Sustainable Planning Strategy of Dairy Farming in China Based on Carbon Emission from Direct Energy Consumption

Abstract

The mechanical and electrical development in dairy farming in China increases energy-related carbon emission (CE). To support the sustainable planning strategy of the department, this study calculated the CE and the carbon emission intensity (CI) of the direct energy consumed in dairy farms from 21 provinces in China. Through four dimensions analysis including the national level, farm scale, inter-provincial distribution, and main producing area, this study illustrates the impact of the environment, production, and management on CE. The total CE of nationwide dairy farming was about 2.4 Tg CO₂ eq. in 2019, and the CIs of the 21 provinces varied from 0.009 to 0.216 kg CO₂ eq. per kg of milk. The results indicate that the management mode applied in large-scale dairy farms (500 heads and above) varies considerably due to inadequate adaptation to climate. In general, semi-arid and semi-humid regions are more suitable for dairy farming than arid and humid regions. In the main milk-producing area, the spatial aggregation effect is visible in the carbon reduction potential. The present study suggests that further steps to promote sustainability and milk productivity are embodied when the replacement of fossil fuels and the management standardization are adapted to regional characteristics.

1. Introduction

Considered as the largest energy consumer and carbon emitter, China accounts for more than 30% of global greenhouse gas (GHG) emissions. To achieve the carbon-neutral goal by 2060, broader and more intensive mitigation actions need to be taken [1,2]. In 2018, the GHG emissions from the livestock sector are about 298 Tg CO₂ eq. in China, nearly half of that from agriculture [3]. As an important part of animal husbandry, dairy farming has received special attention and scrutiny for its environmental impact [4,5].

In 2019, the national milk productivity exceeded 32 Tg, with a stock of 10.45 million head cows [6]. The farms with more than 100 head cows reached 64%, showing an increase of 31% compared with this value in 2011 [7]. The increase in herd size leads to the intensification of dairy farming, which is negatively correlated with ecological efficiency [8]. With the increase in quality control and environmental protection, the proportion of free-range farming decreases notably and intensive farms take on more production tasks [9]. To fit this situation, the application of machines and standardized management spread rapidly [6]. The mechanization and electrification of dairy farming reduce labor costs and result in a significant increase in direct energy consumption and GHG emissions. In on-farm production, direct energy consumption mainly involves feed production, wastewater processing, cattle management, milk extraction, and water supply [10].

While electricity is mainly used for cattle management in barn, milk extraction, and pumping water, diesel is consumed by farm machineries, such as tractors, trucks, and heavy machineries with diesel engines [11]. Houston et al. [12] studied the direct energy consumption of small farms in Canada and reported that electricity is mainly used for cooling, illumination, and ventilation. Through a review of energy consumption in dairy farms, Shine et al. [13] concluded that 48.9 kWh is consumed to produce 1 ton of milk on average and milk cooling is the largest consumer of electricity. Pagani et al. [14] investigated direct and indirect energy consumption of dairy farms in Missouri, USA and in Emilia-Romagna, Italy. They found that on average, an increase of 1 GJ in energy consumption can lead to a rise in milk productivity of 115 kg per cow per year in America, and 149 kg in Italy. Moerkerken et al. [15] studied representative dairy farms in Netherlands and reported that the energy efficiency is 56.3 kWh per ton of milk for small farms, and medium and large farms are 49.3 and 42.6 kWh per ton of milk, respectively. The current situation in Netherlands manifested that the intensification and mechanization deteriorate energy efficiency with 16 kWh per ton of milk, but the promotion of sustainability is expected to eliminate that negative impact.

