Greenhouse Gas Emissions in Maize/Peanut Intercropping Under Water-Limited Semi-Arid Growing Conditions

Abstract

Maize/peanut intercropping is increasingly promoted as a climate-smart strategy for enhancing resource use efficiency and reducing environmental impacts in dryland cropping systems. However, its effects on multi-gas greenhouse emissions and yield-scaled climate performance remain insufficiently understood in semi-arid regions with sandy soil. Here, a two-year field experiment was conducted in western Liaoning, Northeast China, to quantify soil CO₂, CH₄, and N₂O fluxes, cumulative emissions, crop yield, global warming potential (GWP), and greenhouse gas intensity (GHGI) under sole maize (SM), sole peanut (SP), and two maize/peanut intercropping systems. SM produced the highest cumulative CO₂ emissions, whereas SP generated the highest CH₄ uptake and the highest N₂O emissions. Compared with peanut monoculture, maize/peanut intercropping significantly reduced soil N₂O emissions, indicating that the introduction of maize in the intercropping system provided an effective regulatory pathway for reducing N₂O emissions. Peanut yields declined by approximately 47.29–49.41%, leading to total land equivalent ratio (LER) values of 0.83–0.99. Although no significant land use advantage was observed for maize/peanut intercropping at the field scale, when crop yields were taken into account for assessment, the global warming potential (GWP) and greenhouse gas emission intensity (GHGI) were lower than those of monoculture uniformity. CO₂, CH₄ and N₂O fluxes were strongly correlated with soil temperature and moisture, underscoring the dominant role of microclimate rather than soil structure in regulating greenhouse gas (GHG) fluxes in monoculture, while in the intercropping system, the microclimate and the soil stucture together regulate the GHG fluxes. Overall, maize/peanut intercropping has the potential of reducing the climate cost per unit of production and represents a promising strategy for enhancing GHG mitigation potential in semi-arid agroecosystems.

1. Introduction

Soil is a major source of greenhouse gas, with CO₂, CH₄, and N₂O constituting the principal contributors to radiative forcing in agroecosystems [1,2]. In semi-arid regions, pronounced temperature fluctuations and sporadic rainfall events induce strong microclimatic variability, to which soil biogeochemical processes are highly sensitive. As a result, GHG emissions in these environments exhibit strong seasonal variations [3]. Developing cropping systems that can simultaneously sustain productivity and reduce GHG emissions is therefore essential for advancing climate-smart agriculture [4,5]. Among various ecological intensification practices suitable for semi-arid regions, cereal/legume intercropping systems—such as maize/peanut—have gained increasing attention due to their strong adaptability and system stability [6].

As an ecological intensification practice, intercropping has been shown to enhance resource use efficiency, improve nitrogen coordination [7,8], and increase system resilience [9], with particularly notable effects in cereal/legume combinations such as maize/peanut and maize/soybean. These advantages are largely attributed to complementary canopy structures, root stratification, and improved nitrogen synchronization [10,11,12]. However, evidence regarding the influence of intercropping on the “three-gas” GHG emissions (CO₂, CH₄, and N₂O) remains inconsistent, especially in rainfed semi-arid regions where environmental constraints, crop functional traits, and rhizosphere interactions jointly shape system responses [13]. Most existing studies have focused on yield and nitrogen use, while comprehensive assessments of multi-gas GHG emissions, their implications for system-level climate performance (e.g., GWP and GHGI), and the underlying regulatory mechanisms remain limited. A global meta-analysis on the effects of intercropping on soil greenhouse gas emissions demonstrated that cereal/legume intercropping systems can mitigate N₂O emissions. Nevertheless, comprehensive investigations into the three greenhouse gases remain scarce in specific geographical regions [14].

The semi-arid regions of China account for 27.56% of the national territorial area. These regions are mainly distributed in parts of Northwest, North, and Northeast China, where agricultural production is highly dependent on natural precipitation and extremely sensitive to climate change [15]. The agricultural production in this area is mainly based on dryland farming and animal husbandry, with water-saving agriculture as the core [16], and it plays a significant role in global greenhouse gas emissions and the carbon cycle [17]. Semi-arid region such as western Liaoning, Northeast China, is characterized by sandy soils, low organic matter content, and high interannual climatic variability [18], all of which amplify the sensitivity of soil GHG emissions to rainfall pulses and temperature shifts [19,20]. Under such conditions, maize and peanut, the two major crops grown in the region, exhibit pronounced functional contrasts: maize typically produces higher CO₂ emissions due to greater root biomass and carbon inputs [21], whereas peanut increases soil nitrogen through biological N fixation, potentially stimulating N₂O emissions while enhancing CH₄ uptake under low mineral-N conditions [22]. Despite these contrasting behaviors, systematic evaluations of CO₂–CH₄–N₂O dynamics in maize/peanut intercropping systems remain scarce [14], and the relative roles of soil physical structure and microclimatic drivers in shaping emission patterns are not well understood [20,23].

Although maize/peanut intercropping often improves resource use efficiency, it does not always provide a land use advantage under rainfed water-limited region, and its climate benefits may depend on the degree to which species-specific emission patterns offset or reinforce each other [24]. Thus, another knowledge gap concerns whether maize/peanut intercropping can reduce yield-scaled GHG emissions (GHGI) in semi-arid environments. Based on this crucial knowledge gap and the aforementioned research background, we hypothesized that in a semi-arid environment, intercropping of corn and peanuts, compared to monoculture of corn and peanuts, has a certain impact on GHGI.

