1 Introduction

Global warming (or climate change) is a burning issue in the twenty-first century due to the atmospheric accumulation of greenhouse gases (GHGs) dominated by CO2, which is continuing to increase due to the burning of fossil fuels and deforestation (Raihan et al. 2021a; Sultana et al. 2023a; Raihan and Tuspekova 2022a; Raihan et al. 2023a; Shikha et al. 2023). The continuous growth of CO2 in the atmosphere poses a significant risk to human health and the environment (Raihan et al. 2021b; Akter et al. 2023; Isfat and Raihan 2022; Voumik et al. 2022; Sultana et al. 2023b; Sahabuddin et al. 2023). Therefore, the mitigation of the adverse impacts of climate change is commonly acknowledged as a primary focus within national and international development endeavors (Raihan and Bijoy 2023; Raihan and Tuspekova 2022b; Gverdtsiteli 2023). Throughout the last few decades, the effect that rapid economic growth has on the environment has been one of the most discussed subjects in all media (Raihan et al. 2018; Begum et al. 2020; Raihan 2023a; Azam et al. 2023). Over the past 50 years, there has been a significant increase in the consumption of fossil fuels (coal, oil, and gas), with growth of over eight times since 1950 and doubling from 1980 to 2022 (Addai et al. 2023). An environmental problem has surfaced as a consequence of the accelerated consumption of natural resources and the unchecked growth of economies around the world (Raihan and Tuspekova 2022c; Voumik et al. 2023a; Raihan 2023b). Fossil fuels constitute the predominant source of global climate change, responsible for more than three-quarters of global greenhouse gas emissions and close to 90% of CO2 emissions (Voumik et al. 2023b; Raihan 2023c; Kodba et al. 2023). China's economy heavily relies on fossil fuels, accounting for almost 87 percent of its overall energy consumption. Coal has a crucial role in sustaining the Chinese economy by providing essential support to heavy industrial industries, power generation, and the construction industry. The coal consumption of the country had a significant increase from 446 million tons in 1990 to 2.8 billion tons in 2017 (Zaghdoudi et al. 2023). Consequently, China is responsible for approximately 27% of the total global CO2 emissions and approximately one-third of the world's GHG emissions (Teng and Han 2022).

The agriculture, forestry, and other land use (AFOLU) sector is accountable for one-fifth of total anthropogenic GHG emissions (Raihan et al. 2023b; IPCC 2022). Agriculture, apart from its association with GHG emissions and its vulnerability to the impacts of climate change, plays a prominent role in the degradation of the natural environment (Raihan and Tuspekova 2022d; Raihan et al. 2023c; Kirwan et al. 2023). The raising of livestock, the growing of rice, enteric fermentation, and inefficient management of fertilizer are all factors that contribute to GHG emissions from agriculture (Raihan and Himu 2023; Balasundram et al. 2023). The leading causes of CO2 emissions in the agricultural sector are attributed to deforestation resulting from the extension of agricultural land, soil degradation, burning of crop residues, and the utilization of fossil fuel energy for agricultural operations (Jaafar et al. 2020; Vyas and Jhala 2023). In addition to other sources, agricultural operations contribute to the emission of CO2 through several activities, including plowing, planting, and the use of chemical treatments (Alhashim et al. 2021). Therefore, sustainable farming practices that conserve resources such as water and fertilizer inputs while improving soil organic carbon and reducing GHG emissions have to be adopted in response to soil fertility loss and climate change mitigation (Rahman et al. 2021). In this regard, a number of organizations all over the world introduced the climate-smart agriculture (CSA) approach. The core aims of CSA encompass three basic objectives: firstly, to augment production and household income; secondly, to fortify adaptability and resilience; and thirdly, to mitigate GHG emissions originating from agricultural activities (Rahman et al. 2021). Figure 1 illustrates the key elements of CSA.

Fig. 1
figure 1

The key elements of climate-smart agriculture (CSA)

Full size image

Furthermore, agriculture is widely recognized as a significant catalyst for promoting sustainable development in emerging nations (Kumar et al. 2022). Agriculture plays a crucial role in the pursuit of eliminating extreme poverty, as outlined by the sustainable development goals (SDGs). Hence, the establishment of a sustainable production, consumption, and distribution chain is imperative in order to enhance agricultural output, guarantee food supply security, and improve nutritional outcomes. To effectively achieve the SDGs by 2030, a thorough reorganization of the world's food and agriculture system that takes climate change into account must be put into place (Pata 2021). In addition, raising agricultural productivity is necessary to ensure food security and to support the country's economy by increasing the value added by agriculture (Bjornlund et al. 2022). The agricultural sector plays a pivotal role in bolstering economies through the provision of raw resources, food supply, stimulation of competition, and generation of employment opportunities (Christiaensen et al. 2021). Agricultural activities, including fishing, forestry, and hunting, are of significant importance in generating income, particularly in developing nations such as China. When it comes to the production of agricultural goods, China is currently in first place worldwide. Despite possessing just 10% of the global arable land, China exhibits a remarkable capacity for generating 25% of the global grain supply (Liu et al. 2023). In the agriculture, forestry, and fishing sector, value added in China was 1.31 trillion US$ in 2022 which was approximately 7.3% of the country’s GDP (World Bank 2023). Growing economies require an increase in agricultural output, which in turn drives up demand for environmentally friendly goods and services and the ability of the government to implement rules along these lines (Alam et al 2023). Through the utilization of rapid economic expansion, supported by increased agricultural value added, nations have the opportunity to adopt environmentally friendly technologies for agricultural production (Wang et al. 2020). This, in turn, has the potential to contribute to a reduction in national emission levels (Raihan et al. 2022a). The Chinese government has recently developed a policy known as "double carbon" with the aim of achieving carbon neutrality by the year 2060. This initiative raises important questions regarding sustainable agriculture in alignment with the carbon reduction strategy (Liu et al. 2022). Therefore, research on the relationship between agricultural productivity and environmental pollution is required if China is to successfully adopt a climate-smart agriculture policy.

However, the researcher has a predicament when considering the environmental impact of agriculture, as the annual consumption of fossil fuels by this sector results in the release of billions of tons of GHGs into the atmosphere. This, in turn, contributes to global warming and climate change (Raihan et al. 2019; Wijerathna-Yapa and Pathirana 2022). The utilization of outdated technologies and reliance on fossil fuel-based inputs in agricultural practices can lead to adverse environmental consequences and hinder the attainment of the SDGs (Raihan et al. 2022b; Pata 2021). Numerous studies have been conducted to examine the nexus between agricultural value added and environmental deterioration in various nations by utilizing CO2 emissions as a proxy of environmental degradation. Several studies have indicated that agricultural value added has a positive impact on environmental quality by reducing CO2 emissions (Raihan et al. 2023d; Waheed et al. 2018; Ullah et al. 2018; Anwar et al. 2019; Raihan et al. 2022c; Wang et al. 2020; Prastiyo et al. 2020; Koshta et al. 2021; Raihan and Tuspekova 2022e). However, contrasting findings suggest that agricultural productivity, as measured by agricultural value added, might lead to increased environmental degradation through the upsurge of CO2 emissions (Yurtkuran 2021; Usman et al. 2022). Hence, the empirical evidence regarding the impact of agricultural value added on environmental conditions remains equivocal. Due to its economic significance, the dynamic relationship between agriculture and the environment, and the distinct energy consumption patterns observed in the agricultural sector, the investigation of agriculture within the Environmental Kuznets Curve (EKC) paradigm holds considerable importance.

The environmental Kuznets curve (EKC) theory provides a framework for comprehending the correlation between increasing pollution levels and developing economies. Based on the postulates of the EKC, a theoretical framework proposed by Kuznets (1955), there exists a curvilinear relationship between the pace of economic growth and the level of pollution, characterized by an inverted U-shape. The EKC demonstrates a positive relationship between economic growth and environmental degradation during the initial phases of development. However, this correlation undergoes a reversal after certain income threshold is attained. According to Lee et al. (2023), there is a positive correlation between individuals' discretionary income and their inclination toward environmental cleanliness. Within the realm of EKC studies, CO2 emissions are occasionally employed as a proxy for the manifestation of environmental degradation or pollution. In contemporary literature, several studies have examined the correlation between economic growth and carbon emissions in order to confirm the EKC theory. For example, Sultana et al. (2022) provided empirical evidence supporting the presence of the EKC phenomenon in Bangladesh, both in the short-term and long-term. The existence of an inverted U-shaped EKC in the Malaysian economy was proven by Suki et al. (2020), indicating that environmental pollution increases with rapid economic growth but after a certain level of economic growth, people become more aware of the environment and the government would increase financing on improving environmental quality. Furthermore, Ahmad et al. (2021) discovered an inverted U-shaped EKC within the context of Pakistan. Chang et al. (2021) revealed the presence of an inverted U-shaped EKC in China. Genç et al. (2022) established the validity of the EKC in the context of Turkey. Cheikh et al. (2021) discovered an inverted U-shaped pattern across the Middle East and North Africa (MENA) region. Verbič et al. (2021) revealed the presence of an inverted U-shaped EKC across Europe. According to the findings of Pata et al. (2023), the EKC hypothesis has been confirmed to hold true across the countries of the Association of Southeast Asian Nations (ASEAN). However, there has been limited study conducted on the analysis of agricultural value added within an EKC framework. There is a research gap investing the influence of agricultural value added in the EKC framework, particularly in the context of China. Consequently, the present study aims to fill this research gap in the existing literature. The main objective of this research is to examine the impact of agricultural value added, economic growth, and energy use on CO2 emissions in China, specifically within the framework of the EKC. The investigation was expected to discover a positive relationship between energy use and carbon emissions while providing evidence for validating the EKC hypothesis in the context of China. In addition, the study expected to identify a negative relationship between agricultural value added and carbon emissions.

This research made valuable contributions to the existing scholarly discourse and policymaking process in several directions. Firstly, the present study addressed a gap in the existing scholarly literature by offering new insights gained from rigorous empirical analysis on the relationship between agricultural value added and CO2 emissions within the framework of the EKC hypothesis. Secondly, this study utilized the latest and most extensive dataset accumulated during a span of 32 years, specifically from 1990 to 2021. Thirdly, to ensure the validity of the findings, they were subjected to a battery of diagnostic tests and cointegration regression models (DOLS, FMOLS, and CCR). Fourthly, one of the major contributions of the study is that it raises awareness on reducing agricultural emissions by highlighting the necessity to promote clean energy and energy-efficient technologies as well as finance renewable energy technologies throughout agricultural activities in China, which could help achieve SDG 7 and carbon neutrality. Fifthly, the findings of this study would be helpful to the policymakers in China as well as other developing countries that are trying to invest in improving agricultural productivity. Investing in the agricultural sector can not only end hunger and achieve food security (SDG 2) but also contribute to achieving other SDGs, such as clean water and sanitation (SDG 6), decent work and economic growth (SDG 8), responsible consumption and production (SDG 12). Indirectly, agriculture also impacts the community and income-based SDGs addressing poverty (SDG 1), education (SDG 4), gender equality (SDG 5), climate action (SDG 13), life on land (SDG 15), and global partnerships for sustainable development (SDG 17). The outcomes of this research will provide valuable insights for contemporary scholars in the subject, as well as future researchers interested in exploring the relationship between agricultural productivity and environmental degradation. The research findings provide valuable guidance to policymakers in China as well as other agriculture-based emerging nations to achieve carbon neutrality while increasing value-added agriculture by utilizing cutting-edge technologies and promoting sustainable and climate-smart agriculture. The findings from this investigation would be useful for agricultural and environmental policy evaluation as well as further policy formation to prepare China for a 1.5 °C world by strengthening policy and action plans to reduce the climate change impacts thereby ensuring sustainable development and environmental quality in the long term. This study encourages examining the agriculture-economy-energy-environment nexus in other developing countries to establish a balance between agriculture-based economic growth and the quality of the environment. This study’s outcome could provide suggestions and recommendations to other developing countries aiming at building successful strategies to achieve sustainable agriculture while strengthening climate change mitigation and adaptation strategies.

2 Methodology

2.1 Data

The purpose of this study is to examine the relationship between agricultural value-added and CO2 emissions in China, taking into account both short-term and long-term perspectives. The analysis was conducted within the framework of the Environmental Kuznets Curve (EKC) hypothesis, while also including energy consumption as a control variable. The present study utilizes data sourced from the World Development Indicators (WDI) database, as provided by the World Bank (2023). The dataset encompasses a time span of 32 years, specifically from 1990 to 2021. The incorporation of all relevant data into the model was achieved by employing the natural logarithm, hence facilitating this process. Carbon dioxide emission was measured in metric tons per capita, while economic growth (GDP per capita) was measured in constant Chinese yuan. Energy consumption per capita was measured in kilograms of oil equivalent, and agricultural value added (including agriculture, forestry, and fishing) was measured in constant Chinese yuan. The variable names and their logarithmic structures are presented in Table 1. The annual trends of the study variables are presented in Fig. 2. Additionally, Fig. 3 illustrates the geographical distribution of agricultural regions inside China. China is classified into nine distinct agricultural regions according to geographical zoning and the regional characteristics of the grain planting system. These regions include the Northeast China Plain, the Northern arid and semiarid region, the Huang-Huai-Hai Plain, the Loess Plateau, the Qinghai Tibet Plateau, the Sichuan Basin and surrounding regions, the Middle-lower Yangtze Plain, the Yunnan-Guizhou Plateau, and Southern China.

Table 1 Variables with description
Full size table
Fig. 2
figure 2

Annual trends of the study variables

Full size image
Fig. 3
figure 3

A map of China showing the various agricultural regions (Zhang et al. 2021)

Full size image

2.2 Econometric model generation

Through the application of the EKC model, the purpose of this study was to investigate the impact that agricultural value added has on carbon emissions. As a consequence of this, the model can be shown in Eq. (1):

$$\mathrm{CO}_{2\mathrm{t }}=\mathrm{ f }(\mathrm{GDP}_{\mathrm{t}},\mathrm{ GDP}^{2}_{\mathrm{t}},\mathrm{ EU}_{\mathrm{t}},\mathrm{ AVA}_{\mathrm{t}})$$
(1)

where, CO2t is CO2 emissions at time t, GDPt is economic growth at time t, GDP2t is the square of economic growth at time t, EUt is energy use at time t, and AVAt is agricultural value added at time t.

The inclusion of the square of economic growth in the model was undertaken to examine the potential existence of a negative association between this variable and CO2 emissions. The observed inverse relationship between the square of GDP and CO2 emissions suggests that, beyond a certain level of economic development, further expansion in the economy would lead to a reduction in carbon emissions. This finding provides support for the validity of the Environmental Kuznets Curve (EKC) hypothesis in the specific context of China. The econometric model formulation for the current investigation is presented below:

$$\mathrm{CO}_{2\mathrm{t }}=\uptau_{0} +\uptau _{1}\mathrm{GDP}_{\mathrm{t}} +\uptau_{2}\mathrm{GDP}^{2}_{\mathrm{t }}+\uptau_{3}\mathrm{EU}_{\mathrm{t}} +\uptau_{4}\mathrm{AVA}_{\mathrm{t}} +\mathrm{ \varepsilon}_{\mathrm{t}}$$
(2)

where \(\uptau\) 1, \(\uptau\) 2, \(\uptau\) 3, and \(\uptau\) 4 represent the coefficients of the regressors. Additionally, τ0 denotes intercept, and \({\upvarepsilon }_{\mathrm{t}}\) is the error term. Furthermore, it is possible to express Eq. (2) in logarithmic form in the following manner:

$$\mathrm{LCO}2_{\mathrm{t }}=\uptau_{0} +\uptau_{1}\mathrm{LGDP }_{t}+\uptau_{2}\mathrm{LGDP}2_{\mathrm{t }}+\uptau_{3}\mathrm{LEU}_{\mathrm{t}}+\uptau_{4}\mathrm{LAVA}_{\mathrm{t}}+\mathrm{ \varepsilon}_{\mathrm{t}}$$
(3)

where, LCO2t is the logarithmic form of CO2 emissions at time t, LGDPt is the logarithmic form of economic growth at time t, LGDP2t is the logarithmic form of squared economic growth at time t, LEUt is the logarithmic form of energy use at time t, and LAVAt is the logarithmic form of agricultural value added at time t.

The flowchart of the analysis is presented in Fig. 4.

Fig. 4
figure 4

The analysis flowchart

Full size image

2.3 Data stationarity test

To resist a misleading regression, a stationarity test is essential (Raihan et al. 2023e). It guarantees that the series are stationary and offers a regression equation estimate based on the stationary approaches (Raihan et al. 2022d). The factual investigation supports the importance of establishing the integration technique prior to discussing the cointegration. Multiple stationarity tests have been suggested by the literature to ascertain the series integration sequence since the variable sample size affects the validity of the unit root testing (Raihan et al. 2022e; Raihan and Tuspekova 2022f). This analysis applied the Augmented Dickey–Fuller (ADF) test created by Dickey and Fuller (1981), the Dickey–Fuller generalized least squares (DF-GLS) test suggested by Elliott et al. (1992), and Phillips–Perron (P-P) test proposed by Phillips and Perron (1988) to observe the stationarity of data.

