**4. Discussion**

The previous research study conducted by Hussain et al. [14] in the Dhrabi River Catchment indicated that erratic and intensive rainfall during the rainy season generated several peak runoff events, exposing the steep sloped areas to potentially severe soil erosion. For example, in Catchment-25, a total 400 mm rainfall was accumulated from eleven erosive rainstorms in 2009, where a maximum of 108 mm day−<sup>1</sup> rainstorms generated 46.2 mm runoff and a 6.86 t ha−<sup>1</sup> sediment yield. Similarly, the total soil loss during the 2010 investigation period was 31.13 t ha−<sup>1</sup> [14]. In the SWAT, the erosive impact of rainfall is generally estimated in terms of peak runoff generation, so the results obtained during calibration and validation are represented in Figure 6 for surface runoff and sediment yield. The analysis was performed for each total rainfall event and the respective total surface runoff and sediment yield generated by each event. The overall statistical results indicated that the performance of the SWAT was satisfactory and that the simulated values generally matched the corresponding observed values well. However, model adequacy should be further evaluated by how well the model captures high and low rainfall events, specifically regarding the replication of fluctuations in

the resulting hydrographs and sediment yields. The graphical results (Figure 6) revealed that the SWAT was able to satisfactorily reproduce most of the low flow and sediment yield events (due to low rainfall events), although some relatively low sediment yields were considerably overpredicted, e.g., sediment yield events on 7 August 2011. In contrast, it was also found that the SWAT typically underestimated or overestimated high flow and sediment yield events in response to high rainfall events. For example, a maximum intensity rainstorm on 29 July 2010 resulted in the overestimation of surface runoff and sediment yields, while another maximum intensity rainstorm on 29 July 2009 resulted in underestimations. These discrepancies may have occurred due to inaccuracies in observed climate, runoff, and sediment data, such as some of the rainfall events, not being measured properly; this, in turn, could have led to underestimations or overestimations of runoff peaks. Another possible reason could be related to short, rapid rainfall events, which could have led lead to an overestimation discrepancy because small catchments have low times of concentration and thus a low capacity to minimize peak runoff. In addition, the CN technique cannot accurately predict runoff for days that experience several storms. The underestimation and/or overestimation of sediment yield was also observed during high intensity rainstorm events, which may have been due to uncertainties in runoff simulation measurements, as well as uncertainties in model parameterization. This may have also been due to the observed data used for model calibration and validation. Relatively short term events with several storms having high intensity may not have been captured well by the sampling of sediment data, including inaccurately high loads being measured during short term events, which led to an overestimation in sediment yield. The literature data findings indicated that the semi-mountainous region of Pothwar is rainfed and soil erosion is a serious issue due to the steep slope and heavy rainfall. The calibration and validation of the SWAT was successfully performed using parameters mentioned in Table 2 for surface runoff and sediment yield. Similar studies, such as the sediment simulation results by Betrie et al. [61] reported good agreement between the model daily sediment predictions and the observed concentrations at the El Diem gauging station (Ethiopia–Sudan border). SWAT studies for smaller watersheds in the northeast and northwest of Ethiopia have tended to show weaker hydrologic results [73,74], which is an indication that it may be difficult to accurately represent the processes and obtain better results for smaller watersheds. Keeping in view the literature studies, the validated SWAT model was applied to small watersheds of the Pothwar area with and without soil and water conservation structures.

The calibrated and validated model parameters were adopted for ungauged small watersheds for the simulation of sediment yield without the consideration of soil and water conservation structures. The soil and water conservation structures were modeled with the modification of appropriate parameters, and then the effectiveness of these structures in terms of reduction of soil erosion was calculated. The results showed that the soil and water conservation structures constructed by the farmers and the SAWCRI department reduced the soil losses in the small watersheds of the Pothwar region. The results showed that the watersheds in the case of without soil and water conservation structures had higher sediment losses than the watersheds with soil and water conservation structures, given similar climatic and land use patterns. The intervention of soil and water conservation structures measures by the mobilization of the community has a significant soil loss reduction to protect their land from the rainfall-driven soil erosion. To the best of our knowledge, no one has reported the effectiveness of soil and water conservation structures for the reduction of soil erosion in the Pothwar region or in Pakistan. This is the first study in this area where the SWAT model has been used for the evaluation of effectiveness of soil and water conservation structures for soil erosion control. For this purpose, appropriate parameters responsible for soil erosion were modified according to the type of soil and water conservation structures, as performed by different researchers in literature such as Betrie et al. [61] in the Upper Blue Nile River basin, Gebremichael et al. [66] in the northern part of Ethiopia, and Melaku et al. [38] in the Gumara Maksegnit watershed in northwest Ethiopia. The SWAT model has been found to be a useful tool for understanding the hydrologic processes and the sediment dynamic in the study area watersheds, and it assessed the impacts of soil and water conservation

structures on the erosion process. It was observed that severe erosion led to higher soil losses in some watersheds dominated with gullies, such as the Khokar Bala site. This is substantiated by the photo taken in Figure 7, which shows the development of deep gullies in the upper parts of the watershed that were found to contribute higher soil erosion losses and to generate higher sediment load at the outlet.