To address these gaps, a two-year field experiment was conducted in rain-fed semi-arid Western Liaoning, China. The objectives were to: (1) quantify soil CO₂, CH₄ and N₂O fluxes and cumulative emissions under maize, peanut and maize/peanut intercropping systems; (2) quantify GWP and GHGI, and determine whether intercropping reduces yield-scaled GHG emissions relative to monoculture; and (3) evaluate the relationships between gas emissions, microclimatic conditions (temperature and soil moisture), and soil aggregate characteristics. This study is helpful for providing comprehensive evidence for the GHG emission potential of maize/peanut intercropping and offers insights for designing climate-efficient cropping systems in semi-arid regions.

2. Materials and Methods

2.1. Site Description

Field experiments were established in 2022 in the National Agricultural Experimental Station for Agricultural Environment (42°7′ N, 121°44′ E), which is located in Northeast China. The area has a semi-arid climate, with a dry and windy season from April to May and a hot and rainy season from July to August. The precipitation was 332.8 mm and 519.3 mm, and mean air temperatures were 20.5 °C and 21.6 °C during the crop growth seasons in 2023 and 2024, respectively. The daily average air temperature and rainfall during the experiment is shown in Figure 1. The soil type of the experimental land is cinnamon soil, classified as a cambisol (WRB), which is sandy in texture. The soil physical and chemical properties of the 0–20 cm soil layer before planting are shown in

Table 1. Physio-chemical properties of 0–20 cm soil.

Total Nitrogen (TN) (%)	Total Phosphorus (TP) (%)	Total Potassium (TK) (%)	Soil Organic Matter (SOM) (g/kg)	Available Phosphorus (AP) (mg/kg)	Available Potassium (AK) (mg/kg)	Cation Exchange Capacity (CEC) (cmol/kg)	pH	Sand Content (%)	Silt Content (%)	Clay Content (%)
0.10	0.03	1.31	12.20	54.20	80.13	32.36	5.81	57.50	29.63	12.88

.

2.2. Experimental Design and Field Management

Data was collected from 2023 to 2024. The experiment adopted a randomized complete block design with four treatments, SM (sole maize, Zea mays L. Jingke 968), SP (sole peanut, Arachis hypogaea L. Qinghua 6), M4P4 (maize/peanut intercropping with four rows of maize and four rows of peanut) and RM4P4 (maize/peanut intercropping system with four-row alternating strips and in-band rotation). Each treatment was replicated three times. The cropping system is illustrated in Figure 2. Maize was sown at a density of 67,500 plants ha⁻¹ on 6 May 2023 and 26 April 2024. Peanut was sown at a density of 300,000 plants ha⁻¹ on 19 May 2023 and 8 May 2024. Plot area was 84 m² and ridge width was 50 cm. Maize and peanuts are harvested simultaneously on 24 September 2023 and 12 September 2024. A stabilized compound fertilizer (N-P₂O₅-K₂O: 17-17-17) was applied at 750 kg ha⁻¹ across all sole and intercropping plots. The experimental area adopted the conventional tillage system.

2.3. Field Measurement

2.3.1. Greenhouse Gas Flux Measurement

GHG fluxes were monitored from crop emergence to harvest in both 2023 and 2024. Considering the laboratory conditions, two types of instruments were used for the determination of soil greenhouse gases. Both instruments rely on infrared gas analyzers to measure the concentrations of greenhouse gases based on the absorption characteristics of analytes at different concentrations toward infrared wavelengths and further calculate the emission rates from the variation in greenhouse gas concentrations within the gas chamber per unit time. CO₂ and CH₄ fluxes were simultaneously quantified using a GLA132 Series Ultraportable Analyzer (ABB GLA132, Zurich, Switzerland). This instrument consisted of an intelligent measurement chamber and gas analyzer. The soil collars of the instrument base were made of 8-inch thick-walled SDR35 polyvinyl chloride pipe (outer diameter: 20 cm, inner diameter: 18.76 cm, height:15 cm). Soil collars were pre-installed 24 h prior to the first measurement to ensure optimal determination accuracy; they were vertically inserted into the soil at a depth of 12 cm, exposed 3 cm above ground, and the gaps between the collars and the surrounding soil were immediately filled with soil to prevent gas leakage. After the gas analyzer was properly connected to the intelligent measurement chamber, we set the observation length to 120 s, the deadband to 60 s, and the post purge to 30 s, and started the gas measurement. Nitrous oxide (N₂O) fluxes were measured using a N₂O/H₂O Trace Gas Analyzer (LI-COR, LI-7820, Lincoln, NB, USA). The outer diameter of the soil collar was 21.34 cm, the inner diameter was 20.32 cm, and the height was 15 cm. We set the observation length to 120 s, the deadband to 25 s, and the post purge to 45 s, and started the gas measurement. Gas fluxes were calculated using the instrument-specific software.

For sole crops, one chamber was installed at the middle of a plot, between two rows. For intercropping plots, three chambers were deployed in three positions: at the middle of the maize strip (I_m), between maize and peanut adjacent border rows (I_mp), and at the middle of the peanut strip (I_p). The soil collars of the two instruments were placed in close proximity, and the errors in soil environmental conditions (e.g., temperature and humidity) caused by positional differences were considered negligible. Sampling and measurements were performed synchronously, with both instruments configured for identical measurement durations. Gas samples were collected once a week between 09:00 and 11:00 to minimize the influence of diurnal temperature variations on gas emission rates, but if it rained, sampling was conducted after the rain.

2.3.2. Temperature and Soil Properties

Total nitrogen was determined using an elemental analyzer (EA3000, EuroVector, Pavia, Italy). Total phosphorus and total potassium were both measured after acid digestion, with phosphorus quantified by the molybdenum blue colorimetric method (UV-9000S, Shanghai Metash Instruments Co., Ltd., Shanghai, China) and potassium determined using a flame photometer (Z-2000, Hitach, Tokyo, Japan). Soil organic matter content was determined by the potassium dichromate oxidation method. Available phosphorus was extracted using the Olsen method and quantified colorimetrically (UV-9000S, Shanghai Yuanxi, Shanghai, China). Available potassium was determined using a flame photometer (Z-2000, Hitach, Tokyo, Japan). Soil cation exchange capacity was determined using the hexaamminecobalt(III) chloride extraction method. Soil pH was measured in a soil–water suspension (1:2.5, w/v) using a glass electrode pH meter (PHS-3E, Shanghai INESA Scientific Instrument Co., Ltd., Shanghai, China).

