Impacts of Soil Erosion on Soil Quality and Agricultural Sustainability in the North-Western Himalayan Region of India

Abstract

Erosion by water reduces soil quality and hence crop yield. Understanding the response of crop yields to soil erosion is vital in assessing agriculture’s vulnerability to erosion. However, these effects are difficult to quantify. The study presents a quantitative relationship between soil erosion and soil quality and productivity of rainfed wheat (Triticum aestivum) by comparing field plots with different degrees of erosion in some sub-tropical alfisols in the Doon Valley region of India. By comparing the topsoil depth with the reference site as the control, erosion severity was classified into different phases such as slight, moderate, severe, and very severe. A quantitative, weighted additive model was used to evaluate soil quality for different phases of erosion using soil clay content, water holding capacity, soil aggregate, soil organic carbon, pH, CEC, total N, available P, and available K. The synthesis of long-term experimental data revealed that the mean soil erosion rate varied from 5.5 Mg ha⁻¹ yr⁻¹ in slightly eroded plots to 33.4 Mg ha⁻¹ yr⁻¹ in very severely eroded plots. Compared with the reference forest, the soil organic carbon (SOC) declined by 81.4% and water holding capacity by 31% in severely eroded soils. A substantial loss of total N, extractable P, and available K was also observed. Water stable aggregates (WSA) decreased from 86% to 12.6%, and the cation exchange capacity (CEC) from 25 to 12.6 c mol⁽⁺⁾ kg⁻¹. The soil quality index was 0.7 in slightly eroded compared with 0.4 in severely eroded soil. Similarly, the sustainable yield index for wheat was 0.9 and 0.6 for slightly and severely eroded soils, respectively. Thus, there is a strong need to identify land management systems that reduce erosion risks, restore eroded soils, and enhance soil quality.

1. Introduction

Accelerated soil erosion degrades soil quality, and severe degradation of soils has been responsible for the collapse of numerous historic and once thriving civilizations [1,2,3,4]. Furthermore, land misuse and soil mismanagement caused by indiscriminate agricultural expansion [5] can undermine agricultural productivity and environmental quality [6,7]. Erosion-induced soil and environmental degradation have plagued humanity since the dawn of settled agriculture about 10–12 millennia ago [6,8], adversely affecting a large proportion of the world’s arable land [9]. The global soil erosion rate from cropland is estimated at 10.2 Mg ha⁻¹ y⁻¹, which is more than 24 times the estimated average natural erosion rate (0.42 Mg ha⁻¹ y⁻¹) [10]. The problem is more severe for soils of the tropics, which are highly prone to erosion and other degradation processes [4,11].

Erosion depletes topsoil depth (TSD), soil organic carbon (SOC) concentration and stock, and nutrient status. It strongly impacts soil texture and structure, available water holding capacity (WHC), and water retention and transmission characteristics [12], which moderate soil quality and determine crop yield [13]. Soil erosion exacerbates the depletion of N, P, and K, and reduces the cation exchange capacity or CEC [14,15]. It aggravates chemical and nutritional constraints [16,17] on crop production. Specifically, it decreases AWC [18] and SOC concentration [19,20] and increases soil bulk density [7,21].

Erosion-induced changes in soil properties have been documented through simulated erosion by desurfacing [22,23,24,25,26,27] and comparing soil properties and crop yield on different phases of erosion [28,29,30,31,32,33,34]. However, there is a large variability and often an overestimation of the adverse effects because of inconsistent approaches and methodologies [1,35].

Some soils are more vulnerable than others, and the degradative effects are specific to soils, ecoregions, and crops, depending on the management and environmental factors [36]. The north-west Himalayan region of India is one of the hot spots of erosion-induced soil degradation [37,38]. Further, adverse impacts of soil erosion may be aggravated by climate change and the attendant increase in frequency and intensity of extreme events [39]. Thus, there exists large uncertainty about the impacts of future erosion on agronomic productivity and environmental quality [1]. Despite widespread awareness of the severity of erosion and its adverse impacts [40,41,42], quantitative information about the effects of past or future erosion on soil quality and agricultural productivity is scanty, especially for the vulnerable region of the western Himalayas. Thus, this study aimed to quantify the relationship between the severity of erosion and soil quality and productivity of rainfed wheat (Triticum aestivum) in north-western Himalayan India. How much soil erosion would be having what extent of damage to soil quality has been studied by minimizing the confounding effect of landscape on crop yield. Unless the erosion phases are carefully selected within a landscape position, the interaction effects of the landscape position and soil erosion phase make it challenging to establish a cause–effect relationship between soil erosion and crop productivity via alteration of soil quality. This study offers observational evidence from a series of erosional phases to answer the following research questions:

(i)

Can the magnitude of past soil erosion be assessed in comparison with the uneroded phase as the reference point?

