Food, land and greenhouse gases The effect of changes in UK food consumption on land requirements and greenhouse gas emissions. Report for the Committee on Climate Change.

Abstract

EXECUTIVE SUMMARY •1. Key findingsThis study examines the land use and greenhouse gas implications of UK food consumption change away from carbon intensive products. It shows that the UK agricultural land base can support increased consumption of plant-based products arising from the reduced consumption of livestock products. A 50% reduction in livestock product consumption reduces the area of arable and grassland required to supply UK food, both in the UK and overseas. It also reduces emissions of greenhouse gases from primary production by 19%. A switch from beef or sheepmeat (red meat) to pork or poultry (white meat) reduces food consumption related greenhouse gas emissions and the land area required but increases overseas arable land use. With this exception, the release of arable land now used to grow animal feed exceeds the additional arable land required for increased plant based foods in both the UK and overseas. Reducing livestock product consumption also has the potential to enable delivery of other significant environmental benefits, for example, reductions in ammonia and nitrate emissions. A 50% reduction in livestock product consumption reduces UK grassland needs for UK food production by several million hectares. This land could be used to supply livestock products for export markets although our scenarios assume that the proportions of imports, domestic production and exports remain constant. In these circumstances, some of the grassland released could be used to produce arable crops, including crops for biofuel production. Almost all of it could be converted to woodland or managed in other ways for biodiversity and/or amenity purposes. Conversion of this land resource to woodland has significant potential to increase soil carbon storage while supplying biomass for energy. Scenario Cropped area required, kha Grassland area required, kha Total area, kha Greenhouse gas emissions, kt CO2e/ year * UK OS Total UK OS Total UK OS Total Baseline 3,388 4,458 7,846 11,228 1,944 13,172 21,018 51,693 29,001 80,694 50% reduction in livestock with land release priority: Uniform 3,123 4,131 7,254 4,161 700 4,861 12,115 36,282 29,456 65,738 Maximise non-tillable land release 3,123 4,131 7,254 2,905 700 3,605 10,859 36,246 29,451 65,697 Maximise release of tillable land 3,123 4,131 7,254 7,102 700 7,802 15,056 36,282 29,457 65,739 Red to white meat with land release priority: Uniform 3,443 4,908 8,351 3,879 486 4,365 12,716 45,812 27,575 73,387 Maximise release of non-tillable land 3,443 4,909 8,352 2,909 486 3,395 11,747 45,867 27,572 73,439 Maximise release of tillable land 3,443 4,908 8,351 6,947 486 7,433 15,784 45,878 27,575 73,453 50% reduction in white meat consumption: Uniform 3,201 3,735 6,936 11,228 1,944 13,172 20,108 49,525 28,500 78,025 * The greenhouse gas emissions do not include possible effects of land use changeSummary table. The area of land needed to supply UK food and the greenhouse gas emissions from food production under current circumstances and under the seven scenarios studied. In a reduction scenario, concentrating remaining livestock production on different land types (e.g. concentrating on intensive production on lowland farms versus extensive production on lower quality land) has little effect on greenhouse gas emissions from primary production. This indicates that there is relatively little scope to reduce emissions by restructuring production (at least restructuring in relation to land use). It is further noted that concentrating livestock production on higher quality land would cause an almost complete closure of production for UK markets on land not suited to intensive grass or arable production, with biodiversity and economic impacts (discussed further below). The risks of unintended consequences with respect to greenhouse gas emissions are relatively low given the assumptions in the scenarios, but the actuality of such change will depend on future economic, social and political drivers. The report includes detailed analyses of land use and emissions data together with extensive discussion of a wide range of effects based on literature analysis. •2. Study objectivesThis study was conducted for the UK Government's Committee on Climate Change (CCC) to examine if UK agriculture can support consumption change away from carbon- intensive food products. For the purposes of the consumption scenarios, it is assumed the relationships between imports, exports and domestic consumption remain constant for each of the commodities used by the UK food system. The following questions were addressed: 1. Land needs: Given land quality considerations (e.g. land capability and constraints), to what extent is it possible to support a change in the UK consumption of meat and dairy products with a corresponding increase in substitute goods from UK agricultural land? Can a reduction in meat and dairy product consumption release land for other purposes? To what use would this freed-up land be suitable (e.g. food production, biomass production, carbon sequestration, other ecosystem service provision, forestry, etc.)