Soil erosion risk assessment in Moridhal watershed of Assam, India
using Universal Soil Loss Equation and GIS

doi:10.21203/rs.3.rs-3129092/v2

Download PDF

Research Article

Soil erosion risk assessment in Moridhal watershed of Assam, India using Universal Soil Loss Equation and GIS

https://doi.org/10.21203/rs.3.rs-3129092/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this older preprint version

Read the latest preprint version →

Soil erosion is one of the major factors affecting sustainability of agricultural production in watershed. The objective of this paper is to estimate soil loss using the universal soil loss equation (USLE) model and GIS and to suggest soil conservation practices in Moridhal watershed. Soil loss was estimated by USLE. In addition, measurements of randomly selected soil and water conservation structures were done at four physiographic units of watershed. The erodibility of the studied soils was assessed by computing various erodibility indices like clay ratio, silt clay ratio, modified clay ratio, dispersion ratio, erosion ratio and erosion index. The soil loss of watershed was varied from very slight to very severe (range 0.87-67.77 t ha^-1 yr^-1). Among the physiographic units, the soil loss in the upper piedmont plain area was moderately severe to very severe with a value varying from 19.9-67.8 t ha^-1 yr^-1. The dispersion ratio of the soils varied from 0.06 to 1.18. It was observed that 48.82 per cent of the total studied soil samples had dispersion ratio values above 0.15 which may be considered as erodible. The erosion ratio and erosion index of studied soils varied from varied from 0.01-0.55 and 0.03-0.71, respectively. Based on the study it was found that the soils in different physiographic units are suffering from varying degree of soil erosion. The upper piedmont plain areas of the watershed require management practices which could improve the erosion status and soil productivity. The results of soil loss obtained from the study could be effectively used by the planners, watershed managers and policy makers to formulate site specific conservation plans in order to minimize soil loss for sustaining productivity.

Soil loss

USLE model

GIS

and Watershed

Soil erosion is a complex dynamic process which leads to detachment of soil particle from one place to another place by various agents like wind and water. Soil erosion is a major agricultural problem and also one of the major global environmental issues (Parveen and Kumar., 2012). Water-induced erosion involves the process of detachment and transportation of soil by the impact of raindrop and flowing water and subsequently its accumulation to a new area (Wischmeier and Smith, 1978). In India about 90 per cent soil erosion is caused due to water erosion (Saroha, 2017). Researchers involved in soil erosion studies for a long time have developed many models for estimating the soil loss. However, the Universal Soil Loss Equation (USLE) is being considered as the most widely used method to predict long-term rates of soil erosion from field or farm units. The USLE model (Wischmeier and Smith, 1978) computes soil loss based on the product of rainfall erosivity (R), soil erodibility factor (K), slope steepness (LS), cover management parameter (C) and support practice parameter (P). Revised USLE (RUSLE) is an improved version of USLE which include more diverse databases (Renard et al., 1997). Models like Water Erosion Prediction Project (WEPP) and Soil and Water Assessment Tool (SWAT) are being used time to time with its own set of traits and application possibilities (Baigorria et al., 2017; Alitane et al., 2022). Das et al., (2020) observed that GIS-based multi-criteria decision analysis (MCDA) such as Analytical Hierarchical Process (AHP) approach is a useful tool to help decision-makers in obtaining suitable solutions and sustainable ecosystem management. Deka et al., (2011) integrated the USLE factors and GIS to map the erosional soil loss in the Ghiladhari watershed of Assam, India. Parveen & Kumar (2012) used USLE model and GIS for assessing soil loss risk in Upper South Koel Basin of Jharkhand, India and estimated that the annual soil loss in the watershed was 12.2 t ha^-1 yr^-1. Obi Reddy et al., (2016) assessed soil erosion in tropical ecosystem of Goa, India using USLE and found that about 53 per cent of total geographical area of the state was subjected to severe (20-40 t ha^-1 yr^-1), very severe (40-80 t ha^-1 yr^-1) and extremely severe (>80 t ha^-1 yr^-1) soil erosion. Rao et al., (2018) carried out soil erosion assessment in Godavari Basin of South India using USLE and observed an annual soil erosion of 4.34 t ha^-1 yr^-1 in that area during the year 2002 to 2007. Srinivasan et al., (2019) found that 90.90 per cent (22,330 ha) of the Coastal area of Eastern India falls under very low erosion category (<5 t ha^-1 yr^-1), which may be due to level topography and regular vegetation cover. Balakrishna and Balakrishna, (2019) assessed the soils of Hemavathy river basin of Southern India and found that the average annual soil loss of the river basin was 8.00 t ha^-1 yr^-1 which showed moderate soil loss in that area. Tsegaye et al., (2020) carried out soil erosion assessment in Genda-wuha watershed, Ethiopia using Universal Soil Loss Equation with GIS approach. Borgohain et al., (2021) developed the soil erosion risk map of Pabho watershed of North Eastern India using USLE model and GIS technology. The soil erosion risk map could serve as an efficient tool to help decision makers for evolving conservation strategies in the Matmata mountainous regions of South Eastern Tunisia where farmlands are vulnerable and extremely scarce (Jemai et al., 2021). Based on calibration and validation of soil erosion data for the years of 2010 to 2015. Das et al., (2021) concluded that the Revised USLE model can be extrapolated to other watersheds of NE India having similar landforms, soil, and land uses. Chakrabortty et al., (2022) used universal soil loss equation (USLE) and geographical information system (GIS) to evaluate the importance of morphotectonic parameters on soil erosion in Kangsabati river basin of Eastern India. Cetin et al., (2022) used remote sensing data of Landsat and Sentinel data for studying the temporal variations in the vegetation cover of Kırka boron mining site and, thereby, concluded that the Sentinel images are capable of giving more detail information due to its higher spatial resolution. Kumar et al., (2022) found that for improving the accuracy of USLE model in soil erosion mapping of the Himalayan Indian basin, the C factor has to be adjusted based on derived remote sensing maps. Soil erodibility is an indicator of the detachment and transportation of soil particles by rainfall and runoff. It is basically related to various soil properties like sand, silt, clay, organic matter; structure, permeability, water holding capacity, available water content etc. Erodibility indices are important parameters that can be used to describe the intensity of the soil erosion problems. Numbers of erodibility indices have been suggested by various researchers. Some of the efficient indices of soil erodibility are: water stable aggregates, dispersion ratio (Middleton, 1930), clay ratio (Buoyoucos, 1935), modified clay ratio (Kumar et al., 1995) and erosion ratio (Middleton, 1930). Soil erosion is one of the severe problems for the agricultural development threatening entire north eastern region (NER) of India including Assam state. The Brahmaputra valley of Assam is comprised of undulating terrains and, thereby, faces severe problem of soil erosion. The most common type of soil erosion in Assam is loss of topsoil through surface run-off under heavy precipitation which causes decline in productivity of agricultural land. Moridhal watershed, located in the north bank plains of Assam, India has been reported to be erosion prone by the Department of Soil Conservation, Assam. The district suffers from severe erosion hazard amounting 8.64 million ha which is 2.67% of the total geographical area of the district (Dekaraja and Mahanta, 2022). This watershed has a niche of different physiographic units and the soils have different inherent characteristics. However, information on the erosional nature of the soils of watershed is limited. In view of this, the present investigation was carried out to assess soil loss of Moridhal watershed as well as to study the relation of soil loss with the erodibility indices by using USLE model and GIS techniques.

