Citation
Studies of general and sexