Study on the Evolution and Trends of Agricultural Carbon Emission Intensity and Agricultural Economic Development Levels—Evidence from Jiangxi Province

Abstract

The study of the evolutionary patterns and trends of agricultural carbon emission intensity and agricultural economic development levels plays an important role in promoting the green and low-carbon sustainable development of agriculture. This paper adopts the carbon emission factor method to measure the agricultural carbon emissions in Jiangxi Province from 2001 to 2020, uses the LMDI decomposition method to explore the drivers of carbon emissions, and further analyzes the coupling relationship between agricultural carbon emissions and the agricultural economy using the Tapio decoupling model, based on which a GM (1,1) model is used to forecast the agricultural carbon emissions in Jiangxi Province from 2001 to 2015. According to the research results, agricultural carbon emissions in Jiangxi Province show a trend of “rising and then falling”, with the intensity decreasing; the level of economic development is the main factor that increases carbon emissions, while the efficiency of agricultural production, the size of the labor force, and the structure of agricultural production have positive effects in terms of reducing carbon emissions. How to reduce carbon emissions while promoting agricultural economic development is an issue that remains to be addressed in the future. Further analysis found that the decoupling states of Jiangxi Province from 2001 to 2009 switched between strong decoupling and weak decoupling, with weak decoupling dominating the years 2010–2015 and strong decoupling dominating from 2016 onwards. With the continuous promotion of carbon emission reduction, agricultural carbon emissions in Jiangxi Province will continue to show a decreasing trend over the next five years. Three policy recommendations are put forth in order to advance the effort to reduce agricultural carbon emissions in Jiangxi Province: cultivating high-quality and low-carbon rice varieties, switching to green agricultural production, and coordinating the connection between economic growth and agricultural carbon emissions.

1. Introduction

Large amounts of greenhouse gas emissions can damage the environment and cause global warming, resulting in extreme weather, which, in turn, affects human production and life. To cope with the threat posed by global warming to humans, countries at the UN Climate Change Conferences have made carbon (C) peaking and carbon neutrality important national strategies. To ease the huge global pressure on carbon emission reduction, China has adopted more extensive policies and effective measures and has pledged to reach peak CO₂ emissions by 2030 and to achieve carbon neutrality by 2060. Agriculture is one of the most important carbon sources in China. The use of agricultural materials, rice cultivation, farmland utilization, and livestock production activities are the main sources of agricultural carbon emissions. Some research data show that China’s agricultural carbon emissions account for 15–18% of the total national carbon emissions [1]. In this context of carbon emission reduction in Chinese agriculture—i.e., a binding target—how to promote the green and sustainable development of Chinese agriculture while ensuring its quality and efficiency has become an important issue in current agricultural development [2]. Jiangxi, as one of the major agricultural provinces, has a pivotal position in China. Therefore, it is urgent to identify the patterns and trends of the evolution of agricultural carbon emission intensity and agricultural economic development levels, which are significant for the early achievement of carbon peaking and carbon neutrality.

Agricultural carbon emissions, as the name implies, are the emissions of various greenhouse gases caused by agricultural production activities. For example, the use of chemical fertilizers and pesticides, the growth of crops, and the manure produced by livestock and poultry farming all produce large amounts of greenhouse gas emissions such as CH₄ and N₂O. The agricultural carbon emission measurement index system was first constructed by West and Marland [3], using four sources of carbon emissions—fertilizer, pesticides, agricultural irrigation, and seed breeding—to measure agricultural carbon emissions in the United States. The domestic research on agricultural carbon emissions originated from “greenhouse gas emissions”, and Dong Hongmin, Lin Erda, and Yang Qichang [4] used CH₄ and N₂O from rice cultivation and livestock farming as emission sources to measure greenhouse gas emissions in China. Schlesinger [5] confirmed that the carbon emissions generated from plantations in China far exceed the carbon sequestration potential of agricultural land, and many scholars [6,7,8,9] further found that emissions from livestock farming are also an important component of agricultural greenhouse gas emissions in China. The concept of “agricultural carbon emissions” was formally introduced in China by Li Bo [10], who measured agricultural carbon emissions using agricultural material inputs and agricultural land-use activities as carbon sources. Subsequently, many scholars have measured the agricultural carbon emissions of individual agricultural sectors in China using agricultural material inputs [11], carbon emissions from livestock and poultry breeding [12], and carbon emissions from agricultural land use [13]. With continuous research on agricultural carbon emissions, more and more scholars [14,15] believe that agricultural carbon emissions should encompass agricultural material inputs, rice cultivation, livestock and poultry breeding, and agricultural land use. Reviewing the research related to agricultural carbon emissions, we found that the current research on agricultural carbon emissions focuses on the following aspects:

First is the measurement of agricultural carbon emissions and their influencing factors. Agricultural carbon emissions are measured in various ways, including mass balance, measured value, modeling, and emission factor methods [16,17]. For example, some scholars [18] used the carbon emission factor method to measure carbon emissions and found that agriculture in Jiangsu Province has reached its peak and that the future low-carbon development of agriculture and the continuous decline in carbon emissions will help accelerate the achievement of the carbon neutrality target in Jiangsu Province. In addition, some scholars [19] have also used stochastic frontier analysis to study the municipal differences in agricultural carbon emissions. Scholars have also measured the spatial and temporal characteristics of agricultural carbon emissions, the dissipation of agricultural carbon emissions, and the constraining effects of agricultural carbon emissions in different provinces [20,21,22]. Among them, Yun Tian, Zhang Junbiao, and Li Bo [23] constructed an evaluation matrix of regional differences in the effects of agricultural carbon emissions to demonstrate the differences in agricultural economic levels between different provinces. Zhang Songxin [24] further found that the agricultural carbon intensity can be evaluated using eight reliable tools: low-carbon technology, agricultural structure, environmental policies, economic development, environmental awareness, production efficiency, agricultural development status, and labor quality. Gui He and Li Jing et al. [25], on the other hand, argued that both agricultural economic growth and agricultural emission reduction in Ningxia were effective. Chen Yin’e and Chen Wei [26], from an industrial perspective, argued that agricultural carbon emissions showed a negative correlation with the level of agricultural mechanization and a positive correlation with industrial upgrading. Some scholars [27] further found that an important measure to mitigate carbon emissions is to improve agricultural production efficiency and to promote energy-enhancing technological advances.

