Long-Term Effects of Crop Treatments and Fertilization on Soil Stability and Nutrient Dynamics in the Loess Plateau: Implications for Soil Health and Productivity

Abstract

Soil degradation and erosion pose significant threats to agricultural sustainability in fragile ecosystems, such as the Loess Plateau in northern China. This study examines the long-term impacts of fertilization regimes and land-use systems on soil health, focusing on soil aggregate stability, fertility, and crop productivity. Six treatment combinations were evaluated in our study, including three continuous alfalfa fields (AL-CK, AL-P, and AL-NPM) and three continuous wheat fields (WH-NPM, WH-NP, and WH-P), each representing a combination of land use and three fertilization treatments: (1) no fertilization (CK), (2) inorganic fertilization (120 kg ha⁻¹ N, 60 kg ha⁻¹ P-NP), and (3) a combination of organic and inorganic fertilization (75 t ha⁻¹ cow manure-NPM). Soil samples were collected from three depths (0–10 cm, 10–20 cm, and 20–30 cm) to assess physical and chemical properties. We evaluated the long-term effects of different fertilization treatments on soil stability, fertility, and crop yield to explore the interactions among soil’s physical and chemical properties under two land-use types and to assess the effectiveness of combined organic and inorganic fertilization strategies in improving soil health and mitigating erosion in vulnerable landscapes. The study revealed significant depth-specific variations with surface layers (0–10 cm) showing the greatest improvement under NPM treatments, particularly in continuous alfalfa fields, which exhibited higher soil fertility, improved soil structure, and crop yield. In contrast, continuous wheat fields with minimal fertilization demonstrated significantly lower soil quality and productivity. Using the combination of mineral fertilizers and organic amendments, such as cow manure, proved to be the most effective strategy for significantly enhancing nutrient availability and overall soil health. Partial Least Squares Modeling (PLS-M) and Mantel analysis highlighted the critical role of fertilization management in maintaining soil quality, boosting crop productivity, and mitigating erosion in high-risk areas. This study emphasizes the importance of integrated nutrient management for sustainable land use and soil conservation in erosion-prone regions.

1. Introduction

The Loess Plateau in northern China faces significant ecological challenges, primarily driven by soil erosion and degradation [1,2]. This region is characterized by friable, silty loess soil, which is highly susceptible to erosion due to its fragile structure and low resistance to external forces [3]. Over the centuries, human activities such as deforestation, improper agricultural practices, and urbanization have led to the degradation of topsoil, diminishing fertility and reducing soil productivity [4,5,6]. As a result, soil erosion not only compromises crop productivity but also exacerbates water-related issues, such as flooding and reduced water retention capacity [1]. These ongoing ecological and agricultural challenges underscore the urgent need for effective strategies to combat soil degradation and restore soil health.

In response to these issues, a variety of soil conservation strategies have been proposed. Practices such as conservation tillage, integrated soil management, and favorable public policies are essential for preventing soil erosion and promoting sustainable land use [7]. Conservation tillage techniques, including minimum tillage, crop rotation, and cover cropping, help retain topsoil and restore fertility [8]. Additionally, fertilization strategies—particularly the integration of organic and inorganic amendments—improve soil fertility, aggregate stability, and microbial activity, thereby mitigating soil degradation and supporting sustainable agriculture in the region [9,10,11].

The use of organic inputs, combined with precision agriculture methods, has further enhanced soil health and fertility [12]. Organic amendments, such as cow manure, when integrated with mineral fertilizers, improve soil physicochemical properties, increase nutrient availability, and boost microbial activity [13,14,15,16]. Furthermore, the role of plant roots—particularly those of perennial plants—is critical for soil stability. Roots help reinforce soil structure, enhance water infiltration, and reduce runoff and erosion [14,15]. They also contribute to the formation of soil organic matter, improving soil cohesiveness and fertility [16].

Soil health is a key component of sustainable agriculture, defined as the soil’s capacity to function as a living ecosystem [17]. It encompasses essential attributes such as soil fertility, microbial diversity, water regulation, nutrient cycling, and biodiversity maintenance. Healthy soils not only support plant and animal life but also enhance crop growth, improve resilience to environmental stressors like drought, and help soils recover from disturbances [18,19]. These qualities are vital for sustaining long-term agricultural productivity and mitigating land degradation.

Despite significant research on soil conservation practices, there remains a knowledge gap in the comparative and systematic assessment of land-use types (LUTs) and fertilization treatments on soil properties specific to the Loess Plateau. While previous studies have emphasized the need to evaluate the effects of different fertilization regimes on soil health and sustainability, there has been limited attention to systematically comparing their impacts across various LUTs [20,21,22]. Furthermore, there is a growing understanding that sustainable land management strategies must also take into account economic feasibility and environmental impacts to ensure their long-term viability and practical application in regions facing similar challenges.

