Environmental impacts and resource use of milk

Science of the Total Environment 625 (2018) 486–495

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Environmental impacts and resource use of milk production on the North China Plain, based on life cycle assessment Xiaoqin Wang a,⁎, Stewart Ledgard b, Jiafa Luo b, Yongqin Guo c, Zhanqin Zhao c, Liang Guo a, Song Liu a, Nannan Zhang c, Xueqin Duan a, Lin Ma c,⁎ a

College of Natural Resources and Environment, Northwest Agriculture and Forestry University, Yangling 712100, China AgResearch, Ruakura Research Centre, Hamilton 3240, New Zealand Key Laboratory of Agricultural Water Resources, Center for Agricultural Resources Research, Institute of Genetic and Developmental Biology, The Chinese Academy of Sciences, Shijiazhuang 050021, Hebei, China

b c

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Environmental burden of milk production in China was evaluated using LCA. • Feed production significantly affected global warming, energy use, land use and blue water use. • Acidification and eutrophication potentials were determined by manure management. • Environmental burden can decrease with improved cow productivity and herd structure. • Environmental burden shifting was observed with different feeding options.

a r t i c l e

i n f o

Article history: Received 28 September 2017 Received in revised form 21 December 2017 Accepted 21 December 2017 Available online 29 December 2017 Editor: Simon Pollard Keywords: Global warming Eutrophication Acidification Dairy Manure management Feed strategy

a b s t r a c t Life cycle assessment methodology was used to quantify the environmental impacts and resource use of milk production on the North China Plain, the largest milk production area in China. Variation in environmental burden caused by cow productivity was evaluated, as well as scenario analysis of the effects of improvement practices. The results indicated that the average environmental impact potential and resource use for producing 1 kg of fat 2 and protein corrected milk was 1.34 kg CO2eq., 9.27 g PO3− 4 eq., 19.5 g SO2eq., 4.91 MJ, 1.83 m and 266 L for global warming potential (GWP), eutrophication potential (EP), acidification potential (AP), non-renewable energy use (NREU), land use (LU) and blue water use (BWU; i.e. water withdrawal), respectively. Feed production was a significant determinant of GWP, NREU, LU and BWU, while AP and EP were mainly affected by manure management. Scenario analysis showed that reducing use of concentrates and substituting with alfalfa hay decreased GWP, EP, AP, NREU and LU (by 1.0%–5.5%), but caused a significant increase of BWU (by 17.2%). Using imported soybean instead of locally-grown soybean decreased LU by 2.6%, but significantly increased GWP and NREU by 20% and 6.9%, respectively. There was no single perfect manure management system, with variable effects from different management practices. The environmental burden shifting observed in this study illustrates the importance of assessing a wide range of impact categories instead of single or limited indicators for formulating environmental policies, and the necessity of combining multiple measures to decrease the environmental burden. For the North China Plain, improving milking cow productivity and herd structure (i.e. increased proportion of milking cows), combining various manure management systems, and encouraging dairy farmers to return manure to nearby crop lands are promising measures to decrease multiple environmental impacts. © 2017 Elsevier B.V. All rights reserved.

⁎ Corresponding authors. E-mail addresses: [email protected] (X. Wang), [email protected] (L. Ma).

https://doi.org/10.1016/j.scitotenv.2017.12.259 0048-9697/© 2017 Elsevier B.V. All rights reserved.

X. Wang et al. / Science of the Total Environment 625 (2018) 486–495

1. Introduction Milk production plays an important role both for the human diet and the economy in most countries (Muehlhoff et al., 2013). However, the global dairy sector also has a major effect on environmental degradation and resource depletion (Steinfeld et al., 2006; Chobtang et al., 2016), not only as one of the most important anthropogenic sources of greenhouse gas (GHG) emissions (Pirlo, 2012; Gerber et al., 2013), but also as a significant contributor to other impacts including eutrophication, acidification and water scarcity (Steinfeld et al., 2006). China is the third-largest global milk producer (Hagemann et al., 2012) with an increasing milk production and consumption (DAC, 2015), while at the same time facing a huge challenge from various resource and environmental issues. Thus, it is critical to estimate the environmental impacts and resource use of milk production in China and investigate potential measures for decreasing the resource use and environmental burden on the nation. Life cycle assessment (LCA) is a widely used tool to estimate the environmental impact and resource use of products and reveal environmental hotspots through the whole supply chain (Finnveden et al., 2009). Application of LCA in dairy production studies has increased recently, but most focus has been on dairy systems in European or other OECD (Organization for Economic Co-operation and Development) countries (Yan et al., 2011; Chobtang et al., 2016; Baldini et al., 2017). There are very few studies on Chinese dairy farming, which is different from other major milk-producing countries such as European countries, USA, India or New Zealand (Wang et al., 2016). Although several studies have included Chinese dairy farming, only GHG emissions and land use were considered in these studies (Gerber et al., 2013; Wang et al., 2016). Milk production potentially affects a wide range of resource and environmental impact categories (Meul et al., 2014; Battini et al., 2016; Chobtang et al., 2016), and a farm with a low effect on one impact category does not always have a low effect on other categories although they may be correlated (Yan et al., 2013). Environmental burden shifting can also happen between impact categories within the same farm when mitigation strategies are applied (de Boer et al., 2011). Thus, it is critically important to include multiple resource and impact categories (Chobtang et al., 2016) in a life cycle assessment of dairy production, especially when carrying out an analysis on abatement strategies (de Boer et al., 2011). Water is a key resource for dairy farming and water use (WU) has largely been investigated as a single issue. Livestock production is a large consumer of water resources (Ran et al., 2016) and 19% of animal WU is related to dairy production (Mekonnen and Hoekstra, 2010). Therefore, it is important to evaluate WU together with other multiple resource and environmental impact categories to understand the total effects of mitigation measures. The use of blue water (from surface and groundwater sources) is important, particularly for dairy farming in low rainfall regions where there is competition for water (Zonderland-Thomassen and Ledgard, 2012). Therefore, for the present study, resource use and environmental impact categories were selected according to critical issues in China and data availability, to quantify the resources and environmental burden of milk production on the North China Plain, the largest milk production area in China (DAC, 2015). The effects of feed strategies (i.e. feed type and source) and manure management systems on various impact categories were investigated.

