Optimising Nitrogen Fertiliser Management in a Goji Berry–Alfalfa Intercropping System for Dual Benefits of Emissions Reduction and Yield Enhancement in Arid Regions

Abstract

Background: Amidst the pressing need to balance global food security and climate governance, achieving synergistic optimisation between crop yield enhancement and agricultural greenhouse gas reduction has become the central imperative for advancing the transition to green agriculture. Purpose: To investigate the effects of cropping systems and nitrogen fertiliser application on goji berry production systems in arid regions. Method: This study employed two cropping systems (goji berry–alfalfa intercropping (I), goji berry monocropping (M)), and four nitrogen application rates (N0 (0 kg ha⁻¹), N1 (150 kg ha⁻¹), N2 (300 kg ha⁻¹), N3 (450 kg ha⁻¹)). The effects of planting patterns and nitrogen fertiliser regulation on the physicochemical properties of goji berry farmland soil, greenhouse gas emissions, and yield were analysed. Result: (1) Soil temperatures under I were significantly lower than under M, and nitrogen application levels, cropping systems, and the interaction between nitrogen application and cropping systems significantly influenced soil nutrients; (2) Cultivation patterns and nitrogen application levels exerted a highly significant influence on soil greenhouse gas emission fluxes in goji berry fields. CO₂ emission flux peaked under IN3 treatment (annual average: 342.45 mg m⁻² h⁻¹), while N₂O emissions peaked under MN3 (annual average 0.23 mg m⁻² h⁻¹). CH₄ absorption was highest under MN0 (annual average −0.25 mg m⁻² h⁻¹); (3) Cropping systems and nitrogen application rates significantly influence greenhouse gas indicators including cumulative CO₂ emissions, cumulative N₂O emissions, and GWP. At the same nitrogen application level, GWP decreased by 5.63% on average in M compared to I, while under the same cropping system, N3 increased by 62.45% on average in N3 compared to N0; (4) Cropping systems and nitrogen application levels significantly influenced goji berry yield and economic returns. Under the same cropping system, N2 yielded the highest goji berry production and return on investment, with I and M yielding 2768.99 kg ha⁻¹ and 4.06 and 3067.78 kg ha⁻¹ and 3.15, respectively. Conclusions: The IN2 reduced soil greenhouse gas emission fluxes, cumulative emissions, and global warming potential while simultaneously increasing goji berry yield, net revenue, and return on investment. This approach minimises land resource wastage and represents a management model for achieving high yields with reduced emissions in goji berry fields within the Yellow River diversion irrigation districts of Gansu Province and similar ecological zones.

1. Introduction

Currently, agricultural production generally faces the practical constraint of the inefficient utilisation of land resources, which has become a critical bottleneck affecting sustainable agricultural development and system resilience. Taking goji berry (Lycium barbarum L.) as an example, its cultivation pattern is relatively singular, leading to issues such as land resource wastage and low land use efficiency [1]. The fruit of the perennial deciduous goji berry shrub is widely recognised for its pharmacological properties, particularly in supporting liver and kidney function [2]. Due to its robust resilience to cold and drought, goji berry is recognised as a pioneer species in revegetation efforts [3], with its planting area expanding consistently. According to the China Goji Berry Industry Blue Book (2025) [4], by the end of 2024, the total goji berry cultivation area in China reached 1062 km², with Gansu Province accounting for 462 km², or 43.5%. In 2024, China’s dried goji berry output reached 299,600 tons, with Gansu Province contributing 140,700 tons, which accounted for 47% of the national total. The robust expansion of this industry has significantly spurred regional economic development [5] and fostered a nationwide network of more than 30,000 enterprises involved in goji berry production and trade. As a perennial forage, alfalfa also delivers extensive ecological and economic advantages [6]. Alfalfa, often termed the “king of forage” for its high productivity, rich protein content, and broad adaptability [7], can mitigate frequent soil disturbance, preserve soil microbial communities, and enhance soil water retention and nutrient cycling [8]. When intercropped with goji berry, it promotes efficient land use. Economically, alfalfa cultivation boosts unit-area output [9]. This intercropping system reduces soil evaporation, improves water-use efficiency, and through alfalfa’s nitrogen-fixing ability, supplies additional nutrients to goji berry plants [10], ultimately improving both berry yield and quality.

Arid regions account for about 41% of the world’s land area and host more than 38% of its population. Characterised by highly vulnerable ecosystems, these regions are particularly sensitive to global climate change and anthropogenic disturbances. Notably, they have experienced some of the most pronounced warming trends globally over the last century [11]. Consequently, understanding and managing soil greenhouse gas fluxes in these areas is of critical importance for global warming mitigation. In China, agricultural activities account for 6.7% of the country’s total carbon emissions and 13.6% of agricultural emissions worldwide [12], a share that exceeds the global average. According to a United Nations Intergovernmental Panel on Climate Change (IPCC) assessment [13], fertilisation and its associated activities are responsible for 13.5% of global carbon emissions. Nitrogen fertiliser is essential for improving crop yield [14], but excessive application significantly increases greenhouse gas emissions [15]. Excessive fertiliser application compromises nitrogen use efficiency in crops and drives up greenhouse gas emissions from the soil [16]. Agricultural ecosystems rank among the most sensitive and vulnerable regions responding to climate change, capable of influencing global climate dynamics by altering the exchange of three primary greenhouse gases (subsequent use of the GHG acronym)—CO₂, N₂O, and CH₄—between the soil and atmosphere [17]. Therefore, optimising nitrogen fertiliser use in cropland represents a key opportunity to mitigate GHG emissions from agricultural systems. Research on the impact of different nitrogen application rates on farmland GHG emissions is essential to enhance agricultural carbon sequestration and nitrogen fixation, aiding in climate change mitigation. Such efforts constitute a critical strategy for reducing GHG intensity in cropping systems and supporting goals related to carbon peaking and carbon neutrality.

While considerable research has examined agroforestry systems in croplands, with a focus on alterations in soil physicochemical properties [18], nutrient dynamics [19], and microbial communities [20], investigations into the combined influence of cropping patterns and nitrogen application rates on cropland GHG emissions remain scarce. Compared to conventional maize monocropping, corn and soybean intercropping significantly reduced the soil N₂O and CO₂ emissions [21], suggesting its potential as an effective practice for reducing GHG emissions from agricultural soils. A meta-analysis [22] that reviewed 52 published papers and 531 observational datasets assessed how intercropping influenced soil emissions of N₂O, CO₂, and CH₄ under different environmental conditions and management practices. Findings revealed that under alkaline soil conditions, the intercropping of cereals with legumes, the application of moderate nitrogen fertiliser, or intercropping alone substantially decreased soil N₂O emissions. In soils with high organic carbon content, intercropping contributed to lower CH₄ emissions, whereas in soils with moderate total nitrogen levels, it increased the CO₄ emissions. Using the APSIM model to evaluate the effects of straw incorporation combined with optimal nitrogen fertilisation, they concluded that this practice cannot achieve a long-term balance between crop yield enhancement and GHG reduction [23]. A synthesis of 1217 field experiments conducted on farmlands worldwide [24] demonstrated that optimised cropping systems notably improved crop water productivity, achieving an average increase of 17.6%. Furthermore, suitable agricultural management approaches can concurrently raise crop yields and lower GHG emissions from farmland soils. As an efficient cropping system, intercropping leverages the temporal and spatial complementarity of different crops [25] to effectively improve soil fertility and achieve the dual objectives of increased production and enhanced efficiency. In practical agricultural production, intercropping indirectly influences soil gas emissions by increasing surface vegetation cover, altering root growth patterns [26], enhancing rainfall infiltration capacity, and reducing the surface soil salinity levels.

