The IUCN/SSC Canid Specialist Group's
African Wild Dog Status Survey and Action Plan (1997)

Chapter 7
The Rôle of Captive Breeding and Reintroduction in Wild Dog Conservation

Rosie Woodroffe & Joshua R. Ginsberg

The reintroduction of animals raised in captivity has played an important rôle in the conservation and recovery of a number of species.

• Reintroduction of wild dogs is technically possible provided some of the animals released are wild-caught, and that the newly-established population receives adequate protection from persecution and disease. However:
• Reintroduction has limited value for wild dog conservation since
  • Suitable release sites are in short supply: few reserves are sufficiently large and well-protected to sustain viable wild dog populations.
  • Reintroduction is most needed in West and central Africa, but there are no wild dogs with appropriate local genotypes held in captivity, and no local populations large enough to provide a source of wild-caught animals.
  • Reintroduction could be considered in parts of East Africa, but there are few wild dogs of eastern origin held in captivity.
• In southern Africa, reintroduction could create a metapopulation of tiny wild dog populations held in a network of fenced reserves and managed intensively to maintain genetic diversity. This would increase wild dog numbers locally, but would be extremely expensive.
• Protection of existing populations remains a higher priority for Africa-wide wild dog conservation than any attempt at reintroduction.
• Captive wild dogs may still contribute to field conservation by providing subjects for research, and by increasing public awareness and sympathy for wild dogs, both in Africa and abroad.

• Can Wild Dogs Be Reintroduced Successfully?
  • Previous Attempts to Reintroduce Wild Dogs
  • Attempts to Reintroduce other Canid Species
  • What Lessons can we Learn from Previous Reintroduction Attempts?
• Are there Wild Dogs Availabe for Reintroduction?
• Are Suitable Sites Available for Wild Dog Reintrouction?
• What Role can Captive Populations Play in Wild Dog Conservation?
• Conclusions

Background

Captive breeding and reintroduction can play a number of rôles in conservation. First, captive animals may provide insurance against the extinction of species that are threatened in the wild, and be used to reinstate or augment wild populations. Second, captive breeding may serve to increase the world population of a species, providing a source of additional genetic variation that may be fed into wild populations. Third, captive animals may raise public awareness of the species' plight in both range states and donor countries, leading to greater sympathy for field conservation programmes and, in some cases, to financial support for them. Finally, animals held in captivity may be used for research aimed at better management of free-ranging populations.

The release of animals born in captivity has been used to reinstate populations of several species that had become extinct in the wild. For example, in North America wild populations of both black-footed ferrets (Mustela nigripes) and red wolves (Canis rufus) have been restored in this way (Phillips 1995; Seal et al. 1989). Reintroduction - using either wild-caught or captive-bred stock - has also allowed the re-establishment of species which have become locally extinct. For example, swift foxes (Vulpes velox) from the United States have recently been reintroduced to Canada, where they were extirpated in the 1930s (Carbyn et al. 1994). On a larger scale, the reintroduction and translocation of ungulates has formed an extremely important component of the South African National Parks system for many years (Novellie & Knight 1994).

In this chapter, we consider whether reintroduction represents a suitable management option for wild dogs. Although some species have been reintroduced successfully, many programmes fail (Beck et al. 1994), and captive breeding is always expensive (Balmford et al. 1995). Wildlife managers must therefore weigh up the probability of successfully establishing a viable free-ranging population against the costs involved - in some cases protection of the remaining wild populations may represent better value for money. Nevertheless, the reinstatement of wild dogs in areas where they have been extirpated, especially in West and central Africa, is an important goal in their conservation. We therefore consider whether reintroduction could help to attain this goal, and also discuss additional rôles that captive wild dogs might play in field conservation.

Can Wild Dogs be Reintroduced Successfully?

The suitability of reintroduction as a management option for wild dogs depends upon whether viable free-ranging populations can be established from reintroduced animals. Perhaps the best way of assessing this is to review the successes and failures of previous attempts at reintroduction, using both wild dogs and related species. In total, nine attempts have been made to reintroduce or translocate wild dogs, all of them in southern Africa. These attempts are summarized in Table 7.1, and described below.