all dairy farms in the area, animals are housed year-round, and are mainly fed on concentrates and maize silage. Most dairy farms don't have crop land on farm and buy in all feed. There is a large variation in productivity, farm scale and herd structure (the proportion of milking cows in the herd). Twenty-five dairy farms were selected to cover a range of different production systems on the North China Plain. Farm data were collected for one production year (2015/2016) (Table 1). As for most dairy farms on the North China Plain, all 25 farms bought in all their feed. Manure was stacked with infrequent turning for 0–3 months depending on season, and either sold to nearby orchards and vegetable farms or discharged. Female calves were raised as replacements for culled dairy cows, while all bull calves and surplus female calves were sold for meat production after calving. 2.2. The scope definition 2.2.1. System boundary Based on a cradle-to-farm gate perspective, a system boundary (Fig. 1) was defined, including feed crop production, processing of concentrates and transportation of feed to dairy farms. Animal components included enteric methane and manure management. Resource use on dairy farms accounted for land occupation, water and energy use. 2.2.2. Functional unit and allocation The functional unit used was defined as 1 kg of fat and protein corrected milk (FPCM) (IDF, 2015) at the farm gate. The dairy farms also sold surplus calves and culled cows for meat production, and part of the manure was sold to orchard or vegetable farmers. Thus, the total environmental burden was distributed between milk, meat and manure. As frequently applied in LCA of milk production (Baldini et al., 2017), an economic allocation method was used in the present study to handle dairy co-products, as well as for co-products associated with the production of feeds. 2.3. Inventory analysis 2.3.1. Feed production Feed production processes and related activities accounted for included: 1) fertilizer production; 2) application of N fertilizer resulting

Table 1 Mean values and range of farm characteristics of the 25 case–study dairy farms on the North China Plain for 2015/2016.

2. Materials and methods a

2.1. Dairy farming systems on the North China Plain The North China Plain is the second largest plain in China, located in the middle of east China, and is the largest milk production region. For

487

Item

Unit

Mean

SD

Min

Max

Total cattle Proportion of milking cows Milk productiona Surplus calves soldb Cull cows soldc

Number % of total cattle kg/cow number number

531 47 7132 124 51

334 7 1227 73 47

231 35 4753 43 12

1850 61 9157 370 200

Feed consumption (per cattle per year) kg dry matter Concentrated Maize silage kg dry matter Maize straw silage kg dry matter Alfalfa hay kg dry matter Leymus chinensis kg dry matter

1213 837 35 183 353

795 587 96 119 237

491 325 0 0 91

4096 3192 301 427 883

Resource use on farm (per farm per year) Electricity 1000 kWh Coal tonne Diesel l Land occupation ha Water (withdrawn) 1000 m3

156 0.8 6922 7 24

85 2.9 3234 4 2

60 0 2471 2 9

329 10 16,000 17 70

The price of 1 kg milk is 3.8 Yuan. Calves are sold for 500 Yuan per head, and a cull cow for 5500 Yuan. The weight of calves is 38 kg, and cull cows 550 kg. d According to type and brand of concentrate, the ingredients were maize grain, soybean meal, wheat bran, cotton seed meal, rape seed meal and others, accounting for 40–57%, 7–26%, 9–15%, 3–18%, 0–9% and 5–10% of the total, respectively. b,c

488


Fig. 1. System boundary for the dairy farming system on the North China Plain. The emissions of pollutants from milk production came from feed production, transport of feed, enteric fermentation, manure management and resource use on dairy farms.

in direct and indirect emissions of nitrous oxide (N2O), volatilization of ammonia (NH3) and nitrate (NO− 3 ) leaching; 3) irrigation, land occupation and energy consumption for farm operation and processing of concentrates. Data on crop yield and fertilizer input for maize silage and alfalfa were based on a farmer survey due to lack of national or regional data, while data for remaining crops were sourced from NDRC (2015). Data on irrigation water and energy consumption for field operation were based on the literature and farmer surveys (Table 2). For Leymus chinensis harvested from natural pasture by manual labour without fertilizer input, only land occupation and transportation were considered. The electricity consumption for processing rapeseed meal, soybean meal, cotton seed, cotton seed meal, wheat bran and concentrates were 48, 50, 150, 50, 48 and 30 kWh per tonne, respectively (MEPPRC, 2003; Bai, 2007). The coal consumption for processing soybean meal, rapeseed meal, cotton seed, cotton seed meal were 85, 40, 20 and 85 kg per ton, respectively (MEPPRC, 2003). The water use for processing soybean meal and concentrates was 1.2 m3 (MEPPRC, 2003) and 4.6 L (de Boer et al., 2013) per ton, respectively. Nitrous oxide emissions from application of N fertilizers were calculated based on IPCC (2006), using emission factors in Table 3. The inventory data on the fertilizer production were derived from Ecoinvent 3.0 (Ecoinvent Centre, 2013), and on energy source (electricity, diesel and

coal) production and consumption were from CLCD 0.8 (IKE and SCUISCP, 2015), He et al. (1996) and Di et al. (2005). 2.3.2. Feed transportation The maize silage, alfalfa hay and Leymus chinensis were transported to dairy farms by lorry from local area, Gansu province and Inner Mongolia, respectively. The average distances were 8, 1864 and 2134 km, respectively. Concentrates were transported about 30 km from feed factories to dairy farms by lorry, while concentrate ingredients of soybean meal, cotton seed meal and rapeseed meal were transported to feed factories 1621 km from the Heilongjiang province by freight train, 2933 km from the Xinjiang province by lorry, and 1454 km from the Qinghai province by lorry, respectively. Concentrate ingredients of wheat bran and maize grain were transported to feed factories about 5 km from local places, so the transportation effects were omitted. The inventory data on transport were derived from the Ecoinvent Database v3.0. 2.3.3. Enteric methane Enteric methane (CH4) emissions from animals were estimated from gross energy (GE) intake of feeds using the Tier 2 method of IPCC (2006), and a methane conversion factor (Ym) of 6.5%. GE intake of

Table 2 Annual per-hectare inputs and outputs for feed crops production.

Input: Mineral fertilizer, kg N Mineral fertilizer, kg P2O5 Mineral fertilizer, kg K2O Diesel, L Electricity, kWh Irrigation water, m3 Output, based on dry matter: Grain yield, kg Straw or silage yield, kg a b c d e

Liang et al. (2009). From farmer surveys. Knudsen et al. (2010). Zhang et al. (2010). NAHS (2014).

Maize grain

Maize silageb

Wheat

Soybean

Cotton seed

Rapeseed

Alfalfa hayb

163 84 46 45a 732a 1627a

180 74 79 47 730 1600

258 130 54 45a 1177a 2617a

61 67 29 28c 0c 0c

333 208 18 45b 1238b 7750b

110 88 1 30b 0b 3231d

90 140 39 121 1318 7567

6639 2476

– 16,616

5456 –

1923 –

2831 –

2121 –

10,035

Leymus chinensis

2348e

X. Wang et al. / Science of the Total Environment 625 (2018) 486–495 Table 3 Emission factors for calculating environmental burden of feed production.

