Soil Carbon Dynamics, Sequestration Potential, and Physical Characteristics Under Grazing Management in Regenerative Organic Agroecosystems

Abstract

Rotational grazing and cover crops are conservation practices known to improve soil health, particularly soil organic carbon (SOC) and aggregate stability. Combining both practices may enhance these benefits more than either alone. With grazing lands covering 41% of U.S. agricultural land, adopting such methods could significantly impact the soil carbon cycle. A study near Koshkonong, Missouri, examined the effects of regenerative organic grazing with Bubalus bubalis (Linnaeus) on SOC, carbon sequestration, aggregate stability, and soil resistance. The 1620-hectare ranch tested four treatments: rotational grazing with cover crops (RGCC), grazing on native grasses (RGNCC), cover crops without grazing (NGCC), and orchards without cover crops or grazing (NGNCC). Cover crops were seeded twice yearly with diverse species. After three years, SOC increased most in NGNCC (28%), followed by RGCC (13%), NGCC (7%), and RGNCC (4%). Annual carbon gains in surface soils were highest in NGNCC (0.99 Mg ha⁻¹ yr⁻¹). Across all depths, NGCC led (4.88 Mg ha⁻¹ yr⁻¹). Aggregate stability was greatest in non-grazed systems, particularly in fine aggregates, and declined with soil disturbance. Overall, low-disturbance systems like orchards and no-grazing cover crop plots enhanced soil structure and carbon storage. Strategic management is key to improving soil function and ecosystem resilience.

1. Introduction

Soil is the largest terrestrial carbon reservoir and thus a central component of the global carbon (C) cycle (e.g., ~2400 gigatons (Gt) C to 2 m depth) [1]. Accurately assessing whether ecosystems and their soils function as carbon sinks or sources is essential for projecting future atmospheric CO₂ trajectories under changing climate and land use [2]. Climate change and land-use transitions (such as the shift toward more intensive agricultural production) can significantly alter soil organic carbon (SOC) stocks, with substantial implications for the atmospheric carbon budget. The potential amount of SOC sequestered depends on soil mineralogy and texture [3], environmental constraints such as climate and moisture regimes [4], and management practices. Regenerative land management methods (e.g., cover cropping, reduced tillage, integrated livestock, rotational grazing) increasingly show promise for enhancing SOC retention and improving soil health, whereas conventional intensive practices often drive SOC losses and contribute to soil degradation [5]. Nevertheless, the rate and direction of SOC change following management shifts or land conversion vary spatially and temporally. Key factors such as fertilization intensity, grazing pressure, stocking density, and vegetation removal or restoration strongly influence ecosystem carbon storage and SOC composition [6]. The adoption of high-stock density or regenerative grazing schemes (e.g., adaptive multi-paddock grazing) has in some settings led to accelerated SOC gains and associated soil health benefits [7].

Grazing lands currently occupy approximately 3.2 billion hectares, nearly one quarter of Earth’s terrestrial surface [8]. The management of these lands exerts a profound influence on soil physical, chemical, and biological attributes, thereby modulating C storage capacity. Management regimes differ in paddock configuration, livestock density, rotation frequency, and forage species. Adaptive multi-paddock (AMP) grazing is a regenerative approach characterized by elevated stocking densities, brief grazing durations, and extended rest intervals. This flexible strategy adapts to forage dynamics and animal behavior to promote soil recovery and pasture resilience [9]. Emerging evidence indicates that AMP systems may enhance SOC accrual, increase soil organic matter (SOM), and improve aggregate stability, thereby mitigating SOC losses by restricting microbial access and improving SOC mineral associations [3,7,9].

Multiple studies report that AMP regimes tend to reduce soil penetration resistance and curb sediment loss relative to continuous grazing [10] and may also promote greater forage productivity under favorable conditions [11]. Conversely, reviews and meta-analyses generally show that continuous grazing is associated with SOC decline and increased bulk density, whereas rotational or moderated grazing systems often lead to improvements in soil quality and nutrient cycling [1,12,13]. However, in certain contexts, intensive practices can favor accumulation of labile, short-lived C pools rather than more stable SOC fractions, limiting long-term sequestration potential [5,14]. Management-intensive rotational grazing (MIRG) shares key features with AMP, such as small paddock sizes and rapid rotational schedules, but typically employs lower stocking densities and prioritizes maximizing forage regrowth and nutritional quality. In contrast, rotational grazing involves fewer paddocks with extended grazing durations, while continuous grazing offers unrestricted pasture access with variable stocking densities.

