Study on the driving factors and decoupling effect of carbon emission from pig farming in China—based on LMDI and Tapio model

Abstract

The livestock sector accounts for 18% of total anthropogenic carbon emissions and is an important source of global greenhouse gas (GHG) emissions. China occupies a large proportion of total livestock carbon emissions worldwide, especially in the pig industry, which is significant to China's agricultural economy and also a key area for China to achieve the "Carbon peaking and Carbon neutrality goals." This study uses the life cycle approach to calculate the GHG emission status of China's pig farming industry from 2001 to 2020, and then, we establish a logarithmic mean Divisia index (LMDI) model to identify the main driving factors and a Tapio decoupling model to analyze its decoupling status. We decompose the emission sources as well as decoupling index into five drivers: technological progress, livestock structure, policy bias, affluence, and population. The results reveal that the carbon emission of China's pig industry is in a weak growth trend and overall in a weak decoupling state but has volatility, which is closely related to the "Pig Cycle" in China. Decomposition analysis shows that increasing affluence and population growth are the main drivers of GHG emissions; simultaneously, technological progress, livestock structure, and policy bias are the main drivers of emission reduction. Meanwhile, technological and policy factors positively contribute to the decoupling status, while affluence level, population, and livestock structure changes negatively inhibit the decoupling status. The study concludes that technological advances, optimized economic structures, the guidance of green consumption patterns, and the solution to the "Pig Cycle" problem are crucial to further reduce GHG emissions from China's pig industry; meanwhile, technological changes have a dominant role in promoting carbon decoupling in pig farming.

1 Introduction

1.1 Background

Since the beginning of the twenty-first century, the issue of global warming has escalated due to extensive greenhouse gas emissions. In recent times, the conflict between environmental concerns and economic growth has become more pronounced, particularly in China, where rapid development has positioned it as the world's largest emitter of greenhouse gases (Cai et al., 2019). In response to this challenge, the Chinese government has taken significant steps to address the issue. During the "Paris Agreement" in 2016, China pledged to implement a series of carbon emission reduction targets and policies, while repeatedly committing to strive to peak carbon emissions by 2030 and achieve carbon neutrality by 2060.

The significant increase in carbon emissions is mainly due to human activities, as well as the growing global population and meat-driven demand (Han et al., 2019); hence, the livestock industry is a significant source of GHG emissions. According to the survey data of the Food and Agriculture Organization of the United Nations (FAO), GHG emissions caused by livestock production activities already account for about 18% of global GHG emissions from human activities, with the pig industry alone responsible for 5% of total global emissions. And due to the influence of Chinese consumption habits, it is decided that the pig industry is the pillar industry of China. From the supply side, since 2013, China has been accounting for about 50% of global pig production and ranked first in pork production since 1991 (Wang et al., 2017). On the consumption side, the rising economic level over the past two decades has led to increasing demand for livestock products such as meat, eggs, and milk. Except for the impact of special factors such as swine fever in 2018, pork has occupied the top spot in China's meat consumption structure for many years. The growth in pork consumption has strongly contributed to the rapid development of China's rural pig farming industry, so pig farming is both an important source of carbon emissions and a material part of the agricultural economy in China. However, the development of the pig industry and its low-carbon goals have failed to achieve an effective balance. But the increase in pig supply caused by consumption growth also exacerbates GHG emissions and water pollution to a certain extent (Schiffman & Williams, 2005). As a representative industry for high-quality transformation and low-carbon development in China, the pig farming industry, under the trend of continuous scaling up, cannot be ignored for the environmental pollution it produces throughout the farming process. In the meantime, compared with foreign countries, China's pig breeding is scattered and small-scale, and the integration system of agriculture and animal husbandry is not perfect, which leads to a large amount of waste produced in the farming process, often with a high risk of the environment (Li & Liu, 2020).

To effectively reduce GHG emissions in China's livestock sector and promote sustainable development, it is crucial to quantify the emissions from the pig farming industry and investigate the underlying drivers and green development status. Therefore, this research aims to address these issues through a systematic approach. On the one hand, a robust scientific index system will be established to accurately measure GHG emissions. On the other hand, the LMDI model will be employed to analyze the factors influencing changes in GHG emissions within the pig industry. Additionally, the relationship between GHG emissions and industrial development will be analyzed using the decoupling model and decomposition method. Based on the findings from these analyses, targeted strategies will be formulated to enhance GHG emissions management in the pig farming industry, leading to improved environmental sustainability.

2 Literature review

2.1 Review of carbon emission measurement methods for livestock

At present, the calculation methods for carbon emissions from the livestock industry mainly include the emission factor method, life cycle method, and input–output method. The emission factor method mainly accounts for the processes of forage conversion, gastrointestinal fermentation, and manure inside and outside the barn. The life cycle method accounts for the GHG emissions of the whole process of livestock and livestock products production, processing, transportation, and consumption. But the input–output method uses the input–output table to reflect the linkage of carbon emissions of each industry, analyzes both the direct and indirect energy demand of each upstream production stage of livestock production through the Leontief inverse matrix, and projects carbon emissions through energy carbon emission factors.

The emission factor method can be divided into the OECD method and the IPCC method in chronological order. The OECD method was proposed by the Organization for Economic Cooperation and Development (OECD) in 1991 for estimating methane emissions from ruminants raised. Meanwhile, Blaxter (1965) proposed estimating methane emissions based on the proportion of energy converted into methane by animals feeding, but it is rarely used by researchers. However, Dong et al. (1995) used a modified OECD equation to estimate methane emissions from ruminants from 1980 to 1990 for specific situations in China. Meanwhile, the OECD method was gradually replaced by other methods after Lal (2004) argued that it is necessary to include the input of production materials, such as chemical fertilizers, pesticides, irrigation, agricultural film, crop cultivation, and other factors in the agricultural carbon emissions process. The IPCC coefficient method surpasses the OECD method slightly in accounting scope. According to the IPCC Guidelines, for each emission source, it constructs emission factors and records activity data. The product of these two components is utilized as the estimated value for the carbon emission of the emission project. This method is currently the most widely used accounting approach, and numerous scholars have applied it to measure the carbon emissions from animal husbandry farming and livestock products. Ominski et al. (2011) and Cerri et al (2016) used the IPCC factor method to estimate CH4 emissions per head of beef cattle in Canada and annual GHG emissions from large beef cattle farms in Brazil, respectively. Although the emission factor method is simple to use, Lesschen et al. (2011) argue that it has mature accounting formulas, emission factor tables, and many examples for reference and it is less adaptable when faced with changes in the emission system.

