The global human population has dramatically increased in the past 50 years and will continue to do so for at least the next 30 years. Alongside improvements in income, this is leading to significant increases in the demand for livestock products. History shows that livestock production negatively impacts wildlife, some of this is direct through competition for resources, but much of it is associated with land use change and livestock management practices. This will continue unless we create the circumstances in which wildlife and livestock can co-exist, through farmers benefiting from wildlife on their land. The imperative lies, however, in finding ways to improve the efficiency of livestock production so that land is not converted to produce livestock products and remains for wildlife.
The human population of the globe currently sits at around 7.5 Billion and is likely to increase to 9 to 11 Billion in the next 30 years (United Nations, 2017). Much of this increase has happened in the developing world, and Asia and Africa are likely to be the major contributors to population growth by 2050 (United Nations, 2017). The growing population in the developing world is also getting richer, with Gross Domestic Product increasing in lower- and middle-income countries by 180% in the past 10 years (World Bank, 2017). As well as moving people out of poverty, this increase in wealth is associated with changing dietary habits, as people shift from a predominately grain-based diet to one that includes increasing amounts of meat, initially poultry but then pigs, and red meat from goats, sheep and cattle (Delgado et al., Reference Delgado, Rosegrant, Steinfeld, Ehui and Courbois1999; Alexandratos and Bruinsma, Reference Alexandratos and Bruinsma2012). Although grain is increasingly used to provide the feed required for these livestock, land use change is also associated with increased livestock production. For example, it is projected that the land area under permanent pasture in lower- and middle-income countries (particularly in South America and sub-Saharan Africa) will expand by ~320 Mha by 2030 (from 2010; Wirsenius et al., Reference Wirsenius, Azar and Berndes2010), with increases in livestock numbers of between 40% and 50% for buffalo, Bubalus bubalis, and cattle, Bos taurus, and 30% to 45% for sheep, Ovis aries, and goats, Capra aegagrus hircus (Bruinsma, Reference Bruinsma2003).
Livestock production now occupies over a quarter of the land surface area of the globe (Robinson et al., Reference Robinson, Wint, Conchedda, Van Boeckel, Ercoli, Palamara, Cinardi, D’Aietti, Hay and Gilbert2014). Obviously, conversion of land to support livestock production occurs at the expense of land that is currently occupied by forests and other native vegetation (Foley et al., Reference Foley, DeFries, Asner, Barford, Bonan, Carpenter, Chapin, Coe, Daily, Gibbs and Helkowski2005). This change in land use puts pressure on wildlife populations, through either direct effects (e.g. persecution) or indirect effects (e.g. land clearing, fencing). These effects (impacts) are likely to increase as the demand, for particularly, red meat increases. The extent to which they herald the decline in wildlife populations depends upon the nature of the interactions; as discussed below.
Livestock production, and wildlife species are part of a socio-ecological system (Biggs et al., Reference Biggs, Schlüter and Schoon2015), meaning that humans play a major role in determining the outcome of interactions between components of the system. Thus, a focus on the interactions between livestock and wildlife can miss the perceptions and behaviours of, for example, farmers or wildlife managers, who will determine management actions that impact on those interactions. As researchers, we cannot, therefore, approach the issue of the future of wildlife in livestock production dominated landscapes through reductionism-based research; we have to see it within the context of the whole socio-ecological system.
In this review, I will first describe these livestock–wildlife interactions, provide evidence for their effects on wildlife and vice versa, provide a narrative for the extent to which future management interventions can ameliorate these interactions and highlight key areas for future research. I will focus on the interactions between livestock and wildlife that are mediated through nutrition, but to understand the story for the 21st Century, I will draw more broadly on the complexity of interactions within the socio-ecological system within which livestock–wildlife interactions occur. The future of many of our wildlife species is intimately connected to the future of livestock production; I will chart a future that is positive in a world where negative stories dominate.
Before we get into the meat of the review, it is important to clarify the boundaries within which the narrative for the discussion unfolds. First, it is important to define what I mean by livestock and wildlife. In general terms, livestock means any species of animal that has been domesticated and which are kept in an agricultural context. This can be anything from chickens, Gallus gallus domesticus, and guinea pigs, Cavia porcellus, to sheep and cattle. In this context, I will confine myself to domesticated large mammals that use herbage as their primary feed, for example camels, Camelus spp., horses, Equus caballus, goats, sheep and cattle. Second, wildlife is a catch-all term for species of any native, non-domesticated organism, including fungi, plants and animals. This definition is clearly too broad for the purposes of this review and, so, I will confine the discussion to native large mammal species that feed on vegetation, primarily ungulates and macropods.
Ecology is the study of organisms and the way that they interact with the world around them, the ecosystems in which they live. The ecosystems include other organisms of their own species and different species. The organisms can be food, they can be predators, they may be competitors or facilitators, or they may be, occasionally, neutral in terms of their effects on a particular organism or species. Food and predation clearly are direct interactions between organisms, but competition and facilitation generally occur through direct or indirect effects generally mediated, in the case of this review, by changes in vegetation abundance, composition, structure and quality.
Agricultural systems come in all shapes and sizes, from small holders who cut herbage to feed to individual animals in pens, through intensive systems, where the native vegetation has been converted to agronomic species, and feed and water is supplied, either all year or at certain times of forage and water constraints, to extensive systems where inputs are low, livestock roam over large areas feeding upon native vegetation and, to a large extent, fend for themselves. Agricultural systems also have other components that can affect wildlife, for example, humans tend their stock and fences constrain the movement of wildlife onto the farm.