Previous Attempts to Reintroduce Wild Dogs

1) Kalahari Gemsbok National Park, South Africa

In 1975, five wild dogs were translocated from the borders of the Kalahari Gemsbok National Park to the interior of the Park, after two members of their pack had been shot by livestock farmers outside (Frame & Fanshawe 1990). Wild dogs have never been common in this park, which is probably marginal habitat. The translocated pack soon split into two groups, and, within a few months, both pack fragments disappeared.

2) Etosha National Park, Namibia

In 1978, six wild dogs were introduced into Etosha National Park. At 22,270km², Etosha is large enough to sustain a population of wild dogs, and prey densities were considered sufficient (Scheepers & Venzke 1995). The reason for wild dogs' absence from Etosha remains unknown, although some sources suggest that they were never common there (See Chapter3). The dogs introduced in 1978 had been raised in captivity, and were released as yearlings. All six died within four months of their release, mostly from starvation and predation by lions (Scheepers & Venzke 1995).

In 1989, a second attempt was made to reintroduce wild dogs to Etosha. Five captive-born dogs were used: two adult females, two adult males, and one younger animal fostered to one of the pairs. The adults were fitted with contraceptive implants, so breeding would not have been possible until these were exhausted. On release, all five animals died, and were believed to have been killed by lions (Scheepers 1992).

A third attempt was made to reintroduce wild dogs to Etosha in 1990 (Scheepers & Venzke 1995). Eleven animals were involved, all of them bred in captivity from Namibian stock. An attempt was made to teach these animals to hunt before releasing them: live springbok were released into their holding pen. However, the dogs quickly learned to wait until the antelope killed themselves against the pen's perimeter fence. Once the dogs had been released, they were monitored closely and springbok were shot for them every other day if they had not fed. At first, the dogs' hunting attempts were ineffectual and it was five weeks before they made their first kill. When their prey migrated, the dogs did not follow, and had to be lured towards the herds by dragging a carcass ahead of them. By 16 weeks after release, the dogs' hunting skills had improved considerably. Unfortunately, the reintroduction attempt ended in failure. Ultimately, 6 of the 11 dogs were killed by lions, one disappeared, and the last four dogs died of rabies after killing and eating a rabid black-backed jackal (Scheepers & Venzke 1995).

3) Hluhluwe-Umfolozi Park, South Africa

In 1980-1, 22 wild dogs were introduced into the Hluhluwe-Umfolozi Park. Twenty of the dogs (9 females and 11 males) were raised in captivity, but two (one male and one female) were wild-caught adults (Maddock 1992). The 22 dogs were released in four groups between September 1980 and September 1981, with the wild pair being caught up again and re-released with the last group of captive-reared animals.

These releases represent the most successful attempt so far to reinstate a free-ranging wild dog population. Fifteen years later, there are still wild dogs in Hluhluwe-Umfolozi, and 8 litters were recorded there between 1982 and 1993 (Figure 7.1, Maddock 1992). However, the success of this introduction is qualified. At just 960km², Hluhluwe-Umfolozi is a small area that can never sustain a population of wild dogs that will be viable in the long term. Dogs leave the reserve to enter neighbouring ranches and farmland where, fortunately, they are rarely persecuted. Indeed, dogs are welcome on some of the game ranches to the north of Hluhluwe-Umfolozi, and have bred there. Despite this, the wild dog population has not grown and spread into neighbouring areas. While the population has persisted, its numbers have fluctuated considerably (between 3 and 30) and, despite fission and fusion of the existing pack, no new packs have formed. In the long term, extinction seems likely unless intensive management is implemented. Indeed, no pups have been born since 1993. This may be because all members of the population are now close relatives - a plan has therefore been put forward to replace the females with new stock unrelated to the males (Maddock 1996). Alternatively, disease might have contributed to breeding failure (J.van Heerden, pers. comm.). Whether the Hluhluwe-Umfolozi population is viable in the long term or not, the success of this reintroduction will provide extremely useful lessons for future attempts at reintroductions into other areas.

4) Matetsi Safari Area, Zimbabwe

In 1986, nine wild dogs were introduced into the Matetsi Safari Area. Matetsi is contiguous with the Hwange National Park, which sustains a relatively large wild dog population. Thus, this release would have augmented an existing population. The nine dogs - five males and four females - had been raised in captivity and were released at the age of 18 months (Childes 1988). On release, the pack split into two groups, one of four males, and the other of four females and one male. A month after release, the group of four males were starving and injured, apparently, by spotted hyaenas. Two members of the other group disappeared, leaving just three females. These females were in good condition and had been observed hunting successfully on at least two occasions. All of these animals were recaptured and translocated to the Kazuma Pans Forestry Area. There, two more males disappeared but the group was seen to kill a young kudu. The following day, however, the remaining five dogs appeared at the butchery of a livestock farm bordering the Matetsi Safari Area, where the owner of the farm shot them all.