? 2. Greenhouse gas emissions: What are the implications of the transition in production for GHGs both in the UK and abroad (including soil carbon releases, sequestration, reduced production of feed, etc, as well as reductions in direct N2O and CH4 emissions? 3. Other effects: What are the other implications, including for water, other pollutants, farm incomes, availability of manure as a fertiliser input, public health, ecosystem services, biodiversity, and animal welfare? 4. International implications: If UK agricultural land cannot support consumption changes, what are the international implications in terms of agricultural production and land-use displacement (e.g. deforestation, land for biofuels, land for food), and GHGs?•3. MethodsWe developed and used a combination of consumption and production scenarios to examine potential consequences of change. Life-cycle assessment (mainly life cycle inventory analysis) was applied to these scenarios to examine the overall effects of the consumption change on GHG and other emissions from primary production, in the UK and overseas. The production under the various scenarios was allocated to agricultural land resources by a combination of survey-based data analysis and model-derived calculations. Land use change (LUC) emissions (from changing soil C and biomass stocks) were calculated from data in the UK national inventory as well as from the UK Renewable Fuel Agency for overseas land types. Commodity flows as affected by consumption were calculated from FAOSTAT and Defra data. The resulting emissions were allocated to the various inventories in which they are registered, e.g. the UK's GHG inventories for agriculture, LUC, energy use and industry, together with those from overseas that are made up by components from our UK consumption of food and drink. Scientific literature relevant to the wider assessment of these scenarios was analysed (and an ecosystems services method was applied) to enable a qualitative assessment to complement the quantitative analysis. ScenariosWe designed a range of consumption and production scenarios to examine options on both the demand and supply sides. These comprise three consumption and three production scenarios. The consumption scenarios are as follows: Consumption Scenario 1. A 50% reduction in livestock product consumption balanced by increases in plant commodities. Consumption Scenario 2. A shift from red meat (beef and lamb) to white meat (pork and poultry). Red meat consumption is reduced by 75%. Consumption Scenario 3. A 50% reduction in white meat consumption balanced by increases in plant commodities. It must be stressed that the nature of scenarios is such that they contain a variety of assumptions about possible future demands and supplies of agricultural commodities. The scenarios are not forecasts. The focus has been on the technical capacity of land and agricultural production, not on the market changes needed to enable change. It should be noted that the balance of supply from the UK and overseas is assumed to remain as it is now. The 50% reduction in livestock products was not applied uniformly across these commodities. Under the reduction scenario (Consumption scenario 1), consumption of milk and eggs is 60% of current consumption, and meat consumption is 36% of current consumption. Sugar consumption is also reduced to align with healthy eating guidelines. Reduction in consumption of livestock products is balanced by increasing plant consumption on the basis of constant food energy supplied. Fruit and vegetable consumption was increased by 50% and basic carbohydrate (e.g. cereals, potatoes) and oil rich commodities (except palm oil) by 33%. Substitution was estimated on the basis of food energy use at the commodity level using FAOSTAT data. Expert opinion was obtained in relation to the viability of consumption change under Scenario 1. This indicated that diets at the consumer level under this scenario are viable from a nutritional viewpoint. It was also noted that Consumption Scenario 1 aligns with healthy eating guidelines in other countries. The production scenarios are focused on the intensity of use of different types of land. The result is a difference in the quantity and type of land ‘released' from production from change that reduces land needs. The production scenarios are: Production Scenario 1. Uniform land release - ‘pro-rata' changes in land requirements across land types. Production Scenario 2. Maximise release of tillable land - ruminant meat production concentrated on lower quality land. Production Scenario 3. Maximise release of low quality land - ruminant meat production concentrated on high quality land. The combination of consumption scenarios 1 and 2 and three production scenarios gives a total of 6 system scenarios. These are complemented by Consumption Scenario 3 giving a total of 7. •4. ResultsLand needsAll consumption change scenarios reduce the total amount of land estimated as required to support the UK food system. A switch from red to