Comparative Studies on Soil Quality Index Estimation of a Hilly-Zone Sub-Watershed in Karnataka

Abstract

The assessment of soil quality aims to evaluate the utility and health of soils. In agricultural studies, soil productivity can be likened to soil quality. Evaluating the Soil Quality Index (SQI) solely based on surface properties offers an incomplete picture because productivity is influenced by both surface and subsurface characteristics, with the latter associated with pedogenic processes. Additionally, relying on weighted averages of soil properties from a soil profile for the SQI may offer an overall summary, but it can occasionally obscure variations that manifest across different soil horizons. Therefore, the present study was conducted to assess the SQI in the Ganjigatti sub-watershed using data from 27 soil profiles and three different methods: (1) assessment of horizon-wise SQI by subjecting the soil properties of every horizon to principal component analysis (PCA), followed by the calculation of the weighted averages of the SQI for each soil profile (SQI-1); (2) calculation of the weighted averages of the soil properties for each soil profile, subjected to PCA, and followed by an SQI assessment (SQI-2); and (3) SQI assessment considering the properties of the Ap horizon for each soil profile (SQI-3). Additionally, to validate SQI methodologies, correlation studies were conducted against major crop yields in the sub-watershed. The results showed that cation exchange capacity (CEC) has the most significant weight and contribution to the SQI determined using MDS, followed by porosity, exchangeable sodium percentage (ESP), organic carbon (OC), CN ratio, and total N. SQI-1 was most strongly correlated with crop yield; the correlation coefficient ranged from 0.69 to 0.74. Among all the three methodologies, SQI-1 and -2 were better methods for assessment of SQI compared to SQI-3. In the SQI-1 method, the soil quality of pedons ranged from 0.26 (pedon-26) to 0.74 (pedon-11). The majority of the area in the sub-watershed (72.40%) fell within the medium category of SQI (0.35–0.55), followed by the high category of SQI (>0.55), which comprised 12.92%, and the low SQI (<0.35), which comprised 6.45% of the sub-watershed.

1. Introduction

Soil is an important non-renewable natural resource on which humans, plants, and animals rely for survival. Research is being conducted globally to identify the best strategies for protecting soils and using them to increase agricultural productivity while preserving environmental quality through improved management methods. On a global scale, available land resources are decreasing at an alarming rate, largely due to a high population pressure [1]. Approximately 10 hectares of land are lost every minute due to various degradation processes such as erosion, nutrient depletion, salinity, acidity, alkalinity, and compaction [2]. The United Nations Environmental Programme (UNEP) has estimated that during the second half of the 20th century, around 2 billion hectares of farmland had suffered degradation [3]. India has a total geographical area of about 328.8 Mha, of which 180 Mha is agricultural land with different types of soils. It supports 17.5% of the world’s population with only 2.4% of the world’s geographical area and 9% arable land. The demand for food, fuel, and energy has increased many folds, and the growing population needs to be fed with shrinking and deteriorating land and water resources [4,5]. Nearly 120 million hectares of arable land in India is reported to be degraded [6].

The detreating of soil quality is one of the main causes of agricultural productivity stagnation and a serious threat to food supply and environment security. An action plan is needed to minimise natural resource degradation and improve soil quality. It should adhere to sustainability principles to ensure that the soil can be passed on to the next generation in improved conditions compared to what was inherited from the previous generation. Therefore, management approaches must align with ecological integrity, economic viability, and social and political approval. At this time, watershed management is an accepted strategy for the development of rainfed agriculture. This approach is multidisciplinary, broad, and intensive [7,8]. A watershed is a geographical and hydrological region that drains to a common point, and it is regarded as a suitable physical entity for assessing, planning, and managing natural resources. Since the 1980s, watershed-focused development has been the primary approach in India’s rainfed regions, aiming to preserve natural resources, boost agricultural output, and enhance rural livelihoods. In this context, assessing the soil quality of soils in watersheds is crucial to managing soil resources for maximum effectiveness in the here-and-now without compromising their viability for the future [9].

The assessment of soil quality is a sensitive and dynamic method for documenting the status of the soil, as well as the soil’s response to management and its resilience to stress, whether that stress is imposed by natural forces or by human interventions. Soil quality is defined as “the capacity of a specific kind of soil to function, within natural or managed ecosystem boundaries, to sustain plant and animal productivity, maintain or enhance water and air quality, and support human health and habitation” [10]. Soil quality, in the context of agricultural output, is defined as the extent to which it can maintain productivity [11]. In essence, soil quality pertains to the functionality of the soil [12,13], whereas soil health portrays the soil as a dynamic, non-renewable, and finite living resource [14,15]. In order to evaluate the status of soil degradation and the shifting patterns that resulted from various land uses and smallholder management interventions, it is required to conduct a fundamental assessment of soil quality [16]. This is the main cause of Africa’s inability to produce as much food as needed to meet demand, and the continent’s per-capita food production is dropping [17,18], mostly as a result of a decline in the quality of its soil.

Soil quality is such a complicated and multifaceted functional concept. It is not possible to directly measure it in the field or in the laboratory; rather, it can only be inferred from the characteristics of the soil. In order to estimate soil quality, a variety of soil parameters or indicators have been identified. There is currently no precise standard or set of criteria that is commonly acknowledged for use in evaluating soil quality [19,20]. Considering the diverse and intricate nature of soil characteristics, numerous assessment approaches have been created, including soil quality models [21], methods for establishing soil quality indices (SQI) [22,23,24], soil health cards and testing kits [25], fuzzy association rules [26], and the soil multifunctionality index [27]. One of these approaches, the soil quality index (SQI), is typically employed for the assessment of soil quality [28] due to its flexibility and ease of use in quantifying changes in various types of soil.

Soil quality indicators capture variability resulting from land management practices, encompassing a range of chemical, physical, and biological soil characteristics. To evaluate the effects of various management practices, it is crucial to establish a baseline or reference values for these soil quality indicators [29]. In one study [24], multiple physical indicators such as potential available water capacity (AWC), soil penetration resistance, bulk density (BD), mean weight diameter (MWD), aggregate size distributions, the fraction of water-stable aggregates (WSA), and geometric mean diameter (GMD) were employed, alongside other chemical indicators, to assess soil quality. Another study [30] evaluated soil quality by considering various biological and physico-chemical soil quality indicators to examine the sustainability of different management and land-use systems. The study identified soil pH, porosity, cation exchange capacity (CEC), available phosphorus (P), BD, total organic carbon (TOC), earthworm population, and plant available water holding capacity (PAWC) as the most responsive indicators. However, total nitrogen (N), exchangeable potassium (K), total phosphorus (P), and K showed moderate sensitivity, while the percentage base saturation was a weaker indicator. In a separate investigation [31], various soil organic carbon (SOC) fractions and the activities of various soil enzymes like dehydrogenase, phosphatase, aryl sulfatase, and fluorescein diacetate hydrolases (FDAse) were employed as biological indicators to assess soil quality in diverse cropping systems in the northwestern Himalayas. Additionally, ref. [32] utilised macroporosity, microporosity, soil hydraulic conductivity (sHC), moisture saturation (MS), effective saturation, aggregate size distribution, aggregate stability index (ASI), exchangeable calcium (Ca) and magnesium (Mg), exchangeable acidity, potential acidity, aluminium saturation, basal respiration, carbon stock (C stock), and nitrogen stock (N stock) as potential soil quality indicators. Further, ref. [33] conducted a study to assess soil quality in the Lakkampura mini-watershed (Karnataka) based on the interaction of soil carbon stocks with other soil parameters. Similarly, ref. [34] assessed the soil quality of the sub-humid southern plains of Rajasthan by using fertility characters. In the Garhwal Himalayas of Uttarakhand, ref. [35] demonstrated the impact of landslides on soil quality. They utilised Principal Component Analysis (PCA) to determine that in resource-constrained situations, soil organic carbon (SOC), available phosphorus (P), and clay content should be selected as the key indicators for tracking fluctuations in soil quality.

Soil quality can be conceptualised in two aspects, viz., inherent and dynamic soil quality. The inherent soil quality shows little change over time whereas dynamic soil quality changes with respect to soil management. The changes in soil properties may occur within hours to a period of decades with respect to the response level of soil properties. However, the limits to which dynamic soil properties can change are dictated by inherent properties [20]. Previously, many evaluations of soil quality focused on the characteristics of the surface soil [24,31,33,34], and there has been limited research utilising information from soil profiles [36,37,38]. Based on the information provided above, a hypothesis was formulated suggesting that analysing the soil quality index (SQI) across different horizons within a profile yields more comprehensive and accurate insights compared to relying on a weighted average of properties for the SQI. While weighted averages offer a broad overview, they can sometimes obscure variations present in distinct horizons. The assessment of SQI horizon-wise enables the examination of soil quality at each individual horizon, facilitating a better grasp of soil quality nuances. The current study was executed with the objectives: (a) identifying soil quality indicators; (b) evaluating the SQI using horizon-wise soil properties, weighted averages of soil properties, and properties of the Ap horizon in each soil profile; and (c) establishing correlations between the SQI and crop yield.