Site description

Moridhal watershed, located in Dhemaji district of Assam, is a part of the North Eastern India (Fig. 1). The watershed is situated between 94°52’ E to 94°69’ E longitude and 27°38’ N to 27°64’ N latitude. Geocoded FCC of Resourcesat-2 LISS-IV data of 2018 data were visually interpreted and in conjunction with Survey of India toposheets (1:50,000) the physiographic map was prepared. Visual image interpretation based on colour, tonal and textural variations has led to the recognition of four distinct physiographic units which include upper piedmont plain, lower piedmont plain, alluvial plain and flood plain (Fig. 2). The watershed covers 30,730 ha area among which 1,844 ha (6.0 per cent) area was under upper piedmont plain; 2,391 ha (7.78 per cent) by lower piedmont plain; 9,888 ha (32.17 per cent) and 16,607 ha (54.04 per cent) by flood plain and alluvial plain, respectively. The study area experienced humid sub-tropical climate with an average annual rainfall of 3064 mm. The mean annual temperature was more than 23.6°C and the differences between mean summer temperature (26.2°C) and mean winter temperature (23.0°C) was more than 5°C. Hence, the studied area qualified for hyperthermic soil temperature regime. The soil moisture regime of the area was found to be Udic. GPS based 170 numbers of both surface and core soil samples were collected up to a depth of 0-25 cm from each physiographic unit at a regular interval of 2 km (Fig. 3). The number of surface soil samples representing different physiographic units comprises: 19, 50, 26 and 75 from upper piedmont plain, lower piedmont plain, alluvial plain and flood plain, respectively. The sample collection was done during January to February of 2018 and the entire study was completed by the end of 2020.

The collected surface samples were air dried, grinded and passed through 2 mm sieve. The samples were analyzed for various soil parameters which include: Sand, silt and clay by International pipette method (Piper, 1966), organic carbon (Walkley and Black, 1934), Water holding capacity (Piper, 1966), Available water content (Richards, 1948), Soil aggregate (Yoder, 1936) comprising macroaggregate (>0.25 mm diameter) and microaggregate (<0.25 mm diameter). The soil permeability classes were determined from hydraulic conductivity of core samples as estimated by constant head method (Klute, 1965). The soil structure codes were assessed by observing the type, grade and size of soil aggregates in the field. Universal Soil Loss Equation (USLE) model given by Wischmeier and Smith, (1978) was used for estimation of soil loss in the watershed area. This model had the advantage that while estimating soil loss due weight age is given to all the major input parameters like intensity and amount of rainfall, soil erodibility, land use/land cover and slope. USLE can be expressed as: A = R × K × LS × C × P; where, A is the mean annual soil loss (t ha^-1 yr^-1), R is the Rainfall erosivity factor (MJ mm ha yr), K is the Soil erodibility factor (t ha MJ^-1 mm^-1), LS is the Topography factor, C is the Cover and management factor and P is the Conservation practice factor. Rainfall erosivity factor (R) was calculated following the expression given by Bergsma (1980). The derived relationship is: R = 0.1059 a.b.c + 52; where, a is the average annual precipitation, b is the maximum 24 hrs precipitation with recurrence interval of two years and c is one hour maximum precipitation with recurrence interval of two years. The monograph developed by Wischmeier, (1971) was employed to estimate K values in the study area using the following equation:

100 K = 2.1 x 10^-4 x M^1.14 (12-a) + 3.25 (b-2) + 2.5 (c-3)

Where, K is the soil erodibility factor, M is the per cent silt + per cent very fine sand, as is the per cent organic matter, b is the soil structure code and c is the soil permeability code. The LS factor for each physiographic unit was determined based on the degree and length of slope using Survey of India Topographical Map on 1:50,000 scale and by following monograms (Wischmeier and Smith, 1978). C factors were computed on the basis of main crop grown/land use collected during the field survey. The conservation practice factors (P) were selected based on field survey information. The guidelines for assessing ‘P’ values for different land utilization types were taken from published literature (Singh et al., 1981; Potdar et al., 2003). The erodibility of the studied soils was determined by computing three widely used erodibility indices which are: (i) Dispersion ratio (Middleton, 1930) = [% Undispersed (silt + clay)]/ [% Dispersed (silt + clay)]; (ii) Erosion ratio (Middleton 1930) = Dispersion ratio/ [Clay (%) /Available water] and (iii) Erosion index (Sahi et al., 1977) = Dispersion ratio/ [Clay (%) /0.5 Water holding capacity]. The soil loss map was prepared under GIS environment using their numerical values. For this, the computed values of all the sample sites were digitized from the location map using ArcMap 10.4. For unsampled location the soil parameter values were estimated using Inverse Distance Weighted (IDW) function as described by Jensen, (1986). The IDW interpolation technique was used in the study as this geostatistical technique has been frequently used by various authors for estimating surface maps and predicting soil properties (AbdelRahman et al., 2020). The interpolated maps were reclassified to get the map units and legends. The boundary of Moridhal watershed was then marked over the interpolated maps to get the final map. Principal factor/component analysis as described by Mohanty and Filipovski, (1982) was carried out using SPSS-21 software statistical package. Altogether 12 soil erosion related parameters along with three erodibility indices and soil loss were used in the analysis. The analysis provided the important statistical parameters which include: Correlation matrix, Variance-covariance matrix, Eigen values with corresponding eigen vectors and loading. Each sample was assigned a ‘Score’ which is the computed value of the factor at that site and is referred as ‘Factor score’. The samples were graded with respect to the concerned factor.

Soil erosion related properties

The soil properties of Moridhal watershed pertaining to the estimation of soil loss and erodibility indices are presented in Table 1 through statistical parameters viz. minimum, maximum, mean, median, standard deviation, skewness and kurtosis. Various box plots as shown in Fig. 4 display the dataset based on a five-number summary of minimum, maximum, median, first and third quartiles of the physiographic units. The textural properties of soils are described by parameters like total sand, very fine sand, silt and clay content. The total sand and very fine sand content of the studied soils ranged from 5.1 to 86.0 per cent and 2.2-39.2, respectively. The silt content of the soils of watershed varied from 8.0 to 61.0 per cent, while the clay content was in the range of 6.0-46.8 per cent. Among the different physiographic units, the alluvial plain soils showed the highest value of total sand (Mean 48.2 per cent) and very fine sand (Mean 19.6 per cent) content in the upper piedmont plain. The values of total sand were found to be lowest (Mean 41.5 per cent) in the flood plain, while the very fine sand was lowest (Mean 14.8 per cent) in the lower piedmont plain. However, the value of coefficient of variation with respect to total sand (47.14 per cent) and very fine sand (37.68 per cent) were highest in the flood plain. The lowest values of coefficient of variation of total sand (27.42 per cent) and very fine sand (20.32 per cent) were observed in upper piedmont plain and alluvial plain soils, respectively. The highest value of silt content (Mean 31.6 per cent) was found in flood plain and lowest in upper piedmont plain (Mean 27.9 per cent).

The clay content was found to be highest (Mean 27.0 percent) in flood plain and the lowest value of clay content (Mean 21.3 per cent) was found in the alluvial plain. The coefficient of variation for silt (46.10 per cent) and clay (46.96 per cent) content was highest in the lower piedmont plain. The increasing trend of finer materials from upper piedmont plain towards flood plain could be attributed to the fact that these materials were washed down from the upper piedmont plain and, thereby, got deposited in the flood plain. The textural properties of studied soils varied from loamy sand to clay (Table 1). The coarser soil texture ranging from loamy sand to loam was observed in the upper piedmont plain and lower piedmont plain soils. The finer soil texture ranging from clay to clay loam was found in the alluvial plain and flood plain soils. Deka et al., (2009) also reported finer texture in lower elevation areas and coarser texture in higher elevation areas in the soils of north bank plains of Assam. The structural properties of studied soils varied from subangular blocky to massive structure (Table 1). The flood plain and alluvial plain soils were mostly massive in structure which might be due to occasional water stagnation that inhibits aggregation. Being high in silt and clay content, the soils of the upper piedmont plain and lower piedmont plain had mostly single grained structure which may be ascribed to high coarse sand content in these soils. The soils of alluvial plain were dominated by subangular blocky structure. This is in accordance with the findings of Deka et al,. (2017) who also reported similar structures in the rice-growing alluvial soils of the Brahmaputra valley of Assam.

The organic matter content of studied soils varied from medium to high (5.50 to 29.60 g kg^-1) (Table 1) with a standard deviation of 4.96. Among the physiographic units, the highest (Mean 19.10 g kg^-1) and lowest (Mean 13.08 g kg^-1) amount of organic matter was found in flood plain and upper piedmont plain, respectively. The organic matter status of the soils in the flood plain was higher because of intensive cropping associated with the application of organic manures by the farmers during the cultivation of different crops in their field as well as transportation and leaching of bases and clay particles from upper elevation to lower elevation. Debnath et al., (2009) also reported a higher amount of organic matter in rice-growing soils of the Terai zone of West Bengal where the farmers usually apply more organic manure.

The hydraulic conductivity of the studied soils varied from 0.11 to 6.54 hr ^-1 (Table 1). Among the physiographic units, the highest value of hydraulic conductivity was found in the lower piedmont plain (Mean 2.09 cm hr^-1 1), which might be due to a higher amount of coarser material in the area. A significant negative correlation of hydraulic conductivity with clay content (r = -0.742**), mean weight diameter (r = -0.486**) might be due to dispersion and migration of finer particles into conducting pore. Dutta and Barkakoty, (1996) also obtained similar relationships in some soils of Assam.

The water holding capacity of the studied soils varied from 19.88 to 63.12 per cent (Table 1). The highest value of water holding capacity (Mean 40.40 per cent) was found in the flood plain. The upper piedmont plain soils exhibited lowest value of water holding capacity (Mean 29.14 per cent) which might be due to higher sand content. There was an increasing trend of water holding capacity from upper piedmont plain towards flood plain due to the dominance of finer material that provided sufficient capillary pores and maximum surface area to hold water. A significant positive correlation of water holding capacity was observed with silt (r = 0.441**), clay content (r = 0.491**), porosity (r = 0.368**) and organic matter (r = 0.381**). Similar relationships were also reported by Deka et al., (2017) in the soils of the North Bank Plains of Assam, India.

The available water capacity of studied soils varied from 1.0 to 22.5 per cent (Table 1). Among the physiographic units, the highest value of available water capacity (Mean 9.8 per cent) was found in the flood plain which might be due to the dominance of finer material. In contrary, the lowest value of available water capacity (Mean 8.2 per cent) was found in the alluvial plain. The upper piedmont plain soils exhibited highest value of the coefficient of variation (60.62 per cent). A significant negative correlation of available water capacity was observed with total sand content (r = -0.707**), while a positive significant correlation of available water capacity was observed with clay content (r = 0.677**). Borgohain et al., (2021) also reported a similar relationship of pabho watershed.

Estimation of soil loss factors: For estimating soil loss of the Moridhal watershed, the factors associated with the USLE model were first computed. The details about the estimation of the factors are described below:

Rainfall erosivity factor (R): Rainfall data for the period from 1989-2019 was obtained from the Regional Meteorological Centre, North East India. The annual precipitation, maximum 24 hrs precipitation of two year and one hour maximum precipitation interval of two year in the studied area were found to be 306.41 cm, 10.40 cm, and 2.58 cm, respectively. The rainfall erosivity factor (R) for the entire watershed area was determined to be 922.67 mm (Table 2) and it was found to be a driving factor of soil erosion processes in the study area.

Soil erodibility factor (K): The range and mean K value for different physiographic units are given in (Table 2). The soil erodibility factor (K) of the studied soils varied from 0.01 to 0.12. The K value was found to be highest (Mean 0.081 t ha MJ^-1 mm^-1) in upper piedmont plain. The lowest K value (Mean 0.067 t ha MJ^-1 mm^-1) was found in Alluvial plain. This might be due to coarser nature of piedmont plain soils which contributed higher permeability of water and, thereby, lesser runoff. Similar findings were also reported by Deka et al., (2011).

Topography factor (LS): The LS factor of the physiographic units (Table 2) varied based on the degree and length of slope. The lowest LS factor values of 0.58 was observed in the very gently to gently sloping soils of the flood plain areas. In contrary, the high sloppy lands of upper piedmont plain which was at higher elevation showed the highest LS value (3.34).

Cover and management factor (C): The C factor helps in determining the relative effectiveness of soil and crop management systems in reducing or preventing soil loss. The crop cover and management factor (C) values used for different types of land units were: paddy cultivation: 0.29 (for lower relief) and 0.40 (for higher relief), tea cultivation (0.01), waste land (0.50) and fallow land (0.69). These values were determined using information on land use/land cover and following the tables given by Obi Reddy et al. (2016). The C factor indicated the effect of conservation plans on soil loss and the possible distribution of soil loss potential with subsequent adoption of crop rotations or other management practices (Van der Knijff et al., 2000).

Conservation practice factor (P): The P factor reflects the effects of practices that would reduce the quantity and rate of runoff water and, thereby, could reduce the intensity of erosion. The P values were chosen based on field survey information. It was observed that the studied area had no any soil or water conservation measures, and hence the P values were assigned based on land utilization types. Under agricultural land the single crop land had poor bunding as compared to double crop and, hence, the ‘P’ value of 0.54 was assigned to single crop land and 0.30 to double crop land. For tea plantation ‘P’ value was assigned to be 0.70 as growing of plantation crop is a type of conservation practice. Fallow land and wasteland were assigned ‘P’ value as 1.00. Amongst the different physiographic units lower piedmont plain, alluvial plain and flood plain showed the highest P value (Mean 0.54), while the upper piedmont plain exhibited the lowest ‘P’ values (Mean 0.48).

