Impacts of climate change on the livestock food supply chain; a review of the evidence

1Climate change is a major concern for current livestock systems worldwide. Global warming and its associated changes in mean climate variables and climate variability affect feed and water resources as well as animal health and production. Climate change also has implications for the processing, storage, transport, retailing and consumption of livestock products. The ability of current livestock systems to support livelihoods and meet the increasing demand for livestock products is thus threatened.

2The livestock sector currently plays a key role in food supply and food security. Livestock products (meat, milk and eggs) contribute 15% and 31% of global per capita calorie and protein supply, with regional variations (; see Appendix for calculation of estimates presented in the Introduction). About 30% and 6% of global ruminant meat and milk production originates from grazing systems, on land that is often poorly suited for cropping. Furthermore, livestock provides a range of other services, including as a source of draught power, a means of transportation, a source of nutrients for poor soils, a source of income generation and diversification, and a form of financial capital, all of which contribute to the overall well-being and resilience of many communities. Over 844 million people worldwide receive some income from agriculture, and the livestock sector contributes about 40% of agricultural value-added. Livestock contributions to food security and other sustainability dimensions will be affected by climate change, although the full extent and magnitude of the impacts remain unknown.

3Livestock and climate change studies often focus on the climate mitigation potential of livestock and in describing adaptation practices. When studies cover climate impacts, these tend to have a relatively narrow viewpoint, focusing on specific livestock species, primary production, or on selected dimensions of risk of climate-related impacts such as climate hazards without considering vulnerability levels of different communities (e.g. of reviews,;;;; ). In large part this reflects the fact that, compared to crop production, considerably less work has been published on observed and modelled climate impacts on livestock. It also reflects the limited number of synthetic reviews of the issue, as highlighted in.

4In order to fill this gap, we review the risk of climate-related impacts along the land-based livestock food supply chain (i.e. from production to consumption). While not exhaustive, we aim to capture the major trends with direct implications for livestock-sourced food availability, access, utilisation and stability, and highlight key recent literature. We acknowledge that the implications of climate change go well beyond these pillars and affect the provision of goods and services (e.g. wool, hides, skins and manure, animal traction, financial instrument, etc.), human livelihoods and health, ecosystems, economies, cultures, and infrastructure in complex ways. Also, while we recognize that climate adaptation strategies and the impacts of livestock on climate change are significant considerations, these are not covered here but assessed elsewhere (e.g.,,,,,, and ).

5This review is framed around the concept of risk of climate-related impacts, as defined by the Intergovernmental Panel on Climate Change (IPCC) Working Group II. Risk of climate-related impacts results from the interaction of climate-related hazards with the exposure and vulnerability of human and natural systems (Fig. 1). The analysis of this interaction represents the core of the IPCC climate impacts assessments. We use the term hazard to refer to climate-related physical events or trends that impact livestock systems. Exposure refers to the parts of the livestock supply chain that could be adversely affected, while vulnerability encompasses humans’ capacity to cope and adapt to changes. The term impact is used primarily to refer to the effects of extreme weather, climate events and climate change on natural and human systems.

6We first detail the extent to which the livestock supply chain is exposed to climate change, referring to key literature on topics that have received significant past attention and expanding on recent topics of concern. We then discuss the societal ability of the livestock sector to cope or adapt to changes considering broader societal trends before highlighting potential risks of climate-related impacts.

7Anthropogenic greenhouse gas emissions are associated with a range of climate and atmospheric shifts, the impacts of which are already being observed. Major trends that can impact the livestock food supply chain are increases in atmospheric carbon dioxide (eCO2) and tropospheric ozone (O3) concentrations; changes in both mean and variability of temperature and precipitation; sea level rise and storm surges; and increased risk and frequency of extreme weather events. Past and projected changes in these climate variables are detailed in the Supplementary Information.

8These climate change hazards may adversely affect the livestock sector at different stages of the livestock supply chain, as summarised in Fig. 2 and Table 1, and further detailed in the sections below. The potential impacts on labour and prices and overall social dynamics, which affect all stages of the supply chain, are described last.

9Changes to the quantity and quality of livestock feed will be influenced by complex local interactions between eCO2 concentrations, tropospheric O3 levels, temperature, and precipitation. We first provide an inventory of how eCO2, O3, temperature and precipitation can affect livestock feed, then present some model projections under climate change. Livestock consume grains (especially in poultry, pig and intensive ruminant systems), crop above-ground biomass (e.g. in dual purpose crops which are both grazed and harvested), crop residues (e.g., straw or stover – key feed in mixed crop-livestock systems) as well as native and sown pastures (key feed in mixed crop-livestock and grazing-only systems). While not covered here, livestock can also be fed by-products and waste (e.g. oilseed cakes, bran, vegetable waste, brewer waste), concentrates and supplements.

10Research shows that eCO2 may have both positive and negative impacts on livestock feed, although there is recent evidence that the fertilisation effects of eCO2 and nitrogen on plant physiological processes may slow in the future as ecosystems productivity become dominated by the negative effects of higher temperatures and extreme droughts.

