Spatio-temporal patterns of energy consumption-related GHG emissions in China's crop production systems

Energy Policy 104 (2017) 274–284

Contents lists available at ScienceDirect

Energy Policy journal homepage: www.elsevier.com/locate/enpol

Spatio-temporal patterns of energy consumption-related GHG emissions in China's crop production systems

MARK

⁎

Wei Zhena, Quande Qina,b,c, , Yi-Ming Weib,c,d a

College of Management, Shenzhen University, Shenzhen 518060, China Center for Energy & Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China c School of Management and Economics, Beijing Institute of Technology, Beijing 100081, China d Collaborative Innovation Center of Electric Vehicles in Beijing, Beijing 100081,China b

A R T I C L E I N F O

A BS T RAC T

Keywords: Crop production system Energy consumption Greenhouse gas emissions LMDI Spatio-temporal characteristics

This paper aims to reveal the spatio-temporal patterns of energy consumption-related greenhouse gas (ECRGHG) emissions in China's crop production systems (CPSs). The relevant crop production data from 31 provinces during 1997–2014 are utilized. In order to fully reflect the energy consumption and ECR-GHG emissions in CPSs, energy balance techniques are adopted from a consumption perspective. The driving factors behind ECR-GHG emissions are identified by means of a Logarithmic Mean Divisia Index analysis at both national and provincial levels. The results are as follows: (1) The yield of China's CPS is not positively correlated with energy consumption, and China's CPS has the relatively high potential to conserve energy and reduce ECRGHG emissions; (2) Most of China's provinces have experienced enormous growth in ECR-GHG emissions; however there are relatively significant regional disparities; (3) ECR-GHG emissions from CPSs were mostly derived directly from the consumption of chemical fertilizers and diesel oil; (4) Areal productivity is the determining factor in the growth of ECR-GHG emissions, whereas the emission coefficient and energy mix are the main inhibiting factors; (5) Energy intensity has not achieved its full potential to decrease ECR-GHG emissions. This study provides insights into the potential for sustainable crop production in China.

1. Introduction During the past two decades, China has experienced spectacular economic growth, which has come with high levels of fossil energy consumption (Wang et al., 2014). Significant emissions of major greenhouse gases (GHG), including nitrous oxide (N2O), methane (CH4) and carbon dioxide (CO2), are being released into the atmosphere due to energy consumption (Schneider et al., 2007). China currently ranks as the world's largest GHG emitter now. China has enacted a number of laws and regulations to disaggregate energy savings and emission reductions in each region and industry. The role of agriculture cannot be underestimated in the context of climate change (Robaina-Alves et al., 2014). China is one of the largest agricultural countries in the world, and its agricultural industry is responsible for approximately 11% of the nation's total GHG emissions (NCCC, 2012). The crop production system (CPS) is the most important production sector in the agriculture industry. In China, this system is now forced to confront the problem of increased dependency on energy sources such as chemical fertilizers, electricity, diesel oil, etc. (Karkacier et al., 2006). Energy consumption in the CPS increased ⁎

rapidly in response to corresponding increases in population and the limited supply of arable land (Cao et al., 2010; Khoshnevisan et al., 2013b). These factors have encouraged a tendency toward intensive energy use in the CPS, as a means to maximize yield, minimize laborintensive practices, or both (Esengun et al., 2007; Ghorbani et al., 2011). However, energy consumption leads to adverse environmental impacts, such as increasing the potential for global warming, degrading soil quality, and contributing to water, soil and air pollution (Nemecek et al., 2011a). Among these adverse environmental impacts, issues surrounding energy consumption-related GHG (ECR-GHG) emissions have become particularly prominent over the past two decades (Schramski et al., 2011; Khoshnevisan et al., 2013a, 2013b). The central government has issued practical policies targeted reducing ECR-GHG emissions in CPSs (Wang et al., 2015). In order to achieve effective policy-making it is important to understand the evolution of ECR-GHG emissions and the related driving forces of spatio-temporal patterns in China's CPSs. In recent years, the issues of ECR-GHG emissions in CPSs have become a hot topic (Nemecek et al., 2011a, 2011b; Pishgar-Komleh et al., 2012; Yousefi et al., 2014; Zhang et al., 2015). These above-

Corresponding author at: College of Management, Shenzhen University, Shenzhen 518060, China. E-mail address: [email protected] (Q. Qin).

http://dx.doi.org/10.1016/j.enpol.2017.01.051 Received 10 September 2016; Received in revised form 30 December 2016; Accepted 27 January 2017 0301-4215/ © 2017 Elsevier Ltd. All rights reserved.

Energy Policy 104 (2017) 274–284

W. Zhen et al.

national and provincial level, respectively; (2)A new extended KayaPorter identity, which considering the most important factors in China's CPSs, such as chemical fertilizer consumption structures and the energy mix, is proposed; (3) A LMDI decomposition method is employed both national and provincial levels to determine the driving force of ECR-GHG emissions in China's CPSs. The rest of this paper is organized as follows: Section 2 briefly introduces the methodology in detail. Section 3 describes the data sources. Our research results are presented in Section 4. Policy implications are discussed in Section 5. Finally, we draw our conclusions in Section 6.

named studies illustrated that the use of fossil energy increased rapidly, and the related GHG emissions derived from the production and use of fossil energy comprise a major portion of the total emissions of CPSs. These related studies contributed to an understanding of the relationship between energy consumption and the related GHG emissions in CPSs. Due to the imbalanced development of China's CPSs, the spatial and temporal features of ECR-GHG emissions from CPSs must be examined as an important factor for policy-making. Furthermore, previous studies (Liu et al., 2012; Kang et al., 2014; Xu et al., 2014) treated the agricultural sector in a similar manner as other economic sectors in terms of analyzing the factors that influence ECR-GHG emissions. CPSs have certain features that differ from those of other production sectors in terms of ECR-GHG emissions (Margarita and Victor, 2014). Modern crop production is characterized by high inputs of fossil energy types, which are consumed as “direct energy” (the fuel and electricity used on the farm) and as “indirect energy” (the energy used beyond the farm for the manufacturing of fertilizers, plant protection agents, machines, etc.) (Hülsbergen et al., 2001). Previous studies treated CPSs as a common production sector. These studies estimated the ECR-GHG emissions of CPSs by using sector-specific energy consumption data from the Energy Balance Sheet. These data do not account for indirect energy consumption, which may underestimate the ECR-GHG emissions from CPSs. For example, the energy consumption associated with the production of pesticides, agricultural machinery, plastic film and chemical fertilizers was not taken into account in crop production-related emissions. Instead, these processes were accounted for in the various, specific manufacturing industries. Moreover, the agricultural energy consumption data of these studies originated from the Chinese Energy Balance Sheet. These related data include not just farming, but forestry, animal husbandry, fishery and water conservancy. As such, this data cannot accurately reflect the quantity of energy consumption in CPSs. Thus, we believe that previous studies (e.g., Liu et al., 2012; Kang et al., 2014; Xu et al., 2014) did not fully reveal the panorama of energy consumption and ECR-GHG emissions in CPSs. Examining the driving forces behind ECR-GHG emissions is the key to taking effective energy conservation and GHG reduction measures in CPSs (Sanchez and Stern, 2016). The Logarithmic Mean Divisia Index (LMDI) technique is a widely accepted analytical tool, one which can be used to identify the relative impacts of different factors (Ang, 2005; Ma and Stern, 2008; Mulder et al., 2014; O’Mahony et al., 2013; Xu et al., 2014; Zha et al., 2010; Zhao et al., 2012; Zhang and Da, 2015). The Kaya-Porter identity, an extension of the Kaya identify, was introduced to address GHG emissions from CPSs (Bennetzen et al., 2012, 2016a, 2016b). The Kaya-Porter identity provides an effective tool for understanding the components of GHG emissions from crop production (Bennetzen et al., 2012). In view of the characteristics of China's CPSs, an extended Kaya-Porter identity was applied to quantify and explain the major driving forces of ECR-GHG emissions. In this paper, the final consumption point is used as a basis for estimating ECR-GHG emissions. The energy balance technique developed by Hülsbergen et al. (2001) is also employed to estimate energy consumption and ECR-GHG emissions. The goal of this study is to conduct an in-depth comparative analysis of the entire nation and 31 provinces based on their ECRGHG emissions from CPSs during the period 1997–2014 in China. The features of this paper can be summarized as follows: (1) This paper reveals the panorama of energy consumption and ECR-GHG emissions in CPSs to a certain degree; (2) This paper explores the driving forces behind ECR-GHG emissions and investigates regional disparities from the spatial patterns; (3) This paper provides the pertinence policies and suggestions for sustainable crop production development in China. We summarize our contributions as follows: (1) From a consumption perspective, this paper analyzes the evolution of ECR-GHG emissions in China's CPS during the period from 1997 to 2014, and investigates the spacial features of the ECR-GHG emissions from CPSs at the