5) Klaserie Game Reserve, South Africa

A group of eight captive-bred wild dogs were released into the Klaserie Game Reserve in 1991 (M. de Villiers, pers. comm.). Since wild dogs from the adjacent Kruger National Park also use Klaserie (Maddock & Mills 1994), this release would have supplemented an existing population. An attempt was made to teach these animals to hunt before their release, by keeping them in an enclosure where they were given gutted and skinned carcasses, then whole carcasses, and, finally, tranquillized impala. The dogs did hunt after their release, although with rather little success (M. de Villiers pers. comm.). In captivity, the two females in the group had competed over breeding and, once released, they split up with some males accompanying each. The pack re-joined within a few days - but after two weeks they moved out of the reserve onto neighbouring farmland. They were then re-captured to avoid conflict with local farmers.

6) Venetia Limpopo Nature Reserve, South Africa

In 1992, a group of 14 wild dogs were released into the Venetia Limpopo Nature Reserve, a private reserve of 350km² (van Heerden 1993). The animals released were a wild pack which had been captured in the Mthethomusha Game Reserve and translocated to Venetia (English et al. 1993). They were held in a large (>1km²) enclosure in Venetia, from which two pack members escaped - however, the pack re-formed after the remainder were released (van Heerden 1993). One dog was radio-collared, and subsequent monitoring showed that the pack was hunting successfully. They produced a litter of pups within five months of release. However, because warthogs had damaged the electric fence surrounding the reserve, the dogs were able to cross the boundaries and started to use neighbouring farmland. The pack was last seen ten months after their release and, seven months after that, several wild dog skeletons were found lying close to one another on a neighbouring farm - they had probably been poisoned (van Heerden 1993). The rest of the pack has not been seen since.

7) Madikwe Game Reserve, South Africa

The most recent attempt to reintroduce wild dogs was in Madikwe Game Reserve. Madikwe was chosen as a reintroduction site because, while relatively small (600km²), it is securely fenced with predator- and warthog-proof fencing, and prey are abundant (M.Hofmeyr pers. comm.). Lions, cheetah and spotted hyaenas were also reintroduced to Madikwe in the period 1994-5. As in Hluhluwe-Umfolozi, a combination of wild and captive-raised dogs were used: 3 adult males from Kruger National Park and 3 adult females (sisters) from De Wildt cheetah centre were introduced to a boma in Madikwe in February 1995. By March 1995, all the pack members were mating, although no pups were produced. In July 1995, the pack was released. Supplementary feeding was needed at first, but the pack made its first kill five days after release. Two weeks after release, one of the captive bred females was first seen to lead a chase, and by two months after release the pack was hunting daily, and the dogs no longer approached vehicles. The pack did, however, learn to chase prey into the fence, and continue to use this as a hunting technique. The pack now has a home range of 180km², and has made no attempt to escape. At the time of going to press, the pack contained six yearlings born after the release. Since five of these yearlings are females, plans are being considered to release a group of males to try to establish a second pack within the reserve (M.Hofmeyr pers. comm.).

Attempts to Reintroduce other Canid Species

Lessons about wild dog reintroduction can also be learned from attempts to reintroduce related species with similar ecological requirements and social organization. These attempts are summarized in Table7.2.

Grey Wolves

The grey wolf (Canis lupus) is, perhaps, the species most ecologically similar to the African wild dog. Like wild dogs, wolves hunt cooperatively, range over very large areas, and are frequently persecuted when they come into contact with man. As with wild dogs, these characteristics have hampered several attempts at reintroduction.

The first well-monitored attempt at reintroduction involved five captive-bred wolves released in Alaska (Henshaw et al. 1979). The wolves were given some access to small live prey before their release, but showed no aptitude for killing it. After release, they followed caribou several times but were hesitant in their hunting attempts and were never seen to catch live prey. Although the group split up, and three of the five were seen in association with wild wolves, all eventually approached humans in search of food. As a result four were shot, and the fifth returned to the breeding colony, some 280km from the release site, where she was recaptured.