The agricultural practice of livestock production is associated with a number of changes in the landscape, many of which can interact with wildlife. By their very nature these effects are complex and depend upon the circumstance, and the form, that livestock production takes. In the following three sections I will look specifically at the effects on the forage supply for wildlife and vice versa; I will then follow with a broader discussion of the effect of management interventions associated with livestock production on wildlife.
Interactions between livestock and wildlife
In a number of cases, where livestock and wildlife co-exist, research has concluded that there is little impact of livestock on wildlife and vice versa mediated through forage supply (i.e. neutral; Figure 1; see also Schieltz and Rubenstein, Reference Schieltz and Rubenstein2016). For example, a recent comprehensive analysis by Schieltz and Rubenstein (Reference Schieltz and Rubenstein2016) found that of the 155 studies analyses over 20% showed no effect of livestock on wildlife. This may be because livestock are at low densities in these studies and, therefore, do not affect the resources needed by the wildlife species, or there is niche partitioning between the livestock and wildlife species in question. Effective tests of whether the nutritional interactions between livestock and wildlife are really neutral require experiments in which the metabolic/weight densities of the livestock and wildlife are equal in situations where they occur individually or together (e.g. see Hester et al., Reference Hester, Gordon, Baillie and Tappin1999). These have rarely been carried out (see below).
Although, neutral interactions are often recorded in the studies analysed by Schieltz and Rubenstein (Reference Schieltz and Rubenstein2016), it is highly unlikely that there are no direct or indirect positive or negative effects of wildlife on livestock and vice versa (probably the latter). This is because of the scale of the analysis is usually small, plot, paddock, farm. The true interactions within the socio-ecological system happen at a much larger scale. Take, for example, the fact that agriculture tends to displace wildlife from the most productive parts of the landscape (Foley et al., Reference Foley, DeFries, Asner, Barford, Bonan, Carpenter, Chapin, Coe, Daily, Gibbs and Helkowski2005); although there is evidence from Europe that agricultural desertification has resulted in expansion of wildlife populations (Gordon Reference Gordon2009; Acevedo et al., Reference Acevedo, Farfán, Márquez, Delibes-Mateos, Real and Vargas2011; a clear indication that livestock production does displace wildlife). With the expansion of livestock production, wildlife either moves to or is left in the marginal areas, where individual and population performance of non-adapted species is likely to be low relative to the areas of better soils/climate now occupied by livestock production (Newby, Reference Newby2014). This can also lead to increased competition between wildlife species that have moved into marginal areas and those that are more adapted to these areas. Even protected areas that are set aside for wildlife tend to be in marginal areas of a country (e.g. Izquierdo and Grau, Reference Izquierdo and Grau2009). As such, the interaction between livestock production, as an overall activity, and wildlife, is unlikely to be neutral except in the short time scales of local observation.
For direct (interference) competition to take place between livestock and wildlife they have to share a common set of resources, in space and time, that are limiting and the effects are evident through the growth, reproduction and population dynamics of one or both species (negative; Figure 1 and Wiens, Reference Wiens1989). Indirect (exploitative) competition can occur where there is a spatial and temporal separation of the use of common resources that become limiting to one, or both species, in a certain place, at some time/s of the year. Again, for a concrete evidence of direct competition, there should be negative impacts on growth, reproduction and population dynamics of one or both species. Below I will highlight two areas of competition, space and forage.
Livestock production is generally seen as having a negative impact on wildlife (e.g. Western, Reference Western1989; du Toit and Cumming, Reference du Toit and Cumming1999). Again, in the analysis by Schieltz and Rubenstein (Reference Schieltz and Rubenstein2016) over 55% of studies showed negative effects of livestock on wildlife. Many of these effects are caused by direct interference competition whereby wildlife species left an area when livestock, particularly cattle, were present (e.g. Stewart et al., Reference Stewart, Bowyer, Kie, Cimon and Johnson2002). In Stewart et al. (Reference Stewart, Bowyer, Kie, Cimon and Johnson2002) the cattle were free ranging and humans were not present for much of the time that cattle used the range. Still elk, Cervus elaphus nelsoni, and mule deer, Odocoileus hemionus hemionus, avoided areas used by cattle within a 6-h window of analysis (short-term associative competition) and the 7-day window of analysis (longer term competition); whereas elk and mule deer avoided each other in the 6-hour window but not the 7-day window. This suggests it is something about cattle that the deer species were avoiding, and that both direct and indirect competition is taking place. The direct effect appears to be common across different continents and taxa (e.g. Schroeder et al., Reference Schroeder, Ovejero, Moreno, Gregorio, Taraborelli, Matteucci and Carmanchahi2013; Khadka and James, Reference Khadka and James2016) but the effect is confounded in many cases by the presence of humans (Western et al., Reference Western, Groom and Worden2009). It is not clear why wildlife avoid livestock even when humans are not present.