2. Materials and Methods

2.1. Description of the Study Area

The study was conducted in the Ganjigatti sub-watershed, located 37 km away from Dharwad town, Karnataka, during 2021–2022. It is located between 15°10′10.114″ and 15°17′1.147″ N latitudes and 75°0′57.672″ and 75°4′50.525″ E longitudes. This area lies in a hilly zone (zone-9) in Karnataka and a hot semi-arid ecological sub region. It is characterized by ustic soil moisture and isohyperthermic soil temperature regimes (Figure 1) [39]. The study area covers 4323.84 ha and receives an annual average rainfall of 917 mm (average annual rainfall of the zone ranges from 539 to 1256 mm). Annual average kharif (June to September), rabi (October to January), and summer (February to May) rainfall are 616 mm, 139 mm and 162 mm, respectively. The study area’s soils originated from chlorite schist with shale being the predominant parent material. These soils exhibit a coarse texture and are relatively shallow on uplands, and in the lower elements of landscape, they become finer in texture and deeper. The predominant soil types in the area are black and red soils. Black soils are well-suited for cultivating crops such as cotton, wheat, finger millets, sorghum, and oilseeds. Among these, maize and soybean are the primary crops grown during the kharif season, while sorghum and greengram are the main crops cultivated during the rabi season. The length of the growing period (LGP) in this region typically spans from 120 to 150 days.

2.2. Soil Sampling and Analysis

A detailed soil survey of the study area was performed using a high-resolution satellite image (Worldview-2, 50 cm SR) and cadastral maps. The image and scanned topographic sheet were geocoded and subdivided using ArcGIS 10.8.2, resulting in subsets at a scale of 1:7920. During the comprehensive soil survey, landforms and physiographic divisions were identified based on geological characteristics, drainage patterns, surface features, slope attributes, and land usage. Once the landforms were delineated on the satellite image, we conducted thorough traversing of each landform to select representative areas for transect studies. These transects were positioned roughly perpendicular to the contours and aimed to include the variations within each landform. Within each selected transect, profiles were established at closely spaced intervals to account for changes in land features, such as alterations in slope, erosion, presence of gravels, stones, or variations in soil depth. A total of twenty-seven (27) soil profiles were examined in this manner. For each pedon, detailed morphological features were described as per a Soil Survey Manual [39] such as colour, texture, structure, consistency, etc.

Soil samplings were air-dried, ground, sieved (<2 mm), and analysed for soil physical, physico-chemical and chemical properties. Particle size analysis of soil samples was carried out using the international pipette method [40]. Bulk density of soil samples was determined using the clod method [41]. The particle density was assessed using the specific gravity bottle method as outlined in reference [42]. The total porosity was subsequently computed based on the bulk density and particle density. The maximum water-holding capacity and volume expansion were determined using Keen Raczkowski’s method [43]. Moisture retention was determined at −33 kPa and −1500 kPa using pressure plate apparatus and pressure membrane, respectively and available soil moisture content was determined from the difference in moisture held at −33 kPa and −1500 kPa [44].

The pH and electrical conductivity (EC) were measured employing 1:2.5 soil:water ratio [45]. Organic carbon (OC) was determined using the method of [46]. CaCO₃ equivalent (%) was determined using the method described by [40]. Soil inorganic carbon was assessed using the formula in [40]. Exchangeable cations were assessed through a soil extraction process involving 1 N neutral ammonium acetate. The concentrations of calcium (Ca) and magnesium (Mg) in the extract were determined through Versenate titration, while sodium (Na) and potassium (K) concentrations were determined using flame photometry. The quantities of exchangeable cations, specifically Ca²⁺, Mg²⁺, K⁺, and Na⁺, were expressed in units of cmol (p⁺) per kilogram of soil, as detailed in [47]. Additionally, the cation exchange capacity (CEC) of the soils was determined utilising a method involving 1 N sodium acetate at pH 8.2, as outlined in [48]. Base saturation was calculated by summing the bases extracted by dividing by CEC (by 1 N sodium acetate) and multiplying by 100. Exchangeable acidity (Al⁺³ + H⁺) of soil samples was determined using the KCl extractant method, as described by [40]. Exchangeable sodium percentage (ESP) was estimated as the ratio of exchangeable sodium to CEC. Available nitrogen was determined using the modified alkaline permanganate method as described by [49]. The total nitrogen (%) was determined using the Macrokjeldahl digestion cum distillation method as described by [45]. C:N ratio was estimated as the ratio of total carbon (%) to total nitrogen (%). Descriptive statistics for 34 parameters were performed with SPSS ver. 26.

2.3. Assessment of Soil Quality Index

A statistical model based on principal component analysis (PCA) was employed to estimate Soil Quality Index (SQI), as described in [10,22,28,50,51,52,53]. The SQI calculation involved four key steps: (a) defining the management objective, (b) selecting specific indicators as Minimum Data Set (MDS), (c) assigning scores to the chosen indicators, and (d) computing the SQI.

2.3.1. Indicator Selection

The PCA model was applied to establish a Minimum Data Set (MDS) with the aim of reducing the burden of indicators in the model and preventing data duplication [22]. The PCA method is preferred due to its objectivity, as it incorporates various statistical tools such as multiple correlations, factor analysis, and cluster analysis. These tools help in mitigating any potential bias and data redundancy by employing mathematical formulas to select an MDS, as elaborated in references [11,28]. The SQI of the sub-watershed was assessed using three methodologies: (1) horizon-wise SQI assessment followed by weighted averages of the SQI of each soil profile; (2) weighted averages of the soil properties of each soil profile subjected to PCA and followed by SQI assessment; and (3) considering the soil properties of the Ap horizon for SQI assessment. The primary purpose of PCA is to reduce the complexity of a comprehensive dataset containing numerous interconnected variables while preserving as much of the inherent variability as possible. This is accomplished by transforming the data into a fresh set of variables known as principal components (PCs), which are both uncorrelated and arranged in a manner that ensures the initial ones capture the majority of the variation observed across all the original variables [54,55]. In essence, the PCA method was selected as a means of streamlining the data and identifying the most suitable indicator(s) for representing and estimating SQI [8,55].

Principal components (PCs) in a dataset are combinations of variables that capture the maximum variation within the dataset. They describe vectors that closely fit the n observations in a p-dimensional space while being mutually orthogonal. In this study, the approach outlined by [22] was followed. In previous studies, it has been common to assume that PCs with high eigenvalues and variables with high factor loadings best represent system attributes. Consequently, only PCs with eigenvalues greater than 1 were chosen. For each PC, each variable had a corresponding eigenvector weight value or factor loading. Only the variables with the highest weights were retained for inclusion in the Minimum Data Set (MDS). Variables were considered “highly weighted” if they had the highest weight under a specific PC and an absolute factor loading value within 10% of the highest values under the same PC [56]. However, when multiple variables were retained under a specific PC, a multivariate correlation matrix was used to calculate correlation coefficients between these parameters, as detailed in references [22,28]. If the parameters were significantly correlated (with a correlation coefficient r > 0.60 and p < 0.05), then the one with the highest loading factor was included in the MDS, while all others were excluded to prevent redundancy. Parameters that were not correlated under a particular PC were considered important and retained in the MDS, as explained by [28,57].

2.3.2. Scoring of Indicators

The indicators selected for the Minimum Data Set (MDS) were transformed into dimensionless values ranging from 0 to 1 using a linear scoring method [58]. The ranking of indicators was determined by whether a higher value was deemed favourable or unfavourable concerning soil function. For indicators where a higher value indicated better performance, each indicator value was divided by the highest value, ensuring that the highest value received a score of 1. Conversely, for indicators where lower values were preferable, the lowest value was divided by each data value so that the lowest value received a score of 1. In the case of indicators like pH, an “optimum” threshold value was considered. These indicators were initially scored as “higher is better” up to the threshold value (for example, pH 7.5), and beyond the threshold, they were scored as “lower is better”. This scoring method is consistent with the approach outlined in references [22,28].

2.3.3. Soil Quality Index Calculation

Following that, the chosen observations were converted into numerical scores within the range of 0 to 1. These scores were then combined into indices for each soil sample through a weighted additive method. To determine the specific weighting value assigned to each Principal Component (PC), the variance explained by each PC was divided by the maximum total variation among all PCs generated by the PCA.

