Changes in Soil Aggregates and Aggregate-Associated Carbon Following Green Manure–Maize Rotations in Coastal Saline Soil

Abstract

Coastal saline–alkali soils, characterized by poor structures and low fertility, limit sustainable agricultural development. This study aimed to investigate how green manure application influence soil aggregate stability and soil organic carbon (SOC) sequestration in such coastal saline soils. Field experiments were conducted by comparing the following five treatments: (1) control (CK); (2) ryegrass full incorporation (RF); (3) ryegrass mulching (RM); (4) alfalfa full incorporation (AF); (5) alfalfa mulching (AM). The results demonstrated that green manure application significantly increased large macroaggregate (>2 mm) proportions by 20.60–56.70% while reducing microaggregates (<0.25 mm) by 24.35–68.43%. SOC increased across 0–40 cm soil depth, primarily driven by large macroaggregates and microaggregates, which contributed 23.7–73.5% and 34.8–91.4% of the total increase, respectively. Mulching treatments (AM/RM) increased surface SOC sequestration, while full-incorporation practices (AF/RF) boosted subsoil SOC stocks. These results highlight green manure application as an effective strategy to rehabilitate coastal saline soils by enhancing aggregate stability and SOC sequestration, providing technical guidance for saline soil rehabilitation in coastal saline regions.

1. Introduction

With the global population projected to reach 9.7 billion by 2050, agricultural production must increase by 70% to meet rising food demands, necessitating innovative approaches to expand arable land [1]. Saline–alkali soils, distributed across over 100 countries worldwide, have been considered as an important reserve arable land resource [2]. The Food and Agriculture Organization (FAO) designated the theme of World Soil Day 2021 as “Halt soil salinization, boost soil productivity”, highlighting the importance and urgent need for the sustainable utilization of saline soil [3]. In China, about 3.69 × 10⁷ hm² of saline–alkali land is dispersed in the northeast, northwest, central north, north, and coastal area [4]. Of this land, Jiangsu Coast has a saline soil area of 4.5 × 10⁵ hm² [5]. The strategic development of these soils holds immense potential to alleviate land scarcity and safeguard national food security. However, the soil in these areas has a poor structure, high salinity, nutrient deficiencies, and low soil organic carbon (SOC), which directly reduce arable land availability and constrain crop growth [6]. Therefore, the amelioration of saline–alkali soils has become an urgent imperative to enhance cultivated land fertility and ensure sustainable utilization of agricultural resources.

As fundamental components of soil structure and fertility, soil aggregates and SOC are critical for ensuring food production and security. Mechanistically, soil aggregates not only effectively protect SOC from microbial decomposition, thereby enhancing SOC storage [7], but also regulate soil structural quality and hydrological functions through enhanced moisture retention and optimized transport capacity [8]. These properties collectively contribute to enhancing agricultural sustainability by improving soil productivity, reinforcing erosion resistance, and promoting physiological development [9]. Furthermore, SOC is the main binder for the aggregates’ formation, which can enhance the stability of soil aggregates and provide a favorable crop growth environment [10,11]. Hence, the soil aggregates’ structural improvement and SOC enhancement is vital in building soil fertility and increasing food production for saline soil [12,13].

SOC and soil aggregates’ composition and stability are impacted by the addition of amendments. Soil amendments, such as biochar, polyacrylamide, and desulfurized gypsum, can improve the soil structure and enhance soil organic carbon (SOC) content in saline–alkali soils [14,15,16]. However, these approaches often involve high costs and may pose environmental risks. In contrast, green manure application—defined as incorporating fresh plant biomass into the soil—is recognized as a cost-effective and eco-friendly soil improvement strategy, attracting widespread attention. Studies indicated that integrating green manure into rotation systems during winter fallow periods introduces exogenous organic carbon into the soil, promoting SOC sequestration, aggregate stabilization, and subsequent crop yield increases [17]. Zhang et al. [18] demonstrated that green manure introduction significantly elevates SOC content, enhances macroaggregate formation, and improves aggregate stability. Similarly, Song et al. [19] reported that alfalfa (Medicago sativa) rotations increased macroaggregate content by approximately 72.17%, with stability progressively improving with soil depth. However, prior research has predominantly focused on green manure effects in non-saline soils. For instance, studies in conventional agricultural systems suggest that gramineous green manures (e.g., ryegrass) may enhance SOC more effectively than fabaceae green manures (e.g., alfalfa) when incorporated [20]. Despite these insights, comparative analyses of leguminous and gramineous green manures in saline soils was poorly reported. There are also studies showing that different returning methods have different effects. At present, research on returning green manure to the field mainly focuses on the combination of different amounts and methods of returning green manure. The exact impacts of different combinations of green manure types and methods of returning green manure to the field within rotation systems on the properties of soil aggregates and SOC content in saline soils remain unclear.