If we look at negative interactions that occur through the indirect competition for food resources, there is even less broad evidence. Theoretically, we would expect that species that occupy similar niches to livestock (grazers for cattle and sheep, and browsers for goats) would be most affected by the presence of livestock, with less effect on species that occupy different niches (Gordon and Illius, Reference Gordon, Hester and Festa-Bianchet1989). Merrill et al. (Reference Merrill, Teer and Wallmo1957) did indeed find that white-tailed deer, Odocoileus virginianus, predominately a browser, responded more negatively to the presence of goats than they did to cattle or sheep; also guanaco, Lama guanicoe, a specialist mixed feeder, avoided areas used by sheep and goats more than they did those used by cattle (Schroeder et al., Reference Schroeder, Ovejero, Moreno, Gregorio, Taraborelli, Matteucci and Carmanchahi2013).
We would also expect that smaller species would be able to outcompete larger species for shared resources when herbage is in short supply (Illius and Gordon, Reference Illius and Gordon1987). Du Toit and Olff (Reference Du Toit and Olff2014) conclude that indirect competitive dominance by smaller species of herbivores is the predominant mechanism that structures ungulate assemblages when shared resources become limiting. One would, therefore, expect sheep and goats to outcompete larger wildlife species that share their niche and that cattle would be more likely to suffer competition from smaller grazing/generalist wildlife species. There is not sufficient evidence to be able to test this hypothesis, however, the general conclusion is that cattle have a negative impact on many species (Schieltz and Rubenstein (Reference Schieltz and Rubenstein2016), even those smaller than themselves and occupying different niches. One possible explanation for this conclusion is that livestock biomass, predominately cattle, is generally high relative to expectations for wildlife herbivores because of the inputs from livestock management systems (Hempson et al., Reference Hempson, Archibald and Bond2015). This points to the fact that livestock production, as an activity, drives many of the interactions between livestock and wildlife (see below). In a broad-ranging survey of African systems, Fynn et al. (Reference Fynn, Augustine, Peel and de Garine-Wichatitsky2016) provide a set of criteria upon which competition would be expected to occur between livestock and wildlife. Most of these are predicted to result from competition for limited resources during the dry season, or in years with lower than average rainfall.
One other area that indicates competition for food between livestock and wildlife is niche or dietary breadth. A number of studies have found that there is a narrowing of the niche/diet breadth of wildlife in the presence of livestock suggesting that livestock are outcompeting wildlife in certain parts of the shared niche/forage resources. For example, in a recent study Traba et al. (Reference Traba, Iranzo, Carmona and Malo2017) found a narrowing of the diet in guanaco in the presence of domestic livestock (particularly sheep which occupy a similar niche). This suggests, at least an alteration, and more likely, a reduction in the nutrition of guanaco diets in the presence of domestic species. One would expect this effect to be less prevalent where there is greater niche dissimilarity between livestock and wildlife species (Fritz et al., Reference Fritz, De Garine-Wichatitsky and Letessier1996; Herfindal et al., Reference Herfindal, Lande, Solberg, Rolandsen, Roer and Wam2017).
The extent to which livestock interactions with their environment will have a negative effect on wildlife is often a matter of degree. In the majority of situations, involving livestock production, ecological interactions are overridden by the inputs that are associated with the production system itself (see below). These allow livestock to exist at significantly higher densities than if the inputs were not provided or the ecosystems are occupied by wildlife alone (Hempson et al., Reference Hempson, Archibald and Bond2015), leading to excessive grazing pressure being exerted on the vegetation (du Toit et al., Reference du Toit, Cross and Valeix2017). Where this occurs over a long period, there are reductions in vegetation abundance, complexity, diversity and cover, all of which reduce the forage available to wildlife species (Boone et al., Reference Boone, BurnSilver, Thornton, Worden and Galvin2005). These changes in the food resource will have impacts, not only on wildlife species that occupy similar niches to those of livestock but also those which are affected by, for example, changes in vegetation composition (Mishra et al., Reference Mishra, van Wieren, Ketner, Heitkönig and Prins2004; Ogutu et al., Reference Ogutu, Piepho, Reid, Rainy, Kruska, Worden, Meshack Nyabenge and Hobbs2010). A striking example occurs in semi-arid regions of the world where management practices of supply of artificial water have resulted in concentrated livestock around artificial water points. The congregation of livestock around water points can result in a grazing gradient (heavy defoliation gradient near water points; piosphere) (Boone et al., Reference Boone, BurnSilver, Thornton, Worden and Galvin2005) that has a negative impact on wildlife use of the landscape. Wildlife tend to be found further away from water points (possibly because they are also less water-dependent than, e.g. cattle) (de Leeuw et al., Reference de Leeuw, Waweru, Okello, Maloba, Nguru, Said, Aligula, Heitkönig and Reid2001); although browsing and grazing species may differ in their responses, with grazing species tending to be more closely associated with artificial water points than browsing species (Smit et al., Reference Smit, Grant and Devereux2007).
Very few studies have found wildlife to have negative effects on livestock; however, Odadi et al. (Reference Odadi, Karachi, Abdulrazak and Young2011) did find that wildlife reduced the food resource for cattle during the dry season when resource constraints were at their most severe, leading to a reduction in cattle performance. In Australia competition for forage by native grazers, that is, red kangaroos Macropus rufus and grey kangaroos M. giganteus and common wallaroos M. robustus, (Edwards et al., Reference Edwards, Croft and Dawson1996) is estimated to cost the livestock industry ~A$15 million per annum (McLeod, Reference McLeod2004). More than five million kangaroos and wallaroos are culled per year to reduce this conflict, despite the fact that it is not known the extent to which this actually reduces competition (Pople & McLeod, Reference Pople and McLeod2000).
