Analysis of water–energy–carbon coupling and influencing factors in food production

Agriculture, as the foundation of social and economic development, is not only a crucial barrier to the natural environment but also a significant guarantee of national food security, and it has always been the focus of attention (Liu et al. 2022). Agricultural production mainly relies on resource inputs such as water and energy and is also a major source of global anthropogenic greenhouse gas emissions (Abbas et al. 2022a; Elahi et al. 2024). Agriculture uses 70% of the world's fresh water (Feng et al. 2023a), consumes 30% of energy (Wang et al. 2022), and contributes 14% of greenhouse gas emissions (Yu et al. 2022b). In recent decades, population growth, climate change, water scarcity, and energy crises have emerged as significant global challenges (Su et al. 2018), imposing higher demands on agriculture. To meet the needs of the expanding population, agricultural production is projected to increase by around 70% by 2050 (Saray et al. 2022). Nevertheless, the continuous expansion of the agricultural scale may negatively impact on sustainable agricultural development and the attainment of carbon neutrality goals (He et al. 2022). Water, energy, and carbon are vital elements of agricultural production, constituting the foundation of regional agricultural production systems through processes such as water cycling, energy consumption, carbon cycling, and agricultural product trade. The nexus among these components is intricate. Agricultural production necessitates the utilization of water resources and energy, resulting in carbon emissions (CEs) and water pollution due to inputs like fertilizers and pesticides (Xu et al. 2020). Water pollution can suppress crop growth and CEs can exacerbate climate change. This leads to higher temperatures and unfavourable weather conditions, all of which can have a negative impact on yields (Khangar & Thangavel 2024). In addition, unreasonable energy use can contaminate soil, which can have a direct impact on crop growth and, in turn, reduce yields. However, imposing constraints on water and fertilizer usage can lower water consumption and CEs, but it may also diminish crop yields and adversely affect agricultural security (Tian et al. 2018). Moreover, these factors are closely linked to the United Nations Sustainable Development Goals (SDGs), encompassing SDG2, SDG6, SDG7, and SDG13. Currently, agriculture, water, energy, and carbon are associated with different sectors, and a single-sectoral management of resources persists. The water–energy–carbon (WEC) nexus of food production provides a systematic multisectoral perspective to promote agricultural sustainable development and work towards coordinated management of natural resources across sectors. In addition, since agriculture not only has a growing demand for water and energy but also a major source of CE (Yoon et al. 2022), it has been considered as the important influencing factor to climate change risks, such as global warming, water scarcity, and energy shortages (Zhu et al. 2023). Conversely, climate change can have significant impact on agricultural systems (Yu et al. 2022b). Climate alterations are poised to remarkably affect both temperature fluctuations and precipitation patterns, thereby further altering the impact of irrigation on agricultural water, energy, and carbon (Zhu et al. 2023). Meanwhile, water and energy resources are highly vulnerable to climate change (Yoon et al. 2022), which can adversely affect agriculture. Therefore, efficient resource management can contribute to climate change mitigation and adaptation (Yoon et al. 2022). Hence, investigating the WEC nexus in agricultural production is crucial for agricultural security and integrated management. It can also establish a theoretical foundation for the efficient utilization of agricultural water resources and energy, as well as for reducing CEs and addressing climate change.

The ‘footprint theory’ is a crucial concept in the study of the WEC nexus in agricultural production. The agricultural water footprint (WF) refers to the amount of freshwater consumed during crop growth. It includes the blue, green, and grey WF. The blue and green WF refers to the consumption of surface water, groundwater, and precipitation during crop growth. The grey WF refers to the amount of freshwater required to absorb pollutants under given natural background concentrations and existing environmental water quality standards (Cai et al. 2022). The agricultural energy footprint represents the energy consumption from planting to harvesting during the crop's lifecycle (Zhai et al. 2021). The agricultural carbon footprint encompasses direct and indirect carbon dioxide emissions during crop growth (Fan et al. 2022). Therefore, the study of the WEC nexus in agricultural production is essential for identifying pressures from water, energy, and CEs, which would provide theoretical support for water and energy conservation as well as emission reduction.