2.4 ARDL method

The ARDL bounds testing method developed by Pesaran et al. (2001) was utilized in this study to identify the cointegration among the variables. This method has many advantages over the previous cointegration methods (Raihan and Voumik 2022a). For instance, it is applicable in any investigational series integration scenario (Raihan 2023d). Moreover, the integration property of a series needed to be discovered before employing other cointegration procedures, whereas this method did not necessitate any such preliminary testing (Raihan 2023e). Furthermore, by considering the lag length of the variable, the ARDL model can be utilized to account for endogeneity (Raihan 2023f). Finally, the ARDL model maintains validity even with a small number of observations (Pesaran et al. 2001). The ARDL bound testing strategy can be written using the econometric model presented in Eq. (4):

$${\mathrm{\Delta LCO}2}_{\mathrm{t}}={\uptau }_{0}+{\uptau }_{1}{\mathrm{LCO}2}_{\mathrm{t}-1}+{\uptau }_{2}{\mathrm{LGDP}}_{\mathrm{t}-1}+{\uptau }_{3}{\mathrm{LGDP}2}_{\mathrm{t}-1}+{\uptau }_{4}{\mathrm{LEU}}_{\mathrm{t}-1}+{\uptau }_{5}{\mathrm{LAVA}}_{\mathrm{t}-1}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{1}{\mathrm{\Delta LCO}2}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{2}{\mathrm{\Delta LGDP}}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{3}{\mathrm{\Delta LGDP}2}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{4}{\mathrm{\Delta LEU}}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{5}{\mathrm{\Delta LAVA}}_{\mathrm{t}-\mathrm{i}} {+\upvarepsilon }_{\mathrm{t}}$$
(4)

where q is the optimum lag length and Δ is the first difference operator.

This study used the ARDL method developed by Pesaran et al. (2001) to fix the interplay between the factors by in the long and short term. Having shown that the parameters under investigation were cointegrated, the investigation used Eq. (4) to predict the ARDL estimate for the long-run coefficient. The short-run dynamics of the variables were investigated, as well as the short-run adjustment rate toward the long-run rate, using an approximation of the error correction term (ECT). The ECT was used to predict the short-run dynamic parameters from the long-run estimates. To do this, this study incorporated the ECM into the ARDL structure, as shown in Eq. (5):

$${\mathrm{\Delta LCO}2}_{\mathrm{t}}={\uptau }_{0}+{\uptau }_{1}{\mathrm{LCO}2}_{\mathrm{t}-1}+{\uptau }_{2}{\mathrm{LGDP}}_{\mathrm{t}-1}+{\uptau }_{3}{\mathrm{LGDP}2}_{\mathrm{t}-1}+{\uptau }_{4}{\mathrm{LEU}}_{\mathrm{t}-1}+{\uptau }_{5}{\mathrm{LAVA}}_{\mathrm{t}-1}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{1}{\mathrm{\Delta LCO}2}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{2}{\mathrm{\Delta LGDP}}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{3}{\mathrm{\Delta LGDP}2}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{4}{\mathrm{\Delta LEU}}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{i}=1}^{\mathrm{q}}{\upgamma }_{5}{\mathrm{\Delta LAVA}}_{\mathrm{t}-\mathrm{i}}+{\mathrm{\theta ECT}}_{\mathrm{t}-1}{+\upvarepsilon }_{\mathrm{t}}$$
(5)

where ECT stands for the error correction term, and θ is the ECT's coefficient.

2.5 Robustness check

Prior to drawing conclusions from a study, it is crucial to assess the sensitivity of long-run parameters derived from the ARDL model through testing. Consequently, this study employed the FMOLS, DOLS, and CCR methods to assess the reliability of the ARDL results obtained from Eq. (4). Several investigations in extant literature have documented that the FMOLS, DOLS, and CCR approaches are appropriate for assessing the robustness of the ARDL estimation (Raihan et al. 2023f; Raihan and Tuspekova 2022g). These methods offer various advantages, as indicated in the current literature.

The FMOLS methodology utilizes a semi-parametric technique for the estimation of long-run parameters, as proposed by Phillips and Hansen (1990). The utilization of this methodology yields reliable parameters even when working with a limited sample size, effectively addressing issues such as endogeneity, serial correlation, omitted variable bias, and measurement errors (Raihan et al. 2022f). Additionally, it accommodates the presence of heterogeneity in the long-run parameters (Phillips and Hansen 1990). The FMOLS method is employed to estimate a solitary cointegrating relationship that encompasses a combination of integrated order of I(1) variables. The focus of this approach is on the conversion of both data and parameters (Raihan 2023g). According to Phillips and Hansen (1990), the FMOLS method addresses the inference issues seen in classic cointegration techniques, resulting in valid estimated t-statistics for the long-run estimates.

Furthermore, the DOLS method utilizes a parametric approach to estimate a long-run relationship in a model where the variables have varied levels of integration but are nonetheless cointegrated (Stock and Watson 1993). The approach proposed by Stock and Watson (1993) addresses the issues of simultaneity bias and small sample bias through the incorporation of leads and lags (Raihan et al. 2022g). The estimators of DOLS can be derived from the least squares estimates (Raihan 2023h). These estimators possess the desirable properties of being unbiased and asymptotically efficient, even when confronted with endogeneity issues (Raihan et al. 2022h). These factors also serve to modify the potential autocorrelation and deviation from normality in the residuals (Stock and Watson 1993).