Overall, the literature suggests that, where negative interactions occur, there are stronger negative impacts of livestock on wildlife than vice-a-versa and little evidence of two-way interactions (Figure 2a).
In ecology, positive interactions between species can be facultative (commensal) or synergistic (mutualistic) (positive; Figure 1). Facultative interactions are unidirectional, where the ecology of one species benefits another species. For example, egrets, Bubulcus ibis, follow cattle in pastures, feeding off the insects that are disturbed by the cattle’s foraging behaviour; the cattle gain no advantage from the egrets. In synergistic interactions, both species benefit from the presence of each other. For example, lichens are a composite made up of algae or bacteria living with a fungus; the fungus benefits from the carbohydrates produced by the algae or bacteria which themselves benefit from the protection afforded by the fungus. As with negative interactions positive interactions can be direct or indirect.
I can find no evidence for direct positive interactions, for example, where, say, livestock directly reduce the likelihood of predation on wildlife and vice versa which allows the utilisation of preferred resources where predation rate is normally high. However, domestic llamas, Lama glama, have been used to reduce predation by coyotes, Canis latrans, on domestic sheep (Franklin et al., Reference Franklin, Powell and Youngs1994). It would be interesting to see whether guanacos or other wildlife play the same role.
At lower levels of livestock density, than those that cause extensive ecosystem degradation (see above), livestock grazing can increase the structural heterogeneity, change species composition, reduce vegetation component that impedes intake and improves vegetation quality (review by Fynn et al., Reference Fynn, Augustine, Peel and de Garine-Wichatitsky2016). These changes can, in turn, benefit wildlife populations. The study of Gordon (Reference Gordon1988) was seminal in showing that the use of a high biomass, but poor quality, vegetation community during the Scottish winter, stimulated early vegetation regrowth in the spring, and, this regrowth, was used by red deer, Cervus elephus, in preference to other areas of that vegetation community. Through an analysis of areas of the Isle of Rum, which were used by cattle, and those that were not, Gordon (Reference Gordon1988) showed that red deer performance was higher where cattle were present, suggesting that cattle grazing facilitated the foraging of red deer leading to improved population performance. Since that study there have been numerous other examples published of the effect of facilitation by livestock, particularly cattle, to wildlife. Again, in the analysis by Schieltz and Rubenstein (Reference Schieltz and Rubenstein2016) over 20% of studies showed positive effects between livestock and wildlife. For example, across large areas of the USA, livestock grazing is used as a means of improving the habitat for elk (Krausman et al., Reference Krausman, Naugle, Frisina, Northrup, Bleich, Block, Wallace and Wright2009). The emphasis has been on using rotational grazing to enhance the quantity and quality of forage for elk leading to the improved elk population performance (Frisina and Keigley, Reference Frisina and Keigley2004). In Australia, it is thought that cattle grazing increases the ‘green pick’, that is regrowth, which is utilised by kangaroo species, Macropus spp, and has led to significant increases in kangaroo densities in areas of livestock production (Newsome, Reference Newsome1966 and Reference Newsome1971). In a broad-ranging survey of African systems, Fynn et al. (Reference Fynn, Augustine, Peel and de Garine-Wichatitsky2016) provide a set of criteria upon which facilitation would be expected to occur between livestock and wildlife. Most of these are predicted to result from removal of low-quality vegetation biomass and the facilitation of regrowth or increased vegetation heterogeneity. This would require significant management oversight of livestock stocking density and grazing distribution to achieve potential benefits for wildlife (see below).
There are rare examples where wildlife has been demonstrated to benefit livestock; for example, in a study across the dry and wet seasons in Kenya Odadi et al. (Reference Odadi, Karachi, Abdulrazak and Young2011) found that, by removing stem material from grasslands, zebra facilitated grazing by cattle during the wet season (see above for competition in the same system).
As far as I am aware, there are no studies that demonstrate the direct or indirect synergistic effects between livestock and wildlife.
Overall, the literature suggests that, where positive interactions occur, there are stronger positive impacts of livestock on wildlife than vice-a-versa and little evidence of two-way interactions (Figure 2b).
Effects of livestock management on wildlife
As well as the direct and indirect interactions of herbivorous livestock and wildlife species (see above), there are effects of livestock management itself that have an effect on wildlife species. These effects can be direct and indirect, deliberate or unintended (Figure 3). The extent to which management actions impact on wildlife will, in large part, depend upon the perceptions and distribution of livestock farmers. In general, small holders are less likely to perceive wildlife as a cost to their economic and welfare outcomes, as the wildlife are unlikely to directly affect agricultural practices and the supply of feed, unless there is competition for herbage at certain times of year, or there is the potential for wildlife to spread disease (Woodroffe et al., Reference Woodroffe, Thirgood and Rabinowitz2005). However, that does not mean that small holders do not impact upon wildlife. For example, Western et al. (Reference Western, Groom and Worden2009) found an almost 50% reduction in wildlife species in a semi-arid pastoral area in southern Kenya, that had been subdivided, relative to those areas which remained communal.
Generally, all farmers have the potential to be affected by the presence of wildlife on, or near, their farms. They often perceive wildlife to be a threat because of competition for fodder resources (see above) and the potential for the spread of disease from wildlife that affects livestock productivity (Knapp and Fernandez-Gimenez Reference Knapp and Fernandez-Gimenez2009).
