Captive Breeding and Reintroduction

(from Foxes, Wolves, Jackals, and Dogs,
the IUCN/SSC Canid Specialist Group's 1990 Action Plan)

[Ed. - the 6 figures referred to in text to be added, hopefully soon!]

Introduction

Mammals have been bred in captivity since humans first domesticated wild animals approximately 10,000 years ago. Since this time, with few exceptions, most of the effort put into breeding mammals in captivity has been directed at the breeding of domestic mammals (cows, sheep, goats, and pigs) for food, wool, and leather. Yet, to this day, breeding of species which have not been previously domesticated continues.

There are several reasons to breed those species which, for most purposes, are still wild. One aim of such "captive" breeding is meat production (Skinner 1989) or the production of fur; another is to provide specimens for viewing in zoos or game parks without having to capture animals in the wild and further deplete natural populations. Captive populations of wild animals can also serve as latter-day arks.

A commonly stated aim of many captive breeding programmes conducted in zoological parks is to produce founder populations for release into the wild in reintroduction programmes (Anderson 1986; Seal 1986). In the following discussion, we acknowledge that wild animals are often bred in captivity for reasons other than releasing them back into the wild. However, our discussion will focus on the potential of captive breeding as a tool in the conservation of canids, in particular, how the breeding of canids in zoos might be managed with this aim as a priority. We briefly discuss the problems of reintroduction programmes; a rigorous discussion of the various difficulties encountered in reintroduction is badly needed, but cannot be accommodated in this report.

The use of captive animals as founders of reintroduced populations has a long history. Zoos (the New York Zoological Society and the American bison), private ranches (Kenyan ranches and the black rhino), and even cloisters (Père David's deer) can serve as arks. When extinction in the wild is imminent captive breeding offers a last option to save a species. However, captive breeding, as a conservation strategy, is a temporary measure. Even with intensive efforts to preserve genetic variation in a captive population (through the use of genetic screening and studbooks), the inevitable small size of captive populations will make it extremely difficult to preserve sufficient genetic variation such that evolution can occur (Lande 1988).

What constitutes a sufficiently large population for captive breeding depends on many factors, including global weather patterns, genetics, disease, governmental stability and commitment to conservation, and pure chance (references in Soulé 1987a; Lande 1988). We do not discuss these at length, but note that contrary to popular belief, factors other than genetics (e.g. political and economic considerations, habitat destruction, climatic change) may be the critical variables that will determine the fate of an endangered species in the next century.

Captive breeding with the aim of reintroduction is only plausible when the force driving extinction can be reversed. In principle, almost any such force is potentially reversible with biological knowledge, political will, and money. In practice, there are likely to be major difficulties. Nonetheless, potentially reversible threats of extinction include: epidemic disease threatening a sole surviving population, conversion of savanna habitat for agricultural use, extinction driven by active hunting or pest control programmes, or interbreeding with feral domestic dogs. Doubtless there are, in practice, many irreversible causes of extinction such as destruction of fragile and irreplaceable habitat where no suitable refuge exists (e.g. lowland rainforest). If habitat loss is the immediate agent of extinction, and the conservation of suitable habitat is impossible, bringing a species in from the wild may merely delay an inevitable extinction. This is not to say that such an effort is entirely worthless; times change, our understanding of ecology improves, and what appears to be an hopeless case today may not be so tomorrow. As Michael Soulé has said, "(t)here are no hopeless cases, only people without hope and expensive cases" (Soulé 1987b).

Goals and Source of Data

The following analyses aim to tackle three questions specifically with regards to canids:

1. Which species are being bred in captivity?
2. How successful are existing breeding programmes?
3. What accounts for variation in the success of captive breeding of different canid species?

The analyses are based on data extracted from captive breeding records in the International Zoo Yearbooks (IZY). We have included data from the years 1962 to 1984 inclusive, although for certain analyses some years may be excluded. Of course, not all captive breeding of canids is recorded in the IZY. For a species like the arctic fox, the great majority of captive breeding is accomplished on fur farms. The results of captive breeding for economic exploitation of a species have been known for over a century and were first elucidated by Charles Darwin (1859). The goal on a fur farm is to produce standardized, unmarked pelts. Particular traits (colour, size, fecundity) may be selected by the breeder, even if these traits are rare or non-existent in wild populations. In short, while those breeding canids in captivity for conservation aim to maximize genetic variation, the aims of fur breeders will, most probably, minimize variation. While acknowledging that fur breeders may make important contributions to improving breeding technology for rare and endangered species, the success or failure of fur breeders is essentially irrelevant to the following discussion.

The IZY data themselves are, of course, not perfect. Breeding reported in the IZY is listed by zoological collection, not by litter. In most circumstances, for a particular species, only one litter is bred in any particular zoological collection in any particular year. However, if several litters are bred in a single zoo in a single year, the data reported are the summed data of these litters. To simplify language, we refer to each report in the IZY as a "litter," recognizing that in a small number of cases the data are for several litters.

A second problem with IZY data is that, inevitably, there has been variation in reporting standards over the last two decades, both between and within zoos. These variations may distort the results we report and the conclusions we draw from them. Nonetheless, the IZY data do provide a foundation on which to build future research. Recognizing the problems with the IZY data, a similar analysis is planned for data using the ISIS (International Species Information System) database.

In our discussion we combine the information summarized in the Species Status Reports (Chapters 4-8) with the captive breeding data to devise criteria on which species might be judged as suitable for breeding in captivity. On the strength of these criteria, we make policy recommendations for future breeding projects.

Species Bred in Captivity

In the last 26 years, 32 of the 34 canid species have been bred in captivity. The two species for which there are no records of breeding in captivity are the Simien jackal and the Tibetan fox. Of the remaining 32 species bred in captivity, many have produced only a small number of litters (Fig. 1). Of the 32 species, 19 have been bred, on average, in fewer than four collections per year. The remaining 13 species account for the great majority of all captive breeding of canid species. Nearly 80% of all captive breeding can be attributed to the six species most commonly bred in captivity. In decreasing order of frequency, these species are: the grey wolf, the red fox, the raccoon dog, the golden jackal, the arctic fox, and the African wild dog (Fig. 2).

In the last two decades (1965 to 1984), a total number of approximately 4,000 canid litters has been reported in the IZY. There appears to have been a consistent effort to breed canids in captivity, with approximately 180-200 litters being reported world-wide from zoos each year (Fig. 2). Over the last two decades, one might expect that an increased interest in conservation would have led to an increased relative effort in breeding rare species. However, given that only one of the most commonly bred species is endangered (the African wild dog), this does not appear to be the case.

Success, Failure, and Breeding Canids

Someone, somewhere, has at one time or another, attempted to breed nearly every species of canid. However, that people have attempted to breed a species in captivity does not mean that all attempts have been equally successful. The question arises: how do you measure success? One could look at the number of surviving offspring, the size of litters bred in captivity versus those observed in the wild, or at some measure of how well pups survive in captivity.

In our analyses, we have equated breeding success with the percent of pups that survive in a litter. More precisely, we assume that when there is high mortality the attempts to breed a species should be considered less successful than when there is low mortality. We have used this measure for two reasons:

1. Intuitively, poor survival of pups indicates that captive breeding cannot be sustained over a long period of time.
2. The measure has been used in other studies to measure the success of captive breeding efforts (see Ralls et al. 1986, 1988).

To get an idea of what the average levels of mortality are in captively bred canids, we can lump all the data into one analysis. As can be seen in Figure 3, since the late 1960s, approximately 40% of the pups in each litter have died. This pattern is remarkably consistent. What might seem odd is that the earliest reports show much lower levels of mortality. In the first years of the IZY, reported mortality was extremely low, ranging from 0% (1962, 1963) to well under 30% (1965-1968).

Several explanations may account for this result. First of all, in earlier years, zoos may have bred only those species that survived well in captivity. Although this may seem a logical explanation, the data we discussed earlier in the chapter suggest it is not the correct explanation: zoos have bred essentially the same mix of species for the last two decades. A second, and more likely explanation is that there was greater inconsistency in reporting during the first few years in which records were collected. To minimize the effect of potential variations in reporting, we have excluded data from the earlier years (1962-1968) from the following analyses.

[Figure 2. Canid litters bred in captivity, 1962 to 1984. All species compared to those most commonly bred.]

Mean mortality across all species and across all years is approximately 41 %. The level of mortality does not differ significantly from year to year¹. There is, however, great variation in the average mortality for different species (Fig. 4). Species specific mortality ranges from a low of approximately 25% (swift fox, arctic fox) to a high of nearly 70% (bat-eared fox)².

