United Nations Food Systems Summit 2021 Scientific Group https://sc-fss2021.org/

DRAFT July 06, 2021

Food Systems Summit Brief Prepared by Research Partners of the Scientific Group for the Food Systems Summit July 2021

Livestock and sustainable food systems:

status, trends, and priority actions. by Mario Herrero, Daniel Mason-D’Croz, Philip K. Thornton, Jessica Fanzo, Jonathan Rushton, Cecile Godde, Alexandra Bellows, Adrian de Groot, Jeda Palmer, Jinfeng Chang, Hannah van Zanten, Barbara Wieland, Fabrice DeClerck, Stella Nordhagen, Margaret Gill

Abstract

Livestock are a critically important environmental impacts for which the component of the food system, sector is responsible. We propose eight however, the sector needs a profound critical actions for transitioning transformation to ensure that it towards a more sustainable operating contributes to a rapid transition space for livestock. 1. Shifts in the towards sustainable food systems. This consumption of animal source foods, paper reviews and synthesises the recognising that reductions in evidence available on changes in consumption will be required, demand for livestock products in the especially in communities with high last few decades, and the multiple consumption levels, while promoting socio-economic roles that livestock increases in consumption of vulnerable have around the world. We also groups, including the undernourished, describe the nutrition, health, and pregnant women and the elderly. Diet

1 shifts alone will not produce the deep 1. Introduction transformations required, and the following actions need to be deployed There is global consensus of the at scale at the same time. 2. Continue need to transform food systems to work towards the sustainable achieve critical global goals at the intensification of livestock systems, intersection of human and planetary paying particular attention to animal well-being. The Sustainable welfare, food-feed competition, blue Development Goals (SDGs) stress that water use, disease transmission and to meet future needs we need to use perverse economic incentives. land more sustainably, minimise 3. Embrace the potential of circularity negative impacts on the environment in livestock systems as a way of and seek for opportunities to restore partially decoupling livestock from lands that have lost nutrients and/or land. 4. Adopt practices that lead to the biodiversity. Simultaneously it is crucial direct or indirect mitigation of to provide all people with access to a greenhouse gases. 5. Adopt some of more nutritious diet, and hence future the vast array of novel technologies at food systems must provide a diverse scale and design the incentive range of affordable foods to enable all mechanisms for their rapid people to have access to diets of high deployment. 6. Diversify the protein nutritional quality. sources available for human The livestock sector is an consumption and feed, focusing on the important part of these challenges, high-quality alternative protein since on one hand, it is a major user of sources that have low environmental land but on the other hand, it provides impacts. 7. Tackle anti-microbial food with high quality protein and has resistance effectively through a high levels of micronutrients. Over combination of technology and new recent decades, however, livestock regulations, particularly for the fast- production has grown rapidly in growing poultry and pork sectors and response to increasing demand, and its for feedlot operations. 8. Implement environmental footprint has grown to true-cost of food and true-pricing the point that the sector is now approaches to animal source food considered a major disruptor of global consumption. The scale of the efforts biogeochemical cycles, water use, on these actions will depend on the biodiversity loss and others. A large context and needs of each country or reduction in the environmental region, however, these actions will footprint of the livestock sector is need to be deployed simultaneously necessary to facilitate the continuation and in combination to ensure that of conditions that have allowed livestock contribute to sustainable humans to live on the planet and the food systems, leaving no-one behind. Earth’s current ecosystems to thrive.

2 Here we provide a synthesis of the policies, governance processes and current understanding of the dynamics institutions that might minimise of the livestock sector in terms of use negative interactions and maximise of natural resources, trade between positive synergies. We conclude with a countries and the synergies and trade- brief exposition of the possible offs caused by the changing nature of implications for the international the demand and supply of animal agricultural research agenda, along source-food (ASF, including milk, meat, with eight priority actions that need to eggs, and fish in this study). Drivers, be deployed simultaneously and in environmental and social issues are combination to ensure that livestock discussed in detail, and mechanisms contribute to sustainable food systems, for enhancing the synergies are leaving no-one behind. proposed. We discuss the kinds of

Table 1 Glossary of key terms

Key Terms Explanation Livestock sub- Domesticated terrestrial animal sub-sectors that include bovine (beef sectors and buffalo), dairy, sheep, lamb, goat, poultry, egg, and pig production. Livestock Products (food and non-food) derived from terrestrial domesticated Products animal sub-sectors. Animal Food products derived from both terrestrial and aquatic animal sources. Sourced Foods These include livestock food products, as well as food products derived from aquaculture, wild capture seafood, and hunting on land. Ruminants Terrestrial herbivores that have 4 stomach compartments to facilitate the digestion of fibre. Domesticated ruminants can be categorised as large (bovine, buffalo, cows) and small (sheep, goats, lamb/mutton). Monogastric Domesticated animals that have a single compartment stomach, this usually refers to pigs/hogs and fowls, which includes chicken, turkey, duck. Red Meat There are various definitions of red meat depending on geography and if the use is culinary or nutritional/dietary. In this report we follow the WHO (2015) definition where red meat refers to mammalian meat including ruminants and pigs/hogs. White Meat Following nutritional/dietary definitions white meat in this report refers to meat and meat products derived from poultry, other fowl, and seafood. Cropland Area dedicated to the production of food, feed, and biomass crops. This included both area for annual (e.g. cereals) and perennial crops (e.g. fruit trees). Rangeland Land type that can be used for livestock grazing and can vary substantially in terms of productivity, and tends to be characterised by native vegetation, but can vary on level of intensification and management. Pasture Land type that is dedicated for livestock grazing. Vegetation tends to be more managed than for rangelands and is primarily grasses and other forage crops. Feed Crop Crop that is grown primarily to serve as a feed for animals.

3 Food Crop Crop that is grown primarily for direct human consumption. Food crops can have co-products that can be used to feed animals. Feed-food A competition for natural resources (e.g. land) between different competition purposes; feed or food production.

2. Background and trends Figure S1). For example, while there was a nearly 35% increase in per capita In recent years, the analysis of meat demand (+11.27 kg/person/yr), trends of the livestock sector has and total per capita meat demand focused on understanding changes in increased for all regions between 1990 demand, supply, and trade of livestock and 2015, this increase is being driven products, together with its associated intensification and expansion dynamics by large increases in demand for and environmental impacts. Most poultry and pork, which saw increases analyses of demand projections start of 106 and 26% respectively. from Delgado et al's (1999) ‘Livestock Global demand for ruminant meat Revolution’ paper which built on (beef and mutton), however, has evidence that as incomes increase and followed a different trajectory, with societies urbanise, per capita per capita demand having remained consumption of livestock products increases. This, together with increases near 1990 levels (changed less than 1 in population, projected that the total kg/person/year on average globally). demand for livestock products would Within the beef trend we still see grow substantially. This phenomenon, substantial variation regionally, with often generalised, while mostly true, most regions exhibiting much bigger hides substantial heterogeneity in declines in beef demand than the terms of the types of livestock products that are likely to increase in demand global number would suggest. High and the locations of consumption income countries have seen large growth. Below we provide clarity on declines in per capita beef demand the dynamics of ASF demand and since 1990, with Europe, United States, supply. and Australia, with beef demand declining by 8.8, 5.8, and 6.5 2.1. Trends in animal source-food kg/person/yr respectively. Latin demand: 1990-2015 America (excluding Brazil), South Asia,

and Sub-Saharan Africa have also seen Averaged globally, over the last 25 declines in per capita demand for beef. years, per capita food demand of all Globally, this has been balanced out by ASF increased by more than 40 large increases in per capita demand in kg/person/year (FAOSTAT, 2018). China and (4.6 kg/person/yr or >300%), However, this number hides Brazil, (11.8 kg/person per year or substantial variation across regions and >40%), and Western Asia and North by commodity within ASFs, with Africa (2.2 and 3.3 kg/person/yr or several different trends operating in >40% and >50% respectively). Demand opposing directions (Figure 1 and

4 for mutton has followed similar increases in ruminant meat demand. In regional patterns as changes in higher-income countries, we observe demand for beef. small changes (around 5%) in per capita There is much less diversity of total meat demand, masking large trajectories in the trends for poultry. shifts in the makeup of meat demand, Per capita poultry demand has with substantial substitution of beef increased in all regions, with the only and mutton with pork and poultry. difference being the magnitude of the Global demand for dairy products is observed increase. The smallest growing at a similar rate to pork, but increase was in Eastern Africa and the with less regional variation with most United States of America, 27 and 32% regions seeing increasing demand for respectively in per capita demand of dairy products. poultry meat. All other regions Fish demand per capita globally experienced per capita demand of increased by more than 50%, with most poultry meat double. Regional pork regions seeing substantial increases, demand trends are more variable, but with the few exceptions being Eastern resemble poultry more so than beef, and Southern Africa, and the United with non-Muslim-majority regions States of America. However, the generally seeing substantial increases, increase is mainly in farmed fish as particularly in China, Southeast Asia, globally captured fisheries have been South America, and Australia. stagnant or declining. Demand In low-and middle-income satisfied by aquaculture has seen 6% countries, this increase in meat growth per year since 2001 with the demand has led to substantial per majority of the growth in low- and capita meat demand increases driven middle-income countries, especially more by large increases in demand of Asia (FAO, 2018). monogastric meats with only minimal

5

Figure 1: Change in animal source-food demand 1990-2015 (kg/person/yr). Source: Based on authors’ calculations from FAOSTAT (2018). All regional definitions follow FAOSTAT definitions. Regions are inclusive of selected countries (i.e. Eastern Asia includes China), which are reported individually to highlight key trends.

2.2. The role of trade in meeting international trade flows. In these two demand for animal source-foods regions, the value of international dairy and eggs trade is about the same as The increase in consumption in meat. Europe and Oceania are also the some countries has outstripped supply largest exporters of the ASF categories and this has led to substantial increases accounting for almost 85% of exports in international trade in ASF in the last of dairy and eggs. For meat, the main few decades. The value of exports exporting regions at the global level are globally has nearly tripled from around Europe (primarily pork), North and 59 in 1990 to almost 174 billion US$ by South America (beef, pork, and 2010, although total trade value poultry), and Oceania (beef and represents less than 20% of global mutton), which account for more than production (FAO, 2019b). 90% of global meat exports in value Meat in value terms has terms. Nevertheless, global trade contributed nearly two thirds of the statistics do not tell the full story with value of exports of livestock products respect to important regional trade globally. There are only two regions, patterns. Europe and Oceania, where meat does Most trade in ASFs is within the not dominate the value of ASF same region of origin, with most

6 imports coming from nearby countries. United States of America (in North For example, considering trade in America region, which is red). meat, much of the trade of pork in Trade in ASF in volume terms is Europe and mutton of East Asia and small compared to trade of feed. For Pacific is between other countries example, Galloway et al. (2007) within the region (e.g. Europe exports estimated that trade in meat and to Europe; Figure 2). However, while processed meat products accounted regional trade is the primary story in for less than one tenth of the volume of describing meat trade flows, there are trade in feed grains. This is a crucial a number of dominant trading observation, as these dynamics are countries that trade between likely to intensify to supply feed for continents (Figure 2; for example, fuelling the demand for pork and intraregional bovine meat exports are poultry in importing regions. This dominated by the Southern Cone of comes with substantial consequences South America (most of the green for land use and for environmental outside of the Latin America region row impacts, as depending on the land used in Figure 2), particularly Brazil, for producing the feed, it could lead to Australia (in East Asian and Pacific substantial embedded environmental region, which is blue), and the United impacts in overall ASF production. A States of America (in North America clear example is if imports of soybeans region, which is red)). Small ruminant increase in Asia, this could fuel export is dominated by Australia and deforestation in Brazil, a primary New Zealand (in East Asian and Pacific soybean provider. In other regions, region, which is blue), which are the other environmental dimensions primary source of imports for most would take precedence over emissions, countries. Europe and to a lesser with the potential for substantial losses extent North America are the primary of biodiversity and disruption of water exporting regions supplying the bulk of cycles in places (see Searchinger et al. traded intraregional pork. Intraregional (2015) for example, for Sub-Saharan trade in poultry is dominated by Brazil Africa). (in Latin America, which is green) and

7

Figure 2: Composition of 2010 regional imports of meat commodities by source of imports. The source of imports follow the colours given in the final column (i.e. imports from Europe are coloured orange, and from North America are red, etc.), so for example 91% of imports of bovine meat in Europe comes from other countries in Europe, whereas 62% of imports of bovine meat in the Former Soviet Union comes from countries in Latin America (FAO, 2019b).