Currently, it is constrained for the comprehensive analysis of the WEC nexus in agricultural production, and mostly focuses on one or two resource elements. However, comprehensive consideration and analysis of all three elements are found only in a few reports (Feng et al. 2023a; Miao et al. 2023). Yousefi et al. (2017) studied the WEC nexus in sunflower production in Kerman Province, Iran. Xu et al. (2020) analysed the impact of irrigation agriculture on the food–energy–water–CO₂ nexus in the North China Plain. Yu et al. (2022b) evaluated the impact of agricultural activities on the energy–carbon–water nexus in the Qinghai-Tibet Plateau. Feng et al. (2023a) quantified the WEC emission nexus in crop production in the Tarim River Basin. However, these reports rarely considered the quantification of the grey WF in the WF, and the indicator selection of coupling coordination degree did not consider energy output (EO) and carbon sink (CS) data, among other factors. However, Cao et al. (2021) pointed out the importance of the grey WF, and Miao et al. (2023) stated that the CS of the agricultural system should not be ignored. Therefore, it is more comprehensive to incorporate both the grey WF and the CS into the WEC nexus comprehension in this study.

As an agricultural and populous country, China has always considered agriculture as its pillar industry. Food, as the core of agriculture, has been a prominent research topic. From 2011 to 2020, the planting area of food crops in China increased from 1.13 × 10⁸ to 1.17 × 10⁸ ha, a growth rate of 3.54%. Food production also increased from 5.88 × 10⁸ to 6.69 × 10⁸ t, a growth rate of 13.78% (National Bureau of Statistics 2021). This significant increase has contributed to the improvement of food supply and reduction of hunger. As a major agricultural province, Sichuan Province accounts for only 4.8% of the country's arable land but meets the grain needs of 6.7% of the population, making an outstanding contribution to China's food security. However, there was no report on the WEC nexus in food production in Sichuan Province. Previous reports on agriculture in Sichuan Province have focused on the following aspects. Liu et al. (2015) quantified the agricultural WF of Sichuan Province and analysed its influencing factors. Xu (2011) analysed the energy consumption of rice, wheat, and maize and their efficiency. Yang et al. (2023) quantified agricultural CEs and its efficiency in Sichuan Province. Huang (2022) analysed the efficiency of water and energy consumption for food. Yang (2022) accounted for the WF and carbon footprints of food production. Zhen (2017) quantified the energy consumption and CEs of crop production. In addition, agricultural production in Sichuan faces many challenges. With economic development, agriculture and industry are increasingly competing for land, water, and energy. This adversely affects agricultural development in a context of limited resources. Although the use of pesticides and fertilizers in Sichuan is gradually decreasing, the total amount is still large. The environmentally friendly pesticides that are high-efficiency, low-toxicity, and low-residual are not commonly used, which results in serious damage to agricultural water and soil. Furthermore, Sichuan is plagued by natural disasters, such as droughts and floods, which have had severe impacts on food production. Hence, it is imperative to study the WEC nexus in food production in Sichuan Province. The research results can provide a theoretical basis for mitigating the impact of agriculture on the environment and achieving low-carbon and sustainable agricultural development.

The study area, Sichuan Province (26°03′–34°19′N, 97°21′–108°33′E), is located in the southwestern region of China and is situated in the upper reaches of the Yangtze River. The province enjoys abundant precipitation, with an average annual rainfall of over 1,000 mm. As one of the 13 major food-producing areas in China, Sichuan Province plays a crucial role in ensuring stable food production for national food security. In 2020, the planting area of food crops in Sichuan Province was 6.31 × 10⁶ha, with food production of 3.53 × 10⁷t, ranking at the forefront nationwide (Rural Social and Economic Survey Department of the National Bureau of Statistics 2021) and steadily improving its capacity to ensure food security. The per capita food possession was 421.9 kg, ranking 16th nationwide, which was lower than the national average of 474.4 kg (Rural Social and Economic Survey Department of the National Bureau of Statistics 2021), highlighting the ongoing significance of food security in Sichuan Province. Agricultural water usage amounted to 1.54 × 10¹⁰ m³, representing 64.96% of the total water consumption. The utilization coefficient of irrigation water was only 0.484, which is below the national average of 0.565 (Ministry of Water Resources of the People Republic of China 2021). Additionally, the province utilized 2.11 × 10⁶ t of fertilizers, 4.71 × 10⁵ t of diesel, and 4.21 × 10³ t of pesticides (Rural Social and Economic Survey Department of the National Bureau of Statistics 2021), resulting in substantial agricultural energy consumption. The total agricultural CE reached 5.53 × 10⁷ t, accounting for 5.89% of the national total emission (Yin et al. 2023). These numbers underscore the formidable task facing Sichuan Province in improving water and EUE, as well as reducing CEs in agriculture.

