Overview of climate changes affecting livestock production systems

The impact of climate changes on livestock production will depend on the magnitude and nature of the changes and can be mediated through both direct effects on the animal and indirect effects on their environment; for example, through feed quantity and quality.

Changes in climate variables are commonly expressed relative to a 1960–90 ‘climate normal’ baseline; for example, temperature increases of 1.5°C or 2.0°C above this baseline correspond to agreed efforts to limit global warming by 2100 under the Paris Agreement of December 2015 (United Nations Framework Convention on Climate Change, 2015). The impacts of climate changes with 2°C average global temperature rise may be more severe than assessed in earlier studies by the Intergovernmental Panel on Climate Change and may be felt sooner. In particular, the risks associated with increasing frequency and severity of extreme weather events (Tebaldi and Wehner, Reference Tebaldi and Wehner2018) appear to be significantly greater at a global average increase of 2°C than at 1.5°C.

For agriculture, prospects of changes in regional patterns of rainfall as well as temperature are important for developing adaptation strategies. Global precipitation changes and the overall intensification of the hydrological cycle respond to factors other than rising temperature; for example, local soil moisture feedback mechanisms. This results in a higher degree of uncertainty for livestock production.

Regional climate change impacts

Modelling predicts significant differences in regional climate indicators (King and Karoly, Reference King and Karoly2017; King et al., Reference King, Karoly and Henley2017) not only in the severity but also the direction of changes. Results indicate that tropical and subtropical regions will face negative impacts earliest and, in fact, climate changes are already being observed in some areas, for example, Australia (King et al., Reference King, Karoly and Henley2017) and south-east Asia (Thirumalai et al., Reference Thirumalai, DiNezio, Okumura and Deser2017).

The tropics and subtropics are home to much of the world’s grasslands and grazing lands and to some of the world’s poorest peoples who rely on those lands for food security. Ruminant livestock is critical to meeting nutritional needs in these regions where socio-economic factors, including a reliance on local food production and vulnerability to land degradation, add to the risk of food insecurity. While communities have traditionally managed for high climate variability using locally adapted breeds of animals that are more resilient than many temperate phenotypes, the capacity to cope with hotter, often drier and more extreme conditions in the future will require additional investment in adaptation.

Understanding the impacts of climate changes is critical to developing effective adaptation strategies. First, the perceived risk is a major determinant of investment in adaptive actions, and second, the effectiveness of adaptation strategies depends on how well they are matched to the impacts. In livestock production systems there is evidence of incremental changes in management occurring in response to trends in climate averages. Transformative adaptive responses will more likely be driven by extreme conditions that occur less regularly and threaten the survival of animals and those who depend on them. It is the extreme conditions that determine viability, profit and animal productivity and welfare outcomes as recorded in recent events such as the record hot summer of 2003 in Europe, which also caused tens of thousands of human deaths (King and Karoly, Reference King and Karoly2017).

Despite remaining uncertainties in simulations of the impacts of global warming, research is guiding decisions by farmers, industry and policymakers. Regional analyses for Australia and Europe on the effects of different levels of warming on large-scale heat extremes (King and Karoly, Reference King and Karoly2017; King et al., Reference King, Karoly and Henley2017) demonstrate the need for change. For example, under 2°C global warming, large scale heat waves such as that experienced in Europe in the summer of 2003 may occur as often as one in every 2 years. In Australia, the very hot summer experienced in 2012–13 is predicted to be more common under either 1.5°C or 2°C warming. While not precise, these studies highlight the risk of business-as-usual management, and particularly in industries such as ruminant livestock production that are both sensitive to climate impacts and generally located in regions more vulnerable to extreme changes resulting from global warming. A significant gap in knowledge relates to hydro-meteorological extremes that are of high importance for livestock production particularly in arid and semi-arid regions.

Ruminant production and climate vulnerability

Livestock, including ruminant and monogastric species, contribute 18% of global kilocalorie consumption and 25% of global protein consumption (Mottet et al., Reference Mottet, de Haan, Falcucci, Tempio, Opio and Gerber2017). Globally, ruminants (cattle, buffalo, sheep and goats) produce an estimated 36 355 Mt of protein per year (Mottet et al., Reference Mottet, de Haan, Falcucci, Tempio, Opio and Gerber2017) and provide micronutrients essential for human well-being. The ability of ruminants to convert ligno-cellulose to meat and milk that are sources of high-quality protein and essential nutrients contributes significantly to human health. This is particularly important in some of the poorest regions of the world. In addition, ruminant livestock make a positive contribution to food security and well-being of communities by providing manure and draught power for other agricultural production. Livestock is also a reliable source of income from household to national scales. Maintenance or growth of the sector to meet future demand for products and services provided by ruminant livestock must be balanced against a need to manage the well-documented environmental impacts and potential for negative consequences for global food security. These include significant contributions to greenhouse gas emissions, water use and occupation of land suitable for human food production for grazing or feed production (Opio et al., Reference Opio, Gerber and Steinfeld2011; Henry et al., Reference Henry, Charmley, Eckard, Gaughan and Hegarty2012; Mottet et al., Reference Mottet, de Haan, Falcucci, Tempio, Opio and Gerber2017).