SQI = Σ Principal Component Weight ∗ Individual soil parameter score

Principal Component Analysis (PCA) and descriptive statistics were carried out using SPSS version 26 and Microsoft Excel version 16 to compute the Soil Quality (SQ) indices. Crop yield information spanning an eight-year period (2014–2021) was sourced from the Department of Agriculture, Government of Karnataka, Dharwad. The average yield data for soybean, maize, sorghum, and greengram were collected on soil phases basis and then correlated with the Soil Quality Index (SQI).

3. Results

3.1. Physical, Physico-Chemical, and Chemical Properties

Descriptive statistics for the 34 soil properties used for soil quality assessment are given in

Table 1. Descriptive statistics of horizon-wise soil properties used for soil quality assessment.

	Minimum	Maximum	Mean	SD	CV	Skewness
Coarse sand (%)	1.68	44.90	15.47	8.25	53.34	1.000
Fine sand (%)	3.25	35.90	21.62	7.28	33.65	−0.180
Sand (%)	4.93	73.25	37.08	14.18	38.23	0.099
silt (%)	4.20	32.36	17.04	6.22	36.48	0.359
clay (%)	18.35	76.45	45.88	11.86	25.85	−0.184
B.D (Mg m−3)	1.20	1.65	1.38	0.09	5.93	0.362
P.D (Mg m−3)	2.32	2.57	2.44	0.06	2.33	0.169
Porosity (%)	30.67	49.79	43.12	3.42	7.93	−0.868
MWHC (%)	30.47	78.80	50.58	9.44	18.65	0.083
VE (%)	0.38	39.46	9.57	7.43	77.59	1.468
FC (%)	11.46	45.88	27.99	7.47	26.67	−0.056
PWP (%)	6.43	29.10	16.49	5.35	32.40	0.340
AWC (%)	3.64	19.21	11.50	3.09	26.80	0.172
pH	5.15	9.67	7.08	1.12	15.79	0.023
EC (dS m−1)	0.08	0.92	0.25	0.15	58.28	2.122
OC (%)	0.23	1.34	0.62	0.25	39.14	1.029
Ex. Ca [cmol (p+) kg−1]	2.16	40.32	15.51	9.36	60.33	0.681
Ex. Mg [cmol (p+) kg−1]	1.08	19.32	8.43	5.18	61.44	0.663
Ex. Na [cmol (p+) kg−1]	0.13	8.06	0.70	1.15	163.04	4.985
Ex. K [cmol (p+) kg−1]	0.09	1.59	0.32	0.20	60.82	3.674
Sum of Ex. Basic cations	3.89	64.66	24.96	14.88	59.60	0.724
Ex. Al+3 [cmol (p+) kg−1]	0.00	0.39	0.06	0.11	172.29	1.604
Ex. H+ [cmol (p+) kg−1]	0.10	1.78	0.54	0.26	47.54	2.344
Exchangeable Acidity	0.10	2.02	0.60	0.33	55.03	1.832
CEC [cmol (p+) kg−1]	5.20	78.71	31.54	15.23	48.29	0.732
BS (%)	36.93	98.04	75.45	15.34	20.34	−0.585
ESP (%)	0.44	12.77	2.09	1.87	89.11	3.216
CEC/clay ratio	0.28	1.12	0.67	0.24	34.49	0.085
CaCO3 (%)	0.40	20.50	6.73	6.01	89.29	1.018
IC (%)	0.05	2.46	0.81	0.73	89.29	1.018
TC (%)	0.46	2.75	1.43	0.63	43.67	0.774
Available N (kg/ha)	70.00	217.00	137.37	36.54	26.6	0.328
Total N (%)	0.04	0.12	0.08	0.03	26.62	0.363
C:N ratio	3.75	12.84	7.96	1.91	23.94	0.176

,

Table 2. Descriptive statistics of weighted average soil properties used for soil quality assessment.

	Minimum	Maximum	Mean	SD	CV	Skewness
Coarse sand (%)	4.65	43.58	16.72	8.44	50.45	1.302
Fine sand (%)	9.43	34.39	22.86	6.59	28.81	−0.228
Sand (%)	14.22	70.22	39.58	13.58	34.30	0.150
silt (%)	5.99	28.07	15.41	5.40	35.01	0.545
clay (%)	21.23	60.29	45.02	11.00	24.43	−0.492
B.D (Mg m−3)	1.25	1.58	1.39	0.08	5.22	0.591
P.D (Mg m−3)	2.34	2.54	2.44	0.06	2.36	0.092
Porosity (%)	33.57	46.91	42.84	2.92	6.82	−1.310
MWHC (%)	30.85	69.78	48.95	9.26	18.92	−0.019
VE (%)	0.45	26.24	8.95	6.43	71.86	0.911
FC (%)	12.68	40.10	27.04	7.15	26.44	−0.182
PWP (%)	7.64	23.65	15.81	4.68	29.59	0.171
AWC (%)	4.10	16.51	11.23	2.92	25.97	−0.290
pH	5.36	9.10	6.90	1.02	14.76	0.376
EC (dS m−1)	0.06	0.49	0.23	0.12	49.24	0.796
OC (%)	0.33	1.02	0.60	0.15	24.79	0.890
Ex. Ca [cmol (p+) kg−1]	2.66	33.82	13.58	8.73	64.29	0.966
Ex. Mg [cmol (p+) kg−1]	1.21	17.86	7.45	4.82	64.73	0.981
Ex. Na [cmol (p+) kg−1]	0.16	4.56	0.60	0.82	136.74	4.719
Ex. K [cmol (p+) kg−1]	0.13	0.67	0.30	0.13	42.97	1.489
Sum of Ex. Basic cations	4.25	52.12	21.92	13.86	63.24	1.000
Ex. Al+3 [cmol (p+) kg−1]	0.00	0.33	0.08	0.11	135.12	1.117
Ex. H+ [cmol (p+) kg−1]	0.21	1.57	0.56	0.26	45.10	2.486
Exchangeable Acidity	0.21	1.77	0.64	0.33	51.44	1.756
CEC [cmol (p+) kg−1]	6.18	59.33	28.89	14.35	49.66	0.882
BS (%)	40.22	92.99	72.09	15.25	21.16	−0.367
ESP (%)	0.53	7.36	2.00	1.36	68.11	2.698
CEC/clay ratio	0.29	1.00	0.63	0.22	33.84	0.398
CaCO3 (%)	0.45	17.10	5.55	5.39	97.21	1.452
IC (%)	0.05	2.05	0.67	0.65	97.21	1.452
TC (%)	0.62	2.48	1.28	0.58	45.30	1.215
Available N (kg/ha)	99.75	197.53	138.31	31.78	22.98	0.543
Total N (%)	0.06	0.11	0.08	0.02	23.11	0.554
C:N ratio	4.98	10.94	7.76	1.59	20.41	0.063

and

Table 3. Descriptive statistics of Ap horizon soil properties used for soil quality assessment.

	Minimum	Maximum	Mean	SD	CV	Skewness
Coarse sand (%)	6.70	40.27	20.01	7.66	38.27	0.504
Fine sand (%)	11.45	32.88	23.51	6.40	27.20	−0.332
Sand (%)	18.15	62.65	43.52	12.35	28.37	−0.492
silt (%)	5.50	29.88	16.36	6.28	38.40	0.656
clay (%)	21.87	56.65	40.14	9.95	24.77	0.066
B.D (Mg m−3)	1.21	1.52	1.35	0.08	5.75	0.474
P.D (Mg m−3)	2.32	2.54	2.42	0.05	2.06	0.896
Porosity (%)	37.39	49.79	44.30	3.16	7.12	−0.351
MWHC (%)	31.29	59.84	45.90	7.41	16.14	−0.155
VE (%)	0.47	10.95	4.98	3.65	73.23	0.558
FC (%)	13.90	31.93	23.44	5.49	23.40	−0.128
PWP (%)	8.14	20.47	13.28	3.57	26.86	0.465
AWC (%)	4.62	14.12	10.17	2.43	23.85	−0.56
pH	5.15	8.38	6.47	0.93	14.24	0.374
EC (dS m−1)	0.08	0.64	0.23	0.12	53.03	1.834
OC (%)	0.47	1.34	0.81	0.25	29.99	0.616
Ex. Ca [cmol (p+) kg−1]	3.24	28.48	11.38	6.93	60.90	1.013
Ex. Mg [cmol (p+) kg−1]	1.08	17.88	6.34	4.12	64.90	1.351
Ex. Na [cmol (p+) kg−1]	0.13	0.93	0.38	0.22	58.08	0.982
Ex. K [cmol (p+) kg−1]	0.14	0.65	0.34	0.15	42.94	0.705
Sum of Ex. Basic cations	4.66	41.84	18.43	10.86	58.91	1.032
Ex. Al+3 [cmol (p+) kg−1]	0.00	0.32	0.09	0.12	124.73	0.917
Ex. H+ [cmol (p+) kg−1]	0.28	1.10	0.62	0.19	30.53	0.671
Exchangeable Acidity	0.28	1.24	0.71	0.27	37.29	0.364
CEC [cmol (p+) kg−1]	7.30	58.69	25.33	11.80	46.58	1.18
BS (%)	36.93	95.73	70.08	15.91	22.70	−0.308
ESP (%)	0.49	5.44	1.68	1.14	67.91	1.892
CEC/clay ratio	0.30	1.12	0.63	0.23	35.22	0.378
CaCO3 (%)	0.50	12.00	3.71	3.06	82.62	1.571
IC (%)	0.06	1.44	0.45	0.37	82.62	1.571
TC (%)	1.42	0.62	2.04	0.37	29.10	0.338
Available N (kg/ha)	119.00	217.00	165.80	29.31	17.68	−0.042
Total N (%)	0.07	0.12	0.10	0.02	17.99	−0.057
C:N ratio	4.96	12.26	8.6	1.91	22.21	−0.028

