Effects of topographical factors and soil characteristics on gully properties in Kohgiloye and Boyer Ahmad Provinces, Iran

Gully erosion is unequivocally an important form of soil erosion in the arid and semiarid regions of Iran. These lands, which are the main sources of income for farmers, are being rapidly decimated. Moreover, nomadic livestock grazing in such regions is heavily dependent on these lands. Sediments resulting from gully erosion can engender a plethora of environmental problems, such as water quality problems in waterways,rivers and lakes as well as decrease fertility of farmland. This study focuses on the basic implications of the roles of gully watershed and certain physico-chemical soil properties upon gully erosion in a semi-arid region of 210 km2. In this particular study, the Abgendi watershed, which has the largest number of gullies, was selected. In this area, all the gullies (53 gullies) were coded and 35 gullies were chosen randomly (using random numbers in Excel, function of random between) and studied. Each of these gullies has a small watershed, whose properties such as gully volume, gully length, slope above gully, distance between head-cut and borderline, altitude difference, gully watershed area and soil cover were determined. From the main waterway of the gully, soil sampling from head-cut and two walls in two depths (0-30 cm and 30 cm to bottom of gully) was carried out. Physical characteristics of the soil, such as aggregate stability (AS), mean weight diameter (MWD), clay, sand and silt, and soil chemical properties such as electrical conductivity (EC), sodium adsorption ratio (SAR), organic carbon (OC), cation exchange capacity (CEC), calcium (Ca), magnesium (Mg) and sodium (Na) were determined. Moreover, in order to identify the main factors responsible for the development of gully erosion, the gullies were classified according to gully volume in three sizes: big, medium, and small gullies. The three gully volume groups were compared to one another on the basis of their topographical factors such as gully watershed areas, distance between head-cut and borderline, altitude difference of gully watershed area, slope above head-cut and gully bottom slope and then they were compared with the area without gullies (the control group) in terms of their soil’s physico-chemical properties and soil cover such as percentages of vegetation cover, stubble and gravel. The results of the investigation of the soil texture in the gully and control areas indicated that the occurrence probability of gullies is low in soils which tend to have high amount of sand. Soils with silt clay loam texture are the most susceptible types of soils, while sandy loam and loam are the most resistant types to gully erosion in the study area. The soil texture groups and mean values of gully volume and length in soil texture groups in the top and sub-layer of gully head-cut and walls showed that the occurrence probability of gullies in the soil with silt loam and silt clay loam in the top layer, and clay loam and silt clay loam in the sub-layer is high. In contrast, the occurrence probability of gullies in the soil with loam texture is low. In other words, if a gully is formed in soils with clay loam and silt clay loam in the top layer, the gully can be very large and long. The results of the comparison between the gully volume groups and control area indicated that some soil physical properties such as sand and proportion of sand / silt + clay in both top and sub-layers of head-cut (at least one of the gully volume groups),and sand, and proportion of sand / silt + clay in both top and sub-layers of the gully walls (all three volume groups), were significantly lower than those of the same depths in the control area. Furthermore, there was a positive correlation between the percentage of clay in the top and sub-layers of the head-cut and gully volume at the 5% level (2-tailed). Likewise, the proportions of clay / (sand + silt) in the top -layers of the head-cut were correlated positively with the volume and length of gully. A comparison between the gully volume groups and the control area in terms of soil chemical properties showed that, except for EC in the sub-layer of the gully walls (big gully group) that was significantly higher than that in the same depth of the control area, none of the chemical factors was significant. However, there was a significant positive correlation between the volume and/or length of gully and Ca, Mg, EC in the head-cut top layer, EC in the head-cut sub-layer, Mg in the gully walls top layer, and EC and Ca in the gully walls sub-layer at the 5% level (2-tailed). However, comparison between the gully volume groups in terms of soil chemical properties indicated that soil chemical properties of the gully area were approximately uniform. The result of the comparison between gully volume groups in terms of gully watershed properties demonstrated that some topographical factors such as distance between head-cut and borderlines, altitude difference and gully watershed area could explain the different sizes among gully volume groups. The correlation between these topographical factors and gully volume and length was strongly significant at 1% level. The statistical analysis of gully watershed properties not only explained the gully volume and length but also explained the differences among the volume groups in the gully area. The percentage of the vegetation cover in the study area showed that, under these circumstances, not only runoff generation is quickly formed, but raindrops can be quickly dispersed throughout the soil particles as well. This process is attributed to scarcity of soil cover. The result of comparison between control and gully area also showed that the soil covers in the different sizes of gully were uniform, but the vegetation cover of the control area was significantly higher than that in the gully area. In order to develop a model, regressions between the gully length as a dependent variable and other measured factors as independent variables were performed using a stepwise method through SPSS. It should be mentioned that the same process was carried out for the gully volume as the dependent variable and all the measured factors as dependent variable. The gully length model indicated that gully length was affected by the percentages of clay in the top layer of the gully head-cut and walls, gully watershed area and distance between head-cut and borderline. Therefore, these three factors turn out have the greatest impact on gully erosion. Among these effective factors, the most important factor was distance between head-cut and borderline, with a β coefficient of 0.517,whereas the least important factor was clay percentage of the gully walls’ top layer, with a β coefficient of 0.246. On the other hand, the gully volume model indicated that the gully volume was affected by distance between headcut and borderline and clay of the head-cut top-layer. Comparison of the two models revealed that the gully length model could better explain the gully erosion development than the gully volume model for the simple reason that it had stronger R, R2 and adjusted R2.

Preview PDF FP 2013 12R.pdf Download (1MB) | Preview

Actions (login required)

View Item