N2O-Ndirect N2O-Nindirect NH3-N

NO3-N

a b c d

Source

Emission factor (EF)

Reference

Application of fertilizer (N) From NH3 (NH3-N) From leaching (NO3-N) Application of fertilizer (N)

0.0105 0.01 0.0075 0.048a 0.036b 0.192c 0.014d 0.03a 0.035b 0.033c 0.044d

Zhang et al. (2010) IPCC (2006) IPCC (2006) Wang et al. (2002)

Application of fertilizer (N)

Li et al. (2006) Xu et al. (2011) Li et al. (2006) Zhao et al. (2010) Wang et al. (2012)

Maize, maize silage, rapeseed. Wheat. Soybean, alfalfa. Cotton seed.

feeds was calculated based on the actual feed consumption including animal refusals. 2.3.4. Manure management Passive composting of manure was assumed for all 25 farms, based on a known common practice. CH4 emission from manure management was estimated using the IPCC (2006) Tier 2 method. For the maximum CH4 producing capacity (B0), the value 0.13 m3 CH4 kg−1 volatile solids (VS) was used as suggested by IPCC (2006) for Asian dairy cows. Factors for calculating VS are shown in Table A1 and the methane conversion factor (MCF) from IPCC (2006) was used (Table A2). N2O emission from manure management was estimated based on IPCC (2006) Tier 2, with N excreted by animals calculated as the difference between total N intake and N retention of the animals (Meul et al., 2014). Details of calculating N intake from feed and N retained by animals in milk and live weight gain, can be found in Wang et al. (2016). Due to the separation of crop lands and dairy farms, manure from dairy farms was not completely returned to crop lands, and it was reported that less than 60% of animal manure was utilized at a national level (DAC, 2015). Since the specific discharge rate from the 25 dairy farms was not available, it was assumed that 40% of manure after composting was discharged, and the N and P contained in discharged manure was lost by leaching and runoff to water bodies. The amount of N retained in manure after composting was estimated by subtracting the amounts of N in N2O and NH3 emitted during composting from total N excreted. The amount of P retained in manure after composting was calculated as the difference between total P intake from feed and the amount of P in milk and live weight gain. P intake was calculated from the dry matter intake (DMI) and P content in each feed stuff (Table A1). The P contents in milk and live weight gain were 0.09% and 0.92% (Ma et al., 2010), respectively. 2.3.5. On-farm resource use Resource use covered water, energy, and land use on dairy farms, which were obtained from the farm survey (Table 1). Energy was used for milking, refrigeration, feed mixing, feeding, and manure management. Data on land occupation and water use for drinking and cleaning were collected. The environmental impacts and resource use of cleaning chemicals in dairy farms were not included due to the lack of information. 2.4. Impact assessment The environmental impact and resource use categories selected in the present study were global warming potential (GWP in kg CO2eq./kg FPCM), eutrophication potential (EP in kg PO34 − eq./kg FPCM), acidification potential (AP in kg SO2eq./kg FPCM), non-

489

renewable energy use (NREU in MJ/kg FPCM), land use (LU in m2/kg FPCM) and blue water use (BWU in L/kg FPCM). Since the main contributing sources to EP were uncertain for China, a general method that accounted for N and P (Guinee et al., 2002) was used. Characterization factors for GWP, EP and AP are summarized in Table A3. There was no distinction between land types and therefore land use is the same as land occupation. Blue water withdrawn was the sum of irrigation water, drinking and cleaning water on dairy farms, and fresh water required for inputs production (i.e. fertilizers and fossil energy) and transport. Correlation analysis between the environmental burden indicators and farm system data was carried out using SPSS18.0. 2.5. Scenario analysis Scenario analysis was conducted to investigate the effects of changes in feed type, soybean feed source and manure management. Levels of feeding alfalfa hay of less than 3 kg/cow/day are widespread in China and have been identified as limiting to milk production and cow health (Liu et al., 2013). A review of studies by Wang et al. (2016) showed that replacement of 1.5 kg of concentrates with 3 kg of alfalfa hay increased milk production by 2 kg per cow per day. The effect of this change in feed was evaluated. China is one of the largest producers of soybean, but is also the major importer. Some Chinese dairy farms may use soybean meal processed from imported soybeans. The present study tested the effect of imported soybean compared to all soybean sourced locally (as assumed for the 25 farms). Imported soybean was mainly from USA, Brazil and Argentina, which accounts for 42, 39 and 15% of the total import, respectively (Anonymous, 2011; Yang, 2014; Wang et al., 2016). Inventory data for soybean production in Brazil was derived from da Silva et al. (2010) and was also used for soybeans from USA and Argentina due to lack of available detailed data. The GWP of soybean calculated using the inventory data from da Silva et al. (2010), excluding data associated with deforestation and transportation of soybean, is similar to the GWP of American soybean (Adom et al., 2012). Soybeans from Brazil are associated with deforestation, while Argentinian soybeans are partially produced on land converted from grassland and shrubland (FAO, 2010). Thus, GHG emissions related to land use change were added to the GWP of soybean meal sourced from the two countries, with values of 7.69 and 0.93 kg CO2eq·kg−1 of soybean meal (FAO, 2010), respectively. Soybeans from USA were assumed to be transported from Minnesota state to New Orleans by barges over a distance of 2757 km, then shipped to the Tianjin seaport in China over 18,680 km. Soybeans from Brazil were transported from Mato Grosso to Santos by lorry over 1931 km, then shipped to China over 21,483 km. Soybeans from Argentina were transported from Rosario to China by cargo over a distance of 21,746 km. Two manure management systems (a slurry system, widely applied in European countries, and an intensive composting system) were investigated to compare with the passive composting which is common in China, using data in Table A2. In addition, for the slurry system where transportation to land is more difficult, the effects of a discharge rate of 60% and 80% were examined. 3. Results and discussion The total environmental impacts and resource use from dairy production systems were partitioned into milk, meat and manure based on their economic contribution. The average allocation factor applied to milk was 0.95, with a range from 0.92 to 0.98. It is higher than the average allocation factor of 0.89 calculated by the biophysical allocation method (IDF, 2015). Environmental impact potential and resource use were calculated separately for the 25 farms. The means and standard deviations of

490


total impact potentials are presented in Table 4, and the average contribution of each process to the total impact potential is presented in Fig. 2.

3.1. Global warming potential The average GWP resulting from the production of 1 kg of FPCM was 1.34 kg CO2 eq., ranging from 0.95 to 1.88 kg. The largest contributor to this impact category was enteric fermentation, accounting for 50–55% of the total impacts. Feed production was the second-largest at 24–28%, manure management the third-largest at 9–11%, and the contributions of feed transportation and on-farm energy use were 5–7% and 4–8% of total impacts, respectively. The GWP kg−1 FPCM in the present study was within the range of 0.74–2.46 kg CO2 eq. reported by Asselin-Balencon et al. (2013) and Del Prado et al. (2013). It was below the average global value of 2.80 kg CO2 eq. (Gerber et al., 2013) and also below the value of 1.60 kg CO2 eq. reported for the Guanzhong Plain of northwestern China (Wang et al., 2016), but higher than that from research on more high-performing milk production systems (Flysjö et al., 2011; McGeough et al., 2012; Fantin et al., 2012; Battini et al., 2014; O'Brien et al., 2014). The main reason for the difference between the present and previous Chinese study is that the study of Wang et al. (2016) reflected the production in 2011 when the average milk yield per dairy cow was less than 6000 kg for the Guanzhong Plain, while in the present study the average milk yield was about 7000 kg per cow representing the productivity for the North China Plain in 2015.