The Yellow River irrigation district in Gansu Province is situated in the middle and upper reaches of the Yellow River. Goji berry constitutes the primary economic tree species in this region [27], and the vigorous development of the goji berry industry represents a key pathway for improving the local ecological environment and increasing farmers’ incomes [28]. Although the effects of intercropping and nitrogen management on GHG emissions have been relatively well-studied in annual crop systems, their specific responses and quantification in perennial shrub–forage composite systems under arid irrigated conditions remain comparatively scarce. Therefore, this two-year field study in Jingtai County, Gansu Province, aimed to bridge this knowledge gap by exploring the integrated effects of cropping systems and nitrogen fertilisation rates on GHG emissions, soil properties, and system productivity. Specifically, we addressed the following questions: (1) How do intercropping and monocropping, combined with different N rates, affect soil properties and GHG emissions? (2) Which management practice optimises the trade-off between yield, economic return, and climatic impact? We hypothesised that goji berry–alfalfa intercropping with moderate N application would enhance system sustainability by improving the soil nutrients, reducing GHG intensity, and maintaining high profitability. To test this, we analysed the effects on soil physicochemical properties, GHG fluxes, cumulative emissions, global warming potential, yield, and economic benefits to identify the optimal management pattern. The principal components of this study encompass: (1) Quantifying the emission dynamics and cumulative characteristics of soil CO₂, N₂O, and CH₄ under different planting patterns and nitrogen application rates; (2) Assessing the global warming potential, GHG emission intensity, and crop yield responses in goji berry fields under various treatments; (3) Proposing optimised management schemes suitable for arid regions that balance emission reduction and stable yields based on environmental and economic benefits. This research is anticipated to provide empirical data for reducing GHG emissions in similar ecological zones without significantly compromising goji berry yields. It aimed to contribute scientific references for advancing ecological and circular agriculture practices, thereby fostering the development of green and sustainable agricultural production systems.

2. Materials and Methods

2.1. Description of the Experimental Site

The experiment was conducted from May to September 2023 and from May to September 2024 at the Irrigation Experiment Station of the Jingtaichuan Electric Power Lift Irrigation Water Resource Utilisation Centre in Gansu Province, located in the middle and upper reaches of the Yellow River (37°12′ N, 104°5′ E, average elevation 1562 m). The experimental area features a temperate arid continental climate [29], with an annual sunshine duration of 2652 h, solar radiation of 6.18 × 10⁵ J cm⁻², mean annual temperature of 8.6 °C, average annual precipitation of 191.6 mm, evaporation of 2761 mm, and a frost-free period of 191 days. The predominant soil type is sandy loam, with basic physicochemical properties detailed in

Table 1. Basic physical and chemical properties of soil.

Bulk Density (g cm−3)	Organic Matter Content (g kg−1)	Total N Content (g kg−1)	Total P Content (g kg−1)	Total K Content (g kg−1)	Available N Content (mg kg−1)	Available P Content (mg kg−1)	Available K Content (mg kg−1)	Field Capacity (%)	pH
1.63	6.09	1.62	1.32	34.03	74.51	26.31	173	24.1%	8.11

. During the 2023 trial period, the daily average temperature was 20.61 °C, with total effective precipitation of 75.10 mm. In the 2024 trial period, the daily average temperature was 20.06 °C, with total effective precipitation of 259.02 mm. Precipitation and average temperature during the goji berry growing season are shown in Figure 1.

2.2. Experimental Design

The experiment adopted a completely randomised block design with two factors: cultivation pattern and nitrogen application level. The cultivation patterns included goji berry monocropping (M) and goji–alfalfa intercropping (I). Nitrogen application levels comprised four gradients: no nitrogen (0), low nitrogen (150 kg ha⁻¹), medium nitrogen (300 kg ha⁻¹), and high nitrogen (450 kg ha⁻¹). A total of eight treatments were established (

Table 2. Experimental design.

Treatment	Cropping System	Total N Application (kg ha−1)	Nitrogen Application 2023 (kg ha−1)			Nitrogen Application 2024 (kg ha−1)
			5/22	6/10	7/2	5/22	6/11	7/2
IN0	Intercropping	0	0	0	0	0	0	0
IN1		150	90	30	30	90	30	30
IN2		300	180	60	60	180	60	60
IN3		450	270	90	90	270	90	90
MN0	Monocropping	0	0	0	0	0	0	0
MN1		150	90	30	30	90	30	30
MN2		300	180	60	60	180	60	60
MN3		450	270	90	90	270	90	90

). Based on previous research and local practices [30], each treatment was replicated three times. The plot area was 76.5 m² (10.2 m × 7.5 m), and the experimental field layout is shown in Figure 2. On 12 April 2021, two-year-old seedlings of the cultivar ‘Ningqi 5’ were transplanted with a spacing of 1.5 m between plants and 3 m between rows. Alfalfa progressed through three key growth stages: branching, budding, and early flowering (cutting). In 2023, alfalfa was cut on 18 June, 22 July, and 16 September; in 2024, the cutting dates were 16 June, 25 July, and 19 September. Goji berry development was divided into four key growth stages: vegetative growth (late April to mid-June), full bloom (late June to early July), peak fruiting (mid-July to mid-August), and autumn fruiting (late August to mid-September).

A drip-fertigation system was employed. Inline dripper tapes (16 mm diameter, flow rate 2.0 L h⁻¹) were laid in a two-tape-per-row arrangement with a spacing of 0.3 m between tapes. Each experimental plot was fitted with a dedicated valve and water meter (accuracy: 0.001 m³) to enable precise irrigation management. Fertiliser was delivered using a Venturi-type injector. Irrigation scheduling followed thresholds expressed as a percentage of field capacity (θ_fc), maintaining a mild water deficit between 65% and 75% θ_fc, in accordance with prior experimental data [31] and regional farming practices. Basal application of phosphorus (as P₂O₅, 12%) and potassium (as K₂O, 60%) fertiliser was carried out on 22 May. Nitrogen fertiliser was split-applied during the goji berry vegetative, full-bloom, and peak-fruiting stages at a ratio of 6:2:2.

2.3. Measurement and Calculation of Indicators

2.3.1. Meteorological Data

Precipitation and air temperature were recorded throughout the experimental period using a WSTQ Tianqi compact agricultural weather station installed within the field area.

2.3.2. Soil Physicochemical Properties

To measure soil temperature, a thermometer equipped with a 5-cm probe was used. The probe was inserted 10 cm away from the static chamber, which was positioned at the plot centre. This placement targeted an area with relatively dense root distribution to better capture the influence of root activity on temperature. Measurements were taken concurrently with gas sampling.

During the summer fruiting stage of goji berry, three representative goji trees located in the central sampling area of each plot were selected. Three soil cores were obtained from the 0–10 cm layer at distances of 0.3, 0.6, and 0.9 m from the tree trunk using an auger. The collected samples were then air-dried, sieved to 2 mm, and prepared for subsequent physicochemical analysis. Soil total nitrogen was determined by the Kjeldahl method [32], soil available nitrogen was measured using the alkaline hydrolysis diffusion method [33], and soil organic carbon was analysed by the potassium dichromate external heating method [34].