8 2.3. The response of production to ASF (kg) has increased by more than meet the increase in demand: 60%, an increase of almost 2% per year The monogastric “explosion”, (FAOSTAT, 2018). Most regions intensification, and expansion exhibited substantial increases, with dynamics the largest production increases observed in Africa and Asia, which both ASF are produced under a broad increased their production of ASFs by range of production conditions, across more than 160% from 1990 levels, at all agro-ecological zones and under an annual rate of more than 4% per different intensification and resource year (double the global average). use efficiencies. Historically, the Higher-income regions, on the other production trajectories have closely hand, grew at a slower rate, with ASF followed demand with increases production in Europe actually declining observed in the production (Figure S2). by about 15% from 1990 levels. Since the 70s there has been a Across ASF commodities the ‘monogastric explosion’ with rates of fastest growth in production was for growth in animal numbers often poultry meat which nearly tripled exceeding 4% per year, and in meat globally since 1990 (Figure 3). All and eggs production in cases over 6-7% regions on average saw increased per year, globally. Greater availability production, with the global median of feed grains, rapid progress in increase in production across all genetics of animals with improved feed countries at 125% above 1990 levels conversion ratios, coupled with short (~3.3%/year growth). generation intervals and industrial Eggs, pork, and dairy production production methods which have all grew at a slower pace with production been underpinned by improved control increasing by 103%, 72%, and 56%, of infectious and production diseases, respectively. Eggs and pork similar to have made it possible to accelerate the poultry saw increases across most production of eggs, poultry and pork regions, with the median several fold in a short space of time. regional/country increase of 79% and Improvements in crop yields, improved 29% respectively. In low- and middle- feeding rations with high quality income regions dairy production grew feedstuffs, higher production at rates similar to poultry (108 and efficiency, favourable prices and the 203% in Africa and Asia respectively), involvement of the private industry in but saw much smaller growth rates in driving these dynamics played a developed regions, with an 18% significant role, initially in Europe, decline in dairy production in Europe. North America, and Oceania, and later Ruminant meat production grew in Latin America and parts of Asia (FAO, at a much slower pace than dairy and 2006). monogastric productions, with global Since 1990, global production of production of beef and lamb increasing

9 by 30% and 53%, respectively. Beef and intensification, have also taken place. lamb production globally grew about Intensification occurred at different 1/4 and 1/3 the rate of poultry, rates in different parts of the world and respectively, since 1990. For beef, most in some cases led to reductions in regions saw increases in production animal numbers. For example, the with the exception of Europe whose United States of America produces 60% production in 2015 was half their 1990 more milk with 80% fewer cows now levels. Lamb production in low- and than in the 1940s (Capper, Cady and middle-income regions grew at a much Bauman, 2009) through a substantial faster rate than the global average, change in genetics, feeding and with small ruminant production housing systems. Substantial increasing at rates similar to pork in intensification and also expansion of Africa and Asia. However, in developed the livestock sector has occurred regions there were declines in primarily in Latin America and Asia. production, with North America, This is in stark contrast with Sub- Europe, and Oceania seeing declines in Saharan Africa, where productivity has production of 58%, 49%, and 6% remained stagnant for decades, with all respectively from 1990 levels. the growth in production due to While increases in animal numbers increases in animal numbers. These and total production have occurred, general observations hide substantial substantial increases in production heterogeneity, which we disentangle efficiency, often associated with below.

10

Figure 3: Production trends of animal products (kg) from 1990 to 2015. Source: Based on authors’ calculations from FAOSTAT (2018).

11 2.4. Different livestock products and health services in these production production systems, different systems have been limited with patchy dynamics disease control, in particular over remote areas. The production increases in the In contrast, in temperate regions past few decades follow different and in high-income countries, total trajectories for ruminants than for pork production increases were mainly and poultry in smallholder or industrial driven by the increases in productivity operations. Between 2000 and 2011, rather than the increases in animal global milk and meat production numbers. On average, high-income increased by 28% and 11% respectively countries showed a decrease in total (Figure 4). Mixed crop-livestock animal numbers (-4%) while systems contributed to the majority of maintaining modest productivity bovine milk and meat production. In increases (under 1% per yr). 2011, mixed systems produced 505 Mt Only in the highlands of low and of milk and 42 Mt of meat with 608 middle-income countries, did increases million tropical livestock units (TLU). in dairy productivity (28%) outstrip the Grazing systems produced 45 Mt of growth in animal numbers (9%) as the milk and 10 Mt of meat with 192 source of growth in dairy production million TLU. between 2000-2011. This evidence of At the global level, these increases intensification is not surprising, in total production were mainly driven considering that the majority of by the increases in animal numbers Research and Development and (dairy: +19%, meat: +10%), followed by extension efforts have been directed the increases in animal productivities towards these smallholder, mostly (kg of livestock products/TLU/yr, milk: mixed, dairy systems. These regions +9%, meat: +1%). In arid and humid and systems have their own regions, or in low-income countries, constraints, like increasing human total production increases were mainly population densities, shrinking farm driven by the increases in animal sizes, feed deficits and soil fertility numbers rather than the increases in problems. These could limit the productivity. For example, in arid viability of dairy production in the long grazing systems, milk productivity run in these regions (Waithaka et al., stagnated while dairy animal numbers 2006; Herrero et al., 2010, 2014). rose by 27%. This reflects that the It is a concerning trend that feeding systems have remained static, ruminant production increases in many being reliant on animals grazing and regions are still driven mostly by harvesting energy from available land growth in animal numbers. This places instead of greater utilisation of new additional environmental burdens on forage crops or concentrate feeds. land, especially in regions with Similarly, improvements in animal vulnerable ecosystems. A continuing

12 trend could mean further land other hand, efficiencies increase, as degradation in arid regions and seen in e.g. broiler production systems, increase in deforestation or land need to be developed with care to conversion in humid regions. On the avoid animal welfare issues.

Figure 4: Average changes in dairy bovine milk and meat bovine productivities (kg/TLU/yr) and animal numbers in grazing systems (A) and mixed crop-livestock systems (B) by climate and income group. Period: 2000–2011. Data calculated based on productivity and animal number estimates by country, livestock system and climate type from Herrero, Havlík, et al. (2013). The climate category Arid includes semi-arid systems such as northern Australia. Grazing and mixed crop-livestock systems as defined by Robinson et al. (2011), income groups as defined by World Bank (2016). Figure adapted from Godde et al. (2018).

2.5. The role of smallholders in the countries (FAO, 2009) and production of ASF approximately 117 million people work in fisheries and aquaculture (Mills et An important element in the al., 2011). Livestock are responsible for debate of ASF production is who 17–47% of the value of agricultural contributes to it, who is benefiting and production in lower- and middle- whom do we need to target as primary income countries regions (Herrero, beneficiaries of research efforts. Grace, et al., 2013) and contribute Livestock production supports about income to 68% of lower- and middle- 650 million low-income small-scale income country households (FAO, producers in lower- and middle-income 2009), while also playing important

13 cultural roles (Thornton, 2010; towards more plant-rich diets could Herrero, Grace, et al., 2013). While create more jobs than animal men are often most represented in agriculture-based employment, with livestock production and fishing, potential improvements in gender women tend to be highly active in equality and occupational safety processing and sale of animal products (Saget, Vogt-Schilb and Luu, 2020). (Herrero, Grace, et al., 2013). At the In the future, will the smallholders same time, ASF-related livelihoods do be the engine of production growth or not necessarily entail high-quality jobs. will they be superseded by larger, more For example, ASF producers and fishing vertically integrated producers? This communities in lower- and middle- will likely be distinct for different income countries sometimes do not livestock species and products, as the earn enough to eat their own dynamics are very different for production (Thow et al., 2017; Annan ruminant land-based systems than for et al., 2018; Ravuvu et al., 2018). In monogastrics. We attempt to describe high-income countries, poor working it below. conditions in meat processing plant are Bovine milk and meat: Globally, well documented and, considering on- farms smaller than 20 ha produce 45% the-job mortality risk, fishing is among of bovine milk and close to 37% of the deadliest livelihoods. Women in bovine meat (Herrero, Philip K livestock value chains in particular may Thornton, et al., 2017) (Figure 5). lack appropriate recognition and However, important regional remuneration (Agarwal, 2018), and differences exist. Large farms (>50 ha) denial of women’s access to shared ASF dominate bovine milk (>75%) and meat resources, such as fisheries, creates (>80 %) production in North America, power imbalances that expose women South America, and Australia and New to abuse (Fiorella et al., 2019). In Zealand, which are regions with high improperly managed systems, animal levels of exports of these products. handlers can also be exposed to, and Conversely, farms smaller than 20 become the vector for, zoonotic ha produce the majority (>75%) of disease. Exposure to foodborne and bovine milk and meat in China, East zoonotic diseases may be particularly Asia Pacific, South Asia, Southeast Asia, high in settings where workers do not Sub-Saharan Africa, and West Asia and have adequate access to hygiene and North Africa. Very small farms (<2 ha) sanitation services. These jobs are are of particular importance in China, often disproportionately held by the where they still produce more than poorest or most vulnerable in a 60% of bovine milk and meat. These society—making the profile of very small farms are also of importance associated risk similarly inequitable. A in East Asia Pacific, South Asia, recent International Labour Southeast Asia, and Sub-Saharan Organisation study found that a move Africa, where they contribute more

14 than 25% of bovine milk and meat Farms smaller than 50 ha produce production. more than 45% of bovine milk and Bovine milk and meat production meat in Europe and more than 55% in are produced across a range of farm Central America. sizes in Europe and Central America.

Figure 5: The production of bovine milk and meat by farm size and region. Source: Data from (Herrero, Philip K. Thornton, et al., 2017).

The role of smallholders in the contrary, land fragmentation and feed future is uncertain. For dairy, a scarcity are two of the main issues sustainably intensified smallholder confronting these systems if they are to sector could be the engine of remain viable. production growth as there are still For beef, the situation is different. large yield gaps in these systems. In the absence of a clear increase in Furthermore, with demand primarily demand per capita, and with small satisfied by local markets (formal and farm output largely dependent on informal) and demand growing, increased numbers of animals, it is smallholders should benefit from likely that operation size will be more improved cash flow derived from of a constraint. Nevertheless, smaller growth in dairy. For intensification to scale production resulting from culled occur, markets, inputs and services and animals in diversified farming systems increased adoption of key may continue to be economically viable technological packages need to happen even if it will be unlikely to be the main at a faster pace than previously source of income or livelihoods. anticipated (McDermott et al., 2010; Pigs and poultry: A critical Godde et al., 2018). Data from the consideration for understanding the International Farm Comparison dynamics of the pork and poultry sub- Network has also demonstrated that sectors is to distinguish between the there are limited signs of consolidation fast-growing industrial sector and the of land in smallholder dairy (IFCN, smallholder sector in which women are https://ifcndairy.org/). On the strongly represented. The contribution

15 of smallholder systems to monogastric monogastric systems as a source of production based on data from pork, poultry and eggs in several Herrero, Havlík, et al. (2013), shows the regions: notably South and Southeast importance of smallholder Asia and Sub-Saharan Africa (Figure 6).

Figure 6: The proportion of pork, poultry and eggs from smallholder systems in different global regions (Herrero, Havlík, et al., 2013).

Gilbert et al. (2015) (Figure S3) capita. Countries with per capita found a negative relationship between GDP levels in excess of 30,000 USD the proportion of extensively raised tend to raise more than 95% of their chickens and pigs and the GDP per pigs in intensive systems.” (Gilbert capita of different countries. According et al., 2015, p7). to the authors: Although there are large variations “Below 1,000 USD [national between countries, this suggests that GDP] per capita, over 90% of as economies grow, the smallholder chicken are raised under extensive monogastric sector while still systems and the transition from important in some countries, will tend extensive to intensive production to reduce in importance as income really occurs between 1,000 and grows and conditions become more 10,000 USD per capita; above favourable for private industry to which most chickens are raised in industrialise the sector. The reduction intensive systems. For pigs, the in transaction costs and vertical transition zone—within which pigs integration will drive this transition as are raised under a mixture of it has occurred in other regions. The extensive, semi-intensive and question is not if but when? This intensive systems—extends transition presents a whole set of between 1,000 and 30,000 USD per different challenges to the extensive