The required data include the planting area and yield of food crops, the cost and income per unit area of food crops, the input of agricultural materials per unit area of food crops, and meteorological, population, economic, science, and technology data. The planting area yield of food crops and economic data are from the Sichuan Statistical Yearbook. The cost and income and the input of agricultural materials per unit area of food crops are from the National Agricultural Cost and Income Data Compilation and the China Price Yearbook. Meteorological data are from the National Meteorological Science Data Sharing Service Platform (http://data.cma.cn/). Population data are obtained from the China Population and Employment Statistics Yearbook. Science and technology data are from the China Statistical Yearbook of Science and Technology.

The calculation methods for food green WF, grey WF, EI–EO, and CE–CS are similar to those for blue WF and will not be elaborated on one by one.

The corresponding parameters and sources are listed in Table S1.

CE_i are divided into three parts, CO₂ emission from agricultural inputs, CH₄ emission from rice cultivation, and N₂O emission from nitrogen fertilizer application.

The parameters related with CE are listed in Table S2.

The coupling degree describes the interactions among the systems and helps to understand how closely the systems are related. The coupling coordination degree reflects the closeness of the influences and links among systems. Applying them to analyse the WEC nexus in food production is not only able to quantify the interrelationships among WEC but is also significant for the integrated and coordinated management of food production.

Based on the previous research findings (Yu et al. 2022b; Feng et al. 2023a) and combined with our own study, we selected 10 indicators to construct the coupling degree and the coupling coordination degree model, including WF, EI and EO, as well as CE and CS for per unit of area and per unit of value, respectively. The comprehensive evaluation index system of the WEC nexus is listed in Table S4.

This study employs the entropy method to determine the weights of each indicator. The main steps (Feng et al. 2023a) are as follows:

Referred to Guo et al. (2023), the coupling degree and coupling coordination degree are classified in Table S5.

Path analysis can be used to examine the direct and indirect significance of the independent variable on the dependent variable by decomposing the apparent direct correlation between them, thereby separating the correlation coefficients into a direct path coefficient and an indirect path coefficient (Xu et al. 2022). In order to analyse the explanatory power and interaction of different influencing factors on the coupling degree and the coupling coordination degree of the WEC system of food production, nine influencing factors were selected. This selection considered the scientific and comprehensive nature of the indicators, the issue of multicollinearity, and the availability and accessibility of data, building upon previous research results.

The selection and interpretation of the nine influencing factors are shown in Table S6.

Precipitation is critical for food production, and enough precipitation can increase the green WF and reduce the electricity consumption for irrigation, which, in turn, cuts down the EIs and CEs. The high average temperature may increase the consumption of irrigation water, thus increasing EIs and CEs. High yields of food per unit area have a positive impact on EOs and CSs. High agricultural planting structure represents more area sown to food, which consumes more water and energy and leads to more CEs; accordingly, it generates more EO and CSs. The agricultural industrial structure indicates the ratio of agriculture to agriculture, forestry, and livestock and fisheries, and in general, agriculture consumes more water and energy and contributes more to CEs. High rural ageing is likely to lead to a reduction in the labour force of farmers, as well as a decline in the size and output of food, which may result in less resource consumption and EO. The lower the rural Engel's coefficient, the smaller the proportion spent on food. The farmers have more money to invest in agricultural production, such as enlarge food planting area, thus increasing the resource inputs and CEs. Agricultural technology innovation reduces the dependence on pesticides and fertilizers for food production and also decrease water consumption through water-saving irrigation technology. Agricultural economic level will arouse the enthusiasm of farmers and may increase the investment in food production, leading to high resource consumption and CEs. All of these have impacts on the coupling degree and coupling coordination degree by affecting the WEC nexus.

