Report 2009:5
Greenhouse gas emissions
in pig meat production
decision support for climate certification
Written by Ulf Sonesson, Christel Cederberg and Maria Berglund
Translated by Mary McAfee
1
Contents
1Introduction
2Climate impact of pig meat production – summary of existing knowledge
2.1Conventional production
2.2Organic production
3Ways to decrease emissions of methane and nitrous oxide
3.1Improving nitrogen use efficiency
3.2Manure management
3.3Biogas production from manure
3.4Animal welfare – production
3.5suggested measures for decreasing methane and nitrous oxide Emissions
4Energy consumption
4.1Within-farm consumption of energy
4.1.1Heating
4.1.2Ventilation
4.1.3Feeding
4.2energy for transport
4.3Suggested improvement measures
4.3.1Improvements at investment
4.3.2Energy mapping
5Feeding
5.1Improving efficiency
5.2Using feedstuffs with lower emissions
5.3Increasing the proportion of locally grown feed
5.4Suggested improvement measures
6Proposed criteria for pig meat production
6.1Feeding
6.2Manure management
6.3Energy on the farm
6.4Animal welfare
7References
1
1Introduction
This report forms part of the project ‘Climate Labelling of Food’ initiated by KRAV and the IP Sigill quality system in 2007 with the aim of ‘decreasing climate impact by creating a labelling system for food through which consumers can make conscious climate choices and businesses can increase their competitive power’. The project is being run by KRAV and the IP Sigill quality system in partnership with Milko, Lantmännen, LRF, Scan and Skånemejerier. The Swedish Board of Agriculture is also participating as an associate in the project (
In spring 2009, the project commissioned the Swedish Institute for Food and Biotechnology AB (SIK) to draw up decision support for climate certification of beef, pig meat, chicken and eggs. This task was carried out by Ulf Sonesson, and the commissioning agents from the project were Anna Richert at Svenskt Sigill and Zahrah Ekmark at KRAV. In addition, Christel Cederberg, SIK, and Maria Berglund, Halland Rural Economy & Agricultural Society, were involved in producing this report.
Within the project, reports containing proposed criteria for fruit & vegetables, fish and shellfish, cereals and pulses, transport, animal feed and milk production have also been produced. A decision support report on packaging was completed in June 2009. A criteria report on lamb may be produced later in 2009.
The aim of the present report was to identify critical points in the life cycle of pig meat as regards the climate impact of this product. On the basis of this analysis, criteria for climate certification at product level are proposed. The starting point was mainly published Life Cycle Analyses (LCA) of the products, complemented with other relevant research and information.
Chapter 2 gives a detailed description of the climate impact of pig meat production, which provides the starting point for the remainder of the report. Chapter 3 deals with emissions of the biogenic greenhouse gases methane and nitrous oxide and identifies important aspects and measures. Chapter 4 deals with energy consumption on the farm and Chapter 5 with feeding. Chapter 6 then presents proposed criteria.
2Climate impactofpig meatproduction – summary of existing knowledge
2.1Conventionalproduction
Conventional pig meat is a production enterprise that is relatively uniform in Sweden.The animals are kept indoors and fed a ration consisting of grainand protein concentrate. The manure from fattening pig production is mainly managed as slurry. The two main systems in Swedish pig meatproductionareintegrated and specialist rearing, where thespecialistsystem means that piglets are bred by a specialist producer and delivered to the fattening pig producer, who fattens the pigs from around 20 kg live weight to slaughter at around 110 kg liveweight. Theintegrated model means that the pigsare kept by the same producer frombirthto slaughter. In addition, there is a sliding scale between these twosystems, with far-reaching partnerships between producers and integratedproducers who also sell piglets. Consumption of pig meat has increased somewhat since 1990. Between 1990 and 2005, consumptionincreased from 30.6 kg/capita to 35.9 kg/capita. During the same period,Swedishproduction decreased from the equivalent of 34 kg/capita to 30.5kg/capita,withimportedpig meatcoming mainly from Denmark.
Pig meatis aproductthat has been studied relatively well within research onlife cycle analysis. There are studiesfrom a number of countries in northern and western Europe, including several fromSweden. In general, the results do not differ very much between studies,although somedifferencesexist. Thesedifferencesare partly due to choice of method and issue and partly to actual differences inproductionsystems.