Estimation of soil loss: After estimating the different USLE factors (R, K, LS, C, and P), the total soil loss (A) was computed by multiplying all the factors. The soil loss in the studied area varied from 0.87 to 67.77 t ha^-1 yr^-1with a mean value of 16.19 t ha^-1 yr^-1 (Table 2). Based on estimated soil loss values, the watershed area was grouped into different erosion classes (Potdar et al., 2003) viz., very slight (<5 t ha^-1 yr^-1), slight (5-10 t ha^-1 yr^-1), moderate (10-15 t ha^-1 yr^-1), moderately severe (15-20 t ha^-1 yr^-1), severe (20-40t ha^-1 yr^-1) and very severe (40-80 t ha^-1 yr^-1). Among the physiographic units, soil loss in the upper piedmont plain area was moderately severe to very severe with a value varying from 19.9-67.8 t ha^-1 yr^-1 (Mean 46.78 t ha^-1 yr^-1). The amount of soil eroded in lower piedmont plain areas varied from 2.7-52.5 t ha^-1 yr^-1 (Mean 23.06 t ha^-1 yr^-1) which is very slight to very severe erosion. Very slight to moderately severe soil erosion (Mean 9.03 t ha^-1 yr^-1) was found in the alluvial plain areas. Due to gentle slope flood plain areas exhibited very slight to moderate erosion and the soil loss varied from 0.87-12.79 t ha^-1 yr^-1 (Mean 6.14 t ha^-1 yr^-1). The soil loss map (Fig. 5) revealed that nearly 8.35 per cent (2,565 ha) area of the watershed is under very slight erosion hazard having soil loss less than 5 t ha^-1 yr^-1. About 30.15 per cent (9,265 ha) area of the watershed is under slight erosion 5-10 t ha^-1 yr^-1. Moderate erosion (15-20 t ha^-1 yr^-1) and moderately severe erosion (20-40 t ha^-1 yr^-1) covers 21.73 per cent (6,678 ha) and 7.54 per cent (2,318 ha) areas, respectively. Nearly 6.32 per cent (1,943 ha) area of the watershed suffers from very high erosion having soil loss between 40-80 t ha^-1 yr^-1. The studied soils exhibited a decreasing trend of soil loss from upper piedmont plain towards flood plain which might be due to washing away of finer material from upslope towards plain. A strong positive correlation of soil loss with relief as evident from the trend analysis curve (Fig. 6) also corroborates the findings. The positive correlation of soil loss with relief indicated that the increase in degree and length of slope resulted more soil loss (Deka et al. 2011).

Erodibility indices: Soil erodibility depends primarily on the physical characteristics of the soils viz., soil aggregates, nature and amount of organic matter content, and particle size distribution. These physical characteristics of soils are much affected by the land use system. To determine the erosional behavior of soils under different land use systems, various erodibility indices like water-stable aggregates, dispersion ratio, erosion ratio and erosion index were estimated by incorporating the data on basic soil properties in different empirical formulas. The results are presented in Table 3.

Aggregate status

Macroaggregate: The water-stable aggregates (>0.25 mm in diameter) of the studied soils varied from 10.3 to 82.3 per cent (Table 3). There was an increasing trend of macroaggregate from upper piedmont plain (Mean 24.9 per cent) to flood plain (Mean 47.4 per cent). The higher macroaggregate was found in the alluvial plain and flood plain soils which might be due to a higher amount of organic matter and clay content, which acted as binding agents for the formation of macroaggregate. Water-stable macroaggregation (which prevents crusting) and carbon content are important determining factors of soil erodibility (Barthes, 1999). This is supported by the existence of a significant positive correlation of macroaggregate with organic matter (r = 0.234**) and clay (r = 0.588**). The decreasing trend of macroaggregate towards upslope might be due to increase in total sand content as evident from the significant negative correlation of macroaggregate with total sand (r = -0.595**). Similar results were reported by Deka and Dutta, (2016) while assessing erodibility indices in the Northern Brahmaputra plains Assam, India.

Microaggregate: The microaggregate in the studied soils varied from 17.8 to 89.8 per cent (Table 3). Among the physiographic units, the highest value of microaggregate was recorded under upper piedmont plain (Mean 75.1 per cent) and lowest in flood plain (Mean 52.6 per cent) soils, while the highest value of coefficient of variation was observed in the flood plain (45.32 per cent). There was a decreasing trend of microaggregate from upper piedmont plain to flood plain. It might be due to a decrease in clay and organic matter content towards down slope. This is substantiated by the existence of a significant negative correlation with clay content (r = -0.588**) and organic matter (r = -0.234**). Borgohain et al., (2021) also reported similar results while assessing the soils of the Pabho Watershed of Lakhimpur district of Assam, India.

Mean weight diameter: The mean weight diameter in the studied soils varied from 1.00 to 2.74 mm (Table 3). Among the physiographic units, the highest value of mean weight diameter was recorded under flood plain (Mean 1.86 mm) and the lowest in upper piedmont plain (Mean 1.41 mm). The increasing trend of mean weight diameter from upper piedmont plain to flood plain indicated better aggregation which might be brought about by clay and organic matter (Deka et al., 2016). This is supported by the existence of a significant positive correlation of mean weight diameter with clay (r = 0.534**) and organic matter (r = 0.308**). Borgohain et al., (2021) also reported similar results while assessing the soils of the Pabho Watershed of Lakhimpur district of Assam.

Dispersion ratio (DR): As per the data presented in Table 3, the dispersion ratio of the studied soils varied from 0.06 to 0.57. According to the criterion of Middleton, (1930) soils having dispersion ratio >0.15 are erodible in nature. Among the physiographic units, the highest value of dispersion ratio was recorded in upper piedmont plain (Mean 0.31) and the lowest value of dispersion ratio was recorded in alluvial plain (Mean 0.13) soils. Higher dispersion ratio value indicates more proneness to soil erosion. A significant negative correlation of dispersion ratio with total sand (r = -0.331**), organic matter (r = -0.337**), hydraulic conductivity (r = -0.383**) and a positive significant correlation with silt (r = 0.288**) and clay (r = 0.280**) content indicated that soils which were dominated by the finer particles exhibited more dispersion ratio (Deka et al., 2017). The organic matter content was found to be correlated significantly and negatively (r = -0.337**) with the dispersion ratio which indicated that organic matter helped in binding of soil particles and, thereby, prevent dispersion (Singh and Prakash, 1985).

Erosion ratio (ER): The erosion ratio of the studied soils varied from 0.01 to 0.55 (Table 3). All the soils under the present study which had values of erosion ratio of above 0.10 could be considered as erodible as per the criteria of Middleton, (1930). Among the physiographic units, the highest value of erosion ratio was recorded under upper piedmont plain (Mean 0.13). In contrary, the lowest value of erosion ratio (Mean 0.05) was recorded under alluvial plain. The erosion ratio exhibited a significant negative correlation with sand (r = -0.186^*), organic matter (r = -0.234^**) and hydraulic conductivity (r = -0.295**). Higher the hydraulic conductivity more will be the infiltration of water and, thereby, the erosion through runoff is reduced (Singh and Prakash, 1985).

Erosion index (EI): The erosion index of studied soils varied from 0.03 to 0.71 (Table 3). Among the physiographic units, the highest value of erosion index (Mean 0.21) was found in the soils of upper piedmont plain. The lowest value of erosion index (Mean 0.11) was found in the soils of alluvial plain. The erosion index showed significant positive correlation with total sand (r = 0.183^*), fine sand (r = 0.249^**), and relief (r = 0.426^*). However, it exhibited significant negative correlation with clay (r = -0.288^**) and organic matter (r = -0.168^*). Similar results were also reported by Bhatia and Sharma, (1976) for some soils of Assam, India. It was observed that almost all the studied physico-chemical properties influenced the erodibility indices to a great extent. A strong positive correlation of the erodibility indices with relief supported the effect of elevation on soil properties as well as soil erosion. This is similar to the findings of Deka & Dutta, (2016) those who also obtained similar results in the soils of the north bank plain of Assam, India. Based on the calculated values of different erodibility indices the upper piedmont plain soils could be regarded as the most erodible soils of the studied watershed area.

Principal Component Analysis of Soil Parameters: The variability of 11 soil hydro physical parameters along with relief, erodibility indices and soil loss was explained by 2 principal components with eigen values of more than 1.5 explaining 64 % variability in the soil parameters (Table 4). The screen plot (Fig. 7) demonstrated the relation between the eigen values and the principal components numbers. The factor loading matrix of the soil parameters (Table 1) revealed that silt, clay, water holding capacity available water content, macroaggregate and mean weight diameter had a positive loading in principal components 1 (P1), whereas sand, very fine sand, hydraulic conductivity and microaggregate had a negative loading on P1. This P1 component, which was strongly influenced by the inherited mechanical components, was termed as 'Inherent Potentiality Factor'. Five parameters including relief, dispersion ratio, erosion ratio, erosion index and soil loss exhibited positive loading in principal component 2 (P2). In contrary organic matter showed a negative loading in this component. All these erodiblity parameters were significantly correlated with each other and represented a set of related parameters with respect to soil erosion. Thus this second factor was termed as ‘Erosion Factor'. The factor loadings of soil parameters in P1 and P2 are depicted in Fig. 8 which revealed that the studied parameters were distributed in all the 4 quadrants. The first quadrant had parameters with positive loadings of both PCs; while the second quadrant had negative loadings on P1 and positive loadings on P2. The third quadrant had negative loadings of both PCs; and the fourth quadrant had positive loadings on P1 and negative loadings on P2.

Relationship between soil loss and erodibility indices: A significant positive correlation of soil loss with dispersion ratio (r = 0.825**), erosion ratio (r = 0.492**) and erosion index (r = 0.292**) was noticed for the studied soils. It indicated that dispersion ratio (DR), erosion ratio (ER) and erosion index (EI) are reliable measures in acquiring information on the existing or potential erosion prone areas. Among the erodibility indices, dispersion ratio was found to have more precision in predicting soil loss value. The Component Plot generated through PCA analysis also revealed that the erodiblity indices and the soil loss, which occurred closely in the 1^st quadrant, are a set of related parameters. This indicated that the erodibility indices are reliable measures in acquiring information on the existing or potential erosion prone areas. When an inventory of quantitative erosional soil loss is not possible then, the erodibility indices could be used to predict the erosion status of soils (Borgohain et al., 2021). For predicting soil loss from erodibility indices three different regression equations were developed by relating soil loss (SL) with dispersion ratio (SL= 2.86 + 77.00 DR), erosion ratio (SL= 8.13 + 121.86 ER) and erosion index (SL= 7.26 + 69.68 EI) which are represented graphically in Fig. 9, Fig 10 and Fig. 11 respectively.

The study demonstrated the usefulness of USLE model and GIS in assessing soil erosion and identifying the erosion prone areas. Spatial information on physiography, slope, land use/land cover and soils derived through satellite remote sensing and supportive data was used for estimation of various factors contributing to soil loss by water erosion. Based on the study it was found that the soils in different physiographic units of Moridhal watershed of Assam, India are suffering from varying degree of soil erosion. The soil loss in the studied area varied from 0.87 to 67.77 t ha^-1 yr^-1. The study also revealed high degree positive correlation of soil loss with erodibility indices. The linear regression equations relating to the soil loss with dispersion ratio, erosion ratio and erosion index could be used to estimate the soil loss in similar type of soils. The results of soil loss obtained from the study could be effectively used by the planners, watershed managers and policy makers to formulate site specific conservation plans in order to minimize soil loss for sustaining productivity. Based on the study it can be put forwarded that the upper piedmont plain areas of the watershed require management practices for reducing the erosion status and increasing soil productivity. However, flood plain areas are less erodible and, thereby, require no special treatment for soil conservation.

Acknowledgements

This study was supported by the Indian Space Research Organization, Department of Space, Government of India under the RESPOND programme.

Author contribution

Soil sample collection and soil analysis in laboratory performed by Prem Kumar Bharteey and Rituparna Saikia. Text preparation, remote sensing and GIS based mapping, statistical analysis; model validation mapping was done by Bipul Deka, Marami Dutta, D.K. Patgiri and Jonali Goswami

Funding

Not Applicable

Availability of Data and Materials

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Conflict of Interest

The authors declare no competing interests.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

AbdelRahman MAE, Zakarya YM, Metwaly MM, Koubouris G. 2020. Deciphering soil spatial variability through geostatistics and interpolation techniques. Sustainability 13:194 https://doi.org/10.3390/su13010194.

Alitane A, Essahlaoui AEI, Hafyani MEI, Hmaidi AEl, Ouali A, Kassou A, Van Rompaey A. 2022. Water erosion monitoring and prediction in response to the effects of climate change using RUSLE and SWAT equations: Case of R’Dom watershed in Morocco. Land 11: 93 https://doi.org/10.3390/land11010093.

Baigorria GA, Romero CC. 2007. Assessment of erosion hotspots in a watershed: Integrating the WEPP model and GIS in a case study in the Peruvian Andes. Environment Modeling Software 22: 1175-183 https://doi.org/10.1016/j.envsoft.2006.06.012

Balakrishna K, Balakrishna BHD. 2019. Assessment of soil erosion by USLE model using remote sensing and GIS of Hemavathy basin. International Journal of Scientific Engineering and Research 10:106-116

Bergsma E. 1980. Provisional rain-erosivity map of Netherlands: In books on Assessment of Erosion (Eds M Deboodt and D Gabriels). John Willey Sons pp 121-126

Barthes B, Albrecht A, Asseline J, Noni GD., Roose E. 1999. Relationship between soil erodibility and topsoil aggregate stability or carbon content in a cultivated Mediterranean highland (Aveyron, France). Comunication of Soil Science Plant Analysis 30:1929-1938 https://doi.org/10.1080/00103629909370343.

Borgohain S, Deka B, Dutta M, Thakuria RK, Patgiri DK. 2021. Geospatial assessment of water induced soil and nutrient erosion in Pabho watershed of Assam, India using USLE model. Journal of Environment Biology 42:406-413. http://doi.org/10.22438/jeb/42/2(SI)/SI-244.

