Peanut oil is more environmentally sustainable than rapeseed oil from a carbon and nitrogen footprint perspective in China

Study area

Due to the availability and completeness of raw data, this study only considered 13 major peanut-producing regions (Hebei, Liaoning, Jilin, Anhui, Fujian, Jiangxi, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, and Sichuan), and 15 major rapeseed-producing regions (Inner Mongolia, Jiangsu, Zhejiang, Anhui, Jiangxi, Henan, Hubei, Hunan, Chongqing, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, and Qinghai) in China. The aforementioned peanut- and rapeseed-producing regions accounted for 86%–91% and 89%–98% of the national planting area from 2004 to 2023, respectively (National Bureau of Statistics of China, 2025). Additionally, they contributed 89%–93% and 94%–98% of the national production, respectively (National Bureau of Statistics of China, 2025). Due to the various natural conditions and cultivation levels, the planting area, yield, and production of peanut and rapeseed cultivation in different areas in China were different, as shown in Fig. 1. For peanut, Henan and Shandong had higher planting area, yield, and production, while Fujian, Jiangxi, and Hunan had lower planting area, yield, and production. In addition, Anhui had lower planting area and production, but its yield was the highest. For rapeseed, Hubei, Hunan, and Sichuan had higher planting area and production, while Zhejiang, Henan, Shaanxi, Gansu, and Qinghai had lower planting area and production. Jiangsu had low planting area and production, but it had higher yield. The specific geographic information of the study area was presented in Table S1.

Data sources

Diverse farming practices result in varying CF and NF of crop production over different time series. A diachronic analysis can present the temporal dynamics characteristics of the CF and NF in peanut and rapeseed production. Limited to the availability of data, we obtained agricultural input data from 2004 to 2023, which included the amounts of fertilizers, agricultural film, and seeds from the National Agricultural Cost-Benefit Data (NDRC, 2024). The quantities of farm manure, pesticides, diesel used in agricultural machinery, and irrigation electricity were estimated according to the unit cost derived from the National Agricultural Cost-Benefit Data (NDRC, 2024) and the corresponding average unit prices. The average unit prices of farm manure, pesticides, diesel, and electricity were obtained through peer-reviewed literature based on studies conducted in China, which were about 0.62 RMB kg⁻¹ (Huang, Luo & Han, 2021), 61.82 RMB kg⁻¹ (Liu, 2020), 6.00 RMB L⁻¹ (NDRC, 2021), and 0.50 RMB kwh⁻¹ (NDRC, 2021), respectively. Data on crop area and production were collected from the China Rural Statistical Yearbook (National Bureau of Statistics of China, 2025). Limited to the unavailable raw data of the related edible oil processing, energy consumption during the whole industrial chain of oil processing was derived from the national standard on the Norm of Energy Consumption Per Unit Product of Edible Vegetable Oil. According to the standards, the comprehensive energy consumption was set at 185 and 143 kgce t⁻¹ for edible oil production of peanut and rapeseed, respectively (NDRC, 2013).

System boundary

This study adopted a cradle-to-factory gate LCA approach to quantify the CF and NF of edible peanut and rapeseed oils. As illustrated in Fig. 2, the system boundary includes raw materials and oil processing sections. The CF of the crop planting phase comprised upstream production of agricultural inputs (i.e., fertilizers, pesticides, agricultural film, and seeds), N₂O emissions from in-field fertilization, carbon emissions from diesel-burning, and irrigation electricity. Similarly, the NF of the crop planting phase encompassed all direct and indirect Nr losses, including ammonia (NH₃) volatilization, emissions of nitrous oxide (N₂O) and nitric oxide (NO), nitrogen (N) leaching, and N runoff. For the oil processing section, the technological processes of edible peanut and rapeseed oil could be divided into pretreatment, squeezing, filtration, and refining. The oil extraction rate of peanut and rapeseed in China stabilized at around 35% and 34% in spatial–temporal dimensions, respectively, which came from the China National Grain and Oil Information Center. Therefore, it requires 2.86 t of peanut raw material and 2.94 t of rapeseed to produce 1 t of peanut and rapeseed oil, respectively.