Various research shown that AMP grazing systems enhance SOC accumulation, increase SOM, and improve aggregate stability, which in turn slows SOC loss by limiting microbial access and increasing the interaction between mineral surfaces and SOC [15]. Land management strategies play a critical role in determining aggregate stability, which is essential for maintaining soil and environmental quality in agroecological cropping systems [16]. Soil aggregate stability, a fundamental parameter reflecting the structural integrity of soil, is profoundly influenced by intensive agricultural cultivation. Compounding this, soil compaction, primarily induced by mechanized field operations and repeated livestock trampling, adversely affects multiple soil functions. These include root system proliferation, gaseous exchange within soil pores, nutrient mobility, and the soil’s hydraulic conductivity, all of which contribute to diminished soil quality and crop productivity [17].

Concurrently, AMP grazing reduces soil penetration resistance and minimizes sediment loss when compared to continuous grazing [18,19,20] and can lead to increased forage production [21]. Conversely, meta-analytical studies reveal that continuous grazing often leads to declines in SOC alongside increased soil bulk density. Rotational grazing strategies, however, typically induce opposite effects, enhancing soil quality and nutrient cycling [22]. Collectively, these outcomes highlight the profound impact that grazing intensity and management frameworks exert on long-term soil functionality and C cycling dynamics. In certain cases, intensive practices encourage the accumulation of short-lived, unstable C forms rather than more persistent pools [23], limiting the potential for sustained sequestration and climate regulation.

Globally, soil holds an estimated 2500 Gt of C [1], including about 1550 Gt of SOC and 950 Gt of soil inorganic C. The soil C pool is 3.3 times the size of the atmospheric pool (760 Gt) and 4.5 times the size of the biotic pool (560 Gt) [24]. Because soil C represents the largest pool within the terrestrial C cycle, any disruption to its storage can significantly affect global C dynamics [25]. The capacity of soil to sequester C depends on multiple factors, such as texture, climate, topography, and the ecosystem it supports. For this reason, developing management strategies aimed at preserving and enhancing soil resources is a key step toward maintaining ecosystem sustainability [23].

Cover crops are widely recognized for their role in enhancing soil health and promoting C sequestration by increasing SOM, improving aggregate stability, and supporting microbial activity [26,27]. These benefits contribute to improved soil structure, water retention, and nutrient cycling, making cover crops a key strategy in sustainable land management. Similarly, well-managed grazing systems can influence C dynamics and soil quality, particularly through root biomass production and organic matter inputs [18]. Both practices are known as drivers for improved soil function and C storage. However, few studies have investigated the combined effects of cover crops and grazing, especially in grassland or semi-arid systems. Understanding how these two practices interact is critical, as their integration may produce synergistic or antagonistic effects on soil C and structural stability. This study addresses that gap and may be among the first to evaluate their combined impact on aggregate stability, offering valuable insight into soil health management strategies.

This study therefore aims to (1) estimate soil C sequestration, (2) evaluate changes in SOC dynamics, and (3) measure wet aggregate stability and soil penetration resistance under different pasture management strategies under intensive grazing and regenerative agricultural systems in the southern and mid-southern region of the United States.

2. Materials and Methods

2.1. Site Description

This study was conducted from 2022 to 2024 on a ranch located near Koshkonong, Missouri (36.60° N, 91.65° W). The region is characterized by a humid subtropical climate, with hot, humid summers and mild to cool winters [28,29]. Long-term climate data indicate a mean annual temperature of approximately 13.8 °C and a mean annual precipitation of 1200–1250 mm [28,30]. Monthly average temperatures range from −2 to 8 °C in January to 19–32 °C in July, while annual precipitation is relatively evenly distributed, with peaks during late spring and early summer (April-June) [30,31]. Snowfall is limited, averaging ~180 mm yr⁻¹ (178 mm) [31]. These climatic conditions marked by warm, moist summers and variable winter freeze thaw cycles strongly influence SOC dynamics, aggregate stability, and microbial activity. The combination of relatively high precipitation and moderate temperature variability supports productive grassland systems but also increases the potential for seasonal moisture stress and erosion in poorly managed soils [28,29].

The farm encompasses approximately 1620 ha and is characterized by rolling hills in the Ozark Mountains region. The native ecosystem is a mixed hardwood forest, and forestry and agriculture are the main economic activities in the region. Ranches tend to be managed as continuously grazed pasture. The soil at the experiment site was characterized as loamy sand (fine-loamy, siliceous, subactive, thermic Typic Hapludults) [32] with sand contents ranging from 780 to 830 g kg⁻¹, clay ≥ 70 g kg⁻¹ and silt ≥ 150 g kg⁻¹. This soil type and the sloped topography are not conducive for row crops and other higher-value crops and thus are not typical grown in the area. The forest on a portion of the farm had been recently cleared (five years before the experiment) and fencing installed to graze livestock.

