Construction and Application of the Phaeozem Health Evaluation System in Liaoning Province, China

Abstract

Soil degradation has led to a continuous decline in the quality of phaeozem, which is seriously threatening the foundation of national food security. Therefore, precisely evaluating the health status of phaeozem resources and their spatial and temporal variations are crucial for ensuring the effective implementation of soil degradation prevention and control strategies. In this study, soil data from 140 sites were collected, and eight physical and chemical indices (sand content, silt content, pedon thickness, organic matter, total potassium, total nitrogen, total phosphorus, and pH) were utilized to assess the soil health status of phaeozem in Liaoning Province. The results showed the following: (1) The minimum data set is aligned with previous research findings and effectively mirrors the soil’s health condition. (2) The substantial coefficients of variation observed for pedon thickness, sand content, and silt content across regions suggest notable differences, likely influenced by climatic variations, parent material differences, and anthropogenic activities. (3) The study area’s average soil pH of 6.39 indicates an overall acidic nature, potentially attributed to intense soil weathering and suboptimal fertilization practices. (4) The soil health index (SHI) ranged from 0.33 to 0.91, with an average of 0.62, which indicates that the majority of the phaeozem in Liaoning Province are in a sub-healthy state (SHI < 0.6 = unhealthy, SHI 0.6–0.8 = sub-healthy, SHI ≥ 0.8 = healthy). These sub-healthy areas are primarily located in hills, mountains, and the Liaohe Plain, and are significantly impacted by rainfall and wind erosion. Meanwhile, unhealthy areas are concentrated in the south and center of the province, characterized by fragile ecological environments and challenging agricultural conditions.

1. Introduction

Climate change and population growth pose challenges to food security and sustainable utilization of land resources, making them the focus of attention in today’s society [1,2]. In China, food security is particularly important; therefore, farmland is viewed as an essential strategic resource for safeguarding national food security [3]. Despite China’s implementation of the most stringent farmland protection policies, the phaeozem region in Northeast China faces soil issues due to its unique geographical and climatic conditions, such as freeze–thaw alternation, soil erosion, and wind erosion, leading to a decline in soil quality [3,4].

The phaeozem mainly refers to a kind of soil with dark humified substances on the surface layer developed under the conditions of temperate and cold-temperate grassland meadow vegetation, which mainly includes the soil types of black soils, meadow soils, chernozem, brown soil, dark brown soil, and albic soil in the Soil Genetic Classification of China [5]. The phaeozem region in Liaoning Province, situated in the southern part of Northeast China, spans approximately 1.87 × 10⁶ hm², constituting 12.32% of the total phaeozem region in Northeast China, which serves as a crucial commodity grain production hub for the entire country [6]. It also faces soil degradation issues such as a decline in organic matter and thinning of the plow layer in some areas due to unreasonable utilization and development. Therefore, it is necessary to conduct research and investigation on the health status of the phaeozem region in Liaoning Province to understand the spatial and temporal changes in soil properties and cultivated land quality, which is crucial for ensuring soil health and sustainable utilization in this area.

Out of the 17 Sustainable Development Goals set by the United Nations, 13 goals are directly or indirectly related to soil [7,8]. Understanding the current status of soil health, conducting systematic diagnosis, and establishing an early warning system are the foundation for safeguarding healthy soil and also an important component of sustainable global soil management [9,10]. Therefore, whether soil health can be scientifically and accurately assessed directly affects the achievement of global sustainable development goals [11,12]. Existing soil evaluation methods are diverse but each has its own differences. The intricate nature of soil ecosystems, encompassing their composition, characteristics, and roles, coupled with the multitude of interactions between environmental elements and soil components, posed challenges in precisely evaluating soil health [12,13,14]. Furthermore, considering the substantial variations in natural soil formation conditions and human management practices across regions, it became imperative to choose an evaluation method that was both widely applicable and precise.

At present, soil health assessment tools are mainly widely used at the field scale [15,16,17]. However, at larger spatial scales, the evaluation indices and criteria used should be different depending on the soil type and land use pattern [17,18,19]. Therefore, it is necessary to establish a soil health evaluation system at larger spatial scales to scientifically assess soil health.

This study aims to establish a soil health evaluation system for the phaeozem in Liaoning Province based on the principal component analysis (PCA) and minimum data set (MDS) to identify the health status of phaeozem soil in Liaoning Province. The results are expected to provide a scientific basis for the precise management and sustainable utilization of phaeozem, which is conducive to promoting the sustainable utilization of phaeozem resources and ensuring ecological safety.