2.3.3. Greenhouse Gases

Gas samples were collected from each experimental plot using a closed static chamber method. The base frame measured 60 × 60 × 15 cm and was inserted 10 cm into the soil. Following each fertilisation and irrigation (or effective rainfall), we collected gas samples at progressively extended intervals (e.g., Days 1–2, Days 4–5) to accurately capture the peak phase and subsequent decay dynamics of greenhouse gas emissions. During non-fertilisation periods, background sampling was conducted at 1–2 week intervals to monitor long-term trends and baseline levels of greenhouse gas emissions within the system. The static chamber, sized 60 × 60 × 60 cm, was placed on the base frame during gas sampling and sealed by filling the water groove on the upper part of the base with water. A thermometer was inserted through the central port on the top of the chamber. The left side of the chamber was equipped with a gas sampling port and a connector for the fan wiring. Inside the chamber, two 12 V small fans were installed to mix the air, powered by an external battery. The chamber was placed on the base frame only during gas sampling events; otherwise, it was removed to maintain consistent environmental conditions (e.g., light, temperature, and humidity) inside and outside the chamber. Gas sampling was conducted between 9:00 and 11:00 a.m. After setting up the chamber, a 3 min stabilisation period was allowed. A 50 mL syringe was then inserted through the gas sampling port to extract 40 mL of gas. Sampling was repeated at 10 min intervals, yielding four samples per session. During each sampling, the temperature inside the chamber (read from the chamber thermometer) and the soil temperature (measured by a thermometer placed 10 cm from the chamber) were recorded. Collected gas samples were analysed within two days using a gas chromatograph (Shimadzu GC-2010 Pro, Shimadzu Corporation, Kyoto, Japan). Following measurement of the CO₂, N₂O, and CH₄ concentrations, the corresponding fluxes of the GHG were calculated [35] as shown in Equations (1)–(4):

(1)

Fluxes

$$\mathrm{F}=\mathsf{\rho }·\mathrm{H}·{\displaystyle \frac{\mathrm{dc}}{\mathrm{dt}}}·{\displaystyle \frac{273}{273+\mathrm{T}}}$$

(1)

F—emission flux of CO₂, N₂O, or CH₄, mg m⁻² h⁻¹; ρ—density of CO₂, N₂O, or CH₄ under standard conditions, where CO₂ = 1.977 kg m⁻³, N₂O = 1.962 kg m⁻³, CH₄ = 0.717 kg m⁻³; H—effective height inside the static chamber, 0.6 m; dc/dt—gas concentration change, mL m⁻³ h⁻¹; T—average temperature inside the static chamber during sampling, °C.

(2)

Cumulative Emissions

$$\mathrm{CE}={\displaystyle \sum }_{i=1}^n({\displaystyle \frac{{\mathrm{F}}_i{ + \mathrm{F}}_{i+1}}{2}}{) \times (\mathrm{t}}_{i+1}{ - \mathrm{t}}_i) \times 24$$

(2)

CE—cumulative emissions of CO₂, N₂O, or CH₄, mg m⁻²; F_i—emission flux of CO₂, N₂O, or CH₄ at the i-th sampling, mg m⁻² h⁻¹; F_i+1—emission flux of CO₂, N₂O, or CH₄ at the (i+1)-th sampling, mg m⁻² h⁻¹; t_i—time of the i-th sampling, d; t_i+1—time of the (i+1)-th sampling, d.

(3)

Global Warming Potential

$$\mathrm{GWP}=\mathrm{CE}\left({\mathrm{CO}}_2\right)+\mathrm{CE}\left({\mathrm{N}}_2\mathrm{O}\right) \times 273+\mathrm{CE}\left({\mathrm{CH}}_4\right) \times 27$$

(3)

GWP—global warming potential, kg ha⁻¹; CE(CO₂), CE(N₂O), and CE(CH₄) represent the cumulative emissions of CO₂, N₂O, and CH₄, respectively, mg m⁻². According to the Greenhouse Gas Protocol jointly developed by the World Resources Institute and the World Business Council for Sustainable Development, and based on the IPCC Global Warming Potential Values from the Sixth Assessment Report, over a 100 year time horizon, the global warming potential per unit mass of N₂O and CH₄ is 273 times and 27 times that of CO₂, respectively.

(4)

Greenhouse Gas Index

$$\mathrm{GHGI}={\displaystyle \frac{\mathrm{GWP}}{\mathrm{Y}}}$$

(4)

GHGI—greenhouse gas intensity, GWP—global warming potential, kg ha⁻¹; Y—yield, kg ha⁻¹.

2.3.4. Yield Measurement and Economic Analysis

From the beginning of the peak fruiting stage of goji berry, fruits were harvested from each plot every seven days. For every harvest, plants were individually processed through the stages of picking, dewaxing, and drying. The fruits were weighed using a weighing method, and the yield per unit area for each treatment (Y, kg ha⁻¹) was calculated based on the plot area. In the intercropping treatments, alfalfa was cut when approximately 50% of the plants were in bloom, leaving a stubble height of 5 cm. First, a representative 1 m² area of alfalfa was selected and its fresh weight recorded on-site. Following this, the sample was placed in an oven at 105 °C for 30 min to deactivate enzymes. Finally, it was dried at 65 °C until a constant weight was achieved. Alfalfa yield was calculated accordingly.

A comprehensive economic analysis was conducted to evaluate the profitability of each treatment based on the measured yields and prevailing local market prices. Urea: 3.25 CNY kg⁻¹, calcium superphosphate: 1 CNY kg⁻¹, potassium chloride: 4 CNY kg⁻¹, irrigation water and electricity: 0.33 CNY m⁻³; labour costs were calculated at 150 CNY per person per day, fresh goji berry picking fee: 3.5 CNY kg⁻¹, and other costs for each treatment were calculated based on actual conditions; average price of dried goji berries: 36 CNY kg⁻¹, average price of alfalfa hay: 2.1 CNY kg⁻¹. Economic indicators were calculated as shown in Equations (5)–(8):

(1)

Total Revenue (TR)

TR = (Y_g × P_g) + (Y_a × P_a)

(5)

Y_g and Y_a are the yields of dried goji berries and alfalfa hay (kg ha⁻¹), respectively, and P_g and P_a are their corresponding market prices (CNY kg⁻¹).

(2)

Total Cost (TC)

T_C = C_f + C_i + C_l + C_o

(6)

C_f is the cost of fertilisers (urea, calcium superphosphate, potassium chloride), C_i is the cost of irrigation water and electricity, C_l is the total labour cost (including a specific picking fee for goji berries), and C_o represents other operational expenses calculated based on actual conditions.

(3)

Net Revenue (NR)

NR = TR − TC

(7)

(4)

Return on Investment (ROI)

ROI = TCNR

(8)

2.3.5. Entropy Weight-TOPSIS Model

(1)

Determining the weights of indicators using the entropy weight method [36]

Given m evaluation objects and n evaluation indicators, the values of each evaluation indicator Y_ij (i = 1, 2, 3, …, m; j = 1, 2, 3, …, n) are normalised to obtain the normalised indicator values X_ij, and then the weights W_ij of each indicator are calculated, as shown in Equations (9)–(18):

$$P_{ij} ={\displaystyle \frac{X_{ij}}{{{\displaystyle \sum }}_{i=1}^m{X}_{ij}}}$$

(9)

$$e_j=-[\ln m{]}^{-1}\times {\displaystyle \sum }_{i=1}^m[P_{ij}\times \ln P_{ij}]$$

(10)

$$d_j=1-e_j$$

(11)

$$W_j={\displaystyle \frac{1-e_j}{{{\displaystyle \sum }}_{j=1}^n(1-e_j)}}$$

(12)

In the formula, P_ij is the contribution degree of the i-th evaluation object for the j-th indicator; e_j is the information entropy value; d_j is the information utility value; W_j is the weight obtained for each indicator (%).

(2)

Comprehensive evaluation using the TOPSIS method

(1)

Construction of the weighted normalised matrix

$$R_{ij}=W_j\times X_{ij}$$

(13)

(2)

Determination of the positive ideal solution R⁺ and the negative ideal solution R⁻

$$R^{+}=max(R_{i1},R_{i2},\text{…},R_{im})$$

(14)

$$R^{-}=min(R_{i1},R_{i2},\text{…},R_{im})$$

(15)

(3)

Optimal solution distance D⁺ and worst solution distance D⁻

$$D^{+}=\sqrt{{\displaystyle \sum }_{j=1}^n{(R_{ij}-{R_j}^{+})}^2}$$

(16)

$$D^{-}=\sqrt{{\displaystyle \sum }_{j=1}^n{(R_{ij}-{R_j}^{-})}^2}$$

(17)

(4)

Relative Closeness C_i

$$C_i={\displaystyle \frac{D_i^{-}}{D_i^{+}+D_i^{-}}}$$

(18)

Radar charts were created in Origin 2022 to visually compare the relative closeness (Ci) values of different treatments obtained from the entropy weight-TOPSIS analysis.

