SIK-rapport Nr 741 2005
Environmental Assessment
of Future Dairy Farming Systems
- Quantifications of Two Scenarios
from the FOOD 21 Synthesis Work
SIK-rapport Nr 741 2005
Environmental Assessment
of Future Dairy Farming Systems
- Quantifications of Two Scenarios
from the FOOD 21 Synthesis Work
Ulf Sonesson
SR 741
Summary
FOOD 21 is a Swedish research program dealing with sustainable food production in a multi-disciplinary way. The research has been guided by eight main sustainability aspects, covering environmental and resource aspects as well as social-, economic and animal welfare aspects. Within FOOD 21, a synthesis group have developed scenarios for future more sustainable farming systems, based on both research and other expertise. This report presents the environmental assessment of the two scenarios for future dairy farm production developed. The scenarios and a comprehensive analysis are presented in a parallel report (Sonesson et al., 2005).
Life Cycle Assessment (LCA) is a systems analysis method for evaluating the environmental impact and resource use for a product or service. The system studied included all activities on the farm as feed crop production, manure management and animal husbandry. Production of input resources as mineral fertilisers and purchased feed were also included. Production of building and farm machinery was not included, neither were veterinary medicines. The functional unit, i.e. the unit to which all results were correlated was one kg energy corrected milk (ECM) at the farm gate. The on-farm production of meat from culled cows and surplus calves were considered to replace meat and calves from a suckler cow system, hence
emissions and resource use was avoided due to meat and live calves. The impact categories chosen were: use of energy, use of phosphorus, use of land, use of pesticides, climate change, eutrophication and acidification.
The two scenarios studied; “Specialised dairy farming” and “Mixed dairy farming” were developed from different priorities of sustainability aspects. In scenario “Mixed dairy farming”, minimised local environmental impact and resource use together with improved biodiversity and natural behaviour of the animals was prioritised. Scenario “Specialised dairy farming” was characterised by high yields per cow, only roughage feed was locally produced, less pasture, longer calving intervals, and voluntary milking systems. The feed ration contains a high proportion of concentrate feed compared to roughage feed. Both scenarios fulfil the regulations on animal welfare and should give reasonable economic results, according to present prices on milk and input resources.
The use of energy was 1.7 MJ/kg milk for “Mixed dairy farming” and 2.7 MJ/kg milk for “Specialised dairy farming”. This is mainly explained by the higher energy intensity in the purchased feed used in “Specialised dairy farming” compared to locally produced feed in “Mixed dairy farming”. The land use was 2.1 and 0.8 m2/kg milk for Mixed- and Specialised dairy farms respectively. The higher land use in Mixed dairy farming is a combination of more pasture and also that the purchased feed demanded less area per kg to produce. The emissions of gasses causing global warming potential were 0.5 kg CO2-equivalents/ kg ECM
in “Mixed dairy farming” and 0.6 for “Specialised dairy farming”. The differences are largely explained by the higher number of live calves produced in “Mixed dairy farming”, which lead to avoided emissions. The emissions of substances causing eutrophication was higher for “Mixed-” than “Specialised dairy farming”, 60 versus 50 g O2-equivalents/kg ECM. The
difference is due to high ammonia emissions from pasture. Finally, the acidifying emissions were lower for scenario “Mixed dairy farming” than “Specialised dairy farming..
No scenario was best in all environmental and resource aspects covered, but scenario “Mixed Dairy Farming” scored best for most categories. The results also showed the importance of including co-products, calves and meat, in studies of milk production systems.
Table of content
Introduction ... 1
Method ... 1
Goal and Scope Definition ... 2
Goal and Purpose of the study... 2
Scope of the study ... 2
Descriptions of the two scenarios... 3
Delimitations ... 3
Functional units ... 4
Co-product handling... 4
Chosen impact categories... 4
Data quality ... 4
Inventory analysis ... 5
Scenario “Specialised dairy farming” ... 5
General production data ... 5
Use of resources ... 8
Emissions from the systems ... 10
Plant nutrient balance ... 12
Scenario “Mixed dairy farming” ... 12
General production data ... 12
Use of resources ... 16
Emissions from the systems ... 17
Plant nutrient balance ... 20
Data used for emissions from use of oil and electricity ... 20
Data for systems expansion... 20
Avoided emissions and resource use for culled dairy cows... 21
Avoided emissions and resource use for live dairy calves... 21
Results and Impact Assessment ... 22
Use of resources ... 23 Energy ... 23 Use of phosphorus... 24 Land use ... 24 Toxicological effects ... 25 Use of pesticides... 25 Climate change... 25 Eutrophication ... 27 Acidification... 30 Discussion ... 32 References ... 33 Appendix 1. Composition of pre-mixed feed... Appendix 2. Calculations on diesel use for field operations... Appendix 3. Calculations of nitrogen emissions...
Introduction
FOOD 21 is a Swedish research program dealing with sustainable food production in a multi-disciplinary way. The research has been guided by eight main sustainability aspects (FOOD 21, 2004), covering environmental and resource aspects as well as social-, economic and animal welfare aspects.
Within FOOD 21, a synthesis group have developed scenarios for future more sustainable farming systems, based on both research and other expertise. Scenarios for pig production, arable farming, cattle meat production and dairy production have been developed (Stern et al., 2005, Wivstad et al., 2005, Kumm et al., 2005 and Sonesson et al., 2005). The method for scenario development has also been developed within the synthesis group (Sonesson et al., 2003). The scenarios will be or have been evaluated from economic, environmental and animal welfare perspective. In the present report, the environmental assessment of the dairy production scenarios is presented.
I would like to thank Stefan Gunnarsson and Maria Stenberg, SLU for cooperation on the scenario design and quantification and Anna Flysjö and Britta Florén, SIK, for their
calculations of the LCA results. Christel Cederberg contributed with valuable quantification on emissions. The synthesis group of FOOD 21 is also acknowledged for fruitful discussions.
Method
The environmental evaluation of the scenarios developed has been performed with Life Cycle Assessment (LCA) methodology. LCA is a tool for environmental systems analyses of
products and service systems. It is standardised and a thorough description can be found in ISO (ISO 14000, CEN, 1997). LCA have been used for analysing many food products and agricultural systems, and to get an overview of applications proceeding from the two latest conferences on LCA in Foods can be recommended (Mattsson, 2001, Halberg, 2004). One central point in LCA is the cradle-to-grave approach. This means that all environmental impact and resource use, from acquisition of raw materials through processing and transport to consumption, needed to deliver the product or service under study are included. LCA also aims at covering all relevant environmental impact, as emissions and resource use, from the studied system.
The environmental impact quantified is product oriented. This means that it is not limited to the geographical location of the product under study, all emissions and resource use needed to manufacture the product should be included, regardless of where the emissions occur. Hence the environmental impact presented from an LCA can be said to be global.
The method as described in the ISO Standards (CEN, 1997) includes the following steps: Goal and Scope Definition: As for all systems analysis work the goal and scope must be clearly described, i.e. what are the questions posed and for what system or product. The definition decides the systems boundaries employed (i.e. what are included in the study), the time frame, inventory method and the accuracy needed. An important part is the definition of the functional unit. The functional unit is the base that all results are related to. In early LCA the principle was to always follow the product to the grave, but recent studies often study a
long in the life cycle the product is followed. Examples of functional units are 1 kg of milk at the grocery shop or 1 kg of meat at the farm gate.
Inventory Analysis: This part of the method can be performed in different ways, depending on the goal and scope of the study. For example, a real world system can be inventoried by making queries to producing companies and making measurements. Another example is that simulation models can be used to create the data needed, which is often used for analyses of future systems.