2. Materials and Methods

2.1. Overview of the Study Area

Liaoning Province is located in the southernmost part of Northeast China. It is bordered on the south by the Yellow Sea and the Bohai Sea, on the east by North Korea, on the west by the Inner Mongolia Autonomous Region, on the southwest by Hebei Province, and on the northeast by Jilin Province [20]. The eastern and western Liaoning are mountains and hills, which account for 60% of the total area of the province. The central region is the Liaohe Plain. The coast of Bohai Sea in western Liaoning is a long and narrow coastal plain (Figure 1) [20]. Liaoning has a temperate continental monsoon climate. The annual average temperature is between 7–11 °C, which varies greatly due to the influence of monsoon climate. The temperature decreases from southwest to northeast and from plain to mountainous area. The average annual frost-free period is 130–200 days, which gradually increases from northwest to southeast. The annual precipitation is between 600 and 1100 mm. The annual precipitation in the eastern mountainous and hilly areas is above 1100 mm. The mountainous and hilly areas in the west are connected with the Inner Mongolia Plateau, and the annual precipitation is about 400 mm, while the precipitation in the central plain is moderate, with an average annual precipitation of about 600 mm [21].

2.2. Data Sources

The data for this study were derived from sources such as “Soil Series of China—Liaoning Volume” [22], “China Soil Species (Volume 2)” [23], and “Soils of Liaoning” [24], covering the profile location coordinates, profile descriptions, physical and chemical properties, and other relevant data of the phaeozem in Liaoning Province (including soil types such as black soil, chernozem, albic soil, brown soil, dark brown soil, and meadow soil). It is based on the theory of soil genesis, considering soil-forming environmental information such as topography, parent material, and land use patterns in the sample area, and adhering to the principle of relatively uniform spatial distribution for selecting sample points.

2.3. Selection Method of Soil Health Evaluation Index

A literature review was conducted on the selection of soil health/quality evaluation indices in CNKI (China National Knowledge Infrastructure) (China Academic Journals (CD Edition) Electronic Publishing House, Beijing, China) covering a span of 23 years from January 2000 to March 2023 [15]. Statistical analysis was performed on the selected indices mentioned in the literature, and the frequency of each index’s selection was calculated [25]. Based on the characteristics of phaeozem, an initial selection of soil health evaluation indices was made, resulting in the establishment of a comprehensive data set.

2.4. Data Processing and Analysis Methods

2.4.1. Principal Component Analysis (PCA)

The factor analysis (FA) method using PCA was employed to perform dimensionality reduction on the preliminary indices determined by the aforementioned process. Specifically, through linear transformation, a small number of representative comprehensive indices were selected from numerous variables to represent the originally selected multi-dimensional variables, thus preserving key features of the preliminary indices and eliminating redundant indices with significant correlations. By using the dimensionality reduction-factor analysis function in SPSS 26 software (International Business Machines Corporation (IBM), New York, USA), the eigenvalues and contribution rates were derived, which determined the principal components [25].

2.4.2. Construction of Minimum Data Set (MDS)

In a multi-dimensional space, the Norm value represents the length of the vector model of a variable, reflecting the comprehensive loading of the principal component. Specifically, a longer length indicates a greater loading, and a stronger capacity to interpret information [16,26].

The calculation formula is as follows.

$$N_{ik}=\sqrt{\sum _{i=1}^k\left({\mu }_{ik}^2·{\lambda }_k\right)}$$

where “N_ik” denotes the comprehensive load of the i-th index on k principal components prior to the eigenvalue reaching or exceeding 1; “$\mu$_ik” represents the load of the i-th index on the k-th principal component; “λ_k” is the eigenvalue corresponding to the k-th principal component.

By employing PCA, calculating the Norm values of selected indices, and conducting correlation analysis, redundancy can be minimized, resulting in the construction of an MDS that accurately reflects the health condition of the soil and enables the evaluation of the health status of phaeozem [25,26].

2.4.3. Grouping and Score Arrangement

The principal components with eigenvalues greater than or equal to 1 were identified, and the indices were grouped with loading values greater than or equal to 0.5. If an index appears in multiple principal components, its correlation analysis results need to be considered, and it should be assigned to the group with lower correlation. Subsequently, the Norm value of each group of indices is calculated, and the indices with Norm values within 10% of the maximum value in that group are selected as candidate indices. If there are multiple indices in the same group, correlation analysis must be performed to eliminate redundancy, retaining only the index with the largest Norm value. If the correlation coefficients of all indices are ≤0.5, they are all retained in the minimum data set [27].

2.4.4. Bartlett Test

To determine if the selected indices (i.e., variables) are suitable for factor analysis, the independence of each initially selected index in the principal component analysis is tested through the Bartlett test [25].