To reveal the influence of farm management on the carbon footprint of milk production and financial performance for dairy farms in Ontario, Canada, Jayasundara et al. [16] calculated the GHG emissions from intensive farms and pointed out that large-scale farms are more likely to reduce CE and improve productivity. Ilyas et al. [17] studied dairy farms in New Zealand and reported that the carbon footprint of energy consumption in barn system is 11% higher than grazing system. That difference is due to imported feed supplements, machinery usage, and fossil fuel consumption for on-farm activities. Zira et al. [18] studied dairy and beef farming systems in southwestern Europe and indicated that the carbon emission equivalent is 1.4~1.5 kg CO₂ eq. per kg of fat and protein corrected milk (FPCM) at farm gate. Based on survey data in Shandong and Heilongjiang, Huang et al. [19] estimated the nationwide GHG emissions of dairy farming to be about 45 Tg CO₂ eq. They attributed this to that half of dairy farms in China adopt high recycling system and further propelling of which is able to create a reduction of 7~21% in GHG emissions. Wang et al. [20] evaluated the potential mitigation options at farm level in Guanzhong plain, they reported that direct energy usage and feed transportation are responsible for 3~9% and 1~2% of the total GHG emissions, respectively. However, the management mode, the scale of dairy farms, and the breed of cattle vary by climate and geographical conditions [21,22,23]. Simultaneously, to optimize the distribution of milk production, the government integrates the dominant factors and makes the construction focus on the regions with high-quality milk sources, such as northeast regions, Inner Mongolia, and north and central China [7]. Policy bias causes differences in the process of mechanization and management standardization in provinces, which is reflected in a variety of practices in dairy farming [24]. For most cattle farms, technology upgrading and management optimization have a great potential to improve energy efficiency [16,25]. Additionally, by replacing fossil fuels with electricity and taking advantage of clean energy such as biogas and solar energy, farms are able to promote the composition of energy utilization and thus reduce carbon emission (CE) [26]. For the energy-environment performance in China’s agricultural sector, although the technology in Shanxi and Yunnan are at the forefront under contemporaneous frontier, most provinces do not perform efficiently in energy utilization referring to global frontier [27]. From the perspective of improving ecological efficiency, it is difficult to formulate targeted policies and support measures while the differences in CE caused by various practices are unclear [28,29,30]. The existing studies related to dairy farming are insufficient to provide valid conclusions.

Therefore, the aim of this study is to quantify CE from direct energy consumption of dairy farming through on-farm practices. Forage planting, manure application, and energy transposition are not included. To determine the ecological benefit, we calculated and compared the carbon emission intensity (CI) by farm scales and provinces, which refers to CE per kg of milk. These were determined by (1) calculating and illustrating the CE related to direct energy consumption of dairy farming in China, (2) demonstrating the inter-provincial distribution of CE and CI, and (3) analyzing the carbon reduction potential (CRP) and the impact factors in the main producing area. The analysis results also allow us to make recommendations on carbon reduction related to energy utilization and management practices in cattle farms.

2. Materials and Methods

2.1. Data Acquisition

The conditions of cattle breeding and farm management in 2019 were obtained from the China dairy yearbook [7]. The corresponding data of milk productivity and direct energy cost were from the Ministry of Agriculture and Rural Affairs of the People’s Republic of China (MOA) [9]. Based on the current breeding situation, the number of lactating cows is assumed to be 50% of the total stock, and the lactating period is assumed to be 305 days per year [9]. According to the MOA and the Dairy Association of China (DAC), dairy farms are categorized into four groups by herd size: backyard (1~9 heads), small-scale farm (10~49 heads), medium-scale farm (50~499 heads), and large-scale farm (500 heads and above). The environmental temperatures were obtained from the National Meteorological Information Center, which are the average values for provinces from 1981 to 2010 [31]. The regions in China are divided into four classes in terms of arid and humid conditions: humid region (HR), semi-humid region (SHR), semi-arid region (SAR), and arid region (AR) by the People’s Republic of China yearbook [32]. The specific situation of the regional division is shown in

Table 1. Characteristics of four classes of arid and humid regions in China.

Arid and Humid Conditions	Aridity Index	Annual Precipitation (mm)	Natural Vegetation	Land Utilization
HR	≤1.00	>800	Forest	Paddy field mostly
SHR	1.00~1.50	>400	Forest and Grassland	Dryland mostly
SAR	1.50~4.00	<400	Grassland	Animal husbandry and irrigation agriculture
AR	≥4.00	<200	Desert	Alpine animal husbandry and oasis irrigated agriculture

Sources: the Government of China.