2. Methods 2.1. Estimation of energy consumption in CPSs We estimated the energy consumption in CPSs, taking a consumption-based perspective. In this study, energy consumption was calculated using the energy balance technique described by Hülsbergen et al. (2001), Tzilivakis et al. (2005) and Rathke and Diepenbrock (2006). All fossil energy consumption of agricultural production inputs (APIs) (except for manpower, animal power and solar energy inputs) are included. The total amount of fossil energy of APIs used in a CPS has both direct and indirect components. Direct energy consumption includes the diesel oil used on farms and the electricity used for irrigation. Indirect energy consumption includes the energy used in the production of farm machinery, chemical fertilizers, pesticides and plastic film. The calculation of energy consumption is based on Eq. (1):

ECiT =

∑ ECijT = ∑ ITijT × ECCj j

(1)

j

ECiT

refers to the total energy consumption of a CPS in province i where in year T, ECijT represents the total energy consumption of API type j in province i in year T, ITijT denotes the quantity of API type j in province i and ECCj refers to the energy conversion coefficient of API type j. In particular, energy consumption in the production of agricultural machinery was estimated using Eq. (2): T ECAM = i

T = ∑ AMinT × ECCn × DR ∑ ECAM in n

n

(2)

T EC AM i

where refers to the total energy consumption in the production T denotes the of agricultural machinery in province i in year T, EC AM in energy consumption in the production process of agricultural machinery n in province i in year T, AMinT refers to the total quantity of agricultural machinery n in province i in year T, ECCn represents the energy conversion coefficient of agricultural machinery n, and DR refers to the depreciation rate of agricultural machinery, which was set to 10% (Chen, 2002). 2.2. Estimation of ECR-GHG emissions In this study, ECR-GHG emissions are given in Eq. (3). The totals include both direct emissions and indirect emissions. ECR-GHG emissions from a CPS, including carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O), were all converted into carbon dioxide equivalents by multiplying each form of emission by its global warming potential (GWP) parameters, which are 1, 25 and 298 for CO2, CH4 and N2O, respectively (IPCC, 2006).

GHGiT =

∑ GHGijT = ∑ ITijT × EFj j

j

(3)

In Eq. (3), GHGiT represents the total ECR-GHG emissions of a CPS in province i in year T, GHGijT refers to the ECR-GHG emissions of API type j in province i in year T, ITijT denotes the quantity of API type j in province i, and EFj refers to the emissions coefficient of API type j. ECR-GHG emissions caused by the production of agricultural 275

Energy Policy 104 (2017) 274–284

W. Zhen et al.

province i (GHGiT ) can be evaluated using Eq. (7):

machinery are estimated using the following equation: T T GHGAM = ECAM × EFAM i i

GHGiT =

(4)

ECOijT

where refers to the emission coefficient of API type j in province i in year T, EMijT denotes the share of energy consumption of API type j in the total energy consumption of a CPS in province i in year T, EIiT is the energy intensity in province i in year T, PiT refers to the areal productivity of a CPS in province i in year T and PAiT refers to the planted area in province i in year T. The aggregate effects of the various driving factors from the baseline year (GHGi0 ) to the final year (GHGiT ) can be expressed by Eq. (8), as follows:

2.3. Estimation of crop production The total crop production per province was calculated using Eq. (5):

∑ DMijT = ∑ OPijT × DMCCj j

T T T T ΔGHGtot = GHGiT − GHGi0 = ΔGHGECO + ΔGHGEM + ΔGHGEI i i i i

(5)

j

T + ΔGHGPTi + ΔGHGPA i

where DMiT denotes the total dry matter (DM) of a CPS in province i in year T, DMijT refers to the DM of crop type j in province i in year T, OPijT represents the yield of crop type j in province i in year T, and DMCCj refers to the DM conversion coefficient of crop type j.

T = ΔGHGECO i

j

GHGijT ECijT

×

ECijT ECiT

×

ECiT DMiT × × areaiT DMiT areaiT

∑ j

The Kaya-Porter identity is a novel concept and provides a simple framework for understanding the components of GHG emissions from CPSs (Bennetzen et al., 2012, 2016b). Thus, in recent years it is widely accepted to estimate the GHG emissions from crop production (Bennetzen et al., 2012, 2016a, 2016b). The deconstruction of the elements of emissions in the Kaya-Porter identity allows policymakers to assess which elements would be relevant to energy conservation and the reduction of ECR-GHG emissions in CPSs (Bennetzen et al., 2012). According to the features of the Kaya-Porter identity, it is interesting to analyze quantitatively the driving forces behind the ECR-GHG emissions from CPSs and put award the related mitigation policies. Thus, the Kaya-Porter identity combine with LMDI technique enables policymakers to investigate how different elements could be targeted in actions to reduce emissions. Although China is the largest consumer of chemical fertilizers in the world (Lin et al., 2015) and its CPSs becoming the energy-intensive production systems, the reduction effect of structural changes around chemical fertilizers and other energy sources in China's CPSs has not received sufficient attention. To fill this research gap, an extended Kaya-Porter identity is proposed. We intend to utilize the extended Kaya-Porter identity to quantify and explain the major driving forces behind ECR-GHG emissions in China's CPSs. According to the extended Kaya-Porter identity, the ECR-GHG emissions in province i in year T (GHGiT ) can be expressed using Eq. (6):

∑

(8)

The various driving forces can be quantified according to Eqs. (9)– (13).

2.4. Decomposition analysis of changes in ECR-GHG emissions

GHGiT =

(7)

j

T where GHG AM denotes the ECR-GHG emissions from the production of i T refers to the total agricultural machinery in province i in year T, EC AM i energy consumed during the production of agricultural machinery in province i in year T, and EFAM represents the GHG emissions coefficient of agricultural machinery.