Air chamber temperature (AT) was measured using a thermistor probe integrated with chamber of N₂O/H₂O Trace Gas Analyzer. Soil temperature (ST), soil water content (SWC), and soil electrical conductivity (EC) were measured using a Stevens HydraProbe sensor integrated with the N₂O/H₂O Trace Gas Analyzer (LI-COR, LI-7820, Lincoln, NB, USA).

The soil water-stable aggregates were determined by the wet sieving method. Since soil organic matter is one of the main factors affecting the formation and stability of soil aggregates [25], the changes in soil aggregates were induced merely by altering cropping patterns proceed slowly. For this reason, we chose to collect soil samples in 2024. After crop harvest in 2024, intact soil samples from the 0–20 cm layer were collected adjacent to each gas chamber. For monoculture, one sampling point was collected in each plot between the crop rows. For intercropping, samples were taken from each plot in accordance with the maize, peanuts and the boundary point. Bring the intact soil samples back to the laboratory and break them along the natural fracture surface into soil samples with a diameter of 1 cm. The samples were air-dried and dry-sieved, and 100 g subsamples were prepared according to the proportional mass of each size fraction. These subsamples were then subjected to wet-sieving using a sterile sieve set. Water-stable aggregates retained on each sieve were collected and oven-dried at 65 °C to determine the mass of aggregates in the >2 mm, 0.25–2 mm, 0.106–0.25 mm, and <0.106 mm size classes.

2.3.3. Crop Yield

Maize and peanut were harvested from a sub-sampling area of 14 m² (2 × 7 m) and 4 m² (2 × 2 m) in the center of each plot for both sole crops and intercrops at maturity. Ten maize plants and 1 m² of peanut in the sub-sampling area were randomly selected to measure the yield components for each plot after being sun-dried (water content of 14%).

2.4. Data Analysis

2.4.1. Analysis of Greenhouse Effects on the Agroecosystem

(1)

The fluxes of GHGs emissions in intercropping

The fluxes of GHGs emissions in intercropping were calculated using Equation (1) [26].

CF (CO₂, CH₄ or N₂O) = CF(I_m) × LS_m + CF(I_p) × LSp + CF(I_mp) × LS_mp

(1)

where CF (I_m) and (I_p) represent the fluxes for intercropping, and CF (I_mp) represent the fluxes for the junction of maize and peanut planting. LS_m, LS_p and LS_mp denote the land shares of maize and peanut in intercropping, accounting for 3/7, 3/7, and 1/7, respectively.

(2)

Cumulative CO₂, CH₄ and N₂O emissions

Cumulative CO₂, CH₄ and N₂O emissions were calculated using Equation (2) [27].

$$\mathrm{Cumulative} {\mathrm{CO}}_2, {\mathrm{CH}}_4 \mathrm{or} {\mathrm{N}}_2\mathrm{O} \mathrm{emissions}={\displaystyle {\sum }_{\mathrm{i}}^{\mathrm{n}}({\mathrm{F}}_{\mathrm{i}}\times {\mathrm{D}}_{\mathrm{i}}).}$$

(2)

where n represents the number of sampling intervals. F_i represents the ratio of NO₂, CH₄, and CO₂ emission flux (N₂O and CH₄: nmol·m⁻²·s⁻¹, CO₂: µmol·m⁻²·s⁻¹) within the ith sampling interval, and D_i stands for the number of days between F_i and F_i+1.

(3)

GWP

The global warming potential (GWP) was calculated using Equation (3) [28].

GWP = CF(CO₂) + 29.8 × CF(CH₄) + 273 × CF(N₂O)

(3)

where CF(CO₂), CF(CH₄), and CF(N₂O) represent the cumulative emissions of GHGs (kg C ha⁻¹ or kg N ha⁻¹).

(4)

Since there are two different crops in the intercropping system, peanut yields were converted to a unified standard based on market prices.

Maize equivalent yield (MEY) was calculated using Equation (4) [29].

MEY = P_P/P_M × Y_P

(4)

where P_P represents the price of peanut, P_M represents the price of maize, and Y_P represents the yield of peanut. According to the local purchase prices (data from https://www.chinagrain.cn), the price of P_P was 9.6 yuan per kilogram and that of P_M was 2.2 yuan per kilogram in 2023; the price of P_P was 8.2 yuan per kilogram and that of P_M was 1.8 yuan per kilogram in 2024.

(5)

GHGI

GHGI can reflect the CO₂ emissions equivalent of each processing unit output, which was calculated using Equation (5) [30].

GHGI = GWP/Grain yield

(5)

The final grain yield is the sum of the maize yield and the peanut yield (maize equivalent yield).

2.4.2. LER

Land equivalent ratio (LER) was calculated using Equation (6) [7].

LER = (Y_im/Y_sm) + (Y_ip/Y_sp)

(6)

where Y_im and Y_ip represent maize and peanut yield in intercropping systems, respectively. Y_sm and Y_sp represent maize and peanut yield in sole systems, respectively.

2.4.3. Soil Aggregate

The mass percentage of aggregates in each size class (W_i) was calculated using Equation (7). Soil average weight diameter (MWD) and soil geometric mean diameter (GMD) were calculated using Equations (8) and (9) [31].

W_i = G_i/MT × 100%

(7)

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

(8)

$$GMD=\mathrm{EXP} [({\sum }_{i=1}^{\mathrm{n}}{W}_i\ln X_i)/{\sum }_{i=1}^{\mathrm{n}}{W}_i]$$

(9)

where G_i is the air-dry mass of aggregates in size class i. MT is the total mass of the aggregates. X_i is the mean diameter of aggregates in that size class.