(ii)

Does progressive soil erosion significantly impact soil quality and crop yield in a sloping landscape?

2. Materials and Methods

2.1. Soils and Site Description

The present study was conducted for 25 years on four contiguous runoff plots (60 × 7 m) with different degrees of erosion in the Doon Valley region of the lower Himalayas. The Doon valley is located in the Shiwalik Himalayas between latitudes 29°55′ and 30°30′ N and longitudes 77°35′ and 78°24′ E. It is a saucer-shaped valley about 25 km wide and 80 km long. The valley is a longitudinal, intermontane, synclinal depressed bouldery [43,44] filled with coarse clastic fan and gravel of the late Pleistocene and Holocene [45,46]. Geomorphologically, the landforms in the region are formed by variable erosion, deposition, and tectonic activity over time and space [43,47]. The deep alluvium is derived from sedimentary rocks, including shale, sandstone, limestone, siltstone, and conglomerate. The study region as a whole is characterized by a humid subtropical climate. The soil of the study site is alfisol as per the USDA taxonomic classification. The soil texture class of the experimental erosion plots is sandy loam and the reference site texture is sandy clay loam. The average annual rainfall is 1625 mm received over 78 rainy days, and the mean evaporation is 1092.7 mm [48]. Surface runoff is 40 to 50 percent of the total rainfall from the land with an 8-percent slope and 30 to 45 percent for 4-percent slope [48]. Rainfall exceeds evapotranspiration from June to October while water deficit prevails for the other months. The reference site was selected in an undisturbed forest close to the experimental location.

2.2. Experiment

The experimental layout of contiguous runoff plots is shown in Figure 1. Runoff plots were established on a landscape with a natural slope gradient of 0.5, 2.5, 4.5, and 9.5%, causing minimal soil disturbance. Different phases of erosion (e.g., slightly, moderately, severely, and very severely) were created naturally over time for the chosen land slopes. By comparing the topsoil depth with the reference site as control, erosion severity was classified into different phases: slight, moderate, severe, and very severe. Based on the cumulative loss of topsoil depth, the loss of topsoil was estimated at 5%, 13%, 29%, and 41%, respectively, in slight, moderate, severe, and very severe phases. Thus, soil properties under different erosional phases were compared by taking three composite samples as pseudo-replicates within each runoff plot. Maize (Zea mays) and wheat (Triticum aestivum) were grown on each erosional plot under rainfed conditions. Each plot was bermed with earth to prevent outside run-on and facilitate farm operations. Each runoff plot was equipped with a flume, a Coshocton-type runoff sampler [49] which samples 1% of the total flow, and a collection tank that retains runoff and sediment. Runoff was measured for every erosive event, and measurements were made daily during heavy and continuous rains. Runoff was expressed as the equivalent depth of water for the plot. Suspended sediment concentrations were determined gravimetrically using the filtration method [49]. Soil loss (the product of runoff volume and sediment concentration) was expressed as Mg per hectare. Soil loss and runoff data for each sampled runoff event were summed to obtain the annual losses. Soil samples (0–15 cm depth) were collected from each erosion phase as well as from the reference site for the analyses of physical (clay content. WHC, and water stable aggregates or WSA) and chemical properties (pH, CEC, SOC, total N, Olsen P, and available K) and texture. To determine WSA, the soils were sieved using the wet sieving technique [50]. WSA was assessed according to methods described by Cambardella and Elliott [51] and expressed as aggregates > 1.0 mm. A separate set of composite samples were ground and sieved through a 2 mm sieve, and 50 g of the sieved soil was used for particle size analysis (clay content) by the hydrometer method [52]. The remaining soils were used for the determination of water holding capacity [50], soil pH in a 1:2.5 soil: water suspension (Thomas, 2001), SOC by the method of Walkley and Black [53], CEC [54], soil nitrogen by the Kjeldahl method [55], available P [56], and available K [57].