Based on Figure 2(a), a significant disparity in the WF_blue among different crops is evident. The average WF_blue of wheat was as high as 0.927 m³ kg⁻¹, far exceeding that of food at 0.286 m³ kg⁻¹, while the average for maize was only 0.113 m³ kg⁻¹. As per Figure 2(b), the average WF_green of beans was as high as 1.723 m³ kg⁻¹, which is significantly higher than that of food at 0.707 m³ kg⁻¹; meanwhile, wheat had the smallest average at 0.522 m³ kg⁻¹. Figure 2(c) shows that the average WF_grey of tubers was as high as 0.634 m³ kg⁻¹, surpassing that of food at 0.302 m³ kg⁻¹, while beans had an average value of 0.046 m³ kg⁻¹. Furthermore, apart from tubers, all other WF_grey showed a declining trend. Considering Figure 2(a), (b), and (c), the virtual water of rice (the sum of the WF_blue and WF_green) had an average value of 0.847 m³ kg⁻¹, which closely aligns with the result of 0.853 m³ kg⁻¹ reported by Yu et al. (2022a). The virtual water of wheat had an average value of 1.449 m³ kg⁻¹, which is consistent with the results of 1.376 m³ kg⁻¹ reported by Yu et al. (2022a). Similarly, the virtual water content of food had an average value of 0.993 m³/kg, which is close to the result of 1.000 m³ kg⁻¹ reported by Wang et al. (2014). The WF of food had an average value of 1.295 m³ kg⁻¹, demonstrating strong agreement with the result of 1.370 m³ kg⁻¹ reported by Zhang (2022). Furthermore, as shown in Figure 2(d), the blue WF of food exhibited significant fluctuations, with an overall decreasing trend of 18.04%, despite an increase in food yield. This suggests an enhancement in water use efficiency, in line with the rise in the effective utilization coefficient of irrigation water in farmland. The green WF of food showed an upward trend, increasing by 8.57%, which is possibly related to the increase in precipitation in Sichuan Province. The grey WF of food showed a downward trend, with a decrease of 6.58% due to the significant effects of the deep implementation of the ‘one control, two reductions, and three basics’ policy in Sichuan Province and a continuous 5-year negative growth in fertilizer use. The proportion of blue WF of food fluctuated with an overall decreasing trend, averaging 21.98% and decreasing by 15.60%. The proportion of green WF was higher than that of blue WF, which is consistent with the results of Hua et al. (2020), with an average value of 54.69%. The green WF is derived from precipitation, and it does not require engineering measures for transportation and access compared to the blue WF. Therefore, the opportunity cost of the green WF is relatively low and has little negative impact on the environment, so improving its share is conducive to promote the comprehensive utilization of regional water resources. Furthermore, except for the year 2011, the proportion of green WF of food was higher than 50% in other years, further indicating the relatively abundant precipitation resources in Sichuan.

Although Sichuan Province has relatively abundant precipitation resources, they are unevenly distributed throughout the year. As a major water consumer in agriculture, food production exerts great pressure on water resources and is also prone to seasonal water shortages. Therefore, optimizing the planting structure is a good option. The planting area of high-water-consuming crops, such as wheat and beans, can be appropriately reduced. This can be achieved by improving the natural precipitation utilization rate through the construction of terraces and reservoirs along the mountains. In addition, the grey WF can continue to decrease with reduction of chemical fertilizers, the utilization of organic fertilizers, and scientific fertilization. Additionally, there are some other effective measures to reduce water consumption, such as the implementation of a ‘tiered water pricing’ policy and the vigorous promotion of water-fertilizer integration technology (Li & Singh 2020).