. The aridity index refers to the ratio of the annual mean of potential evapotranspiration to the annual mean of precipitation.

Based on the given information, the direct energy costs in this study consist of electricity, coal, and diesel. The direct energy consumption that are less than 1% are not included [13,33]. The electricity prices for provinces are the flat segment prices for the purpose of agriculture, which were collected from the State Grid Corporation of China [34]. The price of coal is the base price, and the price of diesel is the national average in 2019, both of which were collected from the National Development and Reform Commission [35].

2.2. The Calculations of CE and CI

As the direct energy costs per cow were collected by different farm scales in provinces, the direct energy used per cow could be calculated based on the prices of energy. Given the increasing proportion of installed clean energy power generation [36,37], the national carbon emission factor of electricity is 0.5839 t CO₂ eq./MWh, which is used by the Ministry of Ecology and Environment of China in 2020 [38]. The GHG emissions of coal and diesel are the sum of methane (CH₄), nitrous oxide (N₂O), and carbon dioxide (CO₂) emissions. According to the corresponding default emission factors (EF) and the 100-year global warming potential (GWP) value from the Intergovernmental Panel on Climate Change [39,40,41], we calculated the carbon emission equivalents of coal and diesel. CEE, CEC, and CED stand for carbon emissions from electricity, coal, and diesel consumption, respectively. And j = (1, 2, 3) stands for CH₄, N₂O, and CO₂, respectively.

$$\begin{array}{c}CEE=electricity \ast 0.5839\\ CEC=coal \ast \sum E{F}_j\ast GW{P}_j\\ CED=diesel\ast \sum E{F}_j\ast GW{P}_j\\ C{E}_{p,i}=CE{E}_{p,i}+CE{C}_{p,i}+CE{D}_{p,i}\end{array}$$

(1)

The functional unit (FU) of this study is kg CO₂ eq. per kg of milk, referring to the annual CO₂ equivalent generated by direct energy consumption to obtain 1 kg of milk. The carbon emission of each province (CE_p) is calculated based on the stock X_p of lactating cows, the breeding proportion μ_p,i of different scales, and the corresponding CE_p,i per cow, when i = (1, …, n) stands for scale groups of farms.

$$\begin{array}{c}C{E}_p=X_p\ast {\displaystyle \sum }_{i=1}^n{\mu }_{p, i}\ast C{E}_{p,i}\\ \mathrm{s}.\mathrm{t}.{\displaystyle \sum }_{i=1}^n{\mu }_{p,i}=1\\ 0<{\mu }_{p,i}\le 1\\ 1\le n\le 4\end{array}$$

(2)

The carbon emission intensity of each province (CI_p) is the ratio of the weighted average of CE_p,i per cow per year to the weighted average of milk productivity Y_p,i per cow per year, where the weight is the breeding proportion μ_p,i in the province.

$$\begin{array}{c}C{I}_p=\frac{{{\displaystyle \sum }}_{i=1}^n{\mu }_{p,i}\ast C{E}_{p,i}}{{{\displaystyle \sum }}_{i=1}^n{\mu }_{p,i}\ast Y_{p,i}}\\ \mathrm{s}.\mathrm{t}.{\displaystyle \sum }_{i=1}^n{\mu }_{p,i}=1\\ 0<{\mu }_{p,i}\le 1\\ 1\le n\le 4\end{array}$$

(3)

2.3. Main Milk-Producing Area and CRP Analysis

In this study, there are data from 21 provinces collected in detail. We screened out 14 provinces as the main producing area for having the characteristics as follows: (1) they all hold Holstein as the dominating cattle breed according to the DAC, and (2) most of the dairy farms in the province are medium scale and large scale [7]. In consequence, the CRP of energy consumption is capable to be analyzed on the basis of a production year and individual subject. According to the framework of total factor energy efficiency, CE is an undesirable output [42]. Thus, the CRP is expressed as the ratio of the difference value between actual output (AO) and intended output (IO) to IO.