3.2. Eutrophication potential −1 FPCM on avThe eutrophication potential was 9.27 g PO3− 4 eq·kg erage, and varied from 5.88 to 13.32. It was strongly influenced by manure management (average of 90% of the total impacts), which was related to emission of NH3, leaching and discharge of N and P from animal manure. Feed production accounted for 10% of total eutrophication due to emissions of NH3 and leaching of NO− 3 from mineral N fertilizer application. Feed transportation and on-farm energy use affected this category to a lesser degree, accounting for 0.09% and 0.01% of the total impacts, respectively. In total, almost 43% of the total eutrophication potential was associated with NH3 emissions, followed by the leaching and discharge of P (35%) and N (18%). The eutrophication value in this study was similar to results from several other studies (Thomassen et al., 2009; Guerci et al., 2013a; Guerci et al., 2013b), but higher than those in most previous studies (de Vries and de Boer, 2010; Penati et al., 2013; Meul et al., 2014; Battini et al., 2016). With the exception of the effect of allocation methods and milk productivity level, the main reason could be the difference in manure disposal and manure management system. In the present study, it was assumed that 40% of the manure was discharged to waterways from all 25 farms according to the statistical data (DAC, 2015), which is much higher than in most previous studies. Also, the passive composting system used in these 25 farms emits more NH3 than a slurry system with covers, which is commonly used in European countries.

Table 4 Mean values and range of environmental impacts and resource use for the 25 case–study dairy farms, expressed per kg FPCM. Impact/resource category

Unit

Mean

SD

Min

Max

Global warming potential Eutrophication potential Acidification potential Non-renewable energy use Land use Blue water use

kg CO2eq. g PO3− 4 eq. g SO2eq. MJ m2 L

1.34 9.27 19.5 4.91 1.83 266

0.22 1.70 3.41 0.74 0.35 38.4

0.95 5.88 12.98 3.73 1.16 211

1.88 13.32 27.3 6.73 2.49 377

3.3. Acidification potential The average acidification potential from the production of 1 kg of FPCM was 19.5 g SO2 eq., with a variation from 13.0 to 27.3 g SO2eq. As for eutrophication, manure management contributed the greatest proportion to the acidification, accounting for an average of 79% of the total impact, followed by feed production (18% of total impact). Feed transportation and on-farm energy use contributed less to this impact category, accounting for 1.5% and 1.3% of the total impact, respectively. NH3 (92% of the total impact) was the major substance that contributed to this impact category, mainly originating from manure management and the use of mineral N fertilizers. About 5% and 3% of the total impact were attributed to the emissions of NOx and SO2, respectively, mainly originating from the consumption of fossil energy sources for feed production, transportation and on-farm use. The average acidification potential in this study was similar to results of Haas et al. (2001) and Guerci et al. (2013a), lower than that of Penati et al. (2013), but higher than those in most previous studies (van der Werf et al., 2009; Castanheira et al., 2010; de Vries and de Boer, 2010; Fantin et al., 2012; Guerci et al., 2013b; Meul et al., 2014; Battini et al., 2016). The difference was caused partly by allocation methods and milk productivity levels, but especially by the differences in manure management system as described for eutrophication. 3.4. Non-renewable energy use and land occupation On average, the production of 1 kg of FPCM required 4.91 MJ nonrenewable energy, with a range from 3.73 to 6.73 MJ. The greatest contributor to energy use was feed production, accounting for an average of 64% of the total consumption, followed by on-farm energy use (19% of total) and feed transportation (17% of total). Energy use kg−1 FPCM in the present study was within the range of data in the literature (Thomassen et al., 2009; Guerci et al., 2013a; Guerci et al., 2013b; Nguyen et al., 2013; Meul et al., 2014; Battini et al., 2016). The total land used for milk production was 1.83 m2/kg FPCM, ranging from 1.16 to 2.49, of which 98% was required for cultivation of feed crops. Dairy farms only occupied about 2% of the total area, since these farms did not consist of any cropland and all their feed was bought-in, which is common in China (Wang et al., 2016). The value for land use in this study was within the range of results in the literature (de Vries and de Boer, 2010; Kristensen et al., 2011; Guerci et al., 2013b; Penati et al., 2013; Battini et al., 2016; Wang et al., 2016). 3.5. Blue water use The production of 1 kg of FPCM was associated with abstraction of an average of 266 L blue water, ranging from 211 to 377 L. As for land occupation, feed production was the process contributing most to the blue water use (BWU), accounting for 98% of BWU. On-farm use only contributed about 2% of the total BWU, mainly for animal drinking and cleaning. Blue water use from feed transportation and on-farm energy use were negligible, with both of them contributing less than 0.1% of the total use. BWU kg−1 FPCM in the current study was within the range of 16–436 L reported by Sultana et al. (2014), similar to the value of 249 L measured in an irrigated dairy system in a low-rainfall region of New Zealand (Zonderland-Thomassen and Ledgard, 2012), but higher than results evaluated in rain-fed grass-based systems or other dairy systems which depend less on highly irrigated feeds (Ridoutt et al., 2010; de Boer et al., 2013; Palhares and Pezzopane, 2015; Murphy et al., 2017). As discussed in most studies (Mekonnen and Hoekstra, 2012; Sultana et al., 2014; Murphy et al., 2017), BWU was determined by level of irrigation. Compared to studies with lower BWU, the 25 farms and most of their feed crops production area in the present study were located in a low rainfall area, requiring irrigation. However, the BWU for drinking and cleaning on farm in the present study was


491

Fig. 2. Relative contribution of processes in the dairy farm system to the impact categories.

5.1 L per kg of FPCM which is comparable with findings by de Boer et al. (2013) and Murphy et al. (2017). 3.6. Factors affecting the environmental performance of milk production on the North China Plain 3.6.1. Cow productivity, herd structure and size Several farm-attributes, which can easily be understood by farmers and can directly instruct farm practice, were selected for analysis of their association with the impact categories. A correlation analysis (Table 5) showed that both productivity of milking cows and their proportion in the herd (herd structure) had a strong negative relationship with all impact categories except for BWU, which had a weaker correlation. However, the combined impact of milking cow productivity and their proportion in the herd had a more significant correlation with all impact categories than each of them as a sole indicator. This means that farms with a high environmental burden per kg of FPCM on the North China Plain should focus on improving milking cow productivity and also herd structure, which was also suggested by Wang et al. (2016) for decreasing GHG emissions per kg of milk. There was no clear relationship between herd size (total cattle per farm) and all impact categories. Most dairy farms on the North China Plain don't have their own cropland, and pasture hay and most concentrate ingredients come from places more than 1000 km away. The separation of crop production and livestock system interrupts the loop of manure nutrients returning to crop production, in addition to crop farmers preferring mineral fertilizer, which results in a large proportion of manure discharged to the environment (Ma et al., 2012; Li et al., 2016). In the present study, the same discharge percentage was assumed for all 25 farms, no matter their size, due to lack of specific data. However, the larger the farm scale, the more difficult it would be to find enough fields nearby to utilize the manure, potentially resulting

Table 5 Correlation coefficients between environmental impact (per kg FPCM) and cow productivity, herd structure and size of dairy farming system on the North China Plain.