2.4.5. Fuzzy Mathematics Membership Function

The fuzzy mathematic membership function is designed to show the relationship between the soil health evaluation indices selected for a certain study area and the soil health, and to analyze the positive and negative effects of the indices on the soil to determine the critical value [28]. Based on the positive and negative effects of indices on soil health and their critical values [29], different types of membership functions are selected (

Table 1. Piecewise formula of membership function.

Membership Function Type	Calculation Formula	Parameter Descriptions
Upwards membership function	$\mathrm{y}(\mathrm{x})=\left\{\begin{array}{c}1,\hspace{1em}x\ge b \\ {\displaystyle \frac{\mathrm{x}-\mathrm{a}}{\mathrm{b}-\mathrm{a}}},\hspace{1em}a<x<b\\ 0,\hspace{1em}x\le a\end{array}\right.$	Y(x): membership function x: Measured value of index a: Lower threshold b: Upper threshold
Downward membership function	$\mathrm{y}(\mathrm{x})=\left\{\begin{array}{c}1,\hspace{1em}x\le a\\ {\displaystyle \frac{\mathrm{x}-\mathrm{b}}{\mathrm{a}-\mathrm{b}}},\hspace{1em}a<x<b\\ 0,\hspace{1em}x\ge b\end{array}\right.$
Peak-type membership function	$\mathrm{y}(\mathrm{x})=\left\{\begin{array}{c}1,\hspace{1em}{\mathrm{b}}_2\ge x\ge {\mathrm{b}}_1 \\ {\displaystyle \frac{\mathrm{x}-{\mathrm{a}}_1}{{\mathrm{b}}_1-{\mathrm{a}}_1}},\hspace{1em}{\mathrm{a}}_1<x<{\mathrm{b}}_1\\ {\displaystyle \frac{\mathrm{x}-{\mathrm{a}}_2}{{\mathrm{b}}_2-{\mathrm{a}}_2}},\hspace{1em}{\mathrm{a}}_2>x>{\mathrm{b}}_2\\ 0,\hspace{1em}x\le {\mathrm{a}}_1 \mathrm{or} x\ge {\mathrm{a}}_2\end{array}\right.$	y(x): membership function
		x: Measured value of index
		a1: Lower threshold
		a2: Upper threshold
		b1: Lower limit of optimum value
		b2: Upper limit of optimum value

). An Upwards function is adopted for positive correlations, while a Downward function is chosen for negative correlations. If an indicator is beneficial to crops within a specific range but harmful outside that range, a Peak-type membership function is used. The membership degrees of each health evaluation index are obtained through calculations and will be used to evaluate soil health [28,29,30].

2.4.6. Relevant Calculation Formulas

Coefficient of Variation Calculation

The coefficient of variation (CV) is typically utilized to assess the sensitivity of an index [31].

The calculation formula is as follows.

$$\mathrm{C}\mathrm{V}={\displaystyle \frac{\mathrm{S}\mathrm{D}}{\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}}}$$

CV< 10%: the index is not sensitive;

CV between 10% and 50%: the index is low sensitive;

CV between 50% and 100%: the index is moderate sensitive;

CV exceeding 100%: the index is highly sensitive.

As the sensitivity of the index decreased, its fluctuation diminished, resulting in a more stable index.

Soil Health Index (SHI) Calculation

To calculate the SHI, the MDS and PCA are employed to determine the weight of each index and the membership degree derived from the membership function, thereby obtaining the soil health index [29].

The calculation formula is as follows.

$$\mathrm{S}\mathrm{H}\mathrm{I}=\sum _{\mathrm{i}=1}^{\mathrm{n}}\left({\mathrm{H}}_{\mathrm{i}}·{\mathrm{C}}_{\mathrm{i}}\right)$$

where “SHI” signifies the comprehensive fertility index of cultivated land; “H_i” denotes the membership degree of the i-th evaluation factor; “C_i” represents the weight of the i-th combination of evaluation factors.

The Descriptive Statistics and Calculated Variable in SPSS Statistics 26 were used for data statistics. The SigmaPlot 12.5 (Systat Software Inc., San Jose, CA, USA) and ArcGIS 10.7 (Environmental Systems Research Institute (ESRI), Redlands, CA, USA) were used for plotting graphs.

3. Results

3.1. Screening the Health Evaluation Index of Phaeozem

3.1.1. Primary Selection of Soil Health Evaluation Indices

The CNKI was used as the search database for literature analysis. Relevant documents on soil health evaluation were collected, and high-frequency soil health indices were selected as the initial screening indices. Within the set 23-year period, 557 documents were uncovered in the initial search. Following a thorough screening process, articles unrelated to soil health evaluation indices were discarded, leaving 132 relevant papers.