2.2. Farm Management Systems

This ranch grazes water buffalo (Bubalus bubalis (Linnaeus)) in a high-stock density, rotational grazing system where the buffalo are rotated through pasture every sixty-day schedule for 2 days. In this study, fifteen buffalo were moved to new pasture after two days of grazing on one hectare size pasture, resulting in a stock density of 10,500 kg per hectare. In some fields, cover crops are drilled using a Great Plains 1006NT No-Till Drill (Great Plains Manufacturing, Salina, KS, USA) to evaluate soil and pasture responses under different management practices. Cover crops blends are sown twice every year—late in the fall (Hairy vetch Hairy vetch (Vicia villosa Roth), Hazlet cereal rye (Secale cereale L. cv. ‘Hazlet’) and Tetra Sweet perennial ryegrass (Lolium perenne L. cv. ‘Tetra Sweet’). and Medium red clover (Trifolium pratense L.) and early in the spring (Buckwheat (Fagopyrum esculentum Moench), Sunn hemp (Crotalaria juncea L.), Sweet sorghum Sudan (Sorghum bicolor (L.) Moench × Sorghum sudanense (Piper) Stapf), Brown top millet (Urochloa ramosa (L.) T.Q. Nguyen, syn. Brachiaria ramosa L.), and Impact forage collards (Brassica oleracea L. var. acephala DC. cv. ‘Impact’)). Four farm management systems were employed on the farm and evaluated as part of this study. They include,

(1)

grazing ruminants with cover crops (RGCC),

(2)

grazing ruminants on native grasses without seeded cover crops (RGNCC),

(3)

cover crop seeded with no grazing (NGCC) and,

(4)

Orchard with no cover crops and no grazing (NGNCC). The orchard was planted with elderberry trees (Sambucus sp.) at a high density at the start of the experiment in 2022, with trees spacing 1.5 m between rows and 0.5 m within the same row. The orchard was planted with different varieties, however in this study, soil under elderberries were examined.

2.3. Soil Sampling

Prior to the establishment of the experimental plots, 560 soil core samples were taken randomly from 62 plots (Figure 1). Soil samples were collected using a hydraulic soil coring system (Giddings^® Soil Probe, Giddings Machine Company, Windsor, CO, USA), which enables deep and undisturbed core sampling in various soil managements. Sampling was conducted to a depth of 100 cm, segmented into defined depth intervals (e.g., 0–10, 10–20, 20–30, 30–50, and 50–100 cm) to assess soil properties. The Giddings probe was mounted on a tractor, providing the hydraulic force necessary to insert and extract a stainless-steel core barrel. The probe was equipped with a lined coring tube to minimize sample disturbance and facilitate field transport. After extraction, soil cores were labeled and stored in coolers for transport to the laboratory. All equipment was cleaned between sampling points to prevent cross-contamination. All selected fields were sampled twice per year from 2022–2024 in May/June and in November. Samples were collected around the same spot on every sampling date using the marked GPS points. Collected samples from each field were divided into two parts. The first part of each sample location was composited, and the resulting samples were air-dried. Once dried, soils were manually disaggregated using a rubber mallet to break clods and aggregates to pass through 2 mm sieve. No mechanical grinding or milling was performed to preserve soil structure and organic matter integrity. Large roots and visible debris were removed prior to further processing or analysis. Part of these samples were sent to a commercial soil laboratory after processing for total SOC. Total SOC concentration was measured on finely ground air-dried soil samples by dry combustion using LECO TruSpec CN (LECO Corp., St. Joseph, MI, USA) a widely accepted method for quantifying total carbon in soils with high precision and reproducibility [33]. While highly effective, dry combustion at elevated temperatures (typically around 950–1200 °C) can be subject to analytical interferences. These include mass loss from residual moisture, structural water in phyllosilicate clays, and hydration water in minerals such as gypsum [34]. Additionally, the thermal decomposition of magnesium carbonates and dolomite at temperatures below or near combustion range may contribute to overestimation of organic carbon if present [35].

To minimize these potential interferences, all soil samples were air-dried at room temperature and disaggregated using a rubber mallet, without milling, to preserve soil structure and limit artificial exposure of physically protected organic matter. Inorganic carbon was presumed negligible due to the strongly acidic nature of the Typic Hapludult soils at the site, which are typically free of carbonate minerals [32]. To confirm this assumption, a subset of samples was tested with diluted hydrochloric acid; no visible effervescence was observed, indicating the absence of carbonate. Accordingly, carbon measured by dry combustion was interpreted as total soil organic carbon (TOC).

Another portion was used to measure physical properties including bulk density, wet aggregate stability and particle size distribution. Soil bulk density was determined from soil cores collected from the same fields.

2.4. Calculation of Total Organic Carbon (TOC) Sequestration in Soils

TOC was calculated as described by USEPA [36] as follows:

TOC Mg ha⁻¹ = (%C/100) × BD × D × (10,000 m²/ha)

(1)

where

• % C = Mean percent of carbon content in the soil;
• BD = Mean bulk density (in Mg/m³);
• D = Soil layer depth (m);
• m = meters, ha = hectare;
• Mg = mega grams (metric tons).