$$\begin{array}{c}IO=\frac{1}{n}{{\displaystyle \sum }}_{i=1}^{n}C{E}_{cow,i}\\ A{O}_p=C{E}_{cow,p}\\ CR{P}_p=(A{O}_p-IO)/IO\end{array}$$

(4)

In Formula (4), CRP_p is the carbon reduction potential of each province. Additionally, AO_p is the actual CE from direct energy consumption per cow per year in the province, which means the IO is the average of CE_cow in the main producing area. CRPE, CRPC, and CRPD denote carbon reduction potential of electricity, coal, and diesel consumption, respectively.

2.4. Data Process and Analysis

The data were processed in Microsoft Excel (Microsoft Excel, 2019). The results were imported into Python (Pycharm, 2021) for geographic information system (GIS) analysis of the inter-provincial distribution. We used SPSS (IBM SPSS Statistics, 25.0) for correlation analysis and Origin (OriginLab, 2021b) to present the results. All data were tested for normality and homogeneity of variance. The results were expressed as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was used to compare CIs among farms with different herd sizes. The double-tail correlation between CEE and environmental temperature was tested at provincial level in the main milk-producing area, as well as direct energy cost and milk productivity. Within the 95% confidence interval, p < 0.05 is a significant correlation.

3. Results

To illustrate the state of CE from dairy farming in China, we analyzed the results from the four dimensions including the national level, farm scale, inter-provincial distribution, and main producing area. As CE is closely related to direct energy consumption, the characteristics of energy use are also illustrated.

3.1. Overview of CE from Dairy Farming

In 2019, 100% of intensive farms in China are equipped with milking machines, and 95% have applied tractors for total mixed ration. Consequently, the contents of fat and protein in milk sampling from intensive farms are higher than the national average, and the somatic cell count (SCC) and total bacteria count (TBC) are lower than the national level. The quality parameters of milk sampling from intensive farms and the national average are presented in

Table 2. Quality parameters of milk sampling from intensive farms and the national average.

Quality Parameters	Unit	National Average	Intensive Farms
PC	g/100 g	3.25	3.34
FC	g/100 g	3.82	3.91
SCC	104/mL	28.71	22.75
TBC	104 CFU/mL	27.53	6.06

PC is milk protein content, FC is milk fat content, SCC is somatic cell count, and TBC is total bacteria count. Sources: Ministry of Agriculture and Rural Affairs of the People’s Republic of China.

.

The nationwide overview of CE from direct energy consumption of dairy farming is presented in Figure 1. The total CE is 2.4 Tg CO₂ eq. in 2019. The proportion of CEE varies by provinces from 25% to 100%, and CEC from 0 to 75%. For the country, the proportions of CEE and CEC are 60% and 37%, respectively. In some provinces, e.g., Hebei, Shanghai, Zhejiang, and Fujian, CEE accounts for 100% of the total CE, indicating that electricity is the only energy source. Meanwhile, more than 70% of CE is from coal consumption in Ningxia, Shaanxi, and Shanxi. Although CED only accounts for 3% in general, but it is extremely high in Qinghai reaching 54%. In Chongqing and Sichuan, CED accounts for 14% and 21%, respectively. It is worth noticing that almost no CEC is produced in provinces where CED is high, which implies farms tend to exclude coal as energy source in these regions.

3.2. CIs of Dairy Farms with Different Herd Sizes

Among 21 provinces, only four have a certain amount of backyard farms, while nearly half of them maintain small-scale farms. The occupancy of medium-scale and large-scale farms are 81% and 76%, respectively. In Figure 2, the minimum CI is found on small-scale farms in Ningxia (0.007 kg CO₂ eq. per kg of milk), while the maximum occurs on the large-scale farms in Jiangsu (0.233 kg CO₂ eq. per kg of milk). There are a considerable number of dairy farms of all four sizes in both Shanxi and Henan. Although the medium-scale farms in Shanxi and Henan have similar stock and CE, the CI of medium-scale farms is the highest in Shanxi, while it is the lowest in Henan. In Shanghai, Guangdong, and Qinghai, only large-scale farms are abundant, while Chongqing and Sichuan are dominated by medium-scale farms. Notably, the CIs of different farm scales in Xinjiang are obviously different (about 12.4 times at most), while the CIs of all scales in Hebei are low.