. Soil depth among the pedons varied from 20 to 200+. The soil texture of the Ganjigatti sub-watershed is mostly limited to clay, sandy clay, and sandy clay loam, with few exceptions. Clay illuviation is an important soil-forming process. The movement of clay has an important pedological significance as clay bears a good relationship with other properties. Coefficient of variation (CV) is used to interpret soil variability [59].

3.2. Assessment of Soil Quality Index of Ganjigatti Sub-Watershed

3.2.1. Selection of Soil Quality Indicators

To determine the minimum data set (MDS) and key factors affecting soil quality index, PCA was carried out on the total data set. The PCA method is a common and accepted approach for determining MDS and has been widely used in previous studies [22,60,61,62]. The SQI of this sub-watershed was assessed using three methodologies: (1) assessment of horizon-wise SQI by subjecting the soil properties of every horizon to PCA followed by calculating the weighted averages of the SQI of each soil profile (SQI-1); (2) weighted averages of the soil properties of each soil profile subjected to PCA and followed by SQI assessment (SQI-2); and (3) considering the Ap horizon properties of each soil profile for the SQI assessment (SQI-3).

In the SQI-1 method, the soil properties of all 27 profiles horizon-wise were subjected to PCA to reduce the data dimension. The PCA data for the sub-watershed showed that all the eigenvalues were >1, which explained 85.658% of the cumulative variance in the data. The MDS were chosen based on the highly weighted factor loading of variables. In total, eight PCs were extracted for the sub-watershed with eigenvalues >1 (

Table 4. Principal components of soil quality parameters, eigenvalues, and component matrix variables for soil profiles (SQI-1).

Principal Components	PC1	PC2	PC3	PC4	PC5	PC6	PC7	PC8
Eigenvalue	16.401	2.748	2.347	1.992	1.610	1.559	1.400	1.068
% variance	48.237	8.081	6.903	5.860	4.735	4.584	4.118	3.140
% cumulative variance	48.237	56.318	63.221	69.081	73.816	78.400	82.518	85.658
Weightage factor	0.5631	0.0943	0.0806	0.0684	0.0553	0.0535	0.0481	0.0367
Factor loadings (Rotated component matrix)
Coarse sand	−0.830	−0.296	−0.109	0.258	0.132	0.105	−0.051	0.014
Fine sand	−0.669	−0.374	−0.312	0.178	0.168	−0.203	0.141	0.195
Sand	−0.826	−0.364	−0.224	0.241	0.163	−0.043	0.043	0.108
Silt	0.534	0.155	0.040	0.218	−0.238	0.403	−0.504	0.136
Clay	0.708	0.354	0.246	−0.403	−0.071	−0.160	0.213	−0.201
BD	0.228	−0.622	0.247	−0.468	0.270	−0.194	−0.290	0.011
PD	0.349	0.078	0.325	0.142	−0.063	0.038	−0.315	−0.104
Porosity	−0.057	0.660	−0.156	0.504	−0.334	0.241	0.202	−0.065
MWHC	0.894	0.114	0.164	−0.070	−0.059	−0.053	0.138	0.009
VE	0.832	−0.201	0.244	0.012	−0.071	−0.032	0.192	0.159
FC	0.898	0.123	−0.064	−0.144	0.045	−0.124	0.160	−0.004
PWP	0.873	0.088	−0.058	−0.053	−0.046	0.004	0.089	−0.030
AWC	0.662	0.145	−0.054	−0.258	0.188	−0.307	0.233	0.042
pH	0.884	−0.173	−0.148	0.182	−0.058	−0.105	−0.012	−0.074
EC	0.529	−0.348	0.348	0.306	0.157	0.086	0.119	0.212
OC	−0.554	0.392	0.433	0.217	0.310	−0.323	−0.236	0.044
Ex. Ca	0.924	0.099	−0.102	0.005	0.282	0.093	−0.043	0.035
Ex. Mg	0.904	0.149	−0.041	−0.021	0.234	0.003	−0.024	−0.039
Ex. Na	0.563	−0.261	0.544	0.388	0.025	0.049	0.328	−0.068
Ex. K	0.363	0.380	−0.057	0.200	0.220	0.041	−0.147	−0.189
Sum of Ex.cations	0.944	0.099	−0.037	0.028	0.264	0.063	−0.012	0.001
Ex. Al+3	−0.677	0.033	0.369	−0.130	0.134	0.344	0.051	0.162
Ex. H+	−0.608	0.101	0.242	−0.346	0.219	0.478	0.099	−0.103
Exchangeable Acidity	−0.689	0.089	0.306	−0.310	0.212	0.481	0.093	−0.028
CEC	0.892	0.141	−0.031	−0.004	0.328	0.104	0.067	−0.11
BS	0.749	0.051	−0.104	0.077	−0.151	−0.244	−0.218	0.262
ESP	0.298	−0.353	0.583	0.419	−0.171	−0.004	0.36	−0.052
CEC/clay ratio	0.705	−0.001	−0.285	0.269	0.419	0.236	−0.054	0.053
CaCO3	0.880	−0.152	−0.076	0.054	0.157	0.224	−0.103	0.197
Inorganic carbon	0.880	−0.152	−0.076	0.054	0.157	0.224	−0.103	0.197
Total carbon	0.228	0.248	0.352	−0.284	−0.363	−0.004	−0.113	0.632
Available N	−0.655	0.492	0.077	0.138	0.337	−0.190	0.145	0.270
Total N	−0.660	0.487	0.083	0.144	0.329	−0.200	0.144	0.266
C:N ratio	−0.176	0.054	0.569	0.242	0.112	−0.293	−0.475	−0.275

Bold-face factor loadings were considered highly weighted, and the underlined were retained in MDS.

). A representative scree plot showing the variation of eigenvalues with soil components is shown in Figure 2. The parameters in each PC were considered based on higher values of the factor loading. The soil parameters obtained from PCA under PC1 were the sum of exchangeable basic cations, exchangeable Ca, exchangeable Mg, pH, CEC, field capacity, volume expansion, permanent wilting point, CaCO₃, and inorganic carbon. However, a multivariate correlation matrix was utilised to calculate the correlation coefficients between the parameters when more than one variable was retained under a given PC [22,28]. To avoid redundancy, only the parameter with the highest loading factor was kept in the MDS if there was a significant correlation between them (r > 0.60, p 0.05). The non-correlated parameters under a particular PC were considered important and retained in the MDS [22,57]. Among these highly weighted variables, CEC is a parameter that governs nutrients’ holding capacity and is an indicator of soil fertility. It is a very important soil property, influencing the soil’s structural stability, nutrient availability, and pH. Other parameters are highly correlated to each other, so CEC was retained for MDS in PC1 (

Table 5. Correlation coefficient (Pearson’s) for highly loaded parameters in PC1 (SQI-1).