Statistical analysis was then conducted on these 132 selected documents, and a total of 66 specific physical- and chemical-related indices were collected.

Physical Index

Physical indices can reflect various aspects of soil, including its inherent properties, structural characteristics, permeability, water, temperature conductivity, erosion resistance, and suitability for cultivation [32,33]. Based on the results of the literature review, physical indices included bulk density, soil clay content, pedon thickness, texture, water content, aggregate stability, and porosity. Among these, bulk density emerged as the most frequently chosen physical index, with a high selection frequency of 41.7%. Soil clay content, soil layer thickness, and soil texture were selected with frequencies exceeding 20%, whereas water content and aggregate stability had selection frequencies above 15% (Figure 2).

Chemical Index

Chemical indices serve as essential parameters for evaluating the various nutrients present in the soil, and include 35 indices, such as organic matter, total nitrogen, total phosphorus, and pH [34]. Among them, organic matter was selected with the highest frequency, accounting for 87.1%; followed by soil pH (72.0%). In addition, the selection frequencies of total nitrogen, phosphorus, and potassium nutrients in the soil were 50.0%, 48.5%, and 34.1%, respectively (Figure 3).

3.1.2. Preliminary Screening Results of Soil Indices

Of the 66 indices selected from the literature, those with a frequency below 5% were excluded as rare. The remaining indices were categorized as follows. Those with a selection frequency greater than 40% were considered as having a high selection frequency. Those with a selection frequency between 20% and 40% were considered as having a moderate selection frequency. Those with a selection frequency between 5% and 20% were considered as having a low selection frequency. Among them, organic matter had the highest selection frequency, while pH, total nitrogen, total phosphorus, available potassium, and total potassium had moderate selection frequencies (Figure 3 and

Table 2. Primary indices for first and second data sets.

Data	Primary Indices
First group	Black soil thickness	Pedon thickness	Bulk density	Organic matter	pH	Total nitrogen	Total phosphorus	Total potassium	Sand content	Silt content	Clay content
Second group	Black soil thickness	Pedon thickness	Organic matter	pH	Total nitrogen	Total phosphorus	Total potassium	Sand content	Silt content	Clay content

).

Considering the physical and chemical properties of phaeozem and integrating relevant contents of phaeozem evaluation in the literature, 11 indices, including soil bulk density, soil organic matter, black soil thickness, pedon thickness, soil pH, total nitrogen, total phosphorus, total potassium, silt content, sand content, and clay content, were preliminarily selected to construct the full data set.

The data were divided into two groups according to the presence or absence of bulk density. The data in the book of Soil Series of China—Liaoning Volume all had bulk density and were placed in the first group; the data in the books of China Soil Species (Volume II) and the Soils of Liaoning did not have bulk density and were placed in the second group (Table 2).

3.1.3. Minimum Data Set Construction

The PCA was performed on the two sets of indices mentioned above, combining the correlation analysis between Norm value and the preliminary indices, further screening the preliminary indices to construct a minimum data set.

The first set of data was analyzed to produce five principal components with a cumulative contribution of 73.72% and a significance level of less than 0.05 (

Table 3. Explanation of the total variance of the primary selection index in the first data set.

Indicator Name	Characteristic Root	Variance Interpretation Rate (%)	Accumulation (%)	Characteristic Root	Variance Interpretation Rate (%)	Accumulation (%)
Sand content	2.61	23.74	23.74	2.61	23.74	23.74
Silt content	1.57	14.25	37.99	1.57	14.25	37.98
Clay content	1.48	13.52	51.50	1.49	13.52	51.50
Bulk density	1.39	12.68	64.18	1.39	12.68	64.18
Black soil thickness	1.05	9.54	73.72	1.05	9.54	73.72
Total phosphorus	0.88	8.02	81.74
Total potassium	0.62	5.61	87.35
Pedon thickness	0.55	4.97	92.32
Organic matter	0.45	4.05	96.36
pH	0.39	3.56	99.92
Total nitrogen	0.01	0.08	100

and

Table 4. Loadings of the first set of data-priming indices on each principal component.