2.5. Statistical Analysis

All data were subjected to analysis of variance (ANOVA) to evaluate treatment effects on the measured parameters. Treatment means were compared using Duncan’s test at a 5% significance level (p < 0.05). The two-way ANOVA was performed using a general linear model, and when the main effects were significant (p < 0.05), pairwise comparisons were made using Duncan’s test. Year and treatment were treated as fixed factors. Prior to ANOVA, homogeneity of variance test and Shapiro–Wilk (S-W) normality test were conducted to verify the data of each treatment group. The results showed that all data generally satisfied the assumptions of homogeneity of variance and normality (p > 0.05). Statistical analyses were performed using SPSS 29.0 (IBM Corp., New York, NY, USA). The Pearson correlation coefficient analysis (two-tailed test) was performed to assess the relationships between GHG fluxes, soil properties, and environmental factors (Originlab, Northampton, MA, USA). The figures were visualized and presented using OriginPro 2025 (Originlab, Northampton, MA, USA).

3. Results

3.1. Dynamics of Greenhouse Gas Emissions

Across the two growing seasons, the temporal patterns of soil CO₂, CH₄, and N₂O fluxes exhibited clear and treatment-specific differences (Figure 3). Soil CO₂ emissions in all treatments followed a unimodal seasonal trajectory, with gradual increases after crop establishment, peak respiration during mid-vegetative and reproductive stages, and subsequent declines toward crop maturity. Flux magnitudes ranged between 1.21 and 6.99 μmol m⁻² s⁻¹, with monoculture maize (SM) consistently showing the highest CO₂ emissions across both years. In contrast, monoculture peanut (SP) maintained the lowest CO₂ fluxes throughout the growing period, while the two intercropping systems (M4P4 and RM4P4) displayed intermediate levels, reflecting the combined influence of maize- and peanut-derived carbon inputs.

CH₄ fluxes remained negative across all treatments and sampling dates, indicating persistent net atmospheric CH₄ uptake throughout the growing seasons. The minimum fluxes of CH₄ in 2023 and 2024 reached −0.38 nmol m⁻² s⁻¹ and −0.43 nmol m⁻² s⁻¹ in 2023 and 2024, respectively. SP showed the poorest CH₄ absorption capacity. The CH₄ absorption capacity of intercropping lay between the sole cropping patterns.

N₂O emissions also displayed distinct treatment responses. The SP showed substantially higher fluxes than the other treatments with peak emissions reaching 0.84 and 1.66 nmol m⁻² s⁻¹ in 2023 and 2024, respectively. SM consistently produced the lowest N₂O fluxes, while the intercropping treatments exhibited intermediate magnitudes.

3.2. Cumulative GHGs Emissions

Cumulative CO₂, CH₄, and N₂O emissions over the two years reinforced the seasonal flux patterns (

Table 2. Cumulative CO₂, CH₄ and N₂O emissions in 2023 and 2024.

Year	Treatment	Cumulative CO2 Emissions (t CO2–C ha−1)	Cumulative CH4 Uptake (kg CH4–C ha−1)	Cumulative N2O Emissions (kg N2O–N ha−1)
2023	SM	15.12 ± 1.60 a	−4.63 ± 0.16 b	7.56 ± 0.09 b
	SP	10.92 ± 0.03 b	−2.62 ± 0.08 a	26.84 ± 5.39 a
	M4P4	13.10 ± 0.71 ab	−4.12 ± 0.40 b	13.09 ± 1.90 b
	RM4P4	12.80 ± 1.10 ab	−4.07 ± 0.35 b	11.58 ± 3.44 b
2024	SM	17.74 ± 0.24 a	−4.25 ± 0.29 ab	4.37 ± 0.52 b
	SP	12.35 ± 0.72 b	−3.09 ± 0.28 a	23.80 ± 3.89 a
	M4P4	17.69 ± 1.34 a	−3.26 ± 0.31 ab	8.68 ± 2.33 b
	RM4P4	17.28 ± 1.32 a	−4.19 ± 0.32 b	7.44 ± 0.52 b
Mean	SM	16.43 ± 0.80 a	−4.44 ± 0.20 c	5.96 ± 0.29 b
	SP	11.63 ± 0.35 b	−2.86 ± 0.16 a	25.32 ± 2.67 a
	M4P4	15.39 ± 0.95 a	−3.69 ± 0.25 b	10.88 ± 1.95 b
	RM4P4	15.04 ± 1.16 a	−4.13 ± 0.16 bc	9.51 ± 1.77 b
Two-way ANOVA results
F	Year	20.47 **	0.61 n.s.	3.32 n.s.
	Treatment	8.21 **	11.15 **	17.71 **
	Year × Treat	1.11 n.s.	1.97 n.s.	0.03 n.s.

Note: SM and SP indicate sole maize and peanut, respectively. M4P4 indicates maize/peanut intercropping with four rows maize and four rows peanut and RM4P4 indicates maize/peanut intercropping system with four-row alternating strips and in-band rotation. The numbers in the table represent mean ± standard error. Lowercase letters represent the results of one-way ANOVA. Different letters indicate that there was significant difference between treatments at p < 0.05. ** represents significant differences at p < 0.01. n.s. represents not significant differences at p > 0.05.

). The highest cumulative CO₂ emissions occurred under SM, reaching a two-year mean of 16.43 ± 0.80 t CO₂–C ha⁻¹, significantly exceeding those of SP (11.63 ± 0.35 t CO₂–C ha⁻¹). Both intercropping treatments remained between the two monocultures and did not differ significantly from SM, reflecting the dominant contribution of maize to total soil CO₂ respiration. Meanwhile, M4P4 and RM4P4 were significantly lower than the monoculture uniformity in 2023 (p = 0.005 and p = 0.004 respectively) and 2024 (p = 0.007 and p = 0.006 respectively), namely the average emissions of maize monoculture and peanut monoculture.