CF and NF of the crop planting system

Firstly, the accounting framework of carbon emissions and Nr loss from agricultural inputs was consistent with IPCC guidelines (IPCC, 2019), multiplying the activity data with corresponding emission factors as shown in the following equations: (1)${\displaystyle C{F}_{input}=\sum \left(A{D}_i\times E{F}_{C_i}\right)}$ (2)${\displaystyle N{F}_{input}=\sum \left(A{D}_i\times E{F}_{Nr\text{\_}i}\right)}$ where CF_input (kg CO₂eq ha⁻¹) and NF_input(kg Nr ha⁻¹) are the carbon (C) emissions and Nr loss of agricultural inputs, respectively; AD_i is the activity data of agricultural input i (kg ha⁻¹ for fertilizer, pesticides, agri-film, and seeds, L ha⁻¹ for diesel fuel, kWh ha⁻¹ for irrigation electricity); EF_{C_i} is the carbon emission factor of the corresponding agricultural input i (Table S2), and the EF_C of irrigation electricity consumption is local and distinct in different regions of China (Table S3); EF_{Nr_i} is the Nr emission factors of the corresponding agricultural input i (Table S4).

Secondly, the direct and indirect N₂O emissions associated with in-field N fertilizer application were calculated by the following equations: (3)${\displaystyle C{F}_{direct}=\sum \left(F_n\times E{F}_1\times \left(44/28\right)\times 273\right)}$ (4)${\displaystyle C{F}_{indirect}=\sum \left(F_n\times 11\text{\%}\times E{F}_2\times \left(44/28\right)\times 273\right)+\sum \left(F_n\times 24\text{\%}\times E{F}_3\times \left(44/28\right)\times 273\right)}$ where CF_direct is the direct N₂O emissions (kg CO₂eq ha⁻¹); CF_indirect is the indirect N₂O emissions (kg CO₂eq ha⁻¹); F_n is the application rate of N fertilizers (kg ha⁻¹); 44/28 is the mass conversion factor of N₂O-N to N₂O; 273 is the global warming potential of N₂O in a 100-year horizon (IPCC, 2021); 11% represents the fraction of atmospheric deposition of N volatilized from N fertilizer (IPCC, 2019); 24% represents the fraction of leaching and runoff from N fertilizer (IPCC, 2019); EF₁ represents the direct N₂O emission factor from N inputs in different regions of China (Table S5); EF₂ and EF₃ represent the indirect N₂O emission factor caused by N deposition (1%), and N leaching and runoff (0.75%), respectively (IPCC, 2019). These emission factors and fractions follow IPCC guidelines, which are widely used for large-scale estimates of farmland N₂O emissions.

Simultaneously, the different forms of direct Nr loss in the field were calculated using the following equation: (5)${\displaystyle N{F}_{direct}=\sum \left(F_n\times E{F}_j\right)}$ where, NF_direct is the indirect Nr loss in the field (kg Nr ha⁻¹); F_n is the application rate of N fertilizers (kg ha⁻¹); EF_j is the emission factor of the Nr form j (Table S6).

Finally, the CF and NF of the cropping system were calculated by the following equations: (6)${\displaystyle CFa=\left(C{F}_{input}+C{F}_{direct}+C{F}_{indirect}\right)/1000}$ (7)${\displaystyle NFa=N{F}_{input}+N{F}_{direct}}$

where, CFa (t CO₂eq ha⁻¹) and NFa (kg Nr ha⁻¹) are CF and NF per unit area, respectively; 1,000 is the unit conversion factor.

The CF and NF by different functional units were calculated by the following equations: (8)${\displaystyle CFt=\left(CFa\times A\right)/1000000}$ (9)${\displaystyle CFy=CFa/Y\times 1000}$ (10)${\displaystyle NFt=\left(NFa\times A\right)/1000000}$ (11)${\displaystyle NFy=NFa/Y\times 1000}$ where, CFt (Tg CO₂eq) and NFt (Gg Nr) are total CF and NF, respectively; carbon footprint per unit yield (CFy) (kg CO₂eq kg⁻¹) and nitrogen footprint per unit yield (NFy) (g Nr kg⁻¹) are CF and NF per unit yield, respectively; A is the planting area (ha); Y is crop yield per unit area (kg ha⁻¹); 100,0000 and 1,000 are both unit conversion factors.