In terms of the mean value of CI from direct energy consumption (

Table 3. Attributes of CI from direct energy consumption in dairy farms with different herd sizes from 21 provinces. Max = maximum, Min = minimum.

Farm Scale	CI (kg CO2 eq. per kg of Milk)
	Max	Min	Mean ± SD
Backyard	0.104	0.017	0.054 ± 0.038
Small scale	0.076	0.007	0.042 ± 0.022
Medium scale	0.228	0.010	0.062 ± 0.050
Large scale	0.233	0.009	0.079 ± 0.057

Sources: calculation results in this study.

), the small-scale group is the lowest, and its distribution is the most concentrated in the four groups. By contrast, the mean value of the large-scale farms is the highest, it also has the largest range of variation. The CI of the large-scale group is nearly twice that of the small-scale group, the large-scale farms fail to present the due advantage of energy utilization and CE on the whole. The results also suggest that the management mode of small-scale farms is more mature, while the practices of large-scale farms vary greatly among provinces. ANOVA result shows no significant difference among all groups (p > 0.05), indicating that no significant correlation is found between farm scale and CI.

3.3. The Inter-Provincial Distribution of CI and CE

Based on limited data, this study only analyzed dairy farms in 21 provinces. The average CI of these provinces is 0.069 kg CO₂ eq. per kg of milk while the national value is 0.068 kg CO₂ eq. per kg of milk. The difference between the average and the national value is 0.001 kg CO₂ eq. per kg of milk, which is 1.47% of the national value. It is within the acceptable limit. The CI and CE of each province are shown in Figure 3.

The CIs of provinces vary from 0.009 to 0.216 kg CO₂ eq. per kg of milk (Jiangsu to Hebei). From the GIS analysis of the inter-provincial distribution, provinces in SAR and SHR have lower CIs than in AR and HR. The level of CIs is relatively higher in SHR than in SAR. The CIs of coastal provinces are higher than that of inland provinces at the same latitude, such as Shandong versus Henan and Jiangsu versus Anhui.

To the total CE of each province, the stock of dairy cattle is also a crucial factor. Xinjiang has 1.54 million heads of dairy cattle, ranking the highest in the country. Although the CI of Xinjiang is merely 0.082 kg CO₂ eq. per kg of milk, its CE is far ahead and 52.1% higher than that of Shandong. On the contrary, Jiangsu and Guangdong have low CEs (115.19 and 19.34 Gg CO₂ eq., respectively) due to the small stock despite their CIs being relatively high.

3.4. CRP in the Main Milk-Producing Area

According to the criteria in Section 2.3, the 14 provinces that have the criteria of main milk-producing area are distributed mainly in the middle, eastern, and northeast of SAR and SHR, and a small part in AR and HR. The average productivity of milk in the area is (7.19 ± 1.30) ton per cow per year. At the level of high productivity, the baseline of CEE, CEC, and CED are 291.24, 203.40, and 10.52 kg CO₂ eq. per cow per year, respectively. In Figure 4, the negative results signify less environmental impact than the baseline.

According to Figure 4, There are nine provinces in total have replaced diesel with electricity for showing −1 as the result of CRPD. Four provinces have scored −1 at CRPC. Although provinces concentrate in the middle region that are supposed to pay more attention on CRPC, Jiangsu is special for being the only outsider. It is revealed that provinces in northeast region have advantage in CRP while the eastern region is at the inferiority position. Hebei, Liaoning, and Heilongjiang have an excellent performance on CRP for having a negative output in CRPE, CRPC, and CRPD. Further, Hebei is noteworthy for having negative CRPE among the three provinces using pure electricity as energy source. Alternatively, in the five provinces having positive CRPE, Jiangsu has the highest CRPC at 4.81 and Shandong has the highest CRPD at 5.49, representing the extra CE compared with the baseline of CEC and CED, respectively. The CRPD of Jiangsu (3.21) is also unsatisfactory.