Second is the study of the relationship between agricultural carbon emissions and economic decoupling. In a study of agricultural carbon emissions and farmers’ per capita income in the provinces and cities of the Yangtze River Basin, Ding Baogen, Zhao Yu, and Luo Zhihong [28] found an inverted U-shaped EKC curve relationship between the two. Zhang Huanhuan and Wang Qiang [29] used the Tapio decoupling model to find a coupling relationship between agricultural carbon emissions and the agricultural economy in Henan Province, with weak and strong decoupling as the main decoupling relationships. In their study, Jiang Tiancheng, Hu Chun, and Wang Qiaozhi [30] also confirmed the above view and further suggested that limiting the use of pesticides and fertilizers and promoting the resource utilization of stored poultry manure are effective strategies to promote carbon emission reduction in agriculture.

Third is the forecasting of agricultural carbon emission trends. The introduction of the carbon peaking and carbon neutrality goals has accelerated the research on agricultural carbon emissions. Carbon emission projections through environmental Kuznets curves and input–output models have begun to emerge [31,32,33], with the aim of exploring carbon reduction pressures and better proposing targeted reduction strategies. The GM (1,1) grey prediction model is an effective tool for the prediction of small samples due to its ability to avoid the shortcomings of other models in predicting values that can be predicted with a small amount of data [34]. With the wide application of grey prediction theory, more and more scholars [35,36] have made short-term predictions of national or interprovincial carbon emissions through GM (1,1) models, and the prediction results are satisfactory.

Naturally, the results of previous research provide not only a strong basis for the study of agricultural carbon emissions and their reduction in China but also a reference for the measurement methods, index systems, and model construction for agricultural carbon emissions in this paper. However, there are some limitations. For one, with further research, the existing studies only consider the national level or provinces with higher levels of economic development, without considering the interprovincial differences and specificities that may have different impacts on agricultural carbon emissions. Agriculture is an important industrial activity in Jiangxi Province, China’s important food production base. However, due to the geographical environment and economic development level, agricultural production is mainly performed in a rough way and, along with agricultural income increase, is largely dependent on agricultural material inputs, making the task of reducing agricultural carbon emissions in Jiangxi Province unique. For another, most previous studies were conducted too long ago, mainly studying the interprovincial and county-level carbon emissions and influencing factors in Jiangxi through the perspective of agricultural material inputs [37,38,39], and there has been no research to date on the relationships between agricultural carbon emissions, economic decoupling, and carbon emission trends. Therefore, it is necessary to take Jiangxi as a separate research subject to accurately grasp the driving mechanism of its agricultural carbon emissions.

Have the temporal and spatial aspects of Jiangxi Province’s agricultural carbon emissions changed recently? Do the factors affecting agricultural carbon emissions have any recent changes? Has Jiangxi Province similarly attained the decoupling between agricultural carbon emissions and the degree of economic development? Will Jiangxi’s agricultural carbon emissions continue to decrease over time? To this end, based on the existing results, this paper takes Jiangxi Province as the study area, integrates the production of agricultural plantations and livestock from 2001 to 2020, measures Jiangxi’s agricultural carbon emissions using the carbon emission factor method, and uses the Logarithmic Mean Divisia Index model decomposition method to derive the drivers of carbon emissions. Furthermore, the paper explores the coupling relationship between agricultural carbon emissions and the agricultural economy through the Tapio decoupling model. Finally, a GM (1,1) model is constructed to forecast the carbon emissions of Jiangxi’s agriculture for the next five years, with the aim of providing reference and support for the agricultural carbon emission reduction activities in Jiangxi Province and helping to achieve sustainable and high-quality development of Chinese agriculture, along with the carbon peaking and carbon neutrality goals. This paper provides several new contributions: first, it enriches the literature on agricultural carbon emissions in provincial regions with less developed economies, providing new perspectives and ideas for the future; second, on the basis of using the carbon emission factor method and Logarithmic Mean Divisia Index model to study the influencing factors of agricultural carbon emissions, the Tapio decoupling model and GM (1,1) model are used to fill the research gap of the decoupling relationship between agricultural carbon emissions and the economy in Jiangxi Province, along with the forecast of carbon emission trends for the next five years.

2. Data and Study Method Determination

2.1. Indicator Selection

China is a vast country with a history of extensive agriculture, and different regions have different levels of economic growth and agricultural production activities. By reviewing the China Statistical Yearbook and other relevant information, we found that that Jiangxi, the largest agricultural province in the nation, has made significant sacrifices to ensure national food security. Jiangxi is the third largest rice production base in China, a vegetable production base for major cities such as Guangzhou, Shenzhen, and Hong Kong, as well as a major province for the transfer of live pigs and beef cattle. However, the high yield and efficient agricultural production are inevitably accompanied using fertilizers, pesticides, and other agricultural inputs, which produce large amounts of agricultural carbon emissions, causing enormous pressure on the environment.

