Long-Term Flat-Film Hole-Sowing Increases Soil Organic Carbon Stocks and Resilience Under Future Climate Change Scenarios

Abstract

Analyzing the soil organic carbon (SOC) stock in dryland areas of southern Shanxi, particularly under the influence of fertilization and mulching conditions, is crucial for enhancing soil fertility and crop productivity and understanding the SOC pool’s resilience to future climate change scenarios in the region. In a long-term experimental site located in Hongtong County, Shanxi Province, soil samples were collected from the 0–100 cm depth over a nine-year period. These samples were analyzed to evaluate the impact of five treatments: no fertilization and no mulching (CK), conventional farming practices (FP), nitrogen reduction and controlled fertilization (MF), nitrogen reduction and controlled fertilization with ridge-film furrow-sowing (RF), and nitrogen reduction and controlled fertilization with flat-film hole-sowing (FH). The average annual yield of wheat grain, SOC stock, water-soluble organic carbon (WSOC), particulate organic carbon (POC), light fraction organic carbon (LFOC), mineral-associated organic carbon (MOC), and heavy fraction organic carbon (HFOC) stocks were measured. The results revealed that the FH treatment not only significantly increased wheat grain yield but also significantly elevated the SOC stock by 23.71% at the 0–100 cm depth compared to CK. Furthermore, this treatment significantly enhanced the POC, LFOC, and MOC stocks by 106.43–292.98%, 36.93–158.73%, and 17.83–81.55%, respectively, within 0–80 cm. However, it also significantly decreased the WSOC stock by 34.32–42.81% within the same soil layer and the HFOC stock by 72.05–101.51% between the 20 and 100 cm depth. Notably, the SOC stock at the 0–100 cm depth was primarily influenced by the HFOC. Utilizing the DNDC (denitrification–decomposition) model, we found that future temperature increases are detrimental to SOC sequestration in dryland areas, whereas reduced rainfall is beneficial. The simulation results indicated that in a warmer climate, a 2 °C temperature increase would result in a SOC stock decrease of 0.77 to 1.01 t·ha⁻¹ compared to a 1 °C increase scenario. Conversely, under conditions of reduced precipitation, a 20% rainfall reduction would lead to a SOC stock increase of 1.53% to 3.42% compared to a 10% decrease scenario. In conclusion, the nitrogen reduction and controlled fertilization with flat-film hole-sowing (FH) treatment emerged as the most effective practice for increasing SOC sequestration in dryland areas by enhancing the HFOC stock. This treatment also fortified the SOC pool’s capacity to withstand future climate change, thereby serving as the optimal approach for concurrently enhancing production and fertility in this region.

1. Introduction

Drylands, which account for approximately 80% of the world’s arable land, are the cornerstone of global food security, providing 60% of the world’s food supply [1]. As the world population continues to grow and the demand for food intensifies, dryland agriculture emerges as a critical frontier in agricultural development [2]. However, these regions are characterized by low and unevenly distributed rainfall and nutrient-poor soils, particularly with respect to organic matter [3]. These factors significantly constrain the development of dryland agriculture. Film-mulching is a principal strategy to combat drought, conserve water, increase productivity, and secure income in arid regions [2]. Moreover, there are pressing issues related to excessive nitrogen application by farmers in these areas [4]. Therefore, when combined with optimized fertilization, film-mulching not only results in the effective retention of moisture and suppression of evaporation but also mitigates the risk of environmental pollution, making it a widely adopted practice in these areas.

The soil organic carbon (SOC) pool in farmlands is the most pivotal carbon reservoir within the global terrestrial ecosystem. It not only serves as a crucial indicator of soil fertility and quality but also plays a decisive role in maintaining the productivity and ecological stability of farmland. Moreover, given its substantial capacity, with a carbon stock of approximately 170 Pg, it accounts for more than 10% of the global terrestrial carbon stock [5,6]. Even minor fluctuations in this stock can have profound impacts on the ecosystem carbon cycle and global climate. Therefore, increasing the sequestration of SOC in farmlands is vital for improving soil quality, ensuring food security, and achieving the agricultural goals of “carbon peaking” and “carbon neutrality”.

Although SOC is a key determinant of long-term soil productivity and health [7], it is relatively insensitive to short-term changes in soil quality [8]. In contrast, labile organic carbon fractions, including water-soluble organic carbon (WSOC), particulate organic carbon (POC), and light fraction organic carbon (LFOC), tend to be more responsive to changes in soil quality under agricultural management practices [9,10]. These labile fractions represent organic matter that is in transition from fresh plant residue to stable organic matter, with turnover times generally within a decade [11]. Since SOC is a composite of various carbon pools, measurements of a single labile carbon fraction cannot fully reflect the changes in soil quality induced by management practices. Simultaneous measurements of multiple labile fractions can be a more effective indication of how field management practices impact soil quality [12]. In addition to these rapidly cycling labile fractions, there are also fractions with relatively slow turnover times in the soil, such as mineral-associated organic carbon (MOC) and heavy fraction organic carbon (HFOC). These fractions are considered resistant organic carbon and are derived primarily from plant compounds, secondary microbial metabolites, and microbial residues, playing a significant role in SOC sequestration [13].