3.5. Impact Factors in the Main Milk-Producing Area

For intensive farming in barns, ambient temperature has a significant impact on dairy cattle. Generally, the higher the temperature is, the more energy will be consumed by ventilation and cooling to mitigate heat stress, resulting in higher CEE and direct energy cost per cow. Correspondingly, cattle farms in cold regions provide heating measures to calves and lactating cows in winter to reduce the occurrence of diarrhea and other diseases. As little coal and diesel are used for heating in winter, only CEE is analyzed in this study.

The analysis results are shown in Figure 5. In the main producing area, the average temperature of the hottest month is the highest in Zhejiang (28.5 °C), while it is the lowest in Ningxia (17.7 °C). At the same time, the CEEs of them are also the extreme values (645.44 versus 70.44 kg CO₂ eq. per cow per year). In the coldest month, the average temperature of Heilongjiang (−28.2 °C) is the lowest while that of Shanghai (4.3 °C) is the highest. However, the CEEs of them rank 12th and second (90.84 and 618.26 kg CO₂ eq. per cow per year), respectively. The CEE per cow shows a significant correlation with the temperature of the hottest month (p < 0.01), and the analysis with the temperature of the coldest month does not show the expected result. The direct energy cost per cow shows a significant correlation with milk production (p < 0.01), while CE shows no correlation with it (p > 0.05).

4. Discussion

4.1. Nationwide Overview of Dairy Farming

For the sustainable development of dairy farming, targeted policies that are compatible with the current situation should be made. According to previous research, the result verifies that intensive farming in barns is more prevalent in China [17]. It is easier for intensive farms to promote breeding environments and sanitary conditions, which is conducive to guaranteeing animal health and milk quality. In addition, more automation equipment and key technology could be applied to production in intensive farms. Additionally, intensive farming enables farms and regions to arrange centralized processing of feed and manure, which is more efficient and effective, and appealing to preferential policy and inclined resources. Accordingly, intensive and mechanized production became the mainstream in dairy farming. It is the common choice of the government and farmers.

The total CE from direct energy consumption of dairy farming is consistent with previous studies [19,20], which discussed the GHG emissions for dairy farms based on onsite surveys and investigations in Shandong, Heilongjiang, and Guanzhong Plain. Accordingly, there are about 50% of dairy farms in China should pay more attention to the environmental impact based on the current situation. The energy structure varies significantly among provinces. In addition, provinces with similar cattle stock, such as Shanxi and Henan, present different CE when the energy technology are applied on disparate degrees [27].

In a long time, as management practices in small-scale dairy farming adapt to regional characteristics, the impact of climate on farm productivity and efficiency decreases [43]. However, medium-scale and large-scale farms are still new rising things in this department, especially for farms that maintain cows over 10 thousand heads [7]. Due to great differences exist among provinces, the development of management practices is unbalanced and inadequate, which is incompatible with the rapid amplification of farm scale. With dry land and little precipitation, more energy is required to fulfill water supply in AR [44]. Alternatively, farms in HR take on more pressure of cooling and ventilation in cowsheds, resulting in high energy consumption and CI. Influenced by marine monsoon, coastal and inland provinces present differences in CI as well. Therefore, disparate goals and ecological policies should be set to adapt to regional differences.

To adapt to the sufficient sunshine and lack of water, small-scale farms and grazing are common in provinces such as Xinjiang and Xizang, and more than 2/3 of the cow stock is dual-purpose cattle [7]. That leads to low milk productivity and direct energy consumption at the same time. However, the breeding of local livestock breeds maintains the diversity of ecological species. It is also widely proven that suitable intensity and duration of cattle grazing is beneficial to ecosystem function maintenance and carbon sequestration [45,46,47]. In addition, the diverse milk sources and dairy products enrich the supply structure of the milk market. In the long term, they are indispensable to the healthy development of dairy farming. Sichuan, Yunnan, and other places raising milk yak, milk buffalo and other breeds are also suitable for this situation. For agriculture systems with such merits, the relative improvement of their ecological benefits should be pursued, rather than absolute limits [8,48].

