Priorities For The Future

The need for a global captive breeding programme

A global, cooperative captive breeding programme for giant pandas would mean that all holding institutions worldwide would work together to share animals (or germ plasm, as necessary) to maximise genetic heterozygosity, avoid inbreeding and distribute the species' valuable genes among multiple populations (Fig. 21.9). Such a partnership would ensure that the captive population remains viable, healthy and genetically 'insured' while fully capable of supporting species conservation ex situ and, if necessary, in situ and long into the future. Cooperative breeding programmes for some wildlife species have been

Figure 21.9. Metapopulation management of the captive giant panda population showing animal or gene flow among primary breeding centres (large, dark circles) and other captive institutions (medium and small circles) with the potential of future gene flow from the wild as well as animals held outside China (in the west).

Figure 21.9. Metapopulation management of the captive giant panda population showing animal or gene flow among primary breeding centres (large, dark circles) and other captive institutions (medium and small circles) with the potential of future gene flow from the wild as well as animals held outside China (in the west).

Animals or sperm in place for decades. However, unlike what is being proposed here, most exist only at a regional level, administered by regional zoo associations (e.g. the SSP coordinated by the AZA, or the European Endangered Species Programme coordinated by the European Association of Zoos and Aquaria; Princée, 2001). There are a few exceptions, including the global captive management programmes for the golden lion tamarin (Ballou et al., 2002), which can serve as a model for the giant panda.

The global captive breeding programme for the giant panda would consist of multiple regional populations or subpopulations that would be managed as a metapopulation, i.e. an interacting set of populations (see Fig. 21.9). We foresee that the metapopulation would comprise a set of Chinese institutions, institutions outside China and the wild populations themselves. Transfer of animals or germ plasm among captive breeding institutions would maintain genetic diversity within institutions. Eventually (and if warranted) animal movements might also include the release of captive-born animals into vacant but suitable wild habitat, i.e. reintroductions. In theory, pandas living in nature might also occasionally be acquired to bolster the captive gene pool, but given the advances in reproductive technologies (see Chapter 20) it would make more sense to collect surplus germ plasm (i.e. sperm) from wild pandas for infusing new genes into captivity. Lastly, a metapopulation strategy could involve transferring animals between wild populations to maintain maximum genetic diversity in nature.

Most managers of captive wildlife populations have adopted a goal of maintaining 90% of the population's heterozygosity for the next 100 years (Soulé et al., 1986). With the current level of retained heterozygosity at 97% and a realistic population growth rate of about 3% per year (see above), this would require a global captive population of about 340 giant pandas (estimated using PM2000 software; Pollak et al., 2002). Under tighter genetic management, which could increase heterozygosity by successfully breeding wild-caught animals which have so far failed to breed, a smaller overall population would be needed. For example, if retention of 98% of the heterozygosity could be achieved in this manner, population size requirements would decrease to about 270 individuals. Similarly, a smaller population would be needed if the population growth rate was higher than 3%. For example, a doubling to 6% annually could allow the genetic target to be hit with an overall world population of 280 giant pandas.

Regardless of the various scenarios, it is apparent that to maintain the genetic fitness of the giant panda into the future will require space and resources for the management of about 300 individuals. Based on our knowledge of the panda's life table and current age structure, a 3% annual growth would initially require producing about 13 to 17 births per year and would need to increase over time, eventually (in about 30 years) requiring about 28 annual births to maintain the population at 300. This seems attainable, as the average number of births per year over the last ten years has been 13.3.

Demographic versus genetic priorities

Certainly, the population should not be constrained to grow at only 3%. The higher the growth rate, the more demographically and genetically secure the population becomes. However, growth in numbers should not be the only priority. Genetic management must also be considered. Decisions to maximise production now (e.g. by only mating the most willing breeders) could compromise the long-term genetic health of the population. As an extreme example, in a population of 45 giant pandas, the use of only one male as a breeder of most females would cause the inbreeding level to increase at more than twice the rate than if all males were used. The use of only a few highly successful breeders will also hasten the population becoming genetically adapted to the captive environment.

Captive breeding programmes include genetic management priorities to maintain high levels of gene diversity in the population, avoid inbreeding and minimise genetic adaptation to the captive environment (Ballou & Foose, 1996). This can be accomplished by implementing a breeding strategy that minimises mean kinship, as discussed above (Ballou & Lacy, 1995). A mean kinship value (the average kinship between that individual, i, and all individuals) would be calculated for every animal in the population using the equation:

where kij is the kinship coefficient between individuals i and j . This is summed over all N individuals (j — 1 to N) in the genetically viable population and divided by N to give the average kinship of individual i to the entire population. (Note: usually relationships of individuals with wild-caught founders are not included in the calculation because managers are most interested in the genetic diversity already captured

Figure 21.10. Distribution of mean kinship values in the giant panda population.

in the progeny of founders. Wild-caught animals that have not yet produced should not technically be counted as having contributed genetic material to the captive population; Lacy, 1995.)

Priority breeders are those with low mk values as they have fewer relatives in the population than animals with high mk values. By selecting breeders that minimise average mean kinship, the breeding programme maximises retention of heterozygosity and avoids inbreeding to the extent possible. Using a mean kinship strategy also tends to equalise the genetic contribution of founders in the population since descendants of under-represented founders will have low mean kinship and be preferentially bred. The distribution of mk values in the contemporary giant panda population is shown in Figure 21.10.

The 14 giant pandas with mk = 0 are those wild-caught animals that have not yet reproduced at the time of this writing. Priority should be placed on trying to breed these individuals before they become reproductively senescent; in the case of males, at the least multiple semen samples should be collected and used for AI and/or stored in the giant panda Genome Resource Bank (GRB; see Chapter 20). Similarly, the two giant pandas with the highest mk value (0.06) should be kept from contributing further to offspring production, unless their mean kinships decline relative to the rest of the population at some point in the future. If these males continue to breed, their genes will dominate the captive population's genetic pool, causing inbreeding to increase rapidly.

Captive breeding programmes often need to focus on demographic expansion during their early years when the population is extremely small and at a high risk of extinction. As the population grows, however, more emphasis can and should be placed on genetic management (Ralls & Ballou, 1992). The giant panda captive population is at a point where both increased population size and genetic management are needed. This can be accomplished by breeding all females while also attempting to minimise mean kinship. Females with low mean kinships should be placed in the best breeding situations and paired with males with similar low mean kinship values. However, the dilemma faced by this and other programmes is that some of the females and males with low mean kinships are also often the 'problem breeders', e.g. those with behavioural or medical challenges. Lack of reproductive success has led to their low mean kinship values. Thus genetic management becomes a balance of:

1. placing breeding priority on those animals with low mean kinship, some of which may have questionable probability of breeding success, and;

2. establishing a sufficient number of pairs with high probability of breeding success (but not necessarily high mk) to accomplish the desired number of offspring needed for targeted population growth.

Most captive breeding programmes are able to balance these concerns and achieve population growth while undergoing genetic management. Two examples for the reader to review include the black-footed ferret (Russell et al., 1994) and the California condor (Ralls & Ballou, 2004). This same approach is needed for the ex situ population of giant pandas to ensure long-term genetic health and to establish and maintain a demographically secure and self-sustaining population.

Was this article helpful?

0 0

Post a comment