The life cycle assessment (LCA) method of emission is a factor method to evaluate the environmental impact of activities, services, processes, and related products within the life cycle of an economic activity or an industry after defining the boundary of the measurement system. Two delineated boundary approaches to the livestock industry were achieved, namely "cradle-to-farm" and "cradle-to-consumer," where the former was mostly employed by researchers. Important conclusions have been obtained through this method to measure carbon emissions from the livestock industry. The existence of regional differences in carbon emissions from livestock between developed and developing countries was established (Luo et al., 2015; Six et al., 2017; Summary et al., 2012; Vergé et al., 2016). Also, it is believed that the scale of farming has an important impact on the carbon emissions of livestock. Luo et al. (2015) established that the carbon footprint of the free-range model is 3.7, 20, and 5.42 kg of CO₂ per kg of eggs, chicken, and pork, compared to 3.5, 7.9, and 4.3 kg CO₂ for large-scale farms, respectively. This shows that large-scale farming can effectively reduce the carbon emissions of livestock products. Furthermore, The LCA method can concretely measure the carbon emission contribution of each process in the life cycle and the carbon pollution of each livestock product. Zervas and Tsiplakou (2012) estimated that the carbon footprint of each livestock product is from smallest to largest in order of eggs, poultry meat, pork, lamb, and beef using the life cycle assessment method. Daneshi et al. (2014) found that gastrointestinal fermentation, energy consumption, manure discharge, and transportation ranked in descending order according to their contribution to the carbon footprint of milk.

The input–output (I–O) method is a new proposed GHG measurement method in the twenty-first century, but its effect depends on the country and its industry’s statistics. Su et al (2017) used the I–O method to analyze Singapore's urban carbon emissions from the perspective of demand and stated that exports accounted for nearly two-thirds of its total emissions, and the growth of its emissions in the past decade was largely export-driven. Wen and Zhang (2020) also used I–O tables, combined with structural path analysis (SPA) and multidimensional analysis framework (MAF), to analyze the current situation of carbon emission transfer between sectors.

2.2 Review on carbon emissions and economic development

Studies on economic development and GHG emissions have mainly focused on the relationship between the two, namely the environmental Kuznets curve (EKC) and decoupling effects, drawing different conclusions about their relationship. Zhao et al. (2011) concluded that there exists a long-term equilibrium relationship between economic development and GHG emissions, Tian and Zhang (2013) argued that GHG emissions and economic growth are reciprocal causation, while Ghosh (2010), using India as an example, concluded that there is neither a long-term equilibrium relationship nor a long-term causality relationship between the two, but a two-way relationship in the short term. Therefore, human economic activities may be the main factor contributing to the increase of carbon emissions on Earth.

EKC was proposed by Kuznets to show the inverted U-shaped relationship between economic growth and income inequality and then was used in the context of environmental quality to analyze the relationship between carbon emissions and economic growth. Xia and Wang (2020) tested the validity of the environmental Kuznets curve (EKC) hypothesis in China and concluded that EKC considering structural disruption exists in China. In addition, the decoupling model is also widely used because it can measure the relationship between economic development and environmental pollution more scientifically. Zhao et al. (2022) used decoupling models to explore the relationship between economic development and GHG emissions in China, and the results showed that carbon emissions and economic development in China basically were weak decoupling from 2009 to 2019. Weng et al. (2021) concluded that it generally showed weak decoupling characteristics in China's tourism industry. So as to better reflect the contribution of carbon emission drivers in the decoupling process, many scholars have started to combine decomposition methods and decoupling models (Wu et al., 2018), which can reflect not only the impact of emission factors but also their influence on the decoupling process. Li and Qin (2019) used this method to study the decoupling effect between energy GHG emissions and economic growth in China.

2.3 Literature review of carbon emission drivers research

Four approaches have been proposed to establish the driving factors of agricultural carbon emissions. The IPAT model, introduced in the 1970s, considers population, affluence level, and technology level as the main causes of human impact on nature. The Kaya constant equation, proposed by Japanese scholar Kaya(1989), employs exponential decomposition analysis to break down carbon emissions into four drivers: energy carbon intensity, unit GDP energy intensity, GDP per capita, and population. The STIRPAT model builds upon the IPAT model and introduces dimensionless variables, adjusting the coefficients and constant terms compared to the IPAT model. The LMDI model, based on Ang's work in (2004), is the most widely used method for studying carbon emission drivers due to its simplicity and ability to measure the contribution of influencing factors in the process of carbon emission changes. Quan et al. (2020) applies the LMDI model to decompose carbon emission factors in the logistics industry, including carbon emission factors, energy intensity, energy structure, economic level, and population size. The study concludes that economic growth is the primary factor driving carbon emissions in the logistics industry. Similarly, Gong and Song (2015) and Jeong and Kim (2013) utilize LMDI decomposition models to analyze carbon emission drivers in the construction and manufacturing industries, respectively.

Previous literature has predominantly focused on estimating greenhouse gas (GHG) emissions from the livestock industry using the IPCC factor method and LCA method. While these studies have provided mature conclusions regarding the influencing factors of carbon emissions and the decoupling effect to some extent, they still possess certain deficiencies. For instance, carbon emissions from crucial stages of the pig industry's life cycle, such as feed planting and transportation, have been largely overlooked. Additionally, there is a lack of research on carbon emissions across the entire pig industry and a limited emphasis on the national level, with most studies focusing on specific regions or provinces. Moreover, when it comes to investigating the driving factors and decoupling effects of carbon emissions in the pig farming industry, the literature is scarce. Many studies on carbon emissions in China's pig industry fail to consider the significant impact of the "Pig Cycle" phenomenon, which plays a crucial role in the changes of carbon emissions and decoupling status. Furthermore, there is a lack of relevant research on the decoupling effect specific to China's pig industry. While numerous studies have determined the decoupling status in specific years, there is a dearth of further decomposition and research on decoupling elasticity, thus hindering the identification of factors influencing the decoupling status of the pig industry. This article aims to address the aforementioned research gaps and provide additional insights in these areas.