Another reintroduction attempt involved four wolves translocated from Minnesota to Michigan in 1974 (Weise et al. 1979). One female - which may not have been a member of the same original pack as the others - left the group immediately upon release but remained within an area of approximately 900km². The rest of the group wandered over >4,000km² before settling in an area some 90km from the release site. All four wolves died within eight months of release: three were shot and the fourth was killed in a road traffic accident.

Attempts to relocate wolves within Minnesota have met with more success. A total of 107 wolves blamed for depredations upon livestock in Northern Minnesota were translocated to the Superior National Forest and Beltrami Island State Forest in the period 1975-8 (Fritts et al. 1985). Although many of the wolves were shot, or trapped and re-released by U.S. Fish & Wildlife Service control teams, overall the mortality of translocated wolves was no higher than that of wolves already resident in the area. Members of the same pack released together did not remain together after release, and all the wolves moved over very large areas. Furthermore, they tended to 'home': 9/32 wolves (28%) returned to within 10km of their original capture sites (Fritts et al. 1984). Wolves settled, on average, 87km from their sites of release, and animals translocated more than 64km did not return to their capture sites.

Red Wolves

The red wolf resembles the wild dog in that it has a complex social organization, although it takes slightly smaller prey. As a result of persecution and habitat loss, the red wolf was extirpated from the whole of its former range in the Eastern United States earlier this century. However, in 1987 the U.S. Fish & Wildlife Service started a reintroduction programme in the Alligator River National Wildlife Refuge in North Carolina (Phillips 1995). Between 1987 and 1995, a total of 63 captive-bred red wolves were released into the wild. At the beginning of 1995 the release site contained 42 red wolves, 36 of them born in the wild.

Like grey wolves, the red wolves moved over large areas immediately following their release. The U.S. Fish & Wildlife Service intervened when animals moved outside the intended reintroduction area, although 4 animals were still shot by local people. Eventually, the released wolves settled into home ranges of 50-100km², feeding upon deer, racoons and rabbits. Although expensive, intensive monitoring and intervention were considered crucial for the success of the reintroduction, since it avoided conflict between reintroduced wolves and local people, maintaining public support for the project. Following this success, local landowners are now allowing red wolves onto their private land (Phillips 1995).

Swift Foxes

Weighing just 2.3kg, and feeding almost entirely upon rodents, the swift fox might seem to have little in common with the African wild dog. Efforts to reintroduce swift foxes to Canada do, however, provide lessons for wild dog reintroduction. Swift fox releases used a combination of wild-caught and captive-bred stock. As for grey wolves and wild dogs, reintroduction was much more successful when wild-caught animals were used: 32% (6/19) of wild caught foxes bred after release, while 108 foxes reared in captivity produced just 6 breeders after release (6% of those released, Carbyn 1995) This has led to a debate about the usefulness of captive-bred animals in the swift fox reintroduction programme (Carbyn 1995; Smeeton 1995).

The reason for captive-reared foxes' low survival and breeding success is not certain. However, the major cause of mortality in reintroduced swift foxes was predation by coyotes: 34 of 89 foxes (38%) found dead at three release sites in Canada were known or suspected to have been killed by coyotes (Carbyn et al. 1994). Coyotes do not eat swift foxes, but they do compete with them for food. Thus the relationship between swift foxes and coyotes parallels that between African wild dogs and lions.

What Lessons can we Learn from Previous Reintroduction Attempts?

A number of patterns emerge from this survey of previous attempts to reintroduce wild dogs and other canids. These patterns point to important lessons for future reintroduction attempts.

1) In all cases, wild-caught animals survived better than captive-reared ones. There are two reasons for this. First, wild dogs and grey wolves reared in captivity lacked skill in hunting - skill which is essential for the capture of large, fast-moving and often well-armed prey. It is extremely difficult to provide animals with experience of live prey under captive conditions, where space may be limited and local laws (not to mention the zoo-going public) may be unsympathetic. Furthermore, since wild dogs quickly learn to use fences to kill their prey, providing live food may still not mimic conditions in the wild.

The second reason for the high mortality of captive-reared animals involves predation: wild dogs and swift foxes reared in captivity appear unaware of the threats posed by competing predators. It is difficult to imagine a technique whereby captive wild dogs intended for reintroduction could be instilled with a fear of lions and spotted hyaenas.

