Evaluation of soil erosion effects on rainfed wheat (Triticum aestivum) yield using SWAT model

Although soil erosion in natural ecosystems takes place very slowly, yet its cumulative effects on soil productivity are significant in the long term, and even very low rates of erosion cause loss of considerable amounts of soil during the course of many years. Based on statistics presented during the past 40 years, about 30% of agricultural lands in the world have lost their fertility due to erosion and have turned into non-arable lands (Pimentel and Burgess, 2013), and approximately 75 million tons of fertile agricultural soil is lost every year (Eswaran et al., 2001). In another report, it has been stated that about 10 million hectares of agricultural lands in the world lose their productive capacity annually and become non-cultivable lands (Lal, 2001).
SWAT model was used to simulate the effect of soil erosion on dry wheat yield. For this purpose, after model calibration and validation, the following five scenarios were defined to quantify the effects of erosion on rainfed wheat yield: (1) current soil depth, (2) removing 5 cm, (3) removing 10 cm, (4) removing 15 cm, (5) and removing 20 cm of current topsoil depth in soil database.Accordingly, a regression relation was established between surface soil erosion and wheat yield.
Regression analysis showed that for each centimeter of soil erosion, the yield of rainfed wheat was reduced by 15.1 kg equal to 0.5%. The model results showed that the average specific erosion in the agricultural lands of the region is about 10 t/ha (0.77 mm/year). Assuming that the erosion rate is constant over 100 years, about 77 mm of soil is destroyed, equivalent to a loss of 116 kg/ha of dryland wheat yield. Considering the area under wheat cultivation (about 50,000 ha) in the study region, it is estimated that 580 tons of wheat per year will be destroyed by erosion.
SWAT model was used to assess the effect of erosion on crop yield. The gradual decrease in horizon A depth of input soil data to the model showed that with a decrease in the soil depth per 10 cm, the average yield decreased by 5%. It is suggested that in subsequent studies, the results of this method can be compared and validated with the comparative plot method, as the best field method in evaluating the effect of erosion on yield.