DMiT =

× EMijT × EIiT × PiT × PAiT

∑ ECOijT

T = ΔGHGEM i

∑ j

T = ΔGHGEI i

∑ j

ΔGHGPTi =

∑ j

T = ΔGHGPA i

∑ j

GHGijT − GHGij0 ln

GHGijT

− ln

GHGij0

GHGijT − GHGij0 ln

GHGijT

− ln

GHGij0

GHGijT − GHGij0 GHGijT

ln

GHGij0

− ln

GHGijT − GHGij0 ln

GHGijT

− ln

GHGij0

GHGijT − GHGij0 ln

GHGijT

− ln

GHGij0

⎛ ECO T ⎞ i ln ⎜ ⎟ ⎝ ECOi0 ⎠

⎛ EM T ⎞ i ln ⎜ ⎟ ⎝ EMi0 ⎠

(9)

(10)

⎛ EI T ⎞ ln ⎜ i0 ⎟ ⎝ EIi ⎠

(11)

⎛ PT ⎞ ln ⎜ i0 ⎟ ⎝ Pi ⎠

(12)

⎛ PA T ⎞ ln ⎜ i0 ⎟ ⎝ PAi ⎠

(13)

In this paper, the changes in ECR-GHG emissions caused by T variations in the emission coefficients (▵GHGECO ) are only determined i by the proportional change in the amounts of chemical fertilizers (N, P, K and compound fertilizers), i.e., the regional consumption of N, P, K and compound fertilizers varies over the study period because the emission coefficients of other APIs are assumed to remain constant over time. 3. Data This paper includes data from 1997 to 2014 for 31 Chinese regions, including 13 major grain production areas2 (Hebei, Shandong, Liaoning, Jilin, Heilongjiang, Jiangsu, Anhui, Jiangxi, Henan, Hubei, Hunan, Inner Mongolia and Sichuan) and 18 non-major grain production areas (Beijing, Tianjin, Shanghai, Zhejiang, Fujian, Guangdong, Hainan, Shanxi, Guangxi, Chongqing, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang). Hong Kong, Macau and Taiwan were excluded from our study. We selected the period from 1997 to 2014 because this was the most recent period for which common data were available for all regions. These Chinese regions were also chosen because they had the best data availability over the study period. All data were obtained from China's official statistical database, published papers, and governmental reports, thus ensuring that the most reliable data related to CPSs were collected. We calculated ECRGHG emissions in million tons of carbon dioxide equivalents (Mt CO2-

(6)

where T is the time in years, subscript i represents the province, GHGiT refers to the total ECR-GHG emissions in province i in year T, GHGijT refers to the ECR-GHG emissions of API type j in province i in year T, ECijT represents the energy consumption of API type j in province i in year T, ECiT is the total energy consumption of a CPS in province i in year T, DMiT refers to the total crop production of a CPS in province i in year T, and areaiT refers to the total planted area dedicated to crop production in province i in year T. Based on Eq. (6), five potential antecedent factors can be identified as the driving forces behind ECR-GHG emissions from CPSs at the regional level, namely, the emission coefficients, energy mix, energy intensity1 (energy consumptions per unit DM), areal productivity and planted area. Thus, the ECR-GHG emissions from a CPS in year T in

2 According to a statement from “Opinions of the Central Committee of the CPC and the State Council Concerning Several Policies on Promoting the Increase of Farmers' Income” issued in 2014, thirteen provinces are categorized as major grain producing areas.

1 In this paper the concept of energy intensity denotes energy consumptions per unit DM in China's CPSs.

276

Energy Policy 104 (2017) 274–284

W. Zhen et al.

generally rapid rise in Phase 2 and a still rapid but more stable rise in Phase 3. A comparison of the annual growth rates in crop production (from 0.06% to 2.24%) and energy consumption (from 1.17% to 3.22%) between Phase 1 and Phase 2 (Fig. 1) shows that increasing energy consumption is a useful means to improve crop yields. Examining the annual growth rates of crop production (from 2.24% to 2.64%) and energy consumption (from 2.91% to 2.06%) in Phase 2 and Phase 3, respectively (Fig. 1), we can see that the growth rates of energy consumption slowed, but the growth rates of crop production continued to increase. These growth trends, therefore, reveal that energy consumption is not positively correlated to the yield of China's CPS. What's more, this phenomenon indicates that China's CPS has the potential for energy conservation and a corresponding reduction in ECR-GHG emissions. Examining the changes in different APIs in terms of energy consumption and ECR-GHG emissions in China, as illustrated in Table 1, all of the energy sources and ECR-GHG emissions associated with the material inputs of crop production increased during the study period. Energy consumption via the use of chemical fertilizers and diesel oil increased remarkably, by 52.70×104 TJ and 41.22×104 TJ, respectively. ECR-GHG emissions caused by the use of from chemical fertilizers, plastic film and diesel oil also exhibited remarkable growth, reaching totals of 83.64 Mt CO2-eq, 26.94 Mt CO2-eq and 23.72 Mt CO2-eq, respectively. Emissions directly associated with the use of diesel oil, pesticides, plastic film and agricultural machinery increased rapidly, with average annual rates of 4.53%, 3.01%, 7.18%, and 8.40%, respectively. In comparison, ECR-GHG emissions from chemical fertilizers and electricity for irrigation exhibited only small rates of increase during the same period. The total amount of ECR-GHG emissions from the CPS has increased by 41.87%, indicating a gradual shift toward an input-intensive CPS in China.

eq). The fossil energy consumption of APIs was considered in units of tera-joules (TJ). The energy consumption and ECR-GHG emissions from a CPS were combined as an aggregated indicator. Data regarding APIs (diesel oil, irrigation area, pesticides, plastic film, chemical fertilizers and agricultural machinery) were taken from the China Rural Statistical Yearbook (DRES, 1998–2015) and the China Agriculture Yearbook (DREB, 1998–2015). The total crop production dataset in each province was calculated by integrating the amount of DM in each type of crop in each province. These amounts were measured in millions of tons of dry matter (Mt DM). This approach allowed us to compare and aggregate production across different crop types, while still using a common unit of measure. Similar concepts have been presented by Hülsbergen et al. (2001), Raupach et al. (2008), Nemecek et al. (2011a), (2011b) and Kim and Dall'erba (2014). The production of crops in this study includes all cultivated crops, and the total production was represented as the bulk sum of grain crops (cereals and tuber crops) and cash crops (legume, oil crops, fibers, sugar crops, tobacco leaf, vegetables and fruits). The unit of crop land area was 1000 ha. The cropland area and crop production variables were available in the China Rural Statistical Yearbook (DRES, 1998– 2015) and China Agriculture Yearbook (DREB, 1998–2015). In this study, the energy conversion coefficients of APIs, the GHG emission coefficients of APIs and the DM content of major crops were obtained from related published papers, as shown in Appendix A. 4. Results and findings 4.1. Features of ECR-GHG emissions from the CPS 4.1.1. ECR-GHG emissions from the CPS in China From 1997–2014, the total yield of China's crop production grew rapidly, with a total growth increase of 31.96%. The energy consumptions of China's CPS increased by 49.04% during the study period and amounted to 410.85×104 TJ in 2014. The ECR-GHG emissions grew from 383.29 Mt CO2-eq in 1997 to 543.78 Mt CO2-eq in 2014, representing an annual growth rate of 2.46% during the study period. From a national point of view, as shown in Fig. 1, the development of ECR-GHG emissions from CPSs in China since 1997 can be divided into three phases: a slow increasing phase (Phase 1: before 2003), a steeply increasing phase (Phase 2: between 2003 and 2009) and a relatively rapidly increasing phase (Phase 3: after 2009). Consequently, the energy consumed by the CPS followed a similar growth trend. The CPS first exhibited a weaker growth trend in Phase 1, a fluctuating but

4.1.2. ECR-GHG emissions from CPSs in 31 Chinese provinces The spatial distributions of the ECR-GHG emissions from CPSs in China's 31 provinces for the years 1997, 2003, 2009 and 2014 are illustrated in Fig. 2. The figure clearly shows that most of China's provinces experienced tremendous growth in ECR-GHG emissions in their CPSs since 1997, excluding some non-major grain production areas (such as Beijing, Tianjin, Shanghai, Chongqing, Hainan, Guizhou Ningxia, Qinghai and Tibet). In 1997, only the ECR-GHG emissions from the CPS of one major grain production area (Shandong, 38.57 Mt CO2-eq) exceeded 33 Mt CO2-eq. However, in 2014, the ECR-GHG emissions from the CPSs in three major grain production areas (Henan,

Fig. 1. Changes in crop production, energy consumption and ECR-GHG emissions in China's CPS (1997–2014).