Global food security and the growing demand for animal products related to rising population and increasing income emphasise the importance of climate adaptation strategies to minimise adverse impacts on animal production. However, ruminant agriculture is arguably one of the food and fibre production systems most vulnerable to climate changes. Much of the production is in extensive rangelands where inputs are low, and forage production and water supply vary with seasonal climate. Global feed production for all livestock requires 2.5 billion ha of land, but ~90% of livestock protein is derived from feed produced on 2 billion ha of grassland of which 1.3 billion ha is non-arable (Mottet et al., Reference Mottet, de Haan, Falcucci, Tempio, Opio and Gerber2017). Much of the rangelands occupied by ruminant animals is in tropical and subtropical regions sensitive to the impacts of climate changes, with more than 50% of total meat and 60% of milk produced in these zones (Mottet et al., Reference Mottet, de Haan, Falcucci, Tempio, Opio and Gerber2017). Intensive ruminant production is more vulnerable to welfare and health impacts of climate changes through heat stress and there is a threat to both intensive and extensive systems of new or expanded exposure to pests and disease.

Potential impacts of climate changes on livestock include direct effects on animal physiology, production and behaviour (Henry et al., Reference Henry, Charmley, Eckard, Gaughan and Hegarty2012) and indirect effects mediated by changes in water availability and the quantity and quality of forage and feed crops (Craine et al., Reference Craine, Elmore and Angerer2016). Factors influencing these effects include increases in atmospheric levels of carbon dioxide (CO₂), higher temperatures and changes in amount, seasonality and variability of precipitation (Henry et al., Reference Henry, Charmley, Eckard, Gaughan and Hegarty2012). Changes are already being observed by farmers in some regions. For example, in northern Australian rangelands, projections of longer, more severe droughts will increase the risk of land degradation, while heavier rainfall and flooding events are likely to also contribute to extensive soil erosion and to expanding the range of some weeds, pests and diseases (McKeon et al., Reference McKeon, Stone, Syktus, Carter, Flood, Ahrens, Bruget, Chilcott, Cobon, Cowley and Crimp2009).

Changes by farmers already being initiated in response to effects in pasture systems include adjusting inputs (e.g. fertilizer or irrigation), selection of forage species (including deep-rooted grassland mixtures in drought and water-logged soils; Hoekstra et al., Reference Hoekstra, Suter, Finn, Husse and Lüscher2015), managing grazing times and rotations, and farm system choices such as cropping/pasture mix (Cullen et al., Reference Cullen, Eckard and Rawnsley2012). Responses to direct impacts on ruminant animals include changes in feed management (e.g. balancing feed, forage, supplements), stocking rates, genetic selection, disease risk management (e.g. vaccination or selection of resistant genotypes), climate control (e.g. shade, fans, water sprays) and timing of reproduction.

Climate changes and forage production

Climate change effects on productivity and quality of pasture and forages consumed by ruminants are expected to be regionally specific. In temperate regions, the effects of climate changes on grassland production may be positive due to prolonged growth with slight increases in ambient air temperature combined with elevated CO₂, provided water and nutrient supplies are not limiting. In tropical and sub-tropical rangelands, however, both high temperatures and water scarcity appear more likely to result in lower forage quantity under future climate. In both regions, nutritive value of forages may decrease as crude protein (CP) and digestibility decline. More research is needed to understand the species mix most likely to support livestock ranging from high producing dairy cattle to hardy beef cattle, sheep and goats.

With a changing climate, there is potential for changes in both the between-year variability and the within-year seasonality of production. Climate changes are likely to mean more severe drought conditions, through greater evapotranspiration (Hennessy et al., Reference Hennessy, Whetton and Preston2010) which will exacerbate plant water stress, and increased intensity of large-scale climate drivers (e.g. El Niño and Indian Ocean Dipole). These impacts in turn will result in greater variability in forage production between years, with adaptation options focusing on improving longer-range weather forecasting, coupled with more proactive risk management (e.g. increased forage conservation or resting pasture).

Changes in pasture species

In regions where rainfall is evenly distributed or summer-dominant and future scenarios predict that this pattern will continue with warming, then it is likely that C₄ grasses could dominate production in future climates (Howden et al., Reference Howden, Crimp and Stokes2008; Cullen et al., Reference Cullen, Eckard and Rawnsley2012). Typically, this could occur in regions intermediate between winter and summer dominant rainfall, as climate change warms the mid-latitudes and pushes weather systems poleward. Proactive adaptation in these regions would seek to capitalise on the change through a mixture of C₃ and C₄ species in forage systems, thereby spreading risk and capturing the best of both warm and cooler season rainfall (Cullen et al., Reference Cullen, Eckard and Rawnsley2012).

In Mediterranean climates or winter-dominant rainfall regions, where future climate predictions show no shift in rainfall distribution, the productivity of summer dominant forages and pastures will only improve at the shoulders of the production season. Under these conditions, studies show very limited benefit of a shift from C₃ to C₄ species, with adaptation focusing more on selection of heat tolerance and deeper rooting C₃ species (Cullen et al., Reference Cullen, Eckard and Rawnsley2012).