In the Inventory Analysis allocations is used to allocate emissions and resource use between products and co-products. The principles of allocation can differ depending on the Goal and Scope but also availability of data.
Impact Assessment: The data collected in the inventory are used to calculate the potential environmental impact. Hence, the results from an LCA are not amounts of different
substances of emitted, but the potential impact the emissions might have in the environment. Examples of impact categories are Global warming, Eutrophication, Acidification,
Ecotoxicity and Human Toxicity. Use of resources is also part of the results from an LCA. Resource use consists of e.g. MJ of different fuels, land use, water use.
Interpretation: The results from the Impact Assessment are interpreted and discussed in the context of the Goal and Scope definition.
Goal and Scope Definition
Goal and Purpose of the study
The goal of the present study was to perform an environmental assessment of two scenarios for future farm milk production developed by Sonesson et al (2005) using LCA. The purpose was to analyse how the different choices would affect the environment and resource use. The results will be used as one part of the total evaluation of the scenarios made in the report by Sonesson et al (2005), where the scenarios are described in detail.
Scope of the study
The analysis covers all inputs and emissions occurring from the production of milk up to the farm gate (Figure 1). All transports of input resources are also included.
Production of: Feed components (soy, beet fibres etc)
fertilisers diesel minerals
concentrate feed production
Cultivation of cereals, beans, rape seed, ley
Cows + replacement heifers + weaned calves
Feed Dairy Farm Manure
Culled cows Weaned live calves
Milk
Figure 1. Activities included in the analysis.
Descriptions of the two scenarios
Two scenarios for dairy production was developed, “Specialised dairy farming” and “Mixed dairy farming”. The two scenarios prioritise different sustainability goals. In scenario
“Specialised dairy farming” the highest priorities were: Minimised environmental impact and resource use per unit milk produced. In scenario “Mixed dairy farming”, minimised local environmental impact and resource use together with improved biodiversity and natural behaviour of the animals was prioritised. Scenario “Specialised dairy farming” was
characterised by high yields per cow, only roughage feed was locally produced, less pasture, longer calving intervals, and voluntary milking systems. The feed ration contains a high proportion of concentrate feed compared to roughage feed.
Scenario “Mixed dairy farming” was characterised by lower (but still rather high) milk yields, all feed locally produced, normal calving intervals meaning that more calves could be sold and used for meat production. In this scenario the animals spent longer time on pasture to improve their possibilities to perform their natural behaviours. The portion of roughage feed, including pasture, is higher than in scenario “Specialised dairy farming”. Both scenarios fulfil the regulations on animal welfare and should give reasonable economic results, according to present prices on milk and input resources. For detailed descriptions of the scenarios, see Sonesson et al. (2005).
There where more differences between scenarios, but they did not affect the environmental assessment but economic and animal welfare evaluations, hence they are not presented here. Delimitations
Production and maintenance of farm buildings and machinery is not included in the study, neither are the production and use of medicines included. The production of pesticides is not included. Disinfectants and washing detergents are not accounted for.
Functional units
The functional unit is one kg of energy corrected milk (ECM) at the farm gate. Co-product handling
In both scenarios a number of calves are sold from the system to be used in meat production. This meat production is not included in the analysis. However, the avoided production of calves resulting from the dairy production is taken into account by calculating the emissions and resource use if the same number of calves was produced in a suckler cow system. The same approach is used for the culled cows; the emissions and resource use that should have occurred if an equivalent amount of meat was produced from a suckler cow system is calculated and subtracted from the dairy system.
In scenario “Mixed dairy farming” some crop products are sold. There are crop rotation demands; for producing sufficient amounts of protein feed too much grain is produced. This is managed by calculating the emissions and resource use for each crop separately and overall measures in the crop rotation are divided on area basis. In this way, the exported grain can be subtracted form the dairy system.
The growing of rape seed in scenario “Mixed dairy farming” produces both cake, which is used as feed, and oil which is used for other purposes. This rapeseed is assumed to be sold to a crusher and the cake is then delivered to the farm. The emissions and resource load from both growing and processing the seed is divided between oil and cake using economic
allocation, i.e. each product carries the same part of the environmental impact as its part of the economic value.
Chosen impact categories
The environmental impact categories chosen are:
Resource use: Energy, land use, use of phosphorous Toxicity: Use of pesticides
Ecological effects: Climate change, Eutrophication, Acidification
Data quality
The scenarios analysed are hypothetical future systems, which means that no real data are at hand. Instead a combination of data from research reports and expert judgements has been used. The production systems described are in many respects already a reality at some farms, considering milk yield and herd size. The study presented by Cederberg and Flysjö (2004), where 23 dairy farms in south western Sweden was inventoried have been used to check if many of the data are reasonable, and several input data originates from that study.
The data for fertiliser and feed production are contemporary data, but the differences between scenarios are not large with respect to these inflows, so instead of making assumptions about development within these sectors we chose to use contemporary data.
Inventory analysis
The inventory data is presented for each scenario separately, since many nitrogen emissions are caused by the same system and interconnected, so a presentation of both scenarios in parallel would be too complicated to follow.
Scenario “Specialised dairy farming”
General production data
The average number of dairy cows is 150, and 76 replacement heifers are raised yearly (which equals two years replacement, age at first calving is 24 months). Each year 62 calves are sold. The low total number of calves is due to the long calving interval (1.5 year).
The annual milk production is 11 500 kg energy corrected milk (ECM)/cow according to the official milk recording programme. Of that, 300 kg is wasted, partly due to cleaning
operations but mainly as discarded milk from cows being treated with antibiotics. Hence the delivered amount is 11 200 kg/cow and year. The dairy cows spend six hours per day between mid-May and mid-September on pasture, a permanent pasture that can be considered as an exercise yard. The heifers spend the period between May and September on a larger permanent pasture.
Data on delivered amounts per year and feed rations are presented in Table 1 and Table 2. The amounts in Table 2 are equivalent to a daily feed consumption of 10 kg DM roughage feed and 8.1 kg DM concentrate feed for the dairy cows.
Table 1. Delivered amounts from the farm, Scenario “Specialised dairy farming”
Product Annual delivery Comment
Milk 1 680 000 kg 150 cows * 11 200 kg ECM/cow and year Calves 4 650 kg live weight 62 calves * 75 kg = 4 650 kg live weight calves Slaughtered cows 21 812 live weight 38 cows * 574 kg = 21 812 kg live weight
Table 2. Feed rations for cows and replacement heifers Scenario “Specialised dairy farming”
Feed component Kg DM/cow and year Kg DM/heifer between birth and two years age
Silage 2920 2500
Pasture 0 830
Whole crop silage 730 820
Barley 0 600
Protein concentrate feed 2190 0
Pre-mixed concentrate feed 730 0
Crop rotations and acreage
In Table 3 the crop rotation for scenario “Specialised dairy farming” is presented. The acreage of roughage feed crops is calculated to fulfil the demand in the feed rations, and the barley produced will also be used on-farm. However, this barley does not fulfil the demand for grain, additional feed grain must be purchased.
Table 3. Crop rotation in scenario ”Specialised dairy farming”
Year Crop Acreage (ha) Comment
1 Whole crop silage + re-seed 22 oats/peas 50/50
2 Ley I 22 Three harvests
3 Ley II 22 Three harvests
4 Whole crop silage + re-seed Barley + re-seed
14 8
Barley
8*5 t/ha = 40 ton
5 Ley I 22 Three harvests
6 Ley II 22 Three harvests
In addition to the acreage above, 12 hectares permanent pasture for replacement heifers and 7 hectares exercise yard for dairy cows is needed. In total the acreage needed is 151 ha.
In Table 4 the on-farm production of feed and demand are compared.