2.4. Data Processing and Statistical Analysis

Data were processed using Microsoft Excel 2021. Statistical analyses were performed with SPSS 20.0 (IBM Corp., Armonk, NY, USA). A two-way analysis of variance (ANOVA) was employed to evaluate the main effects of cropping system (D), nitrogen application rate (N), and their interaction (D × N). The significance of these effects in the results tables is denoted as: ** for p < 0.01, * for p < 0.05, and ns for p > 0.05. When a significant interaction was detected (p < 0.05), simple effects analysis was conducted to examine the effect of one factor at each level of the other factor. When the interaction was not significant, the main effects were interpreted. For mean separation, the Least Significant Difference (LSD) test was used, with differences considered significant at p < 0.05. Pearson correlation analysis was conducted to explore the relationships between key variables, with significance set at p < 0.05. Figures were generated using Origin 2022 (OriginLab Corporation, Northampton, MA, USA).

3. Results

3.1. Soil Physicochemical Properties

3.1.1. Soil Temperature

During the monitoring of GHG emissions throughout the growth period of goji berry, soil temperature at 5 cm depth was measured simultaneously. Observations indicated a consistent seasonal trend across all treatments: temperatures were highest in summer and lower in spring and autumn (Figure 3). Soil temperatures first rose and then fell throughout the goji berry growth cycle. Throughout the monitoring period, soil temperatures in goji berry intercropping systems remained lower than in monocropping systems, showing a gradual, undulating increase. Soil temperatures rose slightly with increased nitrogen application rates. Following the first and second alfalfa harvests, soil temperatures in the goji berry intercropping system rose rapidly. As alfalfa regrew, these temperatures gradually decreased. Soil temperatures across all treatments peaked in mid-July, with maximum values of 28.33 °C and 26.71 °C recorded in the two years. After September, soil temperatures reached their lowest levels during the observation period.

3.1.2. Soil Nutrients

Different cultivation patterns and nitrogen application rates, as well as their interaction, had significant effects on soil physicochemical properties (p < 0.05,

Table 3. Effects of planting pattern and nitrogen application on the soil physicochemical properties.

Year	Treatment	Total Nitrogen (g kg−1)	Available Nitrogen (mg kg−1)	Soil Organic Carbon (g kg−1)
2023	IN0	1.25 ± 0.07 aD	76.84 ± 2.01 aC	8.07 ± 0.07 aB
	IN1	1.60 ± 0.05 bC	89.07 ± 3.81 aB	8.34 ± 0.11 aA
	IN2	2.39 ± 0.13 aB	97.26 ± 2.95 aAB	8.41 ± 0.13 aA
	IN3	2.61 ± 0.15 aA	104.15 ± 7.02 aA	8.36 ± 0.14 aA
	MN0	1.37 ± 0.04 aD	56.46 ± 2.25 bC	7.84 ± 0.05 bB
	MN1	1.84 ± 0.07 aC	62.98 ± 3.52 bC	8.05 ± 0.16 bB
	MN2	2.03 ± 0.04 bB	79.86 ± 4.96 bB	8.38 ± 0.09 aA
	MN3	2.32 ± 0.11 bA	88.44 ± 3.20 bA	8.34 ± 0.16 aA
	ANOVA
	D	ns	**	*
	N	**	**	**
	D × N	**	ns	ns
2024	IN0	1.12 ± 0.10 bD	70.64 ± 2.37 aC	7.92 ± 0.03 bB
	IN1	1.74 ± 0.17 bC	88.42 ± 3.14 aB	8.43 ± 0.15 aA
	IN2	2.20 ± 0.08 aB	90.15 ± 4.68 aB	8.57 ± 0.07 aA
	IN3	2.76 ± 0.22 aA	99.27 ± 2.05 aA	8.42 ± 0.20 aA
	MN0	1.25 ± 0.05 aC	59.28 ± 3.92 bC	8.08 ± 0.02 aB
	MN1	1.94 ± 0.14 aB	66.40 ± 4.51 bC	8.23 ± 0.18 bAB
	MN2	2.11 ± 0.19 aAB	80.17 ± 2.86 bB	8.25 ± 0.05 bAB
	MN3	2.25 ± 0.11 bA	91.05 ± 3.77 bA	8.39 ± 0.19 aA
	ANOVA
	D	ns	**	ns
	N	**	**	**
	D × N	**	*	*

Note: Within the same year, lowercase letters indicate differences among cropping systems at the same nitrogen application rate; uppercase letters indicate differences among nitrogen application rates within the same cropping system. D represents the planting pattern, N represents the nitrogen application level, and D × N represents their interaction. ** indicates a highly significant difference (p < 0.01), * indicates a significant difference (p < 0.05), and ns indicates no significant difference. Values are means ± standard error (SE) of three replicates (n = 3).

). Under the same cultivation pattern, nitrogen application significantly increased TN, AN and SOC compared to the N0 treatment. Under the same cultivation pattern, TN and AN gradually increased with increasing nitrogen application rates, while SOC increased initially and then decreased with increasing nitrogen application. The TN and AN contents were highest under the IN3 treatment, reaching 2.61 g kg⁻¹ and 104.15 mg kg⁻¹ in the first year, and 2.76 g kg⁻¹ and 99.27 mg kg⁻¹ in the second year, representing average annual increases of 15.94–127.61% and 8.06–75.96%, respectively, compared to other treatments. Nitrogen application level had a highly significant effect on SOC content (p < 0.01), with SOC showing an increasing trend as nitrogen application increased. The SOC content was higher under intercropping conditions than under monocropping. The SOC content was highest under the IN2 treatment, measuring 8.41 g kg⁻¹ and 8.57 g kg⁻¹ in the two years, respectively.

3.2. Soil Greenhouse Gas Emission Flux

3.2.1. CO₂ Flux

CO₂ fluxes in goji berry fields displayed clear seasonal dynamics, consistently showing net emissions over the entire growth period and therefore functioning as a carbon source (Figure 4). Under the same cultivation pattern, CO₂ emission flux followed the order N3 > N2 > N1 > N0, compared to N0, N1, and N2, the N3 treatment increased the annual average CO₂ emission flux by 53.52%, 32.21%, and 13.04%, respectively. Furthermore, the N3 treatment significantly elevated CO₂ emissions from June to mid-August. Under the same nitrogen application level, intercropping resulted in higher CO₂ emissions than monocropping, with the M treatment reducing the annual average CO₂ emissions by 8.11% compared to the I treatment. The CO₂ emission flux under different cultivation patterns and nitrogen application levels displayed a “single-peak curve”, with peaks occurring in mid-July in both years. This pattern aligns with the peak period of soil temperature during the mid-growth season, suggesting that higher temperatures promote soil microbial activity and organic matter decomposition, thereby driving the seasonal peak in CO₂ emission flux. The highest CO₂ emission flux was observed under the IN3 and MN3 treatment, measuring 341.30 mg m⁻² h⁻¹, 343.60 mg m⁻² h⁻¹ and 314.71 mg m⁻² h⁻¹, 326.12 mg m⁻² h⁻¹ in the two respective years. After September, no significant differences in CO₂ emission flux were observed among the treatments.

