ous phenomena and hypotheses for the adaptive signifi-
cance of chemosignals. Given that olfactory cues are ex-
tremely important for most mammals, a series of a priori
hypothesis-testing experiments with a variety of species
from different trophic levels and natural histories should
provide greater insight into the roles of chemosignals in
mammalian societies.
Field tests and ecological issues
One of the key elements that is needed to advance our
knowledge of the possible effects of urinary chemosignals
on reproduction in small rodents is field experimentation.
The theoretical and experimental approaches used by Wolff
and colleagues (Rodd and Boonstra 1988; de la Maza et al.
1999; Mahady and Wolff 2002) were designed specifically
to discern among alternative hypotheses for some behav-
ioral phenomenon. Research must go beyond observation
and description and address current paradigms experimen-
tally to determine the functional and evolutionary signifi-
cance of each trait. We do need, as advocated previously, to
proceed with studies that provide comparative evidence for
or against the occurrence of the chemosignal effects on re-
production in rodents, but those alone will not suffice.
Application of chemosignal treatments to experimental
manipulations of density, reproductive condition, and sex
ratios or age composition can be used to test various hy-
potheses for the functional and adaptive significance of ol-
factory communication. It may be possible to use chemical
treatments to create animals that are incapable of produc-
ing and releasing certain chemosignals. This manipulation
could provide the capacity to change densities without al-
tering the amount of signal present. Castrated males in
natural populations are a probable manipulation, to test
hypotheses about density or male presence, but without
chemosignal production and release.
Mechanism studies
Studies examining the underlying neural and hormonal
mechanisms that regulate production and response to uri-
nary chemosignals are in their infancy. On the hormonal
side, measurements have been made of changes in blood
levels of estrogens, LH, prolactin, and other hormones in
association with some of the chemosignal-mediated phe-
nomena (reviewed by Bronson and Macmillan 1983; Mc-
Clintock 1983; Bronson 1989). While most of this hor-
mone work has been done on house mice, some findings
were reported for voles (e.g., Carter et al. 1986). There is
ample room for additional investigations of the hormonal
bases for production and release of urinary chemosignals,
as well as additional work on the mechanisms of action in
recipients. Modern technologies can accelerate analyses of
hormone levels. Here again, larger-scale comparative stud-
ies are necessary.
As noted earlier in this review, some findings are now
available concerning the chemical composition of the vari-
ous urinary chemosignals (Nishimura et al. 1989; Novotny
et al. 1998, 1999; Mucignat-Caretta et al. 1995). Here also,
modern technology should facilitate investigations via both
better chemical analyses and rapid throughput of samples.
We know virtually nothing about the chemical makeup of
the chemosignals in voles or deer mice. Until such chemi-
cal information is available, we cannot construct any com-
parative analyses to examine questions about ecology, mat-
ing systems, or common evolutionary pathways for urinary
chemosignals.
Developments in the past 20 years now make it possible
to explore the underlying sensory and neural mechanisms
in both recipients of chemosignal cues and also in, for ex-
ample, grouped females, where estrous cycles or estrous in-
duction are prolonged or suppressed, and where group-
ing results in the release of a puberty-delaying substance in
the urine. One such technique, FOS immunoreactivity, has
been used to examine the locations in the brain where the
messages from chemical cues in urine or changes in photo-
period are sorted in prairie voles (Moffatt et al. 1995; Tub-
biola and Wysocki 1997). This technique, and other im-
munocytochemistry procedures, can now provide specifics
on the neural pathways and regions of the brain involved
in the reception and processing of external cues. It should
be possible to trace complete pathways for these primer
chemosignals from external receptors via central nervous
system processing to hormonal mechanisms, altering, for
example, puberty.SummaryInvestigations of effects of chemosignals on rodent repro-
duction over almost 50 years fall into three main catego-
ries: those that accelerate puberty in young females, those
that delay puberty in young females, and the termination of
pregnancy following exposure to a strange male. These re-
productive effects in response to exposure to chemosignals
have been explored most extensively for house mice, but
deer mice, voles, lemmings, and gerbils have also been used
to study some aspects of these phenomena. While consid-
erable information is known about chemosignals and re-
production in the laboratory for several species, sufficient
information is lacking to provide a basis for comparisons
across species, ecological settings, or types of mating sys-Acceleration and Delay of Reproduction in Rodents 113