277

Energy Policy 104 (2017) 274–284

W. Zhen et al.

Table 1 Changes in ECR-GHG emissions and energy consumption from different sources (1997–2014). Emission source

Diesel oil

Fertilizers

Electricity

Pesticides

Plastic film

Agricultural machinery

GHG (Mt CO2-eq) Energy (104 TJ)

23.72 41.22

83.64 52.70

9.47 12.32

11.06 6.14

26.94 14.85

5.65 7.96

2014. Fig. 3 shows the detailed ECR-GHG emissions from CPSs in China's 31 provinces in 2014, including GHG emissions generated by various energy sources. From a national point of view, the ECR-GHG emissions from the CPSs were mainly derived from the consumption of chemical fertilizers and diesel oil, which contributed 352.02 Mt CO2-eq and 54.51 Mt CO2-eq of the total emissions, respectively. However, regional disparities in ECR-GHG emissions patterns are closely linked to the energy consumption structure of the CPSs. The detailed energy consumption structures of CPSs in China's 31 provinces in 2014 are shown in Fig. 4. From Fig. 4, we can clearly see that chemical fertilizers and diesel oil are the two primary energy sources for most provinces, and regional disparities in energy consumption patterns exist. For instance, Hebei and Henan are the most important Chinese grain production areas, but they exhibit different CPS-based energy con-

Shandong and Hebei,) exceeded 33 Mt CO2-eq, where the emissions reached 52.17 Mt CO2-eq, 43.62 Mt CO2-eq and 36.17 Mt CO2-eq, respectively. Out of all China's provinces, two major grain production areas (Inner Mongolia and Henan) and one non-major grain production area (Xinjiang) experienced the largest increments in ECR-GHG emissions from their CPSs during 1997–2014, with increases of 12.23 Mt CO2-eq, 14.79 Mt CO2-eq and 20.81 Mt CO2-eq, respectively. During 1997–2014, the ECR-GHG emissions from the CPSs of China's major grain producing areas increased by 35.97%, reaching 357.80 Mt CO2-eq in 2014. The growth rate of ECR-GHG emissions in non-major grain production areas experienced even more rapid growth. The growth rates in these areas increased by 54.79% during the study period, reaching 185.98 Mt CO2-eq in 2014. From the perspective of geographic area, China's major grain production areas contributed 65.80% of the country's ECR-GHG emissions from CPSs in

Fig. 2. China's provincial ECR-GHG emissions from CPSs (1997, 2003, 2009 and 2014).

278

Energy Policy 104 (2017) 274–284

W. Zhen et al.

Fig. 3. Structure of ECR-GHG emission in CPSs in 31 provinces (2014).

Fig. 4. Structure of energy consumption in CPSs in 31 provinces (2014).

(△GHGP). However, the planted area (△GHGPA) was the main driver behind the decrease in ECR-GHG emissions from CPSs in Phase 1, followed by changes in the energy mix (△GHGEM) and the emission coefficient (△GHGECO). In Phase 2 (2003–2009), areal productivity was the main driver behind the increase in ECR-GHG emissions from the CPS, followed by changes in the planted area and energy intensity. At the same time, the emission coefficient was the main driver behind the decrease in ECR-GHG emissions from the CPS in Phase 2, followed by changes in the energy mix. In Phase 3 (2009–2014), energy intensity was the main driver behind the decrease in ECR-GHG emissions from the CPS, followed by changes in the emission coefficient and the energy mix, while the areal productivity was the main driver increasing ECR-GHG emissions from the CPS in Phase 3, followed by changes in the planted area. From a temporal perspective, as shown in Fig. 5, we can obtain the following results. First, areal productivity (△GHGP) was the main

sumption patterns. In Henan, during the study period, crop production-related energy consumption was mostly associated with the consumption of chemical fertilizers (22.29×104 TJ). In Hebei, the consumption of both diesel oil and chemical fertilizers were the greatest contributors to crop production-related energy consumption (12.66×104 and 11.65×104 TJ, respectively).

4.2. Driving forces of ECR-GHG emissions from CPSs in China By conducting a time series LMDI decomposition analysis, the antecedent factors of ECR-GHG emissions from the CPS for all of China were quantified. Fig. 5 presents the decomposition analysis results for ECR-GHG emissions from China's CPSs for the three previously-mentioned phases. In Phase 1 (1997–2003), energy intensity (△GHGEI) was the main driver behind the increase in ECR-GHG emissions from CPSs, followed by changes in areal productivity 279

Energy Policy 104 (2017) 274–284

W. Zhen et al.

Fig. 5. Decomposition results of ECR-GHG emissions from CPSs (1997–2014) (unit: Mt CO2-eq).

However, energy mix contributed to increases in ECR-GHG emissions in a number of major Chinese grain production areas (such as Inner Mongolia and Henan) and non-major grain production areas (such as Beijing, Tibet, Qinghai, Hainan, Gansu, Guangxi, Yunnan, Shaanxi and Xinjiang). From a national perspective, the planted area played a role in increasing the amount of ECR-GHG emissions from CPSs in major Chinese grain production areas. Correspondingly, however, the planted area played a role in decreasing the amount of ECR-GHG emissions from CPSs in non-major Chinese grain production areas. Meanwhile, the effects of the planted area varied at the provincial scale. In some areas, planted area showed a positive effect (Tibet, Ningxia, Shandong, Guangxi, Sichuan, Hubei, Gansu, Guizhou, Anhui, Liaoning, Hunan, Inner Mongolia, Jilin, Yunnan, Heilongjiang, Henan and Xinjiang). In other areas, planted area had a negative effect (Zhejiang, Guangdong, Fujian, Beijing, Shaanxi, Jiangsu, Shanghai, Hebei, Chongqing, Hainan, Tianjin, Jiangxi, Qinghai and Shanxi). Supporting Figs. 1–5 (see Appendix B) compare the results for the average values of the five antecedent factors at the provincial and national levels. These five figures indicate that the antecedent factors are stronger in China's major grain production areas than in the nonmajor grain production areas. This is especially true for the emission coefficient. The average values of the emission coefficient in non-major grain production areas (such as Shaanxi, Guangxi and Shanxi) are slightly stronger than the national average. At the same time, the average values of the emission coefficient in most major grain production areas (such as Henan, Shandong, Anhui, Jilin, Hubei, Hebei, Jiangxi, Jiangsu, Hunan and Liaoning) are much stronger than the national average. From a national perspective, the functions of each of these five antecedent factors are different. Energy intensity (△GHGEI), areal productivity (△GHGP) and planted area (△GHGPA) increased the level of ECR-GHG emissions coming from China's CPSs during the study period. The emission coefficient (△GHGECO) and energy mix (△GHGEM) offset increases in ECR-GHG emissions from the CPSs.