Supplementation is used to support livestock production, in both intensive and extensive systems. This supplementation can come in many forms, from hay to protein-rich pellets and mineral licks. Although supplementation is focused on livestock, wildlife can take advantage of the materials supplied. For example, red deer and white-tailed deer used salt licks put out for cattle (Schultz and Johnson, Reference Schultz and Johnson1992; Payne et al., Reference Payne, Chappa, Hars, Dufour and Gilot-Fromont2016). Several deer species have also been reported to use feed-in troughs or feeders that were targeting feed for livestock (O’Brien et al., Reference O’Brien, Schmitt, Fitzgerald, Berry and Hickling2006); this is causing significant concerns about the potential for the transfer of disease at these locations (Palmer et al., Reference Palmer, Waters and Whipple2004).
Many farmers see wildlife as a competitor for forage with livestock and/or having the potential to spread disease (du Toit, Reference du Toit2011). Farmers will, therefore, directly persecute wildlife by killing or poisoning those that come on their land (Dunlap Reference Dunlap1988). Although persecution has been most common on predator species (e.g. Kolowski and Holekamp, Reference Kolowski and Holekamp2006), and those herbivore species that crop raid (e.g. Naughton-Treves, Reference Naughton-Treves1997), there are examples of persecution of herbivorous wildlife species because of their perceived impacts on forage availability for livestock (Prins, Reference Prins2000); however, the evidence for the effect of wildlife on livestock forage is slim (see above).
Both indirect and direct
In arid, semi-arid and seasonally dry parts of the world, farmers often supply artificial water to support livestock. The effect of supplementary water can have both a direct and indirect effect on wildlife. For example, these artificial water points can also support wildlife positively or negatively (see above). For example, in France wild boar, Sus scrofa, were frequent visitors to artificial water points during the summer (Payne et al., Reference Payne, Chappa, Hars, Dufour and Gilot-Fromont2016), as are red deer in Spain (Carrasco-Garcia et al., Reference Carrasco-Garcia, Barasona, Gortazar, Montoro, Sanchez-Vizcaino and Vicente2016). In central Australia, during the 1880s, the common wallaroo population expanded significantly with the introduction of water points to support the sheep industry and benefited from changes in vegetation composition associated with heavy grazing by sheep (Newsome, Reference Newsome1971).
Fencing has both a direct and indirect effects on wildlife. Land ownership is one of the fundamentals for most intensive and extensive farmers. In Africa, any areas of pastoral land have recently been subdivided (Western et al., Reference Western, Groom and Worden2009) allowing farmers to invest in the infrastructure and land ‘improvements’ that will benefit their enterprise of livestock production. One key component of infrastructure is fencing as this keeps livestock within the boundaries of the farm. As well as keeping livestock in, fencing can also keep certain species of wildlife out of the property. This reduces competition for herbage and the potential for the spread of a disease that may be resident in the wildlife population. The normal wire fencing that is put in place for domestic livestock deters not all species of wildlife, however, many are impacted (Gordon, Reference Gordon2009; Rey et al., Reference Rey, Novaro and Guichón2012).
Fencing can also be established at a much larger scale than the farm. For example, in Australia, the ‘dingo’ fence (5614 km) was constructed in the 1880s to keep dingoes from predating sheep in southern Queensland, New South Wales, South Australia and Victoria. The dingo fence is only one of many barrier fences that have been constructed in Australia to reduce the impact of wildlife on livestock. These barrier fences have been shown to be major disruptors of the movement of wildlife species such as emu, Dromaius novaehollandiae (Bradby et al., Reference Bradby, Fitzsimons, Del Marco, Driscoll, Ritchie, Lau, Bradshaw and Hobbs2014). There have been perverse outcomes of the fence, for example, the reduction in dingo density on the eastern side of the dingo fence has led to increases in the populations of kangaroos, Macropus spp., and rabbits, Oryctolagus cuniculus (Pople et al., Reference Pople, Grigg, Cairns, Beard and Alexander2000), that compete with sheep and cattle for herbage (Newsome, Reference Newsome1971). Major culling operations are now in place to try to reduce the kangaroo populations. In some areas, the reduction in dingo density has led to an increase in ‘meso-predators’, for example foxes and cats, which reduce rodent and rabbit densities. This leads to shrub encroachment which, again competes with grasses and reduce forage for sheep and cattle (Gordon et al., Reference Gordon, Squire and Prins2017). In Botswana, as a requirement of the European Union to stop the spread of foot-and-mouth disease, a large number of fences were established to keep wildlife away from major beef producing areas (van Oosterzee, Reference van Oosterzee2017). These fences have had a significant detrimental impact on wildlife population, particularly, migratory species (Mbaiwa and Mbaiwa, Reference Mbaiwa and Mbaiwa2006).
Agronomic species and fertilisers
In intensive livestock systems land is cultivated and fertilisers are added to improve the nutritional value and quantity of forage on offer to livestock. These management activities generally improve the grass type of herbage on offer but can involve the addition of legumes to improve the nitrogen content of the forage. The expectation is that these ‘pasture improvements’ would favour wildlife species with grazer or mixed feeder dietary profiles as compared with browsers. Many species of wildlife use agronomic pastures, at least for part of the year (Putman and Moore, Reference Putman and Moore1998), however, the roles that these play in the performance of wildlife populations is, generally, not known.