Table 4. Comparison of feed demand and produced amounts of feed in scenario Specialised dairy farming.
Gross production,
tonnes Net production, tonnes Demand, tonnes Comment
Pasture 78 47 63 A slight lack of pasture, which is covered by the surplus of whole crop silage
Silage 748 636 628
Whole crop silage 264 224 172 Surplus, used to cover the lack in pasture
Barley 40 40 40
Farmyard manure
The amount of farmyard manure produced is presented in Table 5. The amount is calculated from SJV (2003), and the amount of nitrogen is calculated from balance calculations of nitrogen over the cow, i.e. the total amount of nitrogen in the feed is subtracted with total amount of nitrogen in milk and calves. This balance shows a nitrogen surplus of 132 kg/year. Regarding the nitrogen excretion from heifers, statistical data from STANK is used (SJV, 2003).
Table 5. Produced amount of farmyard manure and content of nitrogen (SJV, 2003), Scenario “Specialised dairy farming”
Type of animal Liquid manure for 12 month
storage, including rainwater (m3) kg N/cow gross produced in manure
Dairy cow, producing 10000 kg/year 22 139 a
Heifer < 1 år 5 22
Heifer > 1 år 9.2 47
a This figure calculated as mass balance over the animal
In Table 6 the total amounts of farmyard manure and the proportion produced on pasture is presented. Liquid farmyard manure contains 9.8% dry matter, 0.39% total nitrogen, 0.18% NH4-N, 0.075% phosphorous and 0.4% potassium (SNV, 1999).
Table 6. Calculations of total farmyard manure production, Scenario “Specialised dairy farming”
Type of animal Calculation Total amount (ton) Produced indoors
Dairy cow 150*22 3300 3000 (92 %)
Heifer < 1 år 38*5 190 133 (70 %)
Heifer > 1 år 38*9,2 350 210 (60 %)
Total amount (ton) 3 800 3 300
In Table 7 the use of different fertilisers in the crop rotation is presented. Spreading of liquid manure in ley cause large emissions of mainly NH3 compared to spreading in arable crops,
hence we choose to use mainly mineral fertilisers in ley and pasture. However, manure is spread in second year ley. The use of farmyard manure in the crop rotation is 2 400 ton, and the production is 3 300 ton. This means that 900 ton is sold.
Table 7. Use of fertilisers in the crop rotation, Scenario “Specialised dairy farming”
Year Crop Acreage
(ha) Farmyard manure applied (ton/ha) Total use of farmyard manure (ton) Mineral fertilisers (kg/ha) Total use of mineral fertilisers 1 Oats/peas + re-seed 22 25 550 - 0 2 Ley I 22 0 0 190 4180 3 Ley II 22 30 660 130 2860
4 Whole crop silage /
Barley+ re-seed 22 25 550 30 660
5 Ley I 22 0 0 190 4180
6 Ley II 22 30 660 130 2860
Pasture, heifers 12 0 0 75 900
Exercise yard for dairy cows
Use of resources Electricity
In a study made on 23 dairy farms in Sweden (Cederberg & Flysjö, 2004), the use of
electricity was 1290 kWh/ cow (including replacement heifer). We use a slightly lower figure since small improvements is reasonable, the use of electricity in animal husbandry is 1250 kWh/cow, including replacement heifer. On the farm there are 150 cows, so the total use of electricity in animal husbandry is: 150 * 1250 = 187 500 kWh. Electricity for drying of grain (barley) uses 560 kWh electricity. In total, the use of electricity is 188 000 kWh/year.
Use of Diesel fuel
Diesel fuel is mainly used in the feed crop production, and the calculations for diesel use are presented in Appendix 2. The total diesel use for feed crop production is 13 038 litres, summed from the tables in Appendix 2.
In addition to the use of diesel fuel in the feed crop production, animal farms use their tractors for a number of additional works, as moving and managing animals, fencing and on-farm transports of forage. This fuel consumption varies between farms and is very difficult to calculate. According to a Danish study (Halberg et al., 2000) in addition to crop production, 40% extra diesel use was recorded on animal farms. We chooses to add 40% to the calculated total amount from crop production, which result in a total diesel use of for the farm of 18 250 litres/year.
Oil for grain drying
The only grain that needs drying is the 40 tons of barley, which are dried from on average 18% water content to 14%. According to Törner (2003, pers. comm.), 0.17 litres oil is needed for each kg of water removed. This results in 328 litres of oil for grain drying.
Land use
The land use on farm is presented in Table 8.
Table 8. Land use for different purposes, Scenario “Specialised dairy farming”
Crop Acreage (ha)
Ley for forage 88
Whole crop silage 36
Barley 8
Pasture for heifers 12
Exercise yard for dairy cows 7
Total acreage use on-farm 151
Mineral fertilisers
The use of mineral fertilisers is presented in Table 7; the total use is 15640 kg nitrogen. Purchased feed
The amount of different feeds and their use is presented in Table 9. For the environmental impact and resource use for the purchased feed we use data from Cederberg & Flysjö (2004).
Table 9. Amounts of purchased feed, Scenario “Specialised dairy farming”
Feed type Product purchased
(kg) Protein concentrate feed 365 000 Pre-mixed concentrate feed 122 000
Mineral feed 8 249
Calve feed, calves for replacement 1 900
Calve feed, sold calves 2 728
Concentrate feed calves for replacement 1 037 Concentrate feed, sold calves 911
Pesticides used for crop protection
Pesticides are not used very much, it is mainly in barley and the whole crop silage herbicides are used. Roundup, which is used for perennial weeds are applied on average one in three years in the crop rotation. In Table 10 the on-farm pesticide use is presented (it should be noted that pesticide use for purchased feed is not included).
Table 10. Pesticides used, Scenario “Specialised dairy farming”
Pesticide, used
on Dose applied (l/ha) Activ substance Acreage treated (ha) substance used (g) Amount active
MCPA (Barley) 0.6 MCPA 750 g/l 8 5400
Gratil (Barley) 0.015 Amidosulfuron 75% (weight) 8 90 Basagran,
(Whole crop silage)
1.3 Bentazon, 87 %, (weight) 36 40 700
Roundup
Emissions from the systems Nitrogen emissions
The emissions from the system are calculated using different models. The resulting emissions are then used for the LCA analysis of the farm. The calculations for emissions of nitrogen compounds are presented in Appendix 3, below the resulting emissions are presented (Table 11, Table 12 and Table 13).
Table 11. Emissions of nitrate (NO3--N) from the farm, Scenario ”Specialised dairy farming”
Crop Acreage (ha) Total leaching (kg N/ha) Total leaching (kg N/crop)
Oat/peas + re-seed 22 11 242
Ley I 22 9 187
Ley II 22 25 554
Whole crop silage
/barley + re-seed 22 11 242
Ley I 22 9 187
Ley II 22 25 554
Pasture, exercise yard 7 14 99
Pasture, heifers 12 9 105
Sum 151 14 2171
Table 12. Ammonia emissions (NH3) from the farm, Scenario “Specialised dairy farming”
Source of emission NH3-N (kg)
Housing 1391
Pasture, total 204
Manure storage 185
Manure spreading 711
Mineral fertiliser spreading 156
Total 2647
Table 13. Emissions of nitrous oxide (N2O) from different sources, Scenario “Specialised
dairy farming”
Source of emission N2O-N (kg)
Manure storage 19
Pasture 50 Indirect emissions due to NH3-losses from
storage and pasture
16
Direct emissions from fertilising 448 Indirect emissions due to NH3-losses from
spreading 7
Indirect emissions due to NO3-leachate 54
Methane
Methane (CH4) is emitted from the enteric fermentation of the animals and also from manure
storage. Emissions from enteric fermentation are calculated using factors per animal and year. The default values that are given in IPCC (1997) are based on milk yields significantly lower than the yields in this study. To account for this, a relation given by Kirchgessner et al. (1991) was used to calculate the emissions. For dairy cows producing 11000 kg milk the emissions are assumed to be 138 kg/year and for heifers 50 kg/year, which is the Swedish EPA factor for ”cattle other than cows” (SNV, 2002). This gives: 150*138=20700 kg CH4 for cows, and
(38+38)*50=3800 kg CH4 for heifers.