3.2.2. N₂O Flux

The N₂O flux in goji berry fields remained in an emission state throughout the entire growth period. Under the same cultivation pattern, the N₂O emission flux followed the order N3 > N2 > N1 > N0 (Figure 5). The N3 treatment significantly increased N₂O emissions during the growth period. Compared with the N0, N1, and N2 treatments under the same cropping pattern, the average annual increases were 138.36%, 52.46%, and 22.97%, respectively. Under the same nitrogen application level, the N₂O emission flux in the I treatment was reduced by 12.89% and 15.11% over the two years, respectively, compared to the M treatment. Throughout the growth cycle of goji berry, N₂O emissions displayed three pronounced peaks, each corresponding to one of the three fertiliser applications (which coincided with irrigation). The IN3 and MN3 treatments yielded the highest N₂O flux, with two-year averages of 0.19 mg m⁻² h⁻¹ and 0.23 mg m⁻² h⁻¹, respectively. After August, no significant differences in N₂O emissions were detected among the treatments.

3.2.3. CH₄ Flux

The CH₄ flux in goji berry fields exhibited an overall uptake (absorption) state throughout the growth period. The overall CH₄ flux ranged between −0.25~0.02 mg m⁻² h⁻¹. Under the same cultivation pattern, the CH₄ uptake flux followed the order N0 > N1 > N2 > N3 (Figure 6), indicating that nitrogen application inhibited CH₄ uptake, with N3 significantly reducing CH₄ absorption during the growth period. No significant differences in CH₄ flux were observed among the cultivation patterns. The CH₄ uptake flux and N₂O emission flux exhibited mirroring trends. Peaks in CH₄ uptake occurred following each of the three fertilisation events (coinciding with irrigation). The highest CH₄ uptake fluxes were observed under the IN0 and MN0 treatments, measuring −0.23 mg m⁻² h⁻¹ and −0.24 mg m⁻² h⁻¹ in the first year, and −0.24 mg m⁻² h⁻¹ and −0.25 mg m⁻² h⁻¹ in the second year, respectively. After August, the CH₄ uptake capacity weakened across all treatments, and under the N2 and N3 conditions, CH₄ emissions even occurred.

3.3. Cumulative Emissions of GHG from Soils and Global Warming Potential

Both cultivation patterns and nitrogen application levels had a highly significant impact on cumulative soil CO₂ and N₂O emissions, GWP and GHGI, while nitrogen application level alone significantly affected cumulative soil CH₄ uptake and GHGI (p < 0.01,

Table 4. Effects of planting pattern and nitrogen application on the cumulative soil GHG emissions and global warming potential.

Year	Treatment	CE(CO2) (g m−2)	CE(N2O) (mg m−2)	CE(CH4) (mg m−2)	GWP (kg ha−1)	GHGI (kg CO2-eq kg−1)
2023	IN0	410.78 ± 20.48 aD	118.29 ± 25.38 bD	−228.30 ± 33.15 aC	4369.16 ± 283.06 aD	2.20 ± 0.14 aB
	IN1	476.53 ± 20.83 aC	189.58 ± 20.93 bC	−139.79 ± 26.50 aB	5245.14 ± 272.63 aC	2.38 ± 0.12 aB
	IN2	551.09 ± 24.55 aB	238.06 ± 26.46 aB	−103.50 ± 21.00 aB	6132.87 ± 323.47 aB	2.34 ± 0.12 aB
	IN3	631.64 ± 28.84 aA	289.46 ± 25.22 bA	−40.26 ± 24.13 aA	7095.82 ± 363.81 aA	2.84 ± 0.15 aA
	MN0	380.53 ± 20.28 aD	146.43 ± 18.81 aD	−250.23 ± 27.65 aB	4137.52 ± 261.67 aD	2.06 ± 0.13 bB
	MN1	449.87 ± 21.71 aC	221.14 ± 22.59 aC	−104.23 ± 28.08 aA	5074.35 ± 286.45 aC	2.08 ± 0.12 bB
	MN2	507.97 ± 26.22 bB	269.50 ± 17.42 aB	−98.28 ± 32.90 aA	5788.9 ± 318.72 bB	1.91 ± 0.11 bB
	MN3	570.96 ± 27.35 bA	328.43 ± 27.88 aA	−77.63 ± 24.06 aA	6585.25 ± 356.12 bA	2.38 ± 0.13 bA
	ANOVA
	D	**	**	ns	*	**
	N	**	**	**	**	**
	D × N	ns	ns	ns	ns	ns
2024	IN0	439.12 ± 19.74 aD	122.19 ± 21.99 aD	−236.04 ± 31.28 aC	4661.12 ± 265.88 aD	2.23 ± 0.13 aB
	IN1	494.05 ± 22.68 aC	192.98 ± 24.55 aC	−151.99 ± 33.36 aB	5426.36 ± 302.92 aC	2.23 ± 0.12 aB
	IN2	571.85 ± 27.83 aB	243.70 ± 27.28 aB	−104.83 ± 34.71 aB	6355.57 ± 362.22 aB	2.18 ± 0.12 aB
	IN3	647.63 ± 29.70 aA	306.13 ± 25.10 aA	−41.76 ± 31.91 aA	7300.83 ± 374.16 aA	2.64 ± 0.14 aA
	MN0	375.94 ± 23.03 bD	151.16 ± 25.95 aD	−268.87 ± 31.88 aC	4099.49 ± 309.79 bD	1.79 ± 0.14 bB
	MN1	442.17 ± 22.58 bC	230.99 ± 27.21 aC	−140.86 ± 31.22 aB	5014.3 ± 308.51 bC	1.88 ± 0.12 bB
	MN2	543.82 ± 23.46 aB	285.14 ± 25.40 aB	−91.43 ± 30.81 aAB	6191.95 ± 312.35 aB	1.99 ± 0.10 aB
	MN3	612.52 ± 29.07 aA	354.77 ± 24.51 aA	−80.31 ± 26.70 aA	7072.05 ± 364.91 aA	2.49 ± 0.13 aA
	ANOVA
	D	**	**	ns	*	**
	N	**	**	**	**	**
	D × N	ns	ns	ns	ns	ns

Note: CE(CO₂): Cumulative CO₂ emissions, CE(N₂O): Cumulative N₂O emissions, CE(CH₄): Cumulative CH₄ emissions, GWP: Global warming potential, GHGI: Greenhouse gas intensity. Within the same year, lowercase letters indicate differences among cropping systems at the same nitrogen application rate; uppercase letters indicate differences among nitrogen application rates within the same cropping system. D represents the planting pattern, N represents the nitrogen application level, and D × N represents their interaction. ** indicates a highly significant difference (p < 0.01), * indicates a significant difference (p < 0.05), and ns indicates no significant difference. Values are means ± standard error (SE) of three replicates (n = 3).

). Under the same cultivation pattern, CO₂ and N₂O emissions increased with higher nitrogen application rates, whereas CH₄ uptake decreased with increasing nitrogen application (all comparisons based on absolute values, same below). Compared to N0, N1, and N2, the annual average CE(CO₂) under N3 increased by 13.24%~53.31%, and CE(N₂O) increased by 23.39%~137.66%. The CE(CH₄) under IN0 and MN0 treatments measured −228.3 mg m⁻² and −250.23 mg m⁻² in the first year, and −236.04 mg m⁻² and −268.87 mg m⁻² in the second year, respectively, whereas under N3, the average was only −59.99 mg m⁻². During the goji berry growth period, the GWP ranged from 4137.52~7095.82 kg ha⁻¹ in the first year and from 4099.49~7300.83 kg ha⁻¹ in the second year. GWP showed an increasing trend with higher nitrogen application rates, with N1, N2, and N3 treatments increasing the annual average GWP by 14.65%~62.47% compared to N0. GHGI under N3 was significantly higher than under other treatments, while no significant differences in GHGI were observed among the N0, N1, and N2 treatments. Under the same nitrogen application level, N₂O emissions in the goji berry–alfalfa intercropping system decreased by 13.48% and 15.37% over the two years compared to goji berry monocropping. Meanwhile, CO₂ emissions in monocropping decreased by 7.76% and 8.28%, while CH₄ uptake increased by 3.49% and 8.06% over the two years compared to intercropping. The GWP of the intercropping system was 5.82% and 6.10% higher on average than that of monocropping over the two years during the goji berry growth period. Considering only goji berry yield under different cultivation patterns, GHGI was higher in intercropping than in monocropping.