To address this gap, the present study investigates the influence of six distinct land-use types on soil health and fertility under different fertilization treatments. These land-use types include three continuous alfalfa fields (AL-CK, AL-P, and AL-NPM), which are long-term monoculture systems dedicated to alfalfa cultivation, and three continuous wheat fields (WH-NPM, WH-NP, and WH-P), representing long-term monoculture systems for wheat production. Each field was subjected to various fertilization treatments, allowing for a comparative analysis of their impacts on soil properties. The primary objective of this study is to provide new insights into sustainable crop production and soil management in the Loess Plateau by systematically evaluating the effects of different fertilization strategies on soil properties. The findings are expected to contribute to the development of targeted soil conservation practices and inform policies aimed at mitigating soil erosion, restoring soil health, and enhancing land productivity in vulnerable regions.

2. Materials and Methods

2.1. Study Area

The experiment was conducted in Shilipu Village at the Changwu State Key Agro-Ecological Experimental Station in Shaanxi Province, established in 1984. The station is located 2.5 km west of Changwu County, Xianyang City, Shaanxi Province, P.R. China, with geographic coordinates of longitude 107°40′30″–107°42′30″ E and latitude 35°05′12″–35°16′00″ N. The study field is situated on a flat, expansive plateau characterized by deep loess soils, which are derived from loess deposits and classified as Calcic Cambisols according to the Food and Agriculture Organization (FAO) soil classification system [23]. The continuous wheat plots in the experiment were managed using conventional tillage practices, which involved plowing to a depth of approximately 20 cm, followed by leveling. Changwu County covers an area of 525 km², with altitudes ranging from 946 to 1226 m above sea level. The region experiences an annual mean rainfall of 582 mm and an average temperature of 9.2 °C, with cold, dry winters and hot, humid summers. These climatic and soil characteristics make the area particularly vulnerable to erosion and degradation, underscoring its significance for studying soil conservation practices.

2.2. The Geographical Map of Study Site of Changwu County Loess Plateau

The map (Figure 1) illustrates the Loess Plateau’s topography, including elevation gradients, stream networks, and meteorological and hydrological stations, with Changwu County highlighted as the study site. The bottom panel depicts the experimental plots, showing continuous alfalfa (AL) and wheat (WH) fields under different fertilization treatments (CK, NP, P, NPK).

2.3. Soil Sampling and Fertilization Under Different Land Use

This study assessed the impacts of land-use types (LUTs) and fertilization regimes on soil properties in the Loess Plateau, northern China. The experimental design incorporated two cropping systems: continuous alfalfa fields (AL) and continuous wheat fields (WH), each combined with three fertilization treatments: CK (Control) (no fertilization), NP (inorganic fertilization with 120 kg ha⁻¹ nitrogen and 60 kg ha⁻¹ phosphorus), and NPK (integrated organic-inorganic fertilization with 120 kg ha⁻¹ nitrogen, 60 kg ha⁻¹ phosphorus, and 75 t ha⁻¹ cow manure). These treatments resulted in six combinations: AL-CK, AL-P, AL-NPK, WH-NPK, WH-NP, and WH-P. Each treatment was replicated in three independent plots, resulting in a total of 18 experimental units. Soil samples were systematically collected from each plot at three soil depths (0–10 cm, 10–20 cm, and 20–30 cm) using a stratified random sampling method to ensure representativeness and account for spatial variability. A total of 54 soil samples were obtained for a comprehensive analysis of soil physical and chemical properties. This experimental design, with true replication across plots, provided robust and representative data for evaluating the effects of land use and fertilization regimes on soil health in an erosion-prone agroecosystem.

Soil particle distribution, including sand, silt, and clay, was determined using an LSTM200 laser particle analyzer (Beckman Coulter, Krefeld, Germany), a laser-based technology that analyzes particles without the risk of missing either the largest or smallest particles in the sample. Soil organic carbon (SOC) was assessed using standardized procedures, specifically the Walkley–Black method with dichromate oxidation. The total nitrogen (TN) was measured via the Kjeldahl digestion method followed by colorimetric determination, and the total phosphorus (TP) was determined through acid digestion followed by spectrophotometric analysis. The soil properties analyzed included physical characteristics, such as bulk density, dry-stable aggregates (DSA), and water-stable aggregates (WSA), as well as chemical properties, including the total nitrogen (TN), the total phosphorus (TP), and soil organic carbon (SOC). Focusing on these properties provided valuable insights into soil stability and fertility under different management practices.