Figure 3(a) illustrates a significant difference in the EI for different crops. The EI of tubers is 13,162 MJ hm⁻², which is approximately 1.3 times that of food. In contrast, the EI of beans accounts only for about 20% of that of food (averaging 2,029 MJ hm⁻²). Except for maize, the EI for all other crops showed a slightly increasing trend. Figure 3(b) reveals that the EO for rice was as high as 118,546 MJ hm⁻², which is in good agreement with the results of Bockari-Gevao et al. (2005) and far exceeds that of food at 7,2665 MJ hm⁻². The average for tubers was 13,420 MJ hm⁻², only about 18% of food. Additionally, the EO for all crops showed an increasing trend, closely related to the increase in yield. According to Figure 3(c), the EUE for beans had an average value of 20.88, approximately 3 times that of food at 7.20, and about 20 times that of tubers at 1.04. Although the multi-year average for tubers exceeded 1, half of the periods show an EUE lower than 1, indicating obvious unsustainability. Figure 3(d) indicates that the total EI for food increased from 5.75 × 10¹⁰ MJ in 2011 to 6.63 × 10¹⁰ MJ in 2020, representing a growth of 15.30%. This accounted for 1 and 1.07% of total annual energy consumption, respectively. The increasing EI has led to higher food production and intensified the energy pressure in the region. The average total EI for food was 6.32 × 10¹⁰ MJ. Fertilizers had the highest average, further confirming the conclusions of Feng et al. (2023a) and Abbas et al. (2020), while electricity had the lowest average, accounting for only 4.18% of the total EI. Among all energy types, diesel had revealed a significant increase of 106.45%, while fertilizers and seeds show declining trends. This is closely related to the promotion of fertilizer reduction and improvement of high-quality seed breeding systems in Sichuan Province.

As one of the major food-producing regions in China, Sichuan Province plays a crucial role in ensuring national food security. It is essential to achieve an increase in food production without a corresponding increase or even with a decrease in EI. Although the EI from fertilizers is declining, it still accounted for 43.92% of the total in 2020. Previous studies had shown that less than 50% of the fertilizers applied to farmland are utilized, with the rest seeping into the environment (Sun et al. 2019). Therefore, it is necessary to improve the utilization rate of fertilizers to reduce the amount of application. This can be achieved through enhanced fertilization training and the use of organic fertilizers as a substitute for chemical ones. Furthermore, there is a need to promote new energy sources and transition to clean and renewable energy, optimizing the energy consumption structure. This could make a valuable contribution to the achievement of sustainable energy development goals (Abbas et al. 2022b). Moreover, the conversion of the traditional irrigation method to modern irrigation techniques can reduce energy consumption (Abbas et al. 2020). Additionally, optimizing the planting structure is also crucial. Tubers have the highest average EI, lowest EO, and lowest EUE. Therefore, it may be appropriate to reduce the planting area while balancing the nexus between the government and farmers. According to the ‘National Agricultural Cost and Income Data Compilation’, tubers were reported to have the highest profit per unit area.

From Figure 4(a), it is evident that the CE of rice is significantly higher than that of other crops, which is related to the production of CH₄ during rice growth. The average CE of beans is only 171.41 kgCO₂-eq hm⁻², which is far lower than 2,901.39 kgCO₂-eq hm⁻² for other crops. According to Figure 4(b), the average CS of rice ranks the highest, which is attributed to the higher yield of rice per unit area. The average CS of tubers is the lowest, significantly lower than 16,050.82 kgCO₂-eq hm⁻² for food, which relates to the carbon sequestration coefficient. Figure 4(c) indicates that the SI of all crops was positive, consistent with the research results of Lal (2004). This indicates that all crops have a net carbon benefit, contributing to the positive effects on the green, low-carbon, and sustainable development of agriculture. Figure 4(d) shows that the total CE from food remains relatively stable overall, with a slightly increase (0.23%) during the study period. This is attributed to the implementation of agricultural emission reduction and carbon sequestration actions, the decrease in the area sown to rice, and the increase in the area sown to beans in Sichuan Province. Among all types of CE, CH₄ emission from rice cultivation, N₂O emission from nitrogen fertilizer application, and CO₂ emission from fertilizer input ranked in the top three, accounting for a high proportion of 90.27% of the total CE.

Agricultural production has the dual attributes of CE and CS. The research results indicate that food has a significant net carbon effect, but Wang et al. (2011) emphasized that without reducing the emission from agriculture and forestry, any emission reduction strategy is bound to fail. Considering the types of CE, it may be necessary to appropriately reduce the planting area of rice while ensuring food security. However, since rice is the largest food crop in Sichuan, it is possible to reduce its planting area while safeguarding food self-sufficiency. In addition, it is more important to reduce the use of fertilizers, especially nitrogen fertilizers, because greenhouse gases from fertilizers cause air pollution and harm soil and water (Abbas et al. 2022b). Previous studies had shown that small-scale farmers in China use a large amount of nitrogen fertilizer (Cui et al. 2018). Therefore, it is necessary to improve the fertilization pattern and promote soil testing and formula fertilization to reduce the amount of fertilizer used. Furthermore, conservation tillage (no-till and reduced tillage) and the development of intensive agriculture (Tian et al. 2021; Khangar & Thangavel 2024) can be adopted to reduce CE. In addition, adopting alternative tillage systems can reduce greenhouse gas emissions by 15–29% (Abbas et al. 2022b).

