Erosion tolerance index under different land use units for sustainable resource conservation in a Himalayan watershed using remote sensing and geographic information system ( GIS )

Water erosion induced land degradation is a serious problem in the Himalayan hills of India. This paper presents an approach to analyze and assess the land use-wise soil erosion risks using Revised Universal Soil Loss Equation (RUSLE), fuzzy function, and remote sensing and GIS. Erosion tolerance index (ETI) was computed by integrating potential annual soil loss and soil loss tolerance limits for each landscape unit for the Bino watershed (296.17 km), Ramganga catchment, Uttarakhand. The thematic maps were generated using digital elevation model (DEM), remote sensing and geographic information system (GIS). Soil loss tolerance limits (T) by a set of fuzzy functions and soil characteristics, and potential soil loss rates (A) by RUSLE model. Potential annual soil loss varied from 0.49-3.64, 1.93-8.83, 4.07-55.49 Mg ha in the lower, middle and higher slopes, respectively. The calculated soil loss tolerance limit (SLTL) values in the study watershed ranged from 2.5-10 Mg ha yr. The SLTL values for lower, middle and high slopes varies from 3.1-10.0, 5.0-10.0 and 6.7-10.0 Mg ha, respectively. A total of 135 ETI values have been generated for different landscapes and slope classes. In higher slopes ETI values for open forest, dense forest, rainfed agricultural land, irrigated agricultural land and waste land varied from 0.170-0.249, 0.439-1.0, 0.117-0.130, 0.216-0.389 and 0.1210.137, respectively. The lowest ETI has been found in the Bino sub-watershed, whereas the highest for Syalde sub-watershed in all slopes, however the highest (0.38) weighted ETI value has been observed for Syalde sub-watershed and the lowest (0.212) for Gaya Gadhera sub-watershed. The results revealed the highest priority for Gaya Gadhera and least priority for Syalde sub-watershed. The approach for quantifying ETI and prioritization map developed in the study may serve as a guide for policy makers to decide effective watershed management plan.


INTRODUCTION
Land degradation like climate change is an anthropogenic induced process and poses biggest threat to the sustainable livelihood security of the farming communities across India.All these factors combined with increased rate of land degradation are contributing towards a decline in agricultural productivity leading to food insecurity.Again, per capita availability of inelastic land resource is rapidly declining in relation to the annual population growth of 1.4% in the country (ICAR, 2010).The average rate of soil loss in the steeply sloping Indian Himalayan region is more than 20 Mg ha -1 year -1 with a higher limit of 80 Mg ha -1 year -1 (Samra et al., 2002).The state of Uttarakhand has 92.57% area under hills of Himalayan ranges and 7.43% under valleys and plains.As a consequence of higher erosion in this region, top soil fertility and crop productivity are declining and rivers, canals and reservoirs are silting-up.
Therefore, maintaining and enhancing the productive potential of land resources is vital by progressive introduction of sustainable technologies on a watershed basis, and thereby resilience in crop production.The generation of soil erosion and sediment outflow are the combined effects soil system, rainfall conditions, topography, cropping system and conservation practices or management.However in the soil erosion process, erosion limits need to be defined in order to keep the onsite and off-site impacts at acceptable levels.Soil loss tolerance (T) is 'the maximum rate of annual soil erosion that will permit a level of crop productivity to be obtained economically and indefinitely' (McCormack et al., 1982).The 'T' value is also sometimes called 'permissible soil loss' due to the fact that the rates of soil erosion and soil formation are in equilibrium at this level and must be determined in a scientific manner (Li et al., 2009).However, before any meaningful soil conservation intervention methods could be implemented, there is a need to determine the erosion hazard associated with watersheds for prioritization.In India mostly prioritization of watersheds has been carried out based on the sediment yield index and geomorphologic characteristics using remote sensing and geographic information system (GIS) tools.Conceptually, sediment yield outflow from a watershed does not provide a true representative erosion status of different parts of the large watershed.The overall sediment outflow may be less due to the implementation of conservation measures even though in certain locations of the watershed there could be higher soil erosion (A).While considering watershed conservation work, it is not feasible to take the huge area at once as implementation of watershed management programs involves huge financial and human resources.Further, for the success and sustainability of land use, prioritization must be based on a proper evaluation of all the potential uses of each unit of a landscape (Huston, 2006) after gathering requisite information on physical and biological characteristics (viz.hydrology, soil, land use and topography) of a landscape.
Therefore, it seems logical to include both soil loss tolerance limit (T) and potential soil erosion in the prioritization process and it is better to prevent erosion in vulnerable areas than to rehabilitate already eroded areas and treating the land from the actual soil loss value to soil loss tolerance limit (SLTL) with a better watershed management plan.The increasing use and capability of geospatial tools integration with environmental parameters offer opportunities to utilize these tools for the effective management of information for planning purposes.Thus, the present study is an attempt aimed at comprehensive investigation of environmental aspects of the watershed management in real-time perspective using remote sensing and GIS techniques.
The main objectives of the study were to assess the potential soil loss, to determine the soil loss tolerance limit (SLTL) and to quantify erosion tolerance index from each land use and slope class for comparing SLTL with the expected soil loss and then prioritization of subwatersheds for further planning of soil and water conservation measures implementation in the Bino watershed of Uttarakhand.The working hypothesis for this study was that soil loss from the watershed would be higher than the soil loss tolerance T at each land use and slope class which was found a single value for the study watershed as per Mandal (1999) and thus erosion tolerance index may decrease with increase in slope gradient.

Study area and thematic maps
The study was conducted in the Bino watershed under river Ramganga, a major tributary of the river Ganga that originates in the outer Himalayas in Chamoli district of Uttarakhand and drains into river Bino.It is situated at 79°6′14.4″ to 79°17′16.8″E longitude and 29°47′6″ to 30°02′9.6″N latitude in Almora and Pauri Garhwal district of Uttarakhand.The climate of the watershed varies from Himalayan sub-tropical to sub-temperate with mean annual maximum and minimum air temperature of 30 and 18°C, respectively.The daily mean temperature remains higher during the months of May and June and minimum in December and January.The mean annual rainfall in the area is 931.3 mm.
Remote sensing and GIS were employed in collection, analysis and presentation of all related data for the study because of their synoptic, multi-spectral and multi-temporal nature for effective mapping, monitoring and understanding the geomorphology and landuse system.with UTM zone 44 N and then the area and perimeter of each subwatershed and watershed were calculated using 'calculate geometry' option in geographic information system (GIS).
The downloaded ASTER digital elevation model (DEM) of 30 × 30 m resolution was processed to get the extract DEM and then contour map at 10 m interval was prepared.Slope map of the watershed was created in GIS and then reclassified by dividing into 8 classes, that is, 0-1, 1-3, 3-5, 5-10, 10-15, 15-25, 25-35, 35-50 and >50% following the guidelines of All India Soil and Landuse Survey, New Delhi.The FCC of Indian Remote Sensing (IRS) satellite image LISS-III 1-C collected from Indian Institute of Remote Sensing (IIRS), Dehradun for the year 2006 on a scale of 1: 50,000 was used to prepare landuse map using ERDAS IMAGINE 9.0 software.The landuse map was prepared through unsupervised classification assigning 150 classes initially and then merged into six classes based on the image characteristics like tone, texture, shape, colour, association, background, etc. following standard visual interpretation techniques and ground truth and the information available from landuse reports and field surveys.Land uses were classified as dense forest, open forest, rain-fed agriculture, irrigated agriculture, waste land, and water bodies.