2.4. Bulk Density

Bulk density was measured using the core method. Soil samples were collected with cylindrical metal cores of known volume, which were inserted vertically into the soil to extract undisturbed samples. Excess soil was trimmed to ensure the sample precisely matched the core’s volume. The samples were then oven-dried at 105 °C for 24 h to remove all moisture, and the dry weight was recorded. Bulk density was calculated as the ratio of the dry soil weight to the core volume (g/cm³).

2.5. Root Biomass

Root biomass was determined using the core method described by Böhm (1979) [25]. Undisturbed soil cores were collected with an 8 cm diameter corer from each treatment plot at three depths (0–10 cm, 10–20 cm, and 20–30 cm) using a stratified random sampling approach to ensure representativeness. In the laboratory, roots were separated by washing the soil through a 0.25 mm sieve under running water, and fine roots were carefully handpicked. The roots were then oven-dried at 60 °C for 48 h until a constant weight was achieved and subsequently weighed using a precision balance (0.001 g accuracy).

2.6. Sampling Procedures Used in the Determination of Aggregate Size Fractions

Soil samples were collected from undisturbed soil using a knife, carefully cutting them into small pieces approximately 1 cm in size along natural fractures, and allowing them to air dry. The mass of soil aggregates across different size distributions (<0.25 mm, 0.25–0.5 mm, 0.5–1 mm, 1–2 mm, 2–5 mm, and >5 mm) was determined using wet and dry sieving methods [26,27]. The stability characteristics of the aggregates were analyzed by assessing the aggregate content within the 0.25–5 mm range (R_0.25–5). The formulas for calculating each stability index are as follows:

$$m_i=M_i/M_T\times 100\%$$

(1)

$$R_{0.25\text{–}5}=M_{r>0.25}-M_{r>5}/M_T=1-M_{r>5}-M_{r<0.25}/M_T$$

(2)

$$\mathrm{MWD}={\sum }_{i=1}^n\left(\overline{R_i}{m}_i\right)/{\sum }_{i=1}^n{m}_i$$

(3)

$$\mathrm{GMD}=exp\left\{{\sum }_{i=1}^n{m}_i\mathit{ln}{R}_i\right\}$$

(4)

$${\displaystyle \frac{M\left(r<\overline{R_i}\right)}{M_T}}={\left({\displaystyle \frac{\overline{R_i}}{R_{max}}}\right)}^{3-D}$$

(5)

$$\lg \left[{\displaystyle \frac{M\left(r<\overline{R_i}\right)}{M_T}}\right]=\left(3-D\right)\lg \left({\displaystyle \frac{\overline{R_i}}{R_{max}}}\right)$$

(6)

where m_i is the percentage of the mass of the i-level aggregate, M_i is the mass of the i-level water-stable aggregate, M_T is the total mass of each particle-level aggregate, and R_i is the i-level aggregate. As the average diameter of R, R_max is the maximum particle size of water-stable agglomerates, M(r < R_i) is the mass of agglomerates with a particle size less than R_i, and R_0.25–5 is 0.25~5 mm agglomerates content.

2.7. Statistical Analysis

To analyze the effects of land use and fertilization treatments on soil properties, a one-way analysis of variance (ANOVA) was conducted using SPSS 13.0 (SPSS Inc., Chicago, IL, USA). The significance of differences between treatments was tested at the p ≤ 0.05 and p ≤ 0.01 levels. Pearson correlation analysis was used to examine relationships between soil properties, including soil organic carbon (SOC), total phosphorus (TP), total nitrogen (TN), water-stable aggregates (WSA), mean weight diameter (MWD), and geometric mean diameter (GMD). Partial Least Squares Modeling (PLS-M) was performed to understand the direct and indirect effects of land use, fertilization, and soil depth on soil properties. The PLS models included latent variables for physical properties (e.g., bulk density and WSA) and chemical properties (e.g., TN, TP, and SOC). The data from PLS analysis were used to quantify the contributions of different factors to soil quality and aggregate stability.

3. Results

3.1. Soil Bulk Density, Root Biomass, and Biological Yield in Alfalfa and Grain Yield in Wheat Fields

The results revealed that continuous alfalfa fields (AL-CK, AL-P, and AL-NPM) had lower soil bulk densities and higher root biomass compared to continuous wheat fields (WH-NPM, WH-NP, and WH-P) at the 0–10 cm and 10–20 cm depth intervals. However, these differences were less pronounced at the 20–30 cm depth interval (

Table 1. Different treatments affect soil bulk density and root biomass.