11Increases in eCO2 concentrations stimulate plant primary productivity (see review in ), increasing potential yields of some species. Plants with a C3 photosynthetic pathway such as wheat, rice, soybean and temperate grasses experience greater growth stimulation than C4 plants such as maize, sorghum, sugarcane and tropical grasses. However, the CO2 fertilisation effects can also reduce animal feed quality. For example, reported that C3 crops other than legumes had lower grain protein concentrations under elevated eCO2 concentration in the range 546–586 ppm (−6.3% in wheat grains and −7.5% in rice grains). The impact on C4 crop grain was smaller. Increased eCO2 was also found to decrease the overall mineral concentrations (−8%) and increased the total non-structural carbohydrate (mainly starch, sugars) to mineral ratios in the total biomass of non-leguminous C3 plants. While the nutritional quality of C3 grasses may be the most greatly impacted by eCO2 increases, it may nonetheless remain higher than C4 grasses under elevated eCO2. Increased toxicity has also been reported in some plants, with measuring a 160% increase in the concentration of cyanogenic glycosides (compounds that break down to release toxic hydrogen cyanide when plant tissue is crushed or chewed) in cassava leaves between CO2 concentrations of 360 and 710 ppm in greenhouse experiments. Woody encroachment associated with rising eCO2 levels and changes in fire regimes can also alter grassland ecosystem function and negatively impact the intake and quality of grazing animals’ diets. Woody forages are harder for cattle and sheep to physically access as compared to goats, are less palatable, and have lower dry matter and protein digestibility compared to herbaceous plants.

12The effects of increasing tropospheric O3 on plant productivity at scale and the range of potential secondary effects it might have (e.g. on weeds, pests and diseases, interactions with chemicals such as pesticides) have received less attention than eCO2. However, synthesis of crop responses to O3 finds that O3 pollution reduces crop yields to a similar level as nutrient, heat and aridity stress. For instance, using historical ground-level monitoring data, estimated that, over the past 30 years, O3 pollution reduced U.S. soybean and maize yields by 5–10%. provide a review of the literature on the effect of this air pollutant on plant productivity. The topic is not yet fully understood and remains one of the key uncertainties in crop, grassland and other global terrestrial models, with significant implications on our ability to predict future atmospheric composition and global climate, net primary productivity and provision of ecosystem services.

13Changes in temperature and water availability can greatly affect forage and crop yields and feed quality. Sensitivity to changes in climate depends on the crop type and other environmental factors, but there is strong agreement that air temperatures above approximately 30°C–34°C generally depress cereal yields under water-limited conditions, through accelerating crop development and damaging plant cells (; D. B.; ). The maximum temperature for growth of temperate legumes and pastures is around 30–35°C, increasing to 35–50°C for tropical species. High temperatures are often coupled with water stress, since low soil moisture results in a decrease in evaporative cooling from the landscape and high temperatures increase crop water loss. The combination of warmer temperatures and drier conditions tends to favour C4 rather than C3 species. The concentration of ergot alkaloids, and other potentially toxic secondary compounds (e.g., hydrogen cyanide in cassava and forage sorghum) are also likely to increase in response to a hotter and drier climate.

14Increased instability of feed supply is particularly a concern in grazing systems where it represents a major challenge for herd size and grazing intensity management. Pastures with high year-to-year precipitation variability were found to currently support lower livestock stocking rates than less variable regions. Studies focused on grassland vegetation have also found that changes in seasonal climate patterns can have either positive or negative impacts on above ground biomass, depending on the nature of the change and the agro-ecological context. The arrangement of climate extreme sequences such as drought sequences or number of hot days in a row, could have significant implications for the livestock sector. While less commonly researched than droughts, other hazards such as fires, heavy storms, flooding events, surface melt and icing events, as well as the appearance of new lakes, streams and marshes also disturb crop growth, reduce arable land and restrict animal access to pastures. For instance, in northern Russia, nomadic reindeer herders migrate hundreds of kilometres in spring and autumn to connect summer and winter pastures. The appearance of new water bodies and change in size of existing ones due to melting permafrost can act as barriers, changing migration routes and increasing grazing pressure on the most accessible pastures.

15Changes in precipitation patterns in saline areas will also affect soil salinity and agricultural production potentials. Salinity intrusion and associated reductions in forage area have led farmers across the coastal belt in Bangladesh to look for other sources of livestock feed. suggest that the predicted local sea-level rise for areas such as Carey Island in Malaysia would prevent oil palm production by the 21st century, due to seawater intrusion. This has implications for livestock production through potential reductions in the production of palm kernel meal, which is often fed to cattle in industrial systems. Integrated palm-cattle systems, where cattle graze under trees or are fed palm fronds removed as part of plantation maintenance, would also be impacted. Increases in the frequency, duration and intensity of heavy rainfall events, drought periods and sea level rise will also increase exposure of water, croplands and grasslands to soil contaminants with potential harmful impacts for crop and forage yield quantity and quality.

16At higher levels of warming, crop yields are projected to drop, especially at lower latitudes. This is particularly the case for maize and wheat yields which begin to decline with 1 °C–2 °C of local warming in the tropics, and drop by up to 60% under 5 °C of local warming. Temperate maize is less clearly affected at the 1–⁠2 °C threshold, but would be significantly affected with warming of 3 °C–5 °C. Recent studies also show that global food production has likely already been impacted. estimated that the impact of observed climate change on yields of different crops ranged from −13.4% (oil palm) to +3.5% (soybean), with impacts mostly negative in Europe, Southern Africa and Australia but generally positive in Latin America. Crop yield interannual variability is likely to progressively increase in many regions. For instance,, in a gridded modelling study reported an increase of interannual variability of more than 5% in 64% of grid cells, and a decrease of more than 5% in 29% of cases by 2050.

17Regarding forage availability, as for food crops, the diversity and severity of likely impacts differ considerably by location and species. In an assessment of global rangelands, found that global mean herbaceous biomass is projected to decrease of 4.7% by 2050 under RCP 8.5, with 74% of global rangeland area projected to experience a decline in mean biomass. The largest regional decrease was projected for Oceania while the highest increase was found for Europe. Another study focussed on European grasslands also found projected increases in grassland productivity, mainly attributed to the simulated fertilisation effect of rising CO2. Both studies highlight projected increases in biomass inter-annual variability over some regions. Woody encroachment is also projected to occur on over 51% of global rangeland area by 2050 under RCP 8.5 according to.