To better study and analyze agricultural carbon emissions in Jiangxi, carbon emissions from plantation and animal husbandry were selected as measurement indicators, as these two industries are the most important in the composition of the agricultural economy in Jiangxi, accounting for 45.23% and 28.53% of agricultural gross domestic product from 2001 to 2020, respectively, while the average contributions of forestry, secondary industry, and fishing were 9.03%, 3.45%, and 13.76%, respectively, according to our calculations. The selected indicators were mainly the carbon emissions generated directly and indirectly in the process of agricultural material input, rice field planting, farmland utilization, and livestock and poultry breeding. Among these, carbon emissions from factor inputs are mainly based on end consumption and cultivation scale, while those from livestock and poultry are based on their whole lifecycle. The indicator data are from the 2001–2021 Jiangxi Statistical Yearbook. The process of calculating agricultural carbon emissions is shown in Figure 1.

Agricultural production activities lead to large amounts of agricultural carbon emissions [40,41,42,43]. Therefore, the design of the agricultural carbon emission index system in this paper consists of the following four components: The first is the carbon emissions caused by agricultural material inputs; for this paper, a total of six indicators—fertilizer, agricultural film, pesticides, agricultural land tillage, agricultural irrigation, and agricultural diesel fuel—were selected as the carbon sources of agricultural material inputs, and the calculation method of related carbon emission coefficients was as described in the study of Li Bo [10]. The second is the methane emissions from rice fields induced by rice cultivation; since the area of rice cultivation in Jiangxi Province is large, CH₄ produced by rice was studied separately as a carbon source, and its carbon emission coefficients were assigned and approved separately in combination with the relevant studies of Min Jisheng and Hu Hao et al., in the process of calculating carbon emissions [14]. The third is carbon emissions from farmland utilization; for this paper, a total of six indicators—rice, soybeans, corn, potatoes, oil crops, and vegetables, which are grown on a large scale in Jiangxi Province—were selected to be included in the carbon emissions from farmland use, and the corresponding carbon emission coefficients were derived from the study of Gui He, Li Jing, and Shang Mengyuan [25]. The fourth is carbon emissions from livestock farming; this mainly refers to CH₄ produced by intestinal fermentation of ruminants, along with CH₄ and N₂O produced by livestock excretion. By reviewing the relevant information, five species—pigs, cattle, sheep, rabbits, and poultry—were selected as livestock carbon sources. For the sake of scientific and accurate accounting, the CH₄ emission coefficient of livestock in this study was selected from the United Nations Intergovernmental Panel on Climate Change [44], while the N₂O emission coefficient of livestock excretion was derived from the work of Hu Xiangdong and Wang Jimin [8]. For estimation, both CH₄ and N₂O generated by the above indicators were converted into CO₂ according to 28 and 265 times the 100-year scale CO₂ warming potential, respectively, in the United Nations Intergovernmental Panel on Climate Change Fifth Assessment Report.

2.2. Research Method Selection

2.2.1. Total Agricultural Carbon Emissions

In this study, the following equation was used to account for agricultural carbon emissions:

$$C={\displaystyle \sum }{C}_{it}={\displaystyle \sum }{T}_{it}\times {\delta }_i$$

(1)

where C represents the total agricultural carbon emissions, $C_{it}$ represents the carbon emissions generated by the superscript category of carbon source in the year t, $T_{it}$ is the use or production of the i-th category of carbon source in the year t, and ${\delta }_i$ is the emission factor of each carbon source factor.

2.2.2. Carbon Emission Intensity

In this study, the measurement of agricultural carbon emissions includes both cultivation and livestock; thus, the calculated carbon emission intensity is measured by carbon emission per unit of gross domestic product, which is a very important indicator of the level of economic development of agricultural production and carbon emissions, as this indicator can reflect the carbon emission efficiency. The expression is as follows:

$$I=\frac{C}{P}$$

(2)

where I denote carbon emission intensity per unit of agricultural gross domestic product (t C/10⁴ RMB), C denotes the total agricultural carbon emissions (t C), and P denotes the gross agricultural farming and livestock production in a certain period (10⁴ RMB, comparable prices).

2.2.3. Decomposition Method of Factors Influencing Carbon Emissions

Agricultural carbon emissions are closely related to the economy and society, and agricultural development will inevitably result in the generation of carbon emissions, while the uncontrolled increase of agricultural carbon emissions can cause extreme weather, which is not conducive to the quality and efficiency of the agricultural industry. Therefore, it is important to grasp the driving factors of agricultural carbon emissions, understand the effect of each factor on carbon emissions, and formulate agricultural carbon emission reduction strategies according to the real situation. In this paper, based on the studies of Zhang Xiaoping [13] and Liu Yang [44] on the influencing factors of agricultural carbon emissions, the drivers of carbon emissions are decomposed into four aspects with the following equation:

$$C=\frac{C}{P}\times \frac{P}{A}\times \frac{A}{L}\times L=CEL\times SI\times EDL\times L$$

(3)

where $CEL=\frac{C}{P},SI=\frac{P}{A},EDI=\frac{A}{L}$, C represents the total agricultural carbon emissions, P is the total output value of plantation and livestock, A is the total agricultural output value, and L is the total agricultural labor force. CEL is defined as agricultural production efficiency, SI is defined as agricultural production structure, EDI is defined as agricultural economic development level, and L is defined as labor force size.