Study on soils
Soil mapping units (SMUs) were delineated by taking into account land slope and landuse together using 'decision tree classifier' in environment for visualizing images (ENVI) Version 4.4 software.The land slopes mentioned earlier were divided into three classes, such as lower (<9%), medium (  9% -<33%) and higher (  33%) slopes, and land uses were put into 7 classes (including unclassified area).In this way, by using different expressions at decision nodes, the study watershed was divided into totally 21 classes and executed in landuse classification map of 2006 to get the final soil mapping unit map for the entire study area.
After delineation of soil mapping unit, soil sampling points were decided and soil samples representing all land uses and slopes were collected.Soil samples were collected for 15 different combinations of 3 slopes and 5 land uses (except water body and unclassified areas).For these 15 combinations, a total of 45 sampling points were selected and a point map was generated using GIS.The geographic location of each sampling point was identified in the field using GARMIN Global Positioning System (GPS) and from each location soil samples were collected from two depths (0-15 and 15-30 cm).Bouyoucos Hydrometer method was used to determine the percentage of sand, silt and clay.Bulk density and saturated hydraulic conductivity of soil was derived by Soil Texture Triangle Hydraulic properties calculator, EC by EC Meter, pH (1:2.5) in water by Beckman Glass electrode pH meter and organic C content by modified Walkley and Black method, available N by Alkaline KMnO 4 method using Nitrogen Analyzer (Gerhardt), available P by Colorimetrically by Olsen's extraction method and available K by 1 N Neutral NH 4 OAC extraction using Flame Photometer.

Data analysis
An erosion tolerance index (ETI) was established from the integration of soil loss tolerance (T-value) and average annual soil loss (A).Model functions used for fuzzy membership classification of land attributes are based on the Semantic Import model (SI) approach.With this approach, the most relevant indicators representing various functions were selected first (Table 1), allowing maximum erosion as long as it was able to perform these functions and then transforming measured values of indicators into a common membership grade (0-1) through fuzzy modeling (Wymore, 1993;Mandal et al., 2009) according to the class limits specified by McBratney and Odeh (1997).If MF (xi) represents individual membership function (MF) values for i th land property x, then, the basic SI model function took the following form in the computation process: where, b is central concept and d is the width of transition zone.As there were 'n' soil characteristics to be rated, the membership function values of individual soil characteristics under consideration were then combined using a convex combination function to produce a joint membership function (JMF) of all attributes, Y as follows.
where Y is the convex membership functions (JMF) of all attributes, i  is the weighting factor for the i th soil property x, and is the membership grade for the i th soil property x.Model functions used for fuzzy membership classification of soil attributes are based on the Semantic Import (SI) approach which utilizes a bell shaped curve (Burrough et al., 1992).This approach consists of two basic functions: symmetric and asymmetric.
i. Asymmetric left (Model 3): 'more is better' ii. Asymmetric right (Model 4): 'less is better' Model parameters include lower crossover point (LCP), the central concept (b), upper crossover point (UCP), and width of transition zone (d) (Burrough et al., 1992).The soil parameters taking the model type were shown in Table 2.
Weights were assigned to the potential indicators that reflected their relative importance.The converted values on 0 to 1 scale were then multiplied by the weights assigned to them.Summing the values of the weighted parameters, a quantitative value SLTL indicating the state of soil was obtained for each soil mapping unit.A comprehensive guideline for the estimation of T values based on the favourable rooting depth was followed as per general guidelines for assigning soil loss tolerance "T" (USDA-NRCS, 1999).Soils having higher T-values are more resistant to erosion.In this way, for each soil sampling point, adjusted T values at cell/ polygon/pixel level were determined for different soil types and on a watershed basis with different land uses for better watershed planning and management and then weighted SLTL values were determined for a particular mapping class from all polygons belonging to that class for comparison with the potential soil loss.
After the determination of SI model parameters, the next step followed was to determine potential soil loss.In this study a combination of remote sensing, GIS, and RUSLE model (Renard et al., 1997) was used to estimate the soil erosion rate on a cell-by-cell/pixel basis.Revised Universal Soil Loss equation (RUSLE) is given as: where A is computed average soil erosion per unit area, expressed in the units selected for K and for the period selected for R (Mg ha -1 ); R is a rainfall-runoff erosivity factor (Mg cm ha −1 h −1 per year); K is a soil erodibility factor ((Mg ha -1 per Mg cm ha −1 h −1 ), L is slope length (m), S is the slope steepness (%); C is cover management factor and P is support practice factor.For determination of Rfactor, available daily rainfall data from 1970-1985 for Bungidhar and Jaurasi raingauge stations, from 1970-2008 for Tamadhaun and Kedar rain gauge stations in the watershed were collected from Divisional Forest and Soil Conservation Office, Ranikhet.In the present study, the above relationship of determination of R-factor could not be used as continuous rainfall intensity data in the watershed were not available.Therefore, as an alternate approach relationship proposed by Ram et al. (1969) to determine daily EI 30 values based on daily rainfall for Dehradun in the Garhwal foot hills was used which stated as: where Y is the daily EI 30 value in Mg-cm ha -1 h -1 and X is the daily rainfall (mm) exceeding 12.5 mm.A point map of four rain gauge stations (Bungidhar, Jaurasi, Tamadhaun and Kedar) was prepared in ArcGIS 9.3 by locating it on mosaiced toposheets.Thiessen polygons were created by nearest point method in ArcView 3.2 software and then transferred to ArcGIS 9.3 for calculating area of each thiessen polygon.Thiessen polygon map with R-factor attribute was overlaid on soil mapping unit map and clipped individually for each Thiessen polygon using 'clip' option in GIS and R-factor for each polygon was derived and weighted R-factor was determined using following relationship.
where R i is the calculated rainfall-runoff erosivity factor for i th rain gauge station, A i is the area of i th Thiessen polygon and A is the total area of polygons.Daily rainfall-runoff erosivity (R) factors were determined for all the four rain gauge stations.Then annual erosivity factors were determined and iso-erodent map was generated for the watershed using IDW technique in ArcGIS.K-factor was determined using soil texture and organic matter content (Mandal et al., 2009) at each polygon level and then weighted K-factor was derived for each mapping class.Slope length was determined using ArcGIS by generating flow accumulation and flow length map by raster grid cumulation and maximum downhill slope method (Lu et al., 2004).Polygon-wise slope length and slope steepness was obtained from flow length map and slope map,

