mechanism of this effect would likely be the same as for
intraspecific signals that delay puberty or estrus in adults.
The hypothesis posits that in the presence of heavy preda-
tion, reproduction is delayed to avoid unnecessary waste
of energy when pregnant females could be at greater risk of
predation and progeny likely would be consumed (Cushing
1985; Ylönen 1989; Ylönen and Ronkainen 1994). How-
ever, investigations using gray-tailed voles exposed to urine
and feces from mink (Mustela vison;Wolff and Davis-
Born 1997) and bank voles (Clethrionomys glareolus) ex-
posed to odors from weasels (Mustela nivalis;Mappes et al.
1998) failed to support this hypothesis. Interestingly, in one
field study that examined the response of prairie voles and
pine voles to odors of a mink predator, voles did not re-
duce their scent marking activity compared to control areas
(Wolff 2004).
These tests raise questions about the efficacy of some
of the laboratory studies designed to assess reproductive re-
sponses to chemosignals in rodents. As Wolff (2003c) notes,
we need to remember that laboratory studies are critical for
defining and investigating some aspects (e.g., underlying
mechanisms) of these phenomena. But, we must also be vig-
ilant about testing them in field settings. In fact, it is only
under field conditions that we can judge either the true ex-
istence of the effect in natural populations or test the po-
tential ecological and evolutionary consequences of such ef-
fects. As my concluding section will note, it is in this area
that future work is needed most.
Ecological and Evolutionary Considerations
Bronson (1970, 1979a, 1979b, 1983, 1985; Bronson and
Coquelin 1980) provided multiple reviews of the reproduc-
tive ecology of house mice, with some comments on deer
mice. He details effects of environmental factors such as
space, photoperiod, and food resources on reproduction
and seasonal breeding. He discusses the possible effects on
reproduction of social factors such as aggression and social
signals like urinary chemosignals. Drickamer (1986a) and
Vandenbergh and Coppola (1986) both examined the ecol-
ogy of urinary chemosignals and their possible functions in
house mice. Most of the comments on adaptive significance
in these multiple papers are speculation. As noted earlier,
field tests under even seminatural conditions are few in
number, making it difficult to arrive at any conclusions
about the ecology of the primer urinary chemosignals.
However, given that at least some of the chemosignal-
mediated effects on reproduction are effective under field
conditions and others remain to be tested we can proceed
to ask questions like: What, if any, are the fitness conse-
quences of these urinary chemosignal effects on reproduc-
tion in various rodent species? What might be the func-
tional effects of these various chemosignals on reproduction
under natural ecological conditions? How would these sys-
tems have evolved in terms of both the production /release
of the chemosignals and their reception by conspecifics?
Are there any patterns that develop when these effects are
examined in relation to phylogenetics? Or, do predictable
patterns occur with regard to ecological settings or social
systems?
The urinary chemosignals that accelerate or delay sex-
ual maturation in individuals and, in some cases, alter fe-
male patterns of reproduction, might indirectly affect rates
of population growth (Krebs et al. chap 15, this volume). At
low densities and at the start of a breeding season, male
urine that accelerates female puberty and onset of breed-
ing in adult females would be predominant, leading to in-
creases in reproduction. When density increases or other
conditions change, such that there are greater risks associ-
ated with possible reproductive efforts (see the following in
this section), then a combination of release of maturation-
delaying chemosignal from females overrides the male uri-
nary signal and results in delays in reproduction. I hasten to
note that urinary chemosignals would only be one portion
of the set of cues that are available for the rodents to pro-
cess in terms of the gains from reproduction versus the pos-
sible costs and risks. The findings on urinary chemosignals
need to be embedded in a larger environmental context, and
one that also accounts for the changing seasons in many
locations.
Drickamer (1986c) hypothesized that the active compo-
nent(s) of urinary chemosignals likely originated as urinary
compounds, for example, urinary proteins or components
attached to urinary proteins. At some time in the past, pro-
duction and /or excretion of these particular compounds
could have been influenced by social, environmental, and
hormonal conditions of the animal. If some advantage were
gained by a conspecific receiving a message(s) represented
by these urinary compounds, then natural selection could
favor altering reproductive physiology to conform with the
message. Once the nature of the chemical compounds and
the sensory pathways and modes of action in recipient in-
dividuals are known, the fitness consequences of these phe-
nomena can be assessed.
In addition to social factors affecting the release of these
compounds that influence maturation, daylength and food
resources can affect chemosignal release in house mice.
Shorter daylengths and food restriction result in the release
of delay chemosignal in the urine of grouped females and
the lack of the puberty-accelerating chemosignal in male
urine (Drickamer 1984b,d). In house mice, male urine doesAcceleration and Delay of Reproduction in Rodents 111