Previous research on the effects of management practices, such as fertilization and film-mulching, on SOC levels and organic carbon fractions has been focused predominantly on topsoil (0–20 cm) [14,15,16], but no consensus has been reached. The subsoil contains a substantial portion of the SOC and is responsive to changes in fertilization and other field management practices [17]. In fact, the SOC fraction in the subsoil may surpass that in the topsoil, and this SOC pool can potentially have a large contribution to the overall SOC reserves [18]. Radiocarbon dating showing increased mean residence times of SOC with depth suggests that deep soil C is inherently resistant to decomposition [19]. The subsoil generally contains greater reactive surface areas, and soil organic matter exists there predominately in organo-mineral complexes, which are considered a key mechanism for long-term stabilization of SOC [20,21]. Therefore, it is necessary to pay close attention to changes in deep SOC pools.

Soil organic carbon sequestration is a dynamic process, and model simulation offers a rapid and efficient approach for studying SOC changes. The denitrification-decomposition (DNDC) model is an SOC and total nitrogen (TN) cycle model that relies on inputs of meteorological data, soil properties, and farmland management practices. Its applicability has been validated in numerous studies [22,23,24]. Valkama et al. [22] suggested that the DNDC model can be used to accurately simulate and evaluate the positive effects of conservation tillage on SOC sequestration. Zhang et al. [24] also reported that the DNDC model can effectively simulate the dynamics of SOC under wheat-corn rotation systems in northern China. Therefore, the use of the DNDC model to simulate and predict dynamic changes in SOC plays a significant role in investigating the trends in farmland SOC under various management practices and evaluating changes in SOC sequestration efficiency.

The Loess Plateau, a quintessential semiarid agricultural region in northwestern China, is characterized by a dry climate and nutrient deficiencies that limit crop growth and can even lead to reduced yields. Therefore, plastic film-mulching and the optimization of fertilization practices are instrumental in unlocking the productive potential of dryland agriculture and fostering green transformation and development in this region. In recent years, studies have reported the effects of film-mulching combined with optimized fertilization on crop yield, water and nutrient utilization efficiency, and the mechanisms of organic carbon sequestration within aggregates [25,26,27,28,29,30,31,32]. However, the responses of different SOC fractions in soil profiles under various agricultural management practices are not yet fully understood. Since changes in total SOC, especially in resistant organic carbon pools, involve long-term processes, the complex questions surrounding SOC pool changes can only be addressed through long-term experiments. Therefore, long-term fertilization and film-mulching experiments were conducted on the Loess Plateau starting in 2011. The specific objectives of this study were to (i) assess the impact of long-term fertilization and film-mulching practices on the SOC stock and its fractions, (ii) identify the fractions that contribute most to SOC sequestration, and (iii) evaluate predicted changes in SOC pools over the next 85 years via the DNDC model.

2. Materials and Methods

2.1. Study Site and Experimental Design

A long-term field experiment was established in 2011 in the Hongtong County (36°22′ N, 111°35′ E, with an altitude of 648 m a.s.l.), Loess Plateau, Shanxi, China (Figure 1). The site has a warm temperate continental monsoon climate with a mean annual temperature of 12.6 °C and a mean precipitation of 500 mm (more than 60% of which falls from July to September). Experimental plots (each 360 m²) were cultivated with the winter wheat (Triticum aestivum L.) and summer fallow regime, which represents one of the main cropping systems in China. The soil at the site is classified as a Calcaric Cambisol (IUSS Working Group WRB 2015). The major soil chemical properties from 0 to 20 cm depth in 2011 were SOC content of 8.5 g kg⁻¹, total N content of 0.8 g kg⁻¹, NO₃-N content of 10.4 mg kg⁻¹, Olsen-P content of 10.4 mg kg⁻¹, exchangeable K⁺ content of 208.2 mg kg⁻¹, and a pH of 7.9.