The total effect can be calculated according to the first-order difference, as shown in Equation (4):

$$\Delta C=\Delta CEL+\Delta SI+\Delta EDI+\Delta L$$

(4)

where $\Delta CEL,\Delta SI,\Delta EDI,$ and $\Delta L$ denote the changes in carbon emissions caused by changes in a single factor of agricultural production efficiency, agricultural production structure, agricultural economic development level, and agricultural labor size from year t to t + 1, respectively, with other factors held constant.

2.2.4. Decoupling Analysis of Agricultural Carbon Emissions

We constructed a decoupling model to further illustrate the dynamic relationship between agricultural economic growth and agricultural carbon emission pressure. Summarizing previous studies, we found that the Tapio decoupling model can better reflect the decoupling status of the research object and can better measure the pressure on agricultural economic development, the agricultural production situation, and the resulting ecological environment. Therefore, the Tapio decoupling model was chosen in this paper, and the decoupling model between agricultural carbon emissions and agricultural was constructed as follows:

$$e=\frac{\Delta C/C}{\Delta AGDP/AGDP}$$

(5)

where $e$ denotes decoupling elasticity, $C$ denotes agricultural carbon emissions, $\Delta C$ denotes the rate of change in agricultural carbon emissions, AGDP denotes gross agricultural output, and $\Delta AGDP$ denotes the rate of change in gross agricultural output. The classification of decoupling status and evaluation criteria for agricultural carbon emission trends are shown in

Table 1. Decoupling status classification and evaluation criteria of agricultural carbon emission trends.

Status	Decoupling Status	ΔC/C	ΔAGDP/AGDP	Decoupling Elasticity Factor
Negative decoupling	Expansion negative decoupling	>0	>0	e > 1.2
	Strong negative decoupling	>0	<0	e < 0
	Weak negative decoupling	<0	<0	0 ≤ e < 0.8
Decoupling	Weak decoupling	>0	>0	0 ≤ e < 0.8
	Strong decoupling	<0	>0	e < 0
	Recession decoupling	<0	<0	e > 1.2
Connections	Expansion connection	>0	>0	0.8 ≤ e < 1.2
	Recession connection	<0	<0	0.8 ≤ e < 1.2

.

2.2.5. Forecast

The grey prediction model can use a small amount of original data to form a series and then generate a new series through accumulation, which can weaken the randomness of the original data in the process of operation and can reveal obvious characteristic rules. The specific GM (1,1) model is as follows:

Reduce carbon emissions as the original series, and calculate the rank ratio of the original series x⁽⁰⁾:

$${\lambda }^{\left(0\right)}\left(t\right)=\frac{x^{\left(0\right)}\left(t-1\right)}{x^{\left(0\right)}\left(t\right)}$$

(6)

Of which, t = 1, 2, 3, …, n.

If

$${\lambda }^{\left(0\right)}\left(t\right)\in \left(\frac{-2}{e^{\left(n+1\right)}},\frac{2}{e^{n+1}}\right)$$

(7)

then the original series can build the GM (1,1) grey prediction model.

Establish the calculus equation of carbon emissions:

$$\frac{d^{\left(x\right)}\left(1\right)\left(t\right)}{dt}+\alpha x^{\left(1\right)}\left(t\right)$$

(8)

where x⁽¹⁾ (t) is the cumulative series, and the grey parameters a and μ are solved using the least squares method,

$$\left[\begin{array}{c}\alpha \\ \mu \end{array}\right]={\left(B^{-1}B\right)}^{-1}{B}^T{y}_n$$

(9)

and

$$B=\left[\begin{array}{cc}\begin{array}{c}\begin{array}{c}-\frac{1}{2}\left[x^{\left(1\right)}\left(1\right)+x^{\left(1\right)}\left(2\right)\right]\\ -\frac{1}{2}\left[x^{\left(1\right)}\left(2\right)+x^{\left(1\right)}\left(3\right)\right]\end{array}\\ \cdots \\ -\frac{1}{2}\left[x^{\left(1\right)}\left(n-1\right)+x^{\left(1\right)}\left(n\right)\right]\end{array}& \begin{array}{c}\begin{array}{c}1\\ 1\end{array}\\ \dots \\ 1\end{array}\end{array}\right]$$

(10)

$$y_n={\left[x^{\left(0\right)}\left(2\right),x^{\left(0\right)}\left(3\right),\cdots ,x^{\left(0\right)}\left(n\right)\right]}^T$$

(11)

Substituting the grey parameter into the time function yields the following:

$$x^{\left(1\right)}\left(t\right)=\left[x^{\left(0\right)}\left(1\right)-\frac{\mu }{\alpha }\right]+e^{-a\left(t-1\right)}\frac{\mu }{\alpha }$$

(12)

Then, the corresponding carbon emission projection value is as follows:

$$\stackrel{\wedge (0)}{x}(t+1)=\stackrel{\wedge (1)}{x}(t+1)-\stackrel{\wedge (1)}{x}(t)$$

(13)

Residual tests and classical ratio dispersion tests were used to test the accuracy of the model and to forecast the carbon emissions from 2021 to 2025.

3. Analysis of the Evolution Patterns of Agricultural Carbon Emission Intensity and Agricultural Economic Development Levels—Evidence from the 2001–2020 Data of Jiangxi Province

3.1. Time-Series Characteristics of Agricultural Carbon Emissions in Jiangxi Province

3.1.1. Analysis of Total Agricultural Carbon Emissions

According to the trend of agricultural carbon emissions in Jiangxi Province from 2001 to 2020, carbon emissions have been on a growing trend over time but with a low growth rate during the examination period, increasing from 66.74 million tons in 2001 to 71.11 million tons in 2020—an increase of 11.4%—with an average annual growth rate of 0.56%. The overall trend in carbon emissions has changed over this time from a steady rise to a steady reduction; the lowest value of carbon emissions was 63.40 million tons in 2003, and the peak value was 79.28 million tons in 2015. Since then, carbon emissions have been decreasing at an average annual rate of 1.11%, falling back to 75.11 million tons in 2020.

