Time-Specific and Cohort Life Tables for Belding's Ground Squirrels

Zammuto and Sherman (1986) and Menkens and Boyce (1993) analyzed different data sets. The former considered 1-5 yr old female and 1-3 yr old male Spermophilus beldingi, whereas the latter considered 0-2 yr olds of both sexes. The reasons for these discrepancies are illuminating. Menkens and Boyce (1993) inferred that Zammuto and Sherman excluded the 0-yr age class from their analyses because they "lacked confidence that the estimated litter size, based upon counts following initial emergence from the natal burrow, was unbiased." More pups were undoubtedly born than survived to weaning (3 wk) every year. But there is an additional reason why Zammuto and Sherman (1986) omitted 0-yr-old animals: the original capture data (from Sherman and Morton 1984) did not accurately quantify the total number of juveniles in each year's birth pulse. This was because Sherman, Morton, and their field assistants were unable to capture every litter and pup that appeared aboveground yearly on the large, heterogeneous Tioga Pass Meadow (5 x 10 super(5) m super(2)), especially since the available person-power and the intensity with which pups were sampled differed among years. In addition, some of the weaned juveniles that were captured each year undoubtedly immigrated from outside the study area (especially males; see Holekamp 1984), and so should not figure into the population's birth statistics. These problems create a dilemma for comparative life table analyses. One could either estimate the size of each year's 0-yr age class from the available data, or omit the 0-yr age class from consideration. The first approach increases statistical power but potentially at the expense of accuracy, and the second approach does the reverse. Menkens and Boyce (1993) opted for the former, Zammuto and Sherman (1986) the latter.