Bouyoucos GJ. 1935. The clay ratio as criterion of susceptibility of soil to erosion. Journal of American Society Agronomy 27:738-741.

Bhatia KS, Sarmah N. 1976. Studies on the physical properties of some Assam soils in relation to their erodibility. Journal of the Indian Society of Soil Science 24: 369-373.

Cetin M, Pekkan OI, Ozturk GB, Kurkcuoglu MAS, Kucukpehlivan T, Cabuk A, 2022. Examination of the change in the vegetation around the Kirka Boron mine site by using remote sensing techniques. Water, Air Soil Pollution 233: 1-15 http://dx.doi.org/10.1007/s11270-022-05738-y.

Chakrabortty R, Pal SC, Arabameri A. 2022. Water-induced erosion potentiality and vulnerability assessment in Kangsabati river basin, eastern India. Environment Development Sustainability 24: 3518-3557 https://doi.org/10.1007/s10668-021-01623-6

Das B, Bordoloi R, Thungon LT, Paul A, Pandey PK, Mishra M, Tripathi OP. 2020. An integrated approach of GIS, RUSLE and AHP to model soil erosion in West Kameng watershed, Arunachal Pradesh. Journal Earth System Science 129: 1-18 https://doi.org/10.1007/s12040-020-1356-6.

Debnath P, Mahanta M, Ghosh SK. 2009. Distribution of available boron in the selected surface and subsurfacesoils of Terai zone of West Bengal in relation to physico-chemical properties. Journal of Maharashtra agricultural universities 3:357-358. https://doi.org/10.20546/ijcmas.2017.612.474.

Das S., Deb P., Bora P.-K., and Katre P., 2021. Comparison of RUSLE and MMF Soil Loss Models and Evaluation of Catchment Scale Best Management Practices for a Mountainous Watershed in India. Sustain., 13, 232. https://doi.org/10.3390/su13010232

Deka B, Dutta M. 2016. Principal component analysis of soil properties in assessing erodibility indices in the Northern Brahmputra plains of Assam. Journal of Soil and Water Conservation 15: 277-283 10.5958/2455-7145.2016.00016.3.

Deka B, Maruthi Sankar GR, Dutta M, Baruah TC, Pushpanjali VVK, Mishra PK. 2017. Assessment of fertility potential for the soils of Northern Brahmaputra valley zone using Principal Component Analysis. Indian Journal of Soil Conservation 45: 1-11.

Deka B, Baruah TC, Dutta M, Neog P. 2011. Soil loss estimation on RS data and GIS techniques in Ghiladhari Watershed of Northern Brahmaputra Valley in Assam. Journal of Soil and Water Conservation 10: 123-128.

Deka B, Barua TC, Dutta M, Patgiri DK. 2009. Landscape soil relationship and pedogenic evaluation of soils in Ghiladhari watershed of the Brahmaputra Valley of Assam. Journal of the Indian Society of Soil Science 60: 92-100.

Dekaraja D, Mahanta R. 2022. Exploring the socio-economic impact of riverbank erosion in the Brahmaputra Valley of Assam, India: A Case Study of Two Districts. Landscape Online 97: 1-13 https://doi.org/10.3097/LO.2022.1105.

Dutta M, Barkakoty PK. 1996. Evaluation of some physical characteristics of Charaipani irrigation project command area. Journal Agriculture Science Society North East 9: 135-140.

Jemai S, Kallel A, Agoubi B, Abida H. 2021. Soil erosion estimation in arid area by USLE model applying GIS and RS: Case of Oued El Hamma Catchment, South-Eastern Tunisia. Journal of Indian Society of Remote Sensing 49: 1293-1305. http://dx.doi.org/10.1007/s12524-021-01320-x.

Jensen JR. 1986. In Introductory Digital Image Processing-A Remote Sensing Perspective. Publ. by Prentice Hall, Engelewood Cliff, New Jersey, USA.

Klute A. 1965. Laboratory measurement of hydraulic conductivity of saturated soil. Int. Methods of Soil Analysis Part-I. A Soc Agron Inc.

Kumar A, Singh S, Pramanik M, Chaudhary S, Maurya AK, Kumar M. 2022. Watershed prioritization for soil erosion mapping in the Lesser Himalayan Indian basin using PCA and WSA methods in conjunction with morphometric parameters and GIS-based approach. Envronment Development Sustainability 24: 3723-3761 http://dx.doi.org/10.21203/rs.3.rs-178139/v1.

Kumar K, Tripathi SK, Bhatia KS. 1995. Erodibility characteristics of Rendhar watershed soils of Bundelkhand. Indian Journal of Soil Conservation 23: 200-204.

Middleton HE. 1930. Properties of soils which influence erosion. US Department of Agriculture technical bulletin 178:1-16.

Mohanty K, Filipovski G. 1982. Fertility potential classification of ricegrowing soils using principal component analysis. Journal of Indian Society of Soil Science 30: 312-319.

Obi Reddy GP, Kurothe RS, Sena DR, Harindranath CS, Niranjana KV, Naidu LGK, Singh SK, Sarkar D, Mishra PK, Sharda VN. 2016. Assessment of soil erosion in tropical ecosystem of Goa, India using Universal Soil Loss Equation, geostatistics and GIS. Indian Journal of Soil Conservation 44: 1-7.

Parveen R, Kumar U. 2012. Integrated approach of Universal Soil Loss Equation (USLE) and Geographical Information System (GIS) for soil loss risk assessment in Upper South Koel Basin, Jharkhand. Journal of Geographical Systems 4: 588-596. http://www.scirp.org/journal/PaperInformation.aspx?PaperID=26324

Piper CS. 1966. Soil and Plant Analysis. University of Adelaide, Australia.

Potdar SS, Srivastava R, Nagaraju, MSS, Prasad J, Saxena RK. 2003. Mapping of erosional soil loss in Nanda-Khairi watershed of Nagpur district of Maharasthra using remotely sensed data and GIS techniques. Agropedology 13: 10-18.

Rao SK, Yarrakula K, Mani A, Jayaprakash R. 2018. Soil erosion estimation around Godavari basin in Andhra Pradesh using remote sensing and GIS. Bulletin of Environment, Pharmacology and Life Sciences., 8,119-129.

Richards LA. 1948. Pressure membrane apparatus, construction and use. Agricultural Engineering 28: 451-454.

Renard K, Foster G, Weesies G, McCool D, Yoder D. 1997. Predicting soil erosion by water: A guide to conservation planning with the Revised Universal Soil Erosion Equation (RUSLE), Agriculture Handbook no. 703, Sci. and Educ. Admin., U.S. Dept. Agriculture, Washington, D.C.

Sahi BP, Singh SN, Sinha AC, Acharya B, 1977. Erosion Index – A new index of soil erodibility. Journal of Indian Society of Soil Science 25: 7-10.

Saroha J. 2017. Soil Erosion: Causes, extent and management in India. International Journal Of Creative Research Thoughts 5: 1321-1330.

Singh G R, Prakash O. 1985. Characteristics and erodibility of some hill soils in Uttar Pradesh under varying land use, slope and terracing conditions. Journal of Indian Society of Soil Science 33:858-864.