The objectives of this study are as follows: (1) to investigate the effects of green manure applications on the particle size composition and stability of soil aggregates; (2) to determine the effects of green manure applications on the distribution characteristics of SOC in different size aggregates; (3) to examine the effects of green manure applications on maize yield and yield components.

2. Methods and Materials

2.1. Site Description

The experiment was conducted at the Coastal Research Base of Jiangsu Hydraulic Research Institute, located in Dongtai City, Jiangsu Province, China (120°53′ E, 32°51′ N). This site represents a characteristic coastal alluvial plain within the Yangtze River Delta region. The area exhibits a subtropical monsoon climate, characterized by concurrent heat and precipitation, with mild and humid conditions. The mean annual temperature is 15.1 °C, with an average annual precipitation of 986.0 mm and an average annual evaporation of 1066.6 mm. Particle size distribution was determined using laser diffraction (Mastersizer 3000, Malvern Panalyical, Worcester, UK) and the soil was classified as silt loam by international classification standard [21]. Prior to the experiment, the initial soil properties were as follows: pH of 7.9, SOC of 5.0 g kg⁻¹, soil bulk density of 1.36 g cm⁻³, electrical conductivity of 154.2–202 μs cm⁻¹, and salt content of 3–5 g kg⁻¹. The initial available phosphorus, exchangeable potassium, and mineral nitrogen contents were 8.8 mg kg⁻¹, 114.6 mg kg⁻¹, and 110.7 mg kg⁻¹, respectively. All of the above indicators were determined as described in the literature [22]. The soil, reclaimed from coastal tidal flats, exhibited low fertility. The distribution of daily precipitation and air temperature during the experiment period are depicted in Figure 1. Meteorological data were collected from a small weather station located 500 m from the test site.

2.2. Experimental Design

The experiment was conducted from November 2022 to October 2023. It used a completely randomized design of five treatments replicated three times. The five treatments consisted of (1) control (CK); (2) ryegrass full incorporation (RF); (3) ryegrass mulching (RM); (4) alfalfa full incorporation (AF); (5) alfalfa mulching (AM). Each experimental plot was 8 m × 4 m, and the adjacent plots were separated by 50 cm buffer zones to minimize the influence of neighboring plots. The experiment included 15 experimental plots.

During the winter fallow period, green manure was planted in mid-November. Alfalfa (Medicago sativa L.) and ryegrass (Lolium perenne L.) were sown at the recommended rate of local green manure application, which were 45 kg ha⁻¹ and 45 kg ha⁻¹, respectively. No chemical fertilizer was applied during the winter green manure planting. Before maize sowing, the fresh aboveground biomass of green manure was manually harvested, cut into pieces, and then returned to the same experimental plot. The green manure with mulch was applied to the soil surface, and the green manure with incorporation was incorporated into the soil to a depth of 20–30 cm. Green manure and maize were sown manually in lines. Prior to maize sowing, the fresh aboveground biomass of green manure was manually harvested and cut into pieces for subsequent returning to the field. For full incorporation treatments (AF/RF), the chopped green manure was uniformly spread on the soil surface before mechanical tillage, which buried the material into deeper soil layers. In contrast, surface mulching treatments (AM/RM) applied the chopped material evenly across the soil surface post-tillage. Control plots (CK) underwent identical tillage operations at the same time to ensure consistency across treatments.

Maize was sown on June 25 and harvested on October 8, respectively. The maize variety, “Yufeng 303”, was sown in 55 cm rows with 25 cm spacing between plants (equivalent to 7.27 plants m⁻²). The hill-drop method (3 seeds per hill) was used at a seeding rate of 25 kg ha⁻¹. Before maize sowing, Schickefung long-lasting compound fertilizer (N:P:K = 29:5:6) was applied as a basal fertilizer; the amount of fertilizer was 833 kg ha⁻¹. Urea (N, 46%) was applied at the maize jointing stage; the amount of fertilizer was 833 kg ha⁻¹. The farming method was entirely manual planting, with seeds sown at a depth of 30 cm. Other agronomic and management practices followed local practice in this region.

