Managing predators on livestock producing properties in South Africa and Western Australia – producer perspectives

Introduction

Predation of livestock is a worldwide phenomenon that results in emotional and financial strain on livestock producers (Knowlton et al. 1999; Allen and West 2013; Carruthers and Nattrass 2018). In the Republic of South Africa (RSA), there is a complex assemblage of native predators, many of which are thought to be associated with stock losses (Balfour and Kerley 2018). Control of some of these predators is permitted when they cause damage to livestock (e.g. black-backed jackals, Lupulella mesomelas; South African Government 2004), but such control must be implemented in a way that minimises suffering to the animals (South African Government 1962). In contrast, Western Australia’s (WA) predator assemblage is largely composed of non-native species, including wild dogs (a collective term for dingoes, free roaming dogs and their hybrids; Canis familiaris; Australian Mammalogy Taxonomy Consortium 2024) and European red foxes (Vulpes vulpes), that prey on small stock (e.g. sheep, Ovis aries) regularly (Allen 2014) and cattle (Bos taurus; wild dog predation only). In WA, both predators are a declared pest under the Biosecurity and Agriculture Management Act 2007, enabling management to occur (State Government of Western Australia 2007). However, dingoes are considered a native species under the Biodiversity Conservation Act 2016 (State Government of Western Australia 2016), although exemptions exist for management to occur where agriculture is impacted. Wild dogs are the most contentious species from a management perspective, as there are many groups and individuals who value the dingo for cultural reasons.

Landowners in RSA and WA combat domestic livestock predation using a range of tools. In RSA, the use of poisons is illegal but fencing and corralling of livestock and trapping and shooting of predators are common control tools (Du Plessis et al. 2018). In contrast, in WA, there is a heavy reliance on 1080 (or sodium monofluoroactate (NaFC₂H₂O₂), a naturally occurring toxin found in the native vegetation) in baits to control wild dogs and foxes. In WA, the use of 1080 is especially effective as native Australian species have a high tolerance to the poison (due to coevolution) and, consequently, it is largely target-specific to introduced invasive species (Kreplins et al. 2018). These management tools and the application thereof are a consequence of differential approaches to livestock management, legislative regulations, and social licence in relation to the control of predators.

Predation results in many impacts to producers that could be collectively lumped into two groups: financial and emotional. Financial costs associated with predators, from livestock producers’ point of view, include predator control costs and reduction in stock production due to predation (Minnie et al. 2015; M.S. Kennedy and T.L. Kreplins, unpubl. data). Other variables that impact the viability of producing livestock include fodder availability to the livestock (i.e. competition for grazing or hand feeding livestock), fluctuations in markets prices for livestock and other causes of livestock mortality (e.g. disease) (Binks et al. 2015; Ecker et al. 2015).

Agriculture in RSA and WA is integral to each region’s history and contributes substantially to economies of both countries. Not only does farming engender emotional and community values, but it contributes to the economy by providing direct income streams, jobs and ensuring food security. Predation of livestock in an agricultural context takes an emotional toll on the producers (Binks et al. 2015; Ecker et al. 2017). Attitudes towards predator management can be linked to ‘social identity’ (Van Eeden et al. 2020), past experiences (Trebo et al. 2025) and a range of other emotions and beliefs (Perry et al. 2022). Knowledge of the predator itself also alters the community’s attitude towards predation events (Parker et al. 2014). Both financial and emotional impacts are important motivators of predator management. In this study, we focused on the perceptions of levels of predation experienced, mitigation of predation impacts and financial losses.

To gauge the perspectives of landholders, managers, and livestock producers in relation to livestock predation, we targeted livestock producers in RSA and WA. Individuals from rural localities were interviewed (online or in person) to understand a range of lived experiences, backgrounds and costs related to predator impacts and management. Differences in each country in relation to the following variables were recorded for this study:

Additionally, qualitative data on the producers’ experiences in relation to predation events were recorded for context.

Methods

A survey comprising 26 questions (Supplementary file 1) was disseminated online through Survey Monkey™ and via in-person interviews among WA and RSA livestock producers.