Changes in forage quality

A changing climate is likely to affect forage quality in ways that vary regionally. For example, a meta-analysis by Dumont et al. (Reference Dumont, Andueza, Niderkorn, Lüscher, Porqueddu and Picon-Cochard2015) reported differences in the expected effects in mountain regions, which may benefit from a change in climate through warmer winters reducing low-temperature limitations to pasture growth, as compared with Mediterranean regions where drought and heat may constrain pasture quality in future climates. In species that require vernalisation for flowering quality decline may be associated with both earlier flowering and reduced seed production. Craine et al. (Reference Craine, Ocheltree, Nippert, Towne, Skibbe, Kembel and Fargione2017) reported that increasing temperature and declining precipitation decreased forage CP content and increased the ratio of digestible organic matter to CP content for regions with continental climates. The cost of providing supplemental CP (e.g. soybean meal) to US cattle to compensate for the decline in CP content of forage was estimated at $1.9 billion USD annually (Craine et al., Reference Craine, Ocheltree, Nippert, Towne, Skibbe, Kembel and Fargione2017). Additionally, increasing use of human-edible feed and soybean meal, currently estimated at 6.7 kg/kg protein produced from ruminant livestock (Mottet et al., Reference Mottet, de Haan, Falcucci, Tempio, Opio and Gerber2017) may contribute to environmental costs with soy production considered a major driver of land-use change. In temperate high rainfall climates, particularly in dairy pastures where nitrogen (N) fertiliser is applied, pasture protein levels are typically in excess of animal requirements for much of the year. Hence a reduction in plant CP content may well be beneficial to livestock production while also reducing urinary N loss and thus nitrate leaching and nitrous oxide and ammonia emissions to the atmosphere (Wilkinson and Lee, Reference Wilkinson and Lee2017). However, in more extensive grasslands already marginal in CP content, it is both impractical and ineffective to apply N to redress this protein deficit. Adaptation strategies are likely to require more CP supplementation in these more extensive grazing systems in future climates. In addition to limiting animal growth, a decline in forage quality may also promote greater enteric methane production. High fibre, low digestibility feeds favour biochemical pathways that form more hydrogen and slower rumen passage, both of which raise the potential for methane loss (Wilkinson and Lee, Reference Wilkinson and Lee2017).

In summary, there is a need for greater research on the effects of individual and combined climate change factors (elevated temperature and CO₂ levels, drought) on forage production and quality as the interactions among these factors make it exceedingly difficult to predict responses.

Changes in rangeland composition

While climate change is likely to affect the mix of herbaceous species in a grassland system, there is also evidence that rising CO₂ levels and increasing temperatures may be driving an underlying change in the balance between herbaceous and woody species in grasslands. A recent analysis of growing season integrated Leaf Area Index shows increased greening in over 25% to 50% of the global vegetated area (Zhu et al., Reference Zhu, Piao, Myneni, Huang, Zeng, Canadell, Ciais, Sitch, Friedlingstein, Arneth, Cao, Cheng, Kato, Koven, Li, Lian, Liu, Liu, Mao, Pan, Peng, Peñuelas, Poulter, Pugh, Stocker, Viovy, Wang, Wang, Xiao, Yang, Zaehle and Zeng2016), with multiple global ecosystem models suggesting that CO₂ fertilisation explains 70% of the observed greening trend. Similarly, modelling of the Australian rangelands showed a significant shift in areas of vegetation classes from the period of 1931–70 to 1971–2010 due to the increased rainfall over large areas (Liu et al., Reference Liu, Zhang, Henry, Zhang and Grace2017). This modelling projected a range of changes in vegetation classes for 2050 depending on the climate change scenario used, with Net Primary Productivity likely to increase in Australian rangelands depending on management of cattle and sheep grazing. In the grazed savannas, use of fire and other management strategies to control proliferation of woody vegetation will affect the balance between net carbon accumulation in vegetation and soil, relative to livestock and savannah burning emissions, and thus whether these regions are a net carbon sink or source (Burrows et al., Reference Burrows, Henry, Back, Hoffman, Tait, Anderson, Menke, Danaher, Carter and McKeon2002).

Rising atmospheric CO₂ has been shown to be a key explanatory driver for observed bush encroachment (Scheiter and Higgins, Reference Scheiter and Higgins2009) in sub-Saharan Africa. Further analysis shows that CO₂ fertilisation is not saturated at current ambient levels and would increase in response to increasing CO₂ levels. Together with evidence that rising CO₂ levels are reducing the CP content of unfertilised native grasslands (Craine et al., Reference Craine, Ocheltree, Nippert, Towne, Skibbe, Kembel and Fargione2017), the savannah grasslands of the world will very likely continue to be challenged by bush encroachment into the foreseeable future. With increasing public resistance to expansion of land clearing for livestock production, the active management of bush encroachment could become a contentious issue for livestock industries in future. One adaptation option would be to develop a broader carbon offset market for farmers to be rewarded for avoided deforestation and management of native forest regrowth, combined with practices to more efficiently manage whole farm productivity and potentially growth in intensive pasture or feedlot feeding to maintain or increase total output.

Genetic diversity of plants

Models have been used to predict the response of plant traits to selection under climate change (Hoffmann and Sgrò, Reference Hoffmann and Sgrò2011). These models predict that extinction rather than evolution is likely if environmental conditions change too quickly. However, climate change can also increase the intensity of directional selection for particular traits, theoretically leading to a faster evolutionary response (Hoffmann and Sgrò, Reference Hoffmann and Sgrò2011). It is unclear whether individual forage species can evolve to keep up with the rate of climate change, or if outcomes are more likely through changes in species composition across multi-species grasslands.

