*Agronomy for Sustainable Development*, Vol.28, pp. 139–149

*Agricultural and Biological Engineers*, Vol.49, No.2, pp. 367–374

*Journal of Environmental Quality*, Vol.38, pp. 2273–2284

**8. Acknowledgments** 

**9. References** 

in the model, which can lead to appropriate modelling of nutrients pathways and to

To achieve improvements in water quality in the two research catchments the use of a combination of several measures and a close cooperation with all key stakeholders (environmental, agricultural, spatial planning) would be necessary.