In WA, control of predators is carried out by organisations known as Recognised Biosecurity Groups (RBGs). This online survey was sent via email to the RBGs’ executive officers for further distribution to farmers and landholders. During a field trip to RSA in September 2023, we visited agricultural properties in person and, where the owners/managers were willing, we conducted the surveys in person (using the same questionnaire as that presented online to WA landowners).

Ethics approval to conduct both the online surveys and the in-person interviews was granted by the Human Ethics Committee at Murdoch University 2023/118.

The questionnaire (Supplementary file 1) comprised sections relating to the following aspects of farming enterprises and livestock production:

Producers in RSA estimated losses based on field observations of mortalities due to predation, and on the ratio of pregnant ewes to weaned lambs for a lambing (a very subjective measure of predation). WA producers seldom observed predation events that contributed to the estimation of livestock lost to predation. For cattle production on pastoral stations in WA, animals are often only seen at mustering time so these values are based on cattle value per head and suspected predation impacts. Furthermore, bite damage detected at abattoirs reduces the sale value of cattle and was included in the loss estimates by WA producers.

Mann–Whitney U tests were used to determine the differences in property size and stocking rate of all livestock, cattle alone and sheep alone, by RSA and WA producers. Expected values were calculated utilising the proportion of respondents that stocked livestock in general, cattle alone or sheep only, per 1000 ha. Pearson’s Chi-squared analysis was performed to determine if there was a difference in the estimated financial losses experienced by RSA and WA producers. All statistical tests were conducted using the statistical program R (R Core Team 2024). The outcomes of all statistical tests were evaluated at α = 0.05.

Results

Twenty-six respondents participated in the survey. Thirteen face-to-face surveys were conducted in RSA (September 2023) and a further 13 WA interviewees through the online questionnaire.

On average, RSA properties (x̅ = 3637 ha, range 890–7000 ha) were smaller than WA properties (x̅ = 155,555 ha, range 7000–454,000 ha) but property sizes were not significantly different (P = 0.29). The WA properties encompassed both agricultural farms and pastoral stations. Pastoral stations are typically far larger than the average WA or RSA farm.

Land managers on all but one of the RSA properties included cattle in their production mix and only two properties did not stock dual purpose (meat and wool) sheep. The two RSA properties that did not stock dual purpose sheep stocked sheep breeds that have been developed for either wool or meat production exclusively. In RSA, the average number of cattle stocked on each property was 333 head and the average number of dual-purpose sheep was 1845 head. There were two RSA properties that stocked two individual breeds of sheep for wool or meat production alone. These properties ran on average 667 wool producing sheep and 266 meat producing sheep. In WA, eight of the properties stocked cattle. Two properties stocked dual purpose sheep or meat sheep. Three properties in WA stocked wool sheep exclusively. The WA farms that stocked dual purpose sheep typically maintained the highest stock numbers, averaging 5100 head. Wool sheep and meat sheep were stocked at averages of 5100 and 1750 head, respectively. On average, the properties in WA stocked 800 head of cattle but the range was very large (from 100 to 7500 head) (Table 1).

In RSA, the mean stocking rate for sheep was 539 per 1000 ha and in WA, it was 107 per 1000 ha. The stocking rate for cattle in RSA and WA were 68 and 3 per 1000 ha respectively. There was no statistically significant difference in the two stocking rates.

Compared to WA, RSA has a very wide range of predators potentially present on agricultural properties (Fig. 1). South African respondents all indicated that black-backed jackals and caracals (Caracal caracal) were present on their properties. Other species reported as preying on livestock include warthog (Phacochoerus africanus), bushpig (Potamochoerus porcus), cape fox (Vulpes chama), African wild cat (Felis lybica), aardwolf (Proteles cristata; potentially scavenging not predation), hyaena (both brown, Hyaena brunnea; and spotted, Crocuta crocuta), feral dogs (Canis familaris), otter (Aonyx capensis), baboon (Papio ursinus), yellow mongoose (Cynictis penicillata), leopards (Panthera pardus) and crows (Corvus coronoides) (predator species presented in descending order of perceived frequency of occurrence on South African properties). In WA, 92% of land managers reported wild dogs being present, 77% reported feral cats (Felis catus), 54% reported foxes and 23% reported wedge-tailed eagles (Aquila audax).