The field experiment had a completely randomized block design with five treatments: (1) No fertilization and no mulching (CK). (2) Conventional farming practices (FP), which align with local wheat planting practices, involving the cultivation of winter wheat using conventional flat farming methods. Fertilizer was applied before sowing via rotary tillage without mulching. The amount of fertilizer applied was the average amount of fertilizer applied by local farmers with annual inorganic fertilizer inputs of 150 kg ha⁻¹ nitrogen (N) and 60 kg ha⁻¹ phosphorus (P). (3) Nitrogen reduction (33.33%) and controlled fertilization (MF), in which conventional flat farming was employed with fertilizer being applied via rotary tillage before sowing without mulching. The amount of fertilizer applied was calculated based on the soil nutrient balance and the soil nutrient test value before wheat sowing each year. These plots received 100 kg ha⁻¹ N, 83 kg ha⁻¹ P, and 36 kg ha⁻¹ potassium (K). (4) Nitrogen reduction (33.33%) and controlled fertilization with ridge-film furrow-sowing (RF), in which rotary tillage was used to apply fertilizer before sowing and then ridges were formed. The ridges were covered with film, and seeds were sown in the furrows. The ridge width was 35 cm, and the furrow width was 30 cm. The amount of fertilizer applied was consistent with that applied under MF. (5) Nitrogen reduction (33.33%) and controlled fertilization with flat-film hole-sowing (FH), in which rotary tillage was used to apply fertilizer before sowing and then mulching film was applied over the entire ground surface and covered with 0.5–1 cm of soil. The row spacing was 15–16 cm and the hole spacing was 12 cm, with the amount of fertilizer applied identical to that applied under MF. Nitrogen, phosphorus, and potassium fertilizers were applied in the form of urea, calcium superphosphate, and potassium chloride, respectively. All chemical fertilizers were incorporated into the soil via rotary tilling to a depth of approximately 15–20 cm while winter wheat was sown. Controlled fertilization was employed to ensure that the soil nutrient content in each treatment was consistent. Winter wheat was sown in early October and harvested early in the following June. The sowing rate in all the treatments was 150 kg ha⁻¹. Film-mulching was performed throughout the growth period of the winter wheat with no irrigation. At maturity, crops were harvested manually with sickles, cutting close to the ground, from three areas of 30 m² (15 m × 2 m), to estimate yields for each treatment. All above-ground crop residues were removed after harvest. The fields were conventionally tilled.

2.2. Soil Sampling and Analysis

Soil samples were taken at random from each plot at five depths of 0–20, 20–40, 40–60, 60–80, and 80–100 cm using a soil drill in June 2020. Six soil samples were taken from each plot. First, surface organic materials and fine roots were carefully removed by softly crushing the soil by hand and removing the materials with tweezers. Next, the soil was air-dried and ground to pass through a 2 mm sieve for separation in preparation for the analysis of WSOC, POC, LFOC, MOC, and HFOC. Subsamples of air-dried <2 mm soil were further ground and passed through a 0.15 mm sieve for the determination of SOC and total nitrogen (TN) via a C and N analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).

POC (0.053–2 mm) and MOC (<0.053 mm) were obtained from 2 mm-sieved soil samples following the procedure of Camberdella and Elliott [33]. Briefly, 20 g of air-dried soil was dispersed in 30 mL of sodium hexametaphosphate (Na)₆(PO₃)₆ (5 g L⁻¹) with shaking on a reciprocating shaker for 15 h. The resulting soil suspension was poured over a 0.053 mm screen under a flow of distilled water to ensure separation. All materials remaining on the 0.053 mm screen (POC) and passing through the 0.053 mm screen (MOC) were washed into a pre-weighed aluminum box, dried at 60 °C, weighed, and stored for analysis. The POC and MOC were passed through a 0.15 mm screen and analyzed by a C and N analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).

LFOC and HFOC were determined via methods described by Gregorich and Ellert [34]. A total of 10 g of air-dried soil was placed in a centrifuge tube with 20 mL of NaI solution (with a specific gravity of approximately 1.80 g cm⁻³). The tubes were shaken on a reciprocating shaker for 60 min and then centrifuged at 1000× g for 15 min. The floating material was aspirated and filtered through a 0.45 μm filter membrane. The material on the filter (LFOC) was rinsed with deionized water to remove any sodium iodide that remained, and the material was placed in a pre-weighed aluminum box, dried at 60 °C, and weighed. The sediment remaining in the centrifuge tube was rinsed with deionized water. This procedure was repeated to ensure that all sodium iodide was completely removed. Afterward, the sediments (HFOC) were transferred to a pre-weighed aluminum box, dried at 60 °C, and weighed. The LFOC and HFOC were passed through a 0.15 mm screen and analyzed by a C and N analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).

WSOC was determined via a method described by Li et al. [35]. Briefly, 3 g of air-dried soil (2 mm) was placed in a centrifuge tube with 30 mL of deionized water. The tubes were shaken at 180 rpm at 50 °C for 1 h and centrifuged at 3500 rpm for 15 min. The supernatant was filtered by quantitative filter paper and analyzed by a C and N analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).

2.3. The DNDC Model

2.3.1. Model Parameter Sources

The DNDC model requires the following input data: (1) daily weather data, such as daily maximum and minimum air temperature (°C) and daily precipitation (mm) data, (2) initial soil conditions (e.g., soil organic carbon, soil nitrate, and ammonium N contents), (3) soil data (e.g., soil water at the field capacity, wilting limit, soil bulk density, soil pH, and soil texture), and (4) field management data (e.g., planting dates, harvest dates, crop parameters, tillage rates, fertilization rates, and fertilizer types). In order to run our modeling approach, we collected and projected meteorological data for the study area spanning from 2015 to 2099, including daily maximum temperature (°C), daily minimum temperature (°C), and daily precipitation (mm). Data were obtained from the official website of the CMIP6 “https://esgf-node.llnl.gov/search/cmip6/ (accessed on 26 February 2024)”. The soil and crop management inputs were based on information gathered during each experiment.

The future climate scenarios employed in our study are sourced from the Coupled Model Intercomparison Program, Phase 6 (CMIP6), of the World Climate Research Program (WCRP), specifically from its four Shared Socio-Economic Pathway scenarios (SSP1–2.6, SSP2–4.5, SSP3–7.0, and SSP5–8.5). Utilizing the atmospheric circulation model CanESM2 RCP2.6, we generated a future climate scenario to simulate and project changes in SOC stocks spanning 85 years (2015–2099). The RCP2.6 presents the lowest-emission scenario for greenhouse gases (GHGs) and radiative forcing, capping global mean warming at 2 °C by 2100. This scenario anticipates negative energy emissions in the latter part of the century, accompanied by a reduction in radiative forcing decreases to 2.6 W m⁻² by 2100, and an atmospheric CO₂ concentration of approximately 490 ppm. Detailed information about these climate models and scenarios is available at “https://esgf-node.llnl.gov/search/cmip6/ (accessed on 26 February 2024)”. The weather file was organized to include the date, daily maximum and minimum temperatures (°C), and daily rainfall (mm), formatted according to the DNDC weather specifications, and then used to run the calibrated DNDC model.

2.3.2. Model Performance Statistical Evaluation

To evaluate the DNDC model performance, four deviation statistics were employed to compare the simulated and measured crop SOC values. Variables included the mean bias error (MBE), normalized root mean square error (n RMSE), standard error (E), and index of agreement (d) [36,37].

The MBE value is the difference between the means of the simulated variable and the measured variable. Based on recommendations developed in previous studies [38,39], we considered model performance to be “excellent” when the n RMSE was ≤10%, the performance to be “good” when 10% < n RMSE ≤ 20%, the performance to be “fair” when 20% <n RMSE ≤ 30%, and the performance to be “poor” when the n RMSE was >30%. The standard error E indicates the mean value of the error between the simulated and measured values: a smaller absolute value corresponds to a better model fit, a value of less than 5% corresponds to a high model fit, and a value of 5–10% corresponds to an acceptable model fit [40]. The d (0 to 1) index is a dimensionless and bounded measure, with higher values indicating better agreement between simulations and measurements [36]. Based on recommendations from Liu et al. [41], a value of d ≤ 0.7 was considered to indicate a “poor” agreement, a value of 0.7 ≤ d ≤ 0.8 indicated a “fair” agreement, a value of 0.8 ≤ d < 0.9 was characterized a “good” agreement, and a value of d ≥ 0.9 was classified as an “excellent” agreement between the simulated and measured values:

$$MBE={\displaystyle \frac{{\sum }_{i=1}^n(S_i-M_i)}{n}}$$

(1)

$$nRMSE={\displaystyle \frac{\sqrt{{\sum }_{i=1}^n{(S_i-M_i)}^2/n}}{\overline{M}}}\times 100$$

(2)

$$\mathrm{E}=\sum _{\mathrm{i}=1}^{\mathrm{n}}{\displaystyle \frac{(M_i-S_I)}{\overline{M}}}\times {\displaystyle \frac{100}{n}}$$

(3)

$$d=1-{\displaystyle \frac{{\sum }_{i=1}^n{(S_i-M_i)}^2}{{\sum _{i=1}^n\left(\right|S_i-\overline{M}|+|M_i-\overline{M}\left|\right)}^2}}$$

(4)

where S_i is the simulated value, M_i is the measured value, n is the number of measured values, and $\overline{\mathrm{M}}$ is the mean of the measured values.

2.3.3. Data Calculation and Statistical Analysis

The SOC/WSOC/LFOC/HFOC/POC/MOC pool was calculated according to Holeplass et al. [42] (Equation (5)):

SOC/WSOC/LFOC/HFOC/POC/MOC stock (Mg ha⁻¹) = (g SOC/WSOC/LFOC/HFOC/POC/MOC kg⁻¹)/1000 × bulk density (g cm⁻³) × 10,000 (m² ha⁻¹) × soil depth (m)

(5)

Significant differences between the means were identified via one-way analysis of variance for each year, and the least significant difference (LSD) was computed to compare the means of the above variables (p < 0.05) via SPSS software v.18 (IBM, Armonk, NY, USA). The percentage increase in the mean square error (MSE) of the random forest model was used to predict the main organic carbon fraction affecting soil organic carbon sequestration. Graphs were prepared via Origin Software v.8.1 (OriginLab, Northampton, PA, USA).

3. Results

3.1. Wheat Grain Yield

The annual wheat grain yields in the control (CK), FP, MF, RF, and FH treatments ranged from 1367 to 3797, 1647 to 4917, 1554 to 4614, 2033 to 5816, and 2645 to 7355 kg ha⁻¹, respectively, during the experimental period (2011–2020), and the mean grain yields were 2867, 3453, 3579, 4386, and 5453 kg ha⁻¹, respectively (Figure 2). The annual wheat grain yield trends for each treatment were consistent across the years. On average, compared with the CK treatment, the FP, MF, RF, and FH treatments significantly increased the wheat grain yield by 20%, 25%, 53%, and 79%, respectively. Notably, compared with the MF treatment, the RF and FH treatments further increased yields by 22% and 42%, respectively.