18In an integrated partial equilibrium modelling approach, found that the climate change impacts on crop and pasture forage yields will have little effect on global milk and meat production by 2050 due to trade in animal products, which could compensate for the feed deficits in some parts of the world. However, depending on the scenario, the impacts could be more pronounced at the regional scale. The most uncertain and potentially the most severe effects were found for sub-Saharan Africa, where for example, ruminant meat production could increase by 20% by 2050 but could also decrease by 17%, depending on projected feed supply based on varying the biophysical crop model and CO2 fertilisation assumptions.

19Yield projections as above-mentioned are subject to large uncertainties. Models do not usually consider extreme events and overall increasing variability, or explicitly represent adaptation or effects such as changes in tropospheric O3, pests, pollinators, or agricultural labour. There are also large uncertainties as to climate extremes and trends in the future and changes in management practices, historical and projected land-use patterns. Our understanding of ecosystems responses to climate change is also limited, including the interacting consequences of changes in temperature, precipitation, eCO2 and tropospheric O3, particularly in the context where management-driven yields increases are still occurring across vast areas of croplands. Possible ecosystems transitions from equilibrium to non-equilibrium systems driven primarily by stochastic abiotic factors will likely result in highly variable and less predictable primary production. Land uses such as grazing can also regulate grasslands responses to climate change. For example, sheep grazing has been found to limit CO2 stimulation of grassland productivity by selectively consuming legumes and forbs, plants with the greatest growth responses to CO2. The issue of uncertainty is even more significant for grass yield projections than for crops, as reference data are less available for models’ development and evaluation.

20The effect of climate hazards on pests (e.g. insect pests, pathogens), weed outbreaks and pollinators can have significant consequences for animal feed availability as reviewed in.

21Pests, pathogens and weeds are estimated to currently reduce the production of major crops by 25–40%. Increases in temperature increase winter survival of insect pests and rates of herbivory, and alter the spatial distribution of pests and pathogens. reported an average poleward shift of pest and pathogen distribution of 2.7 km per year since 1960, though there is substantial variation among taxonomic groups. Under climate change, the spatial mismatches between pests and natural predators may be exacerbated in some regions, weakening biological control systems. In some instances, weather extremes can weaken crop defences and create niches for pests and weed outbreaks. In other cases, extreme events can reduce pests and weeds, and as such support crop establishment and growth. Recent intense desert locust outbreaks across East Africa, Asia and the Middle East have been linked to a series of cyclones causing warm and wet conditions. In Ethiopia, as per April 2020, nearly 200,000 ha of cropland were damaged by the insects, leading to the loss of over 356,000 tons of grain and thousands of tons of crop residues, a key livestock feed in the country. An additional 1.3 million hectares of pasture were affected, reducing pastoral areas by as much as 61% in the Somali region. Increases in eCO2 concentrations may also influence the weed composition and crop defences in complex ways, including reducing the effectiveness of herbicides. Shifting pest and disease patterns may increase the use of pesticides, some of which (i.e. dioxins) can pass on to animal products. These toxins can remain in soils for extended periods, and can contaminate animal feeds and water sources, particularly in conditions with alternating periods of drought and floods that are more likely with climate change.

22More unstable weather, including more humid and cloudier conditions will lead to more on-farm post-harvest losses of animal feed, especially in developing countries with hot climates where most smallholders rely on the sun to dry their crops and forages before storage. Contamination by toxins will likely also be higher. For instance, maize and sorghum can become contaminated by aflatoxins, particularly in drought conditions. While concentrations at harvest are usually not poisonous, the storage of grains under damp or poorly aerated conditions can lead to mould and the poisoning of animals and consumers. Increases in pest infestation frequency or intensity under climate change will also result in higher crop losses where storage facilities are inadequate.

23While all grasses and most staple food grains such as maize, wheat, rice and sorghum are wind or self-pollinated, some crops used as livestock feed in industrial and mixed crop-livestock systems are animal pollinator-dependent to varying levels (e.g. soybean, cowpeas, pigeon peas, broad beans, rapeseed, oilseed rape, oil palm and some vegetable and fruit crops). Climate change impacts on pollinators include changes in the abundance and distribution of both flowering plants and pollinators, and the timing of flowering and pollinator emergence and migration, causing a mismatch in pollinator availability and crops to be pollinated. This contributes to reductions in the breadth and nutritional value of feed for pollinators, which in turn decreases pollinator abundance. Increases in eCO2 concentrations also affect the nutritional value of key forages for pollinators. While the net effect of climate change on pollinators remains uncertain, studies indicate that a reduction in animal pollination would decrease yields of numerous pollinator-dependent food crops.

24Water is used at various stages of the livestock supply chain: to grow feed; for animal consumption and cooling; to produce electricity, fertilizers, pesticides and fuel; and to clean animals and infrastructure (see detailed inventory in ). Most water in the livestock value chain is used for feed production, accounting for over 90% of water consumption in many systems. The amount of water required for livestock consumption varies with local climate conditions, with higher consumption under hot conditions. For instance, once air temperatures exceed 30 °C, the expected poultry drinking water intake can increase by 50% above normal rates. In arid and semi-arid regions, the increased frequency, intensity and duration of droughts are a significant challenge for sustaining water supply to livestock and feed production.

25Climate change is also projected to reduce raw water quality, which can decrease animal water intake, feed intake and health. Poor local water quality can be caused by warmer temperatures, sea level rise, or higher sediment, nutrient and pollutant loadings due to heavy rainfall. Reduced dilution of pollutants during droughts and disruptions of treatment facilities during floods are also a concern. High salt concentrations in water and feed depress livestock feed intake and production. The tolerance of animals to different levels of total dissolved solids (including inorganic salts) in water varies considerably among varied species with poultry, buffaloes and dairy cattle having lower tolerance levels (desirable maximum concentration for healthy growth of 2000–2500 ppm) than beef cattle, sheep and pigs (4000 ppm, ).