TheSwedishstudiesthat have been published are by Cederberg & Darelius (2001) and Anon (2002), both descriptive LCAstudies. Withinthe MISTRAS research programme Mat21a futures study was presented of three alternativepig productionsystems in which different sustainability objectives were prioritised (Cederberg & Flysjö, 2004; Stern et al., 2005). Strid Erikssonet al. (2005) constructed and developed a simulation model that wasused to illustrate various environmental aspects of pig production.A completely new study has been produced by Cederberg et al. (2009). Thisstudyis not a conventional LCA of a case study nature but a ‘top-down’ LCAstudy of all Swedish productionof animal-based foods, divided into different animal species. This will allow the climate impactofSwedish mean pig meatto be quantified. The outcome of the study is that, similarly to other studies, it will be possible to distinguish the parts of primary production that make the greatest contribution and also the gases emitted. The study can be regarded as the most comprehensivepresented to date.The study is still not published (August 2009), but will be in 2009 andthe values presented here are the final results.
Internationally, there is a pig study from the UK that forms part of a project on the climate impact of British agricultural products (Williams et al., 2006). These analyses are based on farm and plant production modelling combined with agricultural statistics. A French study analysed thedifferencesbetween efficientconventionalproduction (GAP, Good Agricultural Practice), productionaccording to the LR (Label Rouge) quality system and organicproduction in Brittany. That study was based on data from official statistics, the literature and a panel of experts consisting of researchers and advisors (Basset-Mens & van der Werf, 2003).
The results forgreenhouse gas emissions and energy consumption (secondaryenergy) from the abovestudiesare shown inTable 1.Bysecondaryenergy consumptionis meant directenergy consumption, i.e. the amounts of diesel, electricity and oilused on the farm, as opposed to primary energy consumption, which also includes the energyrequired for production and transport of fuel, or for electricity the amount of fuel requiredto generate the electricity.
Table1. Emissionsofgreenhouse gases and secondaryenergy consumption per kg bone-freemeatin conventional pig production, summary of published studies
Study / CO2-equiv./kg meat / MJ/kg meatSwedish studies / Total / CH4 / N2O / CO2
Cederberg & Darelius (2001) / 4.8 / 1.0 / 2.4 / 1.4 / 22
Anon. (2002) / 4.2 / 0.8 / 2.0 / 1.4 / 23
Cederberg & Flysjö (2004), Scenario A / 4.1 / 1.1 / 2.0 / 1.0 / 16
Cederberg & Flysjö (2004), Scenario B / 3.6 / 1.1 / 1.6 / 0.9 / 15
Cederberg & Flysjö (2004), Scenario C / 4.4 / 1.1 / 2.1 / 1.2 / 18
Strid Erikssonet al. (2005)a,b / 3.2-3.5 / 13-16
Cederberg et al. (2009)c / 5.2 / 1.3 / 2.6 / 1.3
Internationalstudies
Williams et al., 2006 / 5.6-6.4 / 14-17
Basset Mens & van der Werf (2003)a,b, Scenario GAP / 5.3 / 37
Basset Mens & van der Werf (2003)a,b, Scenario RL / 8.0 / 42
aIn this study,emissionsof each greenhouse gas are not presented
b Results converted from liveweight to kg meatwith 73% kill-out from liveweightto slaughterweight and 59% recovery from slaughterweighttobone-free meat
c Results converted from slaughterweight tobone-free meatwith 59% recovery.
As the results inTable 1show, there are variationsbetween studies. These variations are due partly to actual differencesbetween theproductionsystems and partly to differences inchoice of method and sources of data. In the older studies, the former weighting factors for methane and nitrous oxidehave been used (these were changed in 2007). The change involvedmethanereceiving a higher emissionsfactor and nitrous oxide alower factor, so how that affectsthe results varies between studies, but generally the change means that laterstudiesshow somewhat higher results.
Unfortunately,the distributionbetween greenhouse gases is not presented in all thestudies, but a common feature for the studies where this was done was that nitrous oxidemade up almost half the totalgreenhouse gas emissions and methane and carbon dioxideequal proportions of the other half.Nitrous oxide emissionsare mainly caused by feed production and the associated manufacture and use of mineral fertiliser. Methaneoriginates frommanure management, particularly storage.
Table 2showsemissionsofgreenhouse gases from pig productiondivided between activities and together with Table 1provides information on theactivitiesgiving rise to the different types ofemissions. Against this background, the potential for improvement can be identified.
Table2. Proportionofemissionsofgreenhouse gases arising from differentactivities
Study / Proportionofemissions (%)Feed (crop growing, inputs) / Animal rearing (manure,energy)
Cederberg & Darelius (2001) / 69 / 31
Strid Eriksson et al. (2005)a / 67 / 33
Cederberg et al. (2009)b / 57 / 43
Basset Mens & van der Werf (2003)a, Scenario GAP / 70 / 30
Basset Mens & van der Werf (2003)a, Scenario RL / 66 / 34
aResultsconvertedfrom liveweight to kg meatwith 73% kill-outfrom liveweight to slaughterweight and 59% recoveryfrom slaughterweighttobone-free meat
bResultsconvertedfrom slaughterweight tobone-free meatwith 59% recovery.