Computation of erosion tolerance index
To calculate the erosion tolerance index (ETI) for each mapping unit, the comparison was done between calculated annual soil loss rate (A) and soil loss tolerance level (T) in the form of T/A ratio.A unit having a T/A ratio less than 1.0, indicates that annual soil loss could exceed the tolerance level; hence indicating the vulnerability of the land.The ETI was then established using an asymmetric left fuzzy function.This function shows that T/A of 1.0 is selected as an ideal point, while 0.5 (soil loss exceeding two times the threshold value) is for LCP.The result was expressed in continuous values from 0 (high vulnerability) to 1.0 (almost no risk) (Baja et al., 2002).To determine ETI, T/A ratio was determined for each mapping class and for each sub-watershed.For each mapping class, ETI was determined by taking b 1 and d 1 as 1 and 0.5, respectively.Weighted ETI value was then calculated for each class and for each subwatershed and used for prioritization of sub-watersheds.

Thematic maps of the watershed
The delineated Bino watershed and its sub-watershed boundaries have been shown in Figure 1.Area and perimeter of the watershed were found to be 296.17km 2 and 83.24 km, respectively.It was found from the digital elevation model (DEM) (Figure 2) that the elevation of the watershed varied from 802-2884 m above mean sea level (msl) and maximum area (38.15%) was under the elevation range of 1200-1400 m above msl, whereas the lowest 2.39% was under 2400-2884 m amsl.Area covered up to 2000 m above msl was 87.04%, whereas area up to 1600 m elevation was 64.24% of the total geographical area of the watershed.The generated slope map (Figure 3) through GIS showed that 49.22% area was covered under slope up to 35%, whereas 50.78% area was with more than 50% slope.Based on general slopes in different portions of watershed, the land may be categorized as a valley, moderate and steep hill areas.However, the study watershed belonged to steep slope category, as only 4% area was under mild slope (<10%).The soils had coarse textures (sandy loam, loam and sandy clay loam, stony) and with high erodibility.