Singh S, Babu R, Chandra S. 1981. Soil loss prediction research in India. Central Soil and Water Conservation Research and Training Institude, Dehradun, Bull No. T-12, 9-10.

Srinivasan R, Singh SK, Nayak DC, Hegde R, Ramesh M. 2019. Estimation of soil loss by USLE Model using Remote Sensing and GIS Techniques - A Case study of Coastal Odisha, India. Eurasian Journal of Soil Science 8: 321-328.

Tsegaye K, Addis HK, Hassen EE. 2020. Soil erosion impact assessment using USLE/GIS approaches to identify high erosion risk areas in the lowland agricultural watershed of Blue Nile Basin, Ethiopia. International Annals Science 8: 120-129. https://doi.org/10.21467/ias.8.1.120-129.

Van der Knijff J M, Jones RJA, Montanarella L. 2000. Soil Erosion risk Assessment in Europe, European Comm. European Soil Bureau, EUR 19044 EN, p. 34.

Walkey A, Black IA. 1934. Determination of organic matter in soil. Soil Science 37: 549-556.

Wischmeier WH, Smith DD. 1978. Predicting rainfall erosion losses – A guide for Conservation Planning. Agricultural. Handbook No. 537. USDA Washington D. C.

Wischmeier WH, Johnson CG, Cross BV. 1971. Soil Erodibility monograph for farm land conservation sites. Journal of Soil Water Conservation 26: 189-193.

Yoder RE. 1936. A direct method of aggregate analysis and a study of the physical nature of ersosion losses. Journal American Society of Agronomy 28: 337-351.

Table 1 Physiographic distribution of soil erosional properties in Moridhal watershed

		Minimum		Maximum		Mean		Median		Std. Deviation		Skewness		Kurtosis
Physiographic unit: Upper piedmont plain
Sand (%)		24.20		82.20		47.37		47.15		12.99		0.86		1.84
Very fine sand (%)		12.31		39.21		25.21		23.61		7.23		0.48		-0.17
Silt (%)		9.60		42.65		27.87		26.45		8.55		-0.38		-0.07
Clay (%)		8.20		36.45		24.77		27.24		9.05		-0.42		-0.95
Organic matter (%)		0.55		2.10		1.31		1.27		0.41		0.05		-0.52
Hydraulic conductivity (cm hr^-1)		0.34		5.66		1.79		1.36		1.48		1.73		2.72
Water holding capacity (%)		21.04		47.00		29.14		26.58		6.91		1.03		0.75
Available water content (%)		4.00		22.45		8.94		7.40		5.42		1.18		0.59
Macroaggregate (%)		10.25		40.10		24.86		25.91		7.76		-0.24		-0.28
Microaggregate (%)		59.90		89.75		75.14		74.09		7.76		0.24		-0.28
Mean weight diameter (mm)		1.01		2.40		1.41		1.29		0.37		1.34		1.78
Physiographic unit: Lower piedmont plain
Sand (%)		5.20		82.11		47.25		51.13		21.60		-0.03		-0.93
Very fine sand (%)		2.24		26.60		14.76		15.36		5.02		-0.07		0.17
Silt (%)		9.50		60.00		29.21		28.07		13.47		0.31		-0.54
Clay (%)		7.70		46.80		23.53		20.17		11.05		0.33		-1.24
Organic matter (%)		0.56		2.82		1.72		1.72		0.54		0.05		-0.60
Hydraulic conductivity (cm hr^-1)		0.11		6.32		2.09		2.12		1.75		0.84		-0.06
Water holding capacity (%)		19.88		57.91		32.95		31.56		8.79		0.83		0.41
Available water content (%)		3.43		20.71		9.13		5.96		5.42		0.76		-0.91
Macroaggregate (%)		15.50		82.25		31.63		25.03		16.44		1.59		1.68
Microaggregate (%)		17.75		84.50		68.37		74.97		16.44		-1.59		1.68
Mean weight diameter (mm)		1.00		2.45		1.66		1.67		0.34		-0.02		-0.54
Physiographic unit: Alluvial plain
Sand (%)		27.40		76.30		48.16		50.11		13.29		0.00		-0.97
Very fine sand (%)		12.31		30.12		19.64		19.68		3.99		0.48		0.43
Silt (%)		14.30		43.45		30.52		30.20		7.04		-0.12		-0.26
Clay (%)		9.40		38.40		21.32		18.02		7.80		0.63		-0.76
Organic matter (%)		1.10		2.48		1.54		1.51		0.37		0.86		0.22
Hydraulic conductivity (cm hr^-1)		0.14		4.65		1.87		2.13		1.17		0.00		-0.17
Water holding capacity (%)		24.12		48.15		34.69		32.18		6.87		0.52		-0.72
Available water content (%)		1.01		17.55		8.17		7.13		3.91		0.57		-0.04
Macroaggregate (%)		21.50		73.28		46.41		35.99		18.55		0.17		-1.80
Microaggregate (%)		26.72		78.50		53.59		64.02		18.55		-0.17		-1.80
Mean weight diameter (mm)		1.10		2.68		1.81		1.73		0.37		0.46		0.16
Physiographic unit: Flood plain
Sand (%)	5.10		86.01		41.46		39.85		19.55		0.26		-0.56
Very fine sand (%)	2.33		35.22		14.93		14.54		5.62		0.62		1.09
Silt (%)	8.01		61.00		31.58		30.00		11.94		0.27		0.00
Clay (%)	5.98		44.20		26.96		27.00		10.90		-0.12		-1.18
Organic matter (%)	0.89		2.96		1.91		1.93		0.44		0.01		0.05
Hydraulic conductivity (cm hr^-1)	0.12		6.54		1.80		1.76		1.69		0.93		0.14
Water holding capacity (%)	23.12		63.12		40.40		41.12		9.92		0.28		-0.63
Available water content (%)	2.61		18.70		9.77		9.10		4.75		0.36		-1.19
Macroaggregate (%)	14.50		81.85		47.39		40.12		23.84		0.05		-1.68
Microaggregate (%)	18.15		85.50		52.61		59.88		23.84		-0.05		-1.68
Mean weight diameter (mm)	1.02		2.74		1.86		1.82		0.40		0.00		-0.41
Whole watershed
Sand (%)	5.10		86.01		44.85		46.73		18.86		0.09		-0.55
Very fine sand (%)	2.24		39.21		16.75		16.29		6.41		0.68		1.22
Silt (%)	8.01		61.00		30.31		29.26		11.49		0.26		0.04
Clay (%)	5.98		46.80		24.84		24.90		10.47		0.14		-1.18
Organic matter (%)	0.55		2.96		1.73		1.79		0.50		0.08		-0.39
Hydraulic conductivity (cm hr^-1)	0.11		6.54		1.89		2.01		1.61		0.92		0.29
Water holding capacity (%)	19.88		63.12		36.08		33.81		9.72		0.59		-0.28
Available water content (%)	1.01		22.45		9.25		7.18		4.91		0.64		-0.78
Macroaggregate (%)	10.25		82.25		40.09		32.00		21.49		0.63		-1.15
Microaggregate (%)	17.75		89.75		59.91		68.00		21.49		-0.63		-1.15
Mean weight diameter (mm)	1.00		2.74		1.74		1.71		0.40		0.18		-0.47