A consistent conclusion in all studiesis that the total nitrogen use efficiencyis a decisive factor for the final outcome. It is influenced mainly by nitrous oxideemissions, which are the single largest item. The other consistent conclusion is that feed conversion is critical, both for feed consumption per kg live weight gain and feed consumption per kg meat, i.e. a high recovery of meat from the animal. Choice of feedstuff is important, as the feed must be produced with low emissionsofgreenhouse gases. The fourth conclusion given in several of the above-mentioned studies is thatreproduction, i.e. the number of litters per sow, also affects the results. Having more litters weaned per sow ispositive for theclimate impactsince theemissionscaused by feed production for the sow and her manure are spread over a greater number of fattening pigs. However, thiseffect is lower than the first three listed.
2.2Organicproduction
In addition to the use of organic feed, organicpig productionmeans that the pigs are given access to outdoor grazing and forage. The manure is handled as solid manure and deep litter is oftenused. Organicpig production has a slightly lower intensity, which means somewhat fewerlitters weaned per sow and higher weaning age andslaughter age. In general, thenitrogen use efficiencyis worse, since syntheticamino acidsmay not be used, which means that there is some overfeedingof protein to ensure that some essential amino acids are supplied. The buildings used are considerably simpler than in conventionalproductionand, in addition, heating lamps are not used for piglets, both of which result in lowerenergy consumption.
When it comes tointernationalstudiesoforganicpig production, we only found one, Basset Mens & van der Werf (2003), who in addition toconventionalproduction and productionaccording to Label Rouge alsoanalysed anorganic scenario. One Swedish study has been published,withinthe research programme Mat 21. That study is based on two hypothetical farms, which are described with the help of advisors (Cederberg & Nilsson, 2004). We concluded that these two studiesare not enough to base proposed criteria on because they are too old and progress has been rapid, and because the results of the two studies differed greatly.
For this reason, astudyof Swedishorganicpig meatproductionhas been carried out in a sub-projectinitiated by theproject Climate Labelling of Food, co-funded by the Swedish Board of Agriculture within ‘A Food Strategy for all of Sweden’ (LISS). Thisstudyexamined twofarms, an actualproducer and a typical farm constructed in consultation with advisorswithinorganicpig production (Carlssonet al., 2009).
The resultsfrom thesethreestudies are presented inTable 3 and Table 4. It is difficult to definitively determine what caused the highemissionsofgreenhouse gases reported by Basset-Mens & van der Werf (2003), but it was probably a combination of low growth, inefficient feed production and composting of the manure, with subsequent lownitrogen use efficiency and higher ammonia emissions, leading to indirectnitrous oxideemissions. The latter led to a smaller proportion of the nitrogen being plant-available. We consider that this studyis not representative ofSwedishorganicproduction, but opted to include it in this report for the sake of completeness.
Table3. Emissionsofgreenhouse gases and secondaryenergy consumption per kg bone-free meatin organic pig production,summary of published studies
Study / CO2-equiv./kg meat / MJ/kg meatTotal / CH4 / N2O / CO2
Carlssonet al. (2009) / 4.8-4.9 / 0.7 / 2.4 / 1.7 / 21.5
Cederberg & Nilsson (2004) / 4.8-4.9 / 0.75 / 2.4 / 1.7 / 22
Basset Mens & van der Werf (2003) a, b,Scenario OA / 9.2 / 52
aIn this study, emissions of each greenhouse gas are not presented
bResultsconvertedfrom liveweightto kg meatwith 73% kill-outfrom liveweighttoslaughterweight and 59% recoveryfrom slaughterweight to bone-free meat
Table4. Proportionofemissionsofgreenhouse gases arising from variousactivities in organic pig production
Study / Proportionofemissions (%)Feed (growing, inputs) / Animal rearing (manure, energy)
Carlssonet al. (2009)a / 50 / 50
Basset Mens & van der Werf (2003) b, Scenario OA / 61 / 39
aResultsconvertedwith 59% kill-outfrom slaughterweight to bone-free meat
bResultsconvertedfrom liveweight to kg meatwith 73% kill-outfrom liveweight to slaughterweight and 59% recovery from slaughterweight to bone-free meat
Since there are fewer studiesonorganicpig productionthan onconventional, it is more difficult to draw general conclusions with the same certainty, but we consider that the twoSwedishstudies are of sufficiently high qualityto allow areas for improvement to be identified. In addition, the pattern as regardsthe gases that contribute most is the same as for conventionalproduction, approx. 50% nitrous oxide, 15-20% methane and the restcarbon dioxide. The distribution between feed-relatedemissions and farm-specificemissionsis also similar, but for organicproductiona somewhat higher proportionofemissions originate from the farm. The reason is that no commercial fertiliser nitrogen is used, so all nitrogen circulates on the farm.