The sequestration rate of TOC was calculated as the difference in total carbon stocks between sampling dates, divided by the time interval. This calculation was applied separately to each depth interval (0–10, 10–20, 20–30, 30–50, and 50–100 cm) at every sampling location.

2.5. Bulk Density

Bulk density was determined from core samples; each soil depth was removed and dried at 105 °C for 48 h. After drying, the soil was weighed. Soil bulk density was calculated by dividing the mass of dry soil by the volume of the sampler cylinder. However, most samples contained rock fragments. Thus, the USDA formula [37] was used as follows:

Db = (ODW − RF − CW)/[CV − (RF/PD)]

(2)

where

• Db = Bulk density of <2 mm particles (g cm⁻³), ODW = Oven dry samples (g),
• RF = Weight of rocks (g), CW = Weight of empty core (g), CV = Core volume (cm³),
• and,
• PD = Rocks density (g cm⁻³)

2.6. Wet Aggregate Stability

Wet aggregate stability was measured using a wet sieving apparatus (Royal Eijkelkamp, Giesbeek, Netherlands) following methods described by Kemper and Rosenau [38]. Three sieve sizes were used (63, 250, and 500 µm) to assess the distribution of stable aggregates across different size classes. For each sample, 4.0 g of air-dried soil (<2 mm) was placed into the designated sieves and secured in the sieve holder. Distilled water was added to the collection cups until the soil in each sieve was fully submerged. To assess aggregate stability, additional collection cans containing a dispersing solution (2 g NaOH L⁻¹) were included to determine the fully dispersed soil fraction. A wet-sieving apparatus was operated for 10 min, during which sieves were moved vertically in water to separate aggregates by size. Following sieving, the soil-water suspensions were transferred to pre-weighed metal collection cans and oven-dried at 110 °C for 24 h. After drying, each can was reweighed to determine the mass of aggregates in each size fraction. The mass of dry soil was calculated by subtracting the weight of the empty can from the total weight. For samples exposed to the dispersing solution, 0.2 g was subtracted from the final recorded weight to account for solute residue:

Dry Soil Mass = (W₁ − W₀) − 0.2 g

(3)

where W₀ is the weight of the empty can and W₁ is the weight of the can with the dried soil. The mass fractions retained on each sieve were used to determine the distribution of aggregate sizes and calculate the mean weight diameter (MWD), a standard index of soil aggregate stability. MWD was computed according to the method of Kemper and Rosenau [39] as:

MWD = Σ (x_i × w_i)

(4)

where x_i is the mean diameter of each size class and w_i is the proportional mass of aggregates in that class. This approach provides a standardized metric for comparing soil aggregate stability under different treatments.

2.7. Soil Particle Size Distribution

Soil particle size distribution was evaluated by the hydrometer technique [40]. The hydrometer method is based on Stoke’s Law or the fact that particles of different densities have different settling velocities in liquid [41]. A 40 g sample of air-dry, sieved (2 mm) soil was weighed and transferred to a 200 mL container. A 100 mL aliquot of 0.2 M Na-hexametaphosphate was added to the vessel, and the sample was shaken for 16 h at a speed of 250 rpm at room temperature. After 16 h, the container was removed, and the mixed material was transferred to 1000 mL cylinder. Deionized water was added to the cylinder until the water level reached 1000 mL, and the blank sample was included. Once filled, a stopwatch was initialized, and the time was recorded as time 0. A hydrometer was placed floating in the cylinder. The first hydrometer reading was taken at the 40 s on the stopwatch and recorded as Rsand; the temperature was taken and recorded as T. After the readings the hydrometer was removed, moreover, the sample was allowed to continue settling. A second hydrometer reading, Rclay, was chosen according to the temperature from the time 0. The temperature and hydrometer readings were used to calculate the percentage of sand, silt, and clay.

The following equations were used:

Sand (g kg⁻¹) = {[(oven dry soil mass (g)) − (R_sand − R_blank − 40 s)]/oven dry soil mass (g)} × 1000

(5)

Clay (g kg⁻¹) = {(R_clay − R_blank-clay)/oven dry soil mass (g)} × 1000

(6)

Silt (g kg⁻¹) = 1000 − (Sand + Clay)

(7)

2.8. Soil Penetration Resistance

Soil penetration resistance was assessed in January 2023 and 2024 using a FieldScout SC 900 soil compaction meter (Spectrum Technologies, Inc., Aurora, IL, USA). In each plot, measurements were taken at 20 randomly selected locations. Readings were recorded in pounds per square inch (PSI) at depths ranging from 0 cm to 20 cm.