2) Despite attempts to minimize human contact, captive-bred wolves learned to associate human settlements with food, which brought them into conflict with people and led to their being killed. This may also have contributed to the failure of the attempt to release wild dogs in Zimbabwe.

3) Newly-released grey and red wolves wander over very large areas and may settle some distance from the release site - this may also have occurred with the wild dogs released into the Kalahari Gemsbok National Park. Such long-distance movements may bring newly-released animals into contact with humans, leading either to their persecution (as in Venetia) or to the need for recapture (as in Klaserie).

Translocated wolves tend to 'home' to their original capture sites, a problem that has also been encountered in attempts to reintroduce sea otters in California (Estes et al. 1993). The behaviour of released wolves resembles that of dispersing 'lone' wolves seeking mates and territories.

Groups released together tend to split up: this is also characteristic of wild dog releases. Both wandering and group splits may be reduced by keeping the dogs in an enclosure at the reintroduction site for some time before release. For example, at Madikwe wild-caught and captive-raised animals were 'introduced' in an enclosure.

4) Wild dogs released into Etosha died from rabies. This points, once again, to the importance of disease in wild dog conservation: adequate disease control is a crucial consideration for any attempt at reintroduction, especially in areas where disease is believed to have contributed to wild dogs' decline. The measures necessary are discussed in detail in Chapter6.

5) Even successful release programmes may involve high mortality. Twenty-two wild dogs were introduced to Hluhluwe/Umfolozi Park in 1980-1, but by early 1983 just 8 remained (the birth of a litter of 7 then raised the population to 15; Figure7.1). Thus, 14/22 (64%) reintroduced wild dogs died before the new pack bred (Maddock 1996). Similarly, 62% of reintroduced red wolves died before breeding (Phillips 1995), and 79% of swift foxes failed to survive a year after release (Carbyn et al. 1994). This high mortality is a common phenomenon in reintroduction programmes: for example, only 27/71 (38%) golden lion tamarins (Leontopithecus rosalia) and 9/49 (18%) black-footed ferrets survived initial reintroduction (Clark 1994; Kleiman et al. 1991). In each of these cases, most or all of the animals released were captive-bred - but even wild born swift foxes suffered 53% mortality in the first year after release (Carbyn et al. 1994). In contrast, none of the wild dogs introduced to Madikwe died - perhaps because the predator-proof fence protected them from some of the factors which killed wild dogs released elsewhere. Nevertheless, it seems likely that in most cases some mortality is unavoidable - the only solution may be to release more animals, over a longer period (Becket al. 1994).

All of these considerations indicate that future attempts to release wild dogs must use either wild-caught animals or a combination of wild-caught and captive-raised animals. Holding the animals together in a boma prior to release appears to help newly-introduced animals to form a cohesive pack, and might help to prevent the animals from wandering too far once released. Thus, wild dog reintroduction is technically possible if the animals released can be protected from persecution and disease. The measures needed for such protection are discussed in detail in Chapter6.

Are there Wild Dogs Available for Reintroduction?

No wild dog reintroduction attempted so far has established a viable population - thus none can be described as an unqualified success. Nevertheless, the discussion above suggests that wild dog reintroduction is technically possible. The success of any reintroduction programme would, however, depend upon the availability of animals for release.

Wild dogs released in an area should be of the appropriate local subspecies or genotype. Wild dogs from eastern and southern Africa are known to be genetically different, and those from West and central Africa may be different again (Chapter2). Such differences may be a result of random genetic drift, but variation could also be caused by natural selection (Wayne et al. 1994). This could create problems for reintroduction programmes: wild dogs of 'foreign' genotypes might not be adapted to local conditions at the release site.

This need to release animals with local genotypes means that very few wild dogs currently held in captivity are suitable for reintroduction. Reintroduction is most needed in West and central Africa, but there are no captive wild dogs representing these genotypes. Almost all of the world's captive wild dogs are of southern African origin; the only east African dogs in captivity at present are 25 animals captured as puppies in 1995 by the George Adamson Wildlife Preservation Trust to set up a captive breeding programme in the Mkomazi Game Reserve, Tanzania (Fitzjohn 1995).