	MWHC	FC	PWP	pH	Ex.Ca	Ex.Mg	SEBC	CEC	CaCO3	IC
MWHC	1
FC	0.84 **	1
PWP	0.80 **	0.94 **	1
pH	0.75 **	0.74 **	0.73 **	1
Ex. Ca	0.78 **	0.82 **	0.79 **	0.77 **	1
Ex. Mg	0.82 **	0.80 **	0.76 **	0.77 **	0.91 **	1
SEBC	0.82 **	0.83 **	0.80 **	0.80 **	0.98 **	0.96 **	1
CEC	0.79 **	0.80 **	0.75 **	0.72 **	0.95 **	0.93 **	0.97 **	1
CaCO3	0.73 **	0.74 **	0.73 **	0.79 **	0.87 **	0.80 **	0.87 **	0.78 **	1
IC	0.73 **	0.74 **	0.73 **	0.79 **	0.87 **	0.80 **	0.87 **	0.78 **	1.00 **	1

** Correlation is significant at the 0.01 level (2-tailed).

). In PC2, porosity was retained under MDS between bulk density and porosity because of its higher factor loading and better correlation (

Table 6. Correlation coefficient (Pearson’s) for highly loaded parameters in PC 2 (SQI-1).

	BD	Porosity
BD	1
Porosity	−0.90 **	1

** Correlation is significant at the 0.01 level (2-tailed).

). From PC3, exchangeable Na, ESP, and CN ratio were selected, and exchangeable Na and CN ratio were considered for MDS based on correlation (

Table 7. Correlation coefficient (Pearson’s) for highly loaded parameters in PC 3 (SQI-1).

	Ex. Na	ESP	CN
Ex. Na	1
ESP	0.87 **	1
CN	0.11	0.15	1

** Correlation is significant at the 0.01 level (2-tailed).

). Bulk density, CEC/clay ratio, exchangeable acidity, silt, and total carbon were selected as indicators from PC4, PC5, PC6, PC7, and PC8, respectively. Some of the selected parameters were not independent of each other. Exchangeable acidity is correlated with porosity, CN ratio, ESP, BD, and silt (

Table 8. Correlation coefficient (Pearson’s) for selected MDS parameters from PCA results (SQI-1).

	CEC	Porosity	ESP	CN	BD	CEC/Clay	Ex. Acidity	Silt	TC
CEC	1
Porosity	−0.05	1
ESP	0.15	−0.26 *	1
CN	−0.17	0.23 *	0.03	1
BD	0.11	−0.12	−0.04	0.22 *	1
CEC/clay	0.82 **	0.15	0.00	−0.12	0.08	1
Ex. Acidity	−0.00	−0.79 **	0.27 *	−0.33 **	−0.46 **	−0.18	1
Silt	0.36 **	0.31 **	0.03	0.13	0.05	0.48 **	−0.28 **	1
TC	0.06	0.02	0.03	−0.03	0.06	−0.09	−0.06	0.28 *	1

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

). CEC/clay ratio is a derived parameter that is dependent on the CEC of the soil [63].

In the SQI-2 method, the weighted average of the soil properties of all the 27 profiles subjected to PCA was used to reduce the data dimension. The PCA data for the sub-watershed show the eight PCs with eigenvalues > 1, which explain 90.128% of the cumulative variance in the data (

Table 9. Principal components of soil quality parameters, eigenvalues, and component matrix variables for soil profiles (SQI-2).

Principal Components	PC 1	PC 2	PC 3	PC 4	PC 5	PC 6	PC 7	PC 8
Eigenvalue	17.304	2.897	2.343	2.135	1.893	1.561	1.388	1.152
% variance	50.895	8.521	6.891	6.280	5.568	4.591	4.083	3.389
% cumulative variance	50.895	59.416	66.307	72.586	78.154	82.745	86.828	90.218
Weightage factor	0.5641	0.0944	0.0764	0.0696	0.0617	0.0509	0.0453	0.0376
Factor loadings (Rotated component matrix)
Coarse sand	−0.849	−0.224	−0.203	−0.114	−0.046	0.214	−0.056	0.068
Fine sand	−0.694	−0.067	−0.449	−0.125	−0.328	0.040	0.132	0.034
Sand	−0.861	−0.173	−0.338	−0.131	−0.181	0.154	0.026	0.059
Silt	0.640	−0.338	0.047	−0.106	0.400	−0.129	−0.311	0.207
Clay	0.742	0.381	0.393	0.214	0.024	−0.126	0.123	−0.176
BD	0.208	−0.512	0.057	0.663	−0.390	0.106	0.055	0.241
PD	0.373	0.021	0.372	−0.009	−0.020	0.180	−0.496	0.452
Porosity	−0.034	0.529	0.136	−0.668	0.388	−0.062	−0.268	0.056
MWHC	0.865	0.168	0.228	−0.030	0.047	0.002	0.221	−0.088
VE	0.851	−0.071	0.160	−0.089	0.022	0.065	0.272	0.138
FC	0.809	0.269	−0.014	0.054	−0.080	−0.106	0.070	−0.118
PWP	0.803	0.156	0.005	−0.068	0.005	−0.114	0.080	−0.069
AWC	0.780	0.408	−0.043	0.241	−0.204	−0.077	0.043	−0.179
pH	0.901	−0.133	−0.066	−0.165	−0.212	0.024	0.027	−0.072
EC	0.695	−0.538	−0.203	0.006	0.056	0.195	0.004	0.058
OC	−0.466	0.275	0.571	0.067	−0.408	0.325	−0.133	0.044
Ex. Ca	0.927	0.141	−0.163	0.153	0.033	0.188	−0.082	0.064
Ex. Mg	0.928	0.146	−0.044	0.134	−0.020	0.195	0.019	−0.051
Ex. Na	0.570	−0.282	0.302	−0.382	0.076	0.463	0.271	−0.067
Ex. K	0.472	0.215	−0.297	−0.246	−0.044	0.146	−0.153	0.123
Sum of Exchangeable cations	0.961	0.124	−0.102	0.118	0.018	0.215	−0.030	0.020
Ex. Al+3	−0.750	−0.044	0.077	0.106	0.438	0.226	0.147	0.099
Ex. H+	−0.572	−0.016	0.200	0.498	0.452	0.174	0.014	−0.167
Exchangeable Acidity	−0.677	−0.026	0.179	0.418	0.486	0.205	0.057	−0.098
CEC	0.898	0.221	−0.092	0.150	0.079	0.272	−0.127	−0.067
BS	0.781	−0.061	0.034	−0.061	−0.303	−0.205	0.163	0.226
ESP	0.254	−0.455	0.378	−0.527	0.049	0.280	0.432	−0.058
CEC/clay ratio	0.742	0.120	−0.375	0.063	0.073	0.363	−0.231	−0.026
CaCO3	0.877	−0.142	−0.248	0.074	0.166	0.143	0.033	0.048
Inorganic carbon	0.878	−0.142	−0.248	0.074	0.165	0.142	0.034	0.047
Total carbon	0.349	0.000	0.261	0.154	0.200	−0.412	0.311	0.555
Available N	−0.609	0.598	−0.014	0.016	−0.131	0.326	0.228	0.239
Total N	−0.617	0.600	0.004	0.012	−0.146	0.316	0.222	0.237
CN ratio	0.076	−0.337	0.661	−0.036	−0.376	0.096	−0.408	−0.238

Bold-face factor loadings were considered highly weighted, and the underlined were retained in MDS.

). A representative scree plot showing the variation of eigenvalues with soil components is shown in Figure 3. The soil parameters selected from PC1 were the sum of exchangeable basic cations, exchangeable Ca, exchangeable Mg, pH, CEC, CaCO₃, and inorganic carbon. However, multivariate correlation between these parameters indicated a high correlation (

Table 10. Correlation coefficient (Pearson’s) for highly loaded parameters in PC1 (SQI-2).

	SEBC	pH	CEC	CaCO3	IC
SEBC	1
pH	0.81 **	1
CEC	0.96 **	0.73 **	1
CaCO3	0.89 **	0.80 **	0.79 **	1
IC	0.89 **	0.80 **	0.79 **	1.00 **	1

** Correlation is significant at the 0.01 level (2-tailed).

), and only CEC was retained in the MDS because it is a very important soil property influencing soil structural stability, nutrient availability, and soil pH. In PC2, total N was retained under MDS between available N and total N because of its higher factor loading after correlation results. In PC4, bulk density was retained under MDS between bulk density and porosity because porosity is a derived parameter that is dependent on the BD and PD of the soil. CN ratio, exchangeable acidity, exchangeable Na, particle density, and total carbon were selected as indicators from PC3, PC5, PC6, PC7, and PC8, respectively, since they were the only highly weighted parameters. The correlation data between the selected indicators of MDS are presented in

Table 11. Correlation coefficient (Pearson’s) for selected MDS parameters from PCA results (SQI-2).