CF and NF of raw materials and oil processing

This study identified 1 t of edible peanut and rapeseed oil as the functional unit to compare the differences in CF and NF between peanut and rapeseed oil.

In the supply of raw materials, the CF and NF of peanut oil and rapeseed oil were estimated using the following equations: (12)${\displaystyle C{F}_{raw\text{\_}mat}=CFy\times \sigma \times 1000}$ (13)${\displaystyle N{F}_{raw\text{\_}mat}=NFy\times \sigma }$ where, CF_{raw_mat} (kg CO₂eq t⁻¹) and NF_{raw_mat} (kg Nr t⁻¹) are the CF and NF per unit of edible oil, respectively; σ (t) is the raw materials required to produce 1 t of edible oil, estimated as 2.86 for peanut and 2.94 for rapeseed, respectively; 1,000 is the unit conversion factor.

Considering that the main driving force of oil processing is electric energy, comprehensive energy consumption is converted into electricity to estimate the CF and NF during the processing stage of peanut oil and rapeseed oil. The calculation formulas were established as follows: (14)${\displaystyle C{F}_{oil\text{\_}pro}=P\times 8.167\times 1.23}$ (15)${\displaystyle N{F}_{oil\text{\_}pro}=P\times 8.167\times 0.00329}$ where, CF_{oil_pro} (kg CO₂eq t⁻¹) and NF_{oil_pro} (kg Nr t⁻¹) are the CF and NF per unit of edible oil, respectively; P (kgce t⁻¹) is the comprehensive energy consumption per unit edible oil; 8.167 is the conversion factor of energy consumption to electric power (kwh); 1.23 is the GHG emission factor of electric power (kg CO₂eq kWh⁻¹) (Huang et al., 2017); 0.00329 is the Nr emission factor of electric power (kg Nr kWh⁻¹) (Chen et al., 2019).

Therefore, the total CF and NF of the edible oil product were calculated using the following equations: (16)${\displaystyle CF=C{F}_{raw\text{\_}mat}+C{F}_{oil\text{\_}pro}}$ (17)${\displaystyle NF=N{F}_{raw\text{\_}mat}+N{F}_{oil\text{\_}pro}}$ where, CF (kg CO₂eq t⁻¹) and NF (kg Nr t⁻¹) are the total CF and NF per unit of edible oil, respectively.

Data processing and visualization

Data processing and visualization were performed using Microsoft Office Excel 2019 (Microsoft, Redmond, WA, USA) and Origin 2021. To explore the annual changing trends in CF and NF of the crop planting phase, the slope of linear regression at a P-value less than 0.05 for 2004–2023 was conducted with SPSS 22.0 (IBM Corp., Armonk, NY, USA). Therein, linear regression analyses for the total carbon footprint (CFt) and total nitrogen footprint (NFt) and nitrogen footprint per unit area (NFa) of rapeseed and NFy of peanut are not displayed because the P-value for the linear regression slope is greater than 0.05.

Temporal variation in CF and NF by functional unit

The CF and NF presented similar temporal change trends to some extent for both peanut and rapeseed. From 2004 to 2023, the CFt (Fig. 3A), CFy (Fig. 3C), NFt (Fig. 3D), and NFy (Fig. 3F) of peanut were consistently lower than those of rapeseed. Peanut and rapeseed had the lowest CFt and NFt in 2007, with 7.44 and 10.11 Tg CO₂eq of CFt, and 110.66 and 165.57 Gg Nr of NFt, respectively. For peanut, the CFt, carbon footprint per unit area (CFa) (Fig. 3B), NFt, and NFa (Fig. 3E) exhibited a significant increasing trend with an average annual growth value of 0.079 Tg CO₂eq, 0.010 t CO₂eq ha⁻¹, 2.150 Gg Nr, and 0.354 kg Nr ha⁻¹ during the research period, respectively. Meanwhile, the CFy of peanut showed a significant declining trend of 0.005 kg CO₂eq kg⁻¹ yr⁻¹ on average. For rapeseed, CFa, CFy, and NFy displayed a significant downward trend, with an average annual declining value of 0.010 t CO₂eq ha⁻¹, 0.015 kg CO₂eq kg⁻¹, and 0.185 g Nr kg⁻¹, respectively.