Cumulative CH₄ uptake displayed the opposite trend, with SM showing the strongest net uptake (−4.44 ± 0.20 kg CH₄–C ha⁻¹) and SP the weakest (−2.86 ± 0.16 kg CH₄–C ha⁻¹). Meanwhile, M4P4 and RM4P4 were significantly higher than the monoculture uniformity in 2023 (p = 0.001 and p = 0.006, respectively) and 2024 (p = 0.007 and p = 0.006, respectively). Similarly, cumulative N₂O emissions were highest in SP (25.32 ± 2.67 kg N₂O–N ha⁻¹) and lowest in SM (5.96 ± 0.29 kg N₂O–N ha⁻¹), and the intercropping showed significant differences from SP (p < 0.05). Meanwhile, M4P4 (p = 0.033) and RM4P4 (p = 0.018) were significantly lower than the monoculture uniformity in 2024.

According to the two-way ANOVA, year exerted a significant main effect on cumulative CO₂, and there was a significant main effect between treatments on the cumulative emissions of the three gases. However, there was no interaction between year and treatment for the cumulative emissions of the three gases. This absence of interaction suggests that treatment effects on cumulative GHG emissions were consistent despite interannual differences in climatic conditions.

3.3. Yield, Global Warming Potential (GWP) and Greenhouse Gas Emission Intensity (GHGI)

In 2023, neither maize yield nor peanut yield under intercropping treatments showed significant differences compared with monocropping. In 2024, maize yield of RM4P4 was 14.51 ± 0.37 t ha⁻¹, which was 46.39% significantly higher than SM. In contrast, peanut yields of M4P4 and RM4P4 were 3.09 ± 0.32 t ha⁻¹ and 3.33 ± 0.56 t ha⁻¹, which were 54.09% and 50.45% significantly lower than SP, respectively. In 2023, the average yields of maize and peanut in Fuxin City were 7.62 t ha⁻¹ and 2.97 t ha⁻¹, respectively. In 2024, the corresponding figures reached 7.64 t ha⁻¹ and 3.30 t ha⁻¹. Compared with the city-wide average yields, the productivity level of the experimental plots was relatively high (data from https://www.fuxin.gov.cn/). There was no significant difference in the two-year average maize yield among treatments. However, the two-year average peanut yields of M4P4 and RM4P4 were 3.13 ± 0.16 t ha⁻¹ and 3.26 ± 0.50 t ha⁻¹, which were 49.41% and 47.29% significantly lower than SP, respectively. The partial LER of maize was 0.63 over two years, whereas peanut LERp averaged only 0.28. The resulting total LER ranged between 0.83 and 0.99, indicating no consistent land use advantage for intercropping systems (

Table 3. Yield, partial land equivalent ratio of maize (LER_m) and peanut (LER_p), and land equivalent ratio (LER) in sole and intercropping systems.

Year	Treatment	Yield (t ha−1)		LERm	LERp	LER
		Maize	Peanut
2023	Sole	12.80 ± 1.56 n.s.	5.63 ± 1.06 n.s.
	M4P4	13.53 ± 1.93 n.s.	3.16 ± 0.10 n.s.	0.55 ± 0.13 n.s.	0.30 ± 0.06 n.s.	0.86 ± 0.11 n.s.
	RM4P4	14.96 ± 1.59 n.s.	3.18 ± 0.74 n.s.	0.60 ± 0.09 n.s.	0.34 ± 0.15 n.s.	0.94 ± 0.12 n.s.
2024	Sole	9.91 ± 0.92 b	6.73 ± 0.45 a
	M4P4	11.98 ± 1.67 ab	3.09 ± 0.32 b	0.60 ± 0.07 n.s.	0.23 ± 0.02 n.s.	0.83 ± 0.07 n.s.
	RM4P4	14.51 ± 0.37 a	3.33 ± 0.56 b	0.74 ± 0.07 n.s.	0.24 ± 0.03 n.s.	0.99 ± 0.08 n.s.
Mean	Sole	11.35 ± 1.12 n.s.	6.18 ± 0.41 a
	M4P4	12.75 ± 1.80 n.s.	3.13 ± 0.16 b	0.58 ± 0.10 n.s.	0.27 ± 0.03 n.s.	0.85 ± 0.10 n.s.
	RM4P4	14.73 ± 0.77 n.s.	3.26 ± 0.50 b	0.67 ± 0.04 n.s.	0.29 ± 0.08 n.s.	0.96 ± 0.09 n.s.
Two-way ANOVA results
F	Year	1.92 n.s.	0.61 n.s.	1.15 n.s.	1.00 n.s.	0.02 n.s.
	Treatment	2.78 n.s.	15.51 **	1.02 n.s.	0.09 n.s.	1.47 n.s.
	Year × Treatment	0.36 n.s.	0.51 n.s.	0.26 n.s.	0.02 n.s.	0.13 n.s.

Note: SM and SP indicate sole maize and peanut, respectively. M4P4 indicates maize/peanut intercropping with four rows of maize and four rows of peanut, and RM4P4 indicates maize/peanut intercropping system with four-row alternating strips and in-band rotation. The numbers in the table represent mean ± standard error. Lowercase letters represent the results of one-way ANOVA. Different letters indicate that there was significant difference between treatments at p < 0.05. ** represents significant differences at p < 0.01. n.s. represents not significant differences between treatments at p > 0.05.

). According to the two-way ANOVA, there was a significant main effect between treatments on peanut yield.