Name	Principal Component
	PC1	PC2	PC3	PC4	PC5
Sand content	0.933	−0.208	0.051	0.141	−0.072
Silt content	−0.749	0.225	0.125	0.243	0.167
Clay content	−0.696	0.058	−0.212	−0.476	−0.055
Bulk density	0.606	0.225	−0.181	−0.248	0.378
Black soil thickness	0.298	0.706	−0.224	0.082	−0.011
Total phosphorus	0.065	−0.666	−0.45	0.303	−0.091
Total potassium	0.259	0.214	0.628	−0.528	−0.086
Pedon thickness	−0.069	0.566	−0.600	0.224	−0.181
Organic matter	−0.265	0.058	0.576	0.538	−0.207
pH	0.285	0.326	0.215	0.508	−0.059
Total nitrogen	−0.107	−0.040	0.083	0.229	0.88

).

There were five main components in the first group (Table 4). Four primary indices with absolute loadings were greater than 0.50 in PC1, namely sand content, silt content, clay content, and bulk density. Among them, the Norm value of sand content was the largest, with a value of 4.65 (

Table 5. Norm values for the first data set of primary indices.

Name	Sand Content	Silt Content	Clay Content	Bulk Density	Total Potassium	Organic Matter	Pedon Thickness	Total Phosphorus	Black Soil Thickness	Total Nitrogen	pH
Norm	4.65	3.91	3.88	3.47	3.30	3.25	3.24	3.22	3.16	3.91	2.72

). There were no indices with Norm values within 10% of this value in this group, so only sand content was selected into the MDS of the first group of indices in PC1.

The primary indices with absolute loadings greater than 0.50 in PC2 were black soil thickness, total phosphorus, and pedon thickness. Among them, the Norm value of pedon thickness was the largest, with a value of 3.24. The Norm values of black soil thickness and total phosphorus were both within 10% of this value, but only the correlation coefficient of total phosphorus was less than 0.50 (

Table 6. Results of correlation analysis of the first data set.

Name	Black Soil Thickness	Pedon Thickness	Sand Content	Silt Sand	Clay Sand	Bulk Density	pH	Organic Matter	Total Nitrogen	Total Phosphorus	Total Potassium
Black soil thickness	1
Pedon thickness	0.389 **	1
Sand content	0.131	−0.103	1
Silt sand	−0.131	0.078	−0.780 **	1
Clay sand	−0.129	0.076	−0.746 **	0.188	1
Bulk density	0.236 *	0.01	0.364 **	−0.280 *	−0.265 *	1
pH	0.14	0.062	0.18	0.004	−0.299 **	0.148	1
Organic matter	−0.028	−0.093	−0.102	0.255 *	−0.091	−0.391 **	0.149	1
Total nitrogen	−0.005	−0.079	−0.06	0.12	−0.029	0.016	−0.031	0.077	1
Total phosphorus	−0.196	−0.063	0.152	−0.138	−0.087	−0.082	−0.09	−0.102	−0.02	1
Total potassium	0.076	−0.276 *	0.145	−0.179	−0.063	0.137	0.005	0.028	−0.111	−0.390 **	1

Note, “**” represents extremely significant correlation at level of p < 0.01. “*” represents significant correlation at level of p < 0.05.

). Therefore, in PC2, both pedon thickness and total phosphorus were selected into the MDS.

The preliminary indices in PC3 with an absolute loading value greater than 0.50 were total potassium, pedon thickness, and organic matter. Among them, total potassium had the largest Norm value of 3.30, which fell within its 10% range. However, only the correlation coefficient of organic matter was less than 0.50. Therefore, organic matter was selected for inclusion in the MDS in PC3.

In PC4, the preliminary indices with an absolute loading value greater than 0.50 were total potassium and pH. Among them, total potassium had the largest Norm value of 3.3, but the Norm value of pH did not fall within 10% of the Norm value of total potassium. Therefore, only total potassium was selected for inclusion in the MDS in PC4.

In PC5, the only preliminary index with an absolute loading value greater than 0.50 was total nitrogen, which was directly selected for inclusion in the MDS.

The indices finally selected for inclusion in the MDS for the first group were sand content, total phosphorus, pedon thickness, total potassium, organic matter, and total nitrogen.

From the analysis of the second set of data in

Table 7. Explanation of the total variance of the primary selection index of the second data set.

Indicator Name	Characteristic Root	Variance Interpretation Rate %	Accumulation %	Characteristic Root	Variance Interpretation Rate %	Accumulation %
Black soil thickness	2.967	29.67	29.67	2.967	29.67	29.67
Pedon thickness	2.482	24.819	54.489	2.482	24.819	54.489
Sand content	1.533	15.334	69.823	1.533	15.334	69.823
Silt content	0.936	9.356	79.179
Clay content	0.883	8.834	88.013
pH	0.466	4.657	92.669
Organic matter	0.442	4.424	97.093
Total nitrogen	0.239	2.394	99.487
Total phosphorus	0.047	0.472	99.959
Total potassium	0.004	0.041	100

and

Table 8. Loadings priming indices on each principal component of the second data set.

