For walruses, annual survival of juvenile and adult walruses, 4–20 yr of age, must be approximately 0.96–0.99 to compensate for their low fecundity (DeMaster 1984, Fay et al. 1997). Mature females give birth to a single calf, birth intervals are typically ≥2 yr (Fay 1982), and calves are believed to have annual survival rates ranging from 0.5 to 0.9 (Fay et al. 1997). Hence, recruitment into the adult age classes is expected to be low, while adult survival is high; hence, the number of cows should be relatively stable from year to year. We note, however, that small changes in survival and slow rates of decline will likely be undetectable (see Harris et al. 2008). Other issues

that may complicate interpretation of walrus calf:cow ratios Ferrostatin-1 manufacturer Daporinad supplier include classification error, repeatability of surveys, and the timing of surveys: Interpreting the age ratios clearly depends upon

how well observers can classify walruses to different sex and age categories. During the 1981 survey, Fay and Kelly (1989) compared untrained observers with trained observers in their ability to classify walruses based upon Fay’s classification scheme. They determined that all observers differentiated adults from juvenile age classes equally well, but that untrained observers tended to overestimate calves and underestimate yearlings. Fay and Kelly (1989) believed this was likely due to untrained observers not realizing that calves have dark pelage and that yearling tusks are hard to observe medchemexpress unless their heads are tilted up. Training corrected this issue. A more important question is how often trained observers make classification errors. Calves and yearlings are easily recognized; calves due to their small size and dark pelage, yearlings due to their small size, lighter pelage, and “nubbin” tusks. However, classifying adult females requires comparing the ratio of tusk length to snout width and depth as visually depicted in Figure 1. Unfortunately, the range of tusk length to snout depth ratios overlap by 50% between animals 4–5 yr of age and those

6–9 yr of age. For these two age classes, the range of tusk length to snout width ratios also overlap by 47% (Fig. 2). If 4–5-yr-olds are classified as adults, this will bias calf:cow ratios low as it artificially inflates the denominator of the ratio. However, the bias should be low because there are relatively few 4–5-yr-olds in the sample compared to the number of cows. For example, assume the percentage of 4–5-yr-olds misclassified is exactly 50%, representing a worst-case scenario. Across all survey years, 527 calves, 536 4–5-yr-olds, and 5,435 cows were classified. If 50% of the 4–5-yr-olds had been misclassified as adult females then the true number of cows should be 5,435 − 536 = 4,899. Hence, the true calf:cow ratio would be 527/4,899 = 0.108, while the biased calf:cow ratio is 527/5,435 = 0.097. This is an absolute difference of 0.01 or approximately 1 calf per 100 cows.