2.4 Innovation points and contributions

This paper presents a study on the measurement of carbon emissions, the analysis of drivers, and the decoupling effects in the specific context of pig farming. The article contributes to the field of carbon emission research in several innovative ways. First, it extends the IPAT analytical framework by incorporating the Kaya constant equation, which serves as the basis for studying the drivers of carbon emissions in the pig industry, considering the specificity of the Chinese pig industry. Second, unlike previous studies, this paper combines driver decomposition analysis and the decoupling index to explore the individual impact of each driver on the decoupling process, aiming to effectively decouple carbon emissions from pig farming. Furthermore, it investigates whether the "Pig Cycle" exhibits cyclical effects on carbon emissions and the decoupling status within the pig farming industry. These efforts contribute to the theoretical research on emission reduction in the farming industry. Moreover, this paper modifies certain aspects of the carbon emission measurement methodology. For instance, instead of using annual averages, data from 2001 to 2020 are included, incorporating feed consumption unit prices, electricity costs, coal costs, and electricity prices. The calculation of feed carbon emissions is also revised to utilize direct feed consumption from the pig-raising process, rather than relying on livestock product data. These adjustments improve the accuracy of the measurements and bring them closer to the true values.

3 Materials and methods

3.1 Data source and processing

Considering the availability of data, this paper takes 31 provinces in China as the national research subjects, excluding Hong Kong, Macao, and Taiwan. Data are obtained from the "China Rural Statistical Yearbook," "China Statistical Yearbook," "China Animal Husbandry and Veterinary Yearbook," "China Livestock Yearbook," "National Agricultural Product Cost–benefit Data Compilation," "China Energy Statistical Yearbook," and "reports published by IPCC and the Wind database." Considering that the output values calculated in real prices are not vertically comparable, the GDP comparable prices are used to adjust the output value data. Data organization and analysis processing are performed using Excel and Stata, while the creation of figures is done with ArcGIS, Visio, and GraphPad Prism. The research methodology is presented in Fig. 1, illustrating the flowchart of the research method.

Fig. 1

Research methodology flowchart

3.2 Geographical facts and figure

Since the twentieth century, China has experienced overall growth in pig slaughter volume. However, in recent years, due to the acceleration of the national ecological civilization construction process and the impact of African swine fever, the requirements for green development in pig breeding have increased. This has led to a brief decline in pig breeding volume, with the slaughter volume dropping to 527 million pigs in 2020, the lowest in the past 20 years, compared to 549 million pigs in 2001. The highest slaughter volume was in 2014, reaching 735 million head, as shown in Fig. 2. China is not only a largely agricultural country but also a significant player in the animal husbandry sector. The government has always prioritized support for the animal husbandry industry and has implemented various supportive policies. However, in recent years, the development of pig farming has faced challenges due to African swine fever and the "Pig Cycle." Although the price of pigs has increased, the cost of pig farming has also risen. As a result, many pig farmers in China are facing difficulties. Meanwhile, pig farming enterprises, rural cooperatives engaged in pig farming, and other similar entities have emerged. However, with the increasing intensity of environmental regulations and the promotion of large-scale pig farming, the number of pig farmers in China is gradually decreasing. Additionally, the farming model is transitioning toward scale, intensification, and factory-based operations.

Fig. 2

Pig slaughter volume from China, 2001–2020

In terms of industrial layout, China's pig farming is mainly distributed in the central and eastern regions, but with the development of the breeding industry, the distribution of pig breeding is gradually shifting from the southeast to the northeast, northwest, and southwest regions, and the regional layout of pig farming is continuously adjusted and optimized. In addition, due to the high percentage of meat consumption in China, the delivery range of live pigs and pork is constrained by freshness requirements and transportation costs, and regionalized production layouts are formed around the country. After taking into account factors such as environmental carrying capacity, resource endowment, consumption preference and slaughtering and processing, and giving full play to regional comparative advantages, the layout of China's pig industry has been implemented to classify and promote key development areas (Hebei, Shandong, Henan, Chongqing, Guangxi, Sichuan, Hainan), constrained development areas (Beijing, Tianjin, Shanghai, Jiangsu, Zhejiang, Fujian, Anhui, Jiangxi, Hubei, Hunan, Guangdong), potential growth areas (Liaoning, Jilin, Heilongjiang, Inner Mongolia, Yunnan, Guizhou), and moderate development areas (Shanxi, Shaanxi, Gansu, Xinjiang, Tibet, Qinghai, Ningxia) to promote the coordinated development of live pig production. The regional layout of the industry is shown in Fig. 3.

Fig. 3

Regional layout of China's pig industry

3.3 Carbon emission measurement methods

This paper draws on the FAO Greenhouse Gas Emissions Assessment Framework—“IPCC 2006 Guidelines for International Greenhouse Gas Inventories,” referring to the research methods of Hu and Wang (2010) and Li (2022), and citing the research results published by authoritative organizations such as IPCC and FAO and experts and scholars in related fields, this paper uses the LCA method from cradle to farm a basis to measure carbon emissions, covering five links of feed grain planting, feed grain transportation and processing, intestinal fermentation, manure management system and energy consumption in pig breeding process. Due to the inconsistent estimation method of Chinese pig breeding slaughter data before and after 2000, we only measured the carbon emission equivalents from 2001 to 2020, and the GHG emissions of each link are measured as follows:

• 1. Emissions from the conversion of CH₄ to CO₂ produced by gastrointestinal fermentation in livestock

  $$ {\text{C}}_{{\text{gt}}} = {\text{APP}} \times {\text{ef}}_1 \times {\text{GWP}}_{{\text{CH}}4} $$

  (1)

  where APP is the average annual feeding quantity of pigs, i.e., slaughter quantity × feeding cycle/365 (Hu 2010), ${\text{C}}_{{\text{gt}}}$ is the CO₂ emission from gastrointestinal fermentation in pig farming process, ${\text{ ef}}_1$ is the CH₄ emission factor of gastrointestinal fermentation in pigs, and ${\text{GWP}}_{{\text{CH4}}}$ is the CH₄ global warming value. All parameters and coefficients mentioned above are shown in Table 1.

  Table 1 Emission factor table

• 2. Emissions from the conversion of CH₄ to CO₂ from the manure management system amount

  $$ {\text{C}}_{{\text{mc}}} = {\text{APP}} \times {\text{ef}}_2 \times {\text{GWP}}_{{\text{CH}}4} $$

  (2)

  where ${\text{C}}_{{\text{mc}}}$ is the CO₂ emission of manure management system during pig farming process, and ${\text{ef}}_2$ is the CH₄ emission factor of pig manure management system.

• 3. N2O from manure management system is converted to CO₂ emissions amount

  $$ {\text{C}}_{{\text{md}}} = {\text{APP}} \times {\text{ef}}_3 \times {\text{GWP}}_{{\text{NO}}2} $$

  (3)

  where ${\text{C}}_{{\text{md}}}$ is the CO₂ emission of manure management system during pig farming, ${\text{ef}}_3$ is the pig manure management system N₂O emission factor, and ${\text{GWP}}_{{\text{NO2}}}$ is the NO₂ global warming value.