Emissions from manure storage are calculated as a function of the amount of volatile solids in the manure produced and different factors, according to IPCC (1997):
Emissions of CH4 = VS * Bo * k * MFC
Where:
VS = Volatile solids
Bo = Methane generation potential (IPCC, 1997). MFC = Methane conversion factor (IPCC, 1997). k = 0.67 kg/m3 (IPCC, 1997).
In Table 14 the total emissions of methane is presented.
Table 14. Methane emissions from manure storage and enteric fermentation (sources: IPCC, 1997 and Dustan, 2002), Scenario “Specialised dairy farming”
Type of animals Type of
manure Amount of manure (kg) DM content (%) VS (% of DM) Bo k MFC Amount CH4 emitted (kg/animal& year) Dairy cows Liquid 3000 000 9.8 87 0.24 0.67 0.1 4113 Dairy cows Pasture 300 000 9.8 87 0.24 0.67 0.01 41
Heifers Liquid 343000 9.8 87 0.17 0.67 0.1 333 Heifers Pasture 197000 9.8 87 0.17 0.67 0.01 19 Enteric fermentation, cows 20700 Enteric fermentation, heifers 3800 Total 29006
Plant nutrient balance
The plant nutrient balance for the farm is presented in Table 15.
Table 15. Plant nutrient balance for the farm, Scenario “Specialised dairy farming”
Nitrogen (kg/ha) Phosphorus (kg/ha) Potassium (kg/ha)
Purchased feed/seed 100 17 35
Mineral fertilisers 104
N-fixation/precipitation 54
Total inflow 258 17 35
Animal products sold 63 12 18
Farmyard manure, sold 23 4 24
Total outflow 87 16 42
Surplus/deficit, farmgate +171 1 -7
Efficiency (outflow/inflow) 0.34 0.94 1.2
Calculated NO3-N losses 14.5 Calculated NH3-N losses 16.5 Calculated N2O-N losses (direct) 3.4 Total calculated N-losses 34.4
Scenario “Mixed dairy farming”
General production data
The average number of dairy cows is 150, and 76 replacement heifers are raised (which equals two years replacement, age at first calving is 24 months). Each year 112 calves are sold and each cow produces one calf per year.
The annual milk production is 9 000 kg ECM/cow according to the official milk recording programme. Of that, 950 kg is used for calf feed or is wasted, partly due cleaning operations but mainly as discarded milk from cows being treated with antibiotics. Hence the delivered amount is 8 050 kg/cow and year. The dairy cows spend 18 hours per day between May and September on pasture, and the heifers spend all day between May and September on pasture. Data on delivered amounts per year and feed rations are presented in Table 16 and Table 17. The amounts in Table 17 are equivalent to a daily feed consumption of 13.6 kg DM roughage feed and 7.1 kg DM concentrate feed for the dairy cows.
Table 16. Delivered amounts from the farm, Scenario “Mixed dairy farming”
Product Annual delivery Comment
Milk 1 208 000 kg 150 cows * 8 050 kg ECM/cow and year Calves 8 400 kg live weight 112 calves * 75 kg = 8 400 kg live weight calves Slaughtered cows 21 812 live weight 38 cows * 574 kg = 21 812 kg live weight
Table 17. Feed rations for cows and replacement heifers, Scenario “Mixed dairy farming”
Feed component Kg DM/cow and year Kg DM/heifer from birth to two year
Silage 3025 2500
Pasture 1320 830
Whole crop silage 610 820
Grain 1300 600
Faba bean 700 200
Rape seed cake 600 -
Mineral feed - 36,5
The composition of the pre-mixed feeds is presented in Appendix 1. Crop rotations and acreage
In scenario “Mixed dairy farming” two crop rotations are used; one mainly producing pasture feed and one procuring the remaining feed. The acreage feed crops is calculated to fulfil the demand in the feed rations, some crop products are produced in surplus and are sold from the system. In Table 18 the pasture crop rotation is presented and in Table 19 the feed crop rotation is presented.
Table 18. Pasture crop rotation in scenario ”Mixed dairy farming”
Year Crop Acreage (ha) Comment
1 Whole crop silage (oats/peas)+re-seed 12.5 Silage
Barley + re-seed 12.5 Grain, 12.5 ha*3.5 ton/ha
2 Pasture ley I 25 Pasture
3 Pasture ley II 25 Pasture
Table 19. Feed crop rotation, Scenario ”Mixed dairy farming”
Year Crop Acreage (ha) Comment
1 Whole crop silage (Barley) + re-seed Barley + re-seed 20 34 Silage Grain 3,5 ton/ha 2 Ley I 54 Silage 3 Ley II 54 Silage
4 Winter rape 54 Silage
5 Triticale 54 4 ton/ha
6 Faba bean 54 2.5 ton/ha
7 Spring wheat 54 3,5 ton/ha
In Table 20 the on-farm production of feed and demand are compared.
Table 20. Comparison of feed demand and produced amounts of feed, Scenario “Mixed dairy farming”
Gross production,
tonnes Net production, tonnes Demand, tonnes Comment
Pasture 487 292 261
Silage 648 551 644 The deficit is evened out by the surplus in pasture Whole crop silage 162 138 154 Grain 568 Barley: 163 Triticale: 216 Spring wheat: 189
241 All triticale + 25 tons barley is used as feed. The remainder, 138 ton barley and 189 ton spring wheat) is sold.
Faba beans 135 120 The surplus is used as reserve
Rape seed 135 Cake: 90
Oil: 45 90 (cake) The oil is sold
Farmyard manure
The amount of farmyard manure produced is presented in Table 21. The amount of manure is calculated from SJV (2003), and the amount of nitrogen is calculated from balance
calculations of nitrogen over the cow, i.e. the total amount of nitrogen in the feed is subtracted with total amount of nitrogen in milk and calves produced. This balance shows a nitrogen surplus of 165 kg/year. Regarding the heifers, statistical data from STANK is used (SJV, 2003).
Table 21. Produced amount of farmyard manure and content of nitrogen (SJV, 2003), Scenario “Mixed dairy farming”
Type of animal Liquid manure for 12 month storage,
including rainwater (m3) kg N/cow gross produced in manure Dairy cow, producing 8000
kg/year 21.6 165
a
Heifer < 1 year 5 22
Heifer > 1 year 9.2 47
a This figure calculated as mass balance over the animal
In Table 22 the total amounts of farmyard manure and the proportion produced on pasture is presented. Liquid farmyard manure contains 9.8% dry matter, 0.39% total nitrogen, 0.18% NH4-N, 0.075% phosphorous and 0.4% potassium (SNV, 1999, 15 samples analysed).
Table 22. Calculations of farmyard manure production, Scenario “Mixed dairy farming”
Type of animal Calculation Total amount (ton) Produced indoors
Dairy cows 150*22 3 300 2310 (70 %)
Heifer < 1 year 38*5 190 133 (70 %)
Heifer > 1 year 38*9,2 350 210 (60 %)
Total amount (ton) 3 840 2653 (spreading)
In Table 23 the use of different fertilisers in the crop rotation is presented. Spreading of liquid manure in ley cause large emissions of mainly NH3 compared to spreading in arable crops,
hence we choose to use mineral fertilisers in ley and pasture. The use of farmyard manure in the crop rotation is 2 592 ton, and the production is 2653 ton. This minor difference is not accounted for.