26The effects of climate change on resources, as mentioned in this section, will continue to lead to shifts in the global agricultural area as well as changes in seasonality and crop and livestock suitability. For example, estimate a northward shift of the feasible agriculture zone by up to 1200 km by the end of the century, although many of these areas are associated with highly variable water balances. The world's drylands have expanded over the last 60 years, mostly in semiarid regions, and are projected to expand further during this century. In sub-Saharan Africa, more than 20% of the mixed crop-livestock system in arid and semi-arid regions is projected to become unviable for crop agriculture by mid-century. Grass yields may overall be less impacted by climate change than crop yields, which may favour grazing systems and potentially change the current trend towards more intensive systems.

27The impacts of climate change on animal growth, production and welfare include impacts mediated by reduced feed intake, direct physiological and metabolic effects, and changes in behaviour (e.g.,,,,,,,,, ). The potential impacts of climate change on livestock health and production are already well-reviewed in the literature, and we highlight key effects below.

28Heat stress is primarily caused by exposure to high ambient temperatures and relative humidity, which limit the capacity of livestock to shed heat to their environment, on farm or during transportation (see section 2.3.1). Vulnerability to heat stress varies according to species, breed, life stage, genetic potential, nutritional status, size, level of insulation (hide thickness or distribution of feathers) and previous exposure of the animal, with animals with high energy demands (i.e. high-yielding individuals and breeds) most susceptible. For example, dairy cows are more susceptible than beef cattle, and temperate Bos taurus breeds are more susceptible than tropically adapted Bos indicus cattle and their crosses.

29In most cases, the impact of heat stress is reduced productivity and animal welfare. However, under severe or prolonged conditions, mortalities will also occur. Reduced feed intake is one of the first and biggest consequences of heat stress, leading to declines in growth rates and production of milk or eggs.

30Heat stress also affects livestock production through changes to fertility and susceptibility to disease. In both mammals and poultry, the impacts on fertility are caused by reduced ovarian function, reduced motility of spermatozoa, and inhibition of embryonic development. Cattle also show reduced expression of estrus behaviour, further reducing the chances of reproductive success. In addition, heat stress may depress immune function and the effectiveness of some vaccines, increasing the incidence of livestock disease. The reverse is also true, with animals showing clinical signs of disease more susceptible to heat stress. While covered to a limited extent in the literature, there is some evidence that high temperatures may decrease the ability of mammalian herbivores to detoxify plant secondary compounds. While most work to date has been done with rodents and wildlife (e.g. ), high environmental temperatures have been shown to increase the susceptibility of cattle to poisoning from ergot alkaloids in tall fescue.

31The impacts of heat stress on livestock can be both immediate and long-lasting, and can also affect offspring exposed to heat stress in utero. Research in dairy cattle and pigs has shown that heat stress in utero reduces milk yield at first lactation, and alters nutrient partitioning and carcass composition. Animals exposed to heat stress in utero may also be better adapted to heat stress conditions at maturity, with reporting that cows exposed to heat stress in utero are better able to regulate core body temperature.

32Heat stress can also impact the quality of animal products, reducing the size of eggs and thickness of eggshells, decreasing the fat and protein content of milk, and changing the colour and water-holding capacity of both red and white meat. These changes can make livestock products less appealing to customers, increase wastage, and reduce the price that producers receive for their products. Selection of breeds that are better adapted to high temperatures can also have implications for product quality; while Bos indicus cattle are better adapted to high temperatures and humidity than Bos taurus, their meat is less tender, and tends to score lower in meat quality assurance schemes such as the Meat Standards Australia index, receiving lower prices in some markets.

33Heat stress, and other stressors such as feed withdrawal can also cause the dissemination of enteric pathogens such as Salmonella, Escherichia coli O157:H7, and Campylobacter from livestock into human food, and as such is also a major health concern. Indeed, these external stressors can increase animal pathogen carriage and shedding as reviewed in. The underlying mechanisms have however not yet been fully explained.

34In addition to heat stress, the increased frequency and intensity of storms, fires, floods and cold waves can have significant repercussions for the livestock sector. Extreme flooding followed by cold weather in northern Australia in early 2019 caused the loss of approximately half a million livestock (mostly cattle, but also sheep, goats and horses). Similarly, in October 2018, Hurricane Michael destroyed an estimated 84 chicken houses and killed over 2 million chickens in Georgia, U.S. In geographic areas with cold winters, such as in the Northeast US, warmer temperatures may also reduce animal cold stress and maintenance energy requirements, as well as housing heating.

35Climate change may also impact infectious livestock diseases by changing their spatial distributions, affecting annual and seasonal cycles, altering disease incidence and severity, and modifying susceptibility of livestock to illness. Many infectious pathogens that cause disease in livestock are sensitive to changes in climate, primarily moisture, rainfall, temperature and particulate matter – many of these diseases are zoonotic. Some pathogen transmission routes are also more climate-sensitive than others. Vector-borne, foodborne, water-borne and soil-borne pathogens are the most likely to be affected by climate change, while those transmitted directly or by fomite are the least likely to be affected by climate change. In particular, the quantity and spread of insect vectors such as flies, ticks and mosquitoes, but also wild birds, rodents and mammals which can transmit disease to farmed poultry, pigs and ruminants is of concern to the livestock sector.

36It is also probable that increases in tropospheric O3 will affect animal health, though increases in mortality are likely smaller than those reported for heat stress events.