• 4. CO₂ emissions from livestock feeding CO₂ emissions amount

  $$ {\text{C}}_{{\text{me}}} = {\text{APP}} \times \frac{{\cos t_{\text{e}} }}{{{\text{price}}_{\text{e}} }} \times {\text{ef}}_{\text{e}} + {\text{APP}} \times \frac{{\cos t_{\text{c}} }}{{\cos t_{\text{c}} }} \times {\text{ef}}_{\text{c}} $$

  (4)

  where ${\text{C}}_{{\text{me}}}$ is the CO₂ emission from energy demand in the pig rearing process, $\cos t_{\text{e}}$ is the electricity expenditure per pig in the breeding process, ${\text{price}}_{\text{e}}$ is the unit price of electricity used for breeding, $\cos t_{\text{c}}$ is the expenditure of coal per pig in the farming process, ${\text{price}}_{\text{c}}$ is the unit price of coal used in farming, which is mostly used for heating, but there is no uniform price for heating coal, which is estimated at 800 yuan/t, ${\text{ef}}_{\text{e}}$ is the CO₂ emission factor of electricity consumption, and ${\text{ef}}_{\text{c}}$ is the CO₂ emission factor of coal consumption.

• 5. CO₂ emissions from fodder grain cultivation

  $$ {\text{C}}_{{\text{fe}}} {\rm{ = }}\sum {{\text{APP}}} \times f \times t_i \times {\text{ef}}_{i1} $$

  (5)

  where ${\text{ C}}_{{\text{fe}}}$ is the CO₂ emission from pig feed grain production and cultivation, $f$ is the average annual feed consumption per pig, $t_i$ is the proportion of each grain in the feed, and since there is no standard pig feed formulation under actual production conditions, the pig feed inventory in this study is analyzed according to the common pig feed formulation provided by Chen and Yang (2013), which is shown in Table 2. The soybean cake in the feed is a by-product obtained from soybeans after the first treatment, so the GHG emissions from soybean cultivation are excluded from the calculation. ${\text{ef}}_{i1}$ is the emission factor of the ith grain at the planting stage (Table 3).

  Table 2 Feed major composition ratio

  Table 3 Decoupling status table

• 6. CO₂ emissions from transportation and processing of feed grain

  $$ {\text{C}}_{{\text{gp}}} {\rm{ = }}\sum {{\text{APP}}} \times f \times t_i \times {\text{ef}}_{i2} $$

  (6)

  where ${\text{ C}}_{{\text{gp}}}$ is the CO₂ emission from transportation and processing of feed grains, $f$ is the average annual feed consumption per pig, ${ }t_i$ is the proportion of each grain in the feed, and ${\text{ ef}}_{i2}$ is the CO₂ equivalent emission factor of transportation and processing f grain type i.

• 7. Total carbon emissions

  $$ C_{{\text{tot}}} {\rm{ = }}C_{{\text{gt}}} {\rm{ + }}C_{{\text{mc}}} {\rm{ + }}C_{{\text{md}}} {\rm{ + }}C_{{\text{me}}} {\rm{ + }}C_{{\text{fe}}} {\rm{ + }}C_{{\text{gp}}} $$

  (7)

  where $C_{{\text{tot}}}$ represents the total carbon emissions. The whole calculation flowchart is shown in Fig. 4.

  Fig. 4

  

  Carbon emissions measurement flowchart

3.4 Tapio decoupling model

The decoupling coefficient is frequently employed to elucidate the connection between changes in carbon emissions and economic growth, in either single or multiple regions. The Tapio model offers several advantages, including the flexibility to select a suitable base period, stable coefficient results, and clear judgment criteria. Additionally, it is advantageous in mitigating errors caused by base period selection. Therefore, this paper also adopts the Tapio decoupling model to examine the relationship between carbon emissions and the industrial economy of the pig farming industry. The decoupling elasticity index is calculated as follows:

$$e(C,G){\rm{ = }}\frac{{{\raise0.7ex\hbox{${\Delta C}$} \!\mathord{\left/ {\vphantom {{\Delta C} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}}$$

(8)

where e(C, G) represents the decoupling elasticity index of carbon emission of pig farming and its industrial growth, ΔC represents the change of carbon emission in the study years (million t), C represents the carbon emission in the study base period (million t), ΔG represents the change of pig farming output value in the study years (billion yuan), and G is the economic output value of pig farming in the study base period (billion yuan). According to the different values of e, the decoupling states can be divided into the following eight categories according to the correspondence.

3.5 LMDI emission factor decomposition model

The Tapio decoupling model is limited to exploring the simultaneous relationship between economic development and carbon emissions, but it cannot explain the change in carbon emissions. Presently, academics mainly explore the drivers of carbon emission change through factor decomposition methods, including Kaya constant equation, STIRPAT model, and LMDI model. In comparison, the LMDI model has the advantages of eliminating residual terms, more accurate conclusions, deeper theoretical foundations, wider applicability, and fitting the time-series decomposition, which makes it the most widely used method for carbon emission factor decomposition. Therefore, in this paper, we also use the LMDI model to factorize the carbon emissions of pig farming. Specifically, based on the decomposition of carbon emission drivers studied by Dai (2021) and other scholars, this paper intends to formulate a carbon emission decomposition formula for the pig farming industry from five dimensions, including technological progress, livestock structure, industrial structure, income level, and population growth, specific as follows:

$${\text{C}}{\rm{ = }}\frac{{\text{C}}}{{{\text{GDP}}{\,}_{{\text{pig}}}}} \times \frac{{{\text{GDP}}_{{\text{pig}}} }}{{{\text{GDP}}_{{\text{ls}}} }} \times \frac{{{\text{GDP}}{\,}_{{\text{ls}}}}}{{{\text{GDP}}}} \times \frac{{{\text{GPD}}}}{{\text{P}}} \times {\text{P}}$$

(9)

$$\Delta C = \Delta C_T - \Delta C_0 = \Delta T + \Delta S1 + \Delta S2 + \Delta I + \Delta P$$

(10)

T measured as $\frac{{\text{C}}}{{{\text{GDP}}_{{\text{PIG}}} }}$, where ${\text{C}}$ is the carbon emissions from pig farming industry and ${\text{GDP}}_{{\text{PIG}}}$ is the GDP of pig industry, represents the CO₂ emission per unit GDP of the pig industry, and is a comprehensive representation of the technical level of the pig industry, which includes technical aspects such as the construction of energy-saving facilities, animal disease control, the degree of scale, manure management technology, breeding technology, among other technological aspects.