white meat increases the need for overseas arable land, although a larger area of UK land that can be tilled is released. Under a reduction scenario, the amount of extra land required for the direct consumption of plant products is less than the amount of arable land released from livestock feed production. The net effect on total overseas arable land needs is a reduction of about 311,000 ha and a net release of about 265,000 ha arable land in the UK. The need for grassland is greatly reduced. The release of grassland with some arable potential ranges between 1.6 to 3.7 million ha depending on where remaining production is concentrated. The release of grassland with no arable potential ranges from 0.7 to 6.9 million ha. Under a reduction scenario, concentrating remaining production on better quality land would almost entirely eliminate sheep and beef production for the UK from the hills, most uplands and less productive lowland areas. Under Consumption Scenario 2 (a shift from beef and sheepmeat to white meat from pigs and poultry), the diet needs of pigs and poultry result in a net increase in demand for overseas grown crops, although considerably more potentially arable land is released in the UK. More arable cropping is needed both in the UK (an additional 55,000 ha) and to a much greater extent overseas (about an additional 466,000 ha), driven largely by soy. However, the release of arable quality grassland in the UK exceeds the increase in overseas arable landed needed for producing this feed. The result is a net release of between 1.6 and 2.9million ha potentially arable land in the UK plus the release of 1.3 to 6.6 millionha of land suitable only for grassland. Under Consumption Scenario 3 (a 50% reduction in white meat consumption balanced by an increase in plant products) the changes are much less complex with no changes in grassland needs. Increases in demand for arable land for direct human consumption amounted to about 154,000 and 172,000 ha (domestic and overseas respectively), but these are more than compensated for by the release of arable land from feed production (341,000 and 668,000ha domestic and overseas respectively). Focusing a reduced cattle and sheep industry on non- arable land would result in the release of substantially more tillable land (currently grassland). In a 50% livestock production consumption reduction scenario, maximising the use of lower grade land (semi-natural grassland, hill land etc.) releases 3.7 million of tillable grassland (including 1.3million ha of good arable land). The opposite approach of withdrawing production from less capable land releases just 1.7 millionha of potentially arable land, with almost no release of the grassland well suited for to arable production. The land-use trade-off is therefore clear. Under a 50% livestock consumption reduction scenario, 2 million ha of tillable grassland is required to compensate for the withdrawal of cattle and sheep production from 6.9 million ha of non-tillable grassland. A 50% reduction in livestock product consumption opens up the opportunity to release about half of UK land currently used for UK food supplies if remaining production is concentrated on the more capable land. If land is released uniformly, almost two-thirds of this release takes place on grassland not suited to arable production and the remaining third is grassland with some arable potential. There would be with higher levels of land release in Scotland, Wales and Northern Ireland than in England. Depending on where the remaining production takes place, a large proportion of land released may be very unproductive, but it can be assumed that about 5 million ha with potential for other agricultural uses would be available, for example for the production of livestock for export (if they did not reduce their livestock consumption), for producing arable biofuel crops, planted woodland and re-wilding (to natural woodland in many cases). Greenhouse gas emissionsAll consumption scenarios reduce greenhouse gas emissions from primary production. The largest reduction is from a livestock reduction scenario (Consumption Scenario 1): from 81 to 66MtCO2e (19% reduction). The switch from red to white meat reduces emissions by 9% and a 50% reduction in white meat consumption by only 3%. The net effect on emissions depends greatly on the alternative use of the grassland released from food production. The study indicates the range of possible consequences on soil and biomass fluxes. If all tillable grassland released from food production was converted to arable use, 8 to 17MtCO2e per year would be released over 20 years through the effects of land use change. Converting all released land with the potential to support good tree growth to woodland would cause a net carbon uptake equivalent to about 7.5 to 9.5MtCO2e per year in soil and wood per year over 20 years. Land use preference (e.g. focusing remaining production on high quality land) has little effect on emissions. This is an important result indicating that supply chain emissions are unresponsive to changes in industry structure with respect to the land used. The location of