There were no significant differences in GWP among the four cropping systems in either year. But M4P4 and RM4P4 were significantly lower than the monoculture uniformity in 2023 (p = 0.007 and p = 0.005 respectively) and 2024 (p = 0.007 and p = 0.004 respectively). The grain yield under SP was significantly higher than that of the other three treatments over the two-year period (p < 0.05). Meanwhile, M4P4 and RM4P4 were significantly lower than the monoculture uniformity in 2023 (p = 0.006 and p = 0.008, respectively) and 2024 (p = 0.007 and p = 0.004 respectively). In both 2023 and 2024, as well as in the two-year average of GHGI, the ranking consistently followed the order: SM > M4P4 > RM4P4 > SP. The mean GHGI value for SP was significantly lower than those of the other three treatments (p < 0.05). The average value of GHGI showed no significant difference from that of SM, but the p-value for the difference between the RM4P4 and SM treatment groups was 0.051, which is very close to the conventional significance threshold of 0.05. Furthermore, M4P4 (p = 0.013) was significantly lower than the monoculture uniformity in 2023. RM4P4 was significantly lower than the monoculture uniformity in 2023 (p = 0.007) and 2024 (p = 0.004). Additionally, GWP and GHGI exhibited significant interannual variation (p < 0.05), whereas grain yield and GHGI showed significant treatment effects (p < 0.01) (

Table 4. Global warming potential (GWP), grain yield and greenhouse gas emission intensity (GHGI) in 2023 and 2024.

Year	Treatment	GWP (t CO2-eq ha−1)	Grain Yield (t ha−1)	GHGI (CO2-eq kg−1)
2023	SM	17.04 ± 1.59 a	12.80 ± 1.56 b	1.35 ± 0.10 a
	SP	18.17 ± 1.48 a	24.55 ± 4.63 a	0.79 ± 0.12 b
	M4P4	16.55 ± 1.22 a	13.66 ± 1.18 b	1.22 ± 0.04 a
	RM4P4	15.84 ± 2.02 a	14.41 ± 0.83 b	1.10 ± 0.15 ab
2024	SM	18.81 ± 0.38 a	9.91 ± 0.92 c	1.93 ± 0.20 a
	SP	18.75 ± 0.34 a	30.66 ± 2.04 a	0.62 ± 0.04 b
	M4P4	19.96 ± 1.97 a	13.02 ± 1.12 bc	1.57 ± 0.23 a
	RM4P4	19.18 ± 1.44 a	14.85 ± 1.46 b	1.34 ± 0.24 a
Mean	SM	17.92 ± 0.83 a	12.08 ± 1.11 b	1.64 ± 0.09 a
	SP	18.46 ± 0.69 a	27.60 ± 1.79 a	0.70 ± 0.05 b
	M4P4	18.25 ± 1.46 a	13.34 ± 1.07 b	1.39 ± 0.13 a
	RM4P4	17.51 ± 1.61 a	14.63 ± 0.91 b	1.22 ± 0.20 a
Two-way ANOVA results
F	Year	5.05 *	0.27 n.s.	0.27 *
	Treatment	0.17 n.s.	25.34 **	25.34 **
	Year × Treatment	0.45 n.s.	1.71 n.s.	1.71

Note: SM and SP indicate sole maize and peanut, respectively. M4P4 indicates maize/peanut intercropping with four rows of maize and four rows of peanut, and RM4P4 indicates maize/peanut intercropping system with four-row alternating strips and in-band rotation. The numbers in the table represent mean ± standard error. Lowercase letters represent the results of one-way ANOVA. Different letters indicate that there was significant difference between treatments at p < 0.05. * represents significant differences at p < 0.05. ** represents significant differences at p < 0.01. n.s. represents not significant differences between treatments at p > 0.05.

).

3.4. The Relationships Between Emissions and Soil Environment Factors

Soil aggregate composition and stability showed no significant differences among the four cropping systems (

Table 5. The proportion of soil aggregates of different particle sizes in 2024 (%).

Treatment	X ≥ 2 mm	0.25 mm ≤ X < 2 mm	0.106 mm ≤ X < 0.25 mm	X < 0.106 mm	>0.25 mm
SM	9.00 ± 1.12 n.s.	33.74 ± 3.63 n.s.	15.66 ± 4.82 n.s.	41.60 ± 1.03 n.s.	42.74 ± 4.46 n.s.
SP	13.78 ± 1.40 n.s.	40.77 ± 7.02 n.s.	7.41 ± 3.43 n.s.	38.04 ± 4.41 n.s.	54.55 ± 7.52 n.s.
M4P4	14.98 ± 1.53 n.s.	34.74 ± 5.07 n.s.	13.56 ± 1.56 n.s.	36.71 ± 6.13 n.s.	49.72 ± 6.57 n.s.
RM4P4	13.16 ± 3.28 n.s.	36.38 ± 4.78 n.s.	13.26 ± 0.25 n.s.	37.20 ± 7.34 n.s.	49.54 ± 7.32 n.s.

Note: SM and SP indicate sole maize and peanut, respectively. M4P4 indicates maize/peanut intercropping with four rows of maize and four rows of peanut, and RM4P4 indicates maize/peanut intercropping system with four-row alternating strips and in-band rotation. The numbers in the table represent mean ± standard error. n.s. represents not significant differences between treatments at p > 0.05.

). The proportions of aggregates across the four particle-size classes, as well as the content of water-stable aggregates > 0.25 mm, remained statistically similar among treatments, indicating that short-term variation in cropping patterns did not markedly alter soil structural composition. Likewise, the aggregate stability indices MWD and GMD did not differ significantly across treatments; however, both indices exhibited a modest increasing tendency under the intercropping systems, with mean values 11.31% (MWD) and 16.67% (GMD) higher than those under sole maize, suggesting a slight improvement in structural stability that was not large enough to explain treatment-specific differences in greenhouse gas emissions. In addition, the soil electrical conductivity (EC) ranged from 0.13 to 0.14 dS/m and remained highly stable across all treatments, with no significant differences observed (

Table 6. Soil aggregate stability and EC under different treatments in 2024.