	CEC	Total N	CN	BD	Ex. Acidity	Ex. Na	PD	TC
CEC	1
Total N	−0.43 *	1
CN	−0.08	−0.12	1
BD	0.18	−0.33	0.12	1
Ex. Acidity	−0.49 **	0.39 *	−0.07	−0.05	1
Ex. Na	0.47 *	−0.35	0.22	0.08	−0.34	1
PD	0.38	−0.14	0.27	0.23	−0.21	0.23	1
TC	0.16	−0.19	−0.11	0.20	−0.16	0.13	0.23	1

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

.

In the SQI-3 method for soil properties, all 27 soil profiles of the Ap horizon were subjected to PCA to reduce the data dimension. The PCA data for the sub-watershed show the eight PCs with eigenvalues > 1, which explain 87.265% of the cumulative variance in the data. The MDS were chosen based on the highly weighted factor loading of variables (

Table 12. Principal components of soil quality parameters, eigenvalues, and component matrix variables for soil profiles Ap horizons (SQI-3).

Principal Components	PC1	PC2	PC3	PC4	PC5	PC6	PC7	PC8
Eigenvalue	13.800	4.155	3.040	2.282	2.160	1.828	1.231	1.174
% variance	40.589	12.220	8.940	6.711	6.354	5.377	3.622	3.452
% cumulative variance	40.589	52.809	61.749	68.46	74.814	80.191	83.813	87.265
Weightage factor	0.4651	0.1400	0.1024	0.0769	0.0728	0.0616	0.0415	0.0396
Factor loadings (Rotated component matrix)
Coarse sand	−0.814	−0.397	−0.025	0.092	0.012	0.169	0.135	−0.071
Fine sand	−0.613	−0.157	−0.268	0.199	0.450	0.242	−0.319	0.169
Sand	−0.823	−0.328	−0.154	0.160	0.241	0.230	−0.081	0.043
Silt	0.565	−0.276	0.203	−0.278	−0.440	−0.084	0.039	0.115
Clay	0.665	0.582	0.063	−0.023	0.042	−0.233	0.076	−0.127
BD	−0.233	−0.194	0.524	−0.560	0.383	0.212	−0.177	−0.157
PD	−0.013	−0.003	0.259	−0.186	−0.439	0.645	0.186	−0.010
Porosity	0.226	0.198	−0.433	0.505	−0.561	0.033	0.252	0.155
MWHC	0.848	0.280	0.120	0.050	0.107	−0.197	−0.033	−0.161
VE	0.752	−0.030	0.077	0.152	−0.080	−0.130	−0.420	−0.126
FC	0.842	0.239	−0.085	0.040	0.353	0.042	−0.147	0.140
PWP	0.836	0.092	−0.085	0.009	0.270	0.070	−0.131	0.128
AWC	0.676	0.405	−0.067	0.078	0.401	−0.008	−0.140	0.128
pH	0.743	−0.336	−0.251	−0.183	−0.091	0.174	0.114	−0.059
EC	0.129	−0.001	−0.247	−0.199	−0.264	0.293	−0.290	0.717
OC	−0.181	0.834	0.119	−0.266	0.122	0.206	0.185	0.150
Ex. Ca	0.921	−0.093	0.166	0.194	0.076	0.131	0.111	0.026
Ex. Mg	0.906	0.046	0.211	0.070	0.044	−0.031	0.146	−0.065
Ex. Na	0.379	0.066	−0.760	0.020	0.177	−0.247	0.210	0.075
Ex. K	0.551	0.228	−0.042	−0.006	0.152	0.514	0.178	−0.110
Sum of Ex. cations	0.946	−0.037	0.170	0.151	0.071	0.074	0.133	−0.008
Ex. Al+3	−0.688	0.256	0.351	0.096	−0.138	−0.208	0.013	−0.107
Ex. H+	−0.517	0.183	0.480	0.380	0.126	−0.160	0.170	0.363
Exchangeable Acidity	−0.663	0.240	0.493	0.312	0.031	−0.203	0.127	0.214
CEC	0.851	−0.056	0.226	0.368	0.155	0.042	0.142	0.011
BS	0.747	0.121	−0.169	−0.402	−0.163	0.101	−0.004	−0.079
ESP	−0.154	0.137	−0.789	−0.282	0.097	−0.268	0.148	−0.009
CEC/clay ratio	0.660	−0.415	0.141	0.435	0.111	0.180	0.068	0.139
CaCO3	0.844	−0.295	0.181	−0.039	−0.058	−0.010	−0.077	0.085
Inorganic carbon	0.844	−0.295	0.181	−0.039	−0.058	−0.010	−0.077	0.085
Total carbon	0.144	0.527	0.203	−0.118	−0.443	−0.290	−0.442	0.106
Available N	−0.202	0.757	−0.123	0.261	−0.116	0.375	−0.156	−0.191
Total N	−0.206	0.758	−0.128	0.265	−0.117	0.375	−0.149	−0.189
CN ratio	−0.078	0.473	0.236	−0.558	0.277	−0.039	0.357	0.300

Bold-face factor loadings were considered highly weighted, and the underlined were retained in MDS.

). A representative scree plot showing the variation of eigenvalues with soil components is shown in Figure 4. The soil parameters selected from PC1 were the sum of exchangeable basic cations, exchangeable Ca, exchangeable Mg, and CEC. However, multivariate correlation between these parameters indicated a high correlation, and only CEC was retained in the MDS (

Table 13. Correlation coefficient (Pearson’s) for highly loaded parameters in PC1 (SQI-3).

	SEBC	Ex. Ca	Ex. Mg	CEC
SEBC	1
Ex. Ca	0.98 **	1
Ex. Mg	0.95 **	0.88 **	1
CEC	0.94 **	0.93 **	0.89 **	1

** Correlation is significant at the 0.01 level (2-tailed).

). From PC2, organic carbon, available N, and total N were selected, and organic carbon was considered because of its higher factor loading after correlation results (

Table 14. Correlation coefficient (Pearson’s) for highly loaded parameters in PC2 (SQI-3).

	OC	TN	AN
OC	1
TN	0.63 **	1
AN	0.69 **	0.72 **	1

** Correlation is significant at the 0.01 level (2-tailed).

). From PC4, bulk density, porosity, and CN ratio were selected, and bulk density and CN ratio were considered based on correlation (

Table 15. Correlation coefficient (Pearson’s) for highly loaded parameter in PC4 (SQI-3).

	BD	Porosity	CN
BD	1
Porosity	−0.93 **	1
CN	0.28	−0.27	1

** Correlation is significant at the 0.01 level (2-tailed).

). ESP, porosity, particle density, total carbon, and EC were selected as indicators from PC3, PC5, PC6, PC7, and PC8, respectively, since they were the only highly weighted parameters. The correlation data between the selected indicators of MDS are presented in

Table 16. Correlation coefficient (Pearson’s) for selected MDS parameters from PCA results (SQI-3).

	CEC	OC	ESP	BD	CN	Porosity	PD	TC	EC
CEC	1
OC	−0.24	1
ESP	−0.39 *	0.10	1
BD	−0.20	0.10	−0.21	1
CN	−0.17	0.78 **	0.06	0.28	1
Porosity	0.21	−0.05	0.11	−0.93 **	−0.27	1
PD	0.00	0.14	−0.25	0.25	0.04	0.13	1
TC	−0.05	0.28	−0.09	−0.14	0.14	0.15	0.04	1
EC	−0.04	0.03	0.08	−0.11	0.00	0.19	0.20	0.16	1

** Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed).

.

3.2.2. Normalisation of MDS Values

After the selection of parameters for the MDS, all selected observations were transformed using linear scoring functions (less is better, more is better, and optimum) as described in the Materials and Methods. In this study, CEC, CEC/clay ratio, organic carbon, total carbon, total nitrogen, silt, and porosity were considered as more being good from the point of view of soil quality when they were in increasing order; hence, the “more is better” approach was followed. ESP, exchangeable Na, exchangeable acidity, EC, particle density, and bulk density were considered as being less value is better for soil quality; hence, the “less is better” approach was adopted. In the case of the CN ratio, the “optimum is better” approach was followed. After converting the chosen observations into numerical scores within the 0 to 1 range, a weighted additive method was employed to combine them into indices for each soil sample [24,28]. Each Principal Component (PC) explained a specific portion of the variation in the dataset (Table 4, Table 9 and Table 12). To assign a particular weight value to each PC, the variation explained by each PC was divided by the total maximum variation among all PCs selected for the Minimum Data Set (MDS) [28]. For instance, in the case of SQI-1 % variance, which had a value of 48.237, it was divided by the cumulative total variance (85.658) to yield a weight value of 0.563 for PC-1 (Table 4).