37Livestock are often transported long distances by road or sea to markets and slaughter. Climate change, particularly increases in temperature and climate-driven disruptions in the transport network, may increase the risk of heat stress during transport, which in turn, can contribute to poor animal welfare and deaths. This aspect is not often considered in heat stress literature reviews. High stocking density of animals during transport increases ambient temperature and humidity, and reduces ventilation, decreasing opportunities to shed heat loads. In addition, animals transported by truck usually do not have access to water on-board and can become dehydrated faster during hot conditions. In response to this, the live export of sheep from Western Australia to the northern hemisphere is currently prohibited in summer (June to September). Increased scrutiny of animal welfare in other regions may result in additional restrictions on animal transport such as when animals can be moved (time of day and year), and the duration of transport.

38Higher temperatures, increased humidity, increased frequency of extreme weather events, and rising sea levels put additional stress on built-up capital (machinery, transportation infrastructure, electricity networks, telecommunications, etc.). This will lower productivity across many economic sectors, increase construction, operating and maintenance costs, and may shorten the lifespan of critical industry infrastructures.

39Damage and degradation of key transportation infrastructure such as roads, railways, and port infrastructure are of particular concern for the livestock sector. For example, in Australia, 95% of livestock are transported domestically by road from farms to central points such as saleyards, feedlots, abattoirs or ports of embarkation for live export. In early 2019, cattle producers in northern Australia reported over 29,000 km of farm roads destroyed by extreme flooding, in addition to the loss of tools, machinery, homes and 22,000 km of fencing. At a sea level rise of 1.1 m (high end scenario for 2100), the Australian Government projects that coastal assets at risk from the combined impact of inundation and shoreline recession are greater than AU$226 billion and include between 27,000 and 35,000 km of roads and rail, with a value of between AU$51 and AU$67 billion. In the EU, almost 90% of external trade is transported by sea and studies found that 64% of all European seaports are expected to be affected by inundation events by 2060, caused by global mean sea level increases and combined effects of tides, local waves, and storm surges. Reduced polar ice may offer new trade routes that could reduce travel time between Europe and the Northern Pacific, and could reduce the reliance on several key trade passages (e.g. Panama Canal, Suez Canal, and Strait of Malacca) but could also threaten new marine ecosystems through the introduction of invasive species.

40Increasing temperature results in worsening conditions for storage, which leads to degradation of food quality and shelf-life, increased wastage and increases the likelihood of the proliferation of microbes and fungi (i.e. E. coli, salmonella, aflatoxin and mycotoxin producing organisms, etc.), especially under humid conditions. Even in developed economies, increased temperatures have been found to increase the likelihood of food poisoning, suggesting that increasing temperatures will lead to reduced food quality and safety. For instance, chilled (below 10 °C) storage life is halved for each 2–3 °C rise in temperature. In addition, where climate change results in higher levels of microorganisms on meats and produce prior to processing, storage temperatures may need to be lower to preserve required shelf-lives.

41Climate control (air conditioning and refrigeration) can mitigate some of the negative impacts of climate change on food processing and distribution. However, this leads to increased energy usage and costs, and current levels of refrigeration are already estimated to contribute 1% of global CO2 emissions. found that an increased ambient temperature from 17 to 25 °C resulted in an 11% increase in average power consumed in a catering establishment. As a food moves along the cold chain it becomes increasingly difficult to control and maintain its temperature. This is because the temperatures of bulk packs of refrigerated product in large storerooms are far less sensitive to small heat inputs than single consumer packs in open display cases or in a domestic refrigerator or freezer. Fresh milk, in particular, requires significant amounts of cooling and could be the most impacted by increasing costs, or deterioration of cold chains.

42Globally, the use of refrigeration is also far below what is optimal, with only 10–20% of perishable value chains refrigerated. Low- and middle-income (LMIC) countries in already hot environments are particularly vulnerable, due to limited cold chain development. These are also the regions that have seen the largest increases in consumption of livestock products in recent years. It is estimated that if LMIC acquired the same level of refrigerated equipment as that in high income countries, wastage of perishable food would decrease by over 200 million tonnes, or 14% of the current consumption in these countries.

43Changes to the frequency, intensity and duration of extreme events will impact supply chains by increasing uncertainty in input flow and damaging infrastructure. This will make trade patterns less regular, increasing reliance on complex logistic systems. Extended droughts reduce hydropower generation, and can lead to increased wildfires, both of which can lead to brownouts, a particular threat to supply chains heavily reliant on temperature controls to maintain food quality and safety. Extreme events can also damage key transportation (road, rail, ports) infrastructure, limiting the distribution of products. Both low and high flow can affect the navigability of key inland waterways like the Mississippi River and Great Lakes in North America. For instance, a major drought in 1988 drastically reduced barge movement along the Mississippi a critical corridor for transporting grains in the USA to the port of New Orleans.

44In many countries the supply chain for livestock products is highly concentrated and coordinated, and disruptions anywhere along the supply chain can have impacts throughout the chain. For example, the COVID-19 outbreak in the USA prevented the processing of pigs, which forced pork farmers to either hold pigs, or cull them to avoid paying to feed them. There is more scope for keeping animal stocks for beef, but any culling in herds also has a longer impact on animal stocks due to lower reproductive capacity of cattle compared with chickens or pigs.

45We highlight in previous sections the potential impacts of climate change on the supply and processing of animal products, as well as on the quality and safety of animal products through contamination with pathogens or pesticide (sections 2.1.1.4, 2.3.3) and reduced nutritional quality and sensory appeal (section 2.2.1).

