Age-specific survival, abundance, and immigration rates of a Weddell seal ( Leptonychotes weddellii) population in McMurdo Sound, Antarctica

• Published in: Canadian Journal of Zoology
• Main Authors: Cameron, Michael F, Siniff, Donald B
• Format: Article in Journal/Newspaper
• Language: English
• Published: Canadian Science Publishing 2004
• Subjects: Erebus Bay; McMurdo Sound; Weddell; Antarc*; Antarctica; Sea ice; Weddell Seal; Weddell Seals; ice covered areas
• Online Access: http://dx.doi.org/10.1139/z04-025; http://www.nrcresearchpress.com/doi/pdf/10.1139/z04-025

Since the 1960s, Weddell seals (Leptonychotes weddellii (Lesson, 1826)) have been tagged and surveyed annually in McMurdo Sound, Antarctica. Mark–recapture analyses and model selection trials using Akaike's Information Criterion indicate that sex, cohort, and year affect juvenile (ages 1 and 2) survival. In contrast, year and perhaps sex and cohort are less important factors for adult survival. Average annual survival is higher among adults (0.93) than juveniles (0.55–0.59) and there is little evidence for senescence to at least age 17. The oldest known-aged female and male in the study were 27 and 24 years old, respectively. Data suggest that the abundance of a resident population of Weddell seals remains relatively stable over time despite annual fluctuations in Jolly–Seber abundance estimates for the entire population. We argue that this annual variability is likely the result of temporary immigration of animals born outside the study area; mean rates are estimated from a simulation model and tagging data to be between 16.8% and 39.7% for females and between –13.1% and 31.6% for males. Sea ice extent appears to affect immigration where, during times of reduced fast-ice, immigrants are forced, or allowed easier access, into the ice-covered areas of Erebus Bay from surrounding locations. Our findings contradict previous studies reporting lower survival and higher immigration. Model choice is shown to be the most likely cause of these discrepancies and we provide evidence that our models are more appropriate than those used elsewhere.