S1: measured as $\frac{{{\text{GDP}}_{{\text{PIG}}} }}{{{\text{GDP}}_{{\text{LS}}} }}$, where ${\text{GDP}}_{{\text{LS}}}$ is the GDP of livestock sector, reflects the structure of the livestock industry, which to some extent can reflect the level of substitution between pig farming and other livestock farming, as well as changes in the structure of meat consumption in China.

S2: $\frac{{{\text{GDP}}_{{\text{LS}}} }}{{{\text{GDP}}}}$ is the proportion of livestock output in total GDP of China, reflecting the position of the livestock industry in the whole national industry. The amount of structural change can reflect the relevant policy intervention and policy preference for industrial demand to a certain extent.

I: It is the per capita income, reflecting the level of affluence. People tend to adjust their consumption behavior according to changes in their living standards and incomes. In general, as income increased, people tend to increase their intake of foods containing high amounts of protein, such as meat, which in turn affects the demand and supply of pork, thereby affecting GHG emissions.

P: It is the total population, which can reflect the changes in pig demand due to population change to a certain extent.

In this paper, we use the LMDI additive decomposition to decompose the factors and the expression is as follows:

$$ \Delta {\text{CO2}}_{_T } = \frac{{{\text{CO2}}^t - {\text{CO2}}^0 }}{{\ln {\text{CO2}}^t - \ln {\text{CO2}}^0 }} \times \ln \frac{T^t }{{T^0 }} $$

(11)

$$ \Delta {\text{CO2}}_{_{S1} } = \frac{{{\text{CO2}}^t - {\text{CO2}}^0 }}{{\ln {\text{CO2}}^t - \ln {\text{CO2}}^0 }} \times \ln \frac{S1^t }{{S1^0 }} $$

(12)

$$ \Delta {\text{CO}}2_{S1} = \frac{{{\text{CO}}2^t - {\text{CO}}2^0 }}{{\ln {\text{CO}}2^t - \ln {\text{CO}}2^0 }} \times \ln \frac{S2^t }{{S2^0 }} $$

(13)

$$ \Delta {\text{CO}}2_I = \frac{{{\text{CO}}2^t - {\text{CO}}2^0 }}{{\ln {\text{CO}}2^t - \ln {\text{CO}}2^0 }} \times \ln \frac{I^t }{{I^0 }} $$

(14)

$$ \Delta {\text{CO}}2_P = \frac{{{\text{CO}}2^t - {\text{CO}}2^0 }}{{\ln {\text{CO}}2^t - \ln {\text{CO}}2^0 }} \times \ln \frac{P^t }{{p{\,}^{^0 }}} $$

(15)

T, S1, S2, I, and P in the above equation have the same meaning as in Eq. (10). The superscripts 0 and t represent the start and end years of the study period, respectively.

Some of the literature on carbon emission decoupling, based on the measurement of decoupling status, further decomposes and analyzes the drivers of carbon emission decoupling, mostly using a combination of LMDI decomposition and decoupling index. LMDI factor decomposition can identify the influencing factors of decoupling and quantify the degree of contribution of them. Therefore, combined with the above decomposition of emission factors, this paper decomposes the decoupling index into the following form, which can obtain five drivers affecting the decoupling index: technological progress effect, livestock industry effect, policy effect, affluence effect, and population effect, so the specific decomposition form is as follows:

$$E_{C,G} = \frac{{{\raise0.7ex\hbox{${\Delta C}$} \!\mathord{\left/ {\vphantom {{\Delta C} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}} = \frac{{{\raise0.7ex\hbox{${\Delta T}$} \!\mathord{\left/ {\vphantom {{\Delta T} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}} + \frac{{{\raise0.7ex\hbox{${\Delta S1}$} \!\mathord{\left/ {\vphantom {{\Delta S1} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}} + \frac{{{\raise0.7ex\hbox{${\Delta S2}$} \!\mathord{\left/ {\vphantom {{\Delta S2} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}} + \frac{{{\raise0.7ex\hbox{${\Delta I}$} \!\mathord{\left/ {\vphantom {{\Delta I} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}} + \frac{{{\raise0.7ex\hbox{${\Delta P}$} \!\mathord{\left/ {\vphantom {{\Delta P} C}}\right.\kern-0pt}\!\lower0.7ex\hbox{$C$}}}}{{{\raise0.7ex\hbox{${\Delta G}$} \!\mathord{\left/ {\vphantom {{\Delta G} G}}\right.\kern-0pt}\!\lower0.7ex\hbox{$G$}}}} = E_{T,G} + E_{S1,G} + E_{S2,G} + E_{I,G} + E_{P,G}$$

(16)

4 Results and discussion

4.1 Analysis of carbon emissions

The carbon emissions of China's pig farming industry were calculated using Eqs. (1)–(7) (refer to Figs. 5 and 6 and Table 4). As depicted in Fig. 5, the total carbon emissions and carbon emissions of China's pig farming industry between 2001 and 2020 exhibit an overall increasing trend, albeit with a relatively small growth rate, averaging less than 1% annually. Specifically, the changes in total carbon emissions within the pig farming industry over the past 20 years demonstrate a pattern of "continuous growth–sudden decline–continuous growth again–decline again." This pattern can be divided into four stages:

• 1. From 2001 to 2006, the GHG emissions increased from 179.33 million tons of CO₂ equivalent to 221.53 million tons of CO₂ equivalent, with an average annual growth rate of 5%. The reasons for that sharp increase of emissions in this phase may be: first, 2000–2006 was in the phase of restructuring of China's pig industry. During this period, China's pig industry had lower efficiency of feeding volume, higher cost, and faced greater food safety and environmental problems. So, governments subsidized more to promote high quality and scale of breeding, causing large-scale pig breeding enterprises to expand significantly, leading to a significant increase in the quantity of pig breeding. Secondly, China's accession to WTO, further opening to the outside world, and more frequent import and export trades have further improved China's position in the world pig industry. Cooperation among livestock breeding enterprises has been strengthened, indirectly promoting the development of China's pig industry.

• 2. The year 2007 showed a sudden decline phase, with carbon emissions dropping from 221.53 million tons in 2006 to 187.84 million tons, showing a 15% decrease. This can be attributed to the impact of blue ear disease from 2006 to 2007, leading to a significant decline in the slaughter of pigs.