3.2. Soil Organic Carbon Stock

Compared with that of the initial soil, the soil organic carbon (SOC) stock under the CK treatment significantly decreased, ranging from 25.05% to 46.64% at the 20–100 cm depth (Figure 3). The FP treatment notably increased the SOC stock by 4.60% to 14.12% at the 0–40 cm depth, but it led to a significant decrease of 13.26% to 22.28% at the 40–100 cm depth. Conversely, the RF treatment resulted in a significant decrease in the SOC stock by 14.12% to 25.60% at the 60–100 cm depth. In contrast, the FH treatment significantly augmented the SOC stock by 16.61% at the 0–40 cm depth and by 23.71% at the 80–100 cm depth.

3.3. Labile Carbon Fractions

3.3.1. Water-Soluble Organic Carbon

Compared with the CK treatment, the FP, MF, and RF treatments significantly reduced the water-soluble organic carbon (WSOC) stock by 13.46–49.19%, 14.46–50.87%, and 15.49–52.04% at the 0–100 cm depth, respectively, except the RF treatment at the 20–40 cm depth (Figure 4a). The FH treatment also significantly decreased the WSOC stock by 34.32% to 42.81% at the 0–80 cm depth, whereas it notably increased the WSOC stock by 15.40% at the 80–100 cm depth.

3.3.2. Particulate Organic Carbon

In contrast to the CK treatment, the FP treatment substantially increased the particulate organic carbon (POC) stock by 82.37% to 99.80% at the 0–40 cm depth (Figure 4b). Similarly, the RF treatment significantly increased the POC stock by 82.37% to 103.45% at the same soil depth, but it notably reduced the POC stock by 54.58% at the 60–80 cm depth. The MF treatment only significantly increased the POC stock by 59.28% at the 0–20 cm depth. The FH treatment markedly increased the POC stock by 106.43% to 292.98% across the 0–100 cm depth, except for the 60–80 cm depth.

3.3.3. Light Fraction Organic Carbon

Compared with that in the CK treatment, the light fraction organic carbon (LFOC) stock significantly increased in the FP treatment, with increases ranging from 70.64% to 165.86% at the 0–60 cm depth (Figure 4c). The MF treatment substantially increased the LFOC stock by 109.12% at the 20–40 cm depth and by 253.53% at the 60–80 cm depth. Both the RF and FH treatments significantly increased the LFOC stock by 73.96–164.51% and 36.93–158.74% at the 0–40 cm and 60–80 cm depths, respectively.

3.4. Recalcitrant Carbon Fractions

3.4.1. Mineral-Associated Organic Carbon

In contrast to the CK treatment, the FP, MF, RF, and FH treatments significantly increased the mineral-associated organic carbon (MOC) stock by 13.71% to 81.79%, 30.41% to 74.56%, 18.60% to 79.60%, and 17.83% to 103.57%, respectively, at the 0–100 cm depth (Figure 5a).

3.4.2. Heavy Fraction Organic Carbon

Compared with that in the CK treatment, the heavy fraction organic carbon (HFOC) stock significantly decreased by 8.81% to 15.15% across all the treatments at the 0–20 cm depth (Figure 5b). However, the HFOC stock generally increased, by 29.67% to 79.25%, 13.89% to 81.29%, 28.19% to 75.29%, and 72.05% to 101.51% in the FP, MF, RF, and FH treatments at the 20–100 cm depth, respectively, especially for the FH treatment.

3.5. Main Predictors of SOC Sequestration

To investigate the influence of each organic carbon fraction on SOC sequestration under film-mulching treatments (Figure 6), a machine learning approach through the random forest algorithm was conducted to assess the importance of the five organic carbon fractions. The higher the %IncMSE value is, the more significant the independent variable. The results showed that the SOC stock at the 0–100 cm depth was influenced primarily by HFOC, with a %IncMSE of approximately 13.

3.6. Validation of the DNDC Model

Values of the SOC stock at the 0–20 cm depth under various long-term fertilization and mulching treatments, alongside the simulated values generated by the DNDC model, are shown in Figure 7. Based on our experimental data, the n-RMSE and E values for the different fertilization and mulching treatments were calculated (

Table 1. Accuracy of the soil organic carbon stock simulated results by the denitrification–decomposition model.

Treatment	Average Observations Mg ha−1	Average Simulations Mg ha−1	MBE Mg ha−1	n-RMSE (%)	E (%)	d
FP	23.98	24.19	0.21	4.86	−1.16	0.82
MF	21.75	23.55	1.80	9.01	−8.68	0.92
RF	23.62	24.26	0.63	6.28	−3.08	0.85
FH	24.87	24.27	−0.61	2.94	2.42	0.96

). The n-RMSE for each treatment was less than 10%, indicating the ability of the DNDC model to simulate the trends in SOC stock changes accurately under diverse fertilization and mulching conditions. The E value for the FH treatment was positive, suggesting that the measured SOC stock values exceeded the simulated values. Conversely, the E values for the FP, MF, and RF treatments were negative, implying that the simulated SOC stock values were greater than the measured values. However, the absolute E values for each treatment ranged from 1.16% to 8.68%, all below 10%, indicating strong agreement between the simulated and measured values.