The model results indicated that soil and water conservation structures might considerably decrease soil loss by 40–90% in small watersheds of the Pothwar region. Herweg and Ludi conducted a study at the plot scale in the Eritrean highlands of Ethiopia and reported 72–100% sediment yield reductions engendered by stone bunds [63], which was close to the current finding of the soil loss reduction level due to soil and water conservation structures. Similarly, Betrie et al. [61] indicated 6–69% sediment reductions in the Upper Blue Nile River basin caused by stone bunds, and their results were in agreement with our findings. The average annual sediment yield estimated by the SWAT model without the consideration of soil and water conservation structures in all selected sites was in the range of 8.05–30.31 t ha−1. Our findings were in agreement with other studies conducted in the Pothwar region such as those from Hussain et al. [14], who estimated the annual sediment loss (ranged from 2.6 to 31.1 t hm<sup>−</sup>2) in small watersheds of the Dhrabi River Catchment, while Ahmad et al. [16] reported annual soil loss rates of 9–26 t ha−<sup>1</sup> in the Fateh Jang watershed (Attock) with a slope of 1–10%. Similarly, Nasir et al. [11] predicted an annual soil loss that ranged from 0.1 to 28 t ha−<sup>1</sup> at the small mountainous watershed of Rawal Lake in Rawalpindi. The literature data findings indicated that the semi-mountainous region of Pothwar is rainfed and soil erosion is a serious issue due to steep slope and heavy rainfall. The region receive erratic rainfall during a short rainy season and almost 70% precipitation occurs during monsoon [75,76]. The comparative studies on sediment and soil loss confirmed the results of the current study conducted in small watersheds of the Pothwar region.

The soil and water conservation structures are effective measures in reducing the soil erosion problems in the Pothwar region that have varied land slopes. This study reveals that considering the topographic conditions, loose stone soil and water conservation structures should be installed in areas with a slope range of 0–10%, and wire-meshed stone structures should be installed in areas with a slope range of 6–10%. Proper energy dissipation arrangements should be implemented to prevent downstream erosion.

#### **5. Conclusions**

In this research, we performed SWAT watershed modeling to describe the driving hydrological and sediment transport related processes of small watersheds. The effectiveness of soil and water conservation structures for soil erosion control was assessed with a SWAT model. Stone bund-type structure interventions were done in the model through modification of the USLE support P-factor, the curve number, and the SLSUBBSN. The model results revealed that a 40–90% sediment yield reduction could be achieved using soil conservation structures. Thus, soil and water conservation structures are effective options for soil erosion control in rainfed areas. The land use change scenario results revealed that vegetation cover facilitated sediment yield reduction, in addition to soil conservation structures. An all-inclusive interpretation of the quantitative model results may be misleading because no model can fully simulate all the physical processes of soil and water interactions in a real sense. Some assumptions were made during modeling; however, based on the results, we suggest to policymakers and planners that more than 60% of the area in the Attock and Chakwal districts has potential for soil and water conservation structures.

**Author Contributions:** G.N. and F.H. contributed equally to this research. Conceptualization, G.N. and F.H.; data curation, F.H. and R.B.; formal analysis, F.H., R.-S.W., and V.N.; investigation, R.-S.W.; methodology, G.N.; project administration, R.B.; resources, V.N.; software, F.H.; supervision, G.N.; visualization, R.-S.W., V.N., and R.B.; writing—original draft, G.N. and F.H.; writing—review and editing, F.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** This study is part of a research project under the Consultative Group for International Agricultural Research-CGIAR Research Program (CRP) on Dryland Systems. University of Engineering and Technology, Centre for Excellence in Water Resources Engineering, Lahore, Pakistan, The International Center for Agriculture Research in the Dry Areas, Syria, country office Pakistan and Soil and Water Conservation Research Institute, Chakwal, Pakistan performed collaborative research work. The authors particularly thank all colleagues involved in the fieldwork.

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