Compared withconventionalproduction,organicproductionis characterised by higherfeed consumption, which is caused by the animals moving around more and by feed losses being greater outdoors. In addition, nitrogen use efficiency inthe animalsis lowerowing to the fact that it is difficult to get a balanced amino acid composition without the use of meat meal orsyntheticamino acids. According to Carlsson et al. (2009),nitrogen use efficiencyacross the animal is 26% inorganicproduction, while calculations based on data presented by Cederberg et al. (2009)show it to be approx.33% for all Swedishpig productionin 2005. Anotherfactorthat is indirectly linked to the environmental impactis that land use is considerably higherthan forconventionalproduction.
Afactorthat was not included in the studies above is emissionsofgreenhouse gases caused by the construction and upkeep of buildings and on-farm equipment. There is limited information on how this affects the overall results, butaccording to Frishknecht et al. (2007) these emissionsrepresent less than 10% of the total emissions for feed production. There are no data on animal production in that paper. Another studyof this area has been presented by Erzinger & Badertscher Fawaz (2001), who analysed the proportion of the energy inputs for milk production coming from buildings. The resultsshowed that this proportion can be up to50%. Sinceenergy-relatedemissionsconstitute a small proportion of greenhouse gas emissionsand sincepig productionhas not been studied, no far-reaching conclusions can be drawn from thatstudy, other than that it would be good to have a more in-depth studyofproduction under Swedishconditions.
3Ways to decreaseemissionsofmethane and nitrous oxide
Since the greatestproportionofgreenhouse gases from pig productionconsists ofnitrous oxideemissions, partly from the manufacture of commercial fertiliser and partly from nitrogen conversion in the soil, this is a logical area on which to concentrate. The area isrelatively complicated and the level of knowledge as regards nitrous oxide formation in soil is insufficient to allow specific measures for decreasing emissions to be identified. There are probably large variations in the amount ofnitrous oxideformed in arable soil, both between years and between regions or even between fields(Jungkunst et al., 2006). The method used to quantifynitrous oxideemissionsin thestudies presented above was the official method from IPCC (2007), which is a statistical method that calculatesnitrous oxide formation as a function of the amount of total nitrogen added to the soil. This results in measures to decreasenitrous oxideemissionslargely consisting of decreasing nitrogen flows in the system in general, while maintaining production levels. This is not a problem per se, as increasednitrogen use efficiency in agriculture has many advantages and is positive for many environmental targets.
As regardsmethaneemissionsfrom pig production, this is generally a case of manure management, particularly storage.
Nitrogen use efficiency in feed growing is included in the proposed criteria for feed and is only referred to briefly in this report.
3.1Improvingnitrogen use efficiency
In general, the nitrogen content of the feed should be as low as possible without affecting growth. Having a low nitrogen content in the feedgenerally gives alowernitrogen content in the manure, which in turn means that the risks of emissionsofnitrous oxide and ammonia are decreased (see more below under ‘Manure management’).In principle, the less nitrogen circulating in the system, the lower the riskofnitrous oxide formation. In the future scenarios that have been published (Stern et al., 2005) phase feeding and the use ofsyntheticamino acidswere identified as ways to keep the nitrogen content in the feed low. Frequent analyses of the protein in the feedstuffs, in terms of both quantity and amino acid composition, are a prerequisite for optimisingnitrogen supply.
Thestudyby Cederberg & Flysjö reported that nitrogen use efficiency for the entire rearing period (including the sow) wasbetween 37 and 41%, in other words that 37-41% of thenitrogen in the feed was used by the pig.Cederberg & Darelius (2001) reported anitrogen use efficiencyof 38% for the fattening pig phase. For organicproduction Cederberg & Nilsson (2004) reported anitrogen use efficiencyof 29 and 30% for the twofarms studied,whileCarlsson et al. (2009) found 26% for the entire rearing period (including thesow).Calculations ofnitrogen use efficiencyare complicated and sensitive to input data. There are not many studieswhere these have been done in such a way that comparable results are produced. Since we consider the data support to be too weak for a rule that can have a great effect on production, it is our opinion that criteriaonnitrogen use efficiency cannot be introduced at present, but that this is an interesting issue for future reviews, when more information is available.
3.2Manure management
Ammonia, (NH3), which is very volatile, is formed during storage of manure, both solid manure and slurry. Emissionsof ammonia mean two things: 1) The ammonia itself can contribute to nitrous oxide formation when it is oxidised and affects nitrogen turnover in the ecosystem on which it is deposited; and 2) the lower amount of nitrogen left in the manure leads to a greater requirement for supplying other nitrogen to the crop, as mineral fertiliser nitrogen, green manureor biogas digestor residues, which in turn have given rise toemissionsofgreenhouse gases.Covering the slurry tank is an effective way to decrease methane and ammonia emissions (more on this below).