3.4. Goji Berry Yield and Economic Benefits

Both cultivation pattern and nitrogen application rate had a highly significant effect on goji berry yield and farmland economic benefits (

Table 5. Effects of planting pattern and nitrogen application on goji berry yields and farm economic benefits.

Year	Treatment	Wolfberry Yield (kg ha−1)	Total Revenue (×104 CNY ha−1)	Total Cost (×104 CNY ha−1)	Net Revenue (×104 CNY ha−1)	Return on Investment
2023	IN0	1983.96 ± 89.27 aC	9.15 ± 0.41 aC	2.08 ± 0.08 aB	7.07 ± 0.31 aC	3.40 ± 0.19 aB
	IN1	2205.37 ± 134.38 bB	10.62 ± 0.64 aB	2.24 ± 0.12 aB	8.38 ± 0.51 aB	3.74 ± 0.28 aAB
	IN2	2622.94 ± 81.24 bA	12.34 ± 0.38 aA	2.53 ± 0.05 bA	9.81 ± 0.3 aA	3.88 ± 0.15 aA
	IN3	2497.04 ± 134.64 bA	11.5 ± 0.62 aAB	2.66 ± 0.14 aA	8.84 ± 0.47 aB	3.32 ± 0.23 aB
	MN0	2009.48 ± 94.98 aD	7.23 ± 0.34 bD	2.11 ± 0.09 aC	5.12 ± 0.24 bD	2.43 ± 0.16 bC
	MN1	2438.96 ± 121.55 aC	8.78 ± 0.43 bC	2.35 ± 0.13 aB	6.43 ± 0.33 bC	2.74 ± 0.18 bB
	MN2	3028.36 ± 94.78 aA	10.9 ± 0.34 bA	2.62 ± 0.08 aA	8.28 ± 0.25 bA	3.16 ± 0.13 bA
	MN3	2768.14 ± 64.60 aB	9.97 ± 0.23 bB	2.57 ± 0.07 aA	7.40 ± 0.17 bB	2.88 ± 0.09 bB
	ANOVA
	D	**	**	ns	**	**
	N	**	**	**	**	**
	D × N	*	ns	ns	ns	ns
2024	IN0	2094.12 ± 67.84 bC	9.49 ± 0.30 aC	2.19 ± 0.07 bC	7.3 ± 0.23 aC	3.33 ± 0.14 aC
	IN1	2429.48 ± 105.83 bB	11.76 ± 0.51 aB	2.42 ± 0.11 aB	9.34 ± 0.36 aB	3.86 ± 0.21 aAB
	IN2	2915.03 ± 155.91 bA	14.01 ± 0.74 aA	2.68 ± 0.15 aA	11.33 ± 0.6 aA	4.23 ± 0.27 aA
	IN3	2763.87 ± 95.54 aA	12.72 ± 0.43 aB	2.67 ± 0.09 aA	10.05 ± 0.34 aB	3.76 ± 0.16 aB
	MN0	2284.06 ± 90.56 aC	8.22 ± 0.32 bC	2.34 ± 0.13 aB	5.88 ± 0.23 bC	2.51 ± 0.13 bC
	MN1	2672.57 ± 64.02 aB	9.62 ± 0.23 bB	2.56 ± 0.06 aA	7.06 ± 0.16 bB	2.76 ± 0.08 bBC
	MN2	3107.19 ± 158.29 aA	11.18 ± 0.56 bA	2.71 ± 0.10 aA	8.47 ± 0.43 bA	3.13 ± 0.20 bA
	MN3	2839.52 ± 120.67 aB	10.22 ± 0.43 bB	2.63 ± 0.12 aA	7.59 ± 0.32 bB	3.00 ± 0.16 bAB
	ANOVA
	D	**	**	ns	**	**
	N	**	**	**	**	**
	D × N	ns	ns	ns	*	ns

Note: Within the same year, lowercase letters indicate differences among cropping systems at the same nitrogen application rate; uppercase letters indicate differences among nitrogen application rates within the same cropping system. D represents the planting pattern, N represents the nitrogen application level, and D × N represents their interaction. ** indicates a highly significant difference (p < 0.01), * indicates a significant difference (p < 0.05), and ns indicates no significant difference. Values are means ± standard error (SE) of three replicates (n = 3).

, p < 0.01). Under the same cultivation pattern, goji berry yield initially increased and then decreased with increasing nitrogen application, with the maximum yield observed under the N2 condition. Compared to N0, N1, and N3, the average yield under N2 increased by 39.44%, 19.77%, and 7.41%, respectively. Under the same nitrogen application level, the yield of goji berry monocropping was on average 8.39% higher than that of goji–alfalfa intercropping. The highest goji berry yield was achieved under the MN2 treatment, measuring 3028.36 kg ha⁻¹ and 3107.19 kg ha⁻¹ in the two years, respectively. Cultivation pattern and nitrogen application rate also had a highly significant impact on total income, net profit, and benefit–cost ratio (p < 0.01). The yields of goji berry and alfalfa directly determined the total revenue. Based on the economic analysis of each treatment, under the same cultivation pattern, both total revenue and net profit reached their maximum under the N2 condition. For the IN2 treatment, the total revenue and net profit over the two years were 123,400 CNY ha⁻¹ and 98,100 CNY ha⁻¹ in the first year, and 111,800 CNY ha⁻¹ and 113,300 CNY ha⁻¹ in the second year, respectively. Compared to IN0, IN1, and IN3, the total revenue under IN2 increased by an average of 41.36%, 17.74%, and 8.79% per year, while the net profit increased by an average of 46.91%, 19.13%, and 12.05% per year. For the MN2 treatment, the total revenue and net profit over the two years were 109,000 CNY ha⁻¹ and 82,800 CNY ha⁻¹ in the first year, and 111,800 CNY ha⁻¹ and 84,700 CNY ha⁻¹ in the second year, respectively. Compared to MN0, MN1, and MN3, the total revenue under MN2 increased by an average of 42.91%, 20.00%, and 9.36% per year, while net profit increased by an average of 55.27%, 24.17%, and 11.74% per year. Under the same nitrogen application level, both total revenue and net profit were higher for goji–alfalfa intercropping than for goji monocropping, with average increases of 20.32% and 25.23% over the two years, respectively. No significant difference in total cost was observed among cultivation patterns (p > 0.05), with total costs across all treatments ranging from 20,800~27,100 CNY ha⁻¹. The highest return on investment was achieved under the IN2 treatment, with values of 3.88 and 4.23 in the two respective years.

3.5. Correlation Analysis

Pearson correlation analysis was conducted to explore the relationships between key variables, based on data from three replicates per treatment (Figure 7). The results showed that yield was significantly positively correlated (p < 0.05) with TN, ST, N₂OF, CH₄F, N₂OCE, CH₄CE, and GWP. TN exhibited significant positive correlations (p < 0.05) with AN, SOC, and greenhouse gas-related indicators. AN was highly significantly positively correlated (p < 0.01) with SOC, GWP, and GHGI. GWP showed highly significant positive correlations (p < 0.01) with both GHG emission fluxes and emission intensity. A highly significant positive correlation was also observed between GHG emission fluxes and emission intensity (correlation coefficient ≥ 0.86, p < 0.01). Furthermore, ST was significantly positively correlated (p < 0.05) with N₂OF and N₂OCE, indicating that soil temperature is a key environmental factor influencing N₂O emissions.

3.6. Comprehensive Evaluation of the Effects of Cultivation Patterns and Nitrogen Application Rates on Goji Berry

An evaluation and analysis of goji berry indicators under different cultivation patterns and nitrogen management strategies was conducted using the entropy-weighted TOPSIS method. As shown in

Table 6. Weights of each indicator calculated based on the TOPSIS method.