2.9. Statistical Analysis

Analyses of variance (ANOVA) were performed using the General Linear Model (GLM) procedure in Systat software (SigmaPlot 12.3) [42] to evaluate the effects of the experimental factors. Prior to analysis, key ANOVA assumptions were assessed: normality of residuals was evaluated using the Shapiro–Wilk test and visual inspection of Q-Q plots, and homogeneity of variances was tested with Levene’s test. Independence of observations was ensured through experimental design. The means option within the GLM procedure was applied to compute main effect means, and Duncan’s multiple range test was used for pairwise comparisons at a significance level of α = 0.05. Results were summarized in tables that included F-values and p-values. Differences were considered statistically significant when p ≤ 0.05.

3. Results and Discussion

3.1. SOC Dynamics Across Profile Depths

Across the three-year sampling period, there was no significant interaction between time, treatment, and depth for SOC (time × treatment × depth, p = 0.779), nor for time × depth (p = 0.524), treatment × time (p = 0.213), or time alone (p = 0.126) (

Table 1. Analysis of variance (ANOVA) of SOC dynamics in response to different grazing management practices from 2022 to 2024.

Source of Variation	DF	F	p
Treatment	3	7.449	<0.001
Time	5	1.726	0.126
Depth	1	317.898	<0.001
Treatment × Time	15	1.274	0.213
Treatment × Depth	3	3.535	0.015
Time × Depth	5	0.836	0.524
Treatment × Time × Depth	15	0.706	0.779

). This indicates that soil C remained relatively stable across months and years. However, the interaction between grazing management and depth was consistently significant (p = 0.015), showing that different grazing strategies influenced SOC patterns with soil depth.

Increases in SOC under these systems are likely related to reduced soil disturbance, improved water-holding capacity, and enhanced mineralization due to higher plant biomass and litter return to the soil surface. A significant effect of treatment was detected (p <0.001). Post-hoc comparisons indicated that NGNCC exhibited significantly higher mean values than RGNCC (p < 0.001) and RGCC (p = 0.006) but did not differ significantly from NGCC (p = 0.062). NGCC was significantly greater than RGNCC (p < 0.001) and RGCC (p = 0.036). RGCC was also significantly greater than RGNCC (p = 0.006). Depth exerted a strong and consistent influence (power = 1.000). Values at 0–10 cm were substantially greater than those at 10–20 cm across all treatments and times (Figure 2) with values were 2.4 for 0–10 cm and 1.1 for 10–20 cm soil depth. At 0–10 cm, significant differences followed the pattern NGNCC > NGCC, NGNCC > RGCC, RGNCC > NGCC, RGCC > RGNCC, and RGCC > NGCC (p ≤ 0.013). No significant treatment differences were observed at 10–20 cm depth.

By the end of 2024, SOC, an important indicator of soil quality, showed significant variation with grazing managements. The greatest SOC increase in the top 0–10 cm depth was recorded under the orchard (NGNCC) treatment (+28%), followed by RGCC (+13%), NGCC (+7%), and RGNCC (+4%) (Figure 2). Increases in SOC under these systems are likely related to reduced soil disturbance, improved water-holding capacity, and enhanced mineralization due to higher plant biomass and litter return to the soil surface [43]. However, at the 10–20 cm depth, SOC was depleted as follows: under the NGCC (−9%), orchard (NGNCC) treatment (−7%), RGNCC (−6%). followed by RGCC (−2%). These findings align with earlier work show SOC is concentrated in the top 8–15 cm [43]. Undisturbed practices promote slower residue decomposition, greater physical protection of SOC within aggregates, and reduced erosion losses. Higher SOC in topsoil supports microbial activity, enhances nutrient cycling, and improves resilience to climatic extremes. Overall, conservation-oriented orchard (NGNCC) followed RGCC and NGCC management offers the greatest potential for sustaining and enhancing soil C stocks. Furthermore, long-term conservation practices promote slower residue decomposition, greater physical protection of SOC within aggregates, and reduced erosion losses.

Several researchers examined grazed and non-grazed soils and discovered that grazed areas had greater SOC concentrations than non-grazed. For example, Schuman et al. [44] reported that grazing a northern mixed-grass prairie for 12 years, at both light and high stocking rates, increased SOC mass in the top 30 cm compared to ungrazed controls. Povirk [45] observed a marked increase in SOC in alpine meadows grazed by sheep in Wyoming’s Medicine Bow National Forest, with SOC averaging 11% in grazed areas compared to 6.3% in ungrazed areas in the 0–7.5 cm layer. While conducting our research, we discovered that the sort of management used on those soils, which were seasonal or perennial grass rather than cover crops or trees, may have contributed to the disparities.

3.2. Soil Organic Carbon Sequestered (SOCS) in the Soil Profile

The ANOVA results show that both grazing management treatment and soil depth had highly significant effects on soil carbon sequestration (p < 0.001) (

Table 2. Analysis of variance (ANOVA) of soil carbon sequestration under different grazing management practices from 2022 to 2024.