FPCM/milking cow Milking cows/total cattle FPCM/total cattle Total cattle

GWP

EP

AP

NREU

LU

BWU

−0.69 −0.67 −0.84 −0.08

−0.65 −0.56 −0.75 −0.03

−0.64 −0.55 −0.73 −0.04

−0.63 −0.62 −0.78 −0.08

−0.67 −0.61 −0.81 −0.07

−0.35 −0.47 −0.51 −0.01

in a larger proportion of manure discharged than with smaller scale farms. This suggests that limiting farm scale to a moderate size may decrease the eutrophication potential. 3.6.2. Effect of feed ration on environmental performance As discussed above, all impact categories with the possible exception of BWU, were significantly correlated with milking cow productivity which can be increased by improving the feed ration. It is well known that alfalfa hay is a high quality roughage to increase cow productivity, and the Chinese government is encouraging dairy farmers to use more alfalfa in the diet of dairy cows (Nie and Feng, 2013), which is also supported by the data collected from several experimental studies (Li et al., 2003; Yue et al., 2009; Guo, 2010). As in the study of Wang et al. (2016), an alternative feed ration based on adding 3 kg of alfalfa hay instead of 1.5 kg of concentrates was assumed to be used on the 25 dairy farms in the current study. The results showed when using the alternative feed ration, the environmental burden kg−1 FPCM decreased by 2.5%, 4.3%, 1.0%, 2.8% and 5.5% for GWP, EP, AP, NREU and LU, respectively, but BWU increased by 17.2% (Table A4). The decrease of the first five categories was due to their total environmental impacts increasing by less than 7.2%, while total milk production increased by 8.1% on average. However, the total BWU increased by 28% because alfalfa requires much more irrigation water in the major production area of the Gansu province which is an arid region in northwest China, than concentrates. This provides a good example of where a measure for abating some environmental issues such as global warming, doesn't reduce all impact or resource use categories, which was also observed by de Boer et al. (2011) and Yan et al. (2013). It is also important to note that the total impacts for each category increased due to the much greater DMI when using 3 kg of alfalfa to replace 1.5 kg of concentrates. Thus, where the aim is to effectively decrease the total environmental burden of dairy farming in a region, it is not enough to improve productivity solely, which should be combined with limiting the total number of cattle. 3.6.3. Effect of soybean source Compared with domestic soybean in China, imported soybean from the main exporters of USA, Brazil and Argentina, has slightly lower inputs and higher output yields (da Silva et al., 2010; Adom et al., 2012; Wang et al., 2016), which results in a smaller land requirement for producing 1 kg of soybean. However, environmental pressure associated with land use change for soybean production in Brazil and Argentina,

492


in addition to long distance transportation, results in a greater environmental burden from imported soybean versus domestic soybean. The scenario analysis showed that using imported soybeans increased GWP of milk production by around 20% (Table A4), of which 94% was from GHG emissions associated with land use change and the remainder from transportation of imported soybean. Non-renewable energy use increased by 6.9% due to international transportation. The impact categories of EP and AP were influenced less, with increases of 1.9% and 0.5%, respectively. Land use decreased by 2.6%. Thus, importing soybean decreased the pressure on the arable land shortage in China, but increased the environmental impacts and other resource use.

3.6.4. Effect of manure management on environmental performance Manure from the livestock sector is one of the important sources causing eutrophication of surface water bodies in China (MEPPRC, 2010). On most dairy farms, manure was just piled up, with a passive composting system, which results in a high percentage of manure discharged and nutrients lost. The Chinese government is pushing dairy farms to improve manure management systems and decreasing their environmental burden. In this study, the two manure management systems of intensive composting and a slurry system were compared with the current passive composting system. For the intensive composting system, it was assumed that no manure was discharged since the final product was convenient to store, sell and apply. In contrast, slurry is suitable for use in areas nearby dairy farms, but is less convenient for being applied than manure from passive and intensive composting systems, and as a result the percentage of discharge would be higher than the passive composting system. Two scenarios were created for the slurry system, one with a discharge rate of 60% and the other 80%, compared to 40% for the passive composting system. The results (Fig. 3) showed that intensive composting had the highest GWP, but the lowest EP. The slurry systems had a slightly higher GWP than the passive composting system, the lowest AP, but the highest EP. The effect of slurry system on eutrophication increased as the percentage of manure returning to crop lands decreased. However, it is difficult to decrease EP from slurry systems by decreasing the discharge

rate of manure since crop farmers prefer mineral fertilizer or commercial organic fertilizer (Li et al., 2016). There is no one manure management system that can solve all environmental issues and dairy farms should combine various manure management systems according to the season, regional crop structure and land availability. The government also should encourage crop farmers to use manure on crop fields.

4. Conclusions In the present study the variation in environmental burden of milk production on the North China Plain and the effects of mitigation practices were evaluated based on multiple impact and resource use categories. The average GWP, NREU and LU for milk were in the range of those from previous studies, while EP and AP were higher than those in most studies. The latter was due to the high discharge rate of manure and higher NH3 emissions from the passive composting system which is common in the region. Blue water use is higher than in rain-fed regions since on the North China Plain most feed crops are cultivated in arid areas and are irrigated. Feed production and manure management were environmental hot-spots for milk production on the North China Plain, and the combination of milking cow productivity and improved herd structure was the most important farm-attribute affecting the environmental burden per kg of FPCM. Feed production was an important contributor to environmental pressure, but environmental burden shifting was observed in the two feed strategies evaluated in this study. Replacing part of the concentrates with alfalfa decreased GWP, EP, AP, NREU and LU, but significantly increased BWU. Similarly, using imported soybeans decreased LU, but increased GWP and NREU. Due to variable effects from different management practices, there was no single perfect manure management system that could reduce all environmental impacts. Mitigation measures should focus on enhancing milking cow productivity and improving herd structure, combining various manure management systems according to the season, and regional land availability, and encouraging dairy farmers to return manure to nearby crop lands and increase efficiency of nutrient recycling. Since a decrease

Fig. 3. Relative environmental impact potential of different manure management systems compared with composting-passive windrow scenario.


in environmental impacts intensity doesn't mean a decrease in the total environmental burden, ideally, decreased impacts per kg FPCM should be associated with a limit to the total number of cattle to decrease the total environmental impacts of dairy farming in a region. It may also be worthwhile to limit herd size to reduce direct discharge of manure, resulting in decreased eutrophication. Further research is needed on other feed strategies to improve environmental performance of milk production in China and this should include evaluation of a wide range of impact categories to avoid burden shifting between environmental impacts or resources used.