2.3. Soil and Plant Sample Collection and Analysis

Soil samples were obtained after maize harvesting in 2023. To determine the total SOC, soil samples were collected from four depth intervals (0–10, 10–20, 20–30, and 30–40 cm) using a stainless steel auger (4 cm inner diameter), following a random five-point sampling technique. Soil samples were immediately stored in self-sealing bags for transportation to the laboratory. In the laboratory, samples were air-dried under cool, dry conditions, and we manually removed soil stones, plant roots, and impurities. Soil samples were artificially ground and sieved through a 0.15 mm mesh prior to total SOC. The SOC was determined using the potassium dichromate external heating method [23]. The unbroken soil samples were collected at 0–30 cm soil depth with three replicates in each plot using a stainless-steel soil sampler. Samples from each plot were securely housed in a plastic container, and the samples were sent to the laboratory for further processing. The soil samples were air-dried, and, after the removal of soil stones, plant roots, and impurities by hand, the soil samples were gently peeled into small blocks of approximately 10 mm along their natural planes, to avoid soil deformation. The dry sieving method was employed to analyze the soil aggregate fractions [24]. Here, 500 g soil was placed on a sieve mesh, which was arranged in order from top to bottom according to the size of the sieve holes (2, 0.5, and 0.25 mm). The soil was then placed on a shaking table at a speed of 1400 times/minute and shaken for 25 min. Subsequently, soil samples were collected and weighed from each sieve mesh level in order to calculate the percentage content of soil aggregates at each particle level. Soil aggregates were classified as soil large macroaggregates (>2 mm), small macroaggregates (2–0.5, 0.5–0.25), and microaggregates (<0.25 mm). The separate aggregates from the 0–10 cm, 10–20 cm, and 20–30 cm soil layers were subsequently ground and sieved through 0.15 mm sieves. The SOC concentration of aggregates was analyzed using the potassium dichromate external heating method [23].

At maize harvest, the maize ears in each row were manually counted and multiplied by the number of rows to obtain the number of ears in each plot. Three maize plants were randomly selected from each plot to estimate the yield and yield components, i.e., the kernel number per ear and 100-kernel weight. The maize grain yield was recorded at a moisture content of 12%. At the same time, the stem thickness was measured at 5 cm above the ground. Plant height was measured from the ground level to the top of the tasseling. Biomass samples were oven-dried at 105 °C for 30 min to kill the tissues, and then at 80 °C until a constant mass was maintained.

2.4. Calculations

(1)

Soil organic carbon

The formula for calculating soil organic carbon is as follows [23]:

$$SOC={\displaystyle \frac{0.8000\times \raisebox{1ex}{$5.0$}\!\left/ \!\raisebox{-1ex}{$V_0$}\right.\times (V_0-V)\times 0.003\times 1.1}{\mathrm{m}\times \mathrm{k}}}$$

(1)

where 0.8000 is the concentration of potassium dichromate standard solution (mol L⁻¹); V₀ is the volume of FeSO₄ used for blank titration (mL); V is the volume of FeSO₄ used for blank titration (mL); m is the mass of air-dried soil samples (g); k is the coefficient of conversion of air-dried soil to dried soil.

(2)

Analysis of content mechanical stability of soil aggregates

The proportion of each aggregate size fraction was expressed as the percentage of the total dry weight of the sample. Specifically, R_>0.25 represents the mass percentage of aggregates larger than 0.25 mm relative to the total dry weight.

The percentage of mechanically stable aggregates with a particle size greater than 0.25 mm, R_>0.25 (%) was calculated as follows [25]:

$$R_{>0.25}={\displaystyle \frac{W_{>0.25}}{W_t}}\times 100$$

(2)

where W_>0.25 is the mass of aggregates with a diameter greater than 0.25 mm (g).$W_t$ is the total mass of aggregates (g).

To assess soil aggregate stability, four important indicators were analyzed, namely soil mean weight diameter (MWD), geometric mean diameter (GMD), fractal dimension (D), and soil erodibility (K).

The mean weight diameter (MWD (mm)) and geometric mean diameter (GMD (mm)) of soil aggregates were calculated as follows [26]:

$$MWD={\sum }_{i=1}^n({\overline{X}}_i\times W_i)$$

(3)

$$GMD=\exp \left[{\displaystyle \frac{{\sum }_{i=1}^n(W_i\times \ln {\mathrm{X}}_i)}{{\sum }_{i=1}^n{W}_i}}\right]$$

(4)

where ${\mathrm{X}}_i$ is the mean of the $i$ particular size of aggregate (mm). $W_i$ is the mass of the $i$ particular size of aggregate (g).

The fractal dimension (D), and erodibility (K) were calculated as follows [27,28]:

$${\left({\displaystyle \frac{{\overline{d}}_i}{d_{max}}}\right)}^{3-D}={\displaystyle \frac{W\left(\delta <\overline{d_i}\right)}{W_0}}$$

(5)

$$\mathrm{K}=7.954\times \left\{0.0017+0.0494\times exp\left[-0.5\times {\left({\displaystyle \frac{1.675+lgGMD}{0.6986}}\right)}^2\right]\right\}$$

(6)

where ${\overline{d}}_i$ is the average diameter of two sieving diameters ${\overline{d}}_i$ and ${\overline{d}}_{i+1}$ (mm), $W\left(\delta <\overline{d_i}\right)$ is the mass of aggregates with a diameter shorter than average (g), W₀ is the mass of all aggregates (g), and $d_{max}$ is the largest average diameter of an aggregate (mm).