• 3. However, from 2008 to 2018, the carbon emission rate continued to rise, with an average annual growth rate of approximately 2%. During this period, China's pig industry experienced large-scale development, accompanied by rapid social and economic progress. The consumption of pork increased steadily as the pig industry expanded, resulting in a continuous growth in the number of slaughters during this period. Additionally, the Chinese government actively promoted the standardization and scale construction of animal husbandry, implementing a series of new environmental protection policies to guide the livestock industry toward green, low-carbon development. As a result, the carbon emission during this period of pig farming witnessed a slower growth stage.

• 4. In 2019, the rate of carbon emissions once again entered a declining stage, with a drop from 252.21 million tons equivalent in 2018 to 199.56 million tons equivalent. This period was mainly influenced by the outbreak of African swine fever, resulting in a significant decrease in pig slaughter. However, as the impact of swine fever dissipates, the slaughter volume will rise and return to the equilibrium point, influenced by both sides of the pork supply. Based on currently available data, the pig slaughter volume in 2021 is 10% higher than that of 2020, and the pig slaughter volume in the first half of 2022 has already reached 54.5% of the total for 2021. The predictable increase in pig slaughter volume will consequently lead to a rise in carbon emissions from the pig industry.

Fig. 5

CO₂ emissions from pig farming, 2001–2020

Fig. 6

CO₂ emissions from pig farming industry from 2001–2020 (million tons)

Table 4 Carbon emissions from pig farming industry from 2001 to 2020 (million tons)

From the perspective of specific emission links, the two links of feed grain cultivation and manure management system in the pig breeding process are the most important sources, so the focus of GHG emission reduction in the pig industry should be on feed consumption as well as manure treatment links. As shown in Table 4, the average carbon emissions of each link from 2001 to 2020 are in descending order: feed grain cultivation link (56%), manure management link (37%), gastrointestinal fermentation link (3%), feeding energy consumption link (1.7%), and feed grain transportation and processing link (1.2%). Specifically, from 2001 to 2020, the average carbon emission of the feed grain segment of pig farming was 123.83 million tons of CO₂ equivalent, the manure management segment was 83.56 million tons of CO₂ equivalent, and enteric fermentation, feeding energy consumption and feed grain transportation and processing segment reached 7.37, 3.89 and 2.47 million tons of CO₂ equivalent per year, respectively. Among them, the proportion of carbon emissions from the feed grain growing segment and feed grain transportation and processing segment showed an upward trend, while the proportion of gastrointestinal fermentation, manure management and energy consumption segment showed a downward trend. The reason for the increasing proportion of carbon emissions in the feed sector may be due to the continuous reduction of feed costs and the development of the pig industry, resulting in an increasing amount of feed per pig.

4.2 Analysis of the state of decoupling in pig farming

Table 5 shows the decoupling elasticity coefficient of China's pig farming industry from 2005 to 2020, which reflects its industrial development and decoupling status in the past 15 years. The data before 2005 are not included because the comparable output statistics of China's livestock industry were not consistent around 2004. From the table, the output value of the pig industry is in a state of rising trend and fluctuating in a small range, which is mainly influenced by the "Pig Cycle" in China. The fluctuation of the pig price leads to a significant change in the output value in the adjacent years. At the same time, the carbon emission is also affected by the change in feeding amount caused by price changes, but the change range is relatively small (Table 6).

Table 5 Decoupling elastic results

Table 6 Decoupling elastic results (GDP not adjusted)

The decoupling coefficient of the pig farming industry from 2005 to 2020 has exhibited fluctuations, making it challenging to maintain a consistent state over the long term. When considering a 4-year interval as a stage, the overall performance predominantly reflects a negative decoupling state. This means that as China's pig farming industry experiences economic growth, carbon emissions also increase, albeit at a slower rate compared to the growth of the industrial economy. Consequently, it fails to achieve an effective decoupling state. Some years exhibit a strong negative decoupling, while effective strong decoupling states were only achieved in 2016 and 2020. Moreover, the rate of carbon emission growth changes more rapidly during non-decoupled years and exhibits a smaller or negative growth during decoupled years. In the long run, maintaining this decoupling state proves challenging, indicating the absence of genuine decoupling between carbon emissions and pig farming output. Therefore, the key to decoupling lies in effectively controlling carbon emissions. Further observation of the fluctuation in the decoupling state reveals a cyclical pattern over the past 20 years. The emergence of a strong decoupling state is not sustainable and is likely influenced by significant fluctuations in pig prices. Specifically, it indicates a scenario where pig breeding quantities are low, but prices are high. Notably, a detailed comparison of the distribution of the "Pig Cycle" in China demonstrates a significant overlap between the two fluctuation periods. When GDP is taken into account for comparison, the two fluctuation time intervals nearly completely overlap.

Therefore, a gray correlation analysis of the decoupling elasticity index, GDP growth rate, pork price, pig production value growth rate, and carbon emission growth rate was conducted in this paper, as shown in Tables 7 and 8. The results showed that the correlation between the change of the decoupling elasticity index and the change in pig price reached 0.868. therefore, we have reasons to believe that the change in the decoupling status is closely related to the "Pig Cycle" (Table 9).

Table 7 Grey correlation coefficient results

Table 8 Relevance results

Table 9 Decomposition of GHG emission drivers

4.3 Decomposition of carbon emission drivers

Due to the adjustment of the sectors covered by Chinese agriculture around 2004 and the change in the calculation of comparable output value in China since 2004, this paper only calculates the changes of five key variables of carbon emission drivers in the pig industry from 2005 to 2020 and their decomposition results. From Figs. 7, 8, 9, 10, and 11, it can be found that the technology factor (T), the structural change of the livestock sector (S1), and the ratio of livestock to GDP (S2) all show a downward trend, decreasing by 64, 15, and 50%, respectively. At the same time, affluence (I) and population growth (P) both show an upward trend, increasing by 510.0 and 8.6%, respectively.

Fig. 7

The change of technical progress index

Fig. 8

The change of livestock structure index

Fig. 9

The change of policy bias index

Fig. 10

The change of income

Fig. 11

The change of population

Between 2004 and 2020, the technology indicator fell by 64%, indicating a significant decrease in carbon emissions per unit of output value, reflecting technological progress in the pig breeding process, such as the introduction of specialized breeding technology, optimization of feed ratios, improvement of manure management technology, and increase in the rate of large-scale breeding. The index of livestock structure decreased by 15%, indicating that the output value of the pig industry decreased in the livestock industry. This means that the substitution of other livestock farming industries for pig farming has increased for the livestock industry. Besides, it can also reflect the decrease in the consumption of pork in meat food by people, showing the shift in people's dietary structure. The proportion of livestock in total GDP has dropped by 50%, which to a certain extent can reflect the direction of industrial development of China's economy. The position of animal husbandry in the national economy is in a downward trend, and the relevant policy-making is biased toward the secondary and tertiary industries. China's economy has been in rapid development in the past two decades, the per capita income has increased more than five times, and the population has also increased by nearly 10%. The increase in affluence and population growth will both significantly reflect on the per capita meat consumption in China within the past 2 decades, which are the major factors for the increase in annual pig slaughter.