3.1.2. Structural Characteristics of Agricultural Carbon Emissions

The structural characteristics of agricultural carbon emissions depend on the types of carbon sources. Due to the existence of too much data and the limited space of this paper, we selected data from 2001 and 2020 for the following comparison: In

Table 2. Trends in the structure of agricultural carbon emissions in 2001 and 2020.

Indicators	2001		2020
	Emissions (10,000 Tons)	Percentage (%)	Emissions (10,000 Tons)	Percentage (%)
Total carbon emissions	6673.79		7510.95
Agricultural material inputs	1329.62	19.92	1369.60	18.23
Rice cultivation	3358.87	50.33	4082.18	54.35
Farmland utilization	987.56	29.40	1046.21	13.93
Poultry storage	997.74	14.95	1012.95	13.49

, it is easy to see that carbon emissions from agricultural material inputs, rice cultivation, livestock and poultry farming, and farmland use all increased to varying degrees. The largest generator of carbon emissions in 2001 was paddy cultivation, with 33.59 million tons—or 50.32% of the total emissions. Jiangxi is still mostly irrigated with vast amounts of water, and since rice farming requires long-term irrigation water, the anaerobic environment produced by long-term irrigation floods will result in significant CH₄ emissions. Agricultural material inputs were in second, with 13.30 million tons, or 19.92% of the total. Although the use of agricultural inputs can, in certain cases, provide nutrients to crops and decrease disease incidence, their inappropriate usage can also harm the agro-ecological ecosystem and result in significant carbon emissions. Carbon emissions from livestock breeding and farmland use ranked third and fourth, with emissions of 9.98 million tons and 9.88 million tons, accounting for 14.95% and 14.80%, respectively. In 2020, the carbon emissions caused by rice cultivation, agricultural resource input, livestock breeding, and farmland utilization were 40.82 million tons, 13.70 million tons, 10.13 million tons, and 10.46 million tons, respectively, with proportions of 54.34%, 18.23%, 13.49%, and 13.93%, respectively. Carbon emissions from farming exceeded those from livestock. Meanwhile, we found continuous growth of carbon emissions from agricultural material inputs, rice cultivation, livestock and poultry breeding, and farmland use. It is important to note that the issue of “land abandonment” has been resolved as a result of the implementation of high-standard farming and the rise in the transfer of land-use rights in recent years. Moreover, as Jiangxi Province actively responds to the national call for food security in the context of the tense international situation, we have seen a continuously rising scale of rice cultivation, resulting in rapidly increasing carbon emissions. In order to improve the quality of rice agriculture and lower emissions, we must secure national food security based on scientific research efforts to promote the quality of rice cultivation and reduce emissions.

3.1.3. Time-Series Characteristics of Agricultural Carbon Emissions

In Figure 2, we can find inter-year changes in carbon emissions. From 2001 to 2020, agricultural carbon emissions in Jiangxi Province showed a more obvious trend of “continuous climbing—steady declining”, which can be roughly divided into the following four stages: The first stage was the reduction period (2001–2003), in which carbon emissions continued to decrease, and agricultural material inputs and the livestock industry remained stable. A decline in arable land has occurred nearby, and crop cultivation has resulted in a decrease in carbon emissions. On the one hand, the growing income disparity between urban and rural areas brought on by industrialization has dampened farmers’ enthusiasm for production and brought about the emergence of land abandonment. On the other hand, the accelerated urbanization process has taken up a significant amount of arable land.

The second stage was the period of fluctuating climbing (2004–2010), in which carbon emissions generally showed a continuous climbing trend, except for those in 2006 and 2007, when a slight decline in carbon emissions occurred because of changes in the rice-sowing area. In the early part of this period, the national economy was in a stage of rapid development in which several policies were put forward to benefit the agricultural industry, and the demand for agricultural products increased. To obtain continuous economic benefits, farmers expanded the area of cultivated land and the scale of farming, enhancing the level of agricultural production substantially and promoting the rapid growth of agricultural carbon emissions in Jiangxi Province.

The third stage was the smooth transition period (2013–2017), in which the growth rate of carbon emissions slowed down and the average annual growth rate dropped to 0.79%. The main reason for this was that the pure reliance on large amounts of agricultural production materials does not enhance farmers’ income or reduce the input of agricultural production materials; at the same time, the construction of high-standard farmland promotes an improvement in mechanization level, the implementation of strategies to reduce and efficiently use pesticides and chemical fertilizers reduces the use of agricultural materials, and the resourceful and comprehensive utilization of livestock and poultry manure reduces pollution, inhibiting increases in agricultural carbon emissions in Jiangxi Province. We also found that agricultural carbon emissions declined in 2016, while they reached a record high in 2017, due to the transformation of modern agriculture, which stimulated the growth of agricultural carbon emissions in the short term.

The fourth stage was the stable decline period (2018–2020), in which carbon emissions showed a continuous decline. For the most part, China’s economy is moving to the stage of high-quality development, with more emphasis on green and low-carbon development. Although carbon emissions from rice cultivation are increasing, with the progress of agricultural science and technology and the increasing efficiency of agricultural production, the improvement in agricultural mechanization has reduced the use of agricultural materials such as pesticides and fertilizers; at the same time, African swine fever swept the country in 2018, and the scale of livestock and poultry breeding was reduced. Therefore, agricultural carbon emissions achieved negative growth under the combined effects of various factors.