3.7. SOC Stock Simulation Under Long-Term Fertilization and Film-Mulching

To understand the anticipated climate change response of SOC, the DNDC model was utilized to simulate fluctuating SOC stocks from 0 to 50 cm depth over an 85-year period under various fertilization and nitrogen-mulching treatments, and the predicted climate data from 2015 to 2099 were used (Figure 8). The results indicate an increasing trend in the SOC stock from 0 to 50 cm depth under the current climatic conditions for all the treatments. Compared with the initial SOC stock, after 85 years of field management, the SOC stocks for the FP, MF, RF, and FH treatments are expected to increase by 59.15%, 57.15%, 138.94%, and 131.79%, respectively, with the RF treatment demonstrating the most substantial increase. Compared with the MF treatment, the RF and FH treatments are expected to further increase the SOC stock by 45.77% and 45.39%, respectively.

3.8. Soil Organic Carbon Stock Under Future Climate Change

To evaluate the potential influence of future climate change on SOC stocks, we utilized predicted climate data spanning from 2015 to 2099 (Figure 9). Throughout the simulation period, it was presumed that agricultural land management practices would remain unchanged. We modeled the variations in SOC stocks over an extended period of 85 years, considering various projected climate scenarios. These scenarios encompassed temperature rises of 1 °C and 2 °C, coupled with reductions in precipitation levels by 10% and 20%, respectively.

The simulation results revealed that, in comparison to a scenario with a 1 °C temperature rise, a 2 °C increase in temperature would result in a decrease in SOC stock ranging from 0.77 to 1.01 t·ha⁻¹. Conversely, under conditions of reduced precipitation, a 20% decrease in rainfall would lead to an increase in SOC stock by 1.53% to 3.42%, relative to a scenario with a 10% reduction in rainfall. For the FP, RF, and FH treatments, a 10% decrease in rainfall had a minimal impact on the SOC stock, with only a slight increase of approximately 1% compared to the conventional SOC stock under current climatic conditions. However, the MF treatment exhibited a more significant response to reduced rainfall, with a 10% decrease in precipitation resulting in an 11.72% increase in SOC stock compared to conventional levels.

4. Discussion

4.1. Soil Organic Carbon Stock

Organic carbon serves as a crucial indicator for assessing soil fertility. Its primary sources encompass animal and plant residues, microorganisms, as well as organic manure. Although the SOC content within the soil solid phase may not be substantial, it significantly influences soil fertility. The variations in its content and distribution are intimately tied to farmland management practices. In this study, we observed that SOC stock in the topsoil (0–20 cm and 20–40 cm depths) increased relative to the initial SOC stock observed for the FP and FH treatments. In addition, the FH treatment generally increased the SOC stock at the 40–100 cm depth, especially from the 80–100 cm depth, but the FP treatment significantly reduced the SOC stock at the 40–100 cm soil depth (Figure 3). In the subsoil (40–60 cm, 60–80 cm, and 80–100 cm depths), the SOC stock was lower than that in the topsoil under the FP treatment. This decrease could be explained by the possible presence of roots near the soil surface [43]. Subsoil layers receive organic matter inputs in the form of root litter, root exudates, and dissolved organic carbon (DOC) [44], and root C is more stable than shoot C because of the greater recalcitrance of the root tissue [17]. Another study reported that high fertilization rates may lead to soil acidification, which in turn accelerates SOC losses [45]. Moreover, continuous nitrogen fertilization at high rates reduces the C/N value and increases soil carbon consumption [46]. Our results demonstrated that long-term flat-film hole-sowing (FH) could promote SOC sequestration because the FH treatment resulted in water storage and moisture conservation, with water being more effectively used during crop growth. The film can intercept and collect rainfall, providing sufficient moisture for wheat growth. In addition, in the FH treatment, hole spacing was optimized by adjusting the plant spacing and row spacing of wheat to ensure sufficient photosynthesis, maximize the number of plants, and ultimately increase the wheat yield, thereby increasing the amount of crop residue returned to the field and increasing the SOC stock [47]. The FH treatment increased the SOC stock, and this treatment may reduce soil disturbance, aeration, and microbial activity in deeper soils [48], which could result in longer SOC residence times at deeper soil depths. On the other hand, a previous empirical study demonstrated strong mineral protection effects on SOC at deeper soil depths [49], where the Ca_ex and Fe/Al oxide contents are greater than those at the surface depth [50]. Moreover, clay content has been reported to be positively correlated with bacterial diversity [51], and soil texture can affect both the decomposition of plant litter and the retention of SOC in soils [52]. Therefore, the impact of FH treatment enhancing SOC sequestration in deep soil is more pronounced.