2.5. Statistical Analysis

The data were processed using Microsoft Excel software (Version 2018) and the figures were drawn using Origin software (Version 2019). Statistical analyses were executed using IBM SPSS Statistics 21 (IBM Corp., Armonk, NY, USA) with verification of all parametric test assumptions (normality by the Shapiro–Wilk test, homogeneity of variance by Levene’s test). One-way analysis of variance (ANOVA) was conducted, followed by Duncan’s multiple range test for post hoc comparisons at a significance level of p < 0.05. Bivariate Pearson correlation analyses (two-tailed) were conducted with 95% confidence intervals, and correlation coefficients (r) were interpreted according to Cohen’s criteria.

3. Results

3.1. Distribution Characteristics of Soil Aggregates

Figure 2 illustrates the distribution characteristics of soil aggregates under different green manure returning treatments. The dominant aggregate classes were the >2 mm fraction (21.4–63.04%) and <0.25 mm fraction (27.74–82.93%). The distribution of soil aggregates in the 0–10 and 10–20 cm layers was consistent, and the proportion of macroaggregates (>2 mm) increased significantly in all layers with green manure application compared with CK. In the 0–20 cm soil layer, compared with CK treatments, the proportion of the >2 mm aggregates significantly increased, and the increase rates were 523.89%, 464.41%, 360.32%, and 220.27% under the AF, AM, RF, and RM treatments, respectively. The proportion of <0.25 mm aggregates decreased by 63.47%, 64.04%, 36.50%, and 31.70% under the AF, AM, RF, and RM treatments, respectively, compared to CK (Figure 2a). In the 20–30 cm soil layer, except for the AM treatment, AF, RF, and RM treatments all showed an increasing trend in >2 mm aggregates and a decreasing trend in <0.25 mm aggregates, compared with CK.

3.2. Stability of Soil Aggregates

The stability of soil aggregates is shown in

Table 1. R_>0.25, GMD, MWD, D, and K indexes of soil aggregates under different treatments at maize harvest.

Soil Depth (cm)	Treatment	R>0.25 mm	MWD	GWD	D	K
		(%)	(mm)	(mm)
0–10 cm	AF	67.09 ± 2.45 a	3.54 ± 0.10 a	1.38 ± 0.12 a	2.86 ± 0.08 c	0.03 ± 0.01 d
	AM	82.05 ± 12.32 a	4.01 ± 0.68 a	2.14 ± 0.70 a	2.64 ± 0.06 d	0.02 ± 0.01 d
	RF	55.29 ± 2.75 b	2.83 ± 0.13 b	0.86 ± 0.09 b	2.9 ± 0.02 b	0.04 ± 0.01 c
	RM	44.34 ± 1.27 b	2.15 ± 0.21 b	0.56 ± 0.05 b	2.85 ± 0.03 c	0.06 ± 0.02 b
	CK	19.92 ± 3.27 c	0.90 ± 0.20 c	0.23 ± 0.03 b	2.96 ± 0.02 a	0.14 ± 0.04 a
10–20 cm	AF	71.95 ± 2.43 a	3.72 ± 1.12 a	1.60 ± 0.15 a	2.82 ± 0.07 d	0.06 ± 0.01 bc
	AM	57.44 ± 12.31 b	2.73 ± 0.61 b	0.87 ± 0.68 b	2.85 ± 0.09 d	0.04 ± 0.01 c
	RF	50.14 ± 2.78 b	2.59 ± 0.12 b	0.72 ± 0.07 b	2.92 ± 0.03 b	0.05 ± 0.01 bc
	RM	41.83 ± 1.26 b	1.82 ± 0.15 b	0.47 ± 0.04 b	2.89 ± 0.04 c	0.07 ± 0.02 b
	CK	13.46 ± 3.20 c	0.55 ± 0.15 c	0.18 ± 0.04 b	2.96 ± 0.02 a	0.17 ± 0.05 a
20–30 cm	AF	46.18 ± 2.39 a	2.20 ± 0.08 a	0.59 ± 0.09 a	2.89 ± 0.09 c	0.06 ± 0.02 c
	AM	16.77 ± 4.67 b	0.59 ± 0.18 b	0.20 ± 0.54 a	2.94 ± 0.10 b	0.16 ± 0.05 a
	RF	46.47 ± 2.74 a	2.46 ± 0.15 a	0.65 ± 0.06 a	2.93 ± 0.03 b	0.05 ± 0.01 c
	RM	32.06 ± 1.30 a	1.45 ± 0.11 a	0.34 ± 0.02 a	2.93 ± 0.05 b	0.10 ± 0.03 b
	CK	14.30 ± 3.21 b	0.63 ± 0.18 a	0.19 ± 0.05 a	2.97 ± 0.03 a	0.17 ± 0.06 a

Notes: (1) control (CK); (2) ryegrass full incorporation (RF); (3) ryegrass mulching (RM); (4) alfalfa full incorporation (AF); (5) alfalfa mulching (AM). Date are shown as the mean ± standard deviation (SD) (repeated three times). Different lowercase letters in the table indicate significant differences among treatments at the same soil layer (p < 0.05).