	Land Use	Bulk Density	Root Biomass
		(g cm−3)	mg cm−3
Soil Layers (cm)
0–10	AL-CK	1.15 ± 0.01 bc	0.11 ± 1.07 ab
	AL-P	1.09 ± 0.02 c	0.09 ± 0.98 ab
	AL-NPM	1.10 ± 0.06 c	0.31 ± 1.09 a
	WH-NPM	1.16 ± 0.06 b	0.04 ± 0.37 c
	WH-NP	1.18 ± 0.02 ab	0.03 ± 0.28 c
	WH-P	1.20 ± 0.03 a	0.02 ± 0.21 c
10–20	AL-CK	1.32 ± 0.03 a	0.02 ± 0.20 c
	AL-P	1.33 ± 0.03 a	0.06 ± 0.25 b
	AL-NPM	1.24 ± 0.02 c	0.09 ± 0.30 a
	WH-NPM	1.30 ± 0.04 b	0.07 ± 0.69 ab
	WH-NP	1.32 ± 0.00 a	0.03 ± 0.31 c
	WH-P	1.30 ± 0.01 b	0.02 ± 0.19 c
20–30	AL-CK	1.36 ± 0.06 bc	0.04 ± 0.60 a
	AL-P	1.44 ± 0.03 b	0.03 ± 0.11 b
	AL-NPM	1.37 ± 0.02 bc	0.05 ± 0.16 ab
	WH-NPM	1.35 ± 0.03 c	0.02 ± 0.21 c
	WH-NP	1.47 ± 0.00 a	0.01 ± 0.12 c
	WH-P	1.45 ± 0.04 ab	0.01 ± 0.07 c

Note: Bulk density (g cm⁻³) and root biomass (mg cm⁻³) across different land-use treatments and soil layers. Values represent mean ± standard deviation (SD). Statistical significance was determined using one-way ANOVA followed by Tukey’s HSD post hoc test (p < 0.05). Treatments within the same soil layer and parameter (bulk density or root biomass) that do not share the same letter (e.g., “a”, “b”, “c”) are significantly different.

). In the 0–10 cm layer, the bulk density for AL-CK, AL-P, and AL-NPM ranged from 1.09 g cm⁻³ to 1.15 g cm⁻³, whereas for WH-NPM, WH-NP, and WH-P, it ranged from 1.16 g cm⁻³ to 1.20 g cm⁻³. Root biomass values in the AL fields ranged from 0.09 to 0.31 mg cm⁻³, significantly higher than those in the WH fields, where root biomass ranged from 0.02 to 0.04 mg cm⁻³ at the end of the incubation period. These trends were consistent across the other depth intervals, with AL fields exhibiting lower bulk density and higher root biomass at each depth. Notably, the best results for both bulk density and root biomass were recorded under the nitrogen, phosphorus, and manure (NPM) treatment, which most effectively improved these soil properties. The nitrogen and phosphorus (NP) treatment also produced positive outcomes, while the control treatment (CK, no fertilization) exhibited the least desirable results.

The experiment also demonstrated distinct yield patterns for alfalfa (forage crops) and wheat (grain crops) in response to the various fertilization treatments, highlighting their contrasting growth characteristics and nutrient requirements (

Table 2. The table presents the biological yield for alfalfa and the grain yield for wheat (kg/ha).

Plot	Crop Type	Total Yield (kg/ha)
AL-CK	Alfalfa	24,555 b
AL-P	Alfalfa	33,240 a
AL-NPM	Alfalfa	34,900 a
WH-NPM	Wheat	5094.75 c
WH-NP	Wheat	3924.75 d
WH-P	Wheat	1055 e

Note: Total yield (kg/ha) across different treatments and crop types. Statistical significance was determined using one-way ANOVA followed by Tukey’s HSD post hoc test (p < 0.05). Treatments that do not share letters (e.g., “a”, “b”, “c”) are significantly different.

). Alfalfa fields consistently produced higher forage dry matter yields compared to the grain yields of wheat fields across all fertilization treatments. Among the alfalfa treatments, the highest yield was observed in the AL-NPM plot (nitrogen, phosphorus, and manure), with forage dry matter yields ranging from 34,880 to 34,967 kg ha⁻¹, illustrating the synergistic effect of combined organic and inorganic fertilization in promoting biomass production. The AL-P treatment (phosphorus alone) also resulted in substantial alfalfa yields, with dry matter yields recorded at 32,235, 32,353, and 32,547 kg ha⁻¹ across replicates, emphasizing the critical role of phosphorus in enhancing alfalfa productivity. In contrast, wheat fields exhibited lower grain yields, with the highest yield observed in the WH-NPM plot (nitrogen, phosphorus, and manure), ranging from 4981 to 5087 kg ha⁻¹. This indicates that while wheat benefited from the combined application of organic and inorganic fertilizers, its productivity potential remained lower than that of alfalfa, likely due to inherent differences in crop physiology and growth patterns. These findings underscore the suitability of alfalfa for producing higher biomass under fertilized conditions, whereas wheat showed more limited productivity, reflecting the contrasting growth requirements and responses to fertilization between the two crops.

