3.14 lethal equivalents with 50% of these being lethal recessivesd Arbitrary: 2% chance occurring each year;
50% reduction in survival when occurs In unmanaged populations, 20% of males available; in globally managed population, all males available with breeders and matings determined by mean kinship and relatedness of mates using SIMPOP Set at initial population sizes for each scenario a Unless noted, all parameters were derived from analysing data from the International Giant Panda Studbook. Inclusion of these parameters in VORTEX and SIMPOP resulted in a population with a deterministic growth rate of l = 1.027, comparable to the observed life table growth rate of 1.028; b Studbook results showed 21.8% of adult individuals breeding but improving over time;
breeding centres (e.g. at Chengdu or within the Wolong Nature Reserve).
3. A global, cooperatively managed population, initially with 140 individuals. In this scenario, giant pandas were managed as a 'metapopulation' and were moved among holding institutions on the basis of examining pedigrees (via the computer program SIMPOP) that resulted in the best breeding recommendations.
Input parameters for VORTEX and SIMPOP were estimated from studbook data over the interval of 1990 to 2002 described above (see Table 21.2). Some parameters were modified slightly to equate the growth rate of the model population with the growth rate observed from the life table (l — 1.028). All of the observed variation in the demographic rates (percentage females breeding and mortality) could be accounted for solely by demographic variation, so environmental variation effects were set to 0 (Miller & Lacy, 1999).
To model the current practice of repeatedly using successful males, for the first two scenarios only 20% of the males were defined as available for breeding in a polygynous mating system. SIMPOP was used to model the matings in the cooperatively managed population (as in scenario 3 above) because this program allowed identifying the breeding animals according to their mean kinship. In genetically managed captive populations, mean kinship is used to identify priority breeders. Mean kinship is the average kinship between an individual and the living population. Thus, an animal with a low mean kinship value carries genes from under-represented founders, i.e. these animals are priority individuals for mating. A breeding programme that minimises mean kinship maximises retention of heterozygosity (Ballou & Lacy, 1995). While VORTEX selects breeders and mates at random, SIMPOP
30% was used to incorporate some of this improvement and to equate the model l with the observed l of 1.028; c Studbook results revealed an annual female adult mortality of 8%. However, when this was applied to all adult age classes, it produced a survivorship curve that was substantially lower than the curve generated using the observed specific age class mortality rates. The survivorship curve using the rate of 6.5% female adult mortality was a close match between the observed age-specific survivorship and the one using this average; d Average estimated number of lethal equivalents was based on study of 45 mammals (Ralls et al, 1988; Miller & Lacy, 1999).
initial = 140
1 nn initial = 140
Figure 21.8. Probability of giant panda populations surviving over the next 50 years for small (initial N = 15), larger (initial N = 45) and managed (initial N = 140) populations.
mimics the strategy used if the population were being genetically managed using its pedigree.
Although there is no indication of catastrophic effects in the giant panda studbook data over the last 20 years (i.e. no obvious years with extremely high mortality or reduced reproduction across age classes), an arbitrary catastrophe was incorporated (see Table 21.2) to model the real but low probability of occurrence.
In both VORTEX and SIMPOP, simulations start with a set of unrelated founders equalling the number used for the initial population sizes. Since these starting conditions do not represent the level of relatedness in the current populations, results are presented starting at the point in time when level of relatedness in the modelled populations matches the relatedness in the real populations (0.13 for the small population versus 0.09 for the larger populations). By this time, some simulated populations had already gone extinct so that not all the populations have 100% chance of survival in Year 0 (Fig. 21.8).
The risk of extinction in small, isolated populations is directly related to their size. A single population of 15 animals has less than a 30% chance of surviving over the next 50 years compared to a managed population of 140 animals where none of the simulation runs went extinct. These results clearly illustrate that managing isolated populations of giant pandas is not a viable conservation strategy. On the contrary, the overall survival of the captive population is enhanced significantly by establishing a cooperative captive breeding programme. It is worth noting that our results undoubtedly underestimated the effects of population isolation on extinction probability. For example, in our model we only examined the detrimental effects of inbreeding depression on juvenile survival. However, inbreeding influences many other aspects of fitness, including adult survival and reproduction (Keller & Waller, 2002). If these variables had been factored in, the consequences of the isolated management scenario would have been even more severe.
Was this article helpful?