Spatial variation in CF and NF by functional unit

The CF and NF were spatially heterogeneous across different regions in China (Fig. 4). Generally, regions with high CF also tended to have high NF discharge. For peanut, higher CFa (Fig. 4A) and NFa (Fig. 4B) mainly occurred in Hebei, Anhui, Jiangxi, Shandong, and Henan provinces, respectively with 2.12–2.56 t CO₂eq ha⁻¹ and 18.12–43.18 kg Nr ha⁻¹. There were larger peanut CFy (Fig. 4C) and NFy (Fig. 4D) apparent in Jiangxi and Sichuan, respectively with 0.62–0.74 kg CO₂eq kg⁻¹ and 10.01–12.35 g Nr kg⁻¹. Jiangxi had higher CF and NF than the other provinces, largely attributable to lower peanut productivity or larger agricultural inputs. For rapeseed, higher CFa and NFa were mostly found in low-production provinces, including Jiangsu, Zhejiang, Anhui, Yunnan, Shaanxi, and Gansu, where they respectively ranged from 1.84 to 2.67 t CO₂eq ha⁻¹ and 35.45 to 41.72 kg Nr ha⁻¹. Higher rapeseed CFy and NFy were mainly distributed in Jiangxi, Hunan, Guizhou, Yunnan, Shaanxi, and Gansu, respectively with 0.85–1.13 kg CO₂eq kg⁻¹ and 16.03–17.80 g Nr kg⁻¹. In other words, rapeseed production in Yunnan, Shaanxi, and Gansu produced higher GHG emissions and also had higher Nr discharges.

Contribution analysis of CF and NF for peanut and rapeseed

The most significant contributors to the CF and NF of peanut and rapeseed were fertilizer production and application. For the CF, N fertilizer and compound fertilizer emerged as the primary sources, respectively contributing 35.0% and 28.7% for peanut and 74.4% and 17.0% for rapeseed (Fig. 5A). The rapeseed CFy from N fertilizer was 3.5 times that of peanut (678.29 vs. 193.01 g CO₂eq kg⁻¹) (Fig. 5B). In addition, agricultural film and seed also played significant roles in the CF of peanut, while they contributed relatively minor to the CF of rapeseed. The peanut CFy from agricultural film and seed were 97.4 and 23.1 times those of rapeseed, respectively. Moreover, N runoff (30.4% for peanut and 34.0% for rapeseed) and NH₃ volatilization (39.0% for peanut and 33.9% for rapeseed) dominated the NF, followed by N leaching, NO emissions, and N₂O emissions (Fig. 5C). The rapeseed Nfy from N runoff and NH₃ volatilization were 1.8 and 1.4 times those of peanut, respectively (Fig. 5D).

CF and NF of peanut oil and rapeseed oil

The CF of peanut and rapeseed oil, on average, were 3,312.2 and 3,722.4 kg CO₂eq t⁻¹ oil, respectively (Fig. 6). The corresponding NF were 28.5 and 43.4 kg Nr t⁻¹ oil for peanut and rapeseed oil, respectively. The results illustrated that the CF and NF of peanut oil were lower by 11.0% and 34.2% than those of rapeseed oil, respectively. For the CF, raw materials contributed less than oil processing in peanut oil (43.9% vs. 56.1%), whereas the opposite occurred in rapeseed oil (61.4% vs. 38.6%). For the NF, almost all Nr losses were derived from raw materials, which comprised 82.6% and 91.1% in peanut and rapeseed oil, respectively. While the contributions of oil processing to the NF of peanut and rapeseed oil were both very low. Moreover, the CF and NF from raw materials in rapeseed oil were 1.6 and 1.7 times that of peanut oil, respectively. However, the CF and NF from oil processing in peanut oil were 1.3 times that of rapeseed oil.