Moreover, the CCR approach proposed by Park (1992) can also be employed to test for cointegrating vectors inside a model that has an integrated process of order one I(1). The model exhibits a high degree of similarity to FMOLS in terms of its characteristics (Raihan et al. 2022i). However, it is important to note that there exists a distinction between the two approaches. Specifically, CCR primarily emphasizes the transformation of data alone, whereas FMOLS places equal emphasis on the transformation of both data and parameters (Park 1992). Moreover, the CCR is a regression model that can be utilized for both single equation regression and multivariate regression without any need for change, hence ensuring the preservation of efficiency (Park 1992).

3 Results and discussion

3.1 Descriptive statistics

The descriptive statistics of the study's variables are shown in Table 2. The results of multiple normality tests are shown alongside a description of the data. It was clear that the dataset was normal because the skewness values were near zero. In addition, kurtosis values below 3 were indicative of platykurtic variables. All the variables are normally distributed, as shown by the Jarque–Bera and probability values.

Table 2 Statistical summaries of the variables
Full size table

3.2 Result of the unit root tests

The fundamental goal of this study was to investigate the development of a lasting connection between the series under consideration. The evaluation of the unit root test provides crucial information on the integration characteristic of the parameters, which is necessary for employing the approaches in the creation of a long-term interrelationship. Therefore, the integration qualities of the series were investigated using a battery of conventional root tests, including the ADF, DF-GLS, and PP tests. The findings of the stationarity test are summarized in Table 3. According to the canonical tests of unit root output, all variables exhibited the unit root problem at the level before becoming stationary after the first difference.

Table 3 Unit Root test results
Full size table

3.3 Result of the ARDL bound test

Based on the unit root observations, this research found that the series under examination is an I(1) series. Therefore, this study used the ARDL-based bounds test technique to evaluate the long- and short-term connections between the series under investigation. The model's consistent results from each alternative's information criterion led this investigation to settle on the lag specification. Table 4 shows that the calculated F-statistic (5.29) for the cointegration analysis is significantly higher than the upper critical threshold. This investigation therefore concludes that the independent variable and the regressors are cointegrated.

Table 4 ARDL bounds test results
Full size table

3.4 ARDL long- and short-run outcomes

Once it was determined that there existed a cointegration relationship between the variables under investigation, this study proceeded to compute the coefficients of the long-term link between the variables. The long-term and short-term outcomes of the variables that were analyzed with ARDL are presented in Table 5. Both in the short term and the long term, the findings are consistent with one another. The results of a longitudinal study indicate that there is a significant relationship between GDP and energy use (independent variables) and CO2 emission (dependent variable). According to the findings of this study, there is a correlation between GDP and CO2 emissions that is positive when the signs of the connections are considered. If the agricultural value added remains constant, a 1% increase in GDP will result in a 2.42% (long-run) and 1.64% (long-run) increase in carbon emissions, respectively. However, the findings show that the coefficients of GDP squared are negative and significant at a 5% level of significance, indicating that a 1% increase in the square of GDP will lead to a reduction of CO2 emissions by 0.09% in the long run and 0.08% in the short run. The present study’s result on the relationship between economic growth and CO2 emissions indicates that environmental degradation will be reduced in China after a certain level of economic growth. This outcome provides the evidence that EKC hypothesis is valid in the context of China.

Table 5 ARDL long and short-run results
Full size table

Moreover, the statistically significant positive link at the 1% level between energy usage and CO2 emissions reveals that China's rising energy demand is contributing to environmental degradation. The findings indicate that a marginal increase of 1% in energy consumption will lead to a corresponding rise of 1.27% in long-term CO2 emissions and 1.18% in short-term CO2 emissions. At the same time, it was discovered that the value added in agriculture had an inverse association with the amount of carbon emissions produced. According to this study's findings, a 1% increase in the agricultural value added would cut CO2 emissions by 1.37% in the long-run and 0.65% in the short-run. Both the long-run estimation R2 and the adjusted R2 are at a value of 0.9998 and 0.9894, which indicates that the developed regression model fits the data exceptionally well. This would imply that the changes in the dependent variable can be explained to a nearly perfect degree by the independent causes. Consistent with the long-term findings, the short-term results reveal a good influence on GDP and energy consumption from CO2 emissions, but an adverse effect on agricultural value added. The fact that the short-term equilibrium deviations of the model are balanced over the long term is suggested by the fact that the error correction coefficient that was determined by applying the error correction model was -1.2233 and that this value was statistically significant. Additionally, if the coefficient of the error correction variable is bigger than one, the system will fluctuate and then stabilize. The amplitude of this fluctuation is reducing as time passes, which makes it possible for there to be a return to equilibrium.

3.5 Diagnostic tests

Several Diagnostic tests were used to examine and validate the model's viability. The outcomes of the diagnostic tests are presented in Table 6. The results showed that the expected model is acceptable in terms of autocorrelation, normality, shifting variance, and model-building error. In addition to this, this investigation carried out stability tests for the predicted long-term model by utilizing the cumulative sum of recursive residuals (CUSUM) and the cumulative sum of squares of recursive residuals (CUSUMQ). The results of the CUSUM and CUSUMSQ tests are depicted in Fig. 5. The graphs show that the anticipated model is stable during the relevant period since the residuals are contained within the key boundaries at the 5% significance level.

Table 6 The results of the diagnostic tests
Full size table
Fig. 5
figure 5

Plot of CUSUM and CUSUMQ tests

Full size image

3.6 Robustness check

The robustness of the long-run estimation from the ARDL approach was verified by utilizing the FMOLS, DOLS, and CCR tests. Table 7 displays the estimated results from the use of FMOLS, DOLS, and CCR methods. The sign and reliability of the FMOLS, DOLS, and CCR results were all shown to be consistent and dependable. This causes them to produce the same results as the ARDL simulations in the long term. Specifically, the data showed that increasing GDP and energy use increased CO2 while increasing agricultural value added decreased CO2 emissions. Moreover, the positive coefficients of LGDP and the negative coefficients of LGDP2 indicate the validity of the EKC hypothesis in all three models. Hence, the FMOLS, DOLS, and CCR results are duly harmonized with the ARDL outcomes. Furthermore, the higher values of R2 and the adjusted R2 demonstrate the accuracy of the estimation.