Source of Variation	DF	F	p
Year	1	1.375	0.242
Treatment	3	142.415	<0.001
Depth	4	56.195	<0.001
Year × Treatment	3	0.6	0.615
Year × Depth	4	0.577	0.679
Treatment × Depth	12	56.433	<0.001
Year × Treatment × Depth	12	0.337	0.982

). This indicates that different grazing practices and soil layers influence the amount of carbon stored in the soil. The interaction between treatment and depth was also significant (p < 0.001), suggesting that the effect of grazing management on carbon sequestration varies depending on soil depth. In contrast, the effect of year and its interactions with treatment and depth were not statistically significant (p > 0.05), indicating that soil carbon sequestration patterns remained consistent over the study period from 2022 to 2024.

Grazing management significantly affected C storage (p < 0.05). In surface soils (0–10 cm), NGNCC stored the most C (0.99 Mg ha⁻¹ yr⁻¹), more than 100 times than RGNCC (−0.023 Mg ha⁻¹ yr⁻¹). Soil C sequestration rates varied markedly with both management practice and soil depth (Figure 3). In the surface layer (0–10 cm), NGCC and NGNCC had the highest sequestration rates (~1.0–1.2 Mg ha⁻¹ yr⁻¹), followed by RGCC and RGNCC. In the 10–20 cm layer, NGCC maintained high sequestration (~1.6 Mg ha⁻¹ yr⁻¹), whereas both RGCC and RGNCC recorded near-zero to slightly negative values, indicating net C loss at this depth.

At intermediate depths (20–30 cm and 30–50 cm), RGCC had the highest sequestration rates (~0.8 and ~1.8 Mg ha⁻¹ yr⁻¹, respectively), with NGNCC performing similarly at 30–50 cm. NGCC showed moderate values at these depths, while RGNCC remained comparatively low. Notably, RGCC displayed an atypical pattern, maintaining low C Sequestration at shallow and mid-depth but showing a marked increase at 50 cm (Figure 3), exceeding its values at all shallower layers. These depth-dependent patterns indicate that the relative performance of treatments is not constant through the soil profile, and that interpreting treatment main effects without accounting for depth would obscure critical biological and management-relevant differences. In the deepest layer (50–100 cm), NGNCC exhibited the highest sequestration (~2.2 Mg ha⁻¹ yr⁻¹), followed by NGCC (~1.0 Mg ha⁻¹ yr⁻¹), with RGCC and RGNCC showing low values (<0.3 Mg ha⁻¹ yr⁻¹). The results also suggest that RGCC’s advantage is maximized at shallower depths, while certain treatments may perform unexpectedly well under specific deeper conditions.

When summed across all five sampling depths, total soil C ranked as follows: NGCC (4.88 Mg ha⁻¹ yr⁻¹) > NGNCC (4.37) > RGCC (3.13) > RGNCC (0.03). Differences in C storage were also significant through the upper 100 cm of the profile (p < 0.0001). Overall, NGCC and NGNCC tended to outperform RGCC and RGNCC in the surface layers, while RGCC showed strong performance at intermediate depths, and NGNCC excelled at depth.

Grazing impacts persisted in the top 100 cm of the soil profile (p < 0.0001), with NGCC outperforming other managements due to less disturbance and better residue retention. NGNCC and NGCC stored considerably more SOC than RGNCC, consistent with previous reports showing that long-term disturbance accelerates SOC losses and mineralization [46,47]. Derner et al. [48] found greater SOC storage in grazed shortgrass steppe in northeastern Colorado, with 1983 g C m⁻² in grazed plots versus 1321 g C m⁻² in ungrazed plots in the 0–15 cm layer, and no significant differences at 15–30 cm depth. Povirk [45] observed a marked increase in SOC in alpine meadows grazed by sheep in Wyoming’s Medicine Bow National Forest, with SOC averaging 11% in grazed areas compared to 6.3% in ungrazed areas in the 0–7.5 cm layer.

The higher C accumulation under orchard (NGNCC) compared to NGCC and RGNCC likely reflects reduced disturbance, higher root biomass input, and greater residue retention, all of which enhance organic C stabilization [40]. Perennial fruit tree agroecosystems have emerged as significant C sinks, demonstrating substantial potential for mitigating atmospheric CO₂ through C sequestration. Empirical data from Italian apple orchards and vineyards report net ecosystem C gains of up to 4.30 and 7.5 Mg C ha⁻¹ yr⁻¹, respectively [49]. In parallel, research conducted in Sicily indicates that intensively managed high-density orange groves (with sequestration rates of approximately 1.8 Mg C ha⁻¹ yr⁻¹) sequester nearly quadruple the C compared to traditional low-density plantations (0.5 Mg C ha⁻¹ yr⁻¹) [50]. This disparity is primarily attributed to variations in soil respiration rates and C flux dynamics within the soil-plant system [50]. These findings underscore the critical influence of orchard design and management on enhancing soil C stocks and mitigating climate change.