3.2. Soil Nutrient Concentrations

The plot receiving the nitrogen, phosphorus, and manure (NPM) treatment showed the highest total phosphorus (TP) concentration at the 0–10 cm soil depth. Alfalfa (AL-NPM) had the highest TP content at 0.98 mg/kg, followed closely by wheat (WH-NPM) at 0.95 mg/kg at the same depth. In soil depth within treatments, AL-NPM exhibited significantly higher TP at 0–10 cm compared to 10–20 cm and 20–30 cm. In contrast, WH-NPM showed no significant differences across depths, with all layers. Similar patterns were observed for AL-P, where TP at 0–10 cm was significantly higher than at 10–20 cm and 20–30 cm, and for AL-CK, where TP was lower at 0–10 cm compared to deeper layers. TP generally decreased with depth across treatments, except in WH-NP and WH-P, where no significant differences were observed (Figure 2). In Figure 2, the AL treatments were significantly different from the WH treatments. The NPM treatment resulted in the highest total nitrogen (TN) concentration in both alfalfa and wheat fields at the 0–10 cm depth. Among the alfalfa plots, AL-NPM had the highest TN content, averaging 2.12 g/kg, which was significantly higher at the 0–10 cm depth compared to 10–20 cm and 20–30 cm. AL-P followed a similar trend, with TN at 0–10 cm significantly higher than at deeper depths (10–20 cm). In contrast, WH-NPM showed no significant differences across depths, with all depths. Similarly, TN levels in WH-P and WH-NP showed no significant differences across depths, with all depths. In all plots, TN content generally decreased with increasing soil depth (Figure 3). At the 0–10 cm depth, the NPM treatment led to the highest concentrations of soil organic carbon (SOC) in both alfalfa and wheat fields. In soil depths within treatments, AL-NPM had the highest SOC levels at 0–10 cm, which were significantly higher than at 20–30 cm, while no significant difference was observed between 0–10 cm and 10–20 cm. AL-P followed a similar trend, with SOC at 0–10 cm being significantly higher than at deeper depths, with both 10–20 cm and 20–30 cm. Among the wheat treatments, WH-NPM demonstrated the highest SOC at 10.92 g/kg. SOC at 0–10 cm was not significantly different to 10–20 cm, but it was significantly higher than at 20–30 cm (Figure 4).

The significant increase in SOC, along with TN and TP, can be attributed to enhanced microbial activity stimulated by organic manure. The slow-release nature of organic manure further contributes to improved nutrient availability and cycling, thereby enhancing soil fertility and organic carbon accumulation in the upper soil layers. Generally, SOC, TP, and TN concentrations exhibited a decreasing trend with increasing soil depth. This pattern is primarily attributed to the reduced incorporation of organic matter, limited root penetration, and diminished microbial activity in deeper soil layers. Such conditions restrict the accumulation and cycling of nutrients, resulting in lower concentrations in subsurface soils.

3.3. Soil Aggregate Stability

This experiment highlighted the significant impacts of land use and fertilization on the stability of macro-aggregates in the soil. Continuous alfalfa fields (AL) exhibited higher water-stable aggregate (WSA) percentages compared to continuous wheat fields (WH). This improvement in soil aggregate stability is attributed to alfalfa’s extensive root system, which strengthens soil structure, and to organic amendments that enhance microbial activity and nutrient binding. AL-NPM exhibited the highest WSA values across all soil depths, with 0–10 cm significantly higher than 10–20 cm and 20–30 cm. In contrast, AL-CK, which received no fertilization, showed lower WSA values, with 0–10 cm significantly higher than 10–20 cm and 20–30 cm. Among the wheat treatments, WH-NPM exhibited moderate WSA values, with 0–10 cm and 10–20 cm not significantly different from each other, but both were significantly higher than 20–30 cm (Figure 5).