3.2. Measurement of Agricultural Carbon Emission Intensity in Jiangxi Province

Figure 3 shows the trend of agricultural carbon emission intensity in Jiangxi Province from 2001 to 2020. Agricultural carbon emission intensity is a more realistic reflection of the extent of emissions than agricultural carbon emissions [21]. It can be found that with the adjustment and upgrading of the agricultural industry structure in Jiangxi Province, the province’s agricultural emission intensity showed a continuous reduction between 2001 and 2020, from a peak of 11.42 tons of carbon emissions per 10,000 RMB in 2001 to a historical low of 2.67 tons of carbon emissions per 10,000 RMB in 2020, with an average annual decrease of 16.4%. From 2001 to 2020, the carbon emission intensity decreased the most in 2011, falling by 17.56%. In recent years, the carbon emission intensity of agriculture has shown a continuous decreasing trend. This also indicates that the investment in agricultural science and technology has increased in this period, improving the factor input structure; increasing the efficiency of energy and resource utilization; reducing the economic cost of inefficient parts; and thus, reducing the intensity of agricultural carbon emissions.

3.3. Factor Decomposition of Agricultural Carbon Emission Intensity in Jiangxi Province

Table 3. Decomposition results of agricultural carbon emission factors in Jiangxi Province/10⁴ t.

Year	ΔC	ΔCEI	ΔSI	ΔEDI	ΔL
2002	−27.67	−209.73	−49.84	346.62	−114.72
2003	−305.96	−719.22	−28.01	570.96	−129.70
2004	681.24	−1091.60	150.87	1729.18	−107.21
2005	240.75	−271.46	−77.84	711.43	−121.37
2006	−64.44	−277.37	−75.20	424.75	−136.63
2007	−34.82	−1255.37	44.48	1259.84	−83.78
2008	177.61	−1104.63	16.71	1400.53	−135.00
2009	130.63	24.62	−107.42	389.44	−176.01
2010	144.57	−506.70	−58.39	775.43	−65.77
2011	77.96	−1351.12	204.62	1435.20	−210.74
2012	90.41	−379.76	−188.39	888.54	−229.98
2013	69.58	−413.55	−77.74	804.89	−244.02
2014	39.09	−326.11	−62.83	653.64	−225.62
2015	35.05	−362.61	−4.48	673.29	−271.15
2016	−147.11	−891.38	139.54	866.65	−261.92
2017	77.83	101.99	−131.11	369.73	−262.79
2018	−135.30	−215.45	−96.67	464.37	−287.54
2019	−205.65	−1220.14	200.76	1122.92	−309.18
2020	−6.59	−910.23	172.60	977.40	−246.37
Total	837.17	−11,379.82	−28.35	15,864.82	−3619.49

reports the decomposition of each factor of agricultural carbon emissions in Jiangxi. During the examination period, the cumulative agricultural carbon emissions in Jiangxi Province increased by 8.37 million tons, but from 2016 onward, the agricultural carbon emissions began to show a decline. The data show that the level of regional economic development is the main factor in the increase in agricultural carbon emissions; agricultural production efficiency, labor force size, and agricultural structure also play roles in reducing agricultural carbon emissions in Jiangxi Province to a certain extent.

The level of regional economic development is a major factor in the increase in agricultural carbon emissions. The cumulative increase in carbon emissions from 2001 to 2020 was 158.65 million tons. If other factors remain unchanged, economic development can increase carbon emissions by an average of 7.93 million tons per year. It is noteworthy that the carbon increment effect still tended to strengthen in the past two years, indicating that although massive use of agrochemicals and the rapid development of agricultural mechanization driven by the market economy can improve the efficiency of agricultural production and promote the continuous improvement of the agricultural economy, they have also caused large amounts of carbon emissions. However, in the context of rural revitalization, we will continue to vigorously develop the agricultural economy. Meanwhile, intensification, mechanization, and low carbonization are the inevitable trends of agricultural development. Thus, the factor of agricultural economic development level will remain the most important factor leading to the increase in agricultural carbon emissions in Jiangxi Province in the future. Therefore, the sustainable development of agriculture in Jiangxi Province depends on finding ways to reduce agricultural carbon emissions while expanding the agriculture sector.

Agricultural production efficiency is the most important factor in reducing agricultural carbon emissions. From 2001 to 2020, the improvement in agricultural production efficiency has cumulatively reduced emissions by 113.80 million tons and, if other factors remain unchanged, could lead to an average annual reduction of 5.69 million tons. Moreover, the overall high contribution rate of carbon emission reduction continues to rise. The reason for this lies in the continuous innovation of agricultural science and technology under the implementation of rural revitalization strategies in the new era, which has gradually achieved mechanization and green development in the field of agricultural production and suppressed the growth of agricultural carbon emissions in Jiangxi while improving the yield and quality of agricultural production. Therefore, enhancing the efficiency of agricultural production is an important step to promote low-carbon development in Jiangxi Province.

The size of the labor force is also an important factor. The changes in the size of the labor force cumulatively reduced emissions by 36.19 million tons compared with the base period. This is because, in the context of urbanization, the agricultural farming industry in Jiangxi has achieved intensive large-scale operation, less dependence on labor thanks to agricultural modernization and mechanization, and less use of agricultural inputs such as pesticides and fertilizers, reducing agricultural carbon emissions.