16 poultry sector, as the dynamics of feed fertilisers and management of pests sourcing will increasingly play a key and diseases. Although rapid clearing role in the sustainability of the of tropical forests and savannas for industry, as will the impacts of agriculture continues, the current rates increasing density in industrial systems of clearing are relatively small with respect to disease dynamics compared to what happened in the (infectious diseases, antimicrobial temperate latitudes between 1850 and resistance and other issues) and 1950. As an example, Smith et al. managing local pollution. (2010) shows that for the period between 1990 and 2007, global 3. What are the implications of the cropland area increased by 3%, with historical supply and demand the biggest regional changes occurring dynamics of ASF for land use and in Africa (6%) and Latin America (9%). other environmental metrics? The world has around 3 billion ha of suitable land for crop production. A short historical perspective: We already use 1.5 billion ha for Ramankutty et al. (2018) recently feeding the world, with a third of this reviewed the trends in global area used to produce feed for livestock agricultural land use. This section is (FAOSTAT, 2018). The remaining 1.5 largely drawn from their findings. billion ha are currently occupied by Between 1700 and 2000, croplands forests that play a fundamental role in expanded from ~3-4 million km2 to our biogeochemical cycles and in ~15-18 million km2 (Figure S4). providing a broad range of essential Pastures expanded from ~500 million environmental services to humanity. km2 in 1700 to 3100 million km2 in These areas should be reserved, even 2000. Most of the cropland expansion when the short-term economic gains replaced forests, while most of the from conversion may be quite pasturelands replaced grasslands, attractive. Any expansion of croplands savannas, and shrublands, with some into rangelands is likely to be on more notable exceptions (e.g., the North marginal land, in more variable climate American Prairies were replaced by with subsequent lower yields than croplands, while Latin American those observed on current cropland. deforestation today is still mainly for Additionally, rangelands are important grazing. reservoirs of biodiversity and modest The global expansion of agriculture amounts of carbon, which suggests follows the history of human that their conversion would not be settlements and world economic ideal in places like Africa (Searchinger order. Agricultural expansion has et al., 2015). Hence, the pursuit of slowed down since the 1950s, primarily agricultural intensification. as agriculture intensified through Globally, total agricultural improved crop varieties, synthetic greenhouse gas emissions from have

17 risen as a result, primarily, of increases on biodiversity (Herrero et al., 2009; in animal numbers and land-use Barange et al., 2018). In particular, change. Livestock account for the livestock-induced land use conversion majority of greenhouse gas emissions is a major environmental and human from food systems through methane rights concern in some areas (De Sy et from enteric fermentation and manure al., 2015). Many intact ecosystems, management, carbon dioxide from notably carbon-dense and biodiversity- land use change and nitrous oxide from rich tropical forest biomes, have been manure management (Herrero et al., converted to pasture and feed crops 2016; Tubiello et al., 2021). However, for animals (FAO and UNEP, 2020). livestock now use 62% less land and These ecosystems are essential to emit 46% fewer greenhouse gas climate change mitigation (Lennox et emissions to produce one kilocalorie al., 2018). Intact ecosystems currently compared with 1961. These occupy half of the ice-free surface of productivity gains have been observed the earth (Dinerstein et al., 2017), and across the livestock sector, with gains this degree of intactness has been in the ruminant sector and especially proposed as a global limit (Newbold et dairy in Europe and North America, al., 2016; Dinerstein et al., 2017; albeit substantially lower productivity Leclère et al., 2018; Willett et al., gains than those observed for 2019)—implying an urgent halt to land monogastrics. Nevertheless, improved use conversion is needed. In extensive livestock productivity has required an rangeland practices in grassland and increase of 188% in the use of nitrogen savannah biomes, where large grazers fertilisers derived from fossil fuels to (e.g. bison) have been lost, ruminant increase feed production (Davis et al., livestock can be an important means of 2015) (Figure S5). Structural changes in biodiversity conservation and climate the sector, driven by the monogastric mitigation (Olff and Ritchie, 1998; explosion have been partly responsible Griscom et al., 2017). for this trade-off, as a third of the Resource use varies widely by type cropland, which uses most of the of ASF and production practice. Beef fertiliser, is now used to produce feed production tends to be the greatest for livestock. Despite productivity user of land and energy, followed by improvements, due to increased pork, poultry, eggs, and milk demand, the aggregate environmental production (de Vries and de Boer, impacts of livestock have continued to 2010). Fish, shellfish, and molluscs are grow, which will require substantial generally near the low end of the range further reductions in the sector’s (Poore and Nemecek, 2018). environmental footprint. Aquaculture is associated with Animal production practices, emissions and resource use, primarily depending on type and location, can from feed production (FAO, 2013), and have beneficial or detrimental effects can also pollute water and result in

18 habitat destruction (Barange et al., et al., 2018). 2018; FAO, 2019a). Capture fisheries have lower environmental impacts 4. The value of foresight: Were than aquaculture on some fronts but Delgado and colleagues’ put pressure on wild fish populations projections accurate for 2020? and associated ecosystems (Jackson et al., 2001; FAO, 2016, 2018; Barange et It is reasonable to review the 2020 al., 2018) which have been depleted by projections made by Delgado and inequitable natural resource access, others towards the end of the 1990s and poor governance (Leroy et al., against what is happening currently in 2020). The environmental impacts of the livestock sector. capture fisheries and aquaculture vary substantially across context, species, 4.1. Did the livestock revolution and production/harvesting practice really happen in the last 25 years? (Troell, Jonell and Crona, 2019). Overall, energy use per unit protein Globally, their projections of total production of fish/seafood is meat and milk production were 304 comparable to that of poultry and less and 772 million metric tons for 2020, a than other livestock systems (e.g., difference of only -12 and -5% from pork, beef) (FAO, 2019a). what current trends in FAOSTAT Resource use also varies by suggest. When we explore the production system and setting. In many projections by commodity, we observe cases, livestock can be reared in lands that the projections were particularly of low opportunity cost, without accurate for pork, with larger competing with croplands or other deviations for beef and poultry. These land uses (Van Zanten et al., 2018). deviations are offsetting, with an Livestock in grazing systems may have overestimation for beef and an some environmental benefits, such as underestimation of poultry production. conservation of grassland biodiversity, Delgado et al. (1999) were perhaps too though such relationships are complex conservative in their assumptions of and context-specific (FAO, 2009). technological change and the shifts in Animal production systems are often demand for poultry, which has essential to circular production increased its production by a factor of systems (Poux and Aubert, 2018). three rather than doubling, as they However, the intensive production of projected. The faster transition from any animal, including pigs and poultry, smallholder to industrial systems in has substantial environmental impacts, monogastric production, as described especially for surrounding by Gilbert et al. (2015), could have communities and waterways, that played a critical role in accelerating this must be considered (Wing and Wolf, change. The dynamics of this sector 2000; Burkholder et al., 2007; Godfray were simply faster than anticipated.

19

Table 2: Comparing global animal source-food production (million metric tons) in Delgado et al. (1999) to FAOSTAT (2018).

FAOSTAT Delgado et al. (1999) % Difference 1990 2013 2020a 2020 2020 Beef 55 68 72 82 14% Pork 69 113 125 122 -2% Poultry 41 109 127 83 -35% Meat 178 309 346 304 -12% Milk 538 753 813 772 -5% Note: a 2020 projection a linear regression based on FAO production values from 1990-2013

We observe a similar story with projections for China and India are not the per capita demand projections. as large as for milk, we should Overall, the projections are good with recognise a couple of important a difference of only 4 and 10 tendencies in these projections. First, kg/person/year difference for meat that while Delgado et al. (1999) and milk respectively. However, we can correctly projected a strong increase in see that similar to the beef and poultry meat consumption in China (even if projections there are offsetting they underestimated how large this deviations that are masked by only growth would be), the projected looking at the global number (Table 2). increases in meat consumption in India Here the key deviations are for do not appear to have materialized. projections for China and India (Table Income growth in India has not 3). There was an underestimation on translated into the expected increases increased demand of ASFs in China, in consumption across all commodities particularly for dairy products (31 (Alexandratos and Bruinsma, 2012), kg/person/year), with a similarly larger perhaps in part explained by the overestimation of milk demand in India relative slowness of the emergence of (33 kg/person/year). While the the intensive broiler sector in this differences on the meat per capita country compared to east Asia.

Table 3: Comparing per capita consumption of animal source-food (kg/person/year) in Delgado et al. (1999) to FAOSTAT (2018).

FAOSTAT Delgado et al. 1999 % Difference Meat Milk Meat Milk Meat Milk 1990 2013 2020a 1990 2013 2020a 2020 2020 2020 2020 China 25 62 73 6 33 43 60 12 -18% -72% India 4 4 4 53 85 92 6 125 44% 36% World 33 43 46 77 90 95 39 85 -16% -11% Note: a 2020 projection a linear regression based on FAO production values from 1990-2013

20 4.2. Animal source-food consumption underlying Delgado et al. (1999) trends: the 3 key storylines projections continue to be broadly true, recent trends do suggest that Reviewing these projections shifts towards beef may not be highlights that the evolution of the occurring in many countries. global livestock sector over the past Conversely, in many countries, couple of decades can be summarised particularly in high income countries, in a few storylines: there has been a trend towards a) First, demand for poultry has declining consumption of beef. This been the main global driver of decline is especially obvious in Europe increased meat consumption, with per which saw a reduction of more than 10 capita consumption having nearly million metric tons in beef demand doubled since 1990. This is a mix of since 1990. Nevertheless, when we changes in demand and supply. exclude China and Brazil, we can see b) Second, per capita dairy that per capita consumption in low- consumption in high-income regions and middle-income countries has not has stayed constant since 1990, with increased appreciably. any growth in total consumption driven Why is beef demand not growing by changes in population. Low- and with rising incomes like other ASF? middle-income regions have seen Perhaps this can be explained by the substantial increases in dairy price premium of beef vis-à-vis other consumption, with this being driven by meat options. Pork and poultry have both increases in population, and been 50% and 30% cheaper than beef, increasing per capita consumption of respectively, between 2010-2016 dairy products, with the largest according to the IMF (2017). increase observed in China. Additionally, messages suggesting that c) Finally, increases in global beef beef consumption is less healthy than demand is a story of two countries, white meats have been around since China and Brazil, which account for the 1970s. The emergence of Bovine nearly 93% of the 11 million metric ton Spongiform Encephalopathy caused increase in global beef demand, even drops in demand for beef in Europe in as globally per capita beef the 1990s and disrupted trade in North consumption has been declining or America in the 2000s. More recently stagnant in most countries. The key messaging on environmental outcomes role of China and Brazil in the global of beef through methane production beef sector was already identified by and deforestation may also be having Delgado (2003) in an update of their an impact on consumer confidence. 1999 projections.

While the trends for overall meat have largely followed projected trends, and suggest that assumptions

21 5. Animal-sourced foods and 5.1. It is possible for healthy adults to human nutrition and health: the meet their nutrient requirements from well-planned diets based solely need for moderation, not on plant-source foods avoidance. Diets that include few or no ASFs, There is strong and growing including vegetarian and vegan diets evidence that global transitions to have been shown to reduce the risk of healthy diets, as defined in most non-communicable diseases (Tilman national food-based dietary guidelines and Clark, 2014; Springmann et al., would lower climate and land impacts. 2016). Diets with diverse plant sourced In general, healthy plant-rich diets, foods can meet protein requirements including flexitarian, vegan, or (Young and Pellett, 1994), and vegetarian options, have lower climate vegetarian diets can meet adult and land impact than those high in ASF; micronutrients needs (Walker et al., their water and nutrient impacts 2005). However, plant-based foods do depend on the practices used not necessarily equal healthy foods: (Hallström, Carlsson-Kanyama and many highly processed foods are fully Börjesson, 2015; Aleksandrowicz et al., plant-based (e.g., highly processed 2016; Frehner et al., 2021). Reduction snack foods and sugar-sweetened in ASF, notably red meat, consumption beverages) yet have been associated has been shown to reduce with poor health outcomes (Hu, 2013; environmental impacts (e.g., on Marlatt et al., 2016; Mozaffarian, climate, land, and biodiversity), with 2016). some studies suggesting that achieving Controversy exists regarding global climate and biodiversity targets dietary recommendations for some is only achievable through reduced ASF and this has had a polarising effect consumption (Tilman and Clark, 2014; on many scientific and food sector Leclère et al., 2018; Springmann et al., discussions. These foods tend to be rich 2018; Clark et al., 2020). For example, in nutrients, but some specific ASF may transition to healthy plant-rich diets, also increase the risk of diet-related including some meat, would reduce chronic diseases and have harmful food-related emissions by nearly half, impacts on the environment. Most setting them on track to meet the 1.5°C controversial are the climate target (Clark et al., 2020). In recommendations regarding red meat contrast, a global transition to consumption, as beef production has increased consumption of ASF, notably one of the highest environmental red meat, is not feasible within footprints (Willett et al., 2019), but the recommended environmental limits health benefits and consequences (Springmann et al., 2018). remain contested in the literature. Consumption of red meat varies