Table 23. Use of fertilisers in the feed crop rotation, Scenario “Mixed dairy farming”
Year Crop Acreage
(ha) Farmyard manure applied (ton/ha) Total use of farmyard manure (ton) Mineral fertilisers (kg N/ha) Total use of mineral fertilisers (kg N)
1 Whole crop silage + re-seed Barley + re-seed 20 34 15 15 810 - 0 2 Ley I 54 - 0 3 Ley II 54 18 972 - 0 4 Winter rape 54 15 810 81 4374 5 Triticale 54 - 0 6 Faba bean 54 - 0 7 Spring wheat 54 40 2160 Sum 2592 6534
Use of resources Electricity
We use the figure from the scenario “Specialised dairy farming” decreased by 10% due to longer pasture period, hence the use of electricity in animal husbandry is 1125 kWh/cow, including replacement heifer. On the farm there are 150 cows, so the total uses of electricity in animal husbandry is: 150 * 1125 = 168 750 kWh. Electricity needed for drying of grain accounts to 10 260 kWh electricity. This is calculated assuming that on average the grain is dried from 18% water to 14%, the rape seed from 10.8% to 9% and the faba bean from 20% to 14%. It is only drying of the crops used on-farm that is dried; the crops sold are not included in the environmental analysis. The electricity used for pressing of rape seed is 65 kWh/kg seed (Sönnerstedt, 2002, pers. comm.), and half of this is allocated to the cake. This equals 4 388 kWh for the cake.
In total, the use of electricity is 183 428 kWh/year. Use of diesel
Diesel fuel is mainly used in the feed crop production, and the calculations for diesel use are presented in Appendix 2. The total diesel use for feed crop production is 28 981 litres. From that figure the diesel used in the crops exported from the system (spring wheat, rape seed oil and part of the barley, see Appendix 2) must be subtracted. The amounts of diesel used in the exported crops are presented in Table 24. The diesel use for winter rape seed is allocated 50/50 to oil and cake, hence the figures for winter rape seed in Table 24 is half of the total use in Appendix 2.
Table 24. Diesel used in crops exported from the system, Scenario “Mixed dairy farming”
Spring wheat Winter rape seed Barley Sum
Sowing 432 216 272 920 Ploughing 1026 1026 2052 Disc harrowing 1296 1296 2592 Mineral fertiliser application 59 30 89 Harvest 1566 783 986 3335 Sum 4379 3351 1258 8988
If the diesel used in exported crops (8988 litres) are subtracted from the total diesel use (28 981) the net use for feed production is 19993 litres.
In addition to the use of diesel fuel in the feed crop production, animal farms use their tractors for a number of additional works, as moving and managing animals, fencing and on-farm transports of forage. This fuel consumption varies between farms and is very difficult to calculate. According to a Danish study (Halberg et al., 2000) in addition to crop production, 40% extra diesel use was recorded on animal farms. We chooses to add 40% to the calculated total amount from crop production, which result in a total diesel use of for the farm of 27 990 litres/year.
Oil for grain drying
According to Törner (2003, pers. comm.), 0.17 litres oil is needed for each kg of water removed. This results in 328 litres of oil for grain drying. Assuming that on average the grain is dried from 18% water to 14%, the rape seed from 10.8% to 9% and the faba bean from 20% to 14%. It is only drying of the crops used on-farm that is dried. The crops sold are not
included in the environmental analysis. Altogether this result in a total use of oil of 4088 litres. As for field work, 50% of the oil used for drying rape seed is allocated to the feed production.
Land use
The land use on farm is presented in Table 25.
Table 25. Land use for different purposes, Scenario “Mixed dairy farming”
Crop Acreage (ha)
Pasture 75 Ley 108 Roughage, Whole crop silage 33 13 ha oats/peas, 20 ha barley Grain: Triticale
Barley
54 12
Faba beans 54
Winter rape seed 27 50 % allocated to cake
Sum feed crops 363
Mineral fertilisers
The only mineral fertiliser used for the feed crops are nitrogen on winter rape seed, 81 kg N/ha. Of that 50% is allocated to feed (cake), 54 ha are grown. The total use of mineral fertiliser nitrogen is 2187 kg N.
Purchased feed
The only feeds purchased are mineral feed, 2774 kg and calf feed, 2394 kg for replacement heifers. For the environmental impact and resource use for the purchased feed we use data from Cederberg & Flysjö (2004)
Pesticides used for crop protection
In this system, no pesticides are used. The crop rotation, which is generally very varied and dominated by ley and silage crops, means that the need for chemical crop protection is very small. We assume no pesticides are used.
Emissions from the systems Nitrogen emissions
The emissions from the system are calculated using different models. The resulting emissions are then used for the LCA analysis of the farm. The calculations for emissions of nitrogen compounds are presented in Appendix 3, below the resulting emissions are presented (Table 26, Table 27 and Table 28).
Table 26. Nitrate (NO3-) leaching in the crop rotation, Scenario “Mixed dairy farming”
Crop Acreage (ha) Total leachate
(kg N/ha) Total leachate (kg N/crop)
Whole crop silage + re-seed 12.5 8 200
Barley + re-seed 12.5 6 160
Pasture I-III a, b 75 6 160
Barley + re-seed 34 24 600
Whole crop silage + re-seed 20 9 508
Ley I c 54 6 346
Ley II d 54 30 1642
Winter rape seed 54 21 1123
Triticale 54 16 864
Faba beans 54 21 1123
Spring wheat 54 16 864
Gross sum 478 12 7590
NO3- -N allocated to exported crops 1732
Net sum for feed crops 5858
Table 27. Ammonia (NH3) emissions from the farm, Scenario “Mixed dairy farming”
Source of emission NH3-N (kg)
Housing 1329
Pasture, total 671
Manure storage 177
Manure spreading 496
Mineral fertiliser spreading 22
Total 2695
Table 28. Emissions of nitrous oxide (N2O) from different sources, Scenario “Mixed dairy
farming”
Source of emission N2O-N (kg)
Manure storage 18
Pasture 167 Indirect emissions due to NH3-losses from storage and pasture 16
Direct emissions from fertilising 488 Indirect emissions due to NH3-losses from spreading 5
Indirect emissions due to NO3-leachate 146
Sum 840
Methane
Methane (CH4) is emitted from the enteric fermentation of the animals and also from manure
storage. Emissions from enteric fermentation are calculated using factors per animal and year. The default values that are given in IPCC (1997) are based on milk yields significantly lower than the yields in this study. To account for this, a relation given by Kirchgessner et al. (1991) was used to calculate the emissions. For dairy cows producing 9000 kg milk the emissions are
125 kg/year and for heifers 50 kg/year, which is the Swedish EPA factor for ”cattle other than cows” (SNV, 2002). This gives: 150*125=18750 kg CH4 for cows, and (38+38)*50=3800 kg
CH4 for heifers.
Emissions from manure storage are calculated as a function of the amount of volatile solids in the manure produced and different factors, according to IPCC (1997):
Emissions of CH4 = VS * Bo * k * MFC
Where:
VS = Volatile solids
Bo = Methane generation potential (IPCC, 1997). MFC = Methane conversion factor (IPCC, 1997). k = 0.67 kg/m3 (IPCC, 1997)
In Table 29 the total emissions of methane is presented.
Table 29. Methane emissions from manure storage and enteric fermentation (sources: IPCC, 1997 and Dustan, 2002) , Scenario “Mixed dairy farming”
Type of animals Type of
manure Amount of manure (kg) DM content (%) VS (% of DM) Bo k MFC Amount CH4 emitted (kg/animal& year)
Dairy cows Liquid 2310000 9.8 87 0.24 0.67 0.1 3167 Dairy cows Pasture 99000 9.8 87 0.24 0.67 0.01 136
Heifers Liquid 343000 9.8 87 0.17 0.67 0.1 333 Heifers Pasture 197000 9.8 87 0.17 0.67 0.01 19 Enteric fermentation, cows 18750 Enteric fermentation, heifers 3800 Total 26205
Plant nutrient balance
The plant nutrient balance for the farm is presented in Table 30. Since the outflow is larger than the inflow for both phosphorus and potassium the fertilising regime can not be said to be sustainable in the long run, but the clay soil on the farm has large content of both nutrients, so we assume that the system will function well for many years, but in the long run, plant
nutrient deficiencies are not sustainable.
Table 30. Plant nutrient balance for the farm, Scenario “Mixed dairy farming”
Nitrogen (kg/ha) Phosphorus (kg/ha) Potassium (kg/ha)
Purchased feed/seed 4 1 1
Mineral fertilisers 14
N-fixation/precipitation 53
Total inflow 70 1 1
Animal products sold 15 3 4
Vegetable products, sold 12 2 3
Total outflow 27 5 7
Surplus/deficit, farmgate +43 -4 -6
Efficiency (outflow/inflow) 0.39 5 7
Calculated NO3-N losses 16 Calculated NH3-N losses 5 Calculated N2O-N losses (direct) 1,2 Total calculated N-losses 22,2
Data used for emissions from use of oil and electricity
Data for calculating the emissions and resource use for diesel used for tractors are based on Lindgren et al. (2002), but the NOX – emissions are reduced due to coming directives on
agricultural engines in the EU (Hansson et al., 2003). Emissions and resource use from use of oil for grain drying as well as electricity production are calculated using a data base employed in the LCA software tool LCA iT (CIT Ekologik, 2004).
Data for systems expansion
The amount of meat delivered as culled cows are assumed to replace meat produced by a heifer or bull slaughtered at 18 month age from a suckler cow production. The live calves replace live calves from suckle cow production. In a study by Cederberg and Darelius (1999), a cow produces five calves during her life and one of them is used for replacement, hence four calves for slaughter is produced. Data on the suckler cow production are partly from the report by Cederberg & Darelius (1999), where an organic production system is chosen. Some data originates from the description of Scenario “Specialised Dairy Production”. The
calculations of environmental impact for feed production are based on scenario “Specialised dairy farming”. In Table 31 the feed consumption for suckler cows and bulls- and heifers is
presented. In Table 33 data used for calculations of emissions and resource use for meat production is presented.
Table 32. Feed consumption for suckler cows and 2-year bull or heifer
Suckler cow (kg DM/ animal
and year) 2-year bull or heifer (kg DM from birth to two years age)
Pasture consumed 1820 b 1750 a
Silage consumed 1370 b 1500 a
a From scenario “Specialised dairy production”, b Cederberg & Darelius, 1999
Table 33. Input data for feed production for the suckler cow production
Pasture Silage
Yield (kg DM/ha&year) 3900 a 8500 a
Fertilising (kg N/ha&year) 75 a 160 a Emissions, NO3- -N (kg/ha&year) 14 a 14 a Emissions, NH3-N (kg/ha&year) 6 b 10.6 b Emissions, N2O-N (kg/ha&year) 0.6 b 2.1 b Emissions, P (kg/ha&year) 0.25 b 0.3 b
Diesel (l/ha& year) 10 c 51 d
a From scenario “Specialised dairy production”, b Cederberg & Darelius, 1999
c Assumed value
d Assumption based on data from scenario “Specialised dairy production”
Avoided emissions and resource use for culled dairy cows
In Table 34 inventory LCA-data for the suckler cow production is presented. The data in column “Two year bull or heifer” is divided by the amount of bone-free meat at slaughter for the animals to get the data per kg meat. The weight of bone free meat is 243 kg (Average for bulls and heifers, Cederberg & Darelius, 1999). These LCA data per kg meat is used to calculate the avoided emissions and resource use for culled cows in the dairy production systems.
Avoided emissions and resource use for live dairy calves
The data on suckler cows in Table 34 is calculated on basis of feed consumption and other emissions. The emissions and resource use for the meat produced from the culled cow is calculated from the data on “Two year bull or heifer”. Principally, the emissions and resource use for producing one life calf consists of three parts: rearing of a heifer from birth to first calving plus one year feeding (one calf per year is produced) and finally the emissions and resource use for the meat produced from the culled cow is subtracted. Below the calculations is formalised.
A = B+C-D
A = Emissions and resource use for one calf
B = Emissions and resource use for one heifer at two years age (slaughter) divided by slaughter weight
D = Emissions and resource use avoided when slaughtering the suckler cow after its productive life.
If the weight of the culled cow is similar to the weight of a two-year heifer, the first and last (B and D) parts in the equations evens out, leaving only the emissions and resource use for one cow-year as result for one live calf.
A=C
However, the suckler cow needs to be replaced after its productive life to maintain the system. Assuming that each cow produces five calves during its life, one is needed for replacement and the emissions and resources used should be allocated to the remaining calves, hence the final calculation will be:
A=C*5/4
This is how the avoided emissions and resource use for one live calf have been calculated; the data used is presented in Table 34 (column “Suckler cow year”).
Of course the above calculations builds on the assumption that the culled suckler cow and the slaughtered two-year heifer has the same slaughter weight, but that assumption is reasonable considering the relatively small differences in slaughter weight reported by Cederberg & Nilsson (2004).
Table 34. Data for suckler cow and bull/heifer in meat production. Note that the unit is “one cow & year” for the suckler cow, and “one animal at slaughter (two years) age” for bull and heifer
Suckler
cow-year
Two year bull or heifer, values within parentheses represent LCA data per
kg fat and bone free meat NO3- -N (kg/animal) 8.8 8.8 (3.6*10-2)
NH3-N (kg/animal) 18.7 16.3 (6.7*10-2)
N2O-N (kg/animal) 0.9 1.0 (4.2*10-3)
CH4 (kg/ animal) 89 65 (0.27)
P leachate (kg/animal) 0.17 0.17 (6.8*10-4) Nitrogen, mineral fertiliser (kg/animal) 61 62 (0.25)
Diesel use (l/animal 13 13.5 (5.6*10-2)
Electricity (kWh/animal) 35 35 (0,14)
Results and Impact Assessment
In this section the result from the Life Cycle Assessment is presented. For each impact category the weighing factors used are presented.
Use of resources
Energy
The energy use for the two scenarios is presented below. We have chosen to present the use of secondary energy. There are large differences both in total energy use and what the energy is used for. In Figure 2 the energy use per activity is presented, and for “Mixed dairy farming” the largest contributor is the farm, which includes diesel, oil for drying and electricity. There are also contributions from mineral fertilisers and purchased feed but since those inflows are small, the energy use is also small. Scenario “Specialised dairy production” is dominated by purchased feed, which is a result of many things; energy use for soy cultivation and
transports, use of mineral fertilisers and fuel for grain, processing and transport of other feed component as drying of beet pulp and palm kernel expels, and also the use of mineral
fertilisers and fuels in production of other feed components.