Climate change adaptation involves increasing the resilience of the system, on the assumption that future climates will push species to the limits of their tolerance range. This can be achieved through both increasing within and between species diversity in the forage or pasture system. Intensive selection of forage species over multiple generations has resulted in highly productive pastures and the forages species in use today. Improved pasture species such as perennial ryegrass (Lolium perenne), have historically been selected for increased winter production, perhaps at the expense of drought and heat tolerance, and such species may have less genetic diversity than their ancestral parent populations. This makes them more vulnerable to a change in their ideal growing environment; particularly as they are typically sown in mono-specific swards. Cullen et al. (Reference Cullen, Eckard and Rawnsley2012) showed that annual pasture production of a range of C₃ species either increased or showed little change with up to 2°C warming in south-eastern Australia, but production declined rapidly thereafter, and more heat-tolerant C₄ grasses were required to maintain forage production. In contrast, with the natural distribution of the C₄ grass, Themeda triandra, extending throughout Australia, southern and eastern Africa, South-east Asia, India and the Middle East (Snyman et al., Reference Snyman, Ingram and Kirkman2013), this species is likely to have a significantly greater inherent genetic diversity to adapt to a changing climate, albeit with lower productivity than pastures like perennial ryegrass. In Europe, cultivars of Festulolium (Fescue×Lolium hybrids) are being studied for a potential role in adaptation of pastures to climate change through improved growth in winter waterlogged soils or in warmer, drier summers as extreme conditions increase. Overcoming low productivity relative to ryegrass will limit widespread use for some time, but some use to increase diversity and, hence, resilience to climate may occur (Ghesquière et al., Reference Ghesquière, Humphreys and Zwierzykowski2010).

In contrast to mono-specific pastures, species-rich grasslands should have greater between-species diversity to help maintain ecosystem function with changes in climate. Craine et al. (Reference Craine, Angerer, Elmore and Fierer2013) show that most native grasslands contain a high diversity for drought tolerance, concluding that these diverse grasslands throughout the globe have the potential to be resilient to drought in the face of climate changes through local expansion of drought-tolerant species.

Climate change adaptation may therefore require land managers to limit mono-specific forage and pasture systems to regions with low rainfall variability and where temperature changes remain within the thermal tolerance of the selected species. In regions of higher rainfall variability and greater temperature change, a multi-species grassland system is likely to confer greater resilience as the climate changes, but also allowing natural selection to drive the species composition change.

Heat stress

Acute heat stress can lead to morbidity in livestock and the management of long-term high ambient temperatures or short-term acute heat waves is a critical concern for animal welfare as well as for food and fibre production. However, chronic heat stress may lead to greater economic loss through sub-clinical health issues and overall reduced animal performance. Livestock normally maintain their body temperature within a fairly narrow range (±0.5°C). When animals are exposed to combinations of high temperature, solar radiation and humidity and are not able to adequately dissipate accumulated heat, heat stress occurs (Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016; Seijan et al., 2017).

A heat stress response can be influenced by numerous factors: species, breed, previous exposure, health status, level of performance, body condition, age, coat colour, lactation stage and body condition (Dunshea et al., Reference Dunshea, Leury, Fahri, DiGiacomo, Hung, Chauhan and Gaughan2013; Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016). The negative effects of heat stress include compromised metabolic and digestive functions (Gauly et al., Reference Gauly, Bollwein, Breves, Brügemann, Dänicke, Daş, Demeler, Hansen, Isselstein, König and Lohölter2013) that reduce milk production and quality, compromise animal welfare, reduce reproductive efficiency, and in extreme cases, cause death (Kadzere et al., Reference Kadzere, Murphy, Silanikove and Maltz2002; Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016).

The intensity and duration of exposure to hot and humid conditions is also a factor in heat stress (Dunshea et al., Reference Dunshea, Leury, Fahri, DiGiacomo, Hung, Chauhan and Gaughan2013). Whilst there is extensive research into the impact of short-term acute heat stress on dairy cattle (Kadzere et al., Reference Kadzere, Murphy, Silanikove and Maltz2002; Garner et al., Reference Garner, Douglas, Williams, Wales, Marett, DiGiacomo, Leury and Hayes2017), research into extended exposure to heat stress and its cumulative effects under farm conditions is limited with the exception of beef feedlot studies (e.g. Gaughan et al., Reference Gaughan, Bonner, Loxton, Mader, Lisle and Lawrence2010a). Under conditions of acute heat stress, milk productivity has been shown to be negatively affected within 24 h in proportion to heat load (Silanikove et al., Reference Silanikove, Shapiro and Shinder2009). Garner et al. (Reference Garner, Douglas, Williams, Wales, Marett, DiGiacomo, Leury and Hayes2017) found that exposing dairy cows to moderate heat stress over 4 days in a climate-chamber controlled study reduced milk yield by 53% and reduced dry matter intake by 48% (n=6). Heat stress events may be mitigated by cooler night-time temperatures that allow the cow to offload accumulated heat (Silanikove et al., Reference Silanikove, Shapiro and Shinder2009; Dunshea et al., Reference Dunshea, Leury, Fahri, DiGiacomo, Hung, Chauhan and Gaughan2013).