driver behind the increase in ECR-GHG emissions from the CPS, followed by changes in energy intensity (△GHGEI) and planted area (△GHGPA). Second, the decreased emission coefficient (△GHGECO), followed by the change in the energy mix (△GHGEM), were the two most important forces behind the decrease in ECR-GHG emissions from the CPS in China from 1997 to 2014. The results show that the decrease in the emission coefficient played a dominant role, decreasing ECR-GHG emissions from CPSs by 11.86 Mt CO2-eq and accounting for 7.39% of the total change. The change in the energy mix decreased ECR-GHG emissions from CPSs by 9.11 Mt CO2-eq, accounting for 5.68% of the total change in the absolute value. The increase in areal productivity increased ECR-GHG emissions from the CPS by 99.38 Mt CO2-eq, accounting for 61.91% of the total change. The changes in energy intensity and planted area increased ECR-GHG emissions by 45.92 Mt CO2-eq and 36.18 Mt CO2eq, respectively. 4.3. Aggregated effects of the various drivers in 31 provinces The levels of China's ECR-GHG emissions from CPSs have generally grown since 1997. As such, it has become increasingly necessary to identify the drivers behind these increases, in order to design specific solutions for each problem. The aggregated effects of the driving forces behind ECR-GHG emissions from CPSs in various major and nonmajor grain production areas from 1997 to 2014 are presented in Fig. 6. Historically, the changes in ECR-GHG emissions from CPSs in most major and non-major Chinese grain production areas were driven by areal productivity. Areal productivity partially offset the growth in ECR-GHG emissions in some major grain production areas (such as Anhui) and some non-major grain production areas (such as Beijing and Guangdong). The increase of energy intensity is another significant driver behind ECR-GHG emissions from CPSs in China's major and non-major grain production areas. Energy intensity decreased in some of China's major grain production areas (such as Jiangsu, Liaoning, Hebei and Shandong) and non-major grain production areas (such as Qinghai and Guangxi). The decrease in energy intensity played a role in inhibiting the growth of ECR-GHG emissions from CPSs in these regions. In other regions, increases in energy consumption per unit of DM played a role in increasing the level of ECR-GHG emissions. During the study period, the emission coefficient was the major inhibiting factor for ECR-GHG emissions from CPSs based on the factors analyzed in this paper. The changes in the energy mix had only a marginal effect on ECR-GHG emission changes in China's CPSs.

5. Discussion and policy implications The level of both ECR-GHG emissions and energy consumption of China's CPSs increased from 1997 to 2014. Because of the rapidly increasing energy consumption of APIs (especially chemical fertilizers) this trend became increasingly severe during this period. Our results indicate that significant provincial disparities exist in ECR-GHG emission characteristics and the related driving forces in China's 280

Energy Policy 104 (2017) 274–284

W. Zhen et al.

Fig. 6. Aggregated effects of various driving forces behind ECR-GHG emissions from CPSs in 31 provinces (1997–2014) (unit: Mt CO2-eq).

is known about the regional energy efficiency of CPSs. More importantly, our research results indicate that a remarkable disparity exists (in terms of energy intensity) between provinces. In addition, poor energy efficiency directly resulted in increases in ECR-GHG emissions in most provinces during the period 1997–2014. For example, Hebei and Sichuan are ranked as China's major grain production areas. The energy intensity in Hebei in 2014 was estimated to be 127% higher than that in Sichuan. Energy efficiency in Sichuan is 2.27 times of Hebei. Thus, great disparities exist in the technical levels within regional CPSs. In further, these disparities in CPSs have not been fully recognized by policymakers. Technological disparities also exist in the API production processes (such as chemical fertilizers, pesticides, etc.) and equipment (agricultural machinery). Due to the lack of advanced technology and equipment in some areas, the less developed regions must employ low-efficiency technologies and rely on increasing agricultural material inputs to increase crop production (Cao et al., 2010). Under such circumstances, provincial governments should take action to improve the energy efficiency in their CPSs. The speed of technical progress in agriculture depends on market and political incentives for research (Traxler and Byerlee, 2001; Raitzer and Kelley, 2008). Economic and market instruments, such as financial subsidies, can be employed to promote technical progress without negatively impacting farm incomes (Schneider and McCarl, 2003). For instance, investments in CPSs have proven to be powerful tools that maintain productivity growth and can potentially achieve co-benefits, such as mitigating GHG emissions (Burney et al., 2010; Zhen et al., 2016). During the study period, the energy mix did offset increases in ECRGHG emissions from CPSs in most provinces (Fig. 6). Such results indicate the positive effect of changes in the energy structure in most provinces and variations between different provinces. The changes in the energy mix in China's provincial CPSs are illustrated in Fig. 7. It is clear that the proportions of diesel oil, pesticides, plastic film and agricultural machinery increased by small margins in most provinces, while the proportions of electricity and chemical fertilizers considerably decreased in most provinces. Such changes help reduce ECR-GHG emissions from CPSs. The energy consumption and GHG emission intensity associated with producing electricity and chemical fertilizers are much higher than those of other APIs (Popp et al., 2010; Zhang et al., 2015). Because the energy consumption of chemical fertilizers

CPSs. China's major grain production areas exhibited the highest ECRGHG emissions increases in their CPSs over the study period, accounting for more than 65% of China's total ECR-GHG emissions from CPSs. Thus, more comprehensive and stringent policies and standards should be employed in these high emission-producing regions. Meanwhile, ECR-GHG emissions from CPSs in other nonmajor grain production areas are rapidly increasing. This finding indicates that these regions should take immediate action to mitigate their total ECR-GHG emissions. Regional disparities indicate that region-specific policies should be implemented to reduce ECR-GHG emissions. The LMDI analysis suggests that areal productivity is the primary driving force behind the rapid increase in ECR-GHG emissions from CPSs. China's CPSs also face with the challenge of promoting crop production with an increasingly limited supply of cropland. To address these issues, the central government issued a series of policies to promote agricultural modernization and improve areal productivity (NPC, 2004; CPC, 2006). These polices include the promotion of largescale agricultural machinery and agrochemical use on a nationwide basis. These policies have led to an extremely high GHG emissions approach of crop production in China. ECR-GHG emissions mitigation policies should focus on controlling the total energy inputs in the CPSs. However, due to China's growing population and the expanding scale of crop production and investment in crop production infrastructure, it is difficult to limit the total energy inputs in the CPSs. However, some policies should be created and implemented, such as optimizing the planting structure and planting energy-efficient crops. Planting structure changes in the CPSs have been found to directly affect output and energy consumption. Thus, changes in planting structure are closely linked to ECR-GHG emissions. Alternatively, based on regional resource endowments, local governments should promote the planting of energy efficient, rather than energy-intensive, crops (Sabri et al., 1991). Energy intensity is another significant driver behind rising ECRGHG emissions from the CPSs. Currently, China's GHG mitigation policies mainly depend on mandatory energy intensity reductions across all sectors (Wang and Chen, 2010; Liu et al., 2012). Energy intensity can reflect the degree of energy efficiency of CPSs and deliver more information than other indicators (Grassini and Cassman, 2012; van Groenigen et al., 2012; Bennetzen et al., 2016a). Even today, little 281

Energy Policy 104 (2017) 274–284

W. Zhen et al.

Fig. 7. Structural changes in energy sources in 31 provinces (1997–2014).