What accounts for this variation in the litter mortality of captive bred canids? It appears that one important factor is the number of litters bred in captivity (Fig. 5). Practice in captive breeding, while not making things perfect, does seem to reduce mortality greatly. Pup mortality is lower for those species most often bred in captivity.³

Studying Figure 5, however, several further conclusions can be drawn. Several species do not fit this model; in statistical terms they are "outliers." One species, the swift fox, breeds well in captivity, despite few litters being reported. Perhaps, in this case, the great experience gained in breeding other closely related fox species (corsac fox, red fox) has made breeding the swift fox less difficult. Two species, the African wild dog and the fennec, continue to show high mortality despite having been bred in captivity relatively frequently. If these three "outlier" species are removed from the analysis, there is almost a linear decline in mortality with a logarithmic increase in the number of litters bred⁴

The word "logarithmic" may not be familiar to all readers. The log10 scale allows us to compare data where there is a great range of values, in this case from 40 to 1200. Each number is translated into a power of 10. For example, 10=10¹, 100=10², 100=10³. A number between 10 and 100 would have a log value between 1 and 2. Thirty-three, for instance, is equal to 10^1.52, or on the log scale, 1.52.

The use of the log scale may be convenient, both statistically and for viewing data. However, in biological and management terms, what the log scale indicates is that great increases in breeding effort are required to effect a small reduction in mortality. If 20 litters have been bred in captivity, to reduce litter mortality to any great extent we must breed not 20 more litters, but 200 more.

Many of the species most often bred in captivity are abundant in the wild (Fig. 1: red fox, arctic fox, raccoon dog, coyote) or have been abundant until very recently (grey wolf. The status of many species which are rarely bred in captivity is either Vulnerable/Endangered or Insufficiently Known (e.g Simien jackal, island grey fox, Bengal fox, Sechuran zorro). If species are grouped by their conservation status in the wild (see Species Summaries, Chapters 4-8), it is clear that those species most in danger of extinction tend to breed most poorly in captivity (Fig. 6)1. Species which are common in the wild have been bred most often in captivity. Hence, common species have much lower pup mortality when bred in captivity than do those which are classified as Vulnerable, Endangered, or Insufficiently Known.

Footnotes:

1. Effect of year on mortality (p. > 0. 15, 1 way ANOVA).
2. Effect of species on mortality (p. < 0.0001, 1 way ANOVA).
3. Regression of mean mortality per litter, by species, as a function of log₁₀ of litters reported in the IZY. For 18 species with greater than 40 litters bred in captivity, r² =0.25, p. < 0.05.
4. Regression of mean mortality per litter, by species, as a function of log₁₀ of litters reported in the IZY. Excluding three outlying species, for remaining 15 species r² =0.86, p. < 0.001.
5. The grey wolf was, until recently, common throughout much of its range. It is still common in Canada and the Soviet Union. However, given that many sub-populations are endangered, the species as a whole is classified as Vulnerable. We have excluded the wolf from this analysis because of its rapidly changing status. It is the one vulnerable or endangered species that breeds well in captivity, the exception that proves the rule.

[Figure 3. Mean mortality of canid pups in captively bred litters.]

Conclusions and Discussion

Clearly, if captive breeding of endangered species followed by reintroduction into the wild is to be a viable conservation option, we must be able to breed endangered species in captivity. The data we have presented range from hopeful to alarming. To be optimistic, most canids have, at one time or another, been bred in captivity. The apparent importance of experience in husbandry is encouraging insofar as a species that initially was difficult to breed, such as the maned wolf, can be bred successfully.

That the number of canid litters bred in captivity annually has remained more or less constant over the last 20 years is probably nothing more than an indication that the space, time, and money for captive breeding have remained limited. That the grand mean of litter mortality for all species has not changed significantly probably reflects the fact that we have continued to breed various canid species at approximately the same ratios (Fig. 2). In fact, an initial increase in overall mean mortality would be expected if zoos were to shift their emphasis to breeding "difficult" species, those which have not been bred extensively.

Our ability to breed successfully in captivity some species of canids is encouraging. However, accumulated experience in breeding canids does not translate into an ability to breed any particular canid. The data indicate that. experience must be garnered on a species by species, or perhaps genus by genus basis. There may be some exceptions; the data on the breeding success of the swift fox suggest that it may be easier to breed species with con-generic relatives which have frequently been bred in captivity. Perhaps for some rare species such as the island grey fox or the red wolf this will make captive breeding less difficult. However for many rare, vulnerable, or endangered canids there are no closely-related species. The case of the bush dog is a good example. Despite great efforts at several of the world's best zoos, populations are neither self-sustaining nor are sufficient numbers of individuals kept in captivity to form a viable population (Kleiman and Rodden pers. comm.).

From the point of view of conservation, there appears to be insufficient emphasis on breeding those canid species which are most likely to require an ark in the near future. To change this state of affairs will be difficult. The worst problem is one of circularity: there are few specimens of the rarer species in zoos; therefore they breed less frequently and with greater litter mortality; zoo populations never become self-sustaining; hence we do not get experience at breeding these species. And so on ad infinitum.