Management options

The impacts of extreme heat load events on livestock can be mitigated by adjusting their environment and diet, and by selecting animals that are more heat tolerant (Renaudeau et al., Reference Renaudeau, Collin, Yahav, De Basilio, Gourdine and Collier2012). Adjusting the animals’ environment and improving the ability of the animal to dissipate body heat can include providing shade, sprinklers, misters and fans, but ensuring access to adequate drinking water is paramount (Gaughan et al., Reference Gaughan, Bonner, Loxton, Mader, Lisle and Lawrence2010a; Henry et al., Reference Henry, Charmley, Eckard, Gaughan and Hegarty2012). Nutrient deficiencies arising from heat stress can be mitigated by adjusting diets. For example, increasing CP and digestible energy content of diets can help compensate for lower feed intake during heat stress while decreasing fibre content of diets decreases metabolic heat production (Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016). Supplementation with vitamins C, E and A, certain trace minerals (e.g. zinc, selenium) and electrolytes may help relieve oxidative stress and boost immunity (Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016), although more research is needed to understand the unique challenges of meeting the nutrient requirements of heat-stressed ruminants.

Gantner et al. (Reference Gantner, Bobic, Gantner, Gregic, Kuterovac, Novakovic and Potocnik2017) found that some animals within the same breed exhibit greater tolerance to heat stress due to genotypic differences. With the progression of climate change longer-term, however, livestock producers in more marginal regions may need to consider changing the types of animals and genotypes used (Gaughan et al., Reference Gaughan, Lacetera, Valtorta, Khalifa, Hahn and Mader2009).

Beef cattle

Pastured beef cattle are usually able to seek shade, water and air movement to cool themselves, thus heat stress is usually a greater problem for feedlot cattle than breeding cows and growing cattle in rangeland or pasture settings. However, heat stress can negatively affect reproductive function of beef cows. For example, cows exposed to elevated ambient temperature during late gestation experienced premature onset of parturition due to increased body temperature (Wright et al., Reference Wright, Boehmer, Cooper-Prado, Bailey and Wettemann2014). In feedlots, animals experience radiant heat from dirt or concrete surfaces and without appropriate management will show performance loss due to decreased efficiency and feed intake resulting from physiological stress (Mader et al., Reference Mader, Davis and Brown-Brandl2006). Inability to abate the greater heat load can potentially result in death. Management strategies to reduce the risks in feedlots may include regular removal of manure, diet changes to reduce digestive metabolic heat (e.g. increased fat content, lower fibre content), frequent water changes to ensure availability of cool, clean water and provision of additional shade and cooling (e.g. sprinkler) equipment. Warning systems using heat load triggers such as prolonged high temperature and humidity, or forecasted extreme weather are already being used to activate management response in feedlot conditions (Mader et al., Reference Mader, Davis and Brown-Brandl2006; Gaughan et al., Reference Gaughan, Bonner, Loxton, Mader, Lisle and Lawrence2010a; Henry et al., Reference Henry, Charmley, Eckard, Gaughan and Hegarty2012).

Dairy cattle

Dairy cattle are highly susceptible to heat stress, more so than other cattle breeds due to their relatively higher metabolic rate during lactation (Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016). Heat stress reduces milk yield, milk components, fertility and feed efficiency, while increasing culling and mortality rates (Das et al., Reference Das, Sailo, Verma, Bharti, Saikia and Kumar2016). Furthermore, the effects of heat stress increase with parity of dairy cows (Misztal, Reference Misztal2017). An analysis of the US dairy industry projected a 0.60% to 1.35% decline in milk production by 2030 due to heat stress, depending upon the climate model used (Key and Sneeringer, Reference Key and Sneeringer2014). In southern States, losses greater than 2% were predicted.

Dairy animals are able to acclimate to their environment to some extent. Higher producing dairy breeds originating from temperate climates (e.g. Holsteins) have been shown to be less able to cope with heat stress than lower producing breeds from warm and humid climates (e.g. Jerseys) (Nidumolu et al., 2010). This is further compounded by higher producing cows having a greater susceptibility to heat stress, as metabolic heat increases with feed intake required to maintain higher production levels (Dunshea et al., Reference Dunshea, Leury, Fahri, DiGiacomo, Hung, Chauhan and Gaughan2013).

Management strategies to reduce the impact of heat stress on dairy cattle would be similar to those of feedlot cattle, addressed above.

Sheep

Sheep and goats appear generally less susceptible to heat stress than other domesticated ruminants, with unique characteristics such as water conservation capability, higher sweating and respiratory rate and lower basal heat production (Kadzere et al., Reference Kadzere, Murphy, Silanikove and Maltz2002). However, heat stress especially when accompanied by high ambient humidity has been shown to be a major factor in poor reproductive performance of sheep in pastoral environments (Harle et al., Reference Harle, Howden, Hunt and Dunlop2007). When exposed to continuously hot conditions, ewes that are poorly adapted to heat stress may suffer increased levels of embryonic mortality and reduced foetal growth, and mortality in lambs from heat-stressed ewes appears to be higher than in lambs produced during cooler conditions (Harle et al., Reference Harle, Howden, Hunt and Dunlop2007). However, ewes acclimatised to high temperatures and subjected to diurnally fluctuating temperatures do not show these effects indicating potential for adaptation to hotter conditions (McCrabb et al., Reference McCrabb, McDonald and Hennoste1993).

Under future levels of global warming, the combination of lower reproductive rates, poorer growth and survival of lambs and reduced growth of adult sheep may make some areas less suitable for sheep production under climate change. Gaughan (Reference Gaughan2017) suggests that indirect effects of climate change, for example, on nutrition, may be more important for future sheep production than direct impacts of higher temperatures. However, there is evidence for benefits from adaptive management strategies including stronger selection for more heat tolerant animals, preferential supplementary feeding of ewes, ultrasonic scanning of ewes during pregnancy to target beneficial management, and, where appropriate, shifting mating time so that lambing coincides with peak forage availability and more moderate temperatures.