4.2. Labile C Fractions and Recalcitrant C Fractions

Soil water-soluble organic carbon (WSOC), particulate organic carbon (POC), and light fraction organic carbon (LFOC) are components of the soil active organic carbon pool, which is mainly composed of undecomposed or semi-decomposed animal and plant residues and root residues. The soil active organic carbon pool has a fast turnover rate in the soil and is very sensitive to farmland management practices, such as tillage and fertilization. This pool has an important impact on the potential productivity of the soil carbon pool and is often considered important as an indicator for evaluating SOC changes and soil fertility [50]. Compared with the CK treatment, the other treatments generally reduced the WSOC stock, while the FP and FH treatments significantly increased POC and LFOC stocks at the 0–40 cm and 0–80 cm depths, respectively (Figure 4). Results on the effects of mulching application on LFOC have been conflicting, with increases [53,54], decreases [55], or no significant effect [56]. These differences may be attributed to specific differences in study areas, cropping systems, and mulching methods. Our study showed that fertilization decreased the WSOC stocks. We attribute this change to a priming effect of nutrients on fresh organic matter in the soil, which stimulates microbial activity, helps decompose organic matter, and allows water-soluble organic matter to be released quickly [57]. Excessive fertilization causes nitrate nitrogen to leach into deep soil layers with rainfall, causing soil acidification, inhibiting microbial activity, and reducing carbon activation [4]. Therefore, the FP treatment significantly increased the LFOC and POC stocks at the 0–40 cm depth. Mulching also reduces WSOC stocks as a result of reductions in surface evaporation, increases in soil moisture and temperature, and promotion of the soil microbial growth [6], allowing WSOC to be used by soil microorganisms. Mulching promoted the growth of wheat roots, increased root biomass, and provided more fresh carbon sources for the soil. An increase in root biomass resulted in an increase in carbon input to the soil through root deposition [58], increasing the POC and LFOC stocks at the 0–80 cm soil depth.

Both mineral-associated organic carbon (MOC) and heavy fraction organic carbon (HFOC) are recalcitrant pools of organic carbon. MOC mainly refers to organic matter adsorbed by fine particles in the soil through ligand exchange. The main component of HFOC is mineral-associated organic matter [6]. Compared with active organic carbon, recalcitrant organic carbon exhibits a greater degree of resistance to decomposition and has a longer turnover time in soil, reflecting the stability of this type of SOC, which plays an important role in SOC sequestration [59,60]. Compared with the CK treatment, the fertilization and mulching treatments significantly increased the MOC stocks at the 0–100 cm depth, and the FH treatment had the greatest effect on increasing the MOC stocks at greater soil depths. In all treatments, HFOC stock was significantly lower at the 0–20 cm depth but significantly increased those stocks at the 20–100 cm depth (Figure 5). A similar result was reported by Zhang [61], who showed that mulching increased the MOC content of black loess soil. However, there is still controversy surrounding the effect of mulching on the content of HFOC. Some studies have shown that mulching promotes HFOC [58,62], some have demonstrated an inhibitory effect [63], and others have shown that mulching has no significant effect [64]. Mulching may have a positive effect on MOC because it promotes plant root growth [65] and effectively transforms root residues into mineral-organic-associated pools. The input of plant roots also stimulates the turnover of MOC [66]. Moreover, deep soils are increasingly anaerobic under mulching conditions, limiting microbial activity and reducing the oxidative decomposition of soil organic matter [67], which also explains why the FH treatment had a significant effect on increasing the MOC and HFOC stocks in deep soil.

In our study, the fertilization and mulching treatments significantly increased crop yields, thereby increasing the amount of crop root residues returned to the field. Microbial activities provide sufficient energy, especially under the favorable hydrothermal environment provided by mulching, which creates conditions suitable for the life activities of microorganisms. In addition, a previous study by our research group revealed that the use of mulch in dryland wheat fields on the Loess Plateau caused the loss of SOC, mainly by increasing the mineralization of recalcitrant organic carbon and the temperature sensitivity of this pool [30], thereby leading to a decrease in HFOC stocks at the 0–20 cm depth.

4.3. Factors Driving SOC Sequestration

The increase in the SOC stock at the 0–100 cm depth caused by film-mulching can be attributed to the increase in the HFOC stock (Figure 6). This may occur because the number and activity of deep soil microorganisms are expected to be lower than those present in surface soils, hence decreasing the overall decomposition rate of SOC. The HFOC has a relatively high degree of humification, and the relatively stable carbon fraction is anaerobically decomposed by microorganisms, such as lignin-decomposing bacteria or fungi, and this fraction can be transferred to the soil profile or become bound to minerals. A previous study reported that the proportions of galactose, mannose, and rhamnose in soil increased after film-mulching [64], indicating that this management practice increases the efficiency of conversion of plant residues into SOC by soil microorganisms [68,69]. Among the reorganized OC fractions in soil, microbial synthesis products dominate [70,71,72]. Film-mulching increases carbon input and improves soil water and heat conditions, which provides favorable conditions for the growth of microorganisms and accelerates the decomposition of plant-derived carbon by microorganisms [73]. Therefore, under conditions of high carbon input, film-mulching promotes the conversion of plant residues to SOC. A study reported that the proportion of new carbon in the HFOC fraction was lower than that in the LFOC fraction, which also verifies that the HFOC fraction is a more stable part compared to the LFOC fraction [58]. Our results also showed that increasing the content of reorganized organic carbon is conducive to increasing the stability of the SOC pool. Therefore, under existing farmland management measures, FH promotes SOC accumulation.