Fig. 1.

The percentage of properties that reported predators as being present in Republic of South Africa and Western Australia, during the predator impact survey during 2023.

All RSA farmers reported that their stock production efforts were negatively impacted by predators. However, the producers reported differing predator levels (Fig. 2). The species that were reported as being responsible for the most losses were jackals and caracals (Fig. 1). Only two land managers of RSA properties reported that feral dogs had a severe impact on their stock.

Fig. 2.

The perceived level of impact of predators (nil, low, moderate, high and severe) on livestock production from each of the predators based on the predator impact survey in South Africa and Western Australia in 2023.

In WA, all predators were thought to have an impact on livestock production on the respondents’ properties. Low and moderate levels of impact on livestock production were reported for wild dogs, foxes, and feral cats. Some land managers also reported low impacts from eagles. No WA property owners reported high levels of impact from predators. Respondents from four properties rated impacts from wild dogs as severe and respondents from one property rated impacts from foxes as severe. Negligible (‘Nil’) impacts of foxes and cats were reported from two to six respondents respectively (Fig. 2).

Most RSA producers’ annual losses were estimated at > ZAR100,000 (46%), or between ZAR10,000 and ZAR100,000 (46%). A single RSA farmer reported losses between ZAR5000 and ZAR10,000. Most WA producers’ (46%) annual losses to predation events were estimated to range between A$500 and A$1000 (Fig. 3). In WA, two producers reported no losses and two reported A$100–500 losses. One producer reported between A$1000 and A$10,000 losses to predators (Fig. 3). The estimated financial losses incurred by RSA farmers was significantly higher (χ²₄ 17.21, P < 0.001) than those incurred by their WA counterparts (Fig. 3).

Fig. 3.

The estimated annual financial impacts of predators on livestock production for South African and Western Australian livestock producers based on responses to the predator impact survey in 2023.

Nine of 13 respondents from RSA derived their income from livestock production alone. Those not relying exclusively on livestock for income are professional hunters or derived additional income from cash crops, including a variety of grains and/or lucerne (Medicago sativa). In WA, the opposite is true, with only four of the 13 respondents solely relying on livestock production for their income. Carbon sequestration offsets and contracting businesses were income streams cited by those WA landowners not relying solely on livestock for their income.

A variety of control methods were used by stock farmers in RSA and WA (Fig. 4). Shooting is the most common form of predator control in both countries. Trapping with leg-hold traps is also common. Cage trapping is frequently used by RSA farmers (especially for felids), but not WA farmers. Fencing is utilised in both countries by producers to exclude predator species. Almost all the WA respondents use poison for predator control, whereas only four from RSA were willing to admit to using poison. Licensed pest management technicians or hunters were used by five respondents in each of the countries. Driven hunts were relatively common in RSA but not implemented in WA. Corralling of livestock was implemented on five farms in RSA. Guardian animals were used by three farmers in WA and only one in RSA. Canid pest ejectors were used by landowners on two RSA properties and one in WA. Non-lethal tools (i.e. visual and auditory deterrents, e.g. fox lights) were used by two RSA and one WA landowners. The use of hunting dogs was only reported for one property in RSA. The use of snares was not reported in either country (Fig. 4).

Fig. 4.

The number of respondents in South Africa and Western Australia who use a range of control tools, including poisons, leg-hold traps, cage traps, snares, shooting, fencing, corrals, guardian animals, canid pest ejectors (CPEs), non-lethal tools (e.g. light), hunter/licensed pest management technician (LPMTs), hunting dogs, and driven hunts. These data result from a predator impact survey implemented in 2023.

