Info

1990 1992 1994 1996 1998 2000 2002 Year

Figure 21.4. Fecundity of captive-born (•,—) and wild-caught (o, ) giant pandas. Exponential regression is significant for wild-caught (p = 0.005; r2 = 0.56), but not for captive-born (p = 0.065; r2 = 0.30) individuals.

Figure 21.5. Age-specific fecundity of (a) female and (b) male giant pandas (solid line, captive males; dashed line, wild-caught males). Distributions have been smoothed with the Brass function (Gage, 2001).

b Age

Figure 21.5. Age-specific fecundity of (a) female and (b) male giant pandas (solid line, captive males; dashed line, wild-caught males). Distributions have been smoothed with the Brass function (Gage, 2001).

ages of most of these animals were estimated upon their entry into the ex situ population.

Because they were almost identical, wild-caught and captive-born female fecundities were combined (see Fig. 21.5a). Results demonstrated that females begin breeding at five years of age and peak at ten to 11 years; no female over the age of 19 has given birth. The earliest age of reproduction has been six years for captive-born and five years for wildborn males (see Fig. 21.5b). Wild-born male fecundity extends to older age classes than captive-born counterparts because the latter cohort is still relatively young (not having been in captivity long enough to breed at older ages). Wild-born males have successfully mated and produced offspring at 28 years.

Survival

Survivorships (lx), or the proportion of animals surviving from birth to age class x, were calculated using all animals existing in the population from 1 January 1990 to 1 February 2002 (the study window). Animals alive in the population in January 1990 as well as wild-caught animals entering the population during the window were treated as left-censored animals on the date and age they entered the study window. Escapes and animals alive at the end of the window were right-censored at the date and age they left the population (Hosmer & Lemeshow, 1999). Ages of animals at events (entry, exit and death) were recorded in days, and survivorship was calculated using Kaplan-Meier product limit estimators with the SAS Proc PHREG (SAS, 2001). Analyses (including wild-caught animals) were based on estimated dates of birth (as described above). Giant pandas of unknown sex were alternately included as males and females; all mortality of unknown-sex animals occurred before one year of age.

To test for changes in survival rates over time in different animal groupings, the study window was further divided into a pre-1995 window (ending on 31 December 1995) and a post-1995 window (starting 1 January 1996). Survival rates were calculated for males and females and for captive-born and wild-caught groups for both pre- and post-1995 windows. Statistical comparisons used proportional-hazards likelihood ratio tests (Proc PHREG; SAS, 2001).

Overall, there were no differences in survival success between males and females (p = 0.25), the pre- and post-1995 time periods (p = 0.27) or captive-born and wild-caught animals (comparing adult

Was this article helpful?

0 0

Post a comment