Name	Principal Component
	PC1	PC2	PC3
Black soil thickness	0.025	−0.152	−0.16
Pedon thickness	0.056	−0.176	0.168
Sand content	−0.21	0.292	0
Silt content	0.231	−0.221	−0.136
Clay content	0.114	−0.299	0.244
pH	−0.163	−0.019	0.467
Organic matter	0.141	0.117	−0.441
Total nitrogen	0.251	0.205	0.21
Total phosphorus	0.255	0.198	0.214
Total potassium	0.219	0.174	0.133

, three main components were obtained, with a cumulative contribution rate of 69.82% and a significance level of less than 0.05. All of them were suitable for the use of the principal component analysis method.

Although the contribution rates of the principal components with eigenvalues greater than 1 in the preliminary soil evaluation indices of the two groups met the requirements for extracting information, there were differences in the explanatory power of each principal component index on the total variance.

In the second group, there were three main components (Table 8). In PC1, the preliminary indices with an absolute loading value greater than 0.50 were total phosphorus, total nitrogen, silt content, and total potassium. Among them, total phosphorus had the largest Norm value of 4.96 (

Table 9. Norm values for the second data set of primary indices.

Name	Total Phosphorus	Total Nitrogen	Sand Content	Silt Content	Clay Content	Total Potassium	pH	Organic Matter	Pedon Thickness	Black Soil Thickness
Norm	4.96	4.95	4.95	4.7	4.38	4.22	3.86	3.78	2.56	2.15

). Total nitrogen and silt content fell within 10% of the Norm value of total phosphorus. However, the correlation coefficient between total phosphorus and total nitrogen was greater than 0.50 (

Table 10. Results of correlation analysis of the second data set.

Name	Black Soil Thickness	Pedon Thickness	Sand Content	Silt Sand	Clay Sand	pH	Organic Matter	Total Nitrogen	Total Phosphorus	Total Potassium
Black soil thickness	1
Pedon thickness	0.107	1
Sand content	−0.199 *	−0.269 **	1
Silt content	0.148	0.229 *	−0.839 **	1
Clay content	0.193	0.267 **	−0.788 **	0.464 **	1
pH	−0.133	−0.069	0.194	−0.434 **	0.191	1
Organic matter	0.021	−0.252*	−0.111	0.189	−0.217 *	−0.511 **	1
Total nitrogen	−0.085	−0.026	−0.081	0.15	−0.02	−0.146	0.229 *	1
Total phosphorus	−0.094	0.007	−0.095	0.165	−0.004	−0.154	0.226 *	0.995 **	1
Total potassium	−0.157	−0.018	−0.079	0.146	−0.002	−0.207 *	0.231 *	0.633 **	0.631 **	1

Note, **, represents extremely significant correlation at level of p < 0.01. *, represents significant correlation at level of p < 0.05.

), while the correlation coefficient of silt content was less than 0.50. Therefore, in PC1, total phosphorus and silt content were selected for inclusion in the MDS.

In PC2, the preliminary indices with an absolute loading value greater than 0.50 were total nitrogen, silt content, clay content, and sand content. Among them, total nitrogen had the largest Norm value of 4.96. Sand content and silt content fell within 10% of the Norm value of total nitrogen, and the correlation coefficients of both indices with total nitrogen were less than 0.50. Therefore, total nitrogen, sand content, and silt content were all selected for inclusion in the MDS.

In PC3, the indices with an absolute loading value greater than 0.50 were pH and organic matter. Among them, pH had a relatively high Norm value of 3.86, and organic matter fell within 10% of that. However, due to the correlation coefficient between pH and organic matter being greater than 0.50, only pH was selected for inclusion in the MDS in PC3.

In summary, for the second group, the indices finally selected were total phosphorus, total nitrogen, silt content, sand content, and pH.

Therefore, the final selected indices for evaluating soil health were sand content, silt content, pedon thickness, organic matter, total potassium, total nitrogen, total phosphorus, and pH.

3.2. Soil Health Evaluation

3.2.1. Determining Membership Function

Based on the different impact effects of MDS indices on soil, the indices were classified as follows. Pedon thickness, organic matter, total potassium, and total phosphorus belonged to the ascending membership function. The pH, sand content, and silt content belonged to the Peak-shaped membership function. Since there were no indices in MDS that had a negative correlation with soil, the descending membership function was not considered.

