kin recognition and the multiple paternity of litters, could it
be that the male was able to kill selectively the pups that
were not his within a litter? See Ebensperger and Blumstein,
chap. 23 this volume for further discussion on infanticide.
In their feral environment, house mice are not readily
observable, so conclusions about their parental behavior
in these conditions are limited. Females are often observed
with their litters under both feral and commensal situations,
but males are rarely present (Rowe 1981). Communal nest-
ing has been recorded in which at least two litters were be-
ing cared for in the same nest cavity by females related as
sisters or as mother and daughter (L. Drickamer, unpub-
lished data based on the DNA-parental assignment used
by Robinson 2000). Konig (1994) has investigated aspects
of communal nesting in wild house mice in a laboratory
setting. She reported that communally nesting females en-
joy greater lifetime reproductive success than those nest-
ing alone. This effect is pronounced when the communally
nesting mice are related females as opposed to unrelated fe-
males. It is noteworthy that there may be some connections
between the lower levels of female-female aggression, shar-
ing nests, and reproductive success. This notion would re-
quire additional testing for confirmation.
Considerably more is known about parental care in com-
mensal than feral house mice. A system of chemical and
auditory communication cues mediates the relationship be-
tween mother and pups (reviewed by Smith 1981). Com-
munal nesting with nursing of own and other’s pups is re-
corded by a number of investigators studying commensal
house mice (e.g., Crowcroft 1955; Brown 1962; Drickamer
personal observations). There are clear reproductive success
advantages from communal nursing relative to monoga-
mously mated females and those that rear litters alone, and
there may be some overall effects due to related versus un-
related mothers (Konig 1993, 1994). The cost-benefit anal-
ysis of communal nursing likely revolves around several fac-
tors, which could include food abundance and distribution,
external threats such as predators or other house mice, and
the availability of suitable nesting habitat. Further study
would be required to test these factors. As in many polygy-
nous mammals, males have a lesser role with regard to rear-
ing of young. Males are found with mothers and litters in
nests but nothing really is known of their potential paternal
behavior toward pups or young.
Infanticide has been studied extensively in the labora-
tory with mice, but virtually nothing is known from wild
populations. As for rats, the general results from laboratory
studies show that nonsire males will kill unfamiliar pups,
the act of copulation inhibits males from committing infan-
ticide, and that females will come into estrus after losing
a litter thus providing the perpetrator with a mating op-
portunity (e.g., Labov 1980; vom Saal and Howard 1982;
Parmigiani and vom Saal 1994). This “sexual selection” in-
fanticide also occurs in rats (Mennella and Moltz 1988)
and is common among rodents, and may be an adaptive re-
productive strategy for males (see Ebensperger and Blum-
stein, chap. 23 this volume).Diseases
Rat and diseases are synonymous in popular culture; rat-
borne diseases are thought to have contributed to more
deaths in the last 10 centuries than all wars and revolutions
ever fought (Nowak 1999). Recent comprehensive surveys
of rats in rural and urban areas have revealed more than
10 zoonotic agents (i.e., animal pathogens transmissible to
humans), including Leptospira, Toxoplasma, Yersinia, and
Hantavirus (Webster and Macdonald 1995a, 1995b; Bat-
tersby et al. 2002). While this may be of particular relevance
to public health, it should also be of interest to behavioral
ecologists, because of their impact to other wild species gen-
erally (which is essentially unknown), but also on behavior
more specifically. Toxoplasma goondii,the causative agent
of toxoplasmosis, has been shown to increase general ac-
tivity levels in infected rats (Webster 1994) and to reduce
their neophobia (Berdoy et al. 1995; Webster et al. 1994) as
well as their innate aversion to cat odors (Berdoy et al 2000;
fig. 32.7). Such host manipulation is likely to increase rat
predation by cats and therefore will benefit the parasite
that, although capable of infecting all mammals (including
humans), needs to find its way into a cat host to complete
its indirect life cycle. It is noteworthy that while humans
clearly represent a dead-end host for the parasite, reports
ranging from altered personality to higher incidence of car
crashes amongT. gondii-infected patients (Flegr et al. 2002)
may represent the intriguing side effects of a parasite evolved
to manipulate the behavior of another host (Berdoy et al
2000; Webster 2001). Mice also exhibit behavioral changes,
although the effect of T. gondiiinfection is more acute and
sometime lethal; the response of infected mice to the risk of
cat predation has not been tested so far (see Hay et al. 1983;
Hrda et al. 2000; Jackson et al. 1986; and review in Web-
ster 2001).
House mice can also act as both reservoirs of and vectors
for a variety of disease organisms (Blackwell 1981; Moro
et al. 1999; Rowe 1981; Singleton et al. 1993; 2000; Smith
et al. 1993), although there are far fewer recorded instances
of house mice acting as reservoirs or vectors of human-
related diseases than for rats. House mice are associated
with food poisoning (Salmonella;Rowe 1981; Streptoba-
cillus;Taylor et al. 1994) in connections with both humans
and livestock, and are reservoirs for lymphocytic chorio-
meningitis virus (Smith et al. 1993), cryptosporidium (Mor-
gan et al. 1999), and possibly rickettsial diseases, but do390 Chapter Thirty-Two