These results suggest that fertilization positively influences soil aggregate stability, particularly over time, likely due to the beneficial effects of the nutrient components in the fertilization treatment. The soil structure of annual wheat fields, especially those with minimal or no fertilization, was poor, with lower mean weight diameter (MWD) and geometric mean diameter (GMD), as shown in Figure 6 and Figure 7. Lower MWD and GMD values indicate a higher risk of erosion and a reduced capacity for water infiltration in the wheat fields. AL-NPM showed the highest MWD and GMD values across all soil depths, with the 0–10 cm layer being significantly higher than the 10–20 cm and 20–30 cm layers. In contrast, AL-CK, which received no fertilization, showed lower MWD and GMD values, with the 0–10 cm layer also significantly higher than the 10–20 cm and 20–30 cm layers. Among the wheat treatments, WH-P exhibited the lowest values at 0–10 cm, which gradually increased with depth, with all depths being significantly different from each other (Figure 6 and Figure 7). The fractal dimension (D) values, presented in Figure 8, which are inversely related to the stability of macro-aggregates, were significantly higher in wheat fields, indicating a less stable structure. This difference was particularly pronounced in the top 10 cm compared to the alfalfa fields. Among the wheat treatments, WH-NPM, WH-NP, and WH-P consistently exhibited the highest D values across all depths, with no significant differences observed between depths. In contrast, for alfalfa treatments, significant differences were observed across all depths within each treatment (Figure 8). Thus, the results of this study emphasize the importance of maintaining and improving soil aggregate stability to preserve and enhance soil quality in annual cropping systems.

3.4. Correlation Patterns and Partial Least Squares Modeling (PLS-M) Outputs

The correlation matrix in Figure 9 demonstrated significant relationships between soil organic carbon (SOC), water-stable aggregates (WSA), mean weight diameter (MWD), geometric mean diameter (GMD), root biomass (RB), and various aggregate fractions. The Structural Equation Modeling (SEM) analysis revealed that the effects of fertilizer application, soil depth, and treatment, along with their interactions, were significant in influencing the chemical properties, physical characteristics, and yield of the soil. The Partial Least Squares Model (PLS-M) illustrates the relationships among fertilizer, depth, and treatment, and their direct and indirect effects on yield through the mediating variables of physical and chemical soil properties (Figure 10). Fertilizer had the strongest and most significant direct positive effect on yield (4.654, p < 0.01), along with significant positive effects on chemical properties (12.12, p < 0.01) and a weaker, non-significant positive effect on physical properties (0.26). Soil depth had a small but significant positive direct effect on yield (0.35, p < 0.05), with weaker, non-significant effects on physical properties (0.06) and chemical properties (0.03). Treatment had a non-significant positive effect on yield (0.03) but a significant positive effect on physical properties (0.31, p < 0.01) and a significant negative effect on chemical properties (−0.54, p < 0.01). Among the mediating variables, physical properties exerted a strong and significant positive effect on yield (1.55, p < 0.001), while chemical properties had a negative impact on yield. These results identify fertilizer as the most critical factor influencing yield directly and through its significant effects on chemical properties. Depth primarily contributes through its weaker direct and indirect effects on yield. Treatment plays a dual role, positively influencing physical properties but negatively affecting chemical properties, which can ultimately reduce yield. This model emphasizes the interdependent contributions of soil’s physical and chemical characteristics in determining yield and underscores the importance of effective soil management practices.

4. Discussion

4.1. Yield Comparison Between Alfalfa and Wheat

The results of this study revealed that alfalfa consistently produced higher yields compared to wheat, particularly under the NPM (nitrogen, phosphorus, and manure) treatment. This finding aligns with previous research, such as [28,29], which reported that wheat yields decline significantly when fertilizer inputs are reduced, particularly under treatments like WH-NP and WH-P only. This highlights wheat’s greater sensitivity to nutrient availability and underscores the importance of balanced fertilization in maintaining its productivity. The higher performance of alfalfa, even under reduced fertilization, can be attributed to its robust root system, which enhances nutrient uptake and soil health. Alfalfa’s ability to fix atmospheric nitrogen through symbiotic relationships with Rhizobia bacteria further supports its high yields, particularly in phosphorus-enriched treatments like AL-P, where it achieved significant biomass production. In contrast, wheat showed a more pronounced dependence on combined organic and inorganic fertilization to achieve optimal yields. The maximum wheat yield observed in this study, under the WH-NPM treatment, emphasizes the role of manure in improving soil organic matter and nutrient availability, as supported by studies like [29], which demonstrated that organic amendments enhance soil structure and nutrient cycling. However, even under NPM, wheat yields remained considerably lower than those of alfalfa, reflecting the inherent physiological differences between the two crops and their responses to fertilization. These differences underscore the need for crop-specific fertilization strategies to maximize soil health and crop productivity. The lowest nutrient levels recorded in AL-CK emphasize the essential role of fertilizers in enhancing soil fertility