3.2.2. Calculating Index Weight

The impacts of different indices on soil varied, and their respective weights also differed. To avoid the impact of multicollinearity among indices, the PCA was used for dimensionality reduction. The weights were determined by dividing the communality variance of each item by the total communality variance. The table below showed the commonality variances and weights for the total data set and the minimum data set (

Table 11. Common factor variances and weights for the total and minimum data sets.

Name	Total Data Set		Minimum Data Set
	Common Factor Variance	Weights	Common Factor Variance	Weights
Pedon thickness	0.769	4.05%	0.521	9.97%
Sand content	0.94	13.06%	0.851	16.29%
Silt content	0.714	11.66%	0.87	16.65%
Organic matter	0.737	10.29%	0.63	12.06%
Total nitrogen	0.848	13.16%	0.818	15.66%
Total potassium	0.793	9.34%	0.603	11.54%
Total phosphorus	0.75	13.19%	0.61	11.68%
pH	0.495	10.73%	0.322	6.16%
Black soil thickness	0.645	2.98%
Clay content	0.762	11.53%
Bulk density	0.655	8.78%

).

The ranking of weights in MDS was in order of silt content, sand content, total nitrogen, organic matter, total phosphorus, total potassium, pedon thickness, and pH. In TDS, the indices with greater weights were total phosphorus, total nitrogen, silt content, sand content, and clay content. Through comparative analysis, it could be seen that silt content, sand content, and total nitrogen have relatively greater impacts on soil health.

3.2.3. Coefficient of Variation for Soil Health Evaluation Index

The MDS included eight indices, among which total phosphorus was a strongly sensitive index, total nitrogen and organic matter were moderately sensitive indices, while pedon thickness, sand content, silt content, and pH were low-sensitivity indices (

Table 12. Standard deviation, mean, and coefficient of variation for MDS.

Name	Pedon Thickness	Sand Content	Silt Content	Organic Matter	Total Nitrogen	Total Potassium	Total Phosphorus	pH
SD	48.73	17.68	11.6	12.84	0.45	9.41	0.89	0.86
Mean	123.17	42.93	33.68	22.09	0.88	21.69	0.91	6.39
CV	39.56%	41.19%	34.43%	58.13%	50.79%	43.38%	97.50%	13.41%

).

3.2.4. Reasonableness Validation of the Minimum Data Set

Based on the Soil Health Index (SHI) calculation formula, the weights and membership grade of the MDS and TDS indices were substituted respectively to calculate the SHI for the TDS and the MDS. A correlation analysis was performed on the results of the two calculations (Figure 4), and the results showed that R² was 0.7789 (p < 0.001), indicating a significant correlation between the MDS and the TDS. This suggested that the indices in the MDS could adequately represent the TDS for evaluating the health of phaeozem.

3.2.5. Soil Health Index and Grading System

By applying the weights and degrees of membership data of each index into the SHI calculation formula, the SHI was obtained. For TDS, the SHI ranged from 0.40 to 0.81, with an average value of 0.60 and a CV of 10.30%. For MDS, the SHI ranged from 0.33 to 0.91, with an average value of 0.62 and a CV of 12.03%.

Based on the Cornell Soil Health Assessment System, the phaeozem health index (SHI) was graded, and the results showed that an SHI less than 0.6 indicated unhealthy soil; an SHI between 0.6 and 0.8 indicated sub-healthy soil; and an SHI greater than or equal to 0.8 indicated healthy soil [31].

3.3. Soil Health Evaluations

The SHI of the phaeozem in Liaoning Province ranged from 0.33 to 0.91, with an average value of 0.62 and a CV of 12.03% (Figure 5), which means that the soil health status was predominantly in a sub-healthy condition (SHI < 0.6 = unhealthy, SHI 0.6–0.8 = sub-healthy, SHI ≥ 0.8 = healthy).

4. Discussion

Soil health status was influenced by many factors [34]. Comprehensive analysis and diagnosis of soil health status through numerous soil indices allowed for a more systematic and comprehensive reflection of soil health [35]. The more indices selected, the more accurately soil health status could be reflected [36,37,38]. However, in practical applications, it was necessary to consider the cost involved in the selection of indices [39]. Having an excessive amount of data could also lead to complexities in analysis and calculations [37,38]. Additionally, the interaction between soil indices could result in data redundancy [6,26]. Therefore, it was necessary to accurately reflect the soil health status with the least indices [37].

4.1. Minimum Data Set Indices

In this study, pedon thickness, sand content, silt content, organic matter, total nitrogen, total phosphorus, total potassium, and pH were composed into the minimum data set to calculate the health level of phaeozem in Liaoning Province, and the selection of these indices is consistent with the results of existing studies, e.g., [40,41,42].