Conventional tillage was performed using a mould board plow as preparatory tillage, followed by a seed driller for wheat sowing. Wheat variety UP 2572 was sown using a 100 kg ha⁻¹ seed rate at 22.5 cm row spacing to a 5–6 cm soil depth in late September every year. The crop was fertilized with 60 kg N, 13 kg P, and 16 kg K. Half of the N and the total amount of P and K were applied at the sowing time. The remaining quantity of N was applied at the booting stage of the crop. For the control of wheat rust, whenever the appearance of the disease was noticed, propiconazole (tilt) was sprayed with the recommended dose. The hand harvest method was used for wheat yield determination by taking replicate samples (1 × 1 m) at crop harvest. Wheat grain yield was recorded at 12% moisture content.

The sustainability yield index ‘SYI’ of treatment over a period of ‘n’ years was derived by the equation: SYI = (Y_t − σ)/Y_max [38], where Y_t is the treatment mean, σ is the overall standard deviation, and Y_max is the maximum attainable yield in the experiment during the reporting period.

2.3. Soil Quality Evaluation

Soil removal alters soil’s physical and chemical properties and changes overall soil quality. A quantitative, weighted additive model was used to evaluate soil quality for different phases of erosion [58]. The key 9 soil parameters, namely, clay content, WHC, WSA, SOC, pH, CEC, total N, available P, and available K, were converted to unit-less scores (0–1 scale) using appropriate membership functions, and each of these values were multiplied by respective weights. In an erosional landscape, the crop yields are related to erosion-induced differences in soil properties. The most important soil properties are WHC, WSA, SOC, CEC, clay content, and nutrients (pH, N, P, and K). This weighted additive model provides an overall measure of soil quality. A synthesis assessment index was developed to evaluate the soil quality index (SQI) [59]. The SQI represents erosion-related factors on soil quality in one number. The SQI is expressed as follows (Equation (1)):

SQI = ∑Wi Si

(1)

where Wi = Weights of the indicators, and Si = the score of the indicators; thus, the 9 indicators (measured soil attributes) were combined into an SQI.

The evaluation involves combined fuzzy analysis and the analytic hierarchy process (AHP) methods [60,61,62]. The fuzzy analysis was used to calculate the membership functions for each evaluation factor, following the methods suggested by many researchers [58,63,64,65]. Soil parameters were transformed using the following two types of fuzzy membership functions: upper limit function and peak limit function in terms of their relationship with soil functionality. Soil functionality is the evaluation approach that involves properties and processes relating to its ability to function effectively [66,67].

Upper limit function (Equation (2))

$$y_i=\left\{\begin{array}{c}0, u_i\le u_{li}\\ \frac{1}{1+a_i{(u_i-u_{oi})}^2}, u_{li}<u_i<u_{oi}, (i=1, 2, \dots , m)\\ 1, u_i\ge u_{oi}\end{array}\right.$$

(2)

where y_i is the value of the membership function, ai is a constant, u_i is the measured value of the evaluation factor, u_oi is the optimum (or upper) value of the evaluation factor, u_li is the lower limit value of the evaluation factor, and the subscript I denotes the ith value of the factor in the range from 1 to m. This function was chosen for WHC, SOC, WSA, CEC, N, available P, and available K.

The peak function (Equation (3)) is as follows:

$$y_i=\left\{\begin{array}{c}0, u_i\le u_{l1} or u_i \ge u_{l2}\\ \frac{1}{1+a_i{(u_i-u_{oi})}^2}, u_{l1}<u_i<u_{l2}, (i=1, 2, \dots , m)\\ 1, u_i=u_{oi}\end{array}\right.$$

(3)

where u_l1 and u_l2 are the lower and upper limits for u_i. This function was used for clay content and pH.

Parametric values for the above two membership functions for the 9 measured soil attributes are listed in

Table 1. Upper limit membership functions and parameters used for various soil attributes.

Membership Function	Attributes	Parameters of Membership Function
		ai	uli	uoi
Upper limit function	WHC (%)	4.75 × 10−3	10	50
	SOC (%)	5.1	0.09	1.25
	WSA (%)	1.35 × 10−3	9	81
	CEC [c mol(+) kg−1]	2.45 × 10−2	3	20
	N (%)	1450	0.004	0.08
	Available P (Kg ha−1)	1.65 × 10−2	1.4	21
	Available K (Kg ha−1)	9.10 × 10−5	14	270

and

Table 2. Peak membership function and parameters used for various soil attributes.