Goats

Goats are renowned amongst ruminants for their ability to tolerate extremes of temperature and humidity and to function well even in conditions of water deprivation, water restriction and energy restriction (Kaliber et al., Reference Kaliber, Koluman and Silanikove2016). An ability to reduce metabolism allows goats to survive even after prolonged periods of severe limited food availability. In a Turkish study on crossbred goats (a mix of 75% temperate breed with 25% locally adapted breed), the animals showed capacity to endure significant and extended water restriction (Silanikove, Reference Silanikove2000). During summer (July–August), with average daily temperature of 34.2°C (ranging from 23.1°C to 42°C) and an average relative humidity of 68.2%, goats demonstrated a capacity to endure significant water restriction for >30 days, with minimal disruption of plasma electrolyte and metabolite composition, thermoregulation and general well-being (Kaliber et al., Reference Kaliber, Koluman and Silanikove2016).

Adaptive responses to future climate changes may include monitoring for indications of heat stress, increased disease monitoring and increasing access to clean water and feed where coping thresholds are exceeded.

Species selection in response to climate change

Seo et al. (Reference Seo, McCarl and Mendelsohn2010) reviewed the scientific evidence for effects of climate conditions on livestock production. Indicators of animal performance such as weight gain, milk production, wool production and reproductive efficiency were influenced by ambient temperature, humidity and wind speed.

The impacts of climate change in more marginal climatic regions and the sensitivity of animal husbandry to climate may lead to an adjustment of livestock species to maintain the viability of farm enterprises (e.g. Seo and Mendelsohn, Reference Seo and Mendelsohn2008; Seo et al., Reference Seo, McCarl and Mendelsohn2010). Modelling by Seo et al. (Reference Seo, McCarl and Mendelsohn2010) indicated that in South America, adaptation responses to climate change in terms of species selection varied with climate scenario and between countries. Under a hot and dry future scenario, numbers of beef and dairy cattle would decrease by 3.2% and 2.3%, respectively, while sheep numbers would increase by 7%. An increase in the sheep population was predicted to occur mainly in the Andes mountain countries of Chile, Colombia, Ecuador and Venezuela. For a milder and wetter climate scenario beef cattle increased in countries of high current production, such as Argentina and Chile. Using data from 9000 African livestock farmers in 10 African countries, Seo and Mendelsohn (Reference Seo and Mendelsohn2008) reported that farmers were more likely to maintain livestock as temperatures increased and as precipitation decreased, whereas under cooler temperatures and wetter conditions, they favoured crop production. In warmer locations, producers favoured goats and sheep rather than beef cattle and chickens. Thus, farmers may adapt to climate change by shifting from crop production towards livestock production, or mixed farming systems, with a shift in the animal species most suited to the local environmental conditions.

Breed selection: climate tolerance v. performance

Selection of high-performance animals over many years has clearly led to increased productivity in cattle and sheep. However, genetic selection is usually carried out under optimal temperature and humidity conditions and is frequently tailored to more intensive systems in temperate climates or controlled, air-conditioned housing. This approach may result in animals that are less adaptive to less favourable conditions (e.g. Eisler et al., Reference Eisler, Lee, Tarlton, Martin, Beddington, Dungait, Greathead, Liu, Mathew, Miller and Misselbrook2014; Misztal, Reference Misztal2017). There is a strong relationship between production level and metabolic heat generation (Renaudeau et al., Reference Renaudeau, Collin, Yahav, De Basilio, Gourdine and Collier2012) so that selection for improved meat, milk and/or fibre traits may increase animal susceptibility to elevated temperatures. An indication of the potential vulnerability to global warming of high-performance stock can be seen in poor outcomes such as when Holstein dairy cows bred in Europe were imported into hot regions of Sub-Saharan Africa or India (Eisler et al., Reference Eisler, Lee, Tarlton, Martin, Beddington, Dungait, Greathead, Liu, Mathew, Miller and Misselbrook2014). Experience in the Indian state of Kerala has shown that smaller native cows are economically favoured over imported large bodied breeds despite having less than one-third the milk yield as health and survival requires less inputs (Eisler et al., Reference Eisler, Lee, Tarlton, Martin, Beddington, Dungait, Greathead, Liu, Mathew, Miller and Misselbrook2014). Many local breeds of cattle in these tropical latitudes have lower basal metabolic rates and can dissipate a substantial amount of heat by sweating and panting, and thus are more resilient under heat stress (Gaughan et al., Reference Gaughan, Mader, Holt, Sullivan and Hahn2010b).

Natural selection has favoured breeds with higher heat tolerance that are more able to cope with thermal variations and stress in tropical and sub-tropical regions. These tropically adapted breeds also have a greater capacity to grow and reproduce during unfavourable seasonal conditions with poor nutrition and higher disease and parasite pressures than high-performance breeds (Renaudeau et al., Reference Renaudeau, Collin, Yahav, De Basilio, Gourdine and Collier2012). Eisler et al. (Reference Eisler, Lee, Tarlton, Martin, Beddington, Dungait, Greathead, Liu, Mathew, Miller and Misselbrook2014) warn of the unsustainability of replacing locally-adapted animals with those selected for optimal conditions in temperate climates. Nevertheless, despite a high level of selection in some species of domesticated livestock, there is still significant genetic variation for many traits that could provide potential for further adaptation (Hayes et al., Reference Hayes, Lewin and Goddard2013).