As shown in Figure 5(a), the C ranges between 0.4 and 1. The average C of food crops is greater than 0.8, indicating a strong correlation and closeness among the WEC of food crops in Sichuan Province, which is consistent with the results of Feng et al. (2023a). Certain Cs ranged between 0.5 and 0.8, indicating a running-in level of coupling. However, the C of wheat was below 0.5 in 2019, indicating a low level of coupling. Furthermore, different crops exhibited various trends in coupling degree. Wheat, maize, and tubers showed a decreasing trend, while rice and beans showed an increasing trend.

Figure 5(b) shows that the D ranges between 0.2 and 0.5. The D of rice is between 0.3 and 0.5, indicating mild to near incoordination. The D of wheat is between 0.2 and 0.4, indicating moderate to mild incoordination. The average value of 0.26 is lower than the findings of Zhu et al. (2023), possibly due to their exclusive focus on the WEC during the irrigation process. The D of maize, beans, and tubers all range between 0.2 and 0.5, indicating moderate to near incoordination.

The C of WEC in food production in Sichuan Province was at a high coupling stage, but the D had not reached a level of slight coordination, which is consistent with the reports of Yu et al. (2022b). As WEC is associated with different sectors, it is necessary to coordinate the natural resource management across sectors to address the lack of coordination and to balance the nexus among food production, resource conservation, and environmental protection. Feng et al. (2023a) emphasized that water resources are fundamental to the WEC nexus. For water resource systems, drip irrigation can reduce water and fertilizer consumption by 40–50% compared to traditional flood irrigation (Angold & Zharkov 2014). Thus, adopting drip irrigation technology can conserve blue water resources, reduce the grey WF, and decrease energy consumption and CE from fertilizers (Zhang et al. 2021a). For energy systems, reducing the use of fertilizers is essential. Despite a 16.09% decrease in fertilizer usage in Sichuan Province from 2011 to 2020, the energy consumption of fertilizers has remained high. Therefore, the utilization of organic fertilizers can be implemented, thereby enhancing the sustainability of agriculture (Yang et al. 2019). Second, reducing the use of diesel and promoting clean energy sources, such as biomass fuels, is important. This is because the extensive use of fossil fuels not only degrades the environment and depletes natural resources but also has a negative impact on climate change (Abbas et al. 2022b). Nevertheless, the biomass energy technology in Sichuan Province still lags behind that of developed areas domestically and internationally. For the carbon system, methane (CH₄) emission from rice cultivation, nitrous oxide (N₂O) emission from nitrogen fertilizer application, and carbon dioxide (CO₂) emission from fertilizer input contributed to around 90% of total CE. Given that rice is the largest food crop in Sichuan, accounting for approximately 30% of the total food cultivation area and over 40% of the total food production (Feng et al. 2023b), its planting area can be appropriately reduced. It is important to reduce the application of fertilizers (especially nitrogen fertilizers) and promote precision fertilization and soil improvement to enhance fertilizer efficiency. It is also important to promote direct straw returning and the application of organic and biological fertilizers with lower CE coefficients (Yu et al. 2022b). However, the straw returning rate in Sichuan Province is relatively low nationwide (Liu et al. 2018). Fortunately, the promotion area of organic fertilizers in Sichuan Province is growing rapidly, with an annual production of organic fertilizers reaching 2 × 10⁶ t.

The quantified effects of nine influencing factors on the coupling degree and the coupling coordination degree of WEC from 2011 to 2020 are shown in Tables 1 and 2, respectively.