Membership Function	Attributes	Parameters of Membership Function
Peak function		ai	Ul1	Ul2	uoi
	Clay (%)	0.0071	3	76	35
	pH	0.475	3	11	7

.

The AHP was adopted to evaluate weight objectively and accurately. Further, quantitative and qualitative analyses are adequately integrated into a model to assess soil quality. This integration involves analysis, judgment, and the synthesis process. Specifically, a complex problem is partitioned into several layers corresponding to specific dominating factors. These factors were then prioritized based on their relative importance. Different weights were then decided by sensitivity analysis, their maximum eigenvalues, and the consistency test [68]. The model establishment is an iterative process, and both Si and Wi were finalized through model calibrations and validations. Because the focus is to assess the impact of erosion on soil quality, a hierarchy of those factors is established, which have a direct relationship between erosion and soil properties and crop productivity. The judgment matrices and the consistency ratio outcomes are shown in

Table 3. Judgment matrix of various soil parameters used.

	WHC	SOC	WSA	CEC	Clay	pH	N	P	K
WHC	1	2	5	3	6	4	6	8	6
SOC	0.5	1	2.5	1.5	3	2	3	4	3
WSA	0.2	0.4	1	0.6	1.2	0.8	1.2	1.6	1.2
CEC	0.333	0.667	1.667	1	2	1.333	2	2.667	2
Clay	0.167	0.333	0.833	0.5	1	0.667	1	1.333	1
pH	0.25	0.5	1.25	0.75	1.5	1	1.5	2	1.5
N	0.167	0.333	0.833	0.5	1	0.667	1	1.333	1
P	0.125	0.25	0.625	0.375	0.75	0.5	0.75	1	0.75
K	0.167	0.333	0.833	0.5	1	0.667	1	1.333	1

and

Table 4. Range and mean values of runoff and soil loss in different phases of soil erosion.

Erosion Status	Runoff (mm)	Soil Loss (Mg ha−1)	Soil Loss Above Permissible Limits
Slight	153.5 d ± 51.16	5.5 d ± 1.1	Within limits
	(83.3, 329.9)	(3.3, 9.5)
Moderate	255.8 c ± 88.1	10.8 c ± 1.6	Within limits
	(160.4, 517.2)	(8.60, 16.4)
Severe	362.8 b ± 121.1	22.7 b ± 4.4	1.81 times
	(187, 698.8)	(17.7, 35.6)
Very severe	465.3 a ± 153.2	33.4 a ± 4.9	2.67 times
	(288.3, 886.4)	(28.2, 43)

Mean followed by the same letter are not significantly different based on Tukey’s HSD (p < 0.05); values in the parenthesis are a minimum and maximum.

. In the final calibrated and validated model, the total consistency ratio (CR) is 1.531 × 10⁻¹⁶ (=2.22 × 10⁻¹⁶/1.45), which is less than 0.1. Therefore, the consistency test is satisfied [61,69,70]. The consistency test for hierarchy sorting was done by the method of Zhang et al. [60] and Alam et al. [71]. Additional information on the distribution of hierarchy and weights of attributes is available elsewhere [67,69,70]. Relative weights of different soil parameters derived through AHP were in the order of 0.34, 0.17, 0.07, 0.11, 0.06, 0.09, 0.06, 0.04, and 0.06 for WHC, SOC, WSA, CEC, clay content, pH, N, P, and K, respectively.

2.4. Statistical Analysis

All data were subjected to analysis of variance (ANOVA) using a statistical package (SAS 9.3) to assess the impact of erosion on soil quality. The statistical significance of each attribute was assessed using Fisher’s least significant difference (LSD) at the p < 0.05 level. Regression analysis was done between SQI and SYI to evaluate their quantitative relationship.

3. Results and Discussion

3.1. Soil Quality Indicators

The distribution of the soils under the natural ecosystems in the region is generally influenced by the geological substrate and landscape positions. However, several characteristics of the soils in the study region are also affected by erosion and management practices, including land use.

In the present study, soil loss by water erosion was the highest in the very severe erosional phase, and relatively less soil loss occurred in the slight erosion phase (Figure 2). The data presented in

Table 5. Range and mean values of soil parameters in different erosional phases.