Genomic selection that enables prediction of the genetic merit of animals from genome-wide single nucleotide polymorphism markers is expected to continue to be adopted to increase genetic gains in growth rates, milk production and other desirable traits (Hayes et al., Reference Hayes, Lewin and Goddard2013). Alongside more traditional selection and cross-breeding methods the use of whole-genome sequence data and advanced reproductive technologies can assist in accelerating selection to a level needed for adaptation to expected rates of climate changes. However, genetic selection for lower body temperature as a strategy to reduce the magnitude of heat stress effects in dairy cattle may be limited due to adverse association with economic indicators. Dikmen et al. (Reference Dikmen, Cole, Null and Hansen2012) found a significant genetic correlation between rectal temperature in Holstein cattle under heat stress conditions and traits important to productivity, including 305-day yields of milk, fat, and protein, productive life and net merit. Conversely, genetic correlations between rectal temperature and somatic cell count score and daughter pregnancy rate were significant and negative.

The divergence between European and African evolved cattle, Bos taurus, and the tropically adapted Bos indicus (zebu) of Indian origin, in terms of genetic traits, provides some insight into the potential for future breeding for climate change adaptation. Among the genetic adaptations in zebu cattle have been the acquisition of genes for heat tolerance. Zebu cattle are better able to regulate body temperature in response to heat stress due to a number of physiological and cellular level traits (Hansen, Reference Hansen2004; Gaughan et al., Reference Gaughan, Mader, Holt, Sullivan and Hahn2010b) which confer superior ability for regulation of body temperature during heat stress due to a combination of lower metabolic rates and increased capacity for heat loss. Tissue resistance to heat flow from the body core to the skin is lower for zebu cattle, while sweat glands are larger in comparison to European breeds. In addition, the hair coat in zebu cattle enhances conductive and convective heat loss and reduces absorption of solar radiation. At the cellular level, genetic adaptations mean that elevated temperature is less likely to result in preimplantation embryos being inhibited during development for zebu than are embryos from European breeds (Hansen, Reference Hansen2004).

Traditional cross-breeding has been used to introduce the zebu genotype into Bos taurus breeds to develop cattle for beef and dairy production in hot climates. Success has been limited due to co-introduction of other unfavourable genetic characteristics of zebu cattle, particularly those associated with low productivity. In addition, it is now becoming clear that the rate of change in climate is faster than the rate of genetic improvement possible using traditional selection methods (Gaughan et al., Reference Gaughan, Lacetera, Valtorta, Khalifa, Hahn and Mader2009; Hayes et al., Reference Hayes, Lewin and Goddard2013). It is uncertain whether the most effective approach would be to identify phenotypes in a heat-tolerant breed that have favourable production and market traits or to select heat-tolerant animals within breeds that meet market specifications (Gaughan et al., Reference Gaughan, Lacetera, Valtorta, Khalifa, Hahn and Mader2009). An alternative approach is to incorporate specific thermo-tolerance genes from zebu cattle into European breeds while avoiding undesirable genes. This approach relies on firstly being able to identify and map specific genes responsible for thermo-tolerance in zebu, and then to develop breeding strategies such as marker-assisted selection and transgenics to better utilise the zebu genotype for cattle production systems (Hansen, Reference Hansen2004).

The Senepol breed, and some other tropically adapted breeds of Bos taurus cattle with short smooth hair as well as their crosses with Holstein, Charolais and Angus animals can show similar heat tolerance to Bos indicus cattle. The heat tolerance of cattle with slick hair, a trait controlled by a major dominant-acting gene, appears to be mostly associated with coat type although sweating ability may also be enhanced (Olson et al., Reference Olson, Lucena, Chase and Hammond2003). Olson et al. (Reference Olson, Lucena, Chase and Hammond2003) found that slick haired crossbred calves of Carora (a slick-coated breed) and Holstein origin had lower rectal temperature (0.18°C to 0.4°C) than equivalent normal haired animals and these crossbred cows had a production advantage over a full lactation of 810 kg milk. However, the advantage of the slick hair coat trait has been recently contested (Davis et al., Reference Davis, Spelman and Littlejohn2017), highlighting the challenges of genetics-based solutions in adaptation to climate change.

Adaptation to acute heat stress in French cattle

The impact of extreme temperature events on ruminant production systems can be seen in a study by Morignat et al. (Reference Morignat, Perrin, Gay, Vinard, Calavas and Henaux2014) that examined cattle mortality related to two severe heat waves in France. In August 2003, the average maximum temperature in France increased from the normal seasonal mean of 25°C to 37°C where it remained for more than 2 weeks. In 2006, temperature steadily increased from mid-June onwards, exceeding 30°C during most of July, making it the warmest July since 1950, although temperatures were less extreme than in 2003. Overall, cattle mortality increased by 24% in 2003 and by 12% in 2006 (Table 1). In 2003 the mortality occurred during the 2 weeks of the heat wave as well as during the following weeks, whereas in 2006 the mortality occurred during the 2^nd and 3^rd week of the heat wave with no carryover effects afterwards. In the severe conditions of 2003, all categories of cattle were affected with no difference in mortality between dairy and beef cattle. In contrast, in 2006 the mortality rate of mature animals was greater for dairy than beef, consistent with expectations that dairy cattle, especially older animals, are less able to cope with heat stress. Increased mortality in 2003 compared with 2006 undoubtedly reflects the greater intensity of the heat wave in 2003, but it may also reflect increased awareness of livestock farmers of the severe consequences of excess heat as a result of having experienced the 2003 heat wave. The study demonstrates the potential negative consequences of climate variability for the cattle industries and the benefits of increased producer awareness of preventative management practices to help mitigate the extreme effects of heat stress.

