Rodent Societies: An Ecological & Evolutionary Perspective

Arnold 1999). Progesterone decreased, and there was only
slight evidence of infanticide. However, male takeover may
result in infanticide (Perrin et al. 1994). The females that
failed to wean young gained sufficient mass to reproduce
the following year, rather than skipping reproduction. Thus
the males father young at the first opportunity rather than
having to wait an additional year. The females regained
some of the lost fitness by producing a larger litter in the
subsequent mating with the new male (Hackländer and Ar-
nold 1999).
In yellow-bellied marmots there is strong directional
selection for early maturity (Oli and Armitage 2003). De-
spite females reproducing earlier having a higher fitness
than those that delayed maturity, only 53 of 171 (31.0%)
of 2-year-old females (age of first reproduction) successfully
weaned litters. However, only 19.6% reproduced when an
older female was present in the matriline, compared to
47.8% when older adults were absent (Armitage 2003d).
Reproductive suppression acted to delay the age of first re-
production and was independent of the number of females
in the matriline. As age of first reproduction increased, a
smaller percentage of females was suppressed (fig. 30.2).
Reproductive suppression occurred when the mother was
the older female present; however, reproduction by a 2-
year-old was more likely if the mother was the older female
rather than some other adult (Armitage 2003d).
Clearly, reproduction was suppressed more in 2-year-
olds than in older animals (fig. 30.2). Possibly the 2-year-
olds failed to reproduce because they were immature rather
than suppressed. However, when 2-year-olds were matched
for condition (Armitage 2003d) or received supplemental
food (Woods 2001), reproduction was significantly less
likely if older females were present.
The effect of the presence of older adult females on the
reproduction of younger adult females raises the question
of whether the decline in fitness (NRR) in large matrilines
(4 or 5 adult females) could be related to their age distribu-

tion and reproductive suppression. In matrilines in which

one or no female reproduced (low success), half of the fe-

males aged five or older reproduced, whereas only 12%

of females aged two to four reproduced (fig. 30.3). Two

females reproduced in matrilines of moderate success; a

greater percentage of older females than of younger females

reproduced (fig. 30.3). In the matrilines in which three or

four females (high success) reproduced, both old and young

females reproduced at high rates (fig. 30.3). The combined

low- and moderate-success matrilines are characterized by

significant differences in the reproductive rates of old and

young females, whereas reproductive rates of young and

old females in the high success matrilines do not differ (Ar-

mitage 2003d). Thus reproductive suppression is the major

factor decreasing mean NRR in large matrilines; however,

reproductive suppression is not total, which results in high

NRR in some large matrilines. Reproductive suppression

Evolution of Sociality in Marmots: It Begins with Hibernation 365

Figure 30.2 Age and reproductive suppression in yellow-bellied marmots.

Figure 30.3 Reproductive success in large matrilines by young and old adult

females.