493

Acknowledgements The authors gratefully acknowledge the managers of the 25 dairy farms for provision of farm data. This study was supported by the National Natural Science Foundation of China (Grant No. 41201588), the Fundamental Research Funds for the Central Universities (Grant No. QN2012039), the National Science & Technology Pillar Program during the 12th Five-year Plan Period (Grant No. 2012BAD14B11), the Program of International S&T Cooperation (Grant No. 2015DFG91990), and the National Natural Science Foundation of China (Grant No. 31572210).

Appendix A Table A1 Feed characteristics (based on dry matter content) for feeds used on Chinese dairy farmsa Feed

DE (%)b

Ash (%)

CP (%)c

P (%)d

Maize grain Wheat bran Soybean meal Cotton seed meal Rape seed meal Maize silage Maize straw silage Leymus chinensis Alfalfa hay

85.8 77.1 92.2 66.4 66 67.7 67.7 56 62.7

1.3 4.8 6.1 6.6 7.3 6 7 8 8

9.1 16.4 49.7 43.5 38.6 8 6 7 16

0.27 0.28 0.62 1.04 1.02 0.27 0.27 0.15 0.19

a b c d

Data were collected from Chinese Feed Database (IAS, 2013). Digestibility. Crude protein. Phosphorus

Table A2 Factors for calculating GHG emissions from manure managementa Manure management systems

MCF at 13 °C

EF3b

FracGasc

EF4d

Passive composting Intensive composting Slurry with natural crust

0.005 0.005 0.14

0.01 0.1 0.005

0.4 0.4 0.08

0.01 0.01 0.01

a

Sourced from IPCC (2006). Direct emission factor of nitrous oxide. Fraction of manure nitrogen that volatilizes from manure management systems. d Indirect emission factor of nitrous oxide from volatilization (kg N2O-N kg−1 NH3-N + NOx-N volatilized). b c

Table A3 Characterization factors for calculating the environmental impact potential Impact category

Unit

Contributing substance

Characterization factor

Reference

Global warming

kg CO2eq.

1 28 30 265

IPCC (2013)

Eutrophication

kg PO3− 4 eq.

kg SO2eq.

1 3.06 0.42 0.35 0.13 0.13 1.2 1.6 0.5

Guinee et al. (2002)

Acidification

CO2 CH4 (biogenic) CH4 (fossil) N2O PO3− 4 P N NH3 NOx NO− 3 SO2 NH3 NOx

Huijbregts (1999)

Table A4 Mean values of environmental impacts of milk production on the North China Plain under different feed strategies, expressed per kg FPCM. Feed strategies

GWP kg CO2eq.

EP g PO3− 4 eq.

AP g SO2eq.

NREU MJ

LU m2

BWU L

Current farms The alternative feed ration Imported soybeans

1.34 1.31 1.61

9.27 8.87 9.45

19.50 19.31 19.60

4.91 4.77 5.25

1.83 1.73 1.78

266 312 266

494


References Adom, F., Maes, A., Workman, C., Clayton-Nierderman, Z., Thoma, G., Shonnard, D., 2012. Regional carbon footprint analysis of dairy feeds for milk production in the USA. Int. J. Life Cycle Assess. 17, 520–534. Anonymous, 2011. A Report of Soybean Imported to China in 2011. Online at http:// www.doc88.com/p-3813719556956.html. Asselin-Balencon, A.C., Popp, J., Henderson, A., Heller, M., Thoma, G., Jolliet, O., 2013. Dairy farm greenhouse gas impacts: a parsimonious model for a farmer's decision support tool. Int. Dairy J. 31, S65–S77. Bai, L., 2007. The study on LCA of pork Production for Cleaner Production and Key Technologies of Swine Manure Treatment. Si Chuan Agricultural University, Ya'an, China, pp. 60–61 (in Chinese). Baldini, C., Gardoni, D., Guarino, M., 2017. A critical review of the recent evolution of life cycle assessment applied to milk production. J. Clean. Prod. 140, 421–435. Battini, F., Agostini, A., Boulamanti, A.K., Giuntolo, J., Amaducci, S., 2014. Mitigating the environmental impacts of milk production via anaerobic digestion of manure: case study of a dairy farm in the Po Valley. Sci. Total Environ. 48, 196–208. Battini, F., Agostini, A., Tabaglio, V., Amaducci, S., 2016. Environmental impacts of different dairy farming systems in the PoValley. J. Clean. Prod. 112, 91–102. Castanheira, E.G., Dias, A.C., Arroja, L., Amaro, R., 2010. The environmental performance of milk production on a typical Portuguese dairy farm. Agric. Syst. 103, 498–507. Chobtang, J., Ledgard, S.F., McLaren, S.J., Zonderland-Thomassen, M., Donaghy, D.J., 2016. Appraisal of environmental profiles of pasture-based milk production: a case study of dairy farms in the Waikato region, New Zealand. Int. J. Life Cycle Assess. 21, 311–325. CLCD: IKE, SCU-ISCP, 2015. Chinese core Life Cycle Database (CLCD) Version 0.8. Available at eBalance 4.7 software. www.ike-global.com. da Silva, V.P., van der Werf, H.M.G., Spies, A., Soares, S.R., 2010. Variability in environmental impacts of Brazilian soybean according to crop production and transport scenarios. J. Environ. Manag. 91, 1831–1839. DAC (Dairy Association of China), 2015. China Dairy Yearbook. China Agricultural Press, Beijing (in Chinese). de Boer, I.J.M., Cederberg, C., Eady, S., Gollnow, S., Kristensen, T., Macleod, M., Meul, M., Nemecek, T., Phong, L.T., Thoma, G., van der Werf, H.M.G., Williams, A.G., Zonderland-Thomassen, M.A., 2011. Greenhouse gas mitigation in animal production: towards an integrated life cycle sustainability assessment. Curr. Opin. Environ. Sustain. 3, 423–431. de Boer, I.J.M., Hoving, I.E., Vellinga, T.V., Van de Ven, G.W.J., Leffelaar, P.A., Gerber, P.J., 2013. Assessing environmental impacts associated with freshwater consumption along the life cycle of animal products: the case of Dutch milk production in Noord-Brabant. Int. J. Life Cycle Assess. 18, 193–203. de Vries, M., de Boer, I.J.M., 2010. Comparing environmental impacts for livestock products: a review of life cycle assessments. Livest. Sci. 128, 1–11. Del Prado, A., Mas, K., Pardo, G., Gallejones, P., 2013. Modelling the interactions between C and N farm balances and GHG emissions from confinement dairy farms in northern Spain. Sci. Total Environ. 465, 156–165. Di, X.H., Nie, Z.R., Zuo, T.Y., 2005. Life cycle emission inventories for the fuels consumed by thermal power in China. China Environ. Sci. 25 (5), 632–635. Ecoinvent Centre, 2013. Ecoinvent Data v3.0. Swiss Centre for Life Cycle Inventories. www.ecoinvent.org. Fantin, V., Buttol, P., Pergreffi, R., Masoni, P., 2012. Life cycle assessment of Italian high quality milk production. A comparison with an EPD study. J. Clean. Prod. 28, 150–159. FAO, 2010. Greenhouse Gas Emissions From the Dairy Sector, a Life Cycle Assessment. Food and Agriculture Organization of the United Nations, Rome, Italy, p. 55. Finnveden, G., Hauschild, M.Z., Ekvall, T., Guinée, J., Heijungs, R., Hellweg, S., Koehler, A., Pennington, D., Suh, S., 2009. Recent developments in life cycle assessment. J. Environ. Manag. 91 (1), 1–21. Flysjö, A., Henriksson, M., Cederberg, C., Ledgard, S., Englund, J.E., 2011. The impact of various parameters on the carbon footprint of milk production. Agric. Syst. 104, 459–469. Gerber, P.J., Steinfeld, H., Henderson, B., Mottet, A., Opio, C., Dijkman, J., Falcucci, A., Tempio, G., 2013. Tackling Climate Change Through Livestock – a Global Assessment of Emissions and Mitigation Opportunities. Food and Agriculture Organization of the United Nations (FAO), Rome. Guerci, M., Bava, L., Zucali, M., Sandrucc, i A., Penati, C., Tamburini, A., 2013a. Effect of farming strategies on environmental impact of intensive dairy farms in Italy. J. Dairy Res. 80, 300–308. Guerci, M., Trydeman Knudsen, M., Bava, L., Zucali, M., Schönbach, P., Kristensen, T., 2013b. Parameters affecting the environmental impact of a range of dairy farming systems in Denmark, Germany and Italy. J. Clean. Prod. 54, 133–141. Guinee, J.B., Gorree, M., Heijungs, R., Huppes, G., Kleijn, R., de Koning, A., van Oers, L., Wegener Sleeswijk, A., Suh, S., Udo de Haes, H.A., de Bruijn, H., van Duin, R., Huijbregts, M.A.J., Lindeijer, E., Roorda, A.A.H., van der Ven, B.L., Weidema, B.P., 2002. Handbook on Life Cycle Assessment; Operational Guide to the ISO Standards. Institute for Environmental Sciences. Leiden University, The Netherlands. Guo, Z.M., 2010. Effects of alfalfa hay supplementation on milk yield and milk fat of dairy. J. Anim. Sci. Vet. Med. 29 (3), 20–23 (in Chinese). Haas, G., Wetterich, F., Köpke, U., 2001. Comparing intensive, extensified and organic grassland farming in southern Germany by process life cycle assessment. Agric. Ecosyst. Environ. 83, 43–53. Hagemann, M., Ndambi, A., Hemme, T., Latacz-Lohmann, U., 2012. Contribution of milk production to global greenhouse gas emissions. Environ. Sci. Pollut. Res. 19 (2), 390–402.