4.2. Nutrient Management and Soil Dynamics

The application of NPM significantly enhanced the levels of total nitrogen (TN), total phosphorus (TP), and soil organic carbon (SOC) in both AL-NPM and WH-NPM, particularly in the upper soil layers, as illustrated in Figure 2, Figure 3 and Figure 4. This increase in SOC can be attributed to enhanced microbial activity, particularly under the influence of organic amendments such as manure. The organic matter provided by manure serves as a substrate for soil microbes, leading to increased microbial biomass and the stabilization of organic matter through microbial processes [28]. Additionally, the high nitrogen levels in NPM may stimulate soil microbial communities, further accelerating the decomposition of organic residues and contributing to long-term SOC sequestration [27]. These results also align with previous studies by [30,31,32], which have demonstrated similar nutrient enhancements through nutrient management practices. However, the present study provides further insight into the spatial distribution of these nutrients, emphasizing their concentration in the upper soil layers. This finding is consistent with a well-established body of literature indicating that nutrient accumulation tends to be more pronounced in the topsoil, where organic matter and microbial activity are more concentrated [33,34,35,36]. The alfalfa treatments consistently exhibited higher SOC levels compared to the wheat treatments across all soil depths (Figure 4). This trend is consistent with findings from other studies in similar agro-ecosystems, particularly those conducted in the Loess Plateau region and other semi-arid areas, where alfalfa’s deep-rooting nature and nitrogen-fixing capability have been identified as key factors contributing to improved carbon sequestration and soil fertility [33,37]. These findings underscore the role of alfalfa as a beneficial crop for enhancing SOC, especially under practices involving organic amendments or higher mineral fertilizer inputs. In contrast, wheat, with its shallower root system and more limited capacity for nitrogen fixation, showed lower SOC levels and a stronger reliance on mineral fertilizers for nutrient uptake.

The observed decrease in TN, TP, and SOC with increasing soil depth (Figure 2, Figure 3 and Figure 4) further supports the idea that nutrient accumulation is concentrated in the upper soil layers, which are more directly influenced by fertilization and organic inputs. This trend mirrors findings from similar studies conducted in the Loess Plateau and other arid and semi-arid regions, where nutrient stratification in the soil is commonly observed due to fertilization practices and the differential root systems of crops [34,35,36]. The differences in fertilization sensitivity between alfalfa and wheat are due to their distinct root architectures and physiological traits. Alfalfa, a deep-rooted legume, benefits significantly from organic amendments, enhancing microbial activity, nitrogen fixation, and SOC accumulation. In contrast, wheat’s shallower roots are less efficient at utilizing organic inputs and rely more on mineral fertilizers [38,39]. This underscores the need for crop-specific nutrient management to optimize soil health and productivity. The low nutrient levels in the AL-CK (control) treatment highlight the crucial role of fertilization in maintaining soil fertility, particularly in regions like the Loess Plateau, where proper nutrient management is vital for combating soil depletion and ensuring agricultural sustainability.

4.3. Effects of Fertilization on Soil Aggregation and Stability

The results of this study (Figure 5) indicate that water-stable aggregate (WSA) percentages were higher in continuously sown alfalfa treatments compared to wheat treatments, consistent with previous studies [34,37]. Alfalfa treatments treated with nitrogen, phosphorus, and manure (NPM) exhibited the highest WSA values across all soil depths: 0–10 cm (69.33%), 10–20 cm (66.69%), and 20–30 cm (58.22%). In contrast, the unfertilized alfalfa treatment (AL-CK) showed lower WSA values: 0–10 cm (67.87%), 10–20 cm (58.73%), and 20–30 cm (50.87%). Wheat treatments exhibited more variability in WSA, with an increase in WSA at the 20–30 cm depth in the NP-treated plots. These results are consistent with findings from [35], which highlight the positive influence of fertilization on soil aggregate stability. Such improvements in soil structure and stability are particularly important for regions like the Loess Plateau, which are prone to soil erosion and degradation [40,41,42]. The observed biochemical effects in alfalfa, particularly with the addition of NPM, align with those reported by [35,36], who documented significant improvements in soil structure and aggregate stability.

