S3.1). Moreover, the decline
in annual survival rate after 2004 in the year model was not statistically significant, though in the age model, decline after age 17 was. In males, however, the year model produced a better fit to data than the age model (Fig. S3.2). Nevertheless, the survival-year model added information, because it revealed fluctuations in young Cell Cycle inhibitor animals not evident in the age model (Appendix S3). Survival was high in 1986, low in 1987, then increased until 1989 before settling on a long plateau (Fig. S3.3). The 1986–1986 variation was a cohort effect: first-year survival was high for the 1985 cohort relative to the 1986 and 1987 cohorts in both males and females. The cohort difference, however, did not persist in older
animals (Fig. S3.1, Fig. S3.2). The age model produced an intermediate estimate for first-year survival, averaging the three cohorts. Annual survival of adult females was high from age BGB324 5 to 16, averaging 86%/yr, but then declined abruptly. This is a higher rate and a longer duration of prime survival than we expected and the first evidence for senescence in survival rates of northern elephant seals. Our earlier work did not detect the decline in female survival because there were no data on females older than 15 yr (Le Boeuf and Reiter 1988, Reiter and Le Boeuf 1991). Schwarz et al. (2012) found limited power in estimating survival beyond age 15 due to the small number of animals retaining tags. Average male survival was <72%/yr at all ages and lower than female survival after age 3, as reported in earlier studies (Clinton and Le Boeuf 1993). Neither our MCE current analysis nor the earlier work detected senescence in male survival, but high mortality throughout life meant few males were still living at age 12 when senescence would be most likely. On the other hand, our earlier study did detect declining competitive ability in males past age 12 (Clinton and Le Boeuf 1993). Juvenile survivorship in the current study was 31% from weaning to
age 3 and similar in the two sexes, a rate close to the average reported across several previous cohorts (Le Boeuf and Reiter 1988, Le Boeuf et al. 1994). This average masked variation, however, and low survival in 1986–1987 may have been due to poor foraging conditions associated with an El Niño event (Trenberth and Stepaniak 2001, Crocker et al. 2006). Our earlier study of juvenile survival also described substantial year-to-year fluctuations (Le Boeuf et al. 1994). These rates of survivorship, though, began at weaning and omit pup mortality, and 10% of pups in the Año Nuevo mainland colony died before weaning in 1985–1987 (Le Boeuf et al. 2011). In population modeling, the relevant rate of juvenile survivorship (from birth) was thus 28%, not 31%. Dispersal of branded animals to nearby colonies—“prospecting” for alternative breeding sites—also confirms earlier observations (Le Boeuf et al. 1974, 2011).