22 substantially by region and country- staples (Headey and Alderman, 2019). level income classification. Global Relative caloric price varied by income intake of unprocessed red meat is levels ranging from 2.68 in upper- estimated to be 27 g per day (26-28g middle-income countries to 3.72 in per day) (Afshin et al., 2019). This is low-income countries. Regionally, higher than the recommended optimal unprocessed red meat was cheapest in intake established by the Global North America and Australasia and Burden of Disease research group to most expensive in the Middle East and reduce the risk of diet-related chronic North Africa. disease (23 g/day; optimal range: 18-27 The United Nations Food Systems g/ day) (Afshin et al., 2019) and Summit Independent Science Group’s substantially higher than working definition of a healthy diet recommended intake established by recognises that nutrient needs to the EAT Lancet commission for optimal attain ‘healthy’ diets vary across human and planetary health (7 g/day; individuals (Neufeld, Hendriks and optimal range: 0-14 g/day) (Willett et Hugas, 2021), and ASF can be al., 2019). Unprocessed red meat particularly important for reducing consumption was highest in Australasia undernutrition among vulnerable and Latin America and lowest in South groups in resource-poor settings. ASFs Asia and Sub-Saharan Africa (Afshin et are a high-quality source of protein, al., 2019). When comparing the micronutrients and bioactive factors estimated consumption of that are important for development. unprocessed red meat by World Bank Consumption of these foods may be income classification, low-income particularly essential for young countries have a per capita children and pregnant or lactating consumption of 8.2 g per day while women as these individuals have high-income countries have a per increased nutrient requirements due capita consumption of 45 g per day to biological processes (Neumann, (GBD 2017 Mortality Collaborators et Harris and Rogers, 2002; Murphy and al., 2018). Allen, 2003; Semba et al., 2016; Beal et al., 2017). ASFs are considered Differences in consumption may complete sources of protein that be due to cultural preferences, provide all nine essential amino acids. particularly in South Asia, but may also In addition, ASFs are nutrient dense arise from differences in affordability. and have higher bioavailability of key Interestingly, an analysis looking at the nutrients such as iron, vitamin A, and relative caloric price of foods globally zinc compared to plant source foods found unprocessed red meat to be the (Murphy and Allen, 2003). With most affordable ASF globally, but still at regards to undernutrition, most studies least three times higher than the price have assessed the role of ASFs in linear of the equivalent amount of calories growth for children under the age of from a standard basket of starchy

23 five and micronutrient deficiencies in unprocessed red meat consumption both women and children. Recent and type-2 diabetes, cardiovascular systematic reviews have identified disease, and cancer (Pan et al., 2011; limited evidence regarding the Mozaffarian, 2016; Qian et al., 2020). In association between consumption of 2015, the International Agency for ASF and linear growth during early Research on Cancer classified childhood. Both reviews concluded processed meat as a group 1 that substantial heterogeneity in carcinogen and unprocessed red meat definitions of ASFs might have led to as a probable carcinogen (IARC, 2015). inconsistent results (Eaton et al., 2019; On the other hand, in 2019, a Shapiro et al., 2019). On the other systematic review found “low hand, a cross-sectional analysis of certainty” of evidence regarding red Demographic Health Surveys found a meat and poor health outcomes strong association between ASF because of the limited data from consumption and stunting (with ASF randomised control trials and consumption reducing stunting), and heterogeneity in effect size of consumption of multiple ASF sources estimates between studies (Johnston had an additive effect on the et al., 2019). relationship (Headey, Hirvonen and In summary, populations Hoddinott, 2018). This analysis consuming high amounts of red meat, distinguished between dairy, egg, particularly in processed forms, would meat, and fish as types of ASFs, but benefit from decreased consumption authors were unable to look at to improve health and sustainability. associations between stunting and red This mostly applies to consumers in meat specifically. In addition, another higher-income countries but also, to a study found a strong correlation of ASF growing number in lower- and middle- intake and reductions in stunting in income countries, where the burden of Nepal and Uganda, with dairy diet-related non-communicable consumption having the strongest diseases is growing rapidly. For those correlation (Zaharia et al., 2021). vulnerable to undernutrition (whether The association between red meat in lower- and middle-income countries consumption and diet-related chronic or higher-income countries), the diseases is highly debated among nutrient contribution of minimally scientists. Evidence is clear that processed ASF may be beneficial to consumption of processed red meats is reduce risk of micronutrient deficiency detrimental to health, but the and promote growth (Murphy and relationship between unprocessed red Allen, 2003). meat and health needs further research. Evidence from epidemiological cohort studies has found positive associations between

24 6. Essential actions for ensuring needing higher levels of nutrients livestock’s contribution to including pregnant women, the elderly, sustainable food systems children and undernourished populations, particularly those in

lower- and middle-income countries. This section examines some These changes will require an alternative or additional actions that integrated approach that includes a would need to take place for livestock strong regulatory and fiscal framework to contribute to sustainable food and enabling environment in systems, while addressing critical combination with awareness raising aspects of social equity, poverty and and education to encourage other social goals. behavioural changes amongst

consumers, producers, and industry 6.1. Achieve a balance in the including new norms and standards. consumption of animal source foods that improves health and nutrition for all, and that helps reduce the 6.1.1. What sort of environmental environmental pressures of livestock gains could we expect from changes in consumption? production. Several studies have quantified the

potential environmental gains of As discussed in section 5, this will changing dietary patterns. This area of require different actions depending on work started from the need to quantify the context, including: greenhouse gas mitigation potentials  Consumption of ASF at a level of changing diets (Stehfest et al., 2009), appropriate to meet nutritional needs. and has been expanded considerably  A reduction in consumption of red to include health impacts and several and processed meat for populations additional environmental metrics with high risks of diet-related non- (Tilman and Clark, 2014; Leclère et al., communicable diseases or in the 2018; Springmann et al., 2018; Willett context of an unbalanced diet. et al., 2019). As an example, Figure 7  Enable increased consumption by summarises the technical mitigation nutritionally vulnerable populations potential of changing diets.

25

Figure 7: The technical greenhouse gas mitigation potential of changing diets according to a range of scenarios examined in the literature (Mbow et al., 2019).

The features of these studies show 4. The largest technical potential that: comes from reductions in ruminant 1. The upper bound of the meat consumption (most inefficient technical mitigation potential of sub-sector), as most scenarios try to demand-side options is about 7.8 Gt trigger land sparing (reduction of CO2-eq per year (no consumption of carbon dioxide emissions) as the key animal products scenario) (Stehfest et mechanism for reducing emissions. al., 2009). 5. Reductions in livestock product 2. Many dietary scenario variants consumption, especially red meats, have been tested. Key variants include could have both environmental and target kilocalorie levels (i.e., 2500 kcal health benefits (Tilman and Clark, per capita per day), notions of healthy 2014; Willett et al., 2019). diets, swaps between animal products 6. Full vegan diets could meet (red vs. white meat) and/or vegetables, calorie and protein requirements but and stylised diets (Mediterranean, can also be deficient in key nutrients flexitarian, etc.). All fit roughly (vitamin B12, folate, Zinc), a concern between the current emissions and the for vulnerable groups, in particular Stehfest et al. (2009) upper bound. those without access dietary 3. The main impact of reducing the supplements. Therefore, diets with consumption of animal sourced foods some level of animal products may be is to reduce the land footprint of necessary. livestock. This land sparing effect, 7. The economic mitigation coupled with alternative uses of the potential of changing diets is not land (i.e., negative emissions known. This is a crucial research area, technologies), leads to a large together with mechanisms for eliciting mitigation potential. Many of the other behavioural changes. environmental impacts are also 8. Most scenarios so far have associated with the land sparing effect taken kilocalories as the currency for (i.e., biodiversity, Leclère et al., 2018). changing diets, none have dealt with

26 protein or micronutrients, which from perspective, most of these studies use a livestock and a healthy diet fixed environmental impacts per kg of perspective seems like a necessary product and since they do not change step. through time they do not take into 9. Very few key examples of account the potential for food systems legislation and policy-induced shifts in redesigns. consumption exist. There are some examples that have been shown to Attached to livestock production is promote increases in consumption of fruits and vegetables (Garnett et al., an enormous amount of wealth 2015). generation, employment, value chains 10. The social and economic costs and famers livelihoods. Impacts on of reduced demand for ASFs are these are seldomly studied and they unknown. Notably there is little are crucial to create convincing policy information on impacts on farmers cases for a contraction of livestock income, employment, alternative product demand. Global studies that labour markets, reductions in have started to include some of these agricultural GDP, etc. critical feedbacks are only now starting 11. Methodological advances are to emerge (i.e., Mason-D’Croz et al., needed for eliciting simultaneously the 2020). environmental, health and socio- economic impacts of reduced From a nutritional perspective, consumption. there are also important These studies, while important, improvements to be made. All have only told part of the story and scenarios so far have used kilocalories have opened important research areas as the currency. However, livestock’s to complete the picture. These studies contribution to healthy diets are not so tend to lack information on the power much about their kilocalories as their of the private sector to adopt and micronutrients and protein. It is adapt technologies and make them essential to include these in future attractive to consumers. Many food research. Diets in these scenarios are companies are now seeing an also too ‘globalised’, and more advantage to plant-based alternatives realistic, and culturally sensitive to meat and milk as they may become regional variants will need to be more profitable as the technologies examined. The differentiated impacts mature. In addition, all scenarios have of ASF consumption and production modelled the impact of given diets, and across population cohorts, will require have not explored how the diets would that future analysis begin to better be achieved, which makes the ex-ante recognise the heterogeneity of evaluation of policies to shift demand populations (rural/urban, under or patterns difficult, if not impossible. over nourished, gender, age, or by age From a technological change groups), if they are to provide

27 necessary information to improve the adoption and reduce adoption lag targeting of future food policies. times.

Changes in consumption will not Efforts at sustainable be enough to achieve the intensification can have negative transformation required to achieve unintended consequences, which will healthy diets from sustainable food need to be addressed through systems. The next suite of ancillary appropriate regulation and policy actions will be required in tandem with action to ensure sought after consumption changes. environmental benefits are realised. The concept of sustainable 6.2. Sustainably intensify livestock intensification sounds to many as a systems win-win strategy to increase resource use efficiencies, but it is essential that Sustainable intensification has it also improves animal welfare been high on the agenda for some time (Garnett et al., 2013), and does not (Garnett et al., 2013). In livestock contribute to increased food-feed systems, successful examples exist but competition (Van Zanten et al., 2018). all have been associated with the To improve human and planetary availability of inputs (high quality health it is crucial to assess to what feeds, fertilisers, etc.), services extent sustainable intensification (veterinarians, extension) and in many strategies could bring us closer to cases, the development of markets and achieving the SDGs. their associated value chains (McDermott et al., 2010), as these are From a livestock perspective, most key incentives for systems to intensify well managed intensification practices (Herrero et al., 2010; McDermott et al., in the past have led also to improved 2010). Currently, adoption of better systems profitability and leading to feeding practices, such as improved increased production (i.e., pasture forages, have shown low adoption intensification and supplementation in rates. For example, Thornton and the tropics has substantially improved Herrero (2010) found 10-25% adoption milk and meat production). As a result, rates of dual-purpose crops, farmers have often increased the size agroforestry practices and improved of their operation (more animals, more pastures by farmers in selected low- land use changes) to increase even and middle-income regions, over a 10- further the economic returns. This 15-year horizon. Increasing adoption growth in turn has led to increased rates will require significant public and environmental problems (more private investment and institutional deforestation, increased greenhouse change to be able to increase farmer gas emissions, more land degradation, more temperature forcing). A critical

28 challenge ahead is how to regulate investment in low- and middle-income intensification so that it is truly efficient value chains (including market sustainable and equitable, operates development, service provision, within limits of production growth, adequate institutional support, etc.) protects biodiversity and other should be high in the development ecosystems services, and attains net or agenda. near net reductions in the use of resources. This is of particular 6.3. Implement practices that lead to importance, as having fewer animals, greenhouse gas mitigation co-benefits but of higher productivity, is essential explicitly, or indirectly to maximise the environmental benefits (i.e., reductions in greenhouse Mitigating greenhouse gases from gas emissions and land use) of livestock systems is more feasible in productivity growth in livestock some contexts than in others, and this systems (Thornton and Herrero, 2010). largely depends on the livelihoods This would imply reversing the objectives of livestock farmers observed trend of increased ruminant (Herrero et al., 2016). Still, many numbers as the main source of practices that improve productivity or production growth in low- and middle- the production system as a whole, can income regions towards productivity lead to direct and indirect greenhouse increases. gas mitigation co-benefits. These The degree of competitiveness of should be pursued. smallholders against imports from The supply side options for countries that can produce vast mitigating greenhouse gases in the amounts of animal products, at lower livestock sector have been the subject production costs, will be a crucial factor of the recent reviews (Smith et al., to determine the success of livestock 2007, 2014; Hristov et al., 2013; farmers in the low- and middle-income Herrero et al., 2016; Roe et al., 2019). countries, especially as the volume of These options look to: traded livestock products increase. - Reduce enteric methane of ruminants Formal and informal markets will need - Reduce nitrous oxide through manure to ensure the supply of cheaper, locally management of both ruminants and produced, safe livestock products to monogastrics adequately compete. This implies a - Implement best animal husbandry substantial reduction in transaction and management practices (all), which costs for the provision of inputs, would have an effect on major increased resource use efficiencies, greenhouse gases (carbon dioxide, and more responsive, innovative and methane and nitrous oxide) supporting institutions for the livestock - Directly sequester carbon from sector in low- and middle-income pastures (ruminants) countries (FAO, 2009). Hence, - Generally, improve land use

29 practices that also help enhancing soil regions with higher production carbon sequestration. efficiencies, and to spare land for the Excluding land use practices, land use sector to engage in negative Herrero et al. (2016) found that these emissions technologies to mitigate the options have a technical mitigation highest volumes of greenhouse gases. potential of 2.4 GtCO2eq/yr. However, Importantly, this can be done at low they also found that the economic consumption costs in many cases feasibility of these practices is low (10- (Havlík et al., 2014). A prerequisite to 15% of the technical potential, or less trigger the land sparing effect is also to than 0.4 GtCO2eq/yr). substitute the growth in production The largest mitigation from animal numbers for increases in opportunities for the livestock sector productivity and reducing animal occur when livestock are considered numbers, which will not happen unless holistically as part of the agriculture, we develop the appropriate incentives forestry and land use sectors (Havlík et systems that prevent rebound from the al., 2014, Figure 8). This is what gives intensification strategies, which are the flexibility to the ruminant sector to often profitable (Thornton and be able to relocate production to Herrero, 2010).

