Empirical considerations on the stable age distribution

Abstract

We used 24 years of live-trapping data to investigate four questions about the equivalence of time-specific and cohort life tables in the yellow-bellied marmot (Marmota flaviventris). We also reanalyzed parts of the data of Zammuto and Sherman (1986) that claimed equivalence of time-specific and cohort life tables for 8 years of data from Belding s ground squirrel (Spermophilus beldingi). Pulse of births and deaths appeared to move through time for both species. For both marmots and ground squirrels, a k-sample Kruskal-Wallis test showed homogeneity of all time-specific life tables and a x2 test showed heterogeneity. Yearly pairwise comparisons from independent samples revealed no differences between years using Kruskal-Wallis and Kolmogorov Smirnov tests, 19% of the 174 X2 tests showed heterogeneity of samples. Comparing the 24 time-specific life tables to the cohort life table for marmots, pairwise Kruskal-Wallis and Kolmogorov Smirnov test showed no differences, but x2 tests revealed differences for seven of the years. Litter sizes in marmots were not equal among years but were equal among age classes. The yearly crude death rates, gross reproductive rates, and growth rates for marmots were normally distributed without skew or kurtosis, a possible indication of constancy of rates. Kruskal-Wallis and Kolmogorov Smirnov statistics do not seem to have sufficient power for this statistical application, and a x2 test must be used. In marmots, time-specific and cohort specific life tables are not equivalent.