In comparison, research on the Loess Plateau region has also shown that organic amendments, such as manure, enhance soil structure and promote sustainable land use in this fragile ecosystem [43,44,45,46,47]. The incorporation of root exudates from deep-rooted crops like alfalfa plays a critical role in this process. Alfalfa, which has symbiotic relationships with soil microorganisms, releases organic compounds through its roots. These root exudates contribute to soil aggregation by helping to bind soil particles together, forming stable aggregates that reduce soil erosion and enhance water retention. Furthermore, continuous wheat treatments had a higher fractal dimension (D) compared to alfalfa treatments, particularly in the 0–10 cm soil layer, indicating fewer stable aggregates (Figure 8). These findings suggest that alfalfa cultivation, especially when combined with organic-inorganic fertilization, can significantly enhance soil physical properties and structure. This benefit is particularly crucial in areas like the Loess Plateau, where erosion poses a significant challenge to sustainable farming. The results also underscore the importance of including alfalfa in crop rotations and adopting integrated fertilization strategies for long-term soil health and sustainable agricultural practices in vulnerable ecosystems.

4.4. Interpretation of Correlation Patterns and Partial Least Squares Modeling (PLS-M) Outputs

The correlation matrix (Figure 9) reveals significant positive correlations among most soil properties. Soil organic carbon (SOC) was positively correlated with mean weight diameter (MWD) and geometric mean diameter (GMD), suggesting that as SOC increases, the size and stability of soil aggregates improve, as previously recognized by [37]. The strong positive correlation between water-stable aggregates (WSA), SOC, MWD, and root biomass (RB) further highlights the pivotal role of SOC in enhancing soil structural stability and promoting root growth, as noted by [43,44,45,46]. These findings corroborate the existing literature, which emphasizes that SOC is fundamental to improving soil fertility, structure, and overall health, all of which are crucial for sustainable agricultural practices [46,47,48]. Our results also align with those of [49,50,51,52,53,54] in that alfalfa treatments exhibited lower bulk density and higher root biomass compared to annual crop treatments. This can be attributed to alfalfa’s deeper root systems, which are more effective in enhancing soil structure. In contrast, monoculture treatments, such as those commonly employed for crops like rice, may experience soil compaction, which limits the contribution of root biomass to soil structure [55,56].

The Partial Least Squares Modeling (PLS-M) analysis (Figure 10) highlights the critical role of integrated nutrient management in maintaining soil health and enhancing crop productivity. While the model primarily focuses on physical and chemical soil properties, the results suggest that biological processes, such as microbial activity and root biomass, could mediate the effects of fertilization and treatment on soil properties. These biological properties were inferred based on the observed relationships between physical and chemical properties, such as SOC and aggregate stability. The application of organic fertilizers, particularly manure, is thought to significantly enhance microbial activity, nutrient cycling, and soil aggregate stability, while improving soil structure. These inferred biological processes work synergistically with physical and chemical properties to sustain soil fertility and optimize crop yield. Although biological properties were not directly measured in this study, their potential influence is suggested by the positive relationships between root biomass, soil organic carbon, and aggregate stability. Deep-rooted legumes like alfalfa likely contribute to stabilizing soil aggregates, improving soil porosity, and enhancing water retention, which increases soil resilience against erosion. Integrated organic–inorganic fertilization provides a balanced nutrient supply, promoting microbial activity, organic matter decomposition, and nutrient availability, which further enhances soil structure and resilience. As noted by [56,57], the incorporation of organic fertilizers and the cultivation of deep-rooted crops like alfalfa are vital for improving soil surface properties, enhancing root systems, and optimizing nutrient dynamics, all of which are essential for the sustainability of agricultural systems.

5. Conclusions

This study demonstrates that continuous alfalfa cultivation improves soil properties and results in higher crop yields compared to continuous wheat, particularly in the upper soil layers. The application of nitrogen, phosphorus, and manure (NPM) significantly enhanced soil properties, including soil bulk density, root biomass, and nutrient concentrations, particularly total phosphorus (TP) and total nitrogen (TN). Moreover, NPM also increased soil organic carbon (SOC) levels, contributing to improved soil fertility. The stability of soil aggregates was higher in alfalfa treatments, reflecting the positive effects of continuous fertilization on soil structure. In contrast, wheat treatments exhibited lower aggregate stability, especially under minimal fertilization treatments, which increased their vulnerability to soil erosion and reduced water infiltration capacity. These findings underscore the importance of nutrient management and soil conservation practices in agricultural systems, emphasizing that integrated fertilization strategies, such as the combination of organic and inorganic inputs, are crucial for maintaining soil health and optimizing crop productivity. While this study focuses on soil health indicators, future research should incorporate assessments of the economic and environmental implications of these fertilization practices to provide a more comprehensive evaluation of their long-term sustainability. Overall, the study highlights the need for effective treatment practices that prioritize balanced nutrient application and soil conservation to enhance crop productivity, improve soil structure, and ensure long-term sustainability in agricultural systems.