Soil Attribute		Erosion Phase
	Reference Site	Slight	Moderate	Severe	Very Severe
Clay content (%)	22–26 a (23 ± 1.5)	18–23 a (20.8 ± 2.2)	17–20 a (18.4 ± 1.1)	17–21 a (18.5 ± 1.4)	18–24 a (19.3 ± 2.7)
WHC (%)	46–62 a (53 ± 7.4)	44–52 b (48.6 ± 3.4)	42–50 b (45.8 ± 3.03)	40–45 c (42.8 ± 1.92)	35–40 d (36.6 ± 2.07)
WSA (%)	74–92 a (86 ± 5.9)	19.3–22.6 b (20.84 ± 1.2)	18.6–21.8 c (19.8 ± 1.3)	14.4–17.6 d (15.8 ± 1.2)	10.6–14.2 e (12.6 ± 1.4)
SOC (%)	1.94–2.31 a (2.04 ± 0.1)	0.72–0.86 b (0.80 ± 0.05)	0.64–0.76 c (0.70 ± 0.05)	0.52–0.60 d (0.56 ± 0.03)	0.30–0.45 e (0.38 ± 0.06)
pH	6.1–6.3 a (6.2 ± 0.07)	5.6–5.9 b (5.72 ± 0.13)	5.6–5.9 b (5.76 ± 0.11)	5.8–6.2 c (6.02 ± 0.15)	5.9–6.3 d (6.14 ± 0.1)
CEC (c mol(+)kg−1)	20.8–32.4 a (25 ± 4.0)	13.5–16.4 b (15.20 ± 1.11)	13.3–16.0 b (14.52 ± 1.26)	12.7–14.0 c (13.44 ± 0.51)	11.5–13.4 d (12.68 ± 0.7)
Total N (%)	0.16–0.20 a (0.17 ± 0.01)	0.06–0.0.09 b (0.07 ± 0.01)	0.06–0.0.08 b (0.07 ± 0.008)	0.03–0.0.06 c (0.04 ± 0.01)	0.02–0.0.04 c (0.03 ± 0.01)
Available P (Kg ha−1)	13–20 a (16.4 ± 2.4)	28–36 b (32 ± 3.16)	24–35 b (29 ± 4.5)	18–28 c (24 ± 4.0)	15–24 d (19.8 ± 3.5)
Available K (Kg ha−1)	285–292 a (288 ± 2.6)	208–240 b (221.2 ± 13.2)	180–220 b (197.2 ± 17.0)	140–205 c (171.2 ± 27.5)	135–180 d (151.0 ± 17.3)

Different alphabets indicate significant differences between rows ai p < 0.05.

revealed that the mean runoff varied between 153.5 mm and 465.5 mm in the slight and very severe erosion phases, respectively. There were significantly higher water losses through runoff with the progressive increase in slope gradient. Compared to the slight erosion phase (5.5 Mg ha⁻¹), the sediment loss was significantly higher in the severe (22.7 Mg ha⁻¹) and very severe erosion phases (33.4 Mg ha⁻¹) than in other phases and reference sites. Excessive sediment loss and overland flow are of particular concern in Indian Himalayas [11,71]. Being located in high rainfall areas, soil erosion is further aggravated in such fragile terrain. The average soil loss was well below the tolerance level in the case of the slight and moderate phases of erosion. However, the magnitude of soil erosion was about 1.8 times and 2.7 times more in the severe and very severe erosion phases, respectively. Despite the higher erosion rate in the very severe eroded phase, the clay content is similar to that of the reference site. Higher clay content in sub-surface soils than in surface soil is the reason behind this. Because of the relatively larger drop size, tropical rains are characterized by high energy and intensity [6]; thus, a high water infiltration capacity is critical in reducing soil erosion risks.

Data in Table 5 show the range and mean values of the physical and chemical indicators of soil quality for different erosional phases within the same landscape environment. An increase in the severity of erosion (from slight to very severe) decreased the WHC, WSA, SOC, CEC, total N, extractable P, and available K, but increased the clay content and pH. Clay content varied from 17% to 26%, with mean values of 18.4% to 23.0% in the slight and severe erosion phases, respectively (Table 5). The significant difference in clay content among the erosional phases may be due to a steeper slope gradient, more accelerated erosion, and less plant cover in the moderate, severe, and very severe phases than in the reference and slight erosional phases. In addition, the transportation of lighter fractions is relatively less in the slight phase. Shaver et al. [72] related differences in the soil physical properties to the differences in landscape slope gradient and land use management. Low clay concentration in the severe phase (because coarse silt and sand fractions are deposited, while clay fractions are carried with runoff) has been reported by Lal [1]. Li and Lindstrom [73] also observed a linear decrease in clay content and a corresponding increase in silt content due to the selective removal of finer particles by water. Preferential removal may also explain the low clay content in steeply sloping lands [74].

The WHC is relatively lower in the severe (40–45%) and very severe (35–40%) than in the slight (44–52%) and moderate (42–50%) erosional phases (Table 5). The average WHC was 48.6%, 45.8%, 42.8%, and 36.6% in the slight, moderate, severe, and very severe phases, respectively, whereas the average WHC was 53% in the reference site. Therefore, a decrease in WHC with an increase in the severity of erosion may be due to a decline in SOC concentration, as was also reported by Bakker et al. [1] and Sun et al. [75]. Further, the relative proportion of the coarse fraction increases with an increase in the severity of erosion [48]. In the reference site, significantly higher values of SOC and WHC are observed, probably because of less movement of clay and organic matter due to the lower steepness of the slope, rapid recovery of the natural vegetation, less erosion, and slower oxidation of the new organic material [41,76,77].

The WSA ranged from 86% to 22.6% (Table 5), and the largest percentage of WSA (86%) was observed in the reference site. The SOC concentration ranged from 0.3% to 2.3%, and a low SOC concentration was observed in the severe (0.6%) and very severe (0.4%) phases. Thus, a higher WSA in the reference site may be due to more SOC concentration and better root biomass production, which act as binding agents between particles [78]. Similar findings were reported by Motavalli and McConnell [79] and Riezebos and Loerts [80]. A strong decrease in SOC concentration and WSA in the severe erosional phases indicates that the erosional process is the principal determinant of the changes in the surface soil properties in erosion-prone landscapes, affecting soil functions that support plant growth, water retention, and availability [81,82]. The decline in the soil physical quality of sloping lands in the tropics is a major issue. Whereas the plant nutrients can be replenished by applying fertilizers and manures, drastic adverse changes in the soil physical quality and water retention capacity cannot be quickly restored by routine soil management options [13]. Further, the SOC concentration strongly affects the WHC, WSA tilth, and soil biodiversity [83,84]. The rapid decline in the SOC concentration in cultivated soils of the tropics may be due to the continuously high temperatures, severe erosion, and depletion of plant nutrients. A severe loss of SOC concentration degrades the soil structure (WSA) and hydrological parameters and increases soil erodibility [7,15,35].

Soil pH ranged from 5.5 to 6.3 (Table 5) and differed significantly among the erosional phases. Soil pH increased with an increase in the severity of erosion. However, pH did not differ considerably among the slight and moderate erosional phases. A more significant change in pH was recorded with the further severity of erosion. Surprisingly, the soil pH of the reference site with minimum soil erosion was 6.2. This trend indicates that some pedological and edaphological processes are responsible for such a trend. Soil acidification may have occurred in the slight and moderate erosional phases. Erosion appears to expose the sub-surface material containing bases which increased the pH by 1 unit in the soil of the severe and very severe erosion phases [48]. An increase in soil pH with the severity of erosion was also reported by Cihacek and Swan [85], where soil erosion exposed the calciferous horizon.

The data in Table 5 show that the mean value of CEC followed this order: reference site (25 c mol⁽⁺⁾kg⁻¹) > slight eroded (15.2 c mol⁽⁺⁾kg⁻¹) > moderate (14.5 c mol⁽⁺⁾kg⁻¹) > severe (13.4 c mol⁽⁺⁾kg⁻¹) > very severe erosional phases (12.7 c mol⁽⁺⁾kg⁻¹). On average, significantly more CEC was observed in the reference site soil than in the slight, moderate, severe, and very severe erosional phases. These trends in CEC may be in accord with the SOC concentrations. Total N followed trends similar to those of CEC. Extractable P was lower in the reference site, and available K decreased with an increase in the severity of erosion. Average values of available K (kg ha⁻¹) were 221.1, 197.2, 171.2, and 151 in the slight, moderate, severe, and very severe erosional phases, respectively. Pimentel et al. [9] and Mandal et al. [15] also observed that soil erosion causes the loss of basic plant nutrients such as N, P, K, and Ca.

