Life Histories of Mammals, Revived: A Reply to Messier

Abstract

Messier (1990) argues there are six possible biases in Zammuto's (1987) life table study. I agree that some of the problems he discusses are possible biases of Zammuto (1987) or, for that matter, any life table study (see Caughley 1966, 1977, Millar and Zammuto 1983, Harvey and Zammuto 1985). I disagree, however, with the evidence that Messier presents to show these biases exist in Zammuto (1987). In addition, Messier (1990) provides few data or statistical analyses to support his claims. Moreover, Zammuto's (1987) research has been replicated and his conclusions corroborated (Jia and Fox 1990), casting additional doubt on the validity of Messier's criticism. Messier claims Bronson (1979), Murie et al. (1980), and Boag and Murie (1981) do not provide any indirect support that undisturbed populations of montane ground squirrels may be stationary enough to construct time-specific life tables. Zammuto used the mathematical and verbal definition of stationary population (lx and m. constant, Ro = 1, r = 0, no effects) quoted from Pielou (1974:29) as: a population of seasonal breeders ... show[ing] annual fluctuations but ... show[ing] no trend over period of several years. Contrary to Messier, I find: (1) Bronson (1979:275; Table 2) clearly showed 0% of the variance in litter size (mj) was explainable by effects in ANCOVA, and states: year of sampling had no effect on any criterion of litter size, in the montane golden-mantled ground squirrel over three years in several populations; (2) Murie et al. (1980:238) report homogenous means for litter size (mj) among all years and state: comparisons of [litter size] data within each area between years showed no significant differences, enabling us to combine the [annual] data from [7] different years, in three populations of the montane Columbian ground squirrel; and (3) Boag and Murie (1981:2231, 2236) present minor interyear variation in adult annual survival rates (IU 5 of 6 years 0.44-0.56, Table 8) and adult density (5 of 7 years 11.6-12.5 individuals/ha; Table 1) and state: this ... relatively small variation ... in numbers suggest[s] that the density of these ro-