4.4. Dynamics of Soil Organic Carbon

The DNDC model did not predict a smooth increasing curve but rather a wave-like growth trend in SOC (Figure 8 and Figure 9). This finding is consistent with the results of previous studies [24,61,74]. Zhang [61] reported that under a wheat-corn rotation system, balanced fertilization and conservation tillage measures enhanced the “carbon sink” functions of black loess soil and that the net fixation of SOC predicted at the 0–50 cm depth over the next 60 years was 2.81–10.40 Mg ha⁻¹. Zhang et al. [24] and Yu et al. [74] used the DNDC model to simulate the changes in the SOC stocks of a black loess soil from 2016 to 2100 and a latent saline soil from 2020 to 2050, respectively. The authors reported that SOC stocks increased both with mulching and with no mulching. However, Yan et al. [23] used the DNDC model to simulate the changes in the organic carbon stock in the black loess soil of the Weibei dry plateau and predicted that mulching would reduce the SOC stock at the 0–50 cm depth by 25.2 kg ha⁻¹ over the next 50 years. This may be attributed to the fact that the crushing and pressing of straw subsequent to crop harvesting in our study directly augmented the input of exogenous carbon into the soil. Consequently, soil microbial activity was enhanced, straw decomposition was accelerated, and ultimately, SOC stocks increased.

The global terrestrial soil is poised to emit an additional 11–34 Pg of carbon into the atmosphere annually for every 1 °C increase in temperature, a phenomenon that could significantly intensify global warming. Our findings indicate that warming conditions are detrimental to the SOC sequestration (Figure 9), which aligns with the findings observed in other studies. Notably, Yang et al. [75] and Che et al. [76] independently reported that increasing temperatures lead to a reduction in the SOC reserves of both black loess and paddy soils. The projected future temperature increases are predicted to result in a decline in SOC stocks. This outcome is attributed to the fact that elevated atmospheric temperatures promote the proliferation of rhizosphere microbial communities and increase the decomposition of organic matter, thereby accelerating the release of SOC into the environment.

Global climate change significantly influences SOC cycles through alterations in precipitation patterns, and these changes are pivotal in the regulation of SOC dynamics [77]. Our findings indicate that a 20% decrease in rainfall can promote SOC sequestration more effectively than a 10% reduction (Figure 9). This observation aligns with the findings of most researchers, suggesting that a reduction in rainfall increases SOC sequestration [76,78]. However, it is important to note that an alternative study reported a tendency for SOC stocks to decline with reduced rainfall. The effect of rainfall on SOC is context-dependent and varies across different times and locations. In arid regions, an increase in rainfall can increase the soil respiration rate, which may reduce SOC sequestration [79]. Conversely, in humid areas, increased precipitation can suppress soil respiration, thereby facilitating SOC sequestration [78]. Moreover, the conditions of soil pores can impact SOC changes. When water is abundant, the permeability of soil pores tends to decrease, which can partially impede the mineralization and decomposition of SOC [80]. Additionally, the use of plastic film in arid and semiarid regions can bolster the resilience of plant biomass and SOC to fluctuations in temperature and rainfall [81].

5. Conclusions

A nine-year field experiment focusing on a winter wheat–summer fallow cropping system revealed that long-term flat-film hole-sowing (FH) enhanced SOC sequestration and increased wheat grain yield. The MF treatment, which involved a reduction in nitrogen fertilizer application, had no discernible impact on wheat yield. Compared to the CK treatment, all treatments generally led to a decrease in WSOC stock but notably increased MOC stock within the 0–100 cm soil depth and the HFOC stock within the 20–100 cm depth. Generally, the FH treatment resulted in an increase in POC and LFOC stocks at the 0–80 cm soil depth. The increase in the SOC stock within the 0–100 cm depth caused by film-mulching was attributed to an elevation in HFOC stock. The potential response of the SOC pool to future climate conditions was predicted via the DNDC model, and the results showed that warming conditions are unfavorable for SOC sequestration. Furthermore, a 20% reduction in rainfall was found to be more beneficial for SOC sequestration than a 10% reduction. Therefore, our study provides novel information that flat-film hole-sowing can synergistically increase crop yield and SOC pool capacity, buffering the risks of future climate change to arid soils and crop cultivation. However, the capacity of soil to sequestrate OC capacity is limited, and further research is needed on the saturation value of organic carbon in the dryland soil in this study. Our study provides a comprehensive understanding of the distribution of SOC fractions across 0–100 cm soil profiles subjected to different fertilization and film-mulching practices. These findings have significant implications for addressing and mitigating the impacts of future climate change.