The structure of agricultural production has had a positive effect on carbon emission reduction to some extent, cumulatively contributing 283.50 thousand tons of emission reduction compared with the base period. This further demonstrates that, in the context of ensuring national food security, it is more challenging to reduce agricultural carbon emissions by fundamentally altering the structure of agricultural production. Nevertheless, this goal should be actively pursued while remaining within the red line of 1.8 billion mu of arable land, relying on mountain, lake, and river resources to carry out diversified agricultural production activities, which achieves agricultural carbon emission reduction while maintaining agricultural economic development.

3.4. Analysis of Agricultural Carbon Emission and Economic Decoupling State

Table 4. Decoupling characteristics between agricultural carbon emissions and agricultural economic growth in Jiangxi Province, 2001–2020.

Year	ΔC/C	ΔTVFP/TVFP	e	Decoupling Status
2001–2002	−0.0042	0.0266	−0.1568	Strong decoupling
2002–2003	−0.0483	0.0585	−0.8244	Strong decoupling
2003–2004	0.0970	0.2185	0.4440	Weak decoupling
2004–2005	0.0332	0.0680	0.4876	Weak decoupling
2005–2006	−0.0090	0.0285	−0.3143	Strong decoupling
2006–2007	−0.0049	0.1450	−0.0335	Strong decoupling
2007–2008	0.0242	0.1518	0.1594	Weak decoupling
2008–2009	0.0175	0.0142	1.2281	Expansion negative decoupling
2009–2010	0.0190	0.0802	0.2367	Weak decoupling
2010–2011	0.0101	0.1580	0.0641	Weak decoupling
2011–2012	0.0116	0.0576	0.2017	Weak decoupling
2012–2013	0.0089	0.0584	0.1516	Weak decoupling
2013–2014	0.0050	0.0444	0.1115	Weak decoupling
2014–2015	0.0044	0.0480	0.0922	Weak decoupling
2015–2016	−0.0189	0.0858	−0.2203	Strong decoupling
2016–2017	0.0099	−0.0031	−3.1786	Strong negative decoupling
2017–2018	−0.0175	0.0101	−1.7350	Strong decoupling
2018–2019	−0.0274	0.1161	−0.2356	Strong decoupling
2019–2020	−0.0009	0.1073	−0.0082	Strong decoupling

shows that the agricultural economy in Jiangxi Province has maintained continuous positive growth, while agricultural carbon emissions have grown relatively slowly, according to the Tapio decoupling model. Both agricultural carbon emissions and agricultural economic development have shown growth trends over the past 20 years. In general, the decoupling relationship between agricultural carbon emissions and agricultural economic development is mainly characterized by weak decoupling and strong decoupling. To put it differently, the growth rate of the agricultural economy is higher than the growth rate of agricultural carbon emissions, indicating that the work of low-carbon emission reduction in Jiangxi agriculture has achieved some success. Both weak and strong decoupling states occurred nine times, and recent years have been dominated by strong decoupling states, indicating that in recent years the agricultural development model in Jiangxi Province has gradually improved, as evidenced by the mitigation of the imbalance between traditional heavy reliance on investment in agricultural land—e.g., fertilizers and pesticides—and environmental issues. However, the effect of agricultural carbon emission reduction efforts is more limited, and the decoupling status needs to be continuously consolidated and expanded while coping well with the possibility of negative decoupling. The evolution of the decoupling status of agricultural carbon emissions in Jiangxi Province can be divided into the following three stages:

The first stage is the fluctuation stage. From 2001 to 2009, thanks to the rapid development of the country, the agricultural economy of Jiangxi Province maintained continuous growth, with an average annual growth rate of 8.89%—higher than 1.32%, the annual growth rate of agricultural carbon emissions—but in some years, there were “alternating increases and decreases”. Carbon emission decoupling characteristics are concentrated in three decoupling states: weak decoupling, strong decoupling, and expansion negative decoupling, with weak decoupling as the main decoupling state. Between 2001 and 2008, the decoupling status moved from weak to strong decoupling, and from 2008 to 2009, it became expansion negative decoupling. The reason for this fluctuation was that the agricultural development of Jiangxi Province at that stage was more reliant on traditional agricultural inputs such as fertilizers and pesticides. At the same time, the adjustment of unreasonable agricultural restructuring was also a reason for the unstable decoupling status at this stage.

The second stage is the weak decoupling stage. From 2010 to 2015, agricultural economic development maintained a strong development trend, achieving an average annual growth rate of 7.4%. In contrast, agricultural carbon emissions were growing at a slower pace, with an average annual growth rate of only 0.73%. The main decoupling state in this stage is weak decoupling, which is due to the fact that, during this period, the state put forward a number of beneficial agricultural policies for the agricultural industry, which increased the demand for agricultural products, and farmers expanded the area of cultivated land and the scale of farming in order to obtain a constant stream of economic benefits, which enhanced the level of agricultural production; because of the massive use of agricultural inputs such as pesticides and fertilizers, carbon emissions increased; and in the context of rapid economic development, the improvement in the level of agricultural technology has also reduced the intensity of agricultural carbon emissions, thus lowering the growth rate of agricultural carbon emissions below the growth rate of agricultural gross domestic products.

The third stage is the strong decoupling stage. Agricultural economic development from 2016 to 2020 featured a high quality, with an average annual growth rate of 6.32% even after negative growth of 0.3% in 2017. Agricultural carbon emissions achieved a shift from positive to negative growth, decreasing at an average annual rate of 1.1%, and the imbalance between agricultural economic growth and the ecological environment was eased. The characteristics of flexible decoupling in this stage are presented as strong decoupling–strong negative decoupling–strong decoupling. Jiangxi has changed its irrational agricultural industrial structure, increasing standardization and scale of animal husbandry, realizing the comprehensive use of livestock and poultry waste, and gradually reducing the proportion of resource-consuming agricultural enterprises. During this period, the rebound in the use of inputs and materials in the planting and animal husbandry industries also led to a small fluctuation in agricultural carbon emissions.