. In the 0–10 cm soil layer, compared to the CK treatment, the MWD and GMD values followed the order AM > AF > RF > RM > CK, while the R_>0.25 (percentage of aggregates > 0.25 mm) exhibited a similar trend (AM > AF > RF > RM > CK). In the 10–20 cm soil layer, the trends for MWD, GMD, and R_>0.25 were consistent with those observed in the 0–10 cm layer. In the 20–30 cm soil layer, no significant differences in MWD, GMD, or R_>0.25 were observed between the AM and CK treatments (p > 0.05), while the remaining treatments followed the order RF > AF > RM. Among the four treatments, the highest values of MWD, GMD, and R_>0.25 were observed under the AF treatment, indicating that the AF treatment resulted in the most stable soil aggregates. In contrast, the RM treatment showed the lowest values, suggesting poor soil aggregate stability. These results demonstrate that the alfalfa–maize rotation significantly improved soil aggregate stability.

Compared with CK, soil D and K decreased with green manure application. Specifically, the soil D decreased by 4.73–2.69%, 10.81–1.01%, 3.72–0.34%, and 2.03–1.35% in AF, AM, RF, and RM, respectively. The soil K decreased by 78.57–64.71%, 85.71–5.88%, 58.82–41.18%, and 71.43–70.59% in AF, AM, RF, and RM, respectively.

3.3. Distribution of Aggregate-Associated Carbon

Figure 3 illustrates the distribution of SOC in aggregates of different size fractions under different treatments. In the 0–10 cm soil layer, except for the AF treatment, SOC in the <0.25, 0.5–0.25, 2–0.5, and >2 mm aggregates initially increased and subsequently decreased in the AM, RF, RM, and CK treatments. In the 0–30 cm soil layer, except for the AF, AM, and RM in <0.25 mm aggregates in the 0–10 cm soil layer, green manure application increased the SOC in all aggregate sizes compared to CK treatment.

In the 0–10 cm soil layer, the SOC content in >2 mm aggregates significantly increased by 308.24%, 191.04%, and 74.91% under the AM, RF, and RM treatments, respectively, compared to the CK treatment (p < 0.05). For the 2–0.5 mm size fractions, the AM and RF treatments showed significant SOC increments of 7.90 g kg⁻¹ and 6.04 g kg⁻¹ compared to CK, respectively. A similar trend was observed in for the 0.5–0.25 mm fraction, where the AM and RF treatments showed an 8.37 g kg⁻¹ and 7.43 g kg⁻¹ increase, respectively. Notably, no significant differences were observed in the <0.25 mm fraction among the treatments compared to CK. In the 10–20 cm soil layer, the SOC content in >2 mm aggregates significantly increased by 7.19 g kg⁻¹ and 8.06 g kg⁻¹ under the AF and RF treatments, while the SOC content in <0.25 mm aggregates significantly increased by 7.26 g·kg⁻¹ and 6.98 g·kg⁻¹ under the AM and RF treatments, respectively, compared to the CK treatment (p < 0.05). No significant differences were observed in the SOC content of other aggregate size fractions among the treatments compared to CK.

3.4. Total SOC Under Different Treatments

Figure 4 illustrates the vertical distribution of SOC in the 0–40 cm soil profile at maize harvest. In the 0–10 cm soil layer, the AF and AM treatments significantly increased the SOC content by 30.13% and 39.62%, respectively, compared to the CK treatment (p < 0.05). Although the RF and RM treatments also showed an increase in SOC content, the differences were not statistically significant compared to CK. In the 10–20 cm soil layer, the SOC content under the AF and RF treatments increased significantly by 45.79% and 19.94% compared to CK, respectively (p < 0.05), while the increases under the AM and RM treatments were not significant compared to CK (p < 0.05). In the 20–30 cm soil layer, the SOC content under the AF, RF, and RM treatments increased significantly by 22.18%, 32.14%, and 21.27%, respectively, compared to CK (p < 0.05). In the 30–40 cm soil layer, there was no significant difference among treatments, but the total SOC with green manure application was higher than that of CK.