We cannot afford to wait to solve this problem. Unfortunately, if we wait until a species becomes highly endangered to begin captive breeding programmes, the data suggest that such programmes may be doomed to failure. In captive breeding, practice, and lots of practice, makes perfect: mortality declines with the log of breeding effort.

Plans for captive breeding and implementation of these plans must begin before a species reaches an extinction crisis (as outlined in the IUCN Position Statement on Captive Breeding, adopted in 1987). To some extent, greater coordination of breeding programmes may improve the state of affairs: if each zoo attempts to breed a single species of rare or endangered canid, and zoos breeding the same species coordinate efforts, improvement in breeding success may, perhaps, be accelerated. The establishment of studbooks for a variety of canids (Lycaon, Speothos, Chrysocyon), the development of AAZPA/SSP plans, and the U.S. Fish and Wildlife Service recovery plans for maned wolf, red wolf, Mexican wolf, Indian wolf, bush dog, and African wild dog all indicate that such coordination is already well underway (U.S. Seal pers. comm).

Where possible, captive breeding should be attempted in the country in which a species is found, and preferably in the region in which reintroduction will take place. The ecological imperative for such a suggestion is strong: the potential for exposure to unfamiliar diseases will be reduced; semi-natural enclosures can be constructed at a lower cost and native foods can be given to the animals; reintroduction will, most probably, be simpler. Furthermore, costs associated with such projects should be lower.

[Figure 4. Mean litter mortality in all years, by species. Mean] 2 s.e.m.

Of course, arguments can be made against such a strategy: it will limit the ability to use advanced technology (genetic screening, super-ovulation, embryo transplants, hormonal monitoring, etc.); a local epidemic could wipe out the wild and captive population; if poaching is a problem, protection may be, perhaps, more difficult. On balance, however, the potential problems can be mitigated: samples can be analyzed in laboratories away from areas in which breeding occurs and captive individuals can be isolated from wild individuals geographically or with physical barriers.

Of course, there is no single approach that will be the best approach for all species. Ex situ and in situ approaches provide different advantages and can be mutually supporting. If funds are plentiful, and there are relatively large populations still in the wild, captive breeding programmes for the vulnerable and endangered species should be started both locally and in zoological parks that have extensive experience with such programmes.

Species Recommendations

In Table 1, we list species according to their conservation status and captive breeding history. We have tried to group species that we feel present similar problems and need similar attention. Species for which we feel there is not enough information to make definite statements about their present population size, but which are not officially listed by the IUCN as "Insufficiently Known" have been included in the table. Those species that are abundant in the wild and have been shown to breed well in captivity are not included in the table.

Simien jackal

The Simien jackal is the second most endangered canid. As an endemic, population sizes have always been small but they may be declining. The present population is well under a thousand, and probably no more than 500-600. Yet no record exists of the pecies having bred in captivity. We recommend strongly that a captive breeding programme begin as soon as possible. The potential for reintroduction of Simien jackals to areas where they have historically occurred should be pursued so that a second and distinct wild population can be established.

Island grey fox

The island grey fox should not be difficult to breed in captivity considering its close relationship to the North American grey fox, U. cinereoargenteus. The population, at present, appears stable, but is inherently fragile due to its small size. We recommend that trial captive breeding be conducted to ascertain if information gathered in breeding the grey fox will enable breeding of the island grey fox to proceed easily if it were to become necessary. A captive breeding programme is being considered by the Los Angeles Zoo (Garcelon pers comm.).

Red wolf

By 1980, the red wolf was believed to be extinct in the wild (USFWS 1984). Without a doubt, this makes the red wolf the most endangered canid. The eventual demise of wild populations was recognized early enough to allow a captive breeding population to be established at the Point Defiance Zoological Gardens in Tacoma, Washington, United States. In 1977, starting with a founder population of 14 red wolves believed to be genetically pare, a breeding programme was begun: by 1988, there were 80 red wolves in captivity at 8 locations in the United States. A reintroduction programme has begun in the 477 km² Alligator River National Wildlife Refuge in North Carolina and appears to be successful (all information from Phillips and Parker 1988). More reintroduction sites are needed: the U.S. Fish and Wildlife Service recovery plan specifies the need for 8 to 10 additional sites, with a total population of 200 free-ranging animals (M.K. Phillips pers. comm.). We support these goals and emphasize that without further reintroduction efforts, success cannot be assured.

