Biased estimates of fur seal pup mass: origins and implications

The mass of fur seal pups weighed in different years can be used to estimate growth rates or compared with one another to make inferences about the relative condition of a population. However, unless appropriate precautions are taken, many factors can bias estimates of pup mass and lead to incorrect conclusions. Using data collected from tagged and untagged northern fur seal pups ( Callorhinus ursinus ) at the Pribilof Islands, Alaska, I assess how milk consumption, the timing of sampling, and the effects of growth and sample size influence the size of pups captured for weighing. Evidence is presented suggesting that pup mass may increase in a sigmoid fashion, with the most rapid rate of growth occurring when about two months old. This phenomenon can confound efforts to compare the masses of pups weighed on different days in different years, particularly if pups are weighed over the period of rapid growth. Variability in pup mass increases with time because growth rates of individuals vary and because the amount of milk pups consume increases with body size. Thus sample sizes must be increased as the pups grow older in order to detect statistically significant differences in mean body mass. There is also evidence that pups of different ages and sizes are not randomly distributed on the breeding beaches and are not randomly selected for weighing. It appears that the first pups captured for weighing are smaller and younger than subsequent captures, possibly because smaller pups are easier to handle and are segregated to the peripheral rookery regions where sampling begins. These hidden biases, related to sampling error and fur seal biology, must be considered and controlled for when weighing fur seal pups.