Treatment	MWD (mm)	GMD (mm)	EC (ds/m)
SM	0.74 ± 0.06 n.s.	0.26 ± 0.02 n.s.	0.14 ± 0.01 n.s.
SP	0.97 ± 0.10 n.s.	0.37 ± 0.07 n.s.	0.14 ± 0.01 n.s.
M4P4	0.96 ± 0.11 n.s.	0.35 ± 0.08 n.s.	0.13 ± 0.01 n.s.
RM4P4	0.91 ± 0.15 n.s.	0.35 ± 0.09 n.s.	0.13 ± 0.00 n.s.

Note: SM and SP indicate sole maize and peanut, respectively. M4P4 indicates maize/peanut intercropping with four rows maize and four rows of peanut and RM4P4 indicates maize/peanut intercropping system with four-row alternating strips and in-band rotation. MWD indicates soil average weight diameter and GMD indicates soil geometric mean diameter. EC indicates soil electrical conductivity. n.s. represents not significant differences between treatments at p > 0.05.

).

In contrast, the correlation analysis (Figure 4) revealed clear and significant relationships between greenhouse gas emissions and soil environmental conditions. For SM, soil CO₂ fluxes had a significant positive correlation with SWC. CH₄ fluxes had a significant positive correlation with both SWC and EC. N₂O had a significant negative correlation with AT and ST. For SP, CO₂ fluxes had a significant positive correlation with AT, ST, SWC and EC. CH₄ fluxes had a significant positive correlation with both SWC and EC. However, for intercropping, although GHG was significantly correlated with AT, SWC and EC, the response of GHG to microclimate change under intercropping conditions was not as intense as that of sole peanut. Furthermore, the change in CO₂ during intercropping was also significantly negatively correlated with GMD.

4. Discussion

4.1. Species-Specific Carbon–Nitrogen Processes Underpin Divergent GHG Emission Patterns

The contrasting GHG patterns observed among maize, peanut, and their intercropping combinations largely reflect inherent species-specific traits related to carbon allocation, nitrogen cycling, and microbial substrate availability [32]. Depending on soil texture, the lateral spread of mature maize root systems can reach 90 to 120 cm around the plants, and they typically extend to a depth of 150 to 180 cm [33]. Compared with the sorghum root system, the maize root system is denser and has wider root angles [34]. The elevated CO₂ emissions under maize monoculture might be due to the extensive root system of maize, and higher carbohydrate turnover can stimulate microbial respiration, resulting in greater CO₂ fluxes. In contrast, the suppressed CO₂ emissions under peanut monoculture reflect the lower root biomass and reduced carbon input commonly reported for leguminous crops [35].

The significantly higher N₂O emissions under peanut monoculture support previous observations that legume-based systems can stimulate N₂O production by increasing soil ammonium, nitrate, and labile organic N following biological nitrogen fixation [36,37]. Peanut may elevate nitrification–denitrification substrate levels, especially after rainfall-induced soil rewetting, thus accelerating N₂O emissions [38]. Meanwhile, the consistently strong CH₄ uptake under SP parallels earlier reports that low mineral-N conditions favor methanotrophic activity by reducing competition with nitrifiers [39].

The climate benefits of intercropping systems, such as optimizing soil carbon and nitrogen pools and regulating microbial communities, are typically long-term cumulative processes rather than immediate effects. In long-term positioning experiments, legume/cereal intercropping can promote soil organic carbon (SOC) sequestration through continuous input of root exudates and crop residues [40], and enhance nitrogen fixation via rhizobia to reduce reliance on chemical nitrogen fertilizer, thereby inhibiting N₂O emissions [41] or the enhanced diversity of root systems in the intercropping system may have strengthened the capacity for soil nutrient uptake, resulting in an overall decrease in soil NH₄⁺ and NO₃⁻ concentrations—conditions that are not conducive to nitrification and denitrification processes [42]. However, in this short-term experiment, the soil organic carbon content in the intercropping system may not yet have reached a steady state capable of significantly altering greenhouse gas emission fluxes. Therefore, compared with SM, there was no significant difference in CO₂ emissions. Nevertheless, the cumulative N₂O emissions in the intercropping treatments were significantly lower than those in SP, indicating that the introduction of maize in the intercropping system provided an effective regulatory pathway for reducing N₂O emissions.

4.2. Yield-Scaled Climate Benefits of Intercropping Despite Limited Land-Use Advantage

Although land equivalent ratios showed no consistent land use advantage in maize/peanut intercropping, this divergence aligns with observations from maize/soybean and cereal/legume systems worldwide, where intercropping often improves resource use efficiency and nitrogen synchronization but does not always translate into land use advantages due to light competition and legume suppression [9,43,44]. The increased maize yield in the intercropping systems of our study corresponds with well-documented facilitative effects—such as improved light capture and more efficient nutrient use—reported for tall cereals paired with low-stature legumes in semi-arid zones [11,45].

The lack of GWP differences among treatments indicates that compensatory effects among CO₂, CH₄, and N₂O masked treatment-specific emissions. This phenomenon has been documented in other studies where multi-gas responses move in opposite directions, leading to negligible differences in aggregated radiative forcing [46]. Thus, system-level climate assessments based solely on GWP may overlook important mitigation benefits. Greenhouse gas intensity (GHGI) is an evaluation index of low-carbon agriculture at this stage, which takes into account both crop yield and global warming potential [47]. A lower GHGI value indicates a smaller environmental footprint of the production process [48]. In contrast, GHGI captured meaningful differences. Specifically, the GHGI values of the intercropping systems fell between those of SM and SP; although the differences did not reach statistical significance, they were very close to the significance threshold. Previous studies have indicated that intercropping can significantly reduce GHGI [49], but these findings were derived from long-term fixed-site intercropping experiments. By contrast, the duration of intercropping implemented in our study was relatively short, which may explain why no significant reduction in GWP and GHGI were observed when we directly analyzed the data of different treatments. However, through further analysis of the data, the GWP and GHGI were found to be significantly lower than those of monoculture uniformity.