3.5. Analysis and Forecast of Agricultural Carbon Emission Trends

In this paper, we ran the GM (1,1) model through MATLAB software to forecast the trend of agricultural carbon emissions. Because the accuracy of the GM (1,1) model is poor for medium- and long-term prediction, we shortened the prediction years and used the carbon emissions data from 2015 to 2020 to forecast the agricultural carbon emissions from 2021 to 2025. The inverse conversion of the fitting values of the grey prediction model in this study led to the predicted data of agricultural carbon emissions in Jiangxi Province. When the model shows fitting values within the interval [0.982, 1.0098], the data are suitable for model construction, yielding a development coefficient (a) of 0.0164, a grey action quantity (μ) of 8032.9007, an a posteriori ratio value (C) of 0.0760, and a small error probability value (p) of 1.000; the a posteriori difference ratio value (C) of 0.076 ≤ 0.35 implies a very accurate model.

The greatest relative error of the model, calculated from the fitting line between the anticipated and measured values of agricultural carbon emissions in Jiangxi Province from 2010 to 2020, is 0.009 < 0.1, indicating that the model’s fitting effect meets a high standard and that the forecast is accurate. Figure 4 shows us the projected trends in agricultural carbon emissions in Jiangxi Province from 2021–2025. According to the model, agricultural carbon emissions in Jiangxi Province from 2021 to 2025 show a gradually decreasing trend from 77.81 million tons in 2016 to 68.79 million tons in 2025, with a cumulative decrease of 13.11%. This is due to the fact that Jiangxi Province, aware of the importance of the growth of agricultural carbon emissions to the deterioration of the ecological environment, has promulgated a series of bills on ecological environmental protection and green agricultural development, and the implementation of these policies has facilitated the reduction of agricultural carbon emissions in Jiangxi Province. According to the forecast, Jiangxi’s agricultural carbon emissions will continue to show a declining trend but at a low rate. To successfully attains the objectives of carbon peaking and carbon neutrality, Jiangxi Province will need to step up its efforts to reduce carbon emission in the future.

4. Conclusions and Policy Insights

4.1. Conclusions

In this study, we measured the data of agricultural carbon emissions in Jiangxi Province using the carbon emission factor measurement formula, with agricultural carbon emissions in Jiangxi Province as the research object. We aimed to understand the time-series characteristics and driving factors of agricultural carbon emissions in Jiangxi Province, to further analyze the decoupling characteristics of agricultural carbon emissions and agricultural economic growth through the Tapio decoupling model, to predict the carbon emission trends for the next five years through a GM (1,1) model, and to conduct a more objective assessment of agricultural carbon emissions in Jiangxi Province.

First, the time-series characteristics of Jiangxi’s agricultural carbon emissions show a trend of increasing first and then decreasing from 2001 to 2020, and the rate of decline has accelerated in recent years. The GM (1,1) model predicts that agricultural carbon emissions in Jiangxi Province will continue to decline over the next five years. Generally, rice cultivation ranks first in terms of contribution to increasing emissions; agricultural material inputs, livestock breeding, and farmland use are the second-, third-, and fourth-largest carbon sources, respectively. With the adjustment and upgrading of the agricultural industry structure in Jiangxi Province, the province’s agricultural emission intensity showed a continuous reduction during 2001–2020.

Second, among the drivers of agricultural carbon emissions in Jiangxi Province, the level of economic development is the main factor that increases carbon emissions. Agricultural production efficiency, labor force size, and agricultural production structure also have positive effects on carbon emission reduction to a certain extent. How to promote agricultural economic development while reducing carbon emissions is a question that we need to think about in the coming period.

Third, the analysis of the Tapio decoupling model showed that the decoupling states from 2001 to 2009 shifted between strong and weak decoupling states, with weak decoupling dominating the years 2010–2015 and strong decoupling dominating from 2016 onwards. It is not difficult to find that the agricultural economy of Jiangxi Province has maintained continuous growth, while the growth rate of agricultural carbon emissions has slowed down and fallen in recent years, indicating the effectiveness of low-carbon emission reduction work in Jiangxi agriculture.

4.2. Policy Insights

Based on the above analysis, this paper makes the following recommendations:

Cultivate high-quality and low-carbon rice varieties. Rice cultivation is one of the main sources of agricultural carbon emissions in Jiangxi Province, and carbon emissions increase with the expansion of the cultivation scale. The carbon emission reduction potential of rice cultivation should be fully explored, relying on agricultural research institutes, universities, and leading seed production enterprises to cultivate high-quality and low-carbon rice varieties suitable for Jiangxi, thereby alleviating the pressure of carbon emissions.

Transition to green agricultural production. Agricultural material inputs will remain the focus of agricultural carbon emission reduction for a long time to come. We should focus on modern biotechnology research and development, develop ecological technologies for prevention and control, increase the application of organic fertilizers (e.g., in soil testing and composting), and enhance the resource utilization of agricultural film residues to achieve low-carbon development.

Coordinate the relationship between economic development and agricultural carbon emissions. The level of economic development is the main factor in the increase in carbon emissions but, in the context of rural revitalization, agricultural economic development is an inevitable requirement. Therefore, China should make full use of the advantages of economic development; promote the gathering of advantageous rural industries; increase the research and development of agricultural science and technology; promote the integration and scale of agricultural technology; and realize the efficient, green, and low-carbon development of agriculture.