As domesticated livestock species are descended from species that are close relatives of wildlife species alive today, there are many diseases they have in common. Disease is probably still one of the major issues that will impact the interaction between livestock and wildlife across the globe (Hudson et al., Reference Hudson, Rizzoli, Grenfell, Heesterbeek and Dobson2002; Woodroffe et al., Reference Woodroffe, Thirgood and Rabinowitz2005). Although most livestock production includes significant interventions to reduce the impact of disease on animal growth and reproduction, wildlife can still act as reservoirs of those diseases (Bengis et al., Reference Bengis, Kock and Fischer2002), for example, Trypanosomyiasis is spread between wildlife and livestock by the tsetse fly (Glossina spp.) and the disease has constrained the opportunities for livestock production in certain areas of Africa (Alsan, Reference Alsan2015). Dramatic measures have been, and are, taken to reduce the contact between livestock and wildlife (e.g. persecution, fencing (see above)) or to reduce the chances of infection if contact occurs (e.g. vaccination). As well as the cases outlined above, in the Kruger National Park, African buffalo, Syncerus cafer, are managed to reduce the likelihood of them transferring the foot-and-mouth disease to cattle on farms that border the Park (Caron et al., Reference Caron, Cross and Du Toit2003). In Yellowstone National Park, USA, the highly endangered plains bison, Bison bison, which was on the verge of extinction because of hunting pressure, is persecuted outside the Park because of fears that it will spread brucellosis (Keiter Reference Keiter1997, Morris and McBeth, Reference Morris and McBeth2003; see also Alpine ibex in Europe, Hars et al., Reference Hars, Rautureau, Jaÿ, Game, Gauthier and Herbaux2013).
Some have questioned whether wildlife is actually such a risk to livestock, however, because of the perceived risks this is still the main reason for the persecution of wildlife by governments and farmers (Kock, Reference Kock2005).
Since the start of humans raising livestock, predators have been a major threat to livestock production in many parts of the world (Woodroffe and Redpath, Reference Woodroffe and Redpath2015). This has led to livestock farmers persecuting predators, which, in turn, has reduced depredation on wildlife prey species (see dingo fence above). In recent years, persecution of predators in the USA and Europe has been associated with increases in deer and other wildlife species density and distribution (Ward, Reference Ward2005; Harrington and Conover, Reference Harrington and Conover2007). On the savannas of East Africa, Bhola et al. (Reference Bhola, Ogutu, Piepho, Said, Reid, Hobbs and Olff2012) found that the grazing by livestock on ranches produced shorter grass sward which was likely to reduce the predation risk for young warthog, Phacochoerus africanus, and topi, Damaliscus korrigum, relative to that on reserves.
Greenhouse gas emissions
A longer-term impact of livestock production on wildlife will be greenhouse gas emissions. Land use change associated with the conversion of native pasture, and deforestation during the expansion of livestock production, results in greenhouse gas emissions (Gerber et al., Reference Gerber, Steinfeld, Henderson, Mottet, Opio, Dijkman, Falcucci and Tempio2013). Livestock production itself (enteric fermentation and manure) contributes about 20% of total greenhouse gas emissions annually (U.S. Environmental Protection Agency, 2006) and the quantity emitted grows year-on-year as livestock production increases (note that the emissions from livestock are in the region of four to 50 times higher than those from wild ruminant species (Pérez-Barbería, Reference Pérez-Barbería2017)).
These emissions contribute to climate change, which in turn can have positive or negative effects on wildlife distribution and populations (Glick et al., Reference Glick, Stein and Edelson2011; Fullman et al., Reference Fullman, Bunting, Kiker and Southworth2017). On the positive side, the rapid expansion of white-tailed deer into boreal forests in North America is primarily attributed to climate change (Dawe and Boutin, Reference Dawe and Boutin2016); interestingly, their expansion affects predator distributions (Latham et al., Reference Latham, Latham, McCutchen and Boutin2011; with knock on effects on livestock) and also the potential for zoonotic disease transfer to humans (Kilpatrick et al., Reference Kilpatrick, Labonte and Stafford2014). On the negative side, unusually high humidity and temperatures have been associated with the hemorrhagic septicemia caused mass mortality of Saiga antelope, Saiga tatarica tatarica, in Kazakhstan in 2015 (Kock et al., Reference Kock, Orynbayev, Robinson, Zuther, Singh, Beauvais, Morgan, Kerimbayev, Khomenko, Martineau and Rystaeva2018); this demonstrates the potential for the interaction between different indirect effects of livestock production on wildlife.
Implications for future management of livestock and wildlife
From the evidence provided above, the future for wildlife may appear bleak (Figure 3). There are currently three narratives about the ways in which agricultural production and nature conservation can be resolved at the landscape scale. The first has been called the ‘fortress conservation’, or ‘land sparing’, approach where biodiversity is protected within areas that exclude agricultural production and usually people (Adams, Reference Adams2004; Phalan et al., 2011). Agriculture takes place outside protected areas and needs to be intensified to meet the growing demand for food. Second, the ‘land sharing’ approach advocates the integration of agriculture and biodiversity, mainly where, ‘wildlife-friendly’ farming supports biodiversity whilst at the same time meet demands for agricultural products. Finally, the ‘win-win’ approach sees biodiversity as providing ecosystem services (e.g. pollination and pest control) to agricultural and agriculture supports biodiversity outcomes (Gordon et al., Reference Gordon, Squire and Prins2017). Although these are seen as alternative approaches, in effect, they can occur in different parts of the landscape. Gordon et al. (Reference Gordon, Eldridge, Ripple, Crowther, Moore and Letnic2017) highlight the pros and cons of each approach to support wildlife. Below, I will focus on the land sharing and win-win approaches but first, a comment on food production to meet the growing demand for livestock products.