3.2. Soil Quality and Erosional Phases

Mean values of key soil properties (clay content, WHC, WSA, SOC, pH, CEC, total N, available P, and available K) were used to assess soil quality (Table 5). By using the combined fuzzy-AHP scoring technique, overall soil quality was compared among erosional phases (

Table 6. Soil quality index, sustainable yield index, and their relationship in various erosional phases.

Erosion Phase	SQI	SYI	Relationship
Uneroded (forested site)	0.9	-	-
Slight	0.7	0.9	SYI = 0.9 SQI + 0.2 (R2 = 0.86)
Moderate	0.6	0.7
Severe	0.5	0.7
Very severe	0.4	0.6

). The data show that SQI values ranged from 0.7 in the slight to 0.4 in the very severe erosional phase (Table 6), and it was better in the slight erosion phrase and had a higher weighted additive score than the other erosional phases. Compared to the erosional phases, the reference site had the highest SQI value of 0.9, as expected. In other words, the reference site had a better SQI than that for the erosional phases. The direction and degree of the change in soil quality depend on exogenous (climate and land use) and endogenous soil properties. In general, severe progressive soil erosion increases the magnitude of soil-related constraints to agronomic production [14]. Excessive erosion can drastically reduce soil quality, adversely affect agronomic yield [86,87,88,89], and severely jeopardize agricultural production. Soil degradation can render agriculture unsustainable [90], necessitating the conversion to restorative land use and adoption of best management practices (BPMs) [70].

3.3. Wheat Yield and Sustainable Yield Index

The grain yield of wheat was compared in different phases of erosion, and it decreased with an increase in the severity of erosion. Grain yield was significantly higher in slight erosion with a mean yield of 3.6 Mg ha⁻¹ than the other treatments (Figure 3). The lowest yield with a mean value of 2.6 Mg ha⁻¹ was recorded in the very severe eroded phase. Differences in crop yield among the erosion phases may be due to the differences in WHC and available nutrient content of the soil. Very severely eroded soil had significantly lower WHC, available P, and available K than the slight and moderate erosion phases [15].

The data on wheat yield for the different erosion phases (Figure 3 and Table 6) indicate that SYI was more slight (0.9) than the other erosional phases. Thus, the grain yield of wheat is more sustainable in the less eroded or uneroded than in severely eroded soils. Regression analysis showed that SYI is significantly (R² = 0.8) correlated with SQI and that an increase in the severity of erosion decreases SQI and SYI. Further, the increase in erosion severity similarly decreased both SQI and SYI. Although many researchers [91] did not find a significant relationship between SQI and crop yield, our study found a strong relationship between them. Our results are in agreement with Nakajima et al. (2016). Soil erosion under agricultural land use has also been viewed as an important driver impacting soil quality by Nosrati and Collins [92].

4. Conclusions

This study offers observational evidence from a series of erosional phases having different magnitudes of cumulative soil erosion. Based on the cumulative past soil erosion, the loss of topsoil was estimated at 5%, 13%, 29%, and 41%, respectively, in the slight, moderate, severe, and very severe erosion phases. The severe erosion phase had about three times more water loss and six times more soil loss than the slight erosion phase.

Our results showed that soils in the slight and moderate erosional phases were characterized by less variation in soil properties (e.g., WSA, pH, CEC, total N, available P, and available K). The severe erosional phase had significantly lower WHC, WSA, and SOC concentrations than those for the moderate or slight erosional phases. The severity of soil erosion also affected the pH and available nutrients. This study indicates that progressive erosion had a significant impact on SQI. SQI was higher in the reference site soil than in the slight, moderate, severe, and very severe erosional phases. Thus, accelerated soil erosion decreases the magnitude of one or more soil functions. Compared with the reference site, the SQI typically declined by 22.2% and 55.5%, respectively, in the slightly eroded and very severely eroded phases. Accelerated soil erosion had a strong impact on wheat productivity. Wheat productivity significantly reduced with the increase in erosion. However, the results obtained in the study need to be verified in a long-term experiment “with and without” nutrient application, as fertilizer has a masking effect on soil quality. A direct relationship was observed between SQI and SYI in different phases of erosion in the north-western Himalayan region. Further, this study provides a reference for the targeted soil conservation measures and land management techniques so that soil quality can be improved in the north-western Himalayan region of India.