emissions reductions (UK or overseas) was identified. Currently, we estimate that 36% of primary production emissions are overseas. All scenarios reduce UK emissions while Consumption Scenario 1 has little effect on overseas emissions and Consumption Scenario 2 reduces overseas emissions by 5%. None of the scenarios involve a net export of emissions and the GHG reduction benefits in the UK are proportionally greater than those overseas because of the tight link between UK livestock consumption and production. OTHER EFFECTS Other emissions All consumption scenarios are expected to reduce other emissions. Consumption Scenario 1 halves ammonia emissions. Reductions in nitrate emissions, eutrophication emissions generally, and acidification are almost as large (ca 45%). Biodiversity and carbon sequestration It is widely asserted that grassland, especially semi- natural grassland, has a higher biodiversity value compared with other types of vegetation, natural climax vegetation for example. It is often claimed that the retention of these grasslands is important for the continued delivery of some ecosystem services, for example, carbon sequestration. In many other European countries, the uplands and hills are usually wooded. For example, 32% and 29% of the land area in Germany and France respectively are wooded compared with 12% in the UK. Thus conversion to climax woodland or other forms of forestry is one obvious alternative use for released grassland. Our study has identified benefits for carbon sequestration in soil when grassland is converted to woodland (there should also be potential benefits in the use of harvested wood). Our analysis of land use statistics reveals the large proportion of UK land currently occupied by cattle and sheep. Without these livestock, this grassland (much of which is semi-natural grassland) would revert to the natural vegetation - deciduous woodland in many cases. Our results show that the use of livestock to retain semi-natural grasslands is not dependent on the current high level of livestock product consumption. A 50% reduction in demand still leaves a market which is large enough to support this activity. However, given how a declining market affects all suppliers, a livestock reduction scenario presents special challenges to the maintenance of semi-natural grasslands. Livestock systems provide a wide range of services that are currently used by society. In a reduction scenario, rural areas lose skills and employment in the livestock sector and there would be ramifications for linked industries such as the meat processing or veterinary sectors. Culturally important features, for example, hedgerows and stone walls, and much of the fauna and flora associated with grassland would be no longer needed. In the UK as a whole, land that is most likely to be taken out of production is associated with difficult production conditions. In England, upland moorland and common land now in a semi-natural state could change to fully natural vegetation cover. In upland areas, where the majority of re-wilding under Consumption Scenario 1 and 2 would be located, evidence suggests that various natural communities including scrub, bracken, bramble, and woodland with their own assemblage of flora and fauna are likely to develop, with potential increases in wild herbivores such as deer, hares, and rabbits. The majority of SSSIs currently under-grazed occur in lowland areas, for example in southern and eastern parts of England, and a lack of livestock results in difficulty in applying the grazing pressure required to maintain the semi-natural faunal and floral diversity. Recreational access to the uplands, which is now facilitated by open grassland landscapes, may be impaired and evidence suggests that visitors view the loss of traditional semi-natural landscapes, with associated meadows, hedges, and stone walls, negatively. Whilst a reduction in the current ecosystem service provision associated with livestock production from cattle and sheep can be expected under Consumption Scenarios 1 and 2, the net change is also dependent on the alternative use to which land is put. In upland SSSIs, overgrazing is often problematic and reducing grazing pressure may allow semi-natural habitats to recover, in particular dwarf shrub heaths, bogs, acid grassland and upland habitats. The release of large areas of land could also be used to diversify upland areas. For example, semi-natural upland woodlands have declined by 30-40% since the 1950s and the UK Habitat Action Plan has therefore included a target to increase the area of upland oak woodland through planting or natural regeneration of current open ground. In the lowlands, approximately 10% of the current arable land could be released for other activities, such as bioenergy crops, woodlands, recreational land, wetland creation, nature reserves, flood protection, carbon sequestration, and urban development. Each of these land uses will have its own specific range and flow of ecosystem services associated with it. While in general, the release of agricultural land with high environmental value from food production is not viewed as