Figure 8: Total calorie abatement costs for livestock and agriculture and land use at different carbon prices ($5 to $100 / tonCO2) (Havlík et al., 2014).

30 6.4. Embrace the potential for significance as a human roughly needs circularity in the livestock sector 50 g protein per capita per day. This would mean that a global circular Van Zanten et al. (2018) recently livestock system could provide 40% of summarised studies focusing on the human protein needs with circularity in the livestock sector and substantially lower environmental found that at the global level, if impacts and no direct land needed for ruminant livestock were raised only in feed production. areas with no opportunity costs with Van Zanten et al. (2018) showed respect to growing crops, ruminants the consumption of ASF in different would be able to supply 3-7 g of protein regions of the world against the range per capita per day. This protein would of protein produced through circular come mostly from ruminant meat and livestock systems globally (Figure 9). milk grown in rather extensive Under such a system, we could keep conditions, where climate variability or within the circularity bounds: Africa agroecology would preclude crop and Asia could maintain the current production. They also found that if by- levels of ASF consumption and even products and other leftover streams increase them, however, all other from waste could be recycled and regions would require reductions in incorporated in rations for ASF consumption. This adds additional monogastrics, then 13-20 g protein per nuance to often polarised debate on capita per day could be produced and sustainable ASF production and fulfil vitamin B12 and half of the daily consumption, and should be the calcium requirements in a fully subject of a future research. decoupled way from land use. This is of

Figure 9: Animal-source food consumption by region (g protein per capita per day) against the lower (13 g per capita per day) and upper (20 g per capita per day) bounds of ASF supply through circular livestock systems (Van Zanten et al., 2018).

31 6.5. Adopt technological innovations nitrogen losses from croplands and in livestock production at scale agricultural greenhouse gas emissions could be decreased by 6% (0−13%), 8% Technological change in food (3− 8%), and 7% (6− 9%), respecvely. systems is occurring very rapidly and is These are encouraging results, the subject of considerable research considering that this is one of many (see Herrero et al., 2020, 2021 for potential technologies, and could reviews). Innovation in feed contribute towards reducing the production, digital technologies, environmental impacts of burgeoning robotics, genetics and many other monogastric demand. This technology fields are shaping agriculture is also in line with an extended circular considerably. Several of the emerging concept for the food system even as options have the potential to disrupt the next example. the livestock sector and contribute to Superfeeds: Superfeeds, like algae positive changes in the next decade if or grasses with high oil content are appropriate regulatory frameworks currently the subject of significant and social acceptability can be research. Walsh et al. (2015) studied achieved. Below we present a few the technical potential of algae systems examples of these and how they could as feedstock and showed that if increase the sustainability of the production were to be implemented in production methods in the livestock large scale in all regions where there is sector. potential to grow it, it could replace 2 Industrial feed production billion ha of grasslands and croplands. pathways: Engineers have created This could lead to substantial emissions methods to produce high quality reductions through avoided land-use microbial protein (85% protein) by change and land sparing, which could fermenting sewage with a source of be used for afforestation and carbon dioxide and energy. After rewilding. While they only cleaning, drying and pasteurising the demonstrated the technical potential material, this is transformed into a (economically this is still not feasible powder that can be used as an right now), it shows the boundaries of ingredient by the feed industry to what could be possible when the right replace protein sources like soybeans. sets of incentives are developed. Pikaar et al. (2018) recently found that Similarly, CSIRO have been developing by 2050, microbial protein can replace, grass varieties with 30% of oil in them, depending on socioeconomic mostly for biofuels (Vanhercke et al., development and microbial protein 2017). However, they could potentially production pathways, between be fed to livestock. This could disrupt 10−19% of convenonal crop-based the way we think about forage animal feed protein demand. As a improvement in the future, and if result, global cropland area, global deployed in suitable areas it could

32 change how ruminant livestock are Virtual fencing: Virtual fencing raised. Productivity could increase consists of collaring animals with GPS several folds if the energy density of devices with the coordinates of the the diets were to be dramatically areas they graze in (Campbell et al., increased. If coupled with reductions in 2018). If the animals trespass the animal numbers, this could also lead to designated grazing areas, they receive substantial mitigation effects. A a negative stimulus, and through challenge with this approach includes training they learn to keep in the the possibility that these new grasses designated areas. This could contribute could be more prone to pest attacks. to improved grassland management Considerable research is still needed in and pasture restoration, and reduce this area. the cost of extensive systems by Novel anti-methanogenic reducing the need for fencing and compounds: Significant progress has labour to manage herd movement. occurred in the last 4 years in Some of these grazing management identifying plants and/or compounds systems could also lead to higher that could substantially reduce productivity and to improved methanogenesis in ruminants. Two emissions intensities. notable examples, already on the Robotics/digital market but with increasing potential agriculture/sensors: Several start-up for commercialisation are companies are deploying digital Asparogopsis taxiformis algae, technologies in the livestock sector developed by CSIRO, which has shown with great success across a broad range reductions of 60-80% in methane of domains. These include monitoring production in cattle when fed at rates of welfare conditions for pigs and of 2-3 g per day (CSIRO, 2021). This poultry, disease surveillance, precision would be useful for confined animals, feeding, monitoring of physiological like in smallholder systems, or in status and others feedlots or dairy operations. The other (https://animalagtech.com/start-ups- compound is 3-nitrooxypropanol (3- transforming-the-livestock-industry/). NOP), which can decrease methane by up to 40% when incorporated in diets 6.6. Diversity protein production with for ruminants (Hristov et al., 2015). high-quality alternative protein These two examples could potentially sources with lower environmental reduce methane from enteric impacts fermentation, although these additives would have no direct effect on the land Diversifying the protein sources footprint of ruminants and the carbon for human consumption and animal dioxide and nitrous oxide emissions feed will be required as a critical action from ruminants will remain. for transitioning towards a more sustainable food system. Meat and

33 dairy analogues have a long history, Alternative proteins have seen rapid with tofu, seitan, and almond and soy growth in the last decade, and with milk consumed for hundreds if not increasing investments, some thousands of years (Shurtleff and projections suggest they could capture Aoyagi, 2014; Kemper, 2018). ‘Veggie substantial future market share, with burgers’ as we currently know them novel plant-based alternatives (25%) were introduced to mass markets in and cultured meats (35%) potentially the 1970s (Smith, 2014). Nevertheless, capturing the majority of meat as a new generation of protein expenditure by 2040 (Gerhardt et al., alternatives begin to enter the market, 2020). Such technology may be highly the attention being given to alternative disruptive to existing value chains and protein sources for human food and lead to substantial reductions in land livestock feed is burgeoning. These use for pastures and crop-based animal next generation technologies include a feeds (Burton, 2019). The resultant range of novel plant-based meat impacts on greenhouse gas emissions alternatives (e.g., Beyond Burger, depend on the meat being substituted Impossible Meats, etc.), insect-based and the trade-off between industrial proteins, and cultured meat and dairy energy consumption and agricultural products, all of which may displace land requirements (Mattick et al., conventional animal sourced-foods as 2015; Alexander et al., 2017; Rubio, well as first and second generation Xiang and Kaplan, 2020; Santo et al., vegan and vegetarian alternatives. 2020). Those alternative protein sources have Livestock feeds can use a variety of the potential to reduce the sources of protein, such as insect environmental impact (Parodi et al., protein. Insects are generally rich in 2018). protein and can be a substantial source The size of plant-based meat of vitamins and minerals. Black soldier market was between $4-5 billion in fly, yellow mealworm and the common 2018, or about 10 percent of the meat housefly have been identified for market, with rapid growth observed potential use in feed products in the over recent years (Gerhardt et al., European Union, for example 2020). Non-dairy milk alternatives (Henchion et al., 2017). Replacing land- reached $21 billion by 2015, doubling based crops in livestock diets with from the levels in 2009 (Bridges, 2018) some proportion of insect-derived and account for around 13% of the milk protein may reduce the greenhouse market (Sheikh, 2019). Substantial gas emissions associated with livestock investment in alternative proteins has production, though these and other been documented with the sector potential effects have not yet been receiving nearly $3.1 billion in quantified (Parodi et al., 2018). Other investments in 2020, a nearly 4-fold sources are high-protein woody plants increase from 2018 (Keerie, 2021). such as paper mulberry (Du et al.,

34 2021) and algae, including seaweed. use across livestock species and While microalgae and cyanobacteria production systems, and identify the are mainly sold as a dietary supplement main health problems that stimulate in the form of tablets and health drinks their use (Wieland et al., 2019; for human consumption, they are also Gemeda et al., 2020). In some used as a feed additive for livestock countries, this is also being linked to and aquaculture. Nutritionally, they surveillance of pathogens and are comparable to vegetable proteins. antimicrobial resistance profiles (FAO, The potential for cultivated seaweed as OIE and WHO, 2010), yet knowledge a feed supplement may be even gaps remain. There is uncertainty greater, and some red and green about how changes in antimicrobial seaweeds are rich in highly digestible use will impact on livestock production, protein. Novel protein sources may however, studies from Europe and have considerable potential for Southeast Asia indicate that reductions sustainably delivering protein for food and improved management can have a and feed alike, though their nutritional, neutral impact on productivity (Raasch environmental, technological and et al., 2020; Phu et al., 2021). socio-economic impacts at scale need The antimicrobial to be researched and evaluated use/antimicrobial resistance complex further. in livestock and aquatic species has multiple dimensions and multiple 6.7. Tackle anti-microbial resistance outcomes in terms of food production, effectively pathogen management, antimicrobial resistance change and consequent Livestock and aquatic species are a environmental and human health major user of antimicrobials and impacts. Interpretation of antibiotics (Van Boeckel et al., 2015). antimicrobial resistance findings This has raised concerns that poor requires a better understanding of the antimicrobial use in livestock inputs to the system, antimicrobial use, production will lead to increased and residues of antimicrobials in the antimicrobial resistance that will affect environment and animal products. human health and undermine Recognising the complexity of the antimicrobial treatments for humans. system, a study in the aquaculture These concerns have led to legislation sector in Vietnam on antimicrobial changes on the use of antimicrobials resistance risks showed the value of a for growth promotion in livestock, systems thinking approach to obtain starting first in Europe in 2006 and now desired objectives (Brunton et al., widespread and accepted at a global 2019). There is a need for more level. research on human behaviour across There have been efforts to the livestock and farmed aquatic food document the levels of antimicrobial systems, including the drivers and

35 motivators of antimicrobials use and technology appropriate for the context the role of human behaviour in and that is cost-effective and exposure to antimicrobial resistance sustainable; (3) Interventions that can risks. manage immediate problems with a Antimicrobial use occurs within focus on hygiene and waste water the context of regulations and management; (4) Effective enforcement, which includes communication of surveillance and legislation and policing as a framework intervention needs to government, the with actions guided by a combination private sector and wider society; and of private standards, market access, (5) Ensuring that mechanisms are in and social and cultural norms. In place for best practice in antimicrobial addition to intergovernmental use through improved antimicrobials standards, there are powerful stewardship. examples of the use of private standards to manage antimicrobials in 6.8. Implement true-cost of food and the food system. Countries with high true-pricing approaches to animal levels of antimicrobial use in terrestrial source food consumption and aquatic farmed species can be successful in exporting products with Transitioning to a sustainable food no detectable residues. Understanding system will entail reducing the the institutional environment within environmental, social and health costs which antimicrobial use occurs and the of food, while increasing the relative importance of public policy, affordability of food and improving the private company strategy and conditions of people who depend on individual incentives will be critical to food producing systems for their achieving sustainable antimicrobial livelihoods. For livestock systems this use. requires balancing many trade-offs and The antimicrobial simultaneously meeting various use/antimicrobial resistance complex sustainable development goals. is context specific. Achieving Finding pathways that can benefit sustainable antimicrobial use will likely multiple goals is challenging, as the size require substantial education and and value of the various costs and training of multiple actors within the benefits can be hidden. Typically, ASF system, as well as the development environmental, social and health costs of an effective surveillance system. The and benefits are externalised: not process should consider the (1) included in prices (Baker et al., 2020). Importance of understanding flows As a result, sustainable and healthy through the livestock and aquatic food is typically more expensive to systems with a focus on antimicrobials, consumers and less profitable to pathogens and antimicrobial businesses than unsustainable or resistance; (2) Surveillance that uses unhealthy food (Gemmill-Herren,