The availability of captive dogs is, however, only one consideration - the examples discussed above indicate that successful reintroduction depends upon some wild-caught animals being used. It is crucial, however, that collecting wild dogs for translocation should not threaten the population from which they are taken. As discussed in Chapter5, wild dog populations inhabiting small areas are unlikely to be viable in the long term, and any reduction in their numbers could drive them closer to extinction. Thus, reintroduction would depend upon the existence of large, viable populations which could withstand being 'harvested' for animals to be translocated. For example Kruger National Park, together with the reserves that surround it, sustains a population of 350-400 wild dogs which has provided stock for reintroduction attempts elsewhere in South Africa. Selous Game Reserve might provide a source of wild dogs in East Africa. However, there is no obvious source population for wild dogs in West or central Africa. This is a serious barrier to any attempt to reintroduce wild dogs in these areas.

An additional source of wild dogs for translocation would be 'problem' animals in conflict with livestock farmers. As discussed in Chapter 6, wild dogs rarely take livestock. However, genuine problem animals do arise from time to time, and translocation may be one management option for them. For example, the Kenya Wildlife Service captured a group of wild dogs in Laikipia which had killed 137 merino ewes and lambs in a five month period - plans are under consideration to release these animals elsewhere (R.Kock, pers. comm.). However, problem animals should only be introduced to areas where they are unlikely to continue taking livestock: local support is a vital component of successful reintroduction programmes (Beck et al. 1994), which would be seriously compromised if reintroduced animals killed livestock on a regular basis. In practice, translocation will only rarely provide the best way of dealing with problem wild dogs (Chapter6).

Are Suitable Sites Available for Wild Dog Reintroduction?

If wild dogs were available for release, the success of a reintroduction programme would depend upon the availability of suitable reintroduction sites. In particular, the factors which led to the local population's original decline must be removed - otherwise the introduced population is likely to succumb to the same pressures. As discussed in Chapter3, wild dogs' geographic range has contracted through a combination of habitat fragmentation and persecution. However, the immediate causes of local extinction are rarely known for particular areas.

Reintroduction programmes should proceed with caution if the cause of wild dogs' local decline - or, indeed, whether such a decline has occurred - is not known. For example, plans have been put forward to release wild dogs into the Mkomazi Game Reserve, part of the Tsavo ecosystem which extends into Tanzania (Fitzjohn 1995). Extensive poaching and encroachment of livestock into Mkomazi have now been curbed as part of a well-organized programme of rehabilitation, and reintroduction of wild dogs was planned as part of this process. However, the very low density of wild dogs in Tsavo West, and the fact that dogs are least often seen in the southern part of the park which is contiguous with Mkomazi, raises questions about the suitability of Mkomazi as a reintroduction site. Wild dogs are relatively common in the Maasai steppe, some 100km from Mkomazi, where existing dens were dug up to obtain stock for the Mkomazi programme (Fitzjohn 1995). With wild dogs breeding so nearby, it is likely that recolonization of Mkomazi would have occurred naturally if the area represented suitable habitat.

Sites must therefore meet several criteria before they can be considered suitable for wild dog reintroduction.

Size of the Reintroduction Site

In Chapter 6 we established that the highest priority for wild dog conservation is to maintain large populations in extensive protected areas, which require little active management. Ideally, then, the best sites for wild dog reintroduction would be large protected areas, where viable populations could be established by reintroduction and then left to persist naturally. In practice, however, such sites are rare, especially in West and central Africa where there is most need for reintroduction.

If the reintroduction site is too small to sustain a wild dog population in the long term, intensive management is crucial. The maintenance of such small populations has a much lower priority for Africa-wide wild dog conservation than does the protection of larger populations more likely to be viable in the long term. Nevertheless, establishing a network of several small populations, managed together as a metapopulation, would be valuable where a need was seen to increase the numbers of wild dogs in a particular range state or area, and where no larger reserve was available as a release site. It must be emphasized, however, that such metapopulations would require intensive management in the form of fencing, disease control and periodic movement of animals between reserves to maintain genetic diversity (Chapter 6). For this reason, it would be extremely expensive to establish and maintain wild dogs in a network of small reserves. Investing in better protection of existing larger populations might well represent better value for money.

People in the Reintroduction Site

Since persecution represents a very serious threat to wild dogs, the release site must either contain very few human inhabitants, or local people must be unlikely to persecute wild dogs. It must be stressed that, in a survey of 145 attempted reintroductions worldwide, Beck et al. (1994) found that the one factor which most contributed to the success of any particular reintroduction programme was public support for the programme. With animals as wide-ranging and formerly unpopular as wild dogs, the importance of public support cannot be emphasized too strongly. As discussed in Chapter6, local peoples' hostility to wild dogs could be mitigated by a combination of local education, compensation, work on husbandry practices and legislation to control the use of poisons. In practice, the threat of persecution may be minimized by releasing wild dogs only inside protected areas.