He, K.B., Hao, J.M., Fu, L.X., 1996. The status and trend of vehicle pollution in China. J. Environ. Sci. 4, 80–83 (in Chinese). Huijbregts, M., 1999. Life-cycle Impact Assessment of Acidifying and Eutrophying Air Pollutants-calculation of Equivalency Factors With RAINS-LCA. Interfaculty Department of Environmental Science. University of Amsterdam, Amsterdam, the Netherlands. IAS, 2013. Chinese Feed Database. Institute of Animal Science (IAS), Beijing Online at http://www.chinafeeddata.org.cn/CFICTbl/CFIC2013_1.pdf. IDF, 2015. A Common Carbon Footprint Approach for the Dairy Sector-The IDF Guide to Standard Life Cycle Assessment Methodology. International Dairy Federation (IDF), Brussels, Belgium (P.53). IPCC, 2006. IPCC Guidelines for National Greenhouse Gas Inventories. Online at http:// www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.html. IPCC, 2013. Climate Change 2013: The Physical Science Basis. Online at http:// www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_ALL_FINAL.pdf. Knudsen, M.T., Qiao, Y.H., Luo, Y., Halberg, N., 2010. Environmental assessment of organic soybean (Glycine max.) imported from China to Denmark: a case study. J. Clean. Prod. 18 (14), 1431–1439. Kristensen, T., Mogensen, L., Knudsen, M.T., Hermansen, J.E., 2011. Effect of production system and farming strategy on greenhouse gas emissions from commercial dairy farms in a life cycle approach. Livest. Sci. 140 (1–3), 136–148. Li, Z.Q., Han, J.G., Li, S.L., Li, X.S., Huang, Y.Q., Zhang, C.X., Wang, Z.Q., 2003. Study on the optimal amount of alfalfa hay addition in the diet of high producing dairy cow. Sci. Agric. Sin. 36 (8), 950–954. Li, X.X., Hu, C.S., Zhang, Y.M., Dong, W.X., Yang, Z.Y., 2006. Losses of nitrate-nitrogen from a wheat -corn rotation in north China. Agric. Res. Arid Areas. 24 (6), 7–10 (in Chinese). Li, G.H., van Ittersum, M.K., Leffelaar, A., Sattari, Z., Li, H.G., Huang, G.Q., Zhang, F.S., 2016. A multi-level analysis of China's phosphorus flows to identify options for improved management in agriculture. Agric. Syst. 144, 87–100. Liang, L., Chen, Y.Q., Gao, W.S., Sui, P., Chen, D.D., Zhang, W., 2009. Life cycle environmental impact assessment in winter wheat-summer maize system in North China Plain. J. Agro-Environ. Sci. 28 (8), 1773–1776 (in Chinese). Liu, Y.N., Shi, Y.H., Yan, X.B., Wang, C.Z., Liang, M.G., Zhou, L., 2013. Effect of alfalfa hay substituting for part of the concentrate on the production of cows and economic profit. Avta Prataculturae Sinica. 23 (6), 190–197 (in Chinese). Ma, L., Ma, W.Q., Velthof, G.L., Wang, F.H., Qin, W., Zhang, F.S., Oenema, O., 2010. Modeling nutrient flows in the food chain China. J. Environ. Qual. 39, 1279–1289. Ma, L., Velthof, G., Wang, F., Qin, W., Zhang, W., Liu, Z., Zhang, Y., Wei, J., Lesschen, J., Ma, W., 2012. Nitrogen and phosphorus use efficiencies and losses in the food chain in China at regional scales in 1980 and 2005. Sci. Total Environ. 434, 51–61. McGeough, E.J., Little, S.M., Janzen, H.H., McAllister, T.A., McGinn, S.M., Beauchemin, K.A., 2012. Life-cycle assessment of greenhouse gas emissions from dairy production in eastern Canada: a case study. J. Dairy Sci. 95, 5164–5175. Mekonnen, M.M., Hoekstra, A.Y., 2010. The Green, Blue and Grey Water Footprint of Farm Animals and Animal Products. UNESCO-IHE, Delft, the Netherlands. Mekonnen, M.M., Hoekstra, A.Y., 2012. A global assessment to the water footprint of farm animal products. Ecosystems 15, 401–415. MEPPRC, 2003. Cleaner Production Standard-edible Vegetable Oil Industry (Soybean Oil and Soybean Meal). Ministry of Environmental Protection of the People's Republic of China (MEPPRC), Beijing, China, pp. 1–19 (In Chinese). MEPPRC, 2010. The First National Pollution Census Bulletin. Ministry of Environmental Protection of the People's Republic of China (MEPPRC) Online at http:// www.stats.gov.cn/tjsj/tjgb/qttjgb/qgqttjgb/201002/t20100211_30641.html. Meul, M., VanMiddelaar, C.E., de Boer, I.J.M., Van Passel, S., Fremaut, D., Haesaert, G., 2014. Potential of life cycle assessment to support environmental decision making at commercial dairy farms. Agric. Syst. 131, 105–115. Muehlhoff, E., Bennett, A., McMahon, D., 2013. Milk and Dairy Products in Human Nutrition. Food and Agriculture Organization of the United Nations (FAO) Online at http://www.fao.org/3/a-i3396e.pdf. Murphy, E., de Boer, I.J.M., van Middelaar, C.E., Holden, N.M., Shalloo, L., Curran, T.P., Upton, J., 2017. Water footprinting of dairy farming in Ireland. J. Clean. Prod. 140, 547–555. NAHS, 2014. Chinese Grass Industry Statistical Yearbook. National Animal Husbandry Station (NAHS), Beijing, China, p. 82 (in Chinese). NDRC, 2015. Cost and Income of National Agricultural Product. National Development and Reform Commission (NDRC), Beijing, China, pp. 119–145 (in Chinese). Nguyen, T.T.H., Doreau, M., Corson, M.S., Eugene, M., Delaby, L., Chesneau, G., Gallard, Y., van der Werf, H.M.G., 2013. Effect of dairy production system, breed and coproduct handling methods on environmental impacts at farm level. J. Environ. Manag. 120, 127–137. Nie, Y.L., Feng, Q.Y., 2013. Accelerate the implementation of the development of the dairy industry to revitalize the dairy industry to upgrade and promote the transformation and upgrading of modern dairy industry to provide a strong support. Zhonggue Rupin Gongye 6, 4–7 (in Chinese). O'Brien, D., Capper, J.L., Garnsworthy, P.C., Grainger, C., Shalloo, L., 2014. A case study of the carbon footprint of milk from high-performing confinement and grass-based dairy farms. J. Dairy Sci. 97, 1835–1851. Palhares, J.C.P., Pezzopane, J.R.M., 2015. Water footprint accounting and scarcity indicators of conventional and organic dairy production systems. J. Clean. Prod. 93, 299–307. Penati, C.A., Tamburini, A., Bava, L., Zucali, M., Sandrucci, A., 2013. Environmental impact of cow milk production in the central Italian alps using life cycle assessment. Ital. J. Anim. Sci. 12, 584–592. Pirlo, G., 2012. Cradle-to-farm gate analysis of milk carbon footprint: a descriptive review. Ital. J. Anim. Sci. 11 (e20), 09–118.