Potential annual soil loss
The mean weighted annual rainfall determined from the Thiessen polygon developed for the watershed was found to be 931.3mm.It was also observed that the rainfall-runoff erosivity factor for monsoon season (R mon ) and annual (R ann ) ranged from 272.52-498.07 and 390.8-624.72Mg-cm ha -1 h -1 , respectively.K, LS, C and P-factor under different land uses in the watershed varied from 0.19-0.33Mg ha -1 Mg -1 cm -1 ha h, 0.1-34.4,0.001-0.34and 0.13-1.0,respectively.On the basis of the obtained values of these factors, potential annual soil loss rate (A) for each mapping class and sub-watershed was determined by clipping sub-watershed boundary and was then compared with respective soil loss tolerance limit value (T).It was found that potential annual soil loss varied from 0. 49-3.64, 1.93-18.83, 4.07-55.49Mg ha -1 in the lower middle and higher slopes, respectively.

Quantification of SLTL in the study watershed
Soil samples were collected from each represent mapping class (Figure 5) and analyzed in the laboratory.The values of individual soil property representing each   landuse were clubbed and their mean values were obtained.Soil textures found in the watershed include loam, silt loam, sandy loam and sandy clay loam in 0-15 cm solum depth, and loam, silt loam, sandy loam, sandy clay loam, loamy sand and clay loam in 15-30 cm solum depth.The values of pH ranged between 4.52 and 6.60 in 0-15 cm, and 4.66 and 6.71 in 15-30 cm solum depth which indicates that soils in the watershed are acidic in nature.The values of EC for the corresponding soil depths varied from 0.02-0.28and 0.02-0.2ds m -1 .Organic C content ranged from 0.66-2.31 in 0-15 cm and 0.61-2.27 in 15-30 cm depth.Available Nitrogen (N), Phosphorous (P), and Potassium (K) ranged from 37.63-240.14,37.01-68.44,69.58-669.14kg ha -1 in 0-15 cm and from 25. 1-225.79, 34-72.01, 50-328.22kg ha -1 in 15-30 cm solum depth, respectively.The value of K-factor varied from 0.21-0.27and organic C from 1.03-1.87.The organic matter, N, P, and K contents in dense forest were the highest indicating organic matter build up in the surface soil.
Soil groups were determined on the basis of aggregated scores (Table 2) by summing up the weighted scores of all the indicators which represent the level of resistance to erosion.The SLTL values of different soils were determined on the basis of soil group and soil depth.SLTL values for each polygon were determined and then weighted SLTL values for each landuse at a particular slope class were determined.
It was observed from the calculated SLTL values that in the study watershed these values varied from 2.5-10 Mg ha -1 yr -1 .It was also found that SLTL values in the sub-watersheds varied from 3.1-10.0,5.0-10.0 and 6.7-10.0Mg ha -1 in the lower, middle and high slopes, respectively.