46The social licence of the livestock sector is increasingly challenged by the rising awareness of climate change and livestock impacts. These societal changes are also influenced by animal welfare and environmental concerns, and increasingly supported by civil society, governments and other institutions. Changing social norms are likely to impact diets, especially in high-income countries. Research also found that temperature changes can affect dietary preferences of consumers, although the effects on the consumption of livestock products remain largely understudied. Changes in the price of animal products and relative magnitude of price changes across commodities will also alter consumption patterns. These changes will be driven by product availability and quality, as well as costs along the supply chain. The rate of growth in gross domestic product, which can be influenced by climate change, will also alter consumers' purchasing patterns. Indeed, food price sensitivity tends to fall as incomes rise, and a smaller share of income is spent on food (i.e. Engel's Law; ). Sensitivity to primary commodity prices may also decline as supply chains complexify: increasingly the cost of the food is driven by value-added processing, packaging, and branding. For example, the farm share of the food value added in the European Union has fallen from 31% in 1995 to 24% in 2005.

47Labour availability and productivity are key determinants of the efficiency of the livestock supply chain. This is especially the case in less developed regions that do not rely on mechanization. However, human performance and health can be limited by a broad range of direct and indirect effects of climate change as reviewed in previous studies. These impacts relate to temperature, floods and storms, ultraviolet radiation, infections, air quality, nutrition, occupational health, mental health, violence and conflicts. For instance, in 2010, inhalation of climate-altering pollutants other than CO2 caused more than 7% of the global burden of disease. We further detail below two key drivers of labour productivity of particular concern to agricultural labour: heat stress and diseases.

48Above 24–26 °C, labour productivity declines. At 33–34 °C, a worker operating at moderate work intensity loses 50% of their work capacity. In addition, exposure to excessive heat levels can lead to heatstroke and sometimes fatal outcomes. Productivity may decrease by 11–27% by 2080 in hot regions such as Asia and the Caribbean, and labour productivity for high intensity work declines by up to 31%–⁠38% (RCP 4.5–⁠RCP 8.5) in Southeast Asia and the Middle East by 2050, relative to a 2050 baseline without climate change. In some areas, 30–⁠40% of annual daylight hours will become too hot for work to be carried out. Literature focussed on agricultural labour emphasises the challenges caused by physical labour under high temperatures, with dehydration often being the main cause of lost labour productivity. Heat stress is of particular concern in production systems that are dependent on high inputs of human labour and located in environments that are already hot, such as smallholders in sub-Saharan Africa, as highlighted in and. Foreign workers, who represent a large part of the agricultural labour force in some countries such as in the U.S. are also particularly at risk. Foreign farmworkers may have entered the country illegally, often do not speak the local language, have low education levels and incomes and limited access to health care. They often feel they have little control of their workplace, and tend to be reluctant to complain about unsafe work environments.

49In addition to heat stress, communicable diseases such as Malaria, dengue fever, tick borne encephalitis, borreliosis, salmonellosis, cholera and bluetongue are all affected by climate change and affect the health of people working along the livestock supply chain. For instance, excessive rainfall and high temperature results in an increased transmission rate, reproduction rate and the proliferation of the Plasmodium species that causes malaria, and outbreaks of leptospirosis in Sub Saharan Africa. In areas where there is inadequate access to water and sanitation, flooding increases people's exposure to salmonellosis and cholera. Conversely, a drying climate and extended periods of drought in Australia have been conducive to the spread of Q-fever, with the disease-causing bacteria Coxiella burnetii able to survive and travel long distances on dust particles contaminated with animal faeces or birth fluids. Warming temperatures in northern latitudes have also been reported to favour the distribution of anthrax, particularly in association with increased livestock densities. Populations in close contact with livestock and wildlife are particularly at risk of zoonotic pathogens, with climate change further contributing to the risk of zoonosis.

50Climate-related challenges from resource availability to consumption levels mentioned above can result in rises in costs of water, feeding, housing, storing, transport, retailing and insurance, which negatively impact actors throughout the livestock supply chain. An ensemble of crop and economic models with various representations of the global food system and associated assumptions (e.g. related to technological change, land-use policies or consumption patterns) projects a 1–29% increase in cereal prices by 2050 across Shared Socio-economic Pathways 1, 2, and 3 under climate change (RCP 6). This would impact consumers globally through higher food prices, with effects varying regionally. The price of animal products is also projected to increase, but projected prices are about half that of cereals, highlighting that in these models, the impacts of climate change on animal products are felt mainly through changes in costs and availability of feed. These results also highlight scope for feed substitution within the livestock sector. Most climate studies have focused on price levels and not on price volatility. However, with increased spatial and temporal variability in production and supply chain efficiency, it is probable that food prices will be more volatile under climate change. The international agricultural trade system is dominated by regional trade (i.e. within Europe and Mediterranean Basin, North America, etc.), with trans-regional trade dominated by a few major international exporters (e.g. soybeans from Brazil, maize from the US, rice from Southeast Asia, lamb from New Zealand-lamb, etc., ). Major shocks to these major producers could dramatically increase the volatility of international markets and reduce the reliable supply not only of food but also of feeds, which are traded at much higher volumes than livestock products. Furthermore, it can drive policy responses to protect domestic supply that can exacerbate food crises.

51Climate change is contributing to a range of societal changes with implications for the livestock sector. As noted above, changes in climate conditions will challenge productivity, economic viability, structure, and attractiveness of livestock-related enterprises with consequences on the availability, quality, affordability, safety and sensory appeal of livestock products. Climate change will likely amplify existing inequalities with particularly negative impacts on small-scale enterprises, which may struggle to remain competitive. Social learning, the process of learning from and of responding to experience, will also shape social-ecological systems. The mitigation and adaptation actions that food system actors themselves take may also affect social dynamics. For example, rising awareness of the negative consequences of climate change and the sector's environmental footprint may contribute to shifting social norms, particularly in high-income countries, that may favour less resource intensive production, and encourage the further development of novel protein products (plant-based meats, cultured meats, etc.). All these changes will impact both local and global food systems dynamics and food environments.