3.5. Contribution of Soil Aggregates to SOC

Figure 5 illustrates the contribution of SOC in different aggregate size fractions within the 0–30 cm soil layer. For the CK treatment, the <0.25 mm aggregates were the greatest contributor to SOC, but there was no significant difference among >2, 0.5–0.25, and 2–0.5 mm aggregates. Green manure application and soil layer had significant effects on the contribution of soil aggregates to SOC. In the 0–10 cm soil layer, the >2 mm aggregates were the greatest contributor for the AM (73.51%) and RF treatments (51.17%), while the <0.25 mm aggregates were the greatest contributor for the AF (72.17%) and RM treatments (43.28%). In the 10–20 cm soil layer, the >2 mm aggregates was the greatest contributor for the AF (63.96%) and AM treatments (49.57%), while the <0.25 mm aggregates were the greatest contributor for the RF (45.65%) and RM treatments (59.25%). In the 20–30 cm soil layer, the >2 mm aggregates were the greatest contributor for the RF treatment (51.37%), while the <0.25 mm aggregates were the greatest contributor for the AM (91.4%), AF (62.04%), and RM treatments (61.3%).

3.6. Correlation of Soil Aggregates, Erodibility, and Organic Matter Content

As shown in Figure 6, soil MWD exhibited highly significant positive correlations with the proportions of mechanically stable aggregates in the >2 mm fraction (1.000 **, p < 0.01), and soil GMD was positively correlated with the proportions of the 2–0.5 mm fraction (0.978 **, p < 0.01). In contrast, MWD showed a highly significant negative correlation with the proportion of <0.25 mm mechanically stable aggregates (−0.993 **, p < 0.01). Furthermore, soil D and K were significantly negatively correlated with the proportions of 2–0.5 mm (−0.993 **, p < 0.01) and >2mm (−0.971 **, p < 0.01) mechanically stable aggregates, respectively, while K showed a significant positive correlation with the proportion of <0.25 mm mechanically stable aggregates (0.971 **, p < 0.01). However, the correlations between these indicators and the SOC content of aggregates were generally weak, and only GMD exhibited a significantly positive correlation with the SOC content in <0.25 mm aggregates (0.983 *, p < 0.05).

3.7. Effects of Green Manure–Maize Rotations on Maize Yield

Green manure influenced maize yield and yield components (

Table 2. Comparative analysis of maize yield, yield components, and agronomic traits under different green manure rotation systems.

Treatment	Yield and Yield Components			Aboveground Dry Matter (t hm−2)	Plant Height (cm)	Stem Diameter (mm)
	Kernel Number Per Ear	100-Grain Weight (g)	Grain Yield (t hm−2)
AF	528.79 b	30.18 a	6.52 a	19.64 a	216.75 a	2.13 a
AM	567.08 a	31.94 a	6.84 a	18.36 a	220.75 a	2.18 a
RF	537.07 b	30.28 a	6.56 a	17.92 b	216.42 a	2.08 a
RM	572.14 a	32.32 a	6.98 a	17.45 b	217.00 a	1.73 a
CK	529.33 b	29.84 a	6.25 b	16.83 b	215.17 a	1.85 a

Notes: (1) control (CK); (2) ryegrass full incorporation (RF); (3) ryegrass mulching (RM); (4) alfalfa full incorporation (AF); (5) alfalfa mulching (AM). Data are shown as the mean (repeated three times). Different lowercase letters in the table indicate significant differences among treatments (p < 0.05).

). Compared with CK, the highest increase in maize yield was observed in the RM treatment (11.7%), followed by the AM treatment (9.54%) and the RF (5.00%) and AF (4.4%) treatments (p < 0.05). The kernel number per ear and the weight per hundred grains also showed a similar trend. In terms of aboveground biomass, AF and AM treatments significantly exceeded CK by 16.70% and 9.09%, while RM and RF were higher than CK, but the difference was not significant. Moreover, green manure increased maize plant height and stem diameter, but the difference was not significant (p > 0.05).