4.2. Dairy Farming in Main-Milk Producing Area

Similar to other industries [49], the spatial aggregation effect is also visible on dairy farming. To adapt to the rapid amplification of farm scale, more efficient and environmental management modes should be standardized and applied according to regional characteristics.

In recent years, it has become a consensus in scale farms to relieve the heat stress of cows using fan cooling in summer [50,51]. Alternatively, heating practices differ greatly among cattle farms. Heat supplements include hot water pipelines, electric heaters, heating lamps, etc. Some farms supplement heat to cow sheds to relieve cold stress and improve animal welfare, while others do not pay enough attention to this problem. Due to the contradiction between heating and ventilation in winter, neither the calves shed nor dairy shed have formed a widely approved management mode [51]. As cold stress and disease affect the weight and behavior of pre-weaned calves, and poor ventilation increases the risk of airborne disease transmission [52], ventilation strategy and heating technology in winter need to be upgraded, especially in the northeast region.

Normally, more energy cost accompanies more investment in machines and equipment, which implies a farm has better management ability reflected in higher productivity [53]. According to the results, Jiangsu, Zhejiang, Shanghai, and Anhui fit this situation. However, Shanghai manages to perform well in CRP at the same time. Compared with other provinces, Shanghai has advanced CE technology and management systems [42], which is embodied in green energy source and standardized management in dairy farming. That is conducive to high productivity and energy efficiency [7].

According to the research on regional energy efficiency in China, Jiangsu and other eastern provinces are in the superior position of both energy technology and efficiency [54,55]. Consequently, the cooling and ventilation stress and high consumption of fossil fuels should take primary responsibility for unsatisfactory results. For farms in the middle and eastern regions, the replacement of fossil fuels with electricity is helpful to reduce CRP. Further, the production and usage of clean energy in dairy farms, such as biogas fermentation, combined heat and power technology (CHP), and photovoltaic power, is proven to be beneficial both economically and environmentally [7,15]. As reported by DAC, in addition to the advanced management, all dairy farms in Hebei are equipped with manure treatment facilities. Accordingly, biogas fermentation and CHP are promoted on a large scale to recycle waste and reduce direct energy consumption [7]. Although these measures may cause an increase in initial investment, the promised high productivity and energy efficiency are very persuasive [56], which increases the competitiveness of enterprises in the long view.

5. Conclusions

To support the sustainable planning strategy of dairy farming, this study calculated CE from direct energy consumption and analyzed the impact of environment, production, and management on it.

Overall, electricity consumption is the main source of CE for the country and energy structure varies greatly among provinces. By the common choice of the government and farmers, intensive dairy farming in barns is more prevalent in China. From the perspective of farm scale, the management mode of small-scale farms is more mature than others. Alternatively, the development of management practices adapted to regional characteristics is supposed to improve the undesirable situation of large-scale farms. As depicted by the GIS analysis of CI, SAR and SHR are more suitable for dairy farming than AR and HR, and inland provinces have lower CIs than coastal provinces. For provinces breeding local species, characteristic breeding should be encouraged to maintain ecosystem function and various milk supplies.

In the main milk-producing area, the spatial aggregation effect is visible. Provinces concentrated in the middle region are expected to work on CRP. Additionally, provinces in the northeast region have an obvious advantage in CRP while the eastern region is at the inferior position. In addition, high temperature has a significant influence on CEE, and more attention needs to be paid to the environmental regulation of cattle sheds in winter, especially in the northeast region. In general, elevating the direct energy cost is helpful to improve productivity. On this basis, improving and popularizing standardized management will further increase milk productivity and energy efficiency. The replacement of fossil fuels with clean energy is conducive to the optimization of energy structure and the application of energy technology, which could promote the energy-environment performance and the sustainability of dairy farming.