was predominant in the energy structures of most provincial CPSs (as illustrated in Fig. 4), the proportional downward changes in electricity and chemical fertilizers reduced the level of ECR-GHG emissions from CPSs. The most effective measure is the adoption of low-GHG emission energies. This is especially true of clean electricity generated through wind power, hydropower and solar power (Schneider and Smith, 2009; Guo et al., 2014). Since a large proportion of energy use is embedded in the process of producing APIs, low-GHG technologies should be implemented in this productive process, such as the early retirement of energy inefficient machinery (Sabri et al., 1991; Schneider and Smith, 2009). Furthermore, regulating the price of GHG-intensive fossil energy types may induce farmers to use energy sources with lowGHG emissions level. Similarly, manufacturers may be induced to use less energy per unit for APIs due to higher prices (Konyar, 2001; Schneider and McCarl, 2003; Schneider and Smith, 2009). The emission coefficient is another main antecedent factor that reduced ECR-GHG emissions from CPSs in all provinces except Tibet during this period. Due to the inadequate crop production modes which have been used over the past several decades, chemical fertilizers became the main source of energy (Fig. 3) and ECR-GHG emissions (Fig. 4) in China's CPSs. In recent decades, some changes have occurred in China's chemical fertilizer consumption structure. In that regard, Fig. 8 illustrates the proportional changes in the energy consumption of chemical fertilizers in China's provincial CPSs. It is clear that the proportions of N fertilizer and P fertilizer decreased in most provinces. In particular, N fertilizer use shows a considerable decrease, while the proportions of K fertilizer and compound fertilizer increased during the same period. These shifts in fertilizer consumption structures caused significant changes in fertilizer consumptionrelated GHG emissions from CPSs (as illustrated in Fig. 6). This indicates that the optimization of chemical fertilizer consumption modes, especially those of N fertilizer, play a crucial role in reducing ECR-GHG emissions from CPSs because over-fertilization is common in China's CPSs (Peng et al., 2010; Nayak et al., 2015). For example, Ju et al. (2009), Peng et al. (2010), Huang and Tang (2010), Norse et al. (2012) and Nayak et al. (2015) all indicated that the average N fertilizer application rate in China's CPS was from 60% to 150% higher than the recommended rate. To inhibit over-fertilization in China's CPSs, the central government proposed a “zero growth in chemical fertilizer consumption” by 2020 policy (NDRC, 2016). This will be achieved by optimizing the allocation of fertilizer resources and adjusting the

Fig. 8. Proportional changes in energy consumption from chemical fertilizers in 31 provinces (1997–2014).

chemical fertilizer consumption structures (NDRC, 2016). Meanwhile, previous studies conducted by Dong et al. (2013) and Nayak et al. (2015) indicated that reducing the application rates of chemical fertilizers to an optimal level can achieve a sustainable yield and produce GHG reduction benefits. Therefore, mitigation measures, such as improving fertilizer efficiency, are crucial if China is to develop low-GHG emission CPSs in the long term. This is true, because energy consumption and GHG emission intensity from N fertilizer are much higher than those from other chemical fertilizers (Zhang et al., 2015). Improving fertilizer efficiency can be achieved in various ways, such as the use of livestock waste as fertilizer. Additionally, soil testing and formula fertilization have been implemented in Zhejiang Province (ZPG, 2006). Furthermore, the rational utilization of crop residues, such as straw, in cropland regions can provide an enormous source of energy (Pathak and Wassmann, 2007) and increase soil fertility (Nayak et al., 2015). Moreover, the planted area contributed to decreases in ECR-GHG 282

Energy Policy 104 (2017) 274–284

W. Zhen et al.

in China's CPSs. This can be achieved by applying operational management tools, which in turn have direct effects on China's mitigation actions in CPSs. Furthermore, China's CPSs which currently have with relatively high API requirements should optimize energy consumption structures, especially chemical fertilizer consumption structures, which would help conserve energy and reduce GHG emissions.

emissions level in CPSs in several provinces, especially in developed regions such as Zhejiang, Guangdong, Beijing, Shanghai, Tianjin, Jiangsu and Chongqing. Due mainly to rapid urbanization, large amounts of arable land were converted into land for construction purposes (Tan et al., 2005), and the economic gap between rural and urban areas resulted in the mass migration of agricultural laborers to cities (Zhen et al., 2016). The lack of arable land and of labor reduced the overall scale of crop production (Xu et al., 2015), which in turn reduced the level of ECR-GHG emissions from CPSs in impacted provinces (Zhen et al., 2016). In the context of the rapid urbanization seen throughout China, improving the level of cropland management is essential in CPSs. First, from a food security perspective, effective cropland management can result in an increased crop production and make full use of cultivated land resources. Second, improved cropland management can increase soil nutrients, avoid the waste of material inputs, and thereby reduce ECR-GHG emissions from CPSs (Smith et al., 2008).

These results provide policy implications for those responsible for alleviating ECR-GHG emissions from the national and provincial CPSs in China and provide broader policy implications. Different regions and provinces should implement ECR-GHG emission mitigation measures in their CPSs based upon specific considerations, such as local endowments, energy consumption modes, technology levels and financial abilities. It is irrational to implement ECR-GHG emissions reduction targets in different regions and provinces by simply referring to the local crop production and not considering other factors. Similar to the case in China, many regions in a number of countries are facing significant GHG management issues associated with CPSs. This is due to the increasingly limited amount of arable land, rising crop yield demands and increased energy inputs to CPSs. Therefore, our research findings can help policymakers in different regions determine their emissions reduction targets for CPSs. A deeper exploration of the inequalities in ECR-GHG emissions at the regional level deserves further investigation. In addition, it is an interesting task to utilize the input-output analysis to examine the crop-specific embodied energy intensities and energy consumption related GHG intensities of some staple crops in the future.

6. Conclusions China has experienced unprecedented economic prosperity and urbanization over the past few decades. This rapid development has also been accompanied by major challenges at the national level, including improving crop yields and reducing ECR-GHG emissions from CPSs. Due to China's imbalanced regional development, a more detailed study of China's ECR-GHG emissions from CPSs is necessary to determine their spatial and temporal features, the real driving forces and the appropriate energy conservation and ECR-GHG emission reduction policies. This study is intended to fill the existing research gap by providing a detailed and comprehensive analysis of provincial ECR-GHG emissions in China's CPSs. Our research findings present a holistic picture of China's ECR-GHG emissions from CPSs, including total energy consumption, related GHG emissions and energy consumption-related GHG emissions structures, as well as their spatial distribution and historical evolution in different provinces. By conducting LMDI analyses on both a national scale and for the 31 provinces, this study quantifies how the emission coefficient, energy mix, energy intensity, areal productivity and planted area contributed to increases in ECR-GHG emissions in CPSs during the period 1997– 2014. The main research outcomes are summarized as follows:

Acknowledgments The authors thank anonymous referees and an editor of this journal for their valuable comments and gratefully acknowledge the assistance of Prof. Dabo Guan for constructive suggestions and helpful conversations. The work is partly supported by National Natural Science Foundation of China (Nos. 71521002 and 71402103), the MOE Youth Foundation Project of Humanities and Social Sciences at Universities in China (No. 13YJC630123), China Postdoctoral Science Foundation Funded Project (Nos. 2015M580053 and 2016T90042) and Natural Science Foundation of Guangdong Province (No. 2015A030313556).