Between 2004 and 2020, the net change in GHG emissions from the pig industry in China was -3.66 million tons of CO₂. Overall, technological factors, structural factors, and policy bias factors were the main drivers of emission reduction in the pig farming industry, whereas affluence and population factors were the main causes of emission increase. Compared with the base period, the technology factor reduced carbon emissions by 103.4%, and it had an emission reduction effect in most years during the study period, with an average annual emission reduction effect of 130.50 million tons of CO₂. The structural adjustment of livestock and the proportion of livestock in GDP have cumulatively reduced carbon emissions by 16.6 and 69.3%, reaching an average annual emission reduction of 209.35 and 8.74 million tons of CO₂, respectively. Compared to the base period, the affluence and population factors increased carbon emissions by 179.3 and 8.2%, respectively, both of which showed an incremental effect in almost all years studied, with an average annual incremental effect of 22.61 and 1.04 million tons of CO₂. In summary, the order of carbon emission reduction effects of China's pig industry is technology factor > policy bias factor > livestock structure factor > population factor > affluence factor.

In this paper, the technological factors cover all technological advances, and the results show that technological improvements are crucial to reducing GHG emissions from pork production in China, with the cumulative emission reduction effect reaching 103.48%. Specifically, the effect of technology on GHG emission reduction fluctuated over time, with significant emission reduction effects in the early period and decreasing effects after 2014, probably due to the diminishing marginal utility of technological advances in emission reduction, making the mitigation potential of early technological advances on GHG emissions from pig farming higher. However, the emission reductions from subsequent technological improvements tend to decrease. The impact of structural changes in the livestock industry on GHG emissions is generally negative and fluctuating, showing an incremental impact in some years and an abatement impact in others, with weak stability of abatement. It fluctuates over time with more and more years showing negative impacts, which also indicates the increasing substitution effect of other livestock industries on pig farming. The change in policy bias is reducing carbon emissions in most years, accounting for a relatively large share of the overall emission reduction contribution, ranking only after technological progress. Over time, the years in which the policy bias shows a negative impact on carbon emissions are also increasing. Although it rebounded in 2019 and 2020, showing an increase in emissions, it may be the effect of new policies, such as the continuous promotion of large-scale farming of pigs, which resulted in some negative effects. Extensive literature has demonstrated that transforming the economic structure can effectively reduce GHG emissions in the long term (Xiao et al, 2021; Zhang et al., 2020), but most studies have not evaluated livestock or pig farming. Through this paper, we can know that in the field of pig farming, it is possible to guide the restructuring of national agricultural production through industrial restructuring and policy changes to ensure the market supply of agricultural products while producing emission reduction effects.

Overall, carbon emissions tend to increase, mainly due to the positive driving effects of affluence and population growth. Numerous studies have shown that increased income drives a shift in diet toward more meat protein consumption (Schroeder et al, 1996; Vranken et al., 2014). Pork has always been the main source of traditional meat in China, with a significant increase in the resulting greenhouse gas emissions, which increased by 179.32% relative to 2004. In recent years, China's socio-economic development has continued at a high rate, and the per capita income level has increased substantially. The consequent changes in consumption levels have increased the demand for meat and dairy products, prompting further expansion of the livestock industry to meet the growing demand for meat, which will certainly lead to an increase in carbon emissions from livestock farming as well. In comparison, the greenhouse effect due to the population factor is less prominent, with a cumulative increase in carbon emissions of 8.21% relative to the base period. The main reason is that China started to enter an aging society after the twenty-first century, and the new population has remained low for a long time, so the annual population growth rate is not high. Even if the policy of "two and three children policy" has been liberalized, China's future fertility rate will still be very low (Guo et al., 2021), because families are not willing to face the increased costs of raising children, education, and living brought about by having more children. Therefore, in the long run, the decline in the population growth rate will help to reduce the demand for pork. As a whole, the incremental impact of population change is fluctuating, while the incremental effect of affluence is slowly decreasing, indicating that the concept of a green and healthy diet is gradually forming. With further increases in income, the proportion of vegetable consumption will continue to increase. However, affluence is still the largest driver of positive GHG emissions, with the population growth effect ranking second.

According to the relevant equations, we also use the multiplicative model to decompose the contribution of the five driving factors, as shown in Table 10.

Table 10 LMDI multiplicative decomposition results

During the study period, GHG emissions from China's pig industry were reduced to 98% of the 2004 levels. Technological innovation, changes in livestock structure, and proportion of livestock production reduced GHG emissions by about 65, 15, and 50%, respectively; affluence and population growth have increased GHG emissions by 510 and 8%, respectively. These results are consistent with the additive decomposition effects above (Fig. 12).

Fig. 12

Decomposition of GHG emissions from pig farming industry from 2005–2020

5 Drivers and decomposition of the decoupling states

To gain a comprehensive understanding of how carbon emission drivers impact the decoupling process and to identify factors that positively contribute to both decoupling status and carbon emissions reduction, this paper employs Eq. (16) to decompose the decoupling elasticity. The results are illustrated in Fig. 13. Over the period from 2005 to 2020, the decoupling elasticity of per capita income and structural decoupling exhibited substantial fluctuations, while the decoupling elasticity related to technology, policy bias, and demographics showed comparatively smaller variations. Furthermore, by examining the sub-decoupling elasticity index for the five influencing factors, it is evident that technology and income factors consistently held the largest share in most years, signifying their significant influence on decoupling within China's pig farming industry (Fig. 14).

Fig. 13

Elastic contribution of driving factors

Fig. 14

Elastic decomposition

Specifically, the impact of technological progress on decoupling elasticity generally tends to promote decoupling, while the levels of affluence and population typically exhibit a negative decoupling trend in most years. However, the influence of the two structural factors is relatively more complex. The impact of livestock structure on decoupling elasticity is inhibitory. When the output value experiences a positive change, the corresponding carbon emissions from the structural factor also increase, and vice versa. This positive decomposition of the decoupling coefficient indicates an inhibited decoupling effect. In contrast, policy factors, as represented by the proportion of livestock output, generally demonstrate a tendency to promote the decoupling state, resulting in an overall decoupling trend, with only a few individual years exhibiting negative effects. To summarize, based on the decomposition of decoupling coefficients, technological factors and policy bias serve as the primary drivers of decoupling, while affluence, population, and livestock structure act as significant factors inhibiting decoupling (Table 11).