Past evidence demonstrates that most wildlife is extirpated in areas of agricultural expansion (see above). Therefore, an imperative for the future for wildlife is to limit the expansion on land that comes under agricultural production. There are three key areas of agricultural activity and the food supply chain that need to be addressed in order to limit the likely future impact of livestock production on wildlife. All of these focus on reducing or at worst halting the increase, of the land under livestock (and more broadly, agricultural production) to support the demand for livestock products from a growing, wealthier human population. First, intensification of livestock production (more product from the same land area) needs to occur, where it already exists, (Thornton, Reference Thornton2010; Alkemade et al., Reference Alkemade, Reid, van den Berg, de Leeuw and Jeuken2013). There are, however, risks associated with the intensification approach as the areas of high livestock productivity can draw in more people, thereby, increasing the demand for land (e.g. Vale, Reference Vale2014 for the Amazon). Second, a need to reduce the waste of agricultural and food products across the whole supply chain, from farm to waste bin; at present about 30% to 50% of food produced is wasted (Gustavsson et al., Reference Gustavsson, Cederberg, Sonesson, van Otterdijk and Meybeck2011). Finally, the potential of reducing demand that changed eating habits of those in developed countries and/or diverting the growth of demand for meat products in developing countries into more efficient producers of protein than herbivorous livestock (e.g. aquaculture, see Food and Agriculture Organisation, 2014).
As described above, in many extensive livestock systems, where livestock relies on native vegetation for their food supply, inputs (e.g. supplementary feed) and infrastructure (water points) lead to heavy grazing pressure and homogenisation and degradation of the ecosystem. Ultimately, this will have negative consequences, not only for wildlife but also, for livestock production itself. Lighter stocking rates of domestic livestock tend to create great heterogeneity in vegetation structure over small to large spatial scales driving vegetation dynamics, biodiversity and potentially also improving wildlife habitats (Dumont et al., Reference Dumont, Rossignol, Loucougaray, Carrère, Chadoeuf, Fleurance, Bonis, Farruggia, Gaucherand, Ginane and Louault2012; Ren et al., Reference Ren, Han, Ohm, Schönbach, Gierus and Taube2015). Therefore, farmers should be looking to ways of reducing stocking rates and implementing management strategies that both maintain the economic return from livestock (and potentially wildlife) enterprise as well as maintaining vegetation productivity. There is evidence, from, for example, Australia, that reducing stocking densities in extensive systems can improve vegetation condition, thereby improving individual animal performance, reducing the need for inputs, for example supplementation of food, and increase the economic return and sustainability of the enterprise (Gordon and Nelson, Reference Gordon and Illius2007). In Europe, adopting rotational grazing management regimes appears to be able to maintain livestock productivity (generally at less intense grazing pressures) (e.g. Farruggia et al., Reference Farruggia, Dumont, Scohier, Leroy, Pradel and Garel2012; Ravetto Enri et al., Reference Ravetto Enri, Probo, Farruggia, Lanore, Blanchetete and Dumont2017). These reduced stocking rates will result in increased vegetation heterogeneity, providing opportunities for exploitation by a range of wildlife species, and may even be beneficial to livestock production (win-win outcomes; see also Gordon et al., Reference Gordon, Squire and Prins2017).
If we look at areas of the landscape which focus on wildlife management (conservation, hunting) purposes, livestock are often used to create heterogeneity in the landscape with benefits for wildlife that occupy a range of niches (Gordon and Duncan, Reference Gordon and Duncan1988; Fuhlendorf and Engle, Reference Fuhlendorf and Engle2001; Krausman et al., Reference Krausman, Naugle, Frisina, Northrup, Bleich, Block, Wallace and Wright2009). This is generally achieved through low to moderate stocking rates of livestock or rotational grazing where areas are rested from grazing for periods (Krausman et al., Reference Krausman, Naugle, Frisina, Northrup, Bleich, Block, Wallace and Wright2009). Although the impacts on livestock production are not usually measured, as indicated above, this can lead to improved livestock production (Gordon and Nelson, Reference Gordon and Illius2007) as well as benefiting the populations of wildlife.
Although history suggests that wildlife suffers where livestock production dominates, there are potential benefits to integrating livestock and wildlife, particularly in extensive livestock production systems. A combination of wildlife and livestock populations increases ecosystem resilience in the face of climate change and the threats from invasive species (Otieno and Muchapondwa Reference Otieno and Muchapondwa2016; du Toit et al., Reference du Toit, Cross and Valeix2017). For this approach to work there have to be economic benefits to farmers from maintaining wildlife on their land, which need to outweigh the benefits of increasing livestock stocking rates. There are a number of options that would promote conservation on land owned for livestock production, for example, conservation easements and leases, payments for ecosystem services, wildlife tourism, game ranching and hunting (Woodroffe et al., Reference Woodroffe, Thirgood and Rabinowitz2005; Gitahi and Fitzgerald, Reference Gitahi and Fitzgerald2011). As yet, there is a great deal of narrative about this approach from wildlife biologists (see du Toit et al., Reference du Toit, Cross and Valeix2017), however, there are only rare instances demonstrating the benefits of the co-management of wildlife and livestock (Otieno and Muchapondwa, Reference Otieno and Muchapondwa2016). Much more research into the potential and threats from co-management of wildlife and livestock is needed across the continents and the ecosystems on the planet.