Disease in the Reintroduction Site

Disease is known to have caused problems in several potential release sites. For example, an attempt to reintroduce wild dogs to Etosha ended in failure when the last few animals contracted rabies from a jackal (Scheepers & Venzke 1995). Plans have been considered to reintroduce wild dogs to the area of the Masai Mara where rabies is known to have killed wild dog packs in the past. In such circumstances, strategies for disease control would have to form an important component of any reintroduction programme. Recent vaccination programmes for domestic dogs may have ameliorated this threat, at least temporarily. Possible alternative strategies are discussed in detail in Chapter6.

Competitors in the Reintroduction Site

An ideal release site would have abundant prey but low densities of competing predators. Lions killed at least 11 wild dogs released in Etosha, and also represent an important cause of mortality in natural wild dog populations. The presence of lions and hyaenas would be likely to slow the growth of any new wild dog population established by reintroduction (Chapter 5). One option might be to attempt reintroduction on private land where lions and hyaenas have been eliminated (Chapter 6). However, in practice wild dog reintroduction is likely to represent a single component of programmes to rebuild guilds of large carnivores inside reserves - this was certainly the case in Madikwe and in Hluhluwe-Umfolozi. In such circumstances, wild dog reintroduction is more likely to succeed if the wild dogs are released and allowed to establish themselves before lions and hyaenas are introduced to the area.

Suitable Sites for Wild Dog Reintroduction

These observations allow us to suggest a small number of sites which might be suitable for wild dog reintroduction. Etosha National Park, Namibia, would be one possibility. Previous attempts to release wild dogs to Etosha have failed, causing the Namibian government to decide to focus on protecting its existing wild dog population rather than trying to establish another one (Scheepers 1992) - a decision that we strongly support. However, any future attempts might meet with more success. Using a combination of wild-caught and captive-reared animals should avoid the problems of captive-reared dogs' inabilty to hunt or defend themselves against larger predators. In addition, more work on rabies vaccination is likely to establish a safe and effective protocol for use on wild dogs (Chapter 8). At 22,270km², Etosha should be large enough to sustain a viable wild dog population, particularly if several packs could be released there.

Another possible site for reintroduction might be the Serengeti ecosystem, including the Serengeti National Park, the Masai Mara National Reserve, and surrounding lands. At 25,000km², the Serengeti ecosystem is large enough to sustain a viable wild dog population. However, unconfirmed sightings suggest that wild dogs are still present in some parts of the ecosystem (Appendix1). Reintroduction may not, therefore, be necessary. Since disease is known to have contributed to the demise of the study populations in 1989-91, provision for disease control would form a crucial element of any attempt to reintroduce wild dogs to this area.

Another alternative reintroduction site might be the area surrounding Lake Edward on the border between Uganda and the Democratic Republic of Congo (former Zaïre), including the Parc National des Virungas in Congo and the Queen Elizabeth National Park in Uganda. Together, these parks comprise an area of over 9,000km² - although not all of this is suitable habitat for wild dogs. High densities of Uganda kob (Kobus kob thomasi) provide abundant prey, but wild dogs became locally extinct in the 1960s. Most wildlife in Queen Elizabeth was decimated during the civil war in Uganda in the 1970s and 1980s, but, while hippo, elephant and buffalo populations are recovering very successfully, lions remain rare. Two factors argue against this area as a reintroduction site. First, a tarmac road passes through the northern part of Queen Elizabeth, representing a possible threat to wild dogs. Second, it is possible (but by no means certain) that lions' low population density results from persecution by local people living on the park borders. If this were the case, such persecution would also threaten wild dogs. This possibility would need to be investigated thoroughly before wild dog reintroduction could be considered.

Another possible reintroduction site would be the proposed trans-frontier Limpopo National Park. If formed, this park would join parts of the Northern Province of South Africa (including Venetia) with the Tuli Game Reserve in Botswana and conservation areas in Zimbabwe, to protect 6,000km² of habitat suitable for wild dogs (M.G.L. Mills pers. comm.).

Finally, a plan was formulated in 1989 to reintroduce wild dogs to Akagera National Park, Rwanda (J.Kalpers pers. comm.). At 2,800km², Akagera is probably too small to support a viable wild dog population and, indeed, its integrity is now under severe threat. Given the current political climate in Rwanda it is unlikely that this programme will be considered again in the near future.

What Rôle can Captive Populations Play in Wild Dog Conservation?

Captive-reared wild dogs are unlikely to survive long if they are released alone. Thus, future attempts at reintroduction must use at least some wild-caught animals. Nevertheless, captive animals can make important contributions to the conservation of wild populations, even if they are never released.

Perhaps the most important rôle that captive wild dogs can play is as the focus for research. Many possible management strategies for wild dogs are hampered by the need for better information. For example, research is urgently needed into the safety and efficacy of vaccines against diseases such as rabies and canine distemper (Chapter 8), and this research can only be carried out in captivity. Captive animals can also be used to perfect techniques for use on free-ranging animals, allowing protocols for immobilization and designs for equipment such as radio-collars to be tested in captivity before they are used in the field. Captive animals can also be used to refine techniques already in use: for example, the belly scores used to estimate food intake (Appendix 2) could be calibrated using feeding experiments with captive dogs.

Captive wild dogs can also play an extremely important rôle in raising public awareness - and even funds - in both range states and donor countries. For example, several zoos in the United States have 'adopted' reserves in developing countries, formulating education programmes aimed both at people living in and around the reserves, and at people in the U.S., as well as sponsoring technology transfer and raising funds for field conservation (Hutchins & Conway 1995). Captive animals form an integral part of such programmes. The rôle played by zoos in conservation education is a very important one which is often undervalued.

In some cases animals held in captivity have been used to increase genetic variation in wild populations by transferring individuals or gametes from captive populations into the wild (Olney et al. 1994). However, such interactive management would have limited value in wild dog conservation. Population viability analyses (Chapter 5) suggest that loss of genetic variability is unlikely to be an important cause of local extinctions in wild dogs. Furthermore, almost all of the wild dogs currently in captivity are of southern African origin, while the populations in West and central Africa are most in need of augmentation. In addition, genetic studies indicate that wild dogs usually avoid close inbreeding (Chapter 2). It has been suggested that the dogs in Hluhluwe-Umfolozi stopped breeding altogether when the only mates available were relatives (Maddock 1996). If this is the case, artificial insemination of females living in groups at risk of inbreeding would be unlikely to produce litters which would be raised by all group members.

Conclusions

Wild dogs held in captivity are useful to field conservation since they provide subjects for crucial research, and may contribute to public education and fund-raising. However, captive-bred wild dogs lack skills needed to survive in the wild, and can never be released without wild-caught animals to accompany them. Nevertheless, experience in South Africa indicates that wild dogs can survive release if at least some of the pack members are wild-caught. It should, therefore, be technically possible to re-establish wild dog populations by reintroduction if the animals could be protected from persecution and disease after their release.

In practice, however, successful reintroduction would be very difficult where it is most needed, in West and central Africa. Wild dogs of the appropriate local genotypes are not available for release, since there are no captive stocks and no wild populations large enough to provide a source. Furthermore, there are no suitable release sites known to us in these areas. In West and central Africa, then, the protection of the few remaining populations remains a much higher priority for wild dog conservation than any attempt to establish additional populations.

Elsewhere in Africa, reintroduction of wild dogs is also hampered by the availability of suitable release sites. There are very few reserves large enough to sustain viable wild dog populations although Etosha, Serengeti and the proposed cross-border Limpopo National Park are candidates. Reintroduction could, however, be used to establish a network of small sub-populations containing just one or two packs in fenced reserves and private land in southern Africa. While no such sub-population would be viable alone, as discussed in Chapter 6 each could be managed as part of a metapopulation. Such intensive management would be expensive - although, funds permitting, it would be valuable for bringing about local increases in wild dog numbers in highly fragmented landscapes.

To conclude, much of the technical knowledge needed to establish and manage wild dog populations by reintroduction has now been assembled. The usefulness of reintroduction is, however, limited by the availability of wild dogs with the appropriate local genotypes, and by the availability of suitable release sites. Overall, then, the protection of existing populations has a much higher priority for Africa-wide wild dog conservation than does any programme of captive breeding and reintroduction.

Return to Wild Dog Action Plan Table of Contents, AWD Species Account, CSG Publications or CSG Home Page.

© 1997 International Union for the Conservation of Nature and Natural Resources.