Table 7 Robustness results
Full size table

4 Discussion and policy implications

This study investigated the influence of economic growth, energy use, and agricultural value added on carbon emissions in the context of China. The findings revealed a strong positive influence of GDP and energy use on carbon emissions. When the economy grows, more energy, particularly fossil fuels, is required for production, which leads to an increase in carbon dioxide emissions (Raihan et al. 2022j). Results like the positive association between GDP and CO2 emissions in the case of China are congruent with those found by Gessesse and He (2020), Yilanci and Pata (2020), and Kongkuah et al. (2022). The findings suggest that China places a higher priority on economic growth than it does on protecting the environment. However, both in the long and short run, the study revealed a positive coefficient of GDP while a negative coefficient of GDP square. The findings suggest a correlation between environmental pollution and economic growth in the form of an inverted U-shaped connection, confirming the validity of the Environmental Kuznets Curve (EKC) hypothesis in the context of China. This outcome implies that increased economic activity leads to increased levels of environmental pollution, at least in the near run; nevertheless, future advances in per capita income would eventually lead to cleaner environmental conditions (Raihan 2023i). A similar type of relationship between economic growth and environmental degradation to validate the EKC hypothesis has been reported by a number of studies in the literature, including those by Chang et al. (2021) in China, Suki et al. (2020) in Malaysia, Ahmad et al. (2021) in Pakistan, Genç et al. (2022) in Turkey, Sultana et al. (2022) in Bangladesh, Verbič et al. (2021) in Europe, Cheikh et al. (2021) in MENA region, and Pata et al. (2023) in ASEAN countries.

The present study’s findings revealed a positive link between energy use and carbon emissions in the case of China, implying that increased energy consumption triggers environmental degradation. This is particularly true in regard to China's contribution to global warming. It is quite likely that this is the case considering that China derives a significant portion of its energy requirements from fossil fuels. In 2021, coal constituted approximately 55% of China's total energy consumption, while petroleum and other liquid fuels contributed 19% of the nation's total usage of energy (EIA 2022). The findings of this study have been backed up by a number of other researchers, including Gessesse and He (2020), Yilanci and Pata (2020), and Kongkuah et al. (2022) all of whom have reached the same conclusions regarding the positive link between energy use and CO2 emissions in China. Nevertheless, the results of the analysis indicate that increased value added from agriculture helps China to improve the environmental quality by reducing carbon emissions. This finding is in line with those discovered in prior research, such as that conducted by Prastiyo et al. (2020), Rafiq et al. (2016), Raihan and Tuspekova (2022h), Jebli and Youssef (2017), Liu et al. (2017), Waheed et al. (2018), Ullah et al. (2018), Dogan (2016), Anwar et al. (2019), Wang et al. (2020), Koshta et al. (2021), and Raihan (2023c). This finding can be explained by the fact that the agricultural sector in China needs to transition using more energy that comes from renewable sources or using energy in a more efficient manner. Renewable energy not only helps the farmer save money, but it also combats global warming and climate change (Raihan et al. 2023g). Biomass, geothermal, hydroelectric, solar, and wind energy can be used to generate electricity for warmth, lighting, and farm fuel (Al-Ismail et al. 2023). International Renewable Energy Agency (IRENA 2019) projected that electrification, renewable energy, and energy efficiency measures would provide over 90% of the emission reductions required by 2050. Renewable power and electrification of heat and transport alone would reduce emissions by 75% (IRENA 2019).

The production of agricultural goods is beneficial to the environment because the CO2 that would otherwise be released into the air is instead stored as biomass or as carbon in the soil (Ali et al. 2022; Raihan and Said 2022). The production of agricultural goods is considered to be an emitter of carbon dioxide despite the fact that these goods first ameliorate existing environmental conditions before contributing to their deterioration (Raihan and Tuspekova 2022i). To counteract this, however, antiquated and ineffective agricultural practices need to be phased out in favor of more forward-thinking approaches that not only boost agricultural output and cut emissions but also ensure the safety of food and nutrition for China's rapidly growing population. Illustrations of forward-thinking methodologies include regenerative agriculture, low-carbon agriculture, circular agriculture, sustainable agriculture, and climate-smart agricultural practices (Raihan and Tuspekova 2022j). These approaches offer a promising avenue for establishing a sustainable food system that effectively caters to our nutritional needs while simultaneously mitigating ecological repercussions. In response to the growing concern about agricultural emissions, forward-thinking strategies are being implemented globally, including China. In addition to making the environment cleaner and safer, lowering agricultural emissions might make way for new opportunities for economic expansion by stimulating an increase in agricultural activity (Feng et al. 2023). This would be accomplished by encouraging more farming to take place. It is possible to reduce one's carbon footprint by absorbing the carbon that is released as a byproduct of agricultural processes and putting that carbon to productive use, provided that the appropriate treatment and technology are applied (Cui et al. 2022), such as soil conservation for carbon sequestration, organic fertilizer production from livestock manure, and biofuel production from agricultural residues.

According to the findings of this study, China's agricultural development may be able to take advantage of a once-in-a-lifetime opportunity: increased agricultural value added does not always result in an increase in CO2 emissions or in the degradation of the environment. According to the findings of this research, it is suggested that the government of China adopt policies that enhance the quality of the environment by increasing the value that is added to agricultural products. It is a good idea to increase the agricultural sector's contribution to China's economy; however, policymakers should keep in mind that doing so will have a negative impact on the long-term deterioration of the environment if agricultural practices remain as inefficient as they have been for centuries. There is a pressing need to enhance agricultural productivity in order to meet the increasing demand for food. This can be achieved through various strategies, including the advancement of disease-resistant and high-yielding crop varieties, as well as the promotion of more productive farming techniques by encouraging farmers to transition away from traditional methods (Salgotra et al. 2023). In addition, the utilization of contemporary agricultural practices, the distribution of high-quality seeds, and the availability of various other agricultural tools and resources all have the potential to strengthen agricultural expansion and value-added characteristics to a higher level (Song et al. 2023).