Table 2 Range and mean of USLE factors and soil loss in different physiographic units of Moridhal watershed

Parameters		R		K		LS		C		P		Soil loss (t ha^-1 yr^-1)
Physiographic unit: Upper piedmont plain
Range		-		0.03-0.12		2.83-4.06		0.38-0.50		0.30-0.54		19.94-67.77
Mean		922.67		0.081		3.34		0.40		0.48		46.78
SD		0.00		0.02		0.47		0.04		0.11		14.64
SE		0.00		0.01		0.11		0.01		0.02		3.36
CV (%)		0.00		28.19		14.03		8.66		22.77		31.29
Physiographic unit: Lower piedmont plain
Range	-		0.01-0.12		1.49-2.37		0.29-0.38		-		2.74-52.50
Mean	922.67		0.067		2.30		0.30		0.54		23.06
SD	-		0.04		0.23		0.03		0.00		13.06
SE	-		0.01		0.03		0.00		0.00		1.85
CV (%)	-		53.31		9.81		10.42		0.00		56.64
Physiographic unit: Alluvial plain
Range		-		0.01-0.12		0.75-1.10		0.29-0.38		-		0.78-18.60
Mean		922.67		0.064		0.97		0.29		0.54		9.01
SD		-		0.03		0.17		0.02		0.00		4.96
SE		-		0.01		0.03		0.00		0.00		0.97
CV (%)		-		53.17		17.99		6.01		0.00		55.09
Physiographic unit: Flood plain
Range		-		0.03-0.12		0.20-0.85		0.29-0.38		-		0.95-4.18
Mean		922.67		0.07		0.58		0.29		0.54		6.14
SD		-		0.04		0.26		0.01		0.00		4.34
SE		-		0.00		0.03		0.00		0.00		0.50
CV (%)		-		51.72		45.20		4.99		0.00		70.59
Whole watershed
Range		-		0.01-0.12		0.20-4.06		0.29-0.50		0.30-0.54		0.87-67.77
Mean		922.67		0.07		1.45		0.31		0.53		99.39
SD		-		0.04		1.03		0.04		0.04		16.00
SE		-		0.00		0.08		0.00		0.00		1.23
CV (%)		-		50.02		71.00		13.57		7.63		99.39

Table 3 Range and mean value of aggregate and erodibility indices in different physiographic units of Moridhal watershed

Parameters	MAG	MIG	MWD	CR	SCR	MCR	DR	ER	EI
Parameters	(%)	(%)	(mm)	CR	SCR	MCR	DR	ER	EI
Physiographic unit: Upper piedmont plain
Range	10.3-40.1	59.9-89.8	1.0-2.4	1.74 -11.20	0.60 - 3.39	1.67-9.52	0.13-0.57	0.02-0.34	0.07-0.57
Mean	24.9	75.1	1.41	3.88	1.32	3.61	0.31	0.13	0.21
SD	7.76	7.76	0.37	2.65	0.76	2.34	0.11	0.09	0.14
SE	1.78	1.78	0.08	0.61	0.18	0.54	0.03	0.02	0.03
CV (%)	31.22	10.33	26.04	68.31	57.98	64.67	35.79	73.43	64.04
Physiographic unit: Lower piedmont plain
Range	15.5-82.3	17.8-84.5	1.0-2.5	1.14-11.99	0.28-2.85	1.09-10.24	0.08-0.49	0.01-0.34	0.03-0.33
Mean	31.6	68.4	1.66	4.44	1.34	3.97	0.22	0.09	0.17
SD	16.44	16.44	0.34	2.85	0.52	2.41	0.10	0.07	0.07
SE	2.32	2.32	0.05	0.40	0.07	0.34	0.01	0.01	0.01
CV (%)	51.97	24.05	20.44	64.30	38.72	60.62	45.35	71.41	40.85
Physiographic unit: Alluvial plain
Range	21.4-73.3	26.7-78.5	1.1-2.7	1.26-15.76	0.27-2.76	1.19-12.44	0.06-0.26	0.01-0.14	0.05-0.27
Mean	46.4	53.6	1.81	3.98	1.53	3.58	0.15	0.06	0.12
SD	18.55	18.55	0.37	1.88	0.38	1.65	0.04	0.02	0.03
SE	3.64	3.64	0.07	0.37	0.08	0.32	0.01	0.00	0.01
CV (%)	39.97	34.61	20.56	43.64	25.07	41.60	32.94	45.91	25.52
Physiographic unit: Flood plain
Range	14.50-81.85	18.15-85.50	1.02-2.74	1.31-11.41	0.57-2.54	1.25-9.47	0.08-0.23	0.02-0.55	0.06-0.71
Mean	47.4	52.6	1.86	3.44	1.20	3.13	0.13	0.07	0.13
SD	23.84	23.84	0.40	2.79	0.53	2.27	0.05	0.06	0.08
SE	2.75	2.75	0.05	0.32	0.06	0.26	0.01	0.01	0.01
CV (%)	50.31	45.32	21.29	75.73	41.24	68.79	31.10	101.16	63.93
Whole watershed
Range	10.3-82.3	17.8-89.8	1.00-2.74	1.14-15.72	0.27-3.39	1.09-12.44	0.06-0.57	0.01-0.55	0.03-0.71
Mean	40.1	59.9	1.74	4.02	1.35	3.63	0.18	0.08	0.15
SD	21.49	21.49	0.40	2.67	0.54	2.24	0.09	0.07	0.09
SE	1.65	1.65	0.03	0.21	0.04	0.17	0.01	0.01	0.01
CV (%)	53.59	35.86	22.79	66.49	40.21	61.66	50.75	88.25	58.57

Table 4 Eigen values, variance (%), cumulative variance (%) and matrix factor loading of soil parameters

Rotated Component Matrix
	Component
	1	2
Eigen value	6.248	3.929
% of Variance	39.051	24. 556
Cumulative variance (%)	39.051	63.601
Sand	-0.917	-0.088
Very fine sand	-0.437	0.391
Silt	0.749	0.129
Clay	0.832	0.017
Hydraulic conductivity	-0.867	-0.196
Water holding capacity	0.688	-0.312
Available water content	0.796	0.281
Organic matter	0.250	-0.431
Macroaggregate	0.817	-0.277
Microaggregate	-0.817	0.277
Mean weight diameter	0.686	-0.267
Elevation	-0.234	0.704
Dispersion ratio	0.266	0.860
Erosion ratio	0.235	0.849
Erosion index	-0.097	0.704
Soil loss	0.028	0.849

No competing interests reported.

Download PDF

Version 2

posted

You are reading this older preprint version

Read the latest preprint version →

Soil erosion risk assessment in Moridhal watershed of Assam, India using Universal Soil Loss Equation and GIS

Status:

Version 2

Abstract

Figures

Introduction

Materials And Methods

Results And Discussion

Conclusions

Declarations

References

Tables

Table 1 Physiographic distribution of soil erosional properties in Moridhal watershed

Additional Declarations

Status:

Version 2