36 Baker and Daniels, 2021). This creates costs increase from poultry on the low a major barrier for transitions to end to milk and beef on the high end on sustainable livestock systems. average. However, within every One solution is true pricing or True species there is substantial variation of Cost Accounting, the systemic natural capital costs due to measurement and valuation of positive heterogenous production practices. and negative environmental, social, Subsistence systems can be particularly health and economic costs and efficient: these systems supply food to benefits (Baker et al., 2020; Gemmill- the most vulnerable populations, are Herren, Baker and Daniels, 2021). True well adapted to local constraints and pricing can create the right incentives have a low or even positive impact on to enable livestock food chains to biodiversity (Baltussen et al., 2016, reduce their environmental costs and 2019). provide healthy food. By also considering food security, affordability 7. Concluding remarks and and a living income for subsistence recommendations in the context of farmers, it also weighs the interests of the Food Systems Summit the most vulnerable people in the food system. Given the large variation in the The livestock sector will change, externalities of livestock systems, true voluntarily, or as a result of forces pricing can incentivise the most external to the sector. Our paper efficient food systems when provided a synthesis of the demand externalities are considered (Baltussen and supply dynamics of ASF, their et al., 2016). At a global scale, it can nutritional, health, and environmental help balance supply and demand for impacts, and the environmental trade- animal protein, shift consumption offs arising from the uses of land and towards the most sustainable and natural resources. We also showed healthy animal-based protein sources some alternative pathways of how the and shift production to those sector could develop depending on the production types and locations where goals and aspirations of different animals can be held with the lowest countries. In this sense, context is very effects on the environment. important, as what may work in one True Cost Accounting analyses of place may not be suitable for another. livestock have shown that the annual This initial targeting will be environmental costs of livestock fundamental to design actions and systems are substantial and Baltussen policies that profoundly improve and et al. (2016) estimates it over 1 trillion substantially change, in many cases, USD per year. At the same time, there the way we think about the roles of is substantial variation between types livestock. of animal food, regions, and Our study has demonstrated that production system. Natural capital the dynamism of the livestock sector

37 provides a range of avenues for breeding and feeding strategies. This in change, some more relevant to turn would also lead to reduced smallholders than others, and some pressures on land and to the more amenable to public funding than exploration of other greenhouse gas others, and some more likely to mitigation avenues, beyond those alleviate negative environmental explored to date (improved feeds, impacts than others. Picking the most manure management). Eventually this effective and desirable solutions will be could contribute to national mitigation essential for stakeholders associated action plans of specific countries. with the livestock sector to achieve the The future of the smallholder pork desired impacts on sustainable food and poultry sector: our synthesis has systems. The balance between social shown that while there are countries and environmental goals will need to were smallholder pork and poultry be carefully evaluated. The avenues for make an important contribution to the growth, the trade-offs and the supply of these products, in the coming potential actions can be summarised decades, much of the growth in below. production is likely to come from Smallholder dairy: The evidence industrial production, as integrated suggests that demand is growing fast supply chains emerge and the private for milk, and that at least in highland or sector engagement increases. This high potential areas, productivity per suggests that investing in these animal is increasing due to the smallholder systems is at best a adoption of better practices like feeds, medium-term strategy that could animal health management and provide livelihood benefits as these genetics. These systems can be producers diversify or identify new exit competitive, but issues surrounding strategies. Identifying transition land fragmentation and feed options for these producers in the availability need closer attention. future seems necessary. Testing and implementing From an international public good transformational feed technologies or perspective, the future of feed for engaging in developing systems that fuelling the large demand for pork and could increase in circularity, through poultry is a critical researchable issue, increased biomass recycling sound like if the feed is to be sourced sustainably. important next steps to ensure high Biomass value chains, old and new, quality feed at low environmental costs need to be evaluated, developed and in these systems. This needs to go promoted to ensure that competition beyond previous work on crop residues for food with humans is minimised. (e.g. Blummel and colleagues) and may Here, again, the development of need transdisciplinary partnerships circular feed sources, the development with other sectors to develop these of regulations for including a minimum new biomass streams and to adjust amount of recycled feed, and the

38 development of new feed sources most countries, and most of the (superfeeds from industrial production demand is driven by population or others), need to be developed and growth. At the same time, reducing red business cases for local industries to meat consumption could lead to take on these enterprises, well substantial greenhouse gas mitigation, planned. reductions in pressure on land and For monogastrics, there are a lot of biodiversity. Producing red meat only researchable issues, including on from lands of low opportunity costs, or antimicrobial resistance, with priority as a by-product from the dairy industry areas being: would have the lowest environmental 1. Monitoring inputs - what inputs impacts. are used in the system in terms of feed, Identifying the best levels of antimicrobials and other aspects that consumption in relation to other affect the health of the animals and dietary elements for different have implications on the health of population groups should be a high producers, consumers and those priority for the Food Systems Summit, working in the food chain. 2. Surveillance - establishing as well as identifying ways to decouple systems that generate information on red meat production from land, or to current and emerging diseases, create niche products for very specific antimicrobial use and antimicrobial sets of consumers through labelling resistance. systems and certification. 3. Assessment of the economic The livestock sector will change, burden of livestock health and wealth voluntarily, or as a result of forces (see https://animalhealthmetrics.org) external to the sector. If sustainability as a basis to identify interventions that concerns are of paramount impact positively on the economic importance, a critical research area is outcomes of livestock production as to develop economic incentive systems well as minimising impacts on the (price premiums) and regulations to environment and public health. pay for reduced emissions, watershed A central element of a livestock protection, biodiversity protection and agenda in relation to environmental others; and to internalise these in true trade-offs is related to the cost or true pricing schemes, identification of entry points for supported by adequate regulatory and engaging in the beef sector. On one fiscal measures. hand, the existing data shows that most of the growth in red meat 8. References production has been obtained through increases in animal numbers, while Afshin, A. et al. (2019) ‘Health intensification has been influential in effects of dietary risks in 195 countries, only a few countries. Consumption per 1990–2017: a systematic analysis for capita is stagnant, or decreasing in the Global Burden of Disease Study

39 2017’, The Lancet, 393(10184), pp. Tanzania; An application of the 1958–1972. doi: 10.1016/S0140- TEEBAgriFood Evaluation Framework. 6736(19)30041-8. Available at: https://flic.kr/p/3qRfW2 Agarwal, B. (2018) ‘Gender (Accessed: 21 June 2021). equality, food security and the Barange, B. et al. (2018) Impacts of sustainable development goals’, climate change on fisheries and Current Opinion in Environmental aquaculture: synthesis of current Sustainability, 34, pp. 26–32. doi: knowledge, adaptation and mitigation 10.1016/j.cosust.2018.07.002. options. Rome. Aleksandrowicz, L. et al. (2016) Beal, T. et al. (2017) ‘Global trends ‘The impacts of dietary change on in dietary micronutrient supplies and greenhouse gas emissions, land use, estimated prevalence of inadequate water use, and health: A systematic intakes’, PLoS ONE, 12(4), p. e0175554. review’, PLoS ONE, 11(11), p. doi: 10.1371/journal.pone.0175554. e0165797. doi: Van Boeckel, T. P. et al. (2015) 10.1371/journal.pone.0165797. ‘Global trends in antimicrobial use in Alexander, P. et al. (2017) ‘Could food animals’, Proceedings of the consumption of insects, cultured meat National Academy of Sciences of the or imitation meat reduce global United States of America, 112(18), pp. agricultural land use?’, Global Food 5649–5654. doi: Security. Elsevier B.V., pp. 22–32. doi: 10.1073/pnas.1503141112. 10.1016/j.gfs.2017.04.001. Bridges, M. (2018) ‘Moo-ove Over, Alexandratos, N. and Bruinsma, J. Cow’s Milk: The Rise of Plant-Based (2012) World Agriculture towards Dairy Alternatives’, Practical 2030/2050: the 2012 revision, ESA Gastroenterology, 21. Working Paper No. 12-03. Available at: Brunton, L. A. et al. (2019) www.fao.org/economic/esa ‘Identifying hotspots for antibiotic (Accessed: 22 December 2018). resistance emergence and selection, Annan, R. A. et al. (2018) ‘Public and elucidating pathways to human awareness and perception of Ghana’s exposure: Application of a systems- restrictive policy on fatty meat, as well thinking approach to aquaculture as preference and consumption of systems’, Science of the Total meat products among Ghanaian adults Environment, 687, pp. 1344–1356. doi: living in the Kumasi Metropolis’, BMC 10.1016/j.scitotenv.2019.06.134. Nutrition, 4(1), pp. 1–8. doi: Burkholder, J. A. et al. (2007) 10.1186/s40795-018-0209-z. ‘Impacts of waste from concentrated Baker, L. et al. (2020) ‘Prospects animal feeding operations on water for the true cost accounting of food quality’, Environmental Health systems’, Nature Food, 1(12), pp. 765– Perspectives, 115(2), pp. 308–312. doi: 767. doi: 10.1038/s43016-020-00193- 10.1289/ehp.8839. 6. Burton, R. J. F. (2019) ‘The Baltussen, W. et al. (2016) potential impact of synthetic animal Valuation of livestock eco-agri-food protein on livestock production: The systems: poultry, beef and dairy. new “war against agriculture”?’, Wageningen. doi: 10.18174/389545. Journal of Rural Studies, 68, pp. 33–45. Baltussen, W. et al. (2019) doi: 10.1016/j.jrurstud.2019.03.002. Traditional livestock systems in Campbell, D. L. M. et al. (2018)

40 ‘Virtual fencing of cattle using an Feed Science and Technology, 275, p. automated collar in a feed attractant 114766. doi: trial’, Applied Animal Behaviour 10.1016/j.anifeedsci.2020.114766. Science, 200(May 2017), pp. 71–77. Eaton, J. C. et al. (2019) doi: 10.1016/j.applanim.2017.12.002. ‘Effectiveness of provision of animal- Capper, J. L., Cady, R. A. and source foods for supporting optimal Bauman, D. E. (2009) ‘The growth and development in children 6 environmental impact of dairy to 59 months of age’, Cochrane production: 1944 compared with Database of Systematic Reviews, 2007’, Journal of animal science, 87(6), 2019(5). doi: pp. 2160–2167. doi: 10.2527/jas.2009- 10.1002/14651858.CD012818.pub2. 1781. FAO (2006) Livestock Report 2006. Clark, M. A. et al. (2020) ‘Global Rome. Available at: food system emissions could preclude http://www.fao.org/3/a0255e/a0255e achieving the 1.5° and 2°C climate .pdf (Accessed: 5 April 2019). change targets’, Science, 370(6517), FAO (2009) The state of food and pp. 705–708. doi: agriculture: Livestock in balance, Food 10.1126/science.aba7357. and Agriculture Organisation of the CSIRO (2021) FutureFeed. United Nations, Rome 2009. doi: Available at: https://www.future- 10.1016/S0140-6736(75)92740-3. feed.com/ (Accessed: 18 June 2021). FAO (2013) Food wastage Davis, K. F. et al. (2015) ‘Historical footprint: Impacts on natural trade-offs of livestock’s environmental resources. Rome. Available at: impacts’, Environmental Research www.fao.org/publications (Accessed: Letters, 10(12), p. 125013. doi: 18 June 2021). 10.1088/1748-9326/10/12/125013. FAO (2016) The State of World Delgado, C. L. et al. (1999) The Fisheries and Aquaculture 2016 - Coming Livestock Revolution, Contributing to food security and Commission on Sustainable nutrition for all. Rome. Available at: Development. New York. doi: www.fao.org/publications (Accessed: 10.1007/s10680-008-9165-7. 18 June 2021). Delgado, C. L. (2003) ‘Rising FAO (2018) The State of World Consumption of Meat and Milk in Fisheries and Aquaculture 2018 - Developing Countries Has Created a Meeting the Sustainable Development New Food Revolution’, The Journal of Goals. Rome. Available at: Nutrition, 133, pp. 3965–3971. doi: www.fao.org/publications (Accessed: 0022-3166/03. 10 June 2021). Dinerstein, E. et al. (2017) ‘An FAO (2019a) FAO’s work on climate Ecoregion-Based Approach to change - Fisheries & aquaculture. Protecting Half the Terrestrial Realm’, Rome. BioScience, 67(6), pp. 534–545. doi: FAO (2019b) FAOSTAT. Available 10.1093/biosci/bix014. at: Du, Z. et al. (2021) ‘Exploring http://www.fao.org/faostat/en/#data/ microbial community structure and TM (Accessed: 2 July 2021). metabolic gene clusters during silage FAO, OIE and WHO (2010) The fermentation of paper mulberry, a FAO-OIE-WHO Collaboration Sharing high-protein woody plant’, Animal responsibilities and coordinating global