The percentage of time that each country’s farmers estimated that they spent on predator control differed. In RSA, farmers estimated that 12% (range 1–20%) of their time was spent on predator control, whereas WA farmers suggested that 21% (range 5–60%) of their time was spent on predator control. All farmers felt that if they stopped implementing predator control, livestock losses would increase and production would decline substantially. Only one WA farmer disagreed with the statement that the financial benefit of the ‘reduction in predation events (by implementing predator management) were outweighed by the actual costs of predator control’.

The perceptions of impact of stock mortality (through causes other than predation) differed substantially between countries (Fig. 5). Disease was seen as an important cause of stock mortality by 92% of RSA farmers, whereas only 46% of WA farmers indicated that disease caused some mortality; others suggested that disease did not impact production. Poor maternal care by ewes was thought to cause at least some mortalities in both countries. Hypo- and hyperthermia were perceived to cause at least some mortalities by RSA farmers, whereas it was thought to be unimportant by WA farmers. Approximately 33% of farmers in both RSA and WA felt that malnutrition contributed to some of the stock losses. The perceived impact of stock theft was substantially higher for RSA farmers than for WA farmers (Fig. 5).

Fig. 5.

Causes, other than predation, to which South African and Western Australian farmers attributed stock losses, based on data from the predator impact survey conducted in 2023.

Wild or feral herbivores are likely to compete with domestic stock for resources on farms both in RSA and WA (Fig. 6). In RSA, these competitors are primarily native herbivore species, whereas in WA, various birds, invasive mammals and native species compete with domestic stock for the available forage (Fig. 6).

Fig. 6.

The grazing species that compete with livestock for resources in South Africa and Western Australia, as noted by survey respondents to a predator impact survey in 2023.

Discussion

Predation of livestock is a worldwide phenomenon that causes financial and emotional pain to producers (Breck and Meier 2004; Ecker et al. 2015; Kerley et al. 2018). There is a long historical background to the predation of livestock in both RSA (Anthony et al. 2010) and WA (Thomson 1984). This study highlighted how producers report and observe these losses related to the predators that are present, predation experienced, and the control tools available to them. Here we discuss a range of factors associated with livestock farming in RSA and WA, as reported by livestock producers, such as how predation impacts domestic stock production and the historical background.

South African farmers have a wider range of predators on their properties than in WA, although all properties in this study felt impacts from predators. Historically, intensive predator management has been accepted as an integral aspect of small livestock management. In RSA, the predator management efforts by farmers are primarily focused on jackals, while in WA, the focus is on wild dogs. The range of perceived impacts from predation extended from ‘some’ (relatively low levels) to ‘severe’ in both regions. In RSA, there are a range of smaller predators that potentially injure or kill newborn or young lambs. In addition, baboons are known to injure and kill lactating ewes to access the milk in their udders (Kifle 2021). It is therefore not surprising that the perceived impacts on livestock from predators in RSA were generally higher than in WA.

Historically, bounties, trapping, hunting clubs, poison and jackal proof fences were utilised by RSA farmers to reduce jackal numbers. Over 9 years in the early 1900s, more than 300,000 jackal bounties were paid out by the South African government (Nattrass and Conradie 2015). In 1978, only jackals, caracals and vagrant dogs were listed as vermin, as opposed to a range of wildlife previously controlled such as lions, hyaenas and leopards (Nattrass and Conradie 2015). Currently in RSA, there is a reliance on shooting (usually at night with the aid of thermal and/or telescopic sites), cage trapping (for caracals), fencing and leg-hold trapping for jackals. In WA, these approaches are supplemented with the application of broad-spectrum deployment of 1080 poison. The poison 1080 is a naturally occurring secondary chemical found in the native Australian vegetation and, to which, native Australian species have a natural tolerance. However, it is highly toxic to non-native species, including wild dogs, feral cats and foxes (McIlroy et al. 1986; Kreplins et al. 2018; Moseby et al. 2021). Leg-hold trapping is also used for wild dogs in WA, and cage trapping is used for feral cats for conservation purposes. This demonstrates that lethal control is still the prevalent (and often the preferred) approach in both countries (Ecker et al. 2015; Thorn et al. 2015).