52The ability of the livestock sector to cope with or adapt to climate change, i.e., its adaptive capacity, depends on a range of socio-economic, political, institutional and environmental factors from local to global scales that interact in complex ways. Well known demographic, social and economic factors are driving change in the livestock sector. These factors are briefly presented below and treated in more depth in the Supplementary Information.

53A growing human population, together with increasing incomes and shifts in dietary preferences, predominantly in low- and middle-income countries, are driving continued growth in demand for livestock products (especially poultry and pork) and resources needed to produce them, including land and water. Potential enforced production systems transitions due to changes in land suitability will affect the food production landscape. Large-scale demand for plant-based diets and production of synthetic livestock products may also alter livestock production systems dynamics. The development of such food alternatives will be influenced by changing social norms and opportunity-costs of different food production systems under climate change and other global trends. In addition, competition between crop as livestock feed and crop for direct human consumption or biofuel production are increasing. Human migration, partly induced by climate change, can also increase local pressures on scarce natural resources. Rising food safety standards and governments' investments in infrastructure, price support schemes, taxes, credits, subsidies, input and output quotas and the health system, will also influence the vulnerability of livestock supply chains. For instance, the lack of infrastructure and access to knowledge, land, financial services, markets, new dynamic sectors and policy dialogues in rural areas is making the agricultural sector unattractive to youth, with negative implications for the agricultural sector's performance and adaptive capacity. Production challenges associated with harsh climates can also increase youth reluctance to enter or remain in agriculture. Discrimination based on gender, ethnicity, caste, and wealth impedes participation in markets, legal recognition of land and asset ownership, and other rights that play key roles in the ability of stakeholders to adapt to climate change. Civil conflict and institutional disregard of traditional knowledge, institutions and customary practices can also weaken the resilience of livestock systems. While some of the political and institutional forces may contribute to persistent trends (e.g. continued exclusion of pastoralists from political processes), others may result in dramatic shifts in the near future and greatly alter the ‘vulnerability landscape’. The pace at which governance factors can change has been exemplified by governments' actions to stop the spread of diseases (e.g. H5N1 or COVID-19).

54People in different production systems have differentiated exposures and vulnerabilities. Table 2, Table 3 highlight the exposure of people in different regions and production systems based on economic and livelihood indicators of the livestock sector. We summarise in Table 4 key vulnerabilities and adaptative capacity characteristics of people in grazing-only systems, mixed crop-livestock systems and industrial systems, as defined by and mapped in. These vulnerabilities are also discussed in, and a focus on adaptation in mixed systems is provided in,.

55Some characteristics of grazing systems compared with other food production systems are their usual remoteness from population centers and limited access to information, markets, capital, labour and veterinary services. These characteristics, together with market price fluctuations and social considerations (e.g. benefit perception), limit opportunities for management options that could mitigate the effects of climate change and high climate variability such as timely stock adjustments. People in these systems, often both geographically and politically distant from policy makers, tend to be marginalised and receive limited investments from governments and businesses. Grazing systems are often located on land that is poorly suited for cropping, notably due to high climate variability. Herd mobility and collective resource management have been key strategies to take advantage of the spatial and temporal variation in water and forage availability. However, changes in land tenure and increasing landscape fragmentation have become a major concern for grazing systems resilience in many parts of the world. For instance, in Inner Mongolia, the culture had adapted to the harsh climate by using mobility, cooperation, and reciprocity strategies as described in. However, collectivization between the 1950s and mid-1980s followed by market reforms in the early 1980s disrupted these traditions and weakened herd mobility. The state-driven nomad sedentarisation projects in China and transitions from communal to semi-commercial land tenures in southern African rangelands have also limited adaptive capacity of people in these regions. In Australia, where most grazing land is owned or leased, it is common practice to move cattle from a droughted area to other privately run properties that have adequate pasture, and graze for a fee (“agistment” strategy;; ). The shortage of productive land during widespread droughts can however greatly limit the possibility and economic viability of such a mobility strategy.

56Farmers in mixed crop-livestock systems often have more limited economic opportunities as compared to cropping systems. This is particularly the case for small-scale systems. Shrinking farm sizes limit opportunities for management adaptations and more stringent food safety standards limit the ability of smallholder farmers to enter expanding markets as these standards create new knowledge requirements, additional investments and stronger linkages between producers and buyers. The literature has extensively covered the impacts of climate change on crop enterprises in mixed systems, and to a lesser extent, the impacts on livestock enterprises, although the focus has often been on yields rather than livelihood outcomes or other sustainability dimensions. However, relatively little is known about how crop-livestock interactions may be affected or offer buffering capacity to help smallholders adapt to climate change.

57Industrial systems are generally based on high-yielding animals, which are more susceptible to heat stress, so require greater investment in infrastructure to insulate them from climate extremes. Compared to more extensive systems, the close physical proximity of animals to each other and humans can increase the possibility of disease outbreaks. These outbreaks can however be more quickly identified and contained. The limited contact with wildlife also reduces the risk of diseases from endemic carriers. Industrial systems tend to be energy and capital intensive which can make them vulnerable to disruptions to energy supply. Due to greater integration in efficient value chains, they are also more vulnerable to disruptions to transportation and demand given more limited capacity for long run inventory.

58Overall, livestock ownership itself can support farmers resilience in times of climate change and climate extremes. This is particularly the case in smallholder farming systems, where livestock plays multiple roles beyond producing food for the market. These roles include provision of wool, hides, skins, manure, animal traction, transportation and food for home consumption. Livestock can also contribute to the diversification of income and accumulation of capital savings, and as such can also be used as a risk reduction strategy by vulnerable farming communities. In particular, livestock production is often more resilient to high climate variability or short length of growing periods than crop production. Animals can also be sold in times of climate shock to help compensate for crop failures, loss of income or additional expenses. Diversified livestock systems, including the ownership of multiple animal species and breeds, improve dietary adequacy and diversity and food security. Shifts in animals breeds can serve as a climate adaptation strategy (e.g. shifts from cattle to camels in East and North Africa or from British Bos Taurus to Zebu Bos indicus in Australia,;;; ). Nevertheless, climate change may challenge the viability of livestock systems in some contexts, given environmental, socio-economic and institutional constraints to adaptation.