3.3. Aggregate Stability Under Different Managements

Analysis of variance (ANOVA) revealed significant effects of year, grazing management treatment, and aggregate size on wet aggregate stability (WAS) (

Table 3. Analysis of variance of different parameters on wet aggregate stability under different grazing management from 2022 to 2024.

Source of Variation	DF	F	p
Year	2	5.713	0.004
Treatment	3	5.59	<0.001
Size	2	118.191	<0.001
Year × Treatment	6	1.217	0.296
Year × Size	4	1.959	0.1
Treatment × Size	6	2.316	0.033
Year × Treatment × Size	12	1.801	0.046

). Specifically, aggregate size exhibited the strongest effect (p < 0.001), indicating that stability varied markedly among different aggregate size classes. Both years (p = 0.004) and treatment (p < 0.001) significantly influenced WAS, suggesting temporal variability and treatment-driven changes in soil structure. While interactions between year and treatment or year and size were not significant (p > 0.05), the interaction between treatment and size was significant (p = 0.033), indicating that the impact of grazing management on WAS depended on aggregate size. Additionally, the three-way interaction of year, treatment, and size was significant (p = 0.046), reflecting complex temporal and treatment-related dynamics across aggregate sizes. These results suggest that grazing management practices influence soil structural stability differently across aggregate scales over time.

Across all Years and Size fractions, aggregate stability differed significantly among most Treatments (all p < 0.001), except between NGNCC and NGCC (p = 0.314). RGNCC had significantly lower stability compared to NGCC, NGNCC, and RGCC (all p < 0.001). Similarly, RGCC exhibited significantly lower stability than NGCC and NGNCC (both p < 0.001). The RGNCC treatment caused substantial soil disturbance, which reduced surface aggregate formation and stability. In contrast, as SOC concentrations increased, aggregate stability improved more under NGCC and NGNCC systems due to enhanced soil structure and aggregate formation by minimizing disturbance frequency and maintaining high residue cover, thereby reducing erosion and aggregate breakdown [51]. Also, these results indicate that NGCC and NGNCC tended to maintain higher stability than other treatments without cover crops (RGNCC, RGCC), although the magnitude and direction of differences were dependent on Size and Year (Figure 4). Improved aggregate stability specific treatment in soils is most likely owing to considerable increases in soil organic matter. Increasing the amount of organic matter in the soil promotes the production of stable aggregates. This increase in aggregate stability could also be attributed to the large surge in microbial activity reported under examined soils. Microbial activity in soil has been found to promote soil aggregate formation [16,52]. Good aggregate stability increases pore space, which improves water and air infiltration and enables deeper root exploration [53].

Across all three Size levels (63, 250, and 500), the Year × Treatment interaction was statistically significant (p < 0.001 in each case), although the nature of the interaction varied by Size. For Size = 63, the interaction was highly significant, with pronounced Year effects observed within RGNCC (2022 > 2023 > 2024; all pairwise p < 0.001 except 2023–2024, ns) and RGCC (all pairwise Year differences significant, p < 0.001), whereas NGNCC and CCNG showed no significant Year effects. In 2022, RGCC exhibited significantly higher Stability than NGNCC (p = 0.017) and RGNCC (p = 0.004), but no significant Treatment differences were found in 2023 or 2024. For Size = 250, although the interaction reached significance, simple main effects revealed no strong Year differences within any Treatment, with only marginal Treatment effects in 2022 (p = 0.024) and 2024 (p = 0.022) of small magnitude. At Size = 500, a significant Year × Treatment interaction emerged only for RGCC, where 2024 Stability was lower than in both 2022 and 2023 (p < 0.001), while other Treatments showed no Year effects (2022: p = 0.206; 2023: p = 0.602; 2024: p = 0.481), indicating this coarser fraction was less responsive to year-to-year treatment variation. Descriptively, Stability declined sharply with increasing Size (LS means: 63 = 0.496, 250 = 0.323, 500 = 0.139) and showed an overall downward trend across Years (2022 = 0.348, 2023 = 0.335, 2024 = 0.275), though patterns differed across Treatment × Size combinations. On average, NGNCC (0.364) and NGCC (0.349) tended to have higher Stability than RGCC (0.277) and RGNCC (0.287), but these differences were inconsistent. Rotational grazing was found to substantially improve fine aggregate stability but had little effect on coarse aggregate stability, except for a notable reduction under light grazing [54]. Light grazing significantly increased the total content of coarse aggregates in bulk soil but did not affect the proportion of coarse stable aggregates. In contrast, medium and high grazing intensities did not enhance coarse aggregate proportions, likely because higher grazing pressure increases trampling, compressing aggregates.