4. Discussion

4.1. Green Manure–Maize Rotation Enhances Soil Macroaggregate Formation and Stability

Soil aggregates serve as the fundamental units of soil structure, exerting a critical influence on preserving soil health and integrity via their spatial distribution [29]. These aggregates profoundly impact the physical, chemical, and biological attributes of soil ecosystems [30]. Significantly, large soil aggregates can protect SOC from breakdown by physically isolating it from decomposing microbial activity [31]. Aggregates are generally categorized as macroaggregates (>0.25 mm) and microaggregates (<0.25 mm) [32], and a higher proportion of macroaggregates suggests improved soil aggregation, while microaggregate dominance indicates a more dispersed soil structure [33]. Both incorporated and surface-mulched green manure treatments significantly increased the abundance of large macroaggregates (>2 mm) in the soil by 20.6–56.7% and decreased the proportion of microaggregates (<0.25 mm), while the small macroaggregates (2–0.25 mm) changed slightly compared with CK in the 0–20 cm layer (Figure 2). This might be attributed to the fact that green manure returning supplies the soil with exogenous organic matter, thereby facilitating aggregate formation. Furthermore, green manure returning also enhances the accumulation of divalent cations, such as exchangeable calcium (Ca²⁺) and magnesium (Mg²⁺), which actively improve the development of large macroaggregates. Our observation aligns with the mechanistic framework proposed by Elliott et al. [34], who revealed that the introduction of exogenous organic matter into soil ecosystems preferentially promotes the formation of macroaggregates over microaggregates. However, a previous study has also shown that green manure primarily increases the proportion of small macroaggregates (2–0.25 mm) [35]. This discrepancy may be due to differences in farmland types, as their study focused on paddy fields, whereas our research pertains to dryland farmland.

In the 0–20 cm soil layer, the results showed that, compared to ryegrass, alfalfa was more conducive to the increase in large macroaggregates. This might be due to the fact that alfalfa, as a fabaceae plant with a lower C/N ratio, is decomposed faster than ryegrass, which has a higher C/N ratio, and the early-stage decomposition led to more cementing material to promote the formation of large macroaggregates [36]. Consequently, ryegrass was less conducive to the transformation of small aggregates into larger particle size aggregates. For alfalfa, in the surface soil layer (0–10 cm), the large macroaggregate content under the AM treatment exceeds that under AF treatment. In contrast, at the 10–20 cm depth, the AF treatment had a higher large macroaggregate content than AM. This discrepancy may be due to the fact that compared with AM, when green manure residue is not completely exposed to air under AF treatment, the decomposition process results in less organic carbon loss. This leads to the formation of greater amounts of moist microaggregates bound by cementing materials, which become connected with colloidal minerals. Consequently, the content of soil macroaggregates in the green manure residues in the returning soil layer is improved [37].

Our results also showed that both the MWD and GMD of mechanically stable aggregates decreased significantly (p < 0.05) with soil depth across all treatments, indicating reduced aggregate stability in subsurface layers compared to surface soils. These findings align with Wang et al.’s [38] observations of aggregate stability patterns under different land uses in the Longdong Loess Plateau. Green manure application significantly increased MWD and GMD while decreasing D and K in the 0–30 cm soil layer compared with CK (p < 0.05). Thus, green manure made the soil structure more stable and, therefore, reduced the risk of soil erosion in this saline soil area. Notably, the AM treatment showed superior performance, with MWD and GMD values 12–15% higher than under the AF treatment, demonstrating its enhanced capacity for improving aggregate mechanical stability. Soils with a high proportion of macroaggregates (>0.25 mm) typically exhibit improved structural stability, reflected in increased MWD and GMD, alongside decreased D and K [39].

4.2. Green Manure–Maize Rotation Boosts Soil SOC Sequestration in Soil Aggregates

SOC represents a critical element within the soil carbon pool and serves as a cornerstone of the soil material cycle [40]. As an effective indicator of soil quality, its content provides valuable insights into soil health. Maintaining soil health is of paramount importance for ensuring global food security. In the present study, overall, SOC content decreased with the soil depth in saline soil, similar to the results of [41]. Previous studies have indicated that green manure enhances SOC [42]. Consistent with these findings, we found that except for the RF treatment at 10–20 cm and AM at the 20–30 cm soil depth, green manure increased the SOC compared with CK. This phenomenon can be explained by the following three reasons: (1) The application of green manure serves as a critical source for SOC accumulation, increasing organic carbon input and enhancing the total soil carbon pool; (2) green manure reduces SOC loss and minimizes SOC depletion [43]; (3) green manure incorporation enhances soil nitrogen content, which regulates the soil C/N ratio to stimulate organic matter decomposition and nutrient mineralization. This process concurrently modulated soil aggregate formation and stability through enhanced organo–mineral interactions, thereby improving SOC sequestration capacity. It should be noted that the SOC content under the mulching treatments (AM and RM) was higher than that under the full over returning treatments (AF and RF) in the surface layer, while the full over returning treatments (AF and RF) showed higher values than mulching treatments (AM and RM) in subsoil layers below 10 cm. This indicates that the enhancement effect of mulching on SOC is limited to the surface layer, whereas incorporation practices can influence deeper soil layers. This might be explained by the fact that the straw of AM was applied to the soil surface, and a large amount of organic matter was imported into the soil surface, which accelerated the decomposition of green manure by microorganisms and the accumulation of surface SOC. Compared with AM, AF returns alfalfa to the soil subsurface, forming a green manure layer in the deeper soil layer, improving microbial metabolic activity and facilitating the formation of subsurface soil humus and soil carbon fixation.