Table 1 implies that the absolute values of the direct path coefficients that affect the coupling degree are X₄, X₃, X₇, X₅, X₈, X₆, X₂, X₁, and X₉ in proper order. The results indicate that agricultural planting structure, food yield per unit area, and rural Engel's coefficient have significant direct impacts on coupling degree. Comparatively, the agricultural economic level, precipitation, and average temperature have minimal effects on it. Through the analysis of indirect path coefficient among various factors, it can be known that each variable has great influence on the coupling degree through the agricultural planting structure, rural Engel's coefficient, and food yield per unit area. In terms of agricultural planting structure, the agricultural economic level has a strong positive effect on coupling, while rural Engel's coefficient has a strong negative effect. In terms of the rural Engel's coefficient, the agricultural planting structure has a strong positive effect, while agricultural technology innovation has a strong negative effect. In terms of yield per unit area, the agricultural planting structure has a strong positive effect, while the agricultural economic level has a strong negative effect. The three factors that have the greatest total impact on the coupling degree are average temperature, agricultural industry structure, and rural Engel's coefficient.

Table 2 presents that the absolute values of the direct path coefficients that affect the coupling coordination degree are X₃, X₉, X₈, X₄, X₇, X₅, X₆, X₁, and X₂. The results indicate that food yield per unit area, agricultural economic level, and agricultural technology innovation have significant direct impacts on the coupling coordination degree, while the average temperature, precipitation, and rural ageing have minimal effects on it. Through the analysis of indirect path coefficient among various factors, it can be known that each variable has a great influence on the coupling coordination degree through the agricultural planting structure, rural Engel's coefficient, and grain yield per unit area. In terms of agricultural planting structure, the agricultural economic level has a strong positive effect on the coupling coordination degree, while the rural Engel's coefficient has a strong negative effect. In terms of the rural Engel's coefficient, the agricultural planting structure has a strong positive effect, while agricultural technology innovation has a strong negative effect. In terms of yield per unit area, the agricultural planting structure has a strong positive effect, while the agricultural economic level has a strong negative effect. This has similar results to the coupling degree. The three factors that have the greatest total impact on the coupling coordination degree are agricultural planting structure, average temperature, and agricultural economic level.

In summary, the influencing factors show different effects on the coupling degree and the coupling coordination degree. Except for average temperature, which has an inhibitory effect on both total effects, the rest show opposite effects on both. Lowering the average temperature within a moderate range has a promoting effect on both, but this method cannot be achieved under open-air planting conditions. For food production, improving the coupling coordination degree among WEC is particularly important. First, it is possible to reduce the numerical value of the agricultural planting structure, that is, to reduce the area of food planting, but this would have a negative impact on food security, so it is possible to appropriately reduce the area under its planting while ensuring food security. Second, lowering the average temperature over a moderate range, as described above, is almost impossible to achieve; third, the agricultural economic level can be improved, which can be achieved either by increasing the value of agricultural production or by increasing the rate of urbanization in order to reduce the rural population.

As a major agricultural province, Sichuan has made great contributions to the realization of basic self-sufficiency in cereals and absolute safety in food rations in China. However, food production is an intensive industry. It requires large inputs of water and energy and also leads to vast CEs, hindering the green and sustainable development of agriculture in Sichuan Province. In recent years, the Sichuan Provincial Government has taken a number of proactive measures to reduce the consumption of water and the environmental impact of food production, such as strictly controlling the total amount of water used in agriculture, reducing the use of fertilizers and pesticides, and actively promoting highly efficient water-saving irrigation technologies. These measurements have made lots of achievements, but there is still a long way to go. Therefore, this paper comprehensively analyses the WEC nexus in food production, quantifies the degree of interactions and closeness of the WEC nexus, and analyses the factors affecting the coupling degree and the coupling coordination degree. The results of the study show that green WF contributes the most to the WF. Fertilizer and diesel are the most important EIs for food production. Fertilizer has the highest CEs, except for rice cultivation. There is a strong coupling degree while a weak coupling coordination degree of the WEC nexus. The path analysis shows that the rural Engel's coefficient and average temperature are the largest contributing and inhibiting factors for the coupling degree, while the agricultural economic level and agricultural planting structure are the largest influencing factors for the coupling coordination degree.

Based on the existing findings, the following recommendations are provided. First, planting structure optimization is carried out. The planting area with high resource consumption and high CEs crops should be reduced. Second, it is better to actively promote the utilization of organic fertilizers and efficient integrated water-fertilizer irrigation techniques. Third, the development of the agricultural economy and the promotion of urbanization will increase the agricultural economic level, thus improving the coupling coordination degree among the WEC nexus.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.