African wild dog

The African wild dog, despite intensive efforts at breeding, appears to suffer from high mortality in captive breeding. A likely cause of failure is, perhaps, insufficient attention to the species' complex social structure. In the wild, first-year mortality is approximately 35%, just over half that reported in captive breeding. As persecution by humans and disease, rather than habitat destruction, are probably the major causes of population decline, reintroduction of African wild dogs should be possible (but see Childes 1988). We suggest that efforts to breed the species be supported where they already exist and that further attempts to reintroduce wild dogs be investigated. As a large number of individuals arc already in captivity, perhaps lab/zoo research could be initiated to determine the cause of high reproductive failure. The South and East African populations appear to be genetically distinct; zoo populations appear to be predominantly of South African origin (R. Wayne pers. comm.). Hence, genetic screening, and the establishment of' studbooks which incorporate this information will be necessary to ensure that reintroductions using zoo-bred specimens do not mix gene pools.
A studbook for the U.S.A. is being compiled by B. Brewer, Brookfield Zoo (Brewer pers. comm.).

Bush dog

Little is known about the biology of the bush dog, and captive breeding has been fairly unsuccessful. We suggest that both a captive breeding project and a field study be pursued. However, until the ecology and the distribution of the bush dog are better known, the potential for reintroduction, or even the need, remain in doubt.

Dhole

The dhole appears to breed somewhat better in captivity than would be expected. The dhole enjoys a wide distribution despite its low numbers. Uncertainty remains about its status in many parts of its range (Soviet Union, China, Sumatra, Burma). Before embarking on further captive breeding efforts, we would recommend that information be collected from local authorities in these countries.

Maned wolf

A captive breeding programme aimed at improving captive breeding and increasing the potential stock for reintroduction is already underway. We would only suggest that given the recent decline in litter mortality, a programme of in-country breeding and reintroduction be pursued in Brazil and/or Argentina.

Fennec fox

Like the African wild dog, the fennec fox suffers from a high litter mortality when bred in captivity, despite a large number of litters having been bred in captivity. As the fennec does not appear to be in any immediate danger of extinction, captive breeding of the fennec is not a high priority.

Corsac fox

Little is known about the behaviour and ecology of the corsac fox. However, it is commonly bred in captivity, suffers low litter mortality, and should population levels prove to be low, captive breeding would not pose any problems.

Zorros and Other Fox Species

These nine species share little in common, other than that we know little about their biology and none of them is commonly bred in captivity. We would strongly encourage local governments and research institutions to conduct surveys of the distribution and abundance of these species. Where economically possible, captive breeding and captive populations might provide at least basic biological information which, at present, does not exist.

A Final Note on Reintroduction

Carnivores, particularly large carnivores that range over huge areas and live at low densities, are especially vulnerable to extinction. Active efforts to remove predators from agricultural areas may often hasten the demise of a species (red wolf, African wild dog, grey wolf) Nonetheless, many carnivores breed well in captivity and, given the generally flexible nature of many canid societies (Macdonald 1983), reintroduction may well prove easier in many canid species that it has been for other taxa.

Several reintroduction programmes, while still in their initial stages, appear to be successful. The swift fox has been recently reintroduced to Canada. Populations are growing and captive-bred animals appear to be adjusting to living in the wild (see Chapter 6, swift fox, Current Research ). The red wolf reintroduction programme also appears to be a success (see above). However, due to the potential conflict of wolves and humans, the project was nearly abandoned (Phillips and Parker 1988; Wilcove 1987). Not until an isolated reserve was found could reintroduction begin.

The conflict between canids and people makes any reintroduction difficult. Every time a reintroduction of grey wolves has been suggested, controversy has erupted (Wilcove 1987). Even in areas where wolves occur naturally, but are near extinction, efforts to preserve remnant populations meet strong resistance (e.g. Italy, Boitani pers. comm.). Similar problems are voiced concerning the African wild dog. In Zimbabwe, where fewer than 500 wild dogs remain, the mere suggestion of further protection of the wild dog, or reintroduction and captive breeding, evokes a furious response from local ranchers-ironically, one of the most vociferous lobbies against reintroduction of the wild dog in Zimbabwe is game ranchers (Townsend 1988).

As a last resort, captive breeding and reintroduction are useful tools for species conservation. If local and regional objections to reintroduction can be overcome, the success of such projects appears to be good. Captive breeding, in addition to providing a temporary safe haven from persecution or disease can also allow isolation of wild canids from genetic dilution by cross breeding with coyotes (e.g. the red wolf, Phillips and Parker 1988). However, we do not believe a species should or can be preserved in zoos ad infinitwn as the living equivalent of a stuffed dodo or thylacine. The aim of captive breeding, when pursued for conservation purposes, should be to provide a temporary safe haven for a species. For this purpose, zoos have, and should continue to serve as arks, not museums.

© 1990 International Union for the Conservation of Nature and Natural Resources