The aggregate-associated organic carbon is a microscopic characteristic reflecting the balance between soil organic matter accumulation and mineralization rates [44]. It directly mediates both short-term carbon sequestration efficiency and long-term fertility enhancement through physicochemical protection mechanisms [45]. A previous study has shown that the SOC content of aggregates increased with the increase in particle size [46]. However, in our study, we found that the SOC content of aggregates initially increased and then decreased with the increase in particle size. In the present study, our findings reveal that green manure–maize rotation significantly increased the SOC content in >2 mm aggregates, which aligns with the results of Li et al. [47] and Yang et al. [48], who reported that long-term winter green manure cultivation increased SOC content in >2 mm aggregates. Additionally, we found that the contribution of total SOC was mainly from >2 mm and <0.25 mm aggregates, with contribution rates of 23.65–73.51% and 34.79–91.40%, respectively. These results indicated that SOC in saline soils predominantly originates from protected storage within the aggregate classes of >2 mm and <0.25 mm. It should be noted that the SOC content of >2 mm aggregates was higher than that of <0.25 mm aggregates, but <0.25 mm aggregates contributed the most to total SOC. This indicated that aggregate SOC content and its contributing rate do not always match. This discrepancy may occur because the SOC contribution depends on both the proportion of each aggregate size and their specific SOC content. Our findings are consistent with Zhang et al. [49], who found that although <0.25 mm aggregates had the highest nitrogen content, large macroaggregates (>2 mm) contributed more to total nitrogen.

Pearson correlation analysis revealed that the strongest correlation with stability was the proportions of the >2 mm fraction and the 0.5–2 mm fraction (0.978 **), and the higher the soil MWD and GMD but the lower the soil D and K, suggesting that the proportion of these fractions play an important role in the stability of soil aggregates. Furthermore, the SOC content of aggregates was positively correlated with soil MWD and GMD but negatively correlated with soil D and K. Furthermore, the SOC in the <0.25 mm fraction had the most significant effect on the stability of soil aggregates.

4.3. Green Manure–Maize Rotation Boosts Maize Productivity

The accumulation of dry matter in maize directly determines grain yield, especially during the reproductive stage [50]. Our results indicated that, compared to CK, both green manure mulching and incorporation significantly increased the maize yield. Studies have shown that green manure addition enhances maize productivity through four synergistic mechanisms, as follows: (1) increasing soil organic carbon, (2) improving aggregate stability, (3) reducing weed pressure and erosion [51,52], and (4) releasing essential nutrients (such as nitrogen, phosphorus, and potassium) into the agricultural ecosystem through enhanced microbial decomposition and mineralization, making them available for crop uptake [53], thereby providing better growth conditions for subsequent crops [54]. Additionally, research has demonstrated that fabaceae green manure, with its low lignin content and C/N ratio, accelerates microbial decomposition and promotes rapid nitrogen mineralization during critical growth stages [55], improves nitrogen availability through biological nitrogen fixation, and the degradation of its straw and roots releases organic acids, which dissolve recalcitrant forms of phosphorus [56]. Collectively, these mechanisms enhance nutrient bioavailability, particularly for high-demand crops, such as maize. It is worth noting that in this study, although the dry matter accumulation of maize treated with alfalfa was significantly higher than that of ryegrass, the maize yield was lower than that of the ryegrass treatment. This indicates that an increase in dry matter does not necessarily mean an increase in yield. The returning method also affected the maize yield. In our study, we found that the maize yield was higher in the mulching treatment than in the incorporation treatment. This finding was consistent with Shang et al. [57], whose study found that both surface mulching with green manure and full incorporation are beneficial for the accumulation of dry matter and the increase in maize yield, but the yield-increasing effect of surface coverage is greater than that of full tillage incorporation with green manure.

5. Conclusions

This study investigated the effects of different green manure returning treatments on soil aggregates and their associated carbon content in coastal saline soils. Green manure returning increased the proportion of macroaggregate (>2 mm), while reducing microaggregates (<0.25 mm) in the 0–30 cm layer. SOC was mainly contributed by the >2 mm and <0.25 mm aggregate fractions, which validated that >2 mm and <0.25 mm fractions are key to the stability and aggregation of coastal saline soils. These results demonstrate that green manure return strategies effectively improve soil structural stability and SOC sequestration capacity in coastal saline soils. Overall, the AF treatment showed optimal performance in enhancing soil total organic carbon and macroaggregate (>2 mm) formation, whereas the RM treatment exhibited superior performance in crop yield improvement. This study provides theoretical and technical guidance for saline soil rehabilitation in coastal saline farmlands. However, while this study confirms the short-term effects of returning green manure to the field, the long-term sustainability of these improvements requires further investigation. Additional research should address the temporal carbon sequestration patterns, microbial mediation mechanisms, and site-specific management adaptations.