1) From a temporal perspective, along with the rapid increase in crop production, China's ECR-GHG emissions exhibited an increasing trend from 1997 to 2014. This increase was especially sharp after 2003. Energy consumption in China's CPSs is not positively correlated with crop production. In addition, China's CPSs have the potential to conserve energy and to effectively reduce ECR-GHG emissions. 2) From a spatial perspective, significant regional disparities exist between ECR-GHG emissions and energy sources in China's CPSs. China's major grain production areas contributed to more than 65% of total ECR-GHG emissions from their CPSs. However, and the non-major grain production areas saw an even more rapid growth rate in ECR-GHG emissions during 1997–2014. These emissions were mostly caused by the consumption of chemical fertilizers and diesel oil. 3) Whether at national or provincial level, areal productivity was the main driver increasing ECR-GHG emissions from CPSs, followed by changes in energy intensity and planted area. The decreased emission coefficients play a dominant role in ECR-GHG emissions reduction, followed by the change in the energy mix. 4) Of all contributing factors, the energy intensity is not taken completely seriously. Thus, this factor did not achieve its full potential in reducing the level of ECR-GHG emissions from CPSs. The results presented in this study suggest that a reduction in energy intensity or an improvement in energy efficiency is possible

Appendix. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.enpol.2017.01.051. References Ang, B.W., 2005. The LMDI approach to decomposition analysis: a practical guide. Energy Policy 33 (7), 867–871. Bennetzen, E.H., Smith, P., Porter, J.R., 2016a. Decoupling of greenhouse gas emissions from global agricultural production: 1970–2050. Glob. Change Biol. 22 (2), 763–781. Bennetzen, E.H., Smith, P., Porter, J.R., 2016b. Agricultural production and greenhouse gas emissions from world regions—The major trends over 40 years. Glob. Environ. Change 37, 43–55. Bennetzen, E.H., Smith, P., Soussana, J.F., Porter, J.R., 2012. Identity-based estimation of greenhouse gas emissions from crop production: case study from Denmark. Eur. J. Agron. 41, 66–72. Burney, J.A., Davis, S.J., Lobell, D.B., 2010. Greenhouse gas mitigation by agricultural intensification. In: Proceedings of the National Academy of Sciences of the United States of America, vol. 107(26), pp. 12052–12057. Cao, S.Y., Xie, G.D., Zhen, L., 2010. Total embodied energy requirements and its decomposition in China's agricultural sector. Ecol. Econ. 69 (7), 1396–1404. Chen, F., 2002. Agroecology. China Agricultural University Press, Beijing. Department of Rural Social and Economic Investigation of the National Bureau of Statistics (DREB), 1998–2015. China Agriculture Yearbooks. China Statistics Press, Beijing. Department of Rural Social Economical Survey (DRES), 1998–2005. China Rural Statistical Yearbook. China Statistics Press, Beijing.

283

Energy Policy 104 (2017) 274–284

W. Zhen et al.

associated non-CO2 greenhouse gases from agricultural production. Glob. Environ. Change 20 (3), 451–462. Raitzer, D.A., Kelley, T.G., 2008. Benefit-cost meta-analysis of investment in the International Agricultural Research Centers of the CGIAR. Agric. Syst. 96 (1–3), 108–123. Rathke, G.W., Diepenbrock, W., 2006. Energy balance of winter oilseed rape (Brassica napus L.) cropping as related to nitrogen supply and preceding crop. Eur. J. Agron. 24 (1), 35–44. Raupach, M.R., Canadell, J.G., Le Quere, C., 2008. Anthropogenic and biophysical contributions to increasing atmospheric CO2 growth rate and airborne fraction. Biogeosciences 5 (6), 1601–1613. Robaina-Alves, M., Moutinho, V., 2014. Decomposition of energy-related GHG emissions in agriculture over 1995–2008 for European countries. Appl. Energy 114, 949–957. Sabri, H.M., Wilson, H.R., Wilcox, C.J., Harms, R.H., 1991. Comparison of energy utilization efficiency among six lines of White Leghorns. Poult. Sci. 70 (2), 229–233. Sanchez, L.F., Stern, D.I., 2016. Drivers of industrial and non-industrial greenhouse gas emissions. Ecol. Econ. 124, 17–24. Schneider, U.A., McCarl, B.A., 2003. Economic potential of biomass based fuels for greenhouse gas emission mitigation. Environ. Resour. Econ. 24 (4), 291–312. Schneider, U.A., Smith, P., 2009. Energy intensities and greenhouse gas emission mitigation in global agriculture. Energy Effic. 2 (2), 195–206. Schneider, U.A., McCarl, B.A., Schmid, E., 2007. Agricultural sector analysis on greenhouse gas mitigation in US agriculture and forestry. Agric. Syst. 94 (2), 128–140. Schramski, J.R., Rutz, Z.J., Gattie, D.K., Li, K., 2011. Trophically balanced sustainable agriculture. Ecol. Econ. 72, 88–96. Smith, P., Martino, D., Cai, Z., Gwary, D., Janzen, H., Kumar, P., McCarl, B., Ogle, S., O'Mara, F., Rice, C., Scholes, B., Sirotenko, O., Howden, M., McAllister, T., Pan, G., Romanenkov, V., Schneider, U., Towprayoon, S., Wattenbach, M., Smith, J., 2008. Greenhouse gas mitigation in agriculture. Philos. Trans. R. Soc. B-Biol. Sci. 363 (1492), 789–813. Tan, M.H., Li, X.B., Xie, H., Lu, C.H., 2005. Urban land expansion and arable land loss in China – a case study of Beijing-Tianjin-Hebei region. Land Use Policy 22 (3), 187–196. The Central People’s Government of the People’s Republic of China (CPC), 2006. Opinions of the Central Committee of the CPC and the State Council Concerning Several Policies on Promoting the Increase of Farmers' Income. Available at: 〈http:// www.gov.cn/test/2006-02/22/content_207415.html〉. The National People’s Congress of the People’s Republic of China (NPC), 2004. Law of Promotion of Agricultural Mechanization of the People’s Republic of China. Available at: 〈http://www.npc.gov.cn/wxzl/gongbao/2004-07/23/content_5332208.htm〉. The People’s Government of Zhejiang Province (ZPG), 2006. Ecological environment construction planning of Zhejiang Province. Available at: 〈http://www.zj.gov.cn/gb/ zjnew/node3/node24/node648/node651/node658/node679/userobject9ai14981. html〉. Traxler, G., Byerlee, D., 2001. Linking technical change to research effort: an examination of aggregation and spillovers effects. Agric. Econ. 24 (3), 235–246. Tzilivakis, J., Warner, D.J., May, M., Lewis, K.A., Jaggard, K., 2005. An assessment of the energy inputs and greenhouse gas emissions in sugar beet (Beta vulgaris) production in the UK. Agric. Syst. 85 (2), 101–119. van Groenigen, K.J., van Kessel, C., Hungate, B.A., 2012. Increased greenhouse-gas intensity of rice production under future atmospheric conditions. Nat. Clim. Change 3 (3), 288–291. Wang, Q., Chen, Y., 2010. Energy saving and emission reduction revolutionizing China's environmental protection. Renew. Sustain. Energy Rev. 14 (1), 535–539. Wang, W., Guo, L.P., Li, Y.C., Su, M., Lin, Y.B., de Perthuis, C., Ju, X.T., Lin, E.D., Moran, D., 2015. Greenhouse gas intensity of three main crops and implications for lowcarbon agriculture in China. Clim. Change 128 (1–2), 57–70. Wang, W.W., Liu, X., Zhang, M., Song, X.F., 2014. Using a new generalized LMDI (logarithmic mean Divisia index) method to analyze China's energy consumption. Energy 67, 617–622. Xu, X.S., Zhao, T., Liu, N., Kang, J.D., 2014. Changes of energy-related GHG emissions in China: an empirical analysis from sectoral perspective. Appl. Energy 132, 298–307. Xu, Y.J., Huang, K., Yu, Y.J., Wang, X.M., 2015. Changes in water footprint of crop production in Beijing from 1978 to 2012: a logarithmic mean Divisia index decomposition analysis. J. Clean. Prod. 87, 180–187. Yousefi, M., Damghani, A.M., Khoramivafa, M., 2014. Energy consumption, greenhouse gas emissions and assessment of sustainability index in corn agroecosystems of Iran. Sci. Total Environ. 493, 330–335. Zha, D.L., Zhou, D.Q., Zhou, P., 2010. Driving forces of residential CO2 emissions in urban and rural China: an index decomposition analysis. Energy Policy 38 (7), 3377–3383. Zhang, X.H., Pan, H.Y., Cao, J., Li, J.R., 2015. Energy consumption of China's crop production system and the related emissions. Renew. Sustain. Energy Rev. 43, 111–125. Zhang, Y.J., Da, Y.B., 2015. The decomposition of energy-related carbon emission and its decoupling with economic growth in China. Renew. Sustain. Energy Rev. 41, 1255–1266. Zhao, X.L., Li, N., Ma, C.B., 2012. Residential energy consumption in urban China: a decomposition analysis. Energy Policy 41, 644–653. Zhen, W., Qin, Q., Kuang, Y., Huang, N., 2016. Investigating low-carbon crop production in Guangdong Province, China (1993–2013): a decoupling and decomposition analysis. J. Clean. Prod. doi: 〈http://doi.org/10.1016/j.jclepro.2016.05.022〉.

