mice. For prairie deer mice urine from females in dense lab-
oratory populations applied to young females resulted in
smaller uteri and delays in reproduction (Terman 1968a;
Terman and Bradley 1981). Lower metabolic rates and al-
tered adrenal physiology and morphology were further evi-
dence of a physiological effect on growth and development
of young females raised in the presence of adult females
(Cherry et al. 2002). Lastly, Gubernick and Nordby (1992)
reported delayed sexual maturation in Peromyscus califor-
nicuswhen young females were housed with both parents,
though it appears that physical contact and the presence
of the father rather than a urinary chemosignal may be the
modulator of this effect.
Termination of pregnancy
A third type of chemosignal effect on rodent reproduction,
the Bruce effect (Bruce 1959, 1960, 1961), involves the ter-
mination of pregnancy for female rodents of some species
when they are exposed to a strange male or urine from a
strange male early in gestation. Marchlewska-Koj (1983),
Brown (1985b), Storey (1994), de la Maza et al. (1999),
and Mahady and Wolff (2002) have all reviewed these ef-
fects rather extensively.
Termination of pregnancy following exposure of a re-
cently inseminated female to a strange male or its urine was
first reported for house mice, and primarily occurs prior
to the implantation of the fetuses (Bruce 1959, 1960). The
strongest effect occurs when exposure to the strange male
is during the first 24 hours after insemination, with dimin-
ishing effects over the next 3 to 5 days until implantation
occurs. The Bruce effect can be induced by exposure to
male-soiled bedding (Parkes and Bruce 1962) or male urine
(Dominic 1969), indicating that it is based on a chemosig-
nal. Many facets of the Bruce effect have been explored in
house mice, including the neuroendocrine pathways, tests
of discrimination between sire and strange males, and vary-
ing effects of preexposure to the strange male (Parkes and
Bruce 1961; Brown 1985a). The efficacy of the Bruce effect
appears to vary among strains of inbred house mice (Chap-
man and Whitten 1969; Marsden and Bronson 1965a).
The Bruce effect occurs in a number of vole species, in-
cluding field voles (Microtus agrestis), prairie voles, mon-
tane voles (M. montanus), and meadow voles (Clulow and
Clarke 1968; Clulow and Langford 1971; Stehn and Rich-
mond 1975; Jannett 1980; Stehn and Jannett 1981). Un-
like the situation for house mice, in voles there appear to be
interspecific differences in the timing of exposure to strange
male odor, including time periods that extend well beyond
the time of implantation (Stehn and Richmond 1975; Ken-
ney et al. 1977). For example, for meadow voles, exposure
to a strange male was effective in reducing pregnancy rates
up to day 5 after copulation (Clulow and Langford 1971;
Mallory and Clulow 1977), but for prairie voles, exposure
up to 15 days after copulation induced females to abort
pregnancies (Stehn and Richmond 1975; Kenney et al.
1977). There may also be differences in the efficacy of the
Bruce effect that are dependent on whether the test stock
is wild-caught or laboratory bred (Mallory and Clulow
1977). There are likely differences in the neuroendocrine
pathways involved in the pregnancy termination, both rela-
tive to house mice and within voles, depending on when
during gestation the exposure to a strange male occurs (Mil-
ligan 1980).
Prairie deer mice exhibit the Bruce effect (Bronson and
Eleftheriou 1963; Bronson et al. 1964, 1969; Kenney et al.
1977; Dewsbury 1982c). The effect occurs during the first
few days after insemination and involves a chemosignal cue.
Lower rates of pregnancy occur when females are mated
with more than one male sequentially. The effect apparently
does not involve adrenocorticotrophic hormone (ACTH)
or adrenal hormones, but females given doses of prolactin
during exposure to strange males had pregnancy rates simi-
lar to those of unexposed females. In white-footed mice,
young females housed with an adult female and a male do
copulate, but pregnancy is terminated prior to implantation
(Haigh et al. 1988). Additional testing on other species of
Peromyscusmight reveal a general pattern for this genus.Field TestsMany of the findings about chemosignals elucidated in lab-
oratory settings are speculative with respect to their occur-
rence and effects in natural populations of these species.
Only a few of these phenomena have been explored under
field conditions and almost all such studies involve semi-
natural conditions where rodents are housed in some type
of outdoor enclosure (table 9.1).
Delay of sexual maturation and suppression of reproduc-
tion have been reported from field populations of deer mice
(reviewed by Terman 1968b, 1993; Kaufman and Kaufman
1989) and voles (reviewed by Taitt and Krebs 1985). In
white-footed mice, reproductive rates seem to be reduced in
some populations (Terman 1993) but not in others (Wolff
1994b). Reproduction was delayed in some young females
at high densities, but only those that remained in the pres-
ence of their fathers (Wolff 1992). In voles, delayed sexual
maturation and reproductive suppression have been re-
corded at high densities and /or peaks in population cycles
(Krebs et al. chap. 15, this volume); however, one experi-
mental study with gray-tailed voles (M. canicaudus) showed
a minimal and short-term effect of high densities of femalesAcceleration and Delay of Reproduction in Rodents 109