Farmers have a responsibility to stay current on issues pertaining to the environment as well as developments in agricultural technology (Neethirajan 2023). Due to the fact that conventional farming practices are responsible for a significant amount of environmental damage and that damage can be mitigated through the use of precision agriculture, green farming, good farming techniques, and green goods, farmers should be encouraged to take a greater interest in environmental issues (Çakmakçı et al. 2023). Through the use of organic and low-carbon farming practices, sustainable agriculture has the potential to cut emissions and increase carbon storage (Zhou et al. 2023). Agricultural production models need to be rapidly modified in the direction of high-potential export markets such as the European Union in order to get access to those markets (Raihan 2023j; Timpanaro et al. 2023). Farmers should feel more confident adopting forward-thinking, ecologically friendly technologies if the government offers financial support in the form of incentives, awards, and long-term loans that are interest-free. To put it another way, innovative farmers may be confident that the government would support their attempts to adopt technologies that are less detrimental to the environment (Raihan 2023k; Wang et al. 2023).

Long-term structural reforms should be made to increase the percentage of renewable resources in overall energy consumption (Raihan and Tuspekova 2022k; Raihan 2023l). Additionally, China might push the use of renewable and alternative energy sources in other industries as well as a substitute for fossil fuels. Reducing CO2 density in the country can be improved by increasing the share of renewable energy in overall energy use (Ghosh et al. 2023; Raihan 2023m). China currently holds the position of global leader in electricity generation derived from renewable energy sources (Panchenko et al. 2023). The renewable energy sector in China is experiencing a more rapid expansion compared to its fossil fuels and nuclear power capacity (Li et al. 2022). It is projected to provide a significant contribution of 43 percent to the overall global growth in renewable energy capacity (Li and Umair 2023). Despite China's impressive status as the global leader in terms of installed capacity for hydro, solar, and wind power, it is important to note that its substantial energy demands necessitate a significant reliance on non-renewable sources. In 2019, renewable energy sources accounted for 26% of China's electricity generation, with the majority of the remaining power being generated by coal-fired power plants (Yin and Duan 2022). Alternative energy sources have the potential to reduce the amount of coal that is used in power plants and the amount of pollution that is produced as a result.

As of the beginning of 2020, China's total installed electric power capacity was composed of approximately 40% renewable energy sources, while renewable energy accounted for 26% of the country's overall power generation (Yang et al. 2023). As of 2021, the proportion of total electricity generation had increased to 29.4% (Li et al. 2023). It is anticipated that the proportion of renewable energy sources in overall electricity production will persistently rise, reaching 36% by the year 2025 (Wei et al. 2022). This trajectory aligns with China's commitment to attain carbon neutrality prior to 2060 and to reach the peak of its emissions by 2030 (Li et al. 2022). China holds the distinction of being the foremost global investor in renewable energy (Raihan and Voumik 2022b). In the year 2022, the nation allocated a total of $546 billion towards various investment initiatives, encompassing the domains of solar and wind energy, electric vehicles, and batteries (Saqib and Dincă 2023). The aforementioned figure represents an investment amount that is approximately four times greater than the overall investments made by the United States, which amounted to $141 billion.

In addition, carbon capture and storage techniques ought to be put into practice in order to cut down on emissions that are contributing factors to climate change (Raihan and Tuspekova 2023a). The implementation of a carbon tax has the potential to dissuade the utilization of fossil fuels and promote a transition towards less environmentally detrimental energy sources, therefore mitigating the release of CO2 emissions (Chen et al. 2023; Raihan 2023n). The formulation of a carbon tax encompasses two fundamental elements: the magnitude of the tax and the allocation of the generated money (Baranzini et al. 2017). The former concept relies on the utilization of the social cost of carbon (SSC), which endeavors to quantitatively assess the financial implications associated with the externalities stemming from carbon emissions. The exact numerical value is a topic of discussion within academic and policy spheres. A positive correlation exists between the SSC and the assessment of the societal impacts resulting from carbon pollution (Rode et al. 2021). According to conservative estimates, the SSC is approximately $50 (Wagner 2021). The allocation of income is a topic of contention within carbon tax ideas (Povitkina et al. 2021). A government has the option to utilize its money for the purpose of augmenting its discretionary spending or mitigating deficits. Nevertheless, these ideas frequently face the potential drawback of being regressive and inciting public opposition as a result of the elevated energy expenses associated with these levies (Povitkina et al. 2021). In order to mitigate this issue and enhance the acceptance of a carbon tax, a governing body may opt to implement a revenue-neutral approach to the carbon tax (Raihan and Tuspekova 2023b). One potential approach to achieve this objective involves implementing a reduction in income tax that is commensurate with the magnitude of the carbon tax (Yao et al. 2022). Alternatively, another strategy entails distributing the monies generated from the carbon tax back to individuals in the form of a dividend (Maestre-Andrés et al. 2019).