Table 1 Increased cattle mortality (%) during the 2003 and 2006 heat waves in France (Morignat et al., Reference Morignat, Perrin, Gay, Vinard, Calavas and Henaux2014)

Adaptation to extreme heat wave conditions in Australian dairy cattle

Garner et al. (Reference Garner, Douglas, Williams, Wales, Marett, DiGiacomo, Leury and Hayes2017) found that cows exposed to moderate heat stress in a controlled environment-chamber had a milk yield reduction of 53%. However, the fact that this level of impact is seldom observed at an industry level illustrates that operational management adjustments by dairy farmers, in response to experienced climate, over time increases their resilience to climate ‘shocks’.

Milk tanker pick-up data from dairy farms distributed across the state of Victoria, throughout one of the most extreme heat wave events in Australia’s history that occurred from 13 to 18 January 2014, were analysed. The event ranked along with the 2009, 1939 and 1908 heat waves as one of the most significant multi-day heat waves on record and was the hottest 4-day period on record in Victoria, for both maximum and daily mean temperature (Australian Bureau of Meteorology, 2014).

Figure 1 shows the milk tanker pickups from the 640 dairy farms spread across the dairy regions of Victoria, collected every 2^nd day, together with the state-wide average maximum and minimum daily temperatures and the calculated Temperature Humidity Index (THI) (Dunshea et al., Reference Dunshea, Leury, Fahri, DiGiacomo, Hung, Chauhan and Gaughan2013). Notable in the data is that the THI exceeded 72, the threshold for onset of mild heat stress, for 5 days. Additionally, night-time minimum temperatures remained above 21°C for three nights in a row. Relative humidity for the days above THI 72 was below 33%, averaging 61% for the 4 days following the event.

Figure 1 Daily milk tanker volumes (litres of milk/2-day pickup) from 640 dairy farms across the state of Victoria, Australia, shown with state-wide average maximum and minimum temperatures and calculated Temperature Humidity Index (THI). Dairy breeds on these farms included Holstein-Friesian, Jersey and crosses.

While the heat wave impact is clear in the milk production of these dairy cows, the average drop in milk production across the state was only 10% below the average 14-day running mean for the 5 days before the event. Compared with the 53% potential reduction reported by Garner et al. (Reference Garner, Douglas, Williams, Wales, Marett, DiGiacomo, Leury and Hayes2017) for cows in chambers (50% to 60% relative humidity and constant wind speed of 7.8 m³/min), the much lower drop in milk production could demonstrate the potential efficacy of on-farm management intervention, through farmers proactively managing heat stress as per the Australian Cool-Cows (www.coolcows.com.au/) guidelines.

Adaptation options for sheep and goat farmers in Ethiopia

Developing countries are highly vulnerable to climate change since their economy predominantly relies on rain-fed agriculture where production is dependent on weather and other natural factors. In arid and semi-arid regions of Africa, livestock species, particularly sheep and goats are typically a major source of livelihood and survival. They are a source of food for rural people but, additionally, they may be converted to cash to meet household financial needs. In these traditional farming systems a low level of technology adoption increases vulnerability to impacts of climate change and limits options available to develop effective adaptation strategies.

Feleke et al. (Reference Feleke, Berhe, Gebru and Hoag2016) investigated choices being made by sheep and goat farmers in northern Ethiopia to adapt to changes in climate causing thermal, nutritional and water stresses. Of the farmers surveyed, almost 99% had noticed changes in climate with most respondents seeing a rise in temperature and a marked decline in rainfall. Farmers were implementing multiple adaptation responses (Table 2) depending on their circumstances and resources. During extreme events such as long dry periods when feed and water were deficient, the most common strategy was marketing. Despite poor prices, over 96% of farmers surveyed chose to sell their sheep and goats due to poor natural grazing and feed shortages. Home feeding was practised by about 90% of farmers. Practices more common in developed countries during temperature extremes such as provision of shelter or outside shade trees were less used in Ethiopia. Costs and forward planning for establishing trees in previously cleared grasslands and costs related to building housing and providing bedding will likely remain a barrier into the future. In the case of goat farmers, low shade provision may also be related to the relative heat tolerance of these animals.

Table 2 Adaptation options used by sheep and goat farmers surveyed by Feleke et al. (Reference Feleke, Berhe, Gebru and Hoag2016)

¹ Percentages are rounded

Analysis showed that access to information, farming experience, number of households in a village, distance to main market, income and agro-ecological settings influenced farmers’ adaptation choices to climate change. Where adaptation strategies were implemented there was a positive influence on the household income. More effective options such as using crossbred animals were correlated with higher income. Feleke et al. (Reference Feleke, Berhe, Gebru and Hoag2016) propose the development of education and early climate warning systems for local farmers in regions such as northern Ethiopia, but note that support would still be needed to enable adoption of proactive adaptation approaches suited to different agro-climatic conditions. For example, government policy positions and initial support for low-technology regional programs such as planting shade trees and drought-tolerant browse species in grasslands may help to build climate resilience at relatively low cost.