Indicator	Information Entropy	Information Utility Valued	Weighting (%)
GWP	0.87	0.13	18.29
Yield	0.87	0.13	18.18
ST	0.88	0.12	16.10
AN	0.88	0.12	15.86
TR	0.89	0.11	14.87
ROI	0.88	0.12	16.71

Note: GWP (Global Warming Potential), ST (Soil Temperature), AN (Available Nitrogen), TR (Total Revenue), ROI (Return on Investment).

, GWP (18.29%) and yield (18.18%) exhibited the largest variability among all indicators, and therefore were assigned the highest weights in the evaluation. This outcome directly reflects the central research focus of this study—namely, the trade-off between greenhouse gas emissions and crop productivity. These were followed by the benefit–cost ratio (16.71%) and soil temperature (16.10%). Available nitrogen (15.86%) maintained a moderate level of influence, as reflected by its information entropy value. Total revenue (14.87%) had the lowest weight, which corresponds to its relatively high information entropy value (0.89), indicating that its data distribution tended to be more homogeneous. Analysis of the weighted data using the TOPSIS method (Figure 8) revealed that the IN2 treatment achieved the highest comprehensive score (0.68), indicating the best overall performance. This was followed by IN1 and IN3, while MN0 received the lowest comprehensive score (0.38).

4. Discussion

4.1. Effects of Soil Physicochemical Properties on Soil Greenhouse Gas Emissions

The release of GHG from soil is fundamentally a biochemical reaction process driven by microorganisms, and its emission flux is jointly regulated by the dynamic changes in soil temperature and various physicochemical properties [37]. Correlation analysis revealed a marked positive relationship between soil temperature and N₂O emission flux, whereas the influence of soil nutrients on GHG emissions was more pronounced. This suggests that the size of the soil carbon pool serves as a fundamental material basis affecting GHG emissions. As soil organic carbon content increases [38], microbial access to carbon sources becomes more abundant, thereby strengthening soil respiration as well as nitrification and denitrification processes, which in turn enhances GHG release. This observation is consistent with the IPCC report’s conclusion that “the decomposition of soil organic matter constitutes a major source of GHG emissions in agricultural systems” [39]. The synergistic transformation of soil carbon and nitrogen pools is a critical link regulating GHG release. The application of nitrogen fertiliser increases the soil nitrogen availability, indirectly accelerating the decomposition of organic carbon, thereby creating a chain effect of “nitrogen-driven—carbon pool decomposition—GHG emission” [40]. Additionally, hidden regulatory factors such as soil water content and pH may indirectly affect GHG emissions by influencing microbial community structure [41]. Some studies have found that the emission-enhancing effect of nitrogen application is more significant in acidic soils [42]. Therefore, these soil factors must be fully considered in the future to comprehensively analyse the integrated impact of soil physicochemical properties on GHG emissions.

4.2. Effects of Cultivation Patterns and Nitrogen Application Levels on Soil Greenhouse Gas Emissions

Under the same cultivation pattern, CO₂ emission fluxes under medium- to high-nitrogen treatments (N2 and N3) were significantly higher than those under low- and no-nitrogen treatments (N1 and N0). This is likely because nitrogen fertiliser input increases the content of mineralisable nitrogen in the soil, providing an abundant nitrogen source for microbial respiration, thereby promoting the decomposition of soil organic matter and the release of CO₂ [43]. The CO₂ emissions from the intercropping system were higher than those from the monocropping system. This may be attributed to the higher total biomass (especially belowground root biomass) in the intercropping system compared to monocropping. The different root distributions of alfalfa and goji berry increased both the total amount and activity of living roots, leading to greater CO₂ emissions from root respiration. Moreover, intercropping enhances the soil micro-environment, leading to increased soil moisture and improved aeration, which accelerates the decomposition of existing soil organic matter and results in greater CO₂ release. Raseduzzaman et al. [44] found that maize–soybean intercropping systems emitted less CO₂ compared to monocropping, a result that differs from the present study. This inconsistency mainly arises from differences in carbon input mechanisms and microclimate regulation between the two systems. The maize–soybean system features a dense canopy that lowers soil temperature and reduces evaporation, thereby suppressing microbial activity, while its carbon input primarily comes from relatively stable aboveground litter. In contrast, the goji berry–perennial alfalfa combination has a sparse canopy with limited shading, good soil aeration, and the nitrogen-fixing requirement of alfalfa promotes the input of root exudates, which stimulates rhizosphere microbial respiration. This suggests that the effect of intercropping on the soil carbon cycle is determined by the functional group composition of the plants and the micro-environment they create [45].

Cultivation patterns and nitrogen application levels are critical management practices in farmland ecosystems [46]. This study revealed that N₂O emission fluxes from both monocropped and intercropped goji berry systems exhibited an increasing trend with higher nitrogen application. N₂O emission peaks were consistently measured after each of the three nitrogen-fertilisation and irrigation events during the goji berry growing season. This is primarily attributed to the direct stimulation of key soil nitrogen transformation processes by external nitrogen input. As an intermediate product of nitrification and denitrification processes [47], N₂O formation is strongly influenced by the concentration of available nitrogen (NH₄⁺ and NO₃⁻) in the soil. High nitrogen application rates substantially increased the soil inorganic nitrogen pool, supplying abundant substrate for microbial activity and thus stimulating N₂O production. Although the goji berry–alfalfa intercropping system improved nitrogen use efficiency through plant interactions, excessive nitrogen application still resulted in nitrogen accumulation in the soil, leading to higher emissions with increased nitrogen input. This result highlights that the over-application of nitrogen fertiliser [48] is a primary factor driving GHG emissions from agricultural lands, and optimising nitrogen management is crucial for simultaneously achieving high crop yields and GHG reduction. Compared to goji berry monocropping, the goji berry–alfalfa intercropping system reduced both N₂O emission flux and cumulative emissions to some degree. This reduction may be attributed to improved soil structure under intercropping, which enhanced soil gas exchange capacity [49]. As a leguminous plant, alfalfa has nitrogen-fixing capabilities. In the intercropping system, the fixed nitrogen may be more efficiently taken up and utilised directly by goji berry plants, decreasing the accumulation of nitrate and ammonium nitrogen in the soil and consequently reducing N₂O production. This observation is consistent with the findings of Mirzaei et al. [50].

Research indicates that farming systems do not significantly affect CH₄ uptake, yet nitrogen fertilisation exerts a clear influence on CH₄ absorption dynamics. High nitrogen inputs enhance microbial activity in soil [51], accelerating the breakdown of organic matter and consequently raising GHG releases. Nitrogen application predominantly suppresses CH₄ uptake, likely by modifying soil redox potential, which impedes CH₄ oxidation and assimilation processes [52]. Elevated ammonium-N levels inhibit methane-oxidising bacteria while stimulating ammonia oxidation, indirectly boosting N₂O emissions. Concurrently, these conditions provide more substrates for methanogenic archaea, resulting in increased CH₄ production [53].

The emission dynamics of N₂O and CH₄ showed a clear inverse relationship throughout the study, with peaks in one gas corresponding to troughs in the other—a mirrored trend also observed by Li et al. [54]. Within this system, intercropping collectively functioned as a net sink for CH₄. Among the nitrogen treatments, N3 generated higher GHG fluxes compared to N0, N1, and N2. In the context of global climate change—a foremost contemporary challenge [55]—agriculture remains a major source of anthropogenic emissions, underscoring the importance of managing field practices to mitigate this impact. Therefore, appropriate crop production technologies serve as an important approach to reducing soil GHG emissions. This study did not provide an in-depth explanation of the microbial processes and rhizosphere interaction mechanisms driving changes in GHG emissions. Future research is recommended to incorporate molecular biology techniques to analyse the response mechanisms of soil functional microbial communities to GHG emissions.