Overall, these findings indicate that while intercropping may not always increase total yield per unit land, it plays a critical role in reducing the cost of increased productivity for the consortium, aligning with global calls for climate-smart crop diversification strategies.

4.3. The Microclimate and Soil Structure Jointly Determine Greenhouse Gas Emissions in the Intercropping System

Generally, microclimate changes such as soil temperature, soil water content, and soil electrical conductivity have a relative obvious impact on soil greenhouse gases [37,50]. In our study, SWC was significantly and positively correlated with CO₂ and CH₄ fluxes under both monocropping and intercropping systems, and EC was also significantly and positively correlated with CH₄ fluxes across all treatments. Soil moisture and the EC had the strongest impact on the soil greenhouse gases emission. In addition, soil CO₂ emissions are significantly correlated with the GMD of soil aggregates under intercropping conditions, indicating that changes in soil structure significantly affects the soil greenhouse gas emissions for intercropping patterns. Studies have shown that intercropping can improve soil structure and physicochemical properties [51,52]. Therefore, intercropping can not only mitigate the intense response of greenhouse gases to microclimate changes but also has the potential to improve soil structure and thereby reduce greenhouse gas emissions. However, soil aggregate composition and stability showed minimal sensitivity to the short-term cropping treatments, consistent with studies indicating that changes in soil physical structure often require multiple years or continuous organic input to manifest [23]. Therefore, reducing the GHG by improving the soil aggregates may need long periods. This aligns with prior findings that year-to-year variations in aggregation rarely regulate microbial-mediated GHG processes in sandy soils [53], which showed that despite the slight tendency toward improved MWD and GMD under intercropping, the differences were too small to meaningfully influence gas fluxes. In contrast, the strong positive correlations between cumulative CO₂, CH₄ and N₂O emissions, and their significant associations with soil temperature, air chamber temperature, and soil moisture, are consistent with long-established relationships between microclimate and microbial activity. Numerous studies indicate that warming accelerates soil respiration and N transformations [54], while rainfall-induced rewetting events trigger large pulses of CO₂ and N₂O via the Birch effect and denitrification hotspots [20,55]. Our findings fully align with this mechanistic understanding: temperature and moisture dominate controlled emission magnitudes across treatments [53,54].

The semi-arid environment of western Liaoning, characterized by episodic rainfall and strong temperature fluctuations, amplifies the influence of microclimate on GHG emissions in monocultures. Prior work in similar temperate drylands has shown that small rainfall pulses can account for up to 40–70% of seasonal N₂O fluxes [55]. This helps explain why cropping treatments with similar soil structure, but different microclimate exposure, showed clear emission differences. It also underscores the need for management strategies focusing on microclimate stabilization (e.g., mulching, residue retention, optimized irrigation timing) [56], which could synergistically enhance the mitigation benefits of intercropping.

4.4. Limitations and Future Directions

Despite these insights, several limitations should be acknowledged. This experiment was conducted in a semi-arid region. To ensure crop yields and stabilize the experimental system, the nitrogen fertilizer application amount in the intercropping was kept the same as that in the monoculture. This might have masked the GHGI emission reduction effect brought about by intercropping. Furthermore, the research period was short, and it was difficult to observe the inhibitory effect of long-term soil organic carbon accumulation on greenhouse gas emissions, the mitigation potential derived from the evolution of rhizosphere microbial communities under long-term legume-cereal intercropping, and the long-term improvement in emission reduction efficiency of the intercropping system following iterative crop growth cycles. In addition, the row ratio employed in the present study may not represent the optimal pattern for maximizing soil greenhouse gas mitigation effects.

Future continuous research will be conducted, and the combination of long-term observation and model simulation is the necessary path to achieve the research goals. In using relevant models, such as APSIM and DNDC, they will be integrated to better address the spatiotemporal limitations of short-term field experiments, quantify the dynamic evolution of soil carbon and nitrogen cycles in long-term intercropping systems, and assess the comprehensive impacts of the interactions among climatic factors on greenhouse gas emissions. Through model parameter calibration and validation, the dynamic changes in GHGI under different intercropping durations and planting patterns will be accurately predicted, and the critical period and driving threshold for the transition of the intercropping system from non-significant short-term effects to stable long-term emission reduction will be identified. Long-term intercropping systems exhibit remarkable ecological and economic potential in modern agriculture and are particularly aligned with the current dual carbon goals. Despite practical obstacles, such as poor mechanization compatibility and high management complexity, successful typical cases have been achieved through technological innovation, policy guidance and systematic integration. This practice demonstrates high feasibility for popularization and application, especially in small and medium-sized farms, ecological agriculture demonstration zones and even some large-scale farms. In the future, intercropping is expected to become an important component of sustainable intensive agriculture.

5. Conclusions

Under the conditions of this study, maize/peanut intercropping had the potential to reduce the emissions of carbon dioxide and nitrous oxide in the soil compared to monoculture uniformity. Although no significant land use advantage was observed at the field scale, when crop yields were taken into account for assessment, GWP and GHGI were lower than those of monoculture uniformity. Microclimatic conditions and soil structural characteristics played a dominant role in regulating greenhouse gas emissions within this intercropping system. Our results provide empirical evidence on greenhouse gas emission responses of short-term intercropping in semi-arid regions and highlight the importance of optimizing intercropping duration and field management strategies to more fully realize the potential ecological benefits of intercropping systems.