3.2.3. SQI Calculation

Thereafter, to obtain the weighted additive SQI, the weighted MDS indicator scores for each observation were summed up using the following equation:

SQI = Σ Principal Component Weight ∗ Individual soil parameter score

The soil quality index of 27 soil profiles in the Ganjigatti sub-watershed was assessed using three methodologies and is presented in

Table 17. Soil quality index (SQI) values for 27 soil profiles using different methods.

Profile Number	Depth of Soil Profile (cm)	SQI Method
		Horizon Wise	Weighted Average	Ap Horizon
Pedon-1	28	0.46	0.51	0.55
Pedon-2	39	0.45	0.55	0.58
Pedon-3	49	0.52	0.55	0.60
Pedon-4	79	0.73	0.82	0.89
Pedon-5	65	0.49	0.54	0.61
Pedon-6	180	0.70	0.78	0.82
Pedon-7	67	0.37	0.43	0.63
Pedon-8	40	0.44	0.53	0.67
Pedon-9	80	0.60	0.71	0.79
Pedon-10	35	0.44	0.52	0.57
Pedon-11	200	0.74	0.85	0.77
Pedon-12	35	0.30	0.40	0.52
Pedon-13	26	0.37	0.45	0.50
Pedon-14	20	0.41	0.48	0.62
Pedon-15	90	0.42	0.49	0.62
Pedon-16	46	0.38	0.46	0.58
Pedon-17	120	0.37	0.44	0.53
Pedon-18	170	0.41	0.47	0.59
Pedon-19	130	0.65	0.75	0.71
Pedon-20	20	0.39	0.48	0.59
Pedon-21	35	0.43	0.51	0.53
Pedon-22	30	0.34	0.41	0.48
Pedon-23	130	0.64	0.73	0.73
Pedon-24	90	0.45	0.49	0.53
Pedon-25	55	0.43	0.51	0.63
Pedon-26	30	0.26	0.32	0.39
Pedon-27	180	0.57	0.61	0.63

. In the SQI-1 method, the soil quality of pedons ranged from 0.26 (pedon-26) to 0.74 (pedon-11). Pedons 12, 22, and 26 were classified under low SQI (<0.35); pedons 4, 6, 9, 11, 19, 23, and 27 were classified under high SQI (>0.55); and the remaining pedons were under medium SQI (0.35–0.55). All three low SQI pedons were upland pedons, which are characterised by a shallow depth (<35 cm), a low CEC, a low clay content, a low CEC/clay ratio, an acidic pH, and a high exchangeable acidity. Interestingly, all three of these pedons were classified as Entisols. In the high SQI pedons, pedons 4 and 9 were moderately deep (75 to 100 cm), and the remaining pedons were deep to very deep (100 to 200 cm). These pedons were characterised by a high CEC, high clay content, high porosity, low BD, and low exchangeable acidity. Pedons 4 and 9 were located midland, and the remaining pedons (6, 11, 19, 23, and 27) were located in low-land topography positions. The SQI-1 values of the Ganjigatti sub-watershed depicted in maps using GIS software (Figure 5) showed that the majority of the area is under the medium category of SQI (0.35–0.55), comprising about 72.40% of the sub-watershed area. The high category of SQI (>0.55) comprised 12.92%, and low SQI (<0.35) comprised 6.45% of the sub-watershed.

In the SQI-2 method, the soil quality of pedons ranged from 0.32 (pedon-26) to 0.85 (pedon-11). The spatial distribution map of SQI using the SQI-2 method (Figure 6) revealed that the low, medium, and high categories of SQI covered 2.18, 54.11, and 35.49% of the total geographical area of the Ganjigatti sub-watershed, respectively. In the SQI-3 method, the soil quality ranged from 0.39 (pedon-26) to 0.89 (pedon-4), and the spatial distribution map of SQI-3 (Figure 7) indicated that 81.04% of the area was classified under the high category of SQI.

3.2.4. Pearson’s Correlation Coefficients of Soil Quality Index (SQI) versus Crop Yield

The primary aim of this research was to assess the validity of Soil Quality Index (SQI) methodologies in relation to crop yield, which is an aspect that has been relatively understudied with a few exceptions noted in references [22,24,36,38,50]. The challenge lies in establishing a correlation between crop yields and SQI, whether measured at different soil depths or as a combination of soil layers, as there is currently no standardised technique available, and this area has received limited attention until now. However, in order to evaluate the index performance in the current study, SQIs were linked to the end-point variable (i.e., yield). The estimated SQI values were correlated with the recorded yields of soybean, maize, sorghum, and greengram in the Ganjigatti sub-watershed. Results of the present study revealed that (

Table 18. Pearson’s correlation coefficients (r) of soil quality index (SQI) versus crop yield.

	Soybean	Maize	Sorghum	Greengram
SQI-1	0.74 **	0.69 **	0.69 **	0.73 **
SQI-2	0.63 **	0.61 **	0.58 **	0.67 **
SQI-3	0.58 **	0.58 **	0.54 **	0.53 **

** Correlation is significant at the 0.01 level (2-tailed).

) yields of all four crops were significantly correlated with all three SQI methodologies.

4. Discussion

SQI is a composite of a few selected soil indicator characteristics, and it necessitates the selection of the most relevant qualities with a dominant influence on soil functions. It is noteworthy that the MDSs derived from SQI-1, SQI-2, and SQI-3 were very similar, with the exception of the presence or absence of two or three different parameters and the order (or) weightage being different. The SQI computation and ranking of soils based on SQI values reflected these results. It has been argued that a more comprehensive data collection or a larger subset of indicators may more accurately indicate soil quality; nevertheless, this can lead to data duplication when there is a high correlation between the indicators used [64,65]. ESP and CN ratio were included under MDS from PC3 of the SQI-1 method; B.D and CN ratio were included under MDS from PC4 of the SQI-3 method since they were not correlated. Within the MDS of three distinct methods, Cation Exchange Capacity (CEC) had the highest influence on the Soil Quality Index calculation, followed by porosity, Exchangeable Sodium Percentage (ESP), organic carbon, CN ratio, and total nitrogen. These factors have been frequently recognised as significant and sensitive variables in the development of the SQI, as indicated in previous studies [24,30,36,38,62,66,67,68].

Each of these factors plays a pivotal role in influencing the amalgamation of soil physicochemical and biological attributes, as well as soil fertility, productivity, and the various components that contribute to crop yield [69]. CEC is a very important parameter for the Ganjigatti sub-watershed because it influences the nutrient-supplying capacity of soils as it depends on the quantity and type of clay, soil pH, and organic matter [38,70]. On the other hand, CEC provides indications about the clay mineralogy of the soil, which is also responsible for the quality of soil [37]. Porosity significantly influences soil quality by affecting water-holding capacity, soil aeration, nutrient availability, root growth, and microbial activity. Mainly, pore spaces facilitate the availability and movement of air or water within the soil environment. Soil porosity and its capacity to retain water are additional soil characteristics that contribute to creating a favourable environment for microorganisms and moisture retention in the soil. Consequently, this promotes the enhanced decomposition of soil organic carbon by microbial communities. Soil porosity offers space for microbial proliferation and enhances soil aeration, nutrient accessibility, as well as the soil’s capacity for both water drainage and retention [37,71]. Maintaining a well-structured and porous soil allows for better water and nutrient management, supports root development, and promotes the activity of beneficial soil organisms, ultimately leading to improved plant growth and productivity.

Organic carbon is interlinked to all measures of soil quality through a complex web of many factors. This factor has a vital function in the cycling and storage of essential soil nutrients, contributes to the formation of soil structure, and serves as the primary nutrient source for heterotrophic microorganisms within the soil [69]. The C:N ratio is an important parameter that affects soil quality and nutrient availability. It is a measure of the relative amounts of C and N in soil. A higher ratio can lead to nitrogen immobilisation and slower decomposition, while a lower ratio promotes faster decomposition and nutrient release. Achieving a balanced C:N ratio that meets the specific needs of plants can contribute to a healthier soil and improved crop productivity [72,73]. ESP is a measure of the sodium content relative to other cations in the soil. It is an important parameter for assessing soil quality, particularly in terms of soil structure, permeability, and fertility. High ESP in soil can degrade soil structure, reduce water infiltration and drainage, affect nutrient availability, and alter soil pH. Managing high ESP levels through appropriate soil management practices is crucial for maintaining soil quality and supporting healthy plant growth [36,74,75]. Total nitrogen (TN) is a critical component of soil quality and plays a fundamental role in supporting plant growth and overall ecosystem functioning. It is important to note that while nitrogen is crucial for plant growth and soil quality, its excessive application or imbalanced ratios with other nutrients can lead to environmental issues such as nitrogen leaching, eutrophication of water bodies, and air pollution. In summary, total nitrogen is a key factor in soil quality, affecting nutrient availability, plant growth and yield, soil fertility, organic matter decomposition, and microbial activity.