Erosion tolerance index (ETI) for subwatersheds
For each sub-watershed, five land uses and three slope classes, 15 ETI values for all mapping classes (landuse units) were determined and thus, a total of 135 ETI values for the whole watershed were obtained.It was observed from data in Table 3 that in case of lower slope with open and dense forest, rain-fed and irrigated agricultural areas except waste land in the entire sub-watersheds annual soil loss rate was much lower than the soil loss tolerance limit, so no ETI values were considered.ETI values under waste land varied from 0.374-0.930.In middle slope, ETI values for the area under dense forest and irrigated agricultural land were also not considered as annual soil loss rate was much less than the soil loss tolerance limit.
In case of rain-fed agricultural land, open forest and waste land ETI values ranged from 0.160-0.222,0.299-0.700and 0.142-0.180,respectively.On higher slopes ETI values under open forest, dense forest, rain-fed agricultural land, irrigated agricultural land and waste land varied from 0.170-0.249,0.439-1.0,0.117-0.130,0.216-0.389and 0.121-0.137,respectively.The lowest ETI was in Bino sub-watershed, whereas the highest for Syalde sub-watershed in all lower, middle and higher slopes.In some areas of the watershed, annual soil loss rate was less and in some areas it was more than SLTL.Therefore, weighted ETI value was determined for each sub-watershed and it was found that the weighted ETI values for Basolagad, Baya Gadhera, Bino Nadi.Chauna, Gaya Gadhera, Juniya Gadhera, Masangari Nadi, Syalde and Tamadhaun sub-watersheds were 0.218, 0.232, 0.238, 0.238, 0.212, 0.337, 0.244, 0.380 and 0.353, respectively.These weighted ETI values were used for prioritization of subwatersheds by taking the criteria more is the better model.The result of prioritization analysis revealed that the highest priority was given to Gaya Gadhera sub-watershed, second most priority to Basolagad and least priority to Syalde sub-watershed and prioritization map was shown in Figure 6.

DISCUSSION
Potential soil loss and soil loss tolerance limit values in lower, middle and higher slopes revealed that potential soil loss in lower slope was within the permissible limit, but middle and higher slopes may be treated with different conservation measures to bring soil loss within permissible limits in all the land uses except under dense forest and irrigated agricultural land in middle slope.The generally accepted maximum limit of soil loss (or T-value) is 11.2 Mg ha -1 yr -1 (Wischmeier and Smith, 1978), while Rubio (1986) considered a T-value of 20 Mg ha -1 yr -1 .In Northwestern Himalayas of India, T value varied from 5.0-12.5 Mg ha -1 yr -1 .When SLTL value for the study area was overlaid and clipped in GIS environment from the map developed by Mandal et al. (2009) for Uttarakhand state, it was found to be only a single value of 10 Mg ha -1 yr -1 .Because of the large grid size (10 × 10 km), this map failed to provide SLTL values for each landuse and slope class.However, to make the landuse planning more effective, it was considered appropriate to determine and use localized SLTL values.Thus, localized values of SLTL were determined for each landuse to make location specific landuse planning which varied from 2.5-10 Mg ha -1 yr -1 .It was also observed that within the soil group, SLTL values varied with soil depth.The SLTL values depend on many factors and varied from one set of conditions to another.An average soil loss of 5 Mg ha -1 yr -1 has been considered as the limit for shallow soils (Hudson, 1986).Lal (1985) observed that for shallow soils with root restrictive layers at 0.2 to 0.3 m depth, a T-value is set at 1 Mg ha -1 yr -1 .But in this study, for shallow (36-56 cm) soil depth in waste land as low as 2.5-5 Mg ha -1 yr -1 of SLTL values with a maximum potential soil loss of 20.32 Mg ha -1 yr -1 were observed and thus, needed to be protected on priority basis.The variations in soil groups and SLTL values were due to different levels of the most sensitive indicators, viz.saturated hydraulic conductivity, Kfactor and organic carbon.Assignment of site specific SLTL values will be helpful in assessing the vulnerability of soil in the watershed after comparing with a potential soil loss.
It was observed that in middle slope the magnitude of ETI value was in the order of waste land<rain-fed agricultural land<open forest.This showed that in middle slope wasteland is more vulnerable to soil erosion compared to rain-fed agricultural land and open forest that means it needs greater attention.Whereas in higher slope the magnitude of ETI value was in order of rainfed agricultural land<waste land<open forest<irrigated agricultural land<dense forest.This may be due to higher SLTL and less soil erosion in dense forest and irrigated agricultural land compared to other land uses.ETI values were lower in dense forest may be due to more organic matter content in the soil and so the SLTL value.It was also found that ETI value decreases with increase in slope gradient in all the land uses as it is obvious that higher slope is more vulnerable to soil erosion with less SLTL as soil depth is low.It was observed that ETI values for each land use in the Bino sub-watershed were at par with Masangari sub-watershed (Table 3).This could be due to the fact that both sub-watersheds were geo-morphologically similar.Again, the highest priority was given to Gaya Gadhera sub-watershed, whereas least to Syalde sub-watershed as per ETI values.Figure 1 showed that Gaya Gadhera subwatershed is at a higher elevation and almost circular shaped compared to Syalde sub-watershed which is elongated one.This implies the higher soil erosion generation in Gaya gadhera sub-watershed compared to Syalde sub-watershed.Therefore, it corroborated the higher magnitudes of ETI value in Syalde sub-watershed.
The steep slopes of the area have caused soil erosion, resulting in the meager depth of topsoil and the prevalence of gravel and rocks in the surface and causing many exposed rocky areas, poor soil nutrition and lack of organic matter in the soils.There is a continued and ever increasing threat to productivity, food security and environmental quality, especially in ecologically sensitive eco-regions, characterized by fragile soil, high population density and harsh environment.The quantitative ETI values and prioritized map produced in this study may be used as the guide for the policy makers for examining and deciding what land use management practices and which sub-watershed should be adopted on priority on a given individual land units to reduce the degree of soil erosion up to the soil loss tolerance limit.