It is also essential to encourage high-income growth while simultaneously shifting toward less polluting types of energy use, such as renewable energy, and investing in environmentally friendly agricultural practices (Raihan 2023o). In order to achieve sustainable agricultural productivity over the long run, the government may decide to promote a power infrastructure that is more effective and provide assistance with the transition to greener, more intensive agricultural energy sources (Raihan 2023p). Since the use of clean, renewable energy sources like wind, solar, and biofuel can increase agricultural production while also contributing to the alleviating climate change or global warming, the government may decide to incentivize the use of these sources (Raihan 2023q). If the corporation were offered financial incentives to use renewable energy in agricultural production, it would have an advantage over its competitors in global markets (Raihan 2023r). The proportion of energy consumption in the agricultural sector in China shows a steady upward trend, accompanied by a corresponding increase in the input of energy resources (Raihan 2023s). Numerous research (Qiao et al. 2019; Zhang et al. 2019; Ridzuan et al. 2020) have advocated for the reduction of non-renewable energy investment in the agricultural sector through the integration of well-established renewable energy power generating technologies, such as solar and wind power generation, in response to the growing prominence of renewable energy production. In 2017, the Guiding Opinions on the Construction of Clean Energy Utilization Project of Crop Straw Gasification were jointly issued by the National Development and Reform Commission, the Ministry of Agriculture, and the National Energy Administration. It focused on the utilization of pyrolysis technology and anaerobic fermentation technology to produce biogas from agricultural wastes (Zhang et al. 2022). By the year 2018, rural energy policies had placed significant emphasis on biogas and electricity as key focal points. Additionally, there was a notable rise in the number of policies addressing solar and wind energy (Raihan 2023t). It suggests that there has been a prominent shift in the emphasis on agricultural energy policies towards the promotion of environmentally sustainable energy sources (Zhang et al. 2022).

It is possible to achieve a carbon–neutral environment by switching irrigation systems from utilizing non-renewable sources of electricity to utilizing renewable sources of energy (Raihan 2023u). Other significant developments in agriculture include the promotion of the use of solar tube wells for irrigation, organic farming, tunnel farming, the transition from conventional farming to no-till farming, and the reduction of the environmental impact that is caused by the use of fertilizer (Raihan 2023v). Multiple studies have documented the utilization of renewable energy sources across many agricultural domains. In their respective studies, Shoeb and Shafiullah (2018), Sudharshan et al. (2019), and García et al. (2019) have documented the utilization of solar energy panels specifically for the purpose of irrigation. Furthermore, the application of wind energy for irrigation was discussed by Serrano-Tovar et al. (2019) and Caldera and Breyer (2019), while the utilization of geothermal energy for irrigation was reported by Alberti et al. (2018). Nevertheless, the research findings indicate that the selection of the system is heavily influenced by factors such as cost, fluctuations in wind speed, solar radiation, and the size of the system. All of these developments have been made possible by technological advancements. This cutting-edge agricultural technology can be beneficial to large farms in a number of ways, including lowering labor costs, increasing output, and lowering emissions (Raihan 2023w). The World Resources Institute (WRI 2019) reported that the utilization of renewable and cutting-edge agricultural technologies has the potential to cut nearly 65% of projected agricultural energy emissions in 2050. Additionally, investments in intensive farming in China made feasible by increased international cooperation would help cut emissions from the agricultural sector of the country's economy while simultaneously increasing agricultural output. This will be a win–win situation for the country.

However, it should be noted that this study has a few limitations, but that is to be expected with any kind of research. One of the critical drawbacks of the analysis is the unavailability of data beyond the period of study, which limits the power of the econometric techniques used. Furthermore, this study utilized CO2 emissions as a representative measure of environmental deterioration, while omitting to incorporate methane and nitrous oxide, which are other greenhouse gases released by the agricultural industry.

5 Conclusions

The agriculture sector is the second major contributor to global warming and climate change due to the anthropogenic emissions from different agricultural activities. On the other hand, agricultural value added boosts economic growth while ensuring food security. In addition, increased economic growth from agricultural value added helps to increase financing environment-friendly technologies for reducing emissions from the agricultural sector. Therefore, this study aims to ascertain whether or not the value-added agricultural industry is a significant factor in environmental issues. By employing the ARDL approach utilizing the annual time series data between the years 1990 and 2021, this study investigated the dynamic impacts of economic growth, energy use, and agricultural value added on CO2 emissions in China. The results of the ADF, DF-GLS, and P-P unit root tests confirmed the data stationarity. In addition, the ARDL bounds test results provided evidence of long-run cointegration between the variables. Furthermore, the robustness of the ARDL long- and short-run estimation was confirmed by employing the FMOLS, DOLS, and CCR techniques. According to the findings of the ARDL long- and short-run estimation, there is a strong positive connection not just between GDP and CO2 emissions but also between energy use and CO2 emissions. The results revealed that rapid economic growth and increased energy use in China would result in an upsurge of CO2 emissions both in the long- and short-run. Moreover, when the GDP and squared GDP were introduced into the model to investigate the inverted U-shaped relationship between economic growth and pollution, the EKC hypothesis was validated and shown to be correct in the context of China. Moreover, the ARDL outcomes revealed that the agricultural value added does not have a positive impact on CO2 emissions. Instead, the negative relationship between agricultural value added and CO2 emissions indicates that increasing agricultural value added helps to improve the environmental quality by reducing CO2 emissions both in the long short run. Finally, this study provides several policy recommendations for promoting sustainable and climate-smart agriculture as well as utilizing innovative technologies in the agricultural sector which could help to achieve China’s dream of carbon neutrality while ensuring food security.

This investigation was conducted to determine how China's value-added agriculture affects the country's overall CO2 emissions. In the future, studies might be carried out to investigate the impact of agricultural value added on the emissions of other GHGs, such as methane and nitrous oxide. Moreover, further research might be undertaken for other developing countries while utilizing different econometric modeling or micro-disaggregated data to investigate the agriculture-economy-energy-environment nexus. Furthermore, future research can consider other agricultural determinants that have not been addressed in this study, such as agricultural land expansion, crop production, livestock production, fisheries production, energy use in agriculture, and fertilizer consumption. Additionally, future research can compare country-specific results to overall panel outcomes utilizing advanced econometric approaches, in addition to panel estimations. These approaches could provide useful comparisons with the findings of this study, shedding light on future research in the field.