-1 -0.5 0 0.5 1 1.5 2 2.5 3
Mixed dairy farming Specialised dairy farming
MJ/kg e nergy corre cted m ilk Purchased feed Mineral fertiliser Farm Live calves Meat, culled cows
Figure 2. Secondary energy use for the two scenarios, per activity
The type of energy used is presented in Figure 3. The large difference in use of fossil fuel is a result of both more use of mineral fertilisers and purchased feed in “Specialised dairy
farming”, which in turn is a result of the high energy input for producing concentrate feed; the amounts are larger and the energy input per unit feed is also high. The concentrate feed
0 0.5 1 1.5 2 2.5
Mixed dairy farming Specialised dairy farming
M J/ kg milk District heating Renewable fuel Fossil fuel Electricity
Figure 3. Secondary energy use for the two scenarios, per type of energy used
Use of phosphorus
The use of phosphorus in both scenarios is only due to imported feed. Scenario “Mixed dairy farming” use 363 kg phosphorus per year (calculated from Table 25 and Table 30) and scenario “Specialised dairy farming” use 2567 kg/year (calculated from Table 8 and Table 15). The resulting phosphorus use per kg milk is presented in Table 35.
Table 35. Use of Phosphorus for the two scenarios
Scenario g P/kg milk
Specialised dairy farming 0.3
Mixed dairy farming 1.5
Land use
The land needed for production of one kg of energy corrected milk for the two scenarios is presented in Figure 4. The land use is higher for the “Mixed dairy farming”, the reason is the local feed production which demand more land than the imported feed in “Specialised dairy farming”. The low land use for purchased feed is a result of the high proportion of
by-products as beet pulp, which are considered to use very little or no land. The fact that “Mixed dairy farming” also uses rather much more pasture increases the land use. The avoided land use resulting from live calves and meat from culled cows decreases the total land use, but the difference between scenarios are still large. The land use presented in Figure 4 does not reflect the type of land use, only the area needed, hence the land use can be both negative and positive from an environmental point of view.
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Mixed dairy farming Specialised dairy farming
m
2 /yea
r
Purchased feed Protein feed, farm Avoided land use, calves Avoided land use, meat Grain, farm
Roughage feed, farm
Figure 4. Land use for the two scenarios, “roughage feed” includes on-farm pasture
Toxicological effects
Use of pesticides
The use of pesticides is presented as kg of active substance; methods for weighing different substances are today not developed in a way that makes it useful for studies of this kind. The use of pesticides for producing purchased feed is not included, which is important to note since the production of e.g. soy uses significant amount of pesticides. In Table 36 the on-farm use of pesticides is shown. It should also be noted that no pesticides are used in scenario “Mixed dairy farming”.
Table 36. Use of pesticides in scenario “Specialised dairy farming” (in scenario “Mixed dairy farming” no pesticides are used)
Type of pesticides Amount used (mg active substance /kg milk)
Herbicides 56 Fungicides 0.05 Insecticides 0.04
Climate change
The contribution to potential global warming is calculated using the weighing factors (100 years perspective) presented in Table 37.
Table 37. Weighting factors used for global warming potential (100 years perspective), IPCC, 1997
Substance Weighing factor (kg CO2-equiv./kg substance)
CO2 1
CH4 21
N2O 320
The results for the two scenarios are presented in Figure 5. The largest contributor in both scenarios is the farm emissions, but in “Mixed dairy farming” the farm accounts for a larger share. This is largely explained by the lower milk production in scenario “Mixed dairy farming”; the emissions are similar per animal in both scenarios, but the production is higher in “Specialised dairy farming”.
The difference between scenarios, making “Mixed dairy farming” in total slightly better, is the higher number of calves delivered from the system and also the slightly higher meat production. Each calf replaces one calf in a suckler cow system. In such a system, the environmental costs for producing one calf is equivalent to emission occurring for one cow year; one cow produces one calf a year.
-600 -400 -200 0 200 400 600 800 1000
Mixed dairy farming Specialised dairy farming
g CO 2 -eq uiv./kg m ilk Purchased feed Mineral fertilisers Farm Live calves Meat, culled cows
Figure 5. Global warming potential for the two scenarios, for different activities
The distribution between substances causing potential global warming is presented in Figure 6. It shows that the methane emissions are most important, followed by nitrous oxide. The latter points at the importance of nitrogen management on the farm: higher levels of fertilising means higher emissions of nitrous oxides. The higher emissions of CO2 in “Specialised dairy
farming” is due to higher use of mineral fertilisers and also longer transports of feed, e.g. soy meal from Brazil and finally the drying of beet pulp and transport of other ingredients in the concentrate feed causes significant emissions of CO2.
0 100 200 300 400 500 600 700
Mixed dairy farming Specialised dairy farming
g C O2 -e qu iv./kg m ilk N2O CO2 CH4
Figure 6. Global warming potential for the two scenarios, distribution between substances
Eutrophication
The weighing factors used for calculating the potential eutrophication are shown in Table 38. We have chosen to present the maximal potential eutrophication. One important point to mention is that we have not included any leaching of phosphorous from the farm, even if it is an important contributor to eutrophication. The reason is that the connections between supplied phosphorous and leaching are not yet fully described. The method often used of quantifying phosphorous emissions is to use a base leaching per hectare arable land without considering the supply. Since our two scenarios use very different acreage (Figure 4) the phosphorous emissions would heavily affect the results, without considering the phosphorous supply. The calculations of plant nutrients (Table 15 and Table 30) serve as an indication on the long term phosphorous leaching potential.
Table 38. Weighing factors for potential eutrophication, max scenario
Substance kg O2 equivalent per kg
COD 1 NH3 16 NOX 6 N 20 NH4+ 15 NO3- 4.4 P 140
The results for the potential eutrophication per activity are presented in Figure 7. “Mixed dairy farming” shows higher total emissions than “Specialised dairy farming”. This is mainly
both related to area and amount of nitrogen applied, and scenario “Mixed dairy farming” uses more land, mainly due to more pasture. The avoided emissions due to live calves and meat are higher in “Mixed dairy farming” and almost even out the differences; conclusively the
differences are too small to be significant.
-100 -50 0 50 100 150 200
Mixed dairy farming Specialised dairy farming
g O 2 -equiv./kg milk Purchased feed Mineral fertilisers Farm Live calves Meat, culled cows
Figure 7. Potential eutrophication (max scenario) for the two scenarios, for different activities
The results on what substances contributing to eutrophication are presented in Figure 8. As mentioned above, nitrate from land is the largest contributor. The relatively high emissions of phosphorus from “Specialised dairy farming” originate from the growing of soy in Brazil, which causes soil erosions with subsequent phosphorus losses. The negative emissions of phosphorus and ammonia from “Mixed dairy farming” is explained by the fact that producing calves as a by-product from milk production emits less than producing calves in a suckler cow system.
-20 -10 0 10 20 30 40 50 60 70 80
Mixed dairy farming Specialised dairy farming
g O 2 -equ iv. /k g m ilk P NO3 NOx NH3
Figure 8. Potential eutrophication (max scenario) per contributing substance
The environmental impact of eutrophicated emissions is regional or even local, as over-eutrophied watercourses and nitrate in drinking water supply. Hence it is necessary not only to look at the eutrophication per unit milk, but also per area land used. A low impact per unit milk resulting from a high intensity can result in high emissions per unit land used. In Figure 9 the potential eutrophication caused by on-farm activities per unit land used on-farm is presented. Note that no avoided emissions for culled cows and live calves is done, neither is the land used for feed production outside the farm included. The emissions of phosphorus presented originate from the production of purchased feed.
0.00E+00 1.00E+01 2.00E+01 3.00E+01 4.00E+01 5.00E+01 6.00E+01 7.00E+01
Mixed dairy farming Specialised dairy farming
g O 2 -eqiuv. /m 2 NO3 NH3
Figure 9. Potential eutrophication per square meter land, on-farm activities and land used at the farm, no emissions subtracted due to culled cows and live calves.