Dong, G., Mao, X.Q., Zhou, J., Zeng, A., 2013. Carbon footprint accounting and dynamics and the driving forces of agricultural production in Zhejiang Province, China. Ecol. Econ. 91, 38–47. Esengun, K., Gunduz, O., Erdal, G., 2007. Input-output energy analysis in dry apricot production of Turkey. Energy Convers. Manag. 48 (2), 592–598. Ghorbani, R., Mondani, F., Amirmoradi, S., Feizi, H., Khorramdel, S., Teimouri, M., Sanjani, S., Anvarkhah, S., Aghel, H., 2011. A case study of energy use and economical analysis of irrigated and dryland wheat production systems. Appl. Energy 88 (1), 283–288. Grassini, P., Cassman, K.G., 2012. Correction for Grassini and Cassman, High-yield maize with large net energy yield and small global warming intensity. In: Proceedings of the National Academy of Sciences, vol. 109(10), pp. 4021–4021. Guo, B., Geng, Y., Franke, B., Hao, H., Liu, Y., Chiu, A., 2014. Uncovering China's transport CO2 emission patterns at the regional level. Energy Policy 74, 134–146. Huang, Y., Tang, Y.H., 2010. An estimate of greenhouse gas (N2O and CO2) mitigation potential under various scenarios of nitrogen use efficiency in Chinese croplands. Glob. Change Biol. 16 (11), 2958–2970. Hülsbergen, K.J., Feil, B., Biermann, S., Rathke, G.W., Kalk, W.D., Diepenbrock, W., 2001. A method of energy balancing in crop production and its application in a longterm fertilizer trial. Agric. Ecosyst. Environ. 86 (3), 303–321. Intergovernmental Panel on Climate Change (IPCC), 2006. Greenhouse Gas Inventory: 2006 IPCC Guidelines for National Greenhouse Gas Inventories. United Kingdom Meteorological Office, Bracknell, England. Ju, X.T., Xing, G.X., Chen, X.P., Zhang, S.L., Zhang, L.J., Liu, X.J., Cui, Z.L., Yin, B., Christie, P., Zhu, Z.L., Zhang, F.S., 2009. Reducing environmental risk by improving N management in intensive Chinese agricultural systems. In: Proceedings of the National Academy of Sciences of the United States of America, vol. 106(9), pp. 3041–3046. Karkacier, O., Gokalp, G.Z., Cicek, A., 2006. A regression analysis of the effect of energy use in agriculture. Energy Policy 34 (18), 3796–3800. Khoshnevisan, B., Rafiee, S., Omid, M., Mousazadeh, H., 2013a. Reduction of CO2 emission by improving energy use efficiency of greenhouse cucumber production using DEA approach. Energy 55, 676–682. Khoshnevisan, B., Rafiee, S., Omid, M., Mousazadeh, H., 2013b. Applying data envelopment analysis approach to improve energy efficiency and reduce GHG (greenhouse gas) emission of wheat production. Energy 58, 588–593. Kim, T., Dall'erba, S., 2014. Spatio-temporal association of fossil fuel CO2 emissions from crop production across US counties. Agric. Ecosyst. Environ. 183, 69–77. Konyar, K., 2001. Assessing the role of US agriculture in reducing greenhouse gas emissions and generating additional environmental benefits. Ecol. Econ. 38, 85–103. Lin, J.Y., Hu, Y.C., Cui, S.H., Kang, J.F., Xu, L.L., 2015. Carbon footprints of food production in China (1979–2009). J. Clean. Prod. 90, 97–103. Liu, Z., Geng, Y., Lindner, S., Guan, D., 2012. Uncovering China’s greenhouse gas emission from regional and sectoral perspectives. Energy 45 (1), 1059–1068. Ma, C.B., Stern, D.I., 2008. China's changing energy intensity trend: a decomposition analysis. Energy Econ. 30 (3), 1037–1053. Margarita, R.A., Victor, M., 2014. Decomposition of energy-related GHG emissions in agriculture over 1995–2008 for European countries. Appl. Energy 114, 949–957. Mulder, P., de Groot, H.L.F., Pfeiffer, B., 2014. Dynamics and determinants of energy intensity in the service sector: a cross-country analysis, 1980–2005. Ecol. Econ. 100, 1–15. National Coordination Committee on Climate Change (NCCC), 2012. Second National Communication On Climate Change of the People’s Republic of China. Available at: 〈http://qhs.ndrc.gov.cn/zcfg/201404/W020140415316896599816.pdf〉. National Development and Reform Commission (NDRC), 2016. Guidance about Speed up the Development of Agricultural Circular Economy. Available at: 〈http://www. sdpc.gov.cn/zcfb/zcfbtz/201602/t20160204_774444.html〉. Nayak, D., Saetnan, E., Cheng, K., Wang, W., Koslowski, F., Cheng, Y.F., Zhu, W.Y., Wang, J.K., Liu, J.X., Moran, D., Yan, X., Cardenas, L., Newbold, J., Pan, G., Lu, Y., Smith, P., 2015. Management opportunities to mitigate greenhouse gas emissions from Chinese agriculture. Agric. Ecosyst. Environ. 209, 108–124. Nemecek, T., Dubois, D., Huguenin-Elie, O., Gaillard, G., 2011a. Life cycle assessment of Swiss farming systems: i. Integrated and organic farming. Agric. Syst. 104 (3), 217–232. Nemecek, T., Huguenin-Elie, O., Dubois, D., Gaillard, G., Schaller, B., Chervet, A., 2011b. Life cycle assessment of Swiss farming systems: ii. Extensive and intensive production. Agric. Syst. 104 (3), 233–245. Norse, D., Powlson, D., Lu, Y., 2012. Integrated nutrient management as a key contributor to China's low-carbon agriculture. Climate Change Mitigation and Agriculture. Earthscan Publisher, London. O' Mahony, T., Zhou, P., Sweeney, J., 2013. Integrated scenarios of energy-related CO2 emissions in Ireland: a multi-sectoral analysis to 2020. Ecol. Econ. 93, 385–397. Pathak, H., Wassmann, R., 2007. Introducing greenhouse gas mitigation as a development objective in rice-based agriculture: i. Generation of technical coefficients. Agric. Syst. 94 (3), 807–825. Peng, S.B., Buresh, R.J., Huang, J.L., Zhong, X.H., Zou, Y.B., Yang, J.C., Wang, G.H., Liu, Y.Y., Hu, R.F., Tang, Q.Y., Cui, K.H., Zhang, F.S., Dobermann, A., 2010. Improving nitrogen fertilization in rice by site-specific N management. A review. Agron. Sustain. Dev. 30 (3), 649–656. Pishgar-Komleh, S.H., Ghanderijani, M., Sefeedpari, P., 2012. Energy consumption and CO2 emissions analysis of potato production based on different farm size levels in Iran. J. Clean. Prod. 33, 183–191. Popp, A., Lotze-Campen, H., Bodirsky, B., 2010. Food consumption, diet shifts and

284