Conclusions
The main problem of the study area in Indian Himalaya was steep slopes and mismanagement of different land uses, particularly agricultural systems which are ploughed parallel to the slope direction.This problem led to the soil erosion as high as 80 Mg ha -1 yr -1 .Again worldwide the generally accepted maximum limit of soil loss is 11.2 Mg ha -1 yr -1 .Therefore, this paper presented the approaches to analyze potential soil loss and determining localized soil loss tolerance limit and showing their importance and effectiveness of ETI in watershed management planning.Implementation of watershed programs should target not only the reduction in erosion rate, but also reduction in the gap between the potential erosion rate and the soil loss tolerance rate in an area as soil loss tolerance level of all the soils is not the same.The methodological development followed a framework based on RUSLE model, fuzzy set analysis, remote sensing and GIS.The localized values of SLTL were found to be varying from 2.5-10 Mg ha -1 yr -1 for different land uses with maximum 55.4 Mg ha -1 yr -1 of potential soil erosion from rainfed agricultural land.It was found that ETI values ranged from 0.212-0.380and the value decreases with increase in slope gradient in all the land uses as it is obvious that higher slope is more vulnerable to soil erosion with less SLTL as soil depth is low which proved the hypothesis of the study.
The ETI may be used as the basis for examining land use practices for adopting on a given land unit to maintain soil loss within a tolerable limit.The ETI values used for prioritization and results revealed that the highest priority was given to Gaya Gadhera subwatershed and least to Syalde sub-watershed as per ETI values.The quantitative approach, produced in this study may serve as a guide for the policy makers to decide the use of land with different farming system models and landuse management practices.Use of the SLTL will improve conservation planning, help to meet erosion control regulations for development of sustainable watershed management in the study area as well as in this portion of India.Remote sensing and GIS technology can be used as an alternative to the conventional method of prioritization of sub-watersheds for implementing soil conservation practices.This study has provided a package of scientific knowledge that can be used to transfer the technology of land use management from the researcher to the user.

Figure 2 .
Figure 2. Digital elevation model of the Bino watershed.

Figure 3 .
Figure 3. Slope map of the study watershed.

Figure 4 .
Figure 4. Land use map of the study watershed.

Figure 6 .
Figure 6.Prioritization map of sub-watersheds under study watershed.

Table 1 .
Representative indicator for each fuzzy function with corresponding weight.

Table 2 .
Categorical ranking of soil attributes used in the study.
Values in the parenthesis are converted score between 0 to 1, Model 3 (Asymmetric left): more is better, Model 4 (Asymmetric right): Less is better'.

Table 2 .
Categorical ranking of soil attributes used in the study.