Dong et al. [54] showed that both medium term (21 years) and long-term (34 years) grazing exclusion considerably increased soil aggregate quantity and stability. Grassland soil quality is closely linked to plant productivity [55]. Greater above- and belowground biomass-primarily through plant litter inputs-can positively influence soil physical and chemical properties in semi-arid grasslands [56]. Aggregate stability is strongly related to root activity [57], with SOM acting as a key binding agent in aggregate formation [58]. Higher primary productivity leads to increased root exudates [59], further improving aggregate stability. However, soil structure disruption from animal trampling during grazing can increase bulk density and reduce aggregate stability and SOC concentration [60]. These results are consistent with Wang et al. [61], who found that light grazing encourages coarse aggregate development, whereas higher intensities lead to aggregate breakdown. Increased grazing pressure can crush coarse aggregates [15], offsetting potential gains from enhanced root biomass. Overall, our findings suggest that, at least in the short term, rotational grazing has limited effects on coarse aggregate stability, while benefits are more evident in the fine fraction.

3.4. Soil Penetration Resistance

Analysis of variance revealed significant effects of year (F = 110, p < 0.001), depth (F = 544, p < 0.001), and treatment (F = 136, p < 0.001) on penetration resistance (

Table 4. Analysis of variance of soil penetration resistance under different grazing management from 2023 to 2024.

Source of Variation *	DF	F	p
Year	1	110	<0.001
depth	8	544	<0.001
treatments	3	136	<0.001

* No interactions.

). These results indicate that soil compaction was strongly influenced by management practice, sampling depth, and interannual variation. Pairwise comparisons revealed significant differences in soil penetration resistance across all depth intervals (p < 0.001). The greatest differences were observed between the deepest measurement (20 cm) and the shallower depths, with the magnitude of differences progressively decreasing as depths became closer. This pattern indicates a clear and consistent depth-related gradient in penetration resistance, reflecting progressive changes from the soil surface downward.

Soil penetration resistance increased with depth across all treatments and years, frequently reaching the penetrometer limit (350 psi) at depths ≥15 cm (Figure 5). In both years, shallow layers (0–5 cm) exhibited substantially lower resistance than deeper layers. Across treatments, RGCC and NGCC generally maintained lower penetration resistance in the upper 5 cm compared with non-cover crop treatments. This pattern was most pronounced in 2023, when NGNCC and RGNCC recorded the highest values at the surface of the soil. Over the years, penetration resistance at 0–5 cm was generally lower in 2024 than in 2023 for all treatments, suggesting reduced surface compaction. Mid-depth layers (7.5–12.5 cm) also tended to show lower resistance in cover crop treatments, particularly under NGCC.

Soil compaction can negatively alter soil structure, restrict air and water infiltration, reduce porosity, and increase bulk density [62]. Such changes have adverse effects on water and air movement through the root zone [63] and can impede root development [64]. Improved water infiltration combined with greater water-holding capacity can help reduce runoff and erosion in soils. In agricultural systems, compaction is primarily caused by livestock movement and the operation of farm machinery. Even routine machinery traffic can induce mild to severe compaction in the upper soil layer, while increased machinery weight can compact deeper layers [65].

4. Conclusions

This study evaluated the dynamics of soil C and C sequestration and soil physical parameters under various agricultural management practices. Individual parameters used to derive soil quality showed inconsistent responses across different grazing management systems. When examining the effects of soil disturbance on soil properties, the RGNCC system exhibited significantly higher bulk density and lower SOC than the NGNCC system at all sampling depths. NGNCC, RGCC, NGCC, and RGNCC approaches influenced aggregate size distribution and water-stable aggregates, with the most pronounced differences occurring in the aggregate size classes.

Overall, differences in management practices were associated with measurable changes in soil properties, directly influencing soil functionality. Intensive grazing contributed to soil particle breakdown, reduced soil organic matter content, and disrupted nutrient cycling—processes that collectively drive long-term soil degradation and reduced productivity.

Given the extensive coverage of grasslands, even small adjustments in management aimed at reducing carbon emissions could have a substantial impact on national and global carbon budgets. In this study, grazing on blended cover crop plots produced greater carbon inputs to the soil compared with ruminants grazed on native grasses without cover crops treatments, likely due to the combined effects of plant species diversity and higher plant density. Furthermore, perennial fruit tree systems demonstrate a high capacity for carbon sequestration, underscoring their potential role in climate mitigation strategies. These findings collectively highlight that management practices—whether through perennial cropping systems, or grazing regimes—can substantially alter soil structural stability and carbon dynamics. Recognizing these relationships is essential for designing the experimental framework and selecting management treatments that optimize soil function while enhancing carbon sequestration potential.

Collectively, the evidence emphasizes that tailored management practices with cover crops, attentiveness to stocking densities, paddock rotation, and orchard management, are critical to sustaining soil function, enhancing carbon storage, and promoting ecosystem resilience. Prioritizing these management approaches is essential for sustaining productive soils and mitigating climate change through increased carbon storage.