Improve efficiency of livestock production
My starting point is a call for animal scientists to improve the efficiency of livestock production on land that is already under livestock production. This includes improving the use of all of the inputs into the enterprise, reducing the externalities of those inputs and increasing animal production, from growth through to disease management. These advances will limit the need for increasing the land area under livestock production to meet current and future needs for livestock products.
Co-management of livestock and wildlife
Although livestock–wildlife interactions have been the subject of a multitude of studies over the years; there is still a lot that is not understood. The nature of the potential link between livestock or wildlife feeding behaviour on vegetation quantity, composition, structure and quality and the consequential effects on other species of interest requires further experimental research. The majority of the evidence, to date, on livestock–wildlife interactions is observational; an experimental approach (e.g. Hester et al., Reference Hester, Gordon, Baillie and Tappin1999) is needed to truly test the hypotheses about whether the interactions are neutral, negative or positive. Only through these experimental tests will the research field truly move away from observation of pattern to understanding the mechanisms underlying the interaction (e.g. why do many wildlife species avoid livestock even when humans are not present). Also, there is the matter of scale, both in space and in demographics and production. Very few studies test the benefits of co-grazing at the landscape scale (Bhola et al., Reference Bhola, Ogutu, Piepho, Said, Reid, Hobbs and Olff2012); even fewer look at the potential performance improvements in livestock and wildlife (Gordon, Reference Gordon1988; Odadi et al., Reference Odadi, Karachi, Abdulrazak and Young2011).
More broadly, research is needed to provide guidance for ways in which livestock production can go hand-in-hand with wildlife management (e.g. conservation, economic benefits; Gordon et al., Reference Gordon and Nelson2004), within a socio-ecological systems context. This will include research on changing the long-held perception of many farmers that wildlife is a threat to their enterprise and livelihoods (Hulme and Murphree, Reference Hulme and Murphree2001; Knapp and Fernandez-Gimenez, Reference Knapp and Fernandez-Gimenez2009; du Toit, Reference du Toit2011). As stated above, there is extensive research on the ways in which wildlife can be integrated into livestock enterprises, particularly for eco-tourism enterprises (du Toit et al., Reference du Toit, Cross and Valeix2017). Much of this comes from wildlife biologists and sociologists; there is a requirement for more engagement, and potentially research, from the livestock and veterinary sciences part of the system. The potential for the incorporation of wildlife products (meat and hides) into supply chains in middle- and high-income countries requires further research (particularly market research) and policy development. In my country of residence, Australia, the local supermarkets sell kangaroo meat as low cholesterol, green, organic product (Spiegel and Wynn, Reference Spiegel and Wynn2014); the same happens with deer venison in Europe (Hoffman and Wiklund, Reference Hoffman and Wiklund2006). Meat, derived from wildlife species could form a valuable resource to support the growing demand for meat across the globe (Cawthorn and Hoffman, Reference Cawthorn and Hoffman2014). Much more research is needed on product and supply chain development and public attitudes towards no-traditional foods if this potential is to be realised.
As stated above, the disease is still one of the major issues affecting the potential for co-management of wildlife and livestock in many parts of the world. It not only dominates the interactions between wildlife and livestock, directly and indirectly but also drives the perceptions of wildlife by many farmers (de Garine-Wichatitsky et al., Reference de Garine-Wichatitsky, Miguel, Mukamuri, Garine-Wichatitsky, Wencelius, Pfukenyi and Caron2013). Although much is known about the diseases themselves management and government policy needs to be more nuanced if the positive benefits of wildlife on farms is to be realised; for example, managing disease in the wildlife population may reduce the need for disease interventions in livestock (Silk et al., Reference Silk, Croft, Delahay, Hodgson, Boots, Weber and McDonald2017). This is a key area for future research in wildlife/livestock interaction (Miller et al., Reference Miller, Farnsworth and Malmberg2013).
A truly socio-ecological systems approach to livestock–wildlife interactions and achieving win-win outcomes will require biophysical (e.g. ecologists, animal and veterinary scientists and agronomists) to work with the social and economic scientists. With the growing demand for livestock products, the future of wildlife depends upon a transdisciplinary approach to provide practical solutions for livestock farmers, wildlife managers and policy makers. The planet’s unique wildlife biodiversity needs our help, and quickly.
First, the author would like to thank the organisers of the International Symposium on the Nutrition of Herbivores 2018 for inviting me to present a plenary paper in the session on ‘Wild Herbivores’. The research, and ideas, presented in this review were developed whilst the author was employed at the Macaulay Land Use Research Institute, CSIRO, the James Hutton Institute and James Cook University. The author thanks Yvette Williams and two anonymous reviewers for their comments on an earlier version of this manuscript. The author would, however, state that the views presented in this paper are entirely on his own.
Declaration of interest
The authors declare no conflicts of interest.
This review did require any requirement for animal or human ethics.
Software and data repository resources
No primary data was collected to support this review.