Results on the correlation between crop yields and SQI methods revealed that the yields of all four crops were significantly correlated with all three SQI methodologies. Several studies [38,62,68,76,77] have also found a significant correlation between the yield components of various agricultural products and SQIs with different coefficients of variations. A relatively higher correlation coefficient was observed with SQI-1 and -2 than with SQI-3 (Table 17), which indicates the pedological significance and the effect of soil subsurface phenomena on the physiological conditions of plant systems, which take nutrients and water from the subsurface. Since crop production is influenced by both surface and subsurface attributes, which are inherently linked to pedogenic processes, the soil quality index (SQI) assessment using solely surface soil parameters does not provide enough information [78]. However, inherent properties limit the extent to which dynamic soil properties can change [79]. The pedogenic processes have an impact on the inherent properties of the soil, and the changes are particularly noticeable in tropical climates because of physical and chemical weathering that is accelerated by high temperatures and precipitation. Many of the early assessments of soil quality [22,23] used surface (dynamic) soil parameters, while studies utilising soil profile data (dynamic and inherent) are scarce [36,37]. Although surface soil characteristics are simple to measure and assess, they only provide partial information because pedogenic processes in the soil control section drive soil functions. The soil properties that have the higher impact on soil functions can be determined by evaluating the soil quality utilising both surface and subsurface properties. Soil profile parameters that are determined through soil genesis and reflected by taxonomy are the only ones that should be used to evaluate soil quality [80,81]. Therefore, pedogenesis must be taken into consideration when assessing the soil quality in SAT soils, and both the surface and subsurface soil properties should be given the appropriate weight.

In order to improve and restore the soil’s health, ref. [36] used the SQI to evaluate the soil quality of the Indo-Gangetic Plains. There was a moderate association between SQI and yields in the rice-wheat system. They drew a conclusion using the trial-and-error method by taking into account the percentage contribution of SQI for each layer in the calculation. Finally, they came to the conclusion that a composite SQI value was produced by combining 70% of the surface SQI value and 30% of the subsurface SQI value. In the present study, a good and significant correlation between crop yields and the SQI-3 method was probably due to the fact that all four crops (soybean, maize, sorghum, and green gram) considered in this study are shallow-rooted, and their root density is in the top layer of the soil, which can drive the nutrients from the surface horizons. In view of this, we conclude that when subsurface soil variables are considered alongside dynamic surface properties while evaluating the SQI using the weighted index approach, a good correlation between the SQI and defined soil function is established [36,38,80,81,82].

Furthermore, assessing the horizon-wise soil quality index (SQI-1) of a profile can provide more detailed and accurate information compared to using a weighted average (SQI-2). Weighted averages may provide an overall summary measure but can sometimes mask variations that occur at different horizons. Assessing the SQI horizon-wise allows for a more granular understanding of the profile’s soil quality. Here are a few reasons why horizon-wise assessment is advantageous. (1) A soil profile consists of different layers or horizons, each with its unique characteristics and properties. By assessing the SQI horizon-wise, you can account for variations in soil quality across different horizons. This helps capture the heterogeneity within the profile and provides a more comprehensive understanding of soil quality. (2) Soil quality problems, such as compaction, erosion, or nutrient depletion, may occur at specific horizons rather than being evenly distributed across the entire profile. Evaluating the SQI horizon-wise enables the identification of specific horizons that may require attention or management interventions. (3) Horizon-wise assessment allows for more targeted and site-specific management decisions. For example, if the SQI is low in a particular horizon, you can focus on implementing appropriate soil management or nutrient management strategies specific to that horizon. (4) Soil horizons interact with each other, influencing the overall soil quality of the profile. By examining the SQI horizon-wise, you can gain insights into how different horizons interact and affect the overall soil quality. This knowledge is crucial for understanding the complex dynamics within the soil profile. (5) For the assessment of the SQI through PCA, it is better to consider the soil properties of every horizon (SQI-1) instead of the weighted average of the soil properties of a profile (SQI-2). Since PCA uses a number of statistical tools, it could avoid any bias and data redundancy by choosing an MDS using mathematical formulas from huge data sets [11,28]. In conclusion, while weighted averages (SQI-2) can be useful for providing an overall summary measure, horizon-wise assessment of the Soil Quality Index (SQI-1) provides a more detailed and accurate representation of soil quality by accounting for variations, specific issues, targeted management decisions, and interactions between different horizons. The current data suggest that the SQI values were relatively higher in the SQI-2 and SQI-3 methods than in SQI-1. However, SQI-1 appears to be the best method among the three, particularly under the long-term scenario, especially due to its objective approach and relatively higher correlation with crop yield.

According to the SQI-1 method, soil quality in the study area varied from low to high. The large variation in soil quality is due to soil heterogeneity and soil degradation caused by erosion. The SQI of the study area ranged from 0.26 to 0.74, and the spatial distribution of the SQI of the study area (Figure 5) shows that about 6.45% of the sub-watershed has a low category of SQI (<0.35). The low SQI of the Ganjigatti sub-watershed might be due to unfavourable microclimatic conditions, soil erosion, inappropriate land use and management practices, and the high removal of available nutrients by annual crops without replenishment. These soils are located in upland positions with moderate to steep slopes, and, hence, the leaching of basic cations and erosion of the soil to lowlands could be a possible reason for the low soil quality. The cultivation of deep-rooted crops like cotton and pigeon pea is not suitable in these soils because of the limitation imposed by shallow depths (<35 cm). The SQI of pedons 1, 2, 3, 5, 7, 8, 10, 13, 14, 15, 16, 17, 18, 20, 21, 24, and 25 was in the medium category (0.35–0.55), which is the bulk of the area covered in the sub-watershed and comprising about 72.40%. The medium SQI might be due to soil erosion, land use, and management practices fallowed in that area. On the other hand, a few mid- (4 and 9) and low-land pedons (6, 11, 19, 23 and 27) showed a high SQI (>0.55) because of the deposition of soil and nutrients from uplands through the soil erosion process. These soils occupy a sizable area of 12.92% of the sub-watershed.

Soil quality can be used to evaluate cropping systems and recommend alternatives in a particular region [83]. Climate change affects short-term and long-term soil processes, which must be considered when establishing management measures to maintain soil resources and sustain agricultural output. Most crop models, such as InfoCrop and CERES-Wheat, are driven by biophysical parameters, rainfall variability, water balance, and economic implications [84,85] and pay little attention to soil quality [86], especially in India. According to [87], alterations in soil quality over time serve as indicators of whether the soil condition is sustainable or not. Maintaining soil quality at the desired level presents a complex challenge due to the interplay of climatic conditions, soil properties, plant factors, and human activities. This challenge is particularly pronounced in lowland rice cropping systems, largely owing to practices involving soil puddling during preparation [88]. The SQI includes many soil properties as indicators of soil quality, so integrating them into simulation models to predict the effects of climate change on soil functions and crop yield will improve the knowledge and accuracy of models, allowing for better management practices.

5. Conclusions

The Soil Quality Index is a valuable tool for assessing and managing soil health, which is fundamental for sustainable agriculture and environmental preservation. The study highlights the need for a holistic approach that considers both surface and subsurface soil properties, factors that impact soil quality.

1. The current study, within the context of the MDS-determined SQI, has demonstrated that CEC holds the most substantial influence and impact, with porosity, ESP, OC, CN ratio, and total N following in order of importance.

2. By assessing the SQI horizon-wise, we can analyse the quality of the soil at each individual horizon separately. Furthermore, the horizon-wise assessment of SQI provided reliable frameworks for soil quality evaluation in Ganjigatti sub-watershed in the hilly zone of Karnataka. The soil quality within the study area showed significant spatial variability, ranging from low to high SQI. In the sub-watershed, the largest portion of the area, accounting for 72.40%, is categorised as having a medium SQI (ranging from 0.35 to 0.55), while 12.92% is classified as having a high SQI (greater than 0.55), and 6.45% is in the low SQI category (less than 0.35). This approach allows for a more granular understanding of the soil quality.

3. There is a significant correlation between crop yields and all three SQI methodologies, with SQI-1 and SQI-2 showing higher coefficients. This underscores the importance of considering both surface and subsurface soil properties when assessing the SQI.

Factors like soil erosion, microclimatic conditions, land use, and management practices play a crucial role in determining soil quality. SQI can guide management practices in agriculture by identifying areas with a low soil quality and suggesting alternative cropping systems.