4.3. Effects of Cultivation Patterns and Nitrogen Application Levels on Goji Berry Yield and Economic Benefits

As a characteristic cash crop in the arid regions of northwestern China, the enhancement of goji berry yield and economic benefits serves as the core objective of cultivation management [56], with the optimisation of cultivation patterns and nitrogen application levels being a key pathway to achieving this goal. This study found that the intercropping pattern combined with the medium-nitrogen treatment (N2) resulted in significantly higher net profit and return on investment for goji berry compared to monocropping and low- or high-nitrogen treatments. These results are consistent with Qaidam Basin studies on goji berry water and nutrient management, where moderated nitrogen application was shown to enhance economic outcomes [57]. From a yield-formation perspective, optimised cultivation systems improve productivity by facilitating resource complementarity. Intercropping utilises niche differentiation between species to achieve more efficient capture of light, thermal, water, and nutrient resources. This approach not only minimises interspecific nutrient competition but also ameliorates soil micro-environmental conditions [58]. Consequently, root uptake of nitrogen, phosphorus, and other essential nutrients in goji berry is enhanced, leading to higher yields—an outcome that supports the ecological tenet that “well-designed cultivation systems can raise crop productivity through improved resource partitioning” [59]. Yield response to nitrogen application typically exhibits a pattern characterised by “limitation under low nitrogen, suitability under moderate nitrogen, and surplus under high nitrogen” [60]. In the present study, goji berry yield reached its maximum under the N2 treatment, while the high-nitrogen (N3) treatment resulted in reduced yield. This decline can be attributed to excessive nitrogen input, which disturbs soil nitrogen equilibrium, suppresses root development and nutrient-acquisition efficiency, and elevates the likelihood of nutrient leaching [61].

Economic analysis reveals that optimising cultivation patterns and nitrogen application levels enhances both yield and profitability by reducing production costs. In the intercropping system, the combined output of goji berry and alfalfa substantially increases the economic value per unit land area. Compared to the high-nitrogen treatment, the medium-nitrogen treatment maintains comparable yield while reducing fertiliser input, thereby lowering the proportion of fertiliser costs. In this study, the net profit under the N2 treatment was 11.8% higher than under the N3 treatment. Additionally, the ecological benefits of intercropping—such as the nitrogen-fixing capacity of alfalfa—can be translated into potential environmental value, providing further support for the green and sustainable development of the goji berry industry. The selection of inappropriate intercropping partners may intensify plant competition, impair photosynthetic efficiency, and consequently reduce yield [62]. Appropriate intercropping can improve soil structure, reduce the incidence of pests and diseases, and increase both yield and economic returns. Moderate nitrogen application promotes plant growth, enhances chlorophyll content, improves photosynthetic efficiency, and thus boosts yield [63]. Overall, the impact of cultivation patterns and nitrogen application levels on goji berry yield and economic benefits essentially reflects a balance between resource-use efficiency and production input costs. Future research could also explore the spatial dimension of intercropping growth by investigating planting density. The combination of intercropping and medium-nitrogen treatment not only improves production efficiency through resource complementarity but also reduces costs and environmental risks by optimising nutrient input. This optimal management model can provide scientific reference for enhancing quality, efficiency, and green development of the goji berry industry in the arid regions of northwestern China.

4.4. Perspectives for Optimising the Goji–Alfalfa System: Beyond Nitrogen and Planting Pattern

This study established that the IN2 planting pattern can reduce N₂O emissions while maintaining yield and economic returns, providing a foundational management strategy for green goji berry production in arid regions. However, under intercropping conditions, goji berry yield remains slightly lower than that under monoculture, indicating that resource competition persists and the synergistic potential for both production enhancement and emission reduction has not been fully realised. Future research should shift from the singular “planting pattern and nitrogen rate” dual framework toward the synergistic optimisation of multi-dimensional techniques such as cultivar selection, spatiotemporal arrangement, and canopy management, in order to construct intelligent, integrated systems that truly achieve “increased production coupled with emission reduction”.

In annual crop intercropping systems [64,65], it has been demonstrated that variety selection, strip rotation, and sowing date adjustment are effective tools for managing phenological synchrony and alleviating interspecific competition. Future work could explore screening other cultivars, such as upright, narrow-canopy goji berry varieties or dwarf-type alfalfa. By adjusting the sowing date of alfalfa to shift its peak growth period away from the photosynthetic demand peak of goji berry, canopy photosynthetic efficiency and belowground resource allocation can be optimised, potentially narrowing or even reversing the yield gap relative to monoculture while preserving the emission-reduction benefits of intercropping. The intercropping pattern examined in this study represents only a basic form of system design. More sophisticated designs based on spatio-temporal diversification—for instance, strip rotation or the introduction of a third crop—could further stimulate biological complementarity. Research has shown [66] that implementing annual row swapping and complementary root architectures in cotton–soybean strip intercropping systems can reduce interspecific competition, optimise resource distribution, increase yield, and simultaneously further lower GHG emissions. Because goji berry and alfalfa are perennial crops, frequent rotational replanting is neither practical nor cost-effective. Future investigations may explore adjusting alfalfa sowing density or optimising alfalfa cutting schedules in relation to goji berry growth stages. Such dynamic configurations could improve the soil microenvironment and directionally regulate key processes in soil carbon and nitrogen cycling, thereby synergistically enhancing system productivity and carbon-sink function over longer time scales. Canopy management is a critical agronomic lever for modulating the micro-environment in intercropping systems. In cotton–peanut intercropping research [67], timely “topping” of cotton effectively alleviated shading on peanut, thereby boosting the yield of the companion crop without compromising cotton yield. The routine pruning practiced in goji berry production offers an opportunity to actively manipulate canopy structure. Future studies could focus on how the intensity and timing of goji berry pruning alter canopy light transmission, temperature, and humidity profiles. These microclimatic changes could subsequently affect the growth of understory alfalfa, soil moisture dynamics, nitrogen mineralisation rates, and ultimately feedback to GHG emission fluxes. Furthermore, the consistently higher CO₂ emissions observed in the intercropping pattern may be attributed to enhanced rhizodeposition from alfalfa roots, which could accelerate the turnover of soil organic matter. This mechanism warrants direct validation in future studies through ¹³C labelling techniques.

Through this “layered-design, synergistic-optimisation” research paradigm, it is anticipated not only that intercropped yields could match or even surpass those of monoculture, but also that GHG emission intensity could be further reduced. Ultimately, such an approach would provide operable solutions and a theoretical framework to guide perennial agroforestry systems in arid regions toward a green, high-yielding, and low-carbon sustainable future.

5. Conclusions

Cultivation patterns and nitrogen fertilisation rates substantially affect GHG emissions and yield performance in goji berry–alfalfa intercropping systems. Optimised farming methods and nitrogen management not only mitigate GHG releases but also improve economic outcomes. Relative to monocropping, intercropping demonstrates clear benefits in lowering nitrous oxide (N₂O) emission rates. The medium nitrogen application level (N2, 300 kg ha⁻¹) is critical for mitigating GHG emissions in this system, significantly reducing emission flux (F) and global warming potential (GWP) while maintaining high goji berry yields. Furthermore, intercropping with alfalfa provides a profitable pathway with distinct N₂O reduction benefits. Additionally, the intercropping pattern demonstrated superior economic performance compared to the monoculture treatments, with total income and net profit increasing by 20.32% and 25.23%, respectively, relative to the average of all monocropping systems. Based on the comprehensive evaluation using the entropy-weighted TOPSIS model, the goji berry–alfalfa intercropping pattern (I) combined with medium nitrogen application (N2) was identified as a suitable management strategy for achieving emission reduction and high yield in goji berry cultivation in the Yellow River irrigation districts of Gansu and similar ecological regions. Looking forward, the gains established here form a foundational basis upon which future research should build—by integrating varietal selection, spatial arrangement, and adaptive canopy management—to fully unlock the productivity and environmental potential of integrated perennial cropping systems in arid landscapes.