Soil physical properties can reflect the stability of soil structure and directly affect the fertility status of the soil [43]. In this study, the coefficients of variation of pedon thickness, sand content and silt content were 39.56%, 42.93%, and 34.43%, respectively (Table 12). It indicates that the variation between different regions is large. The reason for this phenomenon may be due to the production of different climates and different parent materials [24] or the differences in specific indices related to soil structure and function due to anthropogenic activities such as crop rotation, straw return to the field, and the application of organic fertilizers [42].

For the chemical indices in the MDS, the mean value of pH was 6.39, indicating an acidity trend. This may be due to the high degree of soil weathering in the study area as a result of the dry and windy winter and spring seasons, as well as the pronounced aluminum enrichment of the soil due to the same rain and heat seasons, and the irrational fertilization practices [44]. The CV for organic matter, total nitrogen, total phosphorus and total potassium were 58.13%, 50.79%, 43.38%, and 97.50%, respectively, which are highly variable and may be due to different anthropogenic activities in different areas [42,45].

4.2. The Soil Health Status of the Phaeozem in Liaoning Province

The sub-healthy areas are mainly located in the northeastern part of Liaoning, the Liaohe Plain and the southwestern part of Liaoning. The northeastern Liaoning and southwestern Liaoning areas are mainly hilly and mountainous, and the type of cultivated land is mainly sloping cultivated land [22,24]. Concentrated rainfall in summer is prone to surface runoff erosion, which in turn degrades the physicochemical properties of the soil and ultimately leads to a decline in soil health; at the same time, in winter and spring, the ground cover is reduced due to harvesting, and sand lifting produced by gusty winds also erodes the soil, which in turn degrades the soil health [22,24]. The Liaohe Plain region is the main cultivation area, which is full of river bends and sandbars, with harbors and branches, significant accumulations, and constantly elevated river beds. The summer flood season may lead to poor drainage or riverbank failure, resulting in flooding and a decline in soil health. Similarly, the winter and spring seasons, when there is little ground cover, produce large amounts of sand lifting and wind erosion, leading to a decline in soil health [22,24].

Most of the unhealthy areas were located in the south and center of Liaoning (Figure 5). The ecology of the region was exceptionally fragile, primarily because of its hilly and coastal topography [24]. This distinct terrain had resulted in several environmental issues, specifically inadequate agricultural hydrological conditions, elevated levels of soil erosion and salinization, as well as reduced soil capacity to retain water and fertilizer [22,23].

Meanwhile, the phaeozem in the Liaoning region is dominated by cultivated land [22,23,24]. Another fundamental reason that makes the soils in the phaeozem as a whole show different health status is anthropogenic activities. This was likely due to the long-term adoption of monotonous farming practices and improper use of chemical fertilizers and pesticides, which had led to a decrease in organic matter content and soil compaction in some areas of the phaeozem, thereby hindering its healthy development. However, with increasing attention towards the protection of phaeozem, efforts had been made to improve agricultural management practices. People were actively tackling issues such as soil compaction and promoting crop rotation, which helped balance soil nutrients, improved the health of phaeozem, and positively impacted agricultural economics. Nonetheless, soil erosion has remained a significant problem in practice, requiring further measures for improvement.

In this study, data from 140 sites were collected, and eight physical and chemical indices were utilized to assess the soil health status of phaeozem in Liaoning Province. There might be some deviations; thus, it is necessary to further refine and optimize the assessment system in future extensive research.

5. Conclusions

The pedon thickness, sand content, silt content, organic matter, total nitrogen, total phosphorus, total potassium, and pH were chosen as the essential indices within the MDS to evaluate the health status of the phaeozem in Liaoning Province. This selection aligns with previous research findings and effectively mirrors the soil’s health condition.

The substantial coefficients of variation observed for pedon thickness, sand content, and silt content across regions suggest notable differences, likely influenced by climatic variations, parent material differences, and anthropogenic activities.

The study area’s average soil pH of 6.39 indicates an overall acidic nature, potentially attributed to intense soil weathering and suboptimal fertilization practices. Additionally, high coefficients of variation for organic matter, total nitrogen, total phosphorus, and total potassium underscore the influence of regional anthropogenic activities on soil chemistry.

Sub-Healthy Status of phaeozem in Liaoning Province: The soil health index (SHI) within the range of 0.33 to 0.91, with an average of 0.62, indicates that the majority of phaeozem in Liaoning Province are in a sub-healthy state. These sub-healthy areas are primarily located in hills, mountains, and the Liaohe Plain, significantly impacted by rainfall and wind erosion. Meanwhile, unhealthy areas are concentrated in the south and center of the province, characterized by fragile ecological environments and challenging agricultural conditions.