Table 11 Decomposition of decoupling elasticity factors

However, there are certain limitations in this study. Firstly, the calculation of carbon emissions does not account for the feed transportation process due to challenges in obtaining relevant data. Future research can focus on improving this aspect. Secondly, this paper does not analyze the Chinese pig farming industry on a regional level, even though regional variations exist within the Chinese farming industry. Considering regional differences could provide valuable insights in future studies. Lastly, the number of carbon emission drivers examined in this paper is relatively small. Expanding the scope of drivers considered will be an area for improvement in future research. These limitations highlight areas for further investigation and improvement in future studies.

6 Conclusions

We measured the GHG emissions of the whole farming process (including the feed consumption) in China's pig industry from 2001 to 2020 and analyzed the driving forces, decoupling elasticity, and the factor decomposition of decoupling from 2005 to 2020. The conclusions are as follows:

First, it is observed that carbon emissions exhibited a general upward trend from 2001 to 2020. However, certain disruptive factors such as "blue ear disease," "African swine fever," and the "Pig Cycle" resulted in short-term fluctuations. Among the different emission sources, the feed link not only contributes the largest proportion to annual carbon emissions but also shows a continuous increase over time. Additionally, the emission from pig manure processing ranks second in terms of magnitude, and it poses significant environmental pollution risks by generating both greenhouse gases and contaminating water resources.

Second, the overall decoupling performance is predominantly characterized by a weak decoupling state. However, this state exhibits discontinuity and fluctuations that align with the same cycle as the "Pig Cycle," thus falling short of reaching the desired state of decoupling. In years with high economic growth in the pig industry output value, there is a significant increase in carbon emissions, and the decoupling elasticity fluctuates in tandem with the "Pig Cycle." Furthermore, the strong decoupling or weak decoupling state cannot be consistently sustained over an extended period. Therefore, the impact of the "Pig Cycle" on the decoupling effect becomes highly influential. Nonetheless, a notable benefit is that changes in carbon emissions and the decoupling status can be somewhat predictable.

Third, affluence is the strongest driver of increased GHG emissions from pig farming, followed by population; technological factors are the most important drivers of emission reductions, followed by two types of structural factors.

Fourth, the decomposition of the decoupling coefficients reveals that affluence, demographic factors, and livestock structure act as inhibitors to carbon decoupling, with affluence exerting the greatest influence. Technological factors and policy factors contribute to promoting carbon decoupling, with technology being the primary driver. In comparison with the decomposition of carbon emission drivers, these two structural factors play distinct roles in the decoupling process. The changes in livestock structure often have an inhibitory effect, indicating that while internal structural adjustments in the livestock industry reduce carbon emissions to some extent, they come at the expense of output growth. Conversely, the proportion of livestock output has a positive impact on both emission reduction and the promotion of decoupling effects.

The above measurement and analysis of carbon emissions in China's pig industry help to analyze the main influencing factors of carbon emissions in China's pig industry and effective implementable solutions for future emission reduction, in terms of both macro policies and covering micro subjects. The improvement of carbon emission calculation will also help to further refine the carbon measurement in future, such as setting out differential carbon emission coefficients by province to arrive closer to the real data. In addition, research on decoupling carbon emissions from the pig industry will help to achieve future decoupling in the pig industry and extend future research on emission reduction to the whole livestock industry and even agriculture. And based on the above research conclusions, the following recommendations are made to promote the construction of a low-carbon economic development system for China's pig farming industry.

• (1) For breeding enterprises and farmers, they often face problems such as rising feed prices, fluctuations in pig prices, environmental protection requirements, and epidemic threats. They should continue to play a leading role in the low-carbon development of the pig industry at the technical level, which could help reduce their own environmental pollution, reduce the risk of disease in farming, and cope with increasingly stringent environmental regulations. The technological level is the most important contributing factor in both emission reduction and decoupling. The technological factor has significantly reduced carbon emissions while promoting the increase in the output value of the pig industry. Therefore, the focus of the response approach should be on technological progress. In response to the phenomenon that carbon emissions from feed and manure processing account for a large share of the farming process, we should promote the improvement of feed formulation and appropriate adjustment of the nutritional structure of feed, as well as reduce the use of high-emission fertilizers to reduce carbon emissions from feed grain cultivation, enteric fermentation, and manure management. At the same time, we should improve the professional level of pig farming for farmers and promote the construction of scientific manure management systems, such as improving drainage systems, wet and dry manure separation, and processing into organic fertilizers. Furthermore, we should continue to promote the investment and construction of technical aspects such as large-scale breeding, energy-saving facilities, animal disease control, and learning of green breeding concepts.

• (2) The government should consider the holistic development of the entire industry across all regions of the country, while addressing issues of sustainable development and continually optimizing the industrial structure. In the previous analysis of structural factors, both factors have been found to play a significant role in emission reduction. However, certain adjustments may impose limitations on the growth of the pig industry. Therefore, with the priority of ensuring pork supply safety and recognizing the pig industry's role in the agricultural economy, agricultural and livestock structure adjustments can be implemented in different regions of the country, with the government appropriately reducing the proportion of livestock farming. Simultaneously, it is crucial to analyze the industrial development of each region and provide farming policy support to those regions in need, while avoiding excessive support that may result in significant environmental pollution. In-depth research can be conducted on the interdependence between pig farming and related agricultural industries in different regions, accelerating the integration of farming and breeding methods to achieve coordinated development between the planting and breeding sectors. Additionally, by effectively utilizing manure resources, the carbon emissions from feed grain production and manure treatment can be reduced, emphasizing the importance of implementing efficient manure management practices.

• (3) Conduct guidance on consumption patterns. The growth of per capita income is the most influential factor affecting emissions reduction and decoupling in the pig industry, so the Chinese government should now insist on promoting more green and environmentally friendly food consumption patterns, such as reducing the proportion of meat consumption and increasing diversified protein sources, which will help reduce future demand for pork.

• (4) Take relevant measures to deal with the "Pig Cycle." Our study finds that the "Pig Cycle" leads to cyclical fluctuations in carbon emissions and decoupling status, which makes the government's emission reduction policy and the low-carbon process of the pig industry suffer tremendously. But at the same time, we can flexibly adjust the relevant governance efforts for foreseeable priority years. However, the fundamental problem is to solve the "Pig Cycle," so government should continuously improve the construction of the information platform of pig supply and demand and strengthen the improvement and construction of agricultural insurance related to the breeding process, etc.