Acidification
The weighing factors for acidification are presented in Table 39. We have chosen to present the maximum scenario, i.e. nitrogen compounds contribute.
Table 39. Weighing factors used for acidification, max scenario
Substance mol H+ per kg
NH3 58.7
NO2 21.7
NOX 21.7
SO2 31.2
The acidifying emissions largely originate from on-farm activities, for scenario “Specialised Dairy Production” the purchased feed also contributes. The avoided emissions are high, which means that scenario “Mixed Dairy Farming” has lower total emissions per kg milk, in fact the emissions are negative, i.e. the avoided emissions from the live calves and meat are higher than the total emissions from the dairy farm system. The large avoided emission is a result of the suckler cow system emitting more ammonia both per calf and per kg meat, which in turn can be explained by the choice of meat production system for the systems expansion, which is a system with relatively large emissions of ammonia.
-0.3 -0.2 -0.2 -0.1 -0.1 0.0 0.1 0.1 0.2 0.2
Mixed dairy farming Specialised dairy farming
mol H + - equiv./kg m ilk Purchased feed Mineral fertilisers Farm Live calves Meat, culled cows
Figure 10. Potential acidification (max scenario) for the two scenarios, for different activities
In Figure 11 the substances contributing to acidification is presented. The larger emission of sulphur dioxide in “Specialised dairy farming” is a result of the transport of soy from South America, together with drying of beet pulp and transports of palm kernel expel from Asia. The higher emissions of NOX are also an effect of more transports in that scenario.
-3.E-02 -2.E-02 -1.E-02 0.E+00 1.E-02 2.E-02 3.E-02 4.E-02 5.E-02 6.E-02 7.E-02
Mixed dairy farming Specialised dairy farming
mol H + -eqiuv ./ kg m ilk SO2 NOx NO2 NH3
Discussion
The results show that scenario “Mixed Dairy Farming” is better in most included impact categories; the exception is land use and eutrophication per kg milk. A high land use is not a solely negative result; high land use is important in order to maintain the agricultural
landscape, both for aesthetic values and biodiversity. The type of land use is also important, large monocultures are less valuable than varied crop rotations. The negative side of a high land use is the competition of productive land, if less land is needed for food production, more land can be used for other purposes as energy crops. The lower effect on eutrophication per kg milk for scenario “Specialised Dairy Farming” is a result of several thing, mainly the used emission factors for ammonia from pasture which are very high compared to most recent research (Webb et al., 2005) which favours the Specialised system where grazing is very limited. Moreover, the nitrogen surplus in scenario “Specialised Dairy Farming” is much higher than in scenario “Mixed Dairy Farming”; despite that the calculated emissions are lower.
The perhaps most important conclusion from the present study is that it shows the importance of including the production of live calves and meat, as co-products from the milk production. If this aspect were omitted, the results would be very different and wrong conclusions could be drawn.
Several important environmental effects are not included in the study, as effect on biodiversity and landscape. These aspects are to some extent included in the scenario development; scenario “Mixed Dairy Farming” is developed in order to strengthen the biodiversity and landscape aesthetics by higher degree of variation in the crop rotation and larger share of pasture. It would be interesting to quantify or in a structured way discuss these effects.
The scenarios were developed using different priorities between sustainability goals; each scenario was designed in order to fulfil certain goals. However, these ambitions were not fully reflected in the results. Scenario “Specialised Dairy Farming” partly aimed at minimising the “global environmental impact” which was operationalised as “minimise the environmental impact per unit milk”, the results showed that the scenario “Mixed Dairy Farming”, which was designed to minimise the emissions per area land, turned out to have lower environmental impact per litre milk as well. This is a result of the qualitative working process for scenario design (described in Sonesson et al., 2005), where several choices about the system is made based on the research group’s experience and not on calculations on every choice. This is a methodological problem within scenario design and a discussion about it is presented in Sonesson et al. (2005).
The data used comes from a wide range of sources; research reports, recommendations how to calculate national statistics for emissions, extension services and personal communications with people working in practice. This leads to different accuracy in the data used. However, since the aim of the main study (Sonesson et al., 2005) was to describe the major differences and order of size for environmental impact for the two future scenarios, we judge the data quality to be sufficient. Since the scenarios are describing hypothetical future systems, it is necessary to use more or less hypothetical data. A very important note is that the conclusions must be adjusted to the fact that some data are less accurate than other; hence conclusions must be drawn with consciousness.
The nitrogen turnover is very important for many of the environmental aspects from the system. The modelling of nitrogen emissions is performed with different models; and some of the models use very general data. This is an area for improvements.
One of the most important assumptions made in the study is made in conjunction with the systems expansion. The results clearly show the importance of including the meat from culled cows, but even more important the live surplus calves delivered from the system. In this study we assumed that meat and calves replaced an organic suckler cow system. The suckler cow system was rather extensive in terms of pasture and roughage feed, which lead to high emissions of area based emissions as nitrate and also high ammonia emissions. It would be interesting to make a sensitivity analysis using another meat production system for the systems expansion; unfortunately this was not possible within the present project.
The data on purchased feed is less detailed than the data on the studied farms, and a large share of the purchased feed components are by-products from food industry (as beet pulp, palm kernel expels and spent grain) and allocations have been used witch increases the uncertainty. Since purchased feed are more or less only used in one scenario, the comparison might be affected. An important aspect for some of the imported feed components, as palm kernel expeller and soy meal, is that the production of these crops cause environmental impact not considered in this study, as soil erosion and loss of biodiversity when new land is
cultivated.
References
Cederberg, C. and Darelius, K., 1999, Livscykelanalys (LCA) av nötkött – en studie av olika produktionsformer (Life cycle assessment (LCA) of beef – a study of different production forms, in Swedish), Naturresursforum, Landstinget Halland
Cederberg, C. and Flysjö, A., 2004, Life Cycle Inventory of 23 Dairy Farms in South Western Sweden, SIK-Report 728, SIK – The Swedish Institute for Food and Biotechnology,
Göteborg, Sweden
Cederberg, C. and Nilsson B., 2004, Livscykelanalys (LCA) av ekologisk nötköttsproduktion i ranchdrift (Life Cycle Asessment (LCA) of ecological beef production, in Swedish), SIK-Report 718, SIK – The Swedish Institute for Food and Biotechnology, Göteborg, Sweden CEN, 1997, Environmental management – Life Cycle Assessment – Principles and
Framework, EN ISO 14040, European Commission for Standardisation, Brussels, Belgium CIT Ekologik, 2004, Energy- and Transport Database CIT 3j (Database to LCAiT, LCA-software), CIT Ekologik AB, Göteborg, Sweden
Dustan, A., 2002, Review of methane and nitrous oxide emission factors for manure management in cold climates, JTI-Report 299, JTI – The Swedish Institute of Agricultural and Environmental Engineering, Uppsala, Sweden
FOOD 21, 2004, Annual Report, FOOD 21, Swedish University of Agricultural Sciences, Uppsala Sweden
Halberg N, Kristensen I S, Dalgaard T. 2000. Linking data sources and models at the levels of processes, farm types and regions. In: Agricultural Data for Life Cycle Assessment, Vol 1 pp 16-30. Eds: Weidema B P and Meeusen M J G. Agricultural Economics Research Institute, LEI, The Hague, NL.
Halberg, N. (ed), 2004, Life Cycle Assessment in the Food Sector – Proceedings from the 4th
International Conference, October 6-8, 2003, Bygholm, Denmark, Dias Report 61, DIAS, Danish Institute of Agricultural Sciences, Tjele, Denmark