59The risks of climate-related impacts arise from the interaction of climate-related hazards (including hazardous events and trends) with exposure of the livestock supply chain to these hazards and the vulnerability of the socio-economic systems within which they are embedded (Fig. 1).

60Worldwide, rangeland communities that are projected to see the most negative impacts of climate change on vegetation are also amongst the communities most vulnerable, according to a range of socio-economic vulnerability indicators at local and national levels (see Supplementary Information, ). Across a range of dimensions, including food security, health, socioeconomics, and governance, found that the region with livestock-based food production at most risk was South Asia. They found that LMICs were generally more at risk than higher income regions, with the top decile of nations at risk located in Sub-Saharan Africa, seven of which were in Eastern Africa. The potential impacts emerging from such risks will affect all dimensions of sustainable livelihoods⁠—human, social, natural, physical, and financial capitals (Fig. 2).

61Climate-related risks will be context specific, but also greatly influenced by global socio-economic trends and shocks. Differences in risks arise from climatic and non-climatic factors and from multidimensional inequalities, often due to uneven development processes. These context specificities highlight the importance of renewed attention to diversity within the livestock sector and its multiple socio-economic and environmental contributions. For instance, in Oceania, livestock represents 70% of agricultural added-value (Table 3) and over 90% of livestock feed is from grazing (Table 2, Table 3). Most of Oceania's grazing lands (mainly in Australia) are subject to high and increasing climate variability and frequent and intense fires, droughts and floods. As such, the viability of the sector will greatly depend on the success of the strategies implemented to cope with high climate variability. These strategies will need to account for geographic remoteness, limited economic margins, and over-grazing. Increasing climate variability and drying trends on grasslands is also a key concern in arid and semi-arid African regions (where 60% of feed is from grazing), as are changes in the length and start of the crop growing seasons (over 70% of protein production comes from ruminant mixed systems and grain-fed chicken; Table 2, Table 3; ). Socio-economic and institutional support, including social safety nets, is urgently needed considering the multiple roles livestock plays in such regions. Overall, small-scale farmers in environments that are already hot and with limited resources for adaptation will be the most at risk. We estimate that 228.6 million people worldwide live in low-income countries in grazing-only and mixed crop-livestock regions in arid environments. Agriculture represents approximatively 57% of total employment in these countries (see Supplementary Information for calculations). Although the share is likely to decline in the future, in absolute terms, the number of vulnerable people employed in agriculture will remain high.

62The impossibility of predicting complex interconnected biophysical and social systems far into the future challenges the evaluation and prioritisation of climate risk mitigation and management strategies. Given this fundamental uncertainty, it is crucial to consider a wide range of plausible futures and climate outcomes, which while it may not allow us to predict the future, can allow us to assess the range of possible impacts and to test the robustness of policies and interventions across alternative contexts.

63The rapid growth of the livestock sector and its various contributions to the economy and human livelihoods highlight the importance of better understanding the impacts of climate change on livestock.

64While it is certain that climate change will impact the sector throughout the food supply chain⁠—from farm production to processing operations, storage, transport, retailing and human consumption⁠—large uncertainties remain as to the nature, extent and magnitude of these impacts. These uncertainties relate to the future climate as well as exposures and responses of the interlinked human and natural systems to climatic changes over time. For instance, the impact of climate change on agricultural labour, on the livestock supply chain from farmgate to retailers or on livestock products prices and consumption behaviours have received limited attention in the past. The impacts of climate change on livestock feed production and quality, especially forages, have also been understudied as compared to key staple grains. In addition, the rates of change in global socio-economic and environmental drivers are such that the past may no longer be a reasonable indicator of the future. Further research will help reduce these uncertainties.

65Key hazards to the livestock sector relate not only to climate change trends but also, and importantly, to climate variability and climate extreme events such as heat waves, droughts, floods, cyclones, and wildfires. Increases in inter-annual variability of forage availability is especially a concern for the grazing sector. The increase in frequency, intensity and duration of heat waves also poses a major threat to animal health and human labour where access to mechanization and cooling systems is limited. These climate-related hazards can exacerbate other stressors with negative impacts, especially for people living in poverty. The nature and extent of such impacts are however still largely unknown and warrant further research.

66In the face of global warming, and overall harmful impacts as highlighted in this review, the existing suite of adaptation strategies across ecological, socioeconomic, and institutional systems and coping range that have been developed in response to existing weather patterns may not be enough. More transformative climate adaptation may be required. These range from farm management adjustments, technological developments, income-related responses to institutional changes and, in extreme cases, abandonment of livestock keeping (see reviews by,,,,,, and ). Livestock ownership may also be valued as a climate risk-mitigation strategy in some contexts. Indeed, livestock plays multiple roles beyond producing food for the market, especially in smallholder farming systems. For instance, animals are capital assets that can contribute to income diversification. Livestock are also often more resilient than crops to high climate variability.

67Adaptation choices and risk management actions across temporal and spatial scales and contexts will need to build on robust methods of designing, implementing and evaluating detailed development pathways. Such pathways, yet to be fully elucidated, must strengthen climate-resilience and limit trade-offs between different actors. Choices and actions will need to account for the widest possible range of potential impacts, including those with low probability but large consequences, because of the large future uncertainties related to hazards, exposure and vulnerabilities, and the large potential consequences for the sector.

68The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.