eliminate an entire litter via serial infanticides over several
days (N3 killers).
Infanticide after invasion by a new male is common for
many harem-polygynous mammals, and the payoff is that
victimized mothers come into estrus and conceive more
quickly than do mothers that continue to lactate (Hrdy
1977a, 1979; Fossey 1984; Packer and Pusey 1984). For
prairie dogs, however, females come into estrus only once
each year — so that infanticide by an invading male does
not reduce the time until the next estrus. Further, benefits
from the removal and cannibalism of future competitors are
probably minimal, because infanticidal male black-tailed
and Utah prairie dogs do not seem to experience either im-
proved survivorship or increased reproductive success in the
year following infanticide (Hoogland 1995, unpublished
data). Finally, males of neither species seem to gain via im-
proved condition of victimized mothers in the next breed-
ing season, because victimized females do not survive better
than females that wean a litter, are not more likely to rear
offspring in the following year, and do not produce larger
litters in the following year (Hoogland 1995, unpublished
data). The adaptive significance of infanticide by male black-
tailed and Utah prairie dogs thus remains unclear.
Infanticidal male black-tailed prairie dogs usually kill un-
related juveniles, but 10/39 26% of infanticides involved
males killing half-siblings. In every case, a yearling, non-
breeding male (N6 different killers) invaded a terri-
tory on which his father was the resident breeding male,
and then killed his father’s offspring (i.e., the killer’s half-
siblings). Why has natural selection favored the killing of
half-siblings by invading yearling males? Perhaps the loss
of offspring to invading sons is a secondary consequence of
helping sons to establish residency in a breeding territory.
If so, however, why hasn’t natural selection worked against
such a detrimental secondary consequence?
For 9/48 19% of infanticides among Utah prairie
dogs, a male killed the offspring of a female with whom he
had copulated — that is, the male might have killed his own
offspring. Why? I do not have an answer, but offer the fol-
lowing relevant observations /thoughts. (1) As noted in the
previous section, female Utah prairie dogs usually copulate
with more than one male, and multiple paternity is com-
mon. When males kill offspring of females with whom they
copulated, perhaps they target litters for which the mother
also copulated with more than one other male — so that
multiple paternity is likely and so that the male has a reduced
probability of killing his own offspring. In support of this
hypothesis, in 8/9 89% of cases in which a male killed
offspring of a female with whom he copulated, the female
also copulated with at least one other male. (2) When a
male kills offspring of a female that copulated with him and
other males, perhaps he somehow can discriminate between
his own offspring and juveniles sired by the other male(s),
and then can victimize only the other male’s offspring. I am
investigating this hypothesis via blood /tissue samples from
victims (when anything remains aboveground after canni-
balism) and killers, but the low level of variation among mi-
crosatellites (Haynie et al. 2003) is hindering my analysis.
Multiple mating by females sometimes might reduce the
probability of later infanticide by males. Specifically, be-
cause multiple mating might make it difficult to determine
which male sires offspring, a male might refrain from kill-
ing juveniles that could be his own offspring (Hrdy 1979;
Wolff and Macdonald 2004; but see Buchan et al. 2003 and
Sherman and Neff 2004). As a group, the three species of
prairie dogs that I have studied offer little support for the
notion that multiple mating by females functions to deter
infanticide. Gunnison’s prairie dogs show the lowest fre-
quency of infanticide by males, for example, but Gunnison’s
prairie dogs also have the same high frequency of multiple
mating as do Utah prairie dogs, which show the highest fre-
quency of infanticide by males.
In summary, interspecific differences in the frequency
and details of infanticide among three species of prairie dogs
are salient and bewildering. Data for a better understand-
ing will be elusive. Even though detectable frequencies of
killing among black-tailed and Utah prairie dogs are high,
for example, students and I observe infanticide only about
once every 300 person-hours of observations during the pe-
riod when juveniles are vulnerable.SummaryFrom observations of marked individuals, students and I
studied alarm calling, multiple mating by females, and in-
fanticide among black-tailed, Gunnison’s, and Utah prairie
dogs. For Gunnison’s and Utah prairie dogs, females with
juvenile offspring in the home territory are more likely to
call than females with only nondescendent adult or juve-
nile kin such as siblings, nieces, and cousins nearby. Female
black-tailed prairie dogs, by contrast, call as often when
only nondescendent adult or juvenile kin are in the home
territory as when nearby adult or juvenile offspring are
present. Male prairie dogs with kin in the home territory
are more likely to call than are males without nearby kin,
but only for black-tailed prairie dogs is this difference sig-
nificant. In addition to kinship, demography and body mass
also have been important in the evolution of alarm calling
among prairie dogs.
The percentage of females that copulate with two or
more males is 33% for black-tailed prairie dogs, but is
65% for Gunnison’s and Utah prairie dogs. One possible
reason for these differences is that female Gunnison’s prai-Alarm Calling, Multiple Mating, and Infanticide among Black-Tailed, Gunnison’s, and Utah Prairie Dogs 449