41 activities to address health risks at the (London, England), 392(10159), pp. animal-human-ecosystems interfaces. 1684–1735. doi: 10.1016/S0140- FAO and UNEP (2020) The State of 6736(18)31891-9. the World’s Forests 2020: Forests, Gemeda, B. A. et al. (2020) biodiversity and people. Rome. ‘Antimicrobial Use in Extensive Available at: Smallholder Livestock Farming Systems http://www.fao.org/3/ca8642en/ca86 in Ethiopia: Knowledge, Attitudes, and 42en.pdf (Accessed: 18 June 2021). Practices of Livestock Keepers’, FAOSTAT (2018) FAOSTAT Food Frontiers in Veterinary Science, 7, p. 55. Balance Sheets. Available at: doi: 10.3389/fvets.2020.00055. http://www.fao.org/faostat/en/#data/ Gemmill-Herren, B., Baker, L. E. FBS (Accessed: 16 June 2021). and Daniels, P. A. (2021) True Cost Fiorella, K. J. et al. (2019) ‘A review Accounting for Food: Balancing the of transactional sex for natural Scale. Available at: resources: Under-researched, https://www.routledge.com/True- overstated, or unique to fishing Cost-Accounting-for-Food-Balancing- economies?’, Global Public Health, the-Scale/Gemmill-Herren-Baker- 14(12), pp. 1803–1814. doi: Daniels/p/book/9780367506858 10.1080/17441692.2019.1625941. (Accessed: 21 June 2021). Frehner, A. et al. (2021) ‘How food Gerhardt, C. et al. (2020) ‘How will choices link sociodemographic and cultured meat and meat alternatives lifestyle factors with sustainability disrupt the agricultural and food impacts’, Journal of Cleaner industry?’, Industrial Biotechnology, Production, 300, p. 126896. doi: 16(5), pp. 262–270. doi: 10.1016/j.jclepro.2021.126896. 10.1089/ind.2020.29227.cge. Galloway, J. N. et al. (2007) Gilbert, M. et al. (2015) ‘Income ‘International Pork Chop Trade in disparities and the global distribution Meat: The Tip of the’, Ambio, 36(8), pp. of intensively farmed chicken and pigs’, 622–629. doi: 10.1579/0044- PLoS ONE, 10(7), pp. 1–14. doi: 7447(2007)36[622:ITIMTT]2.0.CO;2. 10.1371/journal.pone.0133381. Garnett, T. et al. (2013) Godde, C. M. et al. (2018) ‘Grazing ‘Sustainable Intensification in systems expansion and intensification: Agriculture: Premises and Policies’, Drivers, dynamics, and trade-offs’, Science Magazine, 341(July), pp. 33– Global Food Security, 16(November), 34. doi: 10.1126/science.1234485. pp. 93–105. doi: Garnett, T. et al. (2015) ‘Policies 10.1016/j.gfs.2017.11.003. and actions to shift eating patterns: Godfray, H. C. J. et al. (2018) ‘Meat What works?’, FCRN/Chatham House, consumption, health, and the p. 85. Available at: environment’, Science, 361(6399). doi: http://www.fcrn.org.uk/sites/default/f 10.1126/science.aam5324. iles/fcrn_chatham_house_0.pdf. Griscom, B. W. et al. (2017) GBD 2017 Mortality Collaborators, ‘Natural climate solutions’, D. et al. (2018) ‘Global, regional, and Proceedings of the National Academy national age-sex-specific mortality and of Sciences, Early edit, pp. 1–6. doi: life expectancy, 1950-2017: a 10.1073/pnas.1710465114. systematic analysis for the Global Hallström, E., Carlsson-Kanyama, Burden of Disease Study 2017.’, Lancet A. and Börjesson, P. (2015)

42 ‘Environmental impact of dietary States of America, 110(52), pp. 20888– change: A systematic review’, Journal 93. doi: 10.1073/pnas.1308149110. of Cleaner Production, 91, pp. 1–11. Herrero, M., Grace, D., et al. (2013) doi: 10.1016/j.jclepro.2014.12.008. ‘The roles of livestock in developing Havlík, P. et al. (2014) ‘Climate countries’, in Animal. Elsevier, pp. 3– change mitigation through livestock 18. doi: 10.1017/S1751731112001954. system transitions’, Proceedings of the Herrero, M. et al. (2014) ‘Exploring National Academy of Sciences of the future changes in smallholder farming United States of America, 111(10), pp. systems by linking socio-economic 3709–14. doi: scenarios with regional and household 10.1073/pnas.1308044111. models’, Global Environmental Headey, D. D. and Alderman, H. H. Change, 24(1), pp. 165–182. doi: (2019) ‘The Relative Caloric Prices of 10.1016/j.gloenvcha.2013.12.008. Healthy and Unhealthy Foods Differ Herrero, M. et al. (2016) Systematically across Income Levels ‘Greenhouse gas mitigation potentials and Continents’, Journal of Nutrition, in the livestock sector’, Nature Climate 149(11), pp. 2020–2033. doi: Change, 6(5), pp. 452–461. doi: 10.1093/jn/nxz158. 10.1038/nclimate2925. Headey, D., Hirvonen, K. and Herrero, M., Thornton, Philip K, et Hoddinott, J. (2018) ‘Animal Sourced al. (2017) ‘Farming and the geography Foods and Child Stunting’, American of nutrient production for human use: Journal of Agricultural Economics, a transdisciplinary analysis’, Lancet 100(5), pp. 1302–1319. doi: Planet Health, 1(1), pp. e33–e42. doi: 10.1093/ajae/aay053. 10.1016/S2542-5196(17)30007-4. Henchion, M. et al. (2017) ‘Future Herrero, M., Thornton, Philip K., et protein supply and demand: Strategies al. (2017) ‘Farming and the geography and factors influencing a sustainable of nutrient production for human use: equilibrium’, Foods, 6(7), pp. 1–21. doi: a transdisciplinary analysis’, The Lancet 10.3390/foods6070053. Planetary Health, 1(1), pp. e33–e42. Herrero, M. et al. (2009) doi: 10.1016/S2542-5196(17)30007-4. ‘Livestock, livelihoods and the Herrero, M. et al. (2020) environment: understanding the trade- ‘Innovation can accelerate the offs’, Current Opinion in Environmental transition towards a sustainable food Sustainability, 1(2), pp. 111–120. doi: system’, Nature Food, 1(5), pp. 266– 10.1016/j.cosust.2009.10.003. 272. doi: 10.1038/s43016-020-0074-1. Herrero, M. et al. (2010) ‘Smart Herrero, M. et al. (2021) Investments in Sustainable Food ‘Articulating the effect of food systems Production: Revisiting Mixed Crop- innovation on the Sustainable Livestock Systems’, Science, 327(5967), Development Goals’, The Lancet pp. 822–825. doi: Planetary Health, 5(1), pp. e50–e62. 10.1126/science.1183725. doi: 10.1016/S2542-5196(20)30277-1. Herrero, M., Havlík, P., et al. (2013) Hristov, A. N. et al. (2013) ‘Biomass use, production, feed Mitigation of greenhouse gas efficiencies, and greenhouse gas emissions in livestock production. A emissions from global livestock review of technical options for non-CO2 systems.’, Proceedings of the National emissions, FAO Animal Production and Academy of Sciences of the United Health Paper. Edited by P. J. Gerber, B.

43 Henderson, and H. P. S. Makkar. Rome. Kemper, B. (2018) ‘Nut Milks Are doi: 10.1016/j.agee.2005.08.009. Milk, Says Almost Every Culture Across Hu, F. B. (2013) ‘Resolved: There is the Globe’, Smithsonian Magazine. sufficient scientific evidence that Available at: decreasing sugar-sweetened beverage https://www.smithsonianmag.com/his consumption will reduce the tory/nut-milks-are-milk-says-almost- prevalence of obesity and obesity- every-culture-across-globe- related diseases’, Obesity Reviews, 180970008/ (Accessed: 21 June 2021). 14(8), pp. 606–619. doi: Leclère, D. et al. (2018) ‘Towards 10.1111/obr.12040. pathways bending the curve of IARC (2015) IARC Monographs terrestrial biodiversity trends within evaluate consumption of red meat and the 21 st century’, Preprint. doi: processed meat. Available at: 10.22022/ESM/04-2018.15241. http://www.iarc.fr/en/media- Lennox, G. D. et al. (2018) ‘Second centre/iarcnews/pdf/Monographs- rate or a second chance? Assessing Q&A_Vol114.pdf (Accessed: 18 June biomass and biodiversity recovery in 2021). regenerating Amazonian forests’, IFCN (no date) IFCN - Dairy Data · Global Change Biology, 24(12), pp. Knowledge · Inspiration - IFCN Dairy 5680–5694. doi: 10.1111/gcb.14443. Research Network. Available at: Leroy, A. et al. (2020) Seafood https://ifcndairy.org/ (Accessed: 18 Sustainability Stability and Security: June 2021). How the sustainability of fisheries IMF (2017) IMF Primary management impacts the stability and Commodity Prices. Available at: security of countries, and what we must https://www.imf.org/en/Research/co do about it. Brussels. doi: mmodity-prices (Accessed: 18 June 10.1371/journal.pone.0004570. 2021). Marlatt, K. L. et al. (2016) Jackson, J. B. C. et al. (2001) ‘Breakfast and fast food consumption ‘Historical overfishing and the recent are associated with selected collapse of coastal ecosystems’, biomarkers in adolescents’, Preventive Science, 293(5530), pp. 629–637. doi: Medicine Reports, 3, pp. 49–52. doi: 10.1126/science.1059199. 10.1016/j.pmedr.2015.11.014. Johnston, B. C. et al. (2019) Mason-D’Croz, D. et al. (2020) ‘Unprocessed red meat and processed ‘Modelling the global economic meat consumption: Dietary guideline consequences of a major African swine recommendations from the nutritional fever outbreak in China’, Nature Food, recommendations (NUTRIRECS) 1(4), pp. 221–228. doi: consortium’, Annals of Internal 10.1038/s43016-020-0057-2. Medicine, 171(10), pp. 756–764. doi: Mattick, C. S. et al. (2015) 10.7326/M19-1621. ‘Anticipatory Life Cycle Analysis of In Keerie, M. (2021) Record $3.1 Vitro Biomass Cultivation for Cultured billion invested in alt proteins in 2020 Meat Production in the United States’, signals growing market momentum, Environmental Science and The Good Food Institute. Available at: Technology, 49(19), pp. 11941–11949. https://gfi.org/blog/2020-state-of-the- doi: 10.1021/acs.est.5b01614. industry-highlights/ (Accessed: 21 June Mbow, C. et al. (2019) ‘Food 2021). Security’, in IPCC Special Report on

44 Climate Change and Land. IPCC, pp. 1– 10.1126/science.aaf2201. 200. Olff, H. and Ritchie, M. E. (1998) McDermott, J. J. et al. (2010) ‘Effects of herbivores on grassland ‘Sustaining intensification of plant diversity’, Trends in Ecology and smallholder livestock systems in the Evolution, 13(7), pp. 261–265. doi: tropics’, Livestock Science, 130(1–3), 10.1016/S0169-5347(98)01364-0. pp. 95–109. doi: Pan, A. et al. (2011) ‘Red meat 10.1016/j.livsci.2010.02.014. consumption and risk of type 2 Mills, D. J. et al. (2011) ‘Under- diabetes: 3 Cohorts of US adults and an reported and undervalued: Small-scale updated meta-analysis’, American fisheries in the developing world’, in Journal of Clinical Nutrition, 94(4), pp. Small-Scale Fisheries Management: 1088–1096. doi: Frameworks and Approaches for the 10.3945/ajcn.111.018978. Developing World. CABI Publishing, pp. Parodi, A. et al. (2018) ‘The 1–15. doi: potential of future foods for 10.1079/9781845936075.0001. sustainable and healthy diets’, Nature Mozaffarian, D. (2016) ‘Dietary Sustainability, 1(12), pp. 782–789. doi: and Policy Priorities for Cardiovascular 10.1038/s41893-018-0189-7. Disease, Diabetes, and Obesity’, Phu, D. H. et al. (2021) ‘Reducing Circulation, 133(2), pp. 187–225. doi: Antimicrobial Usage in Small-Scale 10.1161/CIRCULATIONAHA.115.01858 Chicken Farms in Vietnam: A 3-Year 5. Intervention Study’, Frontiers in Murphy, S. P. and Allen, L. H. Veterinary Science, 7, p. 1244. doi: (2003) ‘Nutritional Importance of 10.3389/fvets.2020.612993. Animal Source Foods’, in Journal of Pikaar, I. et al. (2018) ‘Decoupling Nutrition. American Institute of Livestock from Land Use through Nutrition, pp. 3932–3935. doi: Industrial Feed Production Pathways’, 10.1093/jn/133.11.3932s. Environmental Science and Neufeld, L. M., Hendriks, S. and Technology, 52(13), pp. 7351–7359. Hugas, M. (2021) The Scientific Group doi: 10.1021/acs.est.8b00216. for the UN Food Systems Summit Poore, J. and Nemecek, T. (2018) Healthy diet: A definition for the United ‘Reducing food’s environmental Nations Food Systems Summit 2021. impacts through producers and Available at: https://sc-fss2021.org/ consumers’, Science, 360(6392), pp. (Accessed: 18 June 2021). 987–992. doi: Neumann, C., Harris, D. M. and 10.1126/science.aaq0216. Rogers, L. M. (2002) ‘Contribution of Poux, X. and Aubert, P. (2018) An animal source foods in improving diet agroecological Europe in 2050: quality and function in children in the multifunctional agriculture for healthy developing world’, Nutrition Research, eating Findings from the Ten Years For 22(1–2), pp. 193–220. doi: Agroecology (TYFA) modelling exercise. 10.1016/S0271-5317(01)00374-8. Paris. Available at: www.iddri.org Newbold, T. et al. (2016) ‘Has land (Accessed: 18 June 2021). use pushed terrestrial biodiversity Qian, F. et al. (2020) ‘Red and beyond the planetary boundary? A processed meats and health risks: How global assessment’, Science, 353(6296), strong is the evidence?’, Diabetes Care, pp. 291–288. doi: 43(2), pp. 265–271. doi:

45 10.2337/dci19-0063. Public Health and Food Systems Raasch, S. et al. (2020) Perspective’, Frontiers in Sustainable ‘Effectiveness of alternative measures Food Systems, 4, p. 134. doi: to reduce antimicrobial usage in pig 10.3389/fsufs.2020.00134. production in four European Searchinger, T. D. et al. (2015) countries’, Porcine Health ‘High carbon and biodiversity costd Management, 6(1), pp. 1–12. doi: from converting Africa’s wet savannahs 10.1186/s40813-020-0145-6. to cropland’, Nature Climate Change, 5, Ramankutty, N. et al. (2018) pp. 481–486. doi: ‘Trends in Global Agricultural Land Use: 10.1038/NCLIMATE2584. Implications for Environmental Health Semba, R. D. et al. (2016) ‘Child and Food Security’, Annual Review of Stunting is Associated with Low Plant Biology, 69(1), pp. 789–815. doi: Circulating Essential Amino Acids’, 10.1146/annurev-arplant-042817- EBioMedicine, 6, pp. 246–252. doi: 040256. 10.1016/j.ebiom.2016.02.030. Ravuvu, A. et al. (2018) ‘Protocol Shapiro, M. J. et al. (2019) ‘A to monitor trade agreement food- Systematic Review Investigating the related aspects: The Fiji case study’, Relation between Animal-Source Food Health Promotion International, 33(5), Consumption and Stunting in Children pp. 887–900. doi: Aged 6-60 Months in Low and Middle- 10.1093/heapro/dax020. Income Countries’, Advances in Robinson, T. P. et al. (2011) Global Nutrition, 10(5), pp. 827–847. doi: livestock production systems, Food and 10.1093/advances/nmz018. Agriculture Organization of the United Sheikh (2019) ‘Got Impossible Nations (FAO) and International Milk? The Quest for Lab-Made Dairy’, Livestock Research Institute (ILRI),. The New York Times, 2 August. Roe, S. et al. (2019) ‘Contribution Available at: of the land sector to a 1.5 °C world’, https://www.nytimes.com/2019/08/0 Nature Climate Change, 9(11), pp. 817– 2/science/lab-grown-milk.html 828. doi: 10.1038/s41558-019-0591-9. (Accessed: 21 June 2021). Rubio, N. R., Xiang, N. and Kaplan, Shurtleff, W. and Aoyagi, A. (2014) D. L. (2020) ‘Plant-based and cell-based History of Meat Alternatives (965 CE to approaches to meat production’, 2014): Extensively Annotated Nature Communications, 11(1), pp. 1– Bibliography and Sourcebook. Soyinfo 11. doi: 10.1038/s41467-020-20061-y. Center. Available at: Saget, C., Vogt-Schilb, A. and Luu, https://www.soyinfocenter.com/pdf/1 T. (2020) Jobs in a Net-Zero Emissions 79/MAL.pdf (Accessed: 21 June 2021). Future in Latin America and the Smith, A. (2014) ‘The History of the Caribbean. Washington D.C. and Veggie Burger’, Smithsonian Magazine. Geneva. Available at: Available at: https://www.ilo.org/wcmsp5/groups/ https://www.smithsonianmag.com/art public/---americas/---ro- s-culture/history-veggie-burger- lima/documents/publication/wcms_7 180950163/ (Accessed: 21 June 2021). 52069.pdf (Accessed: 18 June 2021). Smith, P. et al. (2007) ‘Policy and Santo, R. E. et al. (2020) technological constraints to ‘Considering Plant-Based Meat implementation of greenhouse gas Substitutes and Cell-Based Meats: A mitigation options in agriculture’,

46 Agriculture, Ecosystems and Reversal of an import ban on Turkey Environment, 118(1–4), pp. 6–28. doi: tails’, Bulletin of the World Health 10.1016/j.agee.2006.06.006. Organization, 95(10), pp. 723–725. doi: Smith, P. et al. (2014) ‘Agriculture, 10.2471/BLT.17.192468. Forestry and Other Land Use’, in Tilman, D. and Clark, M. (2014) Intergovernmental Panel on Climate ‘Global diets link environmental Change Working Group (ed.) Change sustainability and human health’, 2014: Mitigation of Climate Change, p. Nature, 515(7528), pp. 518–522. doi: 94. doi: 10.1016/j.phrs.2011.03.002. 10.1038/nature13959. Springmann, M. et al. (2016) Troell, M., Jonell, M. and Crona, B. ‘Analysis and valuation of the health (2019) The role of seafood in and climate change cobenefits of sustainable and healthy diets The EAT- dietary change’, Proceedings of the Lancet Commission report through a National Academy of Sciences of the blue lens. Stockholm. United States of America, 113(15), pp. Tubiello, F. N. et al. (2021) ‘Carbon 4146–4151. doi: emissions and removals from forests: 10.1073/pnas.1523119113. New estimates, 1990-2020’, Earth Springmann, M. et al. (2018) System Science Data, 13(4), pp. 1681– ‘Options for keeping the food system 1691. doi: 10.5194/essd-13-1681- within environmental limits’, Nature, 2021. 562(7728), pp. 519–525. doi: Vanhercke, T. et al. (2017) ‘Step 10.1038/s41586-018-0594-0. changes in leaf oil accumulation via Stehfest, E. et al. (2009) ‘Climate iterative metabolic engineering’, benefits of changing diet’, Climatic Metabolic Engineering, 39, pp. 237– Change, 95(1–2), pp. 83–102. doi: 246. doi: 10.1007/s10584-008-9534-6. 10.1016/j.ymben.2016.12.007. De Sy, V. et al. (2015) ‘Land use de Vries, M. and de Boer, I. J. M. patterns and related carbon losses (2010) ‘Comparing environmental following deforestation in South impacts for livestock products: A America’, Environmental Research review of life cycle assessments’, Letters, 10(12), p. 124004. doi: Livestock Science, 128(1–3), pp. 1–11. 10.1088/1748-9326/10/12/124004. doi: 10.1016/J.LIVSCI.2009.11.007. Thornton, P. and Herrero, M. Waithaka, M. M. et al. (2006) ‘Bio- (2010) Development and Climate economic evaluation of farmers’ Change: The inter-linkages between perceptions of viable farms in western rapid growth in livestock production, Kenya’, Agricultural Systems, 90(1–3), climate change, and the impacts on pp. 243–271. doi: water resources, land use, and 10.1016/j.agsy.2005.12.007. deforestation. Nairobi. Walker, P. et al. (2005) ‘Public Thornton, P. K. (2010) ‘Livestock health implications of meat production production: Recent trends, future and consumption’, Public Health prospects’, Philosophical Transactions Nutrition, 8(4), pp. 348–356. doi: of the Royal Society B: Biological 10.1079/phn2005727. Sciences. Royal Society, pp. 2853–2867. Walsh, B. J. et al. (2015) ‘New feed doi: 10.1098/rstb.2010.0134. sources key to ambitious climate Thow, A. M. et al. (2017) ‘Food targets’, Carbon Balance and supply, nutrition and trade policy: Management, 10, pp. 1–8. doi:

47 10.1186/s13021-015-0040-7. and quality of life among eastern North WHO (2015) Cancer: Carolina residents’, Environmental Carcinogenicity of the consumption of Health Perspectives, 108(3), pp. 233– red meat and processed meat. 238. doi: 10.1289/ehp.00108233. Available at: World Bank (2016) ‘List of https://www.who.int/news-room/q-a- economies’. Available at: detail/cancer-carcinogenicity-of-the- http://databank.worldbank.org/data/ consumption-of-red-meat-and- download/site-content/CLASS.xls. processed-meat (Accessed: 1 July Young, V. and Pellett, P. (1994) 2021). ‘Plant protein in relation to human Wieland, B. et al. (2019) AMUSE protein and amino acid nutrition’, Livestock, version 2-Antimicrobial use American Journal of Clinical Nutrition, in livestock production A tool to 59, pp. 1203S-1212S. Available at: harmonise data collection on http://pilarmartinescudero.es/Ene201 knowledge, attitude and practices. 8/Plan protein in relation to human Available at: protein .pdf (Accessed: 21 June 2021). https://cgspace.cgiar.org/bitstream/h Zaharia, S. et al. (2021) ‘Sustained andle/10568/107443/AMUSE_livestoc intake of animal-sourced foods is k_v2.pdf?sequence=2&isAllowed=y associated with less stunting in young (Accessed: 21 June 2021). children’, Nature Food, 2(4), pp. 246– Willett, W. et al. (2019) ‘Food in 254. doi: 10.1038/s43016-021-00259- the Anthropocene: the EAT–Lancet z. Commission on healthy diets from Van Zanten, H. H. E. et al. (2018) sustainable food systems’, The Lancet, ‘Defining a land boundary for 393, pp. 447–429. doi: 10.1016/S0140- sustainable livestock consumption’, 6736(18)31788-4. Global Change Biology, 24(9), pp. Wing, S. and Wolf, S. (2000) 4185–4194. doi: 10.1111/gcb.14321. ‘Intensive livestock operations, health,

48 Supplementary information

Figure S1: Percent change in per capita animal source-food demand 1990-2015. Source: Based on authors’ calculations from FAOSTAT (2018).

Figure S2: Evolution of livestock numbers by region during the period of 1961-2013. Data from FAOSTAT(2018), as presented in Ramankutty et al. (2018).

49

Figure S3: Modelled proportions of extensive, semi-intensive and intensive pig production in different parts of the world in relation to the gross domestic product (GDP, in Purchasing Power Parity (PPP)) (Gilbert et al., 2015).

Figure S4: Trends in land use for cropland and permanent pastures 1700-2013 (Ramankutty et al., 2018).

50

Figure S5: Historical trends in land use, greenhouse gas emissions and nitrogen (N) use intensities of the livestock sector 1961-2010 (Davis et al., 2015).

51 Food Systems Summit Briefs are prepared by researchers of Partners of the Scientific Group for the United Nations Food Systems Summit. They are made available under the responsibility of the authors. The views presented may not be attributed to the Scientific Group or to the partner organisations with which the authors are affiliated.

Authors: Mario Herrero Commonwealth Scientific and Industrial Research Organisation, Australia. Daniel Mason-D’Croz Commonwealth Scientific and Industrial Research Organisation, Australia Philip K. Thornton CGIAR Research Programme on Climate Change Agriculture and Food Security, ILRI, Kenya. Jessica Fanzo Bloomberg School of Public Health and the Nitze School of Advanced International Studies, Johns Hopkins University, USA. Jonathan Rushton Department of Livestock and One Health, Institute of Infection, Veterinary and Ecological Sciences, University of Liverpool, UK Cecile Godde Commonwealth Scientific and Industrial Research Organisation, Australia Alexandra Bellows Bloomberg School of Public Health and the Nitze School of Advanced International Studies, Johns Hopkins University, USA Adrian de Groot Impact Institute, The Netherlands. Jeda Palmer Commonwealth Scientific and Industrial Research Organisation, Australia Jinfeng Chang College of Environmental and Resource Sciences, Zhejiang University, China Hannah van Zanten Farming Systems Ecology group, Wageningen University & Research, the Netherlands. Barbara Wieland International Livestock Research Institute, Addis Ababa, Ethiopia and Institute of Virology and Immunology, University of Bern, Switzerland. Fabrice DeClerck Agricultural Biodiversity and Ecosystem Services, Bioversity International, Rome, Italy and EAT, Oslo, Norway. Stella Nordhagen Global Alliance for Improved Nutrition (GAIN), Geneva, Switzerland. Margaret Gill School of Biological Science, University of Aberdeen, United Kingdom.

For further information about the Scientific Group, visit https://sc-fss2021.org or contact [email protected] @sc_fss2021

52