positive, Defra has concluded that there are likely to be situations where positive outcomes can occur. Economic considerations The reduction in the amount of land needed to supply the UK goes hand-in-hand with a reduction in the value added by agriculture supplying UK consumed food. A 50% reduction in livestock product consumption (Consumption Scenario 1) reduces the UK farm-gate value of livestock products from £7.6 to 3.5 billion. The farm-level economic impact of a change along these lines will depend crucially on what replacement output is found for the land released and on market effects that are beyond the scope of this study. One economic response scenario is that the land resource released remains in agriculture serving export markets. Another strategy is to use the land for non-food purposes. Using biomass energy cropping as a benchmark and assuming a price of £40/tonne dry matter biomass wood, we estimate that replacing the value of the food output of higher quality land released will be challenging, although it is reported that biomass energy is an economically viable alternative to sheep production on uplands.[1] Potential unexpected or unintended consequences Changes to UK crop production The general conclusion that a reduction in livestock production consumption will have little effect in total arable land requirements masks some important regional effects. This scenario will reduce arable crop production for livestock feed and increase arable production for direct human consumption, including a 50% increase in fruit and vegetables. The increase of 0.6 million ha of UK crops for human consumption includes an increase of about 0.2 million ha in potatoes, field vegetables and fruit. Research indicates that agricultural change driven by healthy eating recommendations will result in expansion of production of these crops particularly in the south and east of England.[2] Many of these crops are irrigated and some are protected using for example poly-tunnels. Whilst the change in land use is small in absolute terms, the local effects on water resources and landscape could be significant. It should be noted however that the increase in fresh fruit and vegetable consumption in these scenarios arise from the full implementation of current UK healthy eating guidelines (‘five-a- day') and are not just a consequence reduced livestock product consumption. Potential unexpected or unintended consequencesUneven distribution of economic effects The effect of a contraction in the value of farm output for UK markets will be unevenly distributed. There will be many losers, but also some winners. Given regional land quality characteristics, almost all Welsh, Scottish and Northern Irish farmers would be affected by output contraction counterbalanced by output growth in the south and east of England. Effects on overseas land use The reduction in livestock product consumption will have little effect overall on net overseas land needs. Release of land in South America and the USA used for animal feed, especially soy, will be counter-balanced by increases in a wide range of crops elsewhere. The consumption changes also reduce the need for overseas grassland. This affects three countries in particular: Ireland (dairy products, beef), New Zealand (butter and lamb), and South America (beef). The effect on Brazil is now small as imports have dwindled in recent years but the change would close off the UK as a growth market for Brazilian beef in the longer term. The effects on Ireland are particularly significant. •5. ConclusionThis study has clearly shown that UK land can support consumption change that reduces greenhouse gas emissions from the food system. The reduction in land needed to supply the UK that comes with a reduction in livestock product consumption brings potential environmental benefits and significant opportunities to deliver other products, including other ecosystem services, from UK agricultural land. The study has shown that some risks currently argued as arising from consumption change are small. In particular the study shows that arable land needs will not increase if the consumption of livestock products is decreased. The risk that emissions will be exported is also shown to be small. The identification of the significant potential benefits of consumption change combined with the low risks of unintended consequences has far-reaching implications for guidance to consumers and the development of agricultural policy. The results are broadly applicable to other European countries which means they are relevant to international policy development, for example the reform of the Common Agricultural Policy. [1] Heaton, R.J., Randerson, P.F., Slater, F.M. 1999. The economics of growing short rotation coppice in the uplands of mid-Wales and an economic comparison with sheep production. Biomass and Bioenergy 17: 59-71. [2] Jones, P.J. and Tranter, R.B. 2007. Modelling the impact of different policy scenarios on farm business management, land use and rural employment Project Document No. 13. Implications of a nutrition driven food policy for land use and the Rural Environment. Work package No