X. Wang et al. / Science of the Total Environment 625 (2018) 486–495 Ran, Y., Annerstad, M.L., Herrero, M., VanMiddelaar, C.E., De Boer, I.J.M., 2016. Assessing water resource use in livestock production: a review of methods. Livest. Sci. 187, 68–79. Ridoutt, B.G., Williums, S.R.O., Baud, S., Fraval, S., Markst, N., 2010. Short communication: the water footprint of dairy products: case study involving skim milk powder. J. Dairy Sci. 93, 5114–5117. Steinfeld, H., Gerber, P., Wassenaar, T., Wassenaar, T., Castel, V., Rosales, M., de Haan, C., 2006. Livestock's Long Shadow: Environmental Issues and Options. Food and Agriculture Organisation/Livestock Environment and Development (FAO/LED), Rome, Italy. Sultana, M.N., Uddin, M.M., Ridoutt, B.G., Peters, K.J., 2014. Comparison of water use in global milk production for different typical farms. Agric. Syst. 129, 9–21. Thomassen, M.A., Dolman, M.A., van Calker, K.J., de Boer, I.J.M., 2009. Relating life cycle assessment indicators to gross value added for Dutch dairy farms. Ecol. Econ. 68, 2278–2284. Van der Werf, H.M.G., Kanyarushoki, C., Corson, M.S., 2009. An operational method for the evaluation of resource use and environmental impacts of dairy farms by life cycle assessment. J. Environ. Manag. 90, 3643–3652. Wang, Z.H., Liu, X., Ju, X.T., Zhang, F.S., 2002. In situ determination of ammonia volatilization from wheat-maize rotation system field in north China. Acta Ecol. Sin. 22 (3), 359–365 (in Chinese). Wang, X.J., Wei, C.Z., Zhang, J., Dong, P., Wang, J., Zhu, Q.C., Wang, J.X., 2012. Effects of irrigation mode and N application rate on cotton field fertilizer N use efficiency and N losses. Chin. J. Appl. Ecol. 23 (10), 2751–2758.

495

Wang, X.Q., Kristensen, T., Mogensen, L., Knudsen, M., Wang, X.D., 2016. Greenhouse gas emissions and land use from confinement dairy farms in the Guanzhong Plain of China - using a life cycle assessment approach. J. Clean. Prod. 113, 577–586. Xu, W.L., Tang, M.Y., Ma, Z.Q., Zhang, Y.S., Liu, H., 2011. In Situ Monitoring of the losing of ammonia volatilization from soil in Xinjiang oasis cotton cropping systems. Hua Bei Nong Xue Bao 20 (9), 197–201. Yan, M.J., Humphreys, J., Holden, N.M., 2011. An evaluation of life cycle assessment of European milk production. J. Environ. Manag. 92, 372–379. Yan, M.J., Humphreys, J., Holden, N.M., 2013. Life cycle assessment of milk production from commercial dairy farms: the influence of management tactics. J. Dairy Sci. 96, 4112–4124. Yang, S.G., 2014. Economics of Soybean Industry in China From Industry Chain Perspective. China Agricultural University, Beijing, China, pp. 144–145 (in Chinese). Yue, C.W., Sun, M.H., Sun, Q.W., Wang, N., 2009. Effects of appended alfalfa hay on performance of lactating dairy cows. Hubei Agric. Sci. 48 (5), 1204–1207 (In Chinese). Zhang, Q., Ju, X.T., Zhang, F.S., 2010. Re-estimation of direct nitrous oxide emission from agricultural soils of China via revised IPCC 2006 guideline method. Chin. J. EcoAgric. 18 (1), 7–13 (in Chinese). Zhao, W., Wang, H.Y., Chen, Y.J., Wang, D.Q., Yu, J., Song, L.J., 2010. Effects of manure and chemical fertilizers application on nitrogen leaching in black soil. J. Northeast. Agric. Univ. 41 (11), 47–52 (in Chinese). Zonderland-Thomassen, M.A., Ledgard, S.F., 2012. Water footprinting - a comparison of methods using New Zealand dairy farming as a case study. Agric. Syst. 110, 30–40.