Predator control and management is an important part of any domestic livestock production enterprise. South African property managers and owners had a different approach to the timing of their control efforts to those implemented in WA. South African farmers stated that they were often reactive to losses (i.e. performing control only when an animal was predated on), rather than implementing proactive preventative control. This is likely due to financial constraints of the country (Thorn et al. 2015) and tentative control where high profile predator species were concerned (Anthony et al. 2010). A professional hunter was typically only called in on RSA farms when losses to jackal predation were already occurring. In WA, licensed pest management technicians perform control year-round (i.e. baiting, trapping and shooting) as a preventive measure on cattle and sheep production properties, as part of the Recognised Biosecurity Group’s (RBGs; landholder groups that work collaboratively to control a range of invasive species) activities. There are some RSA properties where workers are employed exclusively to implement preventive trapping, but this was only implemented on two properties visited during this study. Landowners from both countries anecdotally recorded that landscape scale collaborative control efforts resulted in fewer livestock losses overall. For example, when RSA producers in a local area implemented a range of management actions simultaneously, local eradication of black-backed jackals was achieved. Areas with absentee landholders found it trickier to achieve this outcome. In WA, this has been achieved for foxes in some areas (Thomson et al. 2000) but not for wild dogs. Alternative forms of control, including the use of guardian animals and non-lethal tools (e.g. fox light), were used by some landholders.

Anecdotally, individual jackals that have survived persecution for an extended period on farms have been christened ‘University’ or ‘Professor’ jackals by RSA farmers. These individuals are ‘wise’ to the control methods and are particularly difficult to eliminate. Different areas around RSA implement varying levels of jackal control. For example, it was reported by the participants that on the northern Free State farms, jackals were easier to remove, with up to 15 animals being culled in a single night. Whereas in the southern Free State, where control practices have been implemented for a longer period, hunters were lucky to remove one jackal a night. Learned aversion has been noted when long term control practices for predator management (i.e. baiting) go unchanged (Allsop et al. 2017; Kreplins et al. 2018) in Australia. Some of the RSA farmers interviewed during this study participated in aerial shooting of jackals. This practice was reported by the survey participants as typically very successful in the first year, but success declined in subsequent years. The respondents suggested that the jackals had possibly rapidly developed a learned aversion to the helicopter noise. This serves as a stark reminder to all those participating in predator management to not underestimate the predators’ intelligence and adaptability.

In addition to predation, a variety of other factors (e.g. hypo-/hyperthermia, malnutrition, poor maternal care, theft, and market prices) influenced stocking rate decisions and productivity of livestock farms in both countries. Severe cold reduced lamb survival in RSA (only one property recorded it in WA). Those lambs raised in barns or feedlots did better than those free ranging in paddocks, according to RSA farmers in this study. The barn and feedlot method (i.e. hand feeding livestock) are much more expensive initially but resulted in faster attainment of marketable mass by the livestock (Van Der Merwe et al. 2020). Poor maternal care, malnutrition, and theft were recorded as having an influence on livestock production in both areas. Disease was recorded as a production limitation more commonly in RSA (all properties but one) than in WA (seven properties). Market price and the cost of living has a strong influence on the profitability of any farming enterprise (Department of Agriculture, Land Reform and Rural Development 2022; Australian Bureau of Agricultural and Resource Economics and Sciences (ABARES) 2024). Western Australia is better positioned geographically for accessing international export markets and the WA meat commands a premium price relative to that produced in RSA (Deards et al. 2014). Unlike RSA, where disease outbreaks occur sporadically, Australian meat production is seldom (if ever) impacted by disease. Consequently, Australia is one of the few places from which the exporting of livestock or meat is safe (Australian Bureau of Agricultural and Resource Economics and Sciences (ABARES) 2024). However, the market prices for a sheep (A$150; ZAR1800 per head) or cattle (A$1000; ZAR12,000 per head) was similar in both South Africa and Australia. The cost of raising livestock (and the cost of living) was substantially lower in RSA (Nkonki-Mandleni et al. 2019) than Australia. Consequently, the profitability of livestock production may be higher in RSA than WA. However, this is likely to vary based on the scale and intensity of the farming enterprise in RSA.

Enterprise choice was reported by the participants as not only impacted by the predation levels but also the grazing availability and potential stocking rate (Nkonki-Mandleni et al. 2019). Farms were on average the same size in both countries and had similar stocking rates. However, WA had some very large stations with high stocking rates that are above average in size. The average RSA farm (3637 ha), as described by the survey participants, is typically divided into several 50–100 ha camps (paddocks). Each camp is grazed for 10–14 days (2 weeks) and rested for 12–16 weeks. In the Karoo, the shrubs are grazed until 50% of the above ground material remains and then the livestock are shifted to another camp. In some cases, the camps are sub-divided using electric fences to provide a lambing kraal (enclosure). Electrified fencing works very well for larger livestock production enterprises to combat predation, unless warthogs dig underneath the fences, providing access points for predators. In WA, farms range in size from 5000 to 10,000 ha, but pastoral stations are far larger and may cover 240,000 ha. Farms in WA are intensively managed, much like those in RSA, except for the use of electrified fencing for lambing camps. WA stations detailed by the survey participants have little management or fencing, and the cattle or sheep are left to roam throughout large paddocks (~30,000 ha), where they graze on native vegetation. The productivity of the WA vegetation on smaller farms is much lower than that of the farms in RSA due to the difference in annual rainfall. In the WA, Morawa, the annual rainfall for 2020 was 285 mm (Bureau of Meteorology 2025) and in Craddock, the Eastern Cape of RSA, the annual rainfall for 2020 was 422 mm (Tutiemp Network S. L. 2024). Grazing capacity is related to annual rainfall and vegetation production in both localities. The competition for grazing by wild or feral species was reported by farmers in both countries to impact the forage availability for livestock. In RSA, there are a wide range of indigenous species (game) that may compete with domestic stock for natural forage, however, these can also be used for economic advantage. In WA, those species competing with domestic livestock for grazing (i.e. donkeys or macropods) are not seen as additional income streams. It is very difficult to predict rainfall, future competition for grazing, the amount of food on offer for your current stocking load, and future production by the average farmer in each country.

Sheep are very vulnerable to predation by jackals or wild dogs, whereas cattle can withstand higher densities of predators (Allen 2014) and are less susceptible to predation. Landowners from both countries reported that sheep production returns higher profits than cattle production. Participants noted that the vegetation in the Karoo region of RSA (where this study was conducted) and in WA was better suited to sheep than cattle production. In RSA, the land managers of cattle properties confirmed that they carried out little control in comparison to those responsible for sheep production enterprises. This is in contrast with the situation in WA, where a lot of wild dog control occurs on cattle stations. However, it was noted by a sheep farmer amidst cattle properties that jackals impacted the survival of calves. Regardless, enterprise choice was strongly impacted by predation levels, and even if sheep were the more profitable choice, they could not withstand predation in either country.

There have been political and farming land ownership changes in both countries. Landowners typically retain fewer staff and stock fewer animals than in the past (Hall et al. 2017). Land tenure is fluid, with the number of absentee landholders increasing, with people purchasing weekend properties, and with the submission of claims or properties being returned to indigenous land tenure. Changes in enterprise from sheep to cattle production has meant that jackal (Nattrass and Conradie 2015) and wild dog (Allen 2014) control has been reduced, as these mid-sized predators typically have relatively little impact on cattle production. In the past, RSA farmers benefited from government subsidies and tax deductions (Anthony et al. 2010; Nattrass and Conradie 2015). This is not dissimilar to WA farmers, who enjoyed the benefits of the Agricultural Protection Board (Thomson 1984) and their activities to remove predators from their lands. Since the 1980s, support for predator control has decreased (Nattrass and Conradie 2018) in RSA, and in Australia, there is a strong push for increased implementation of non-lethal forms of predator control (Smith et al. 2021). Despite this, it is apparent from these data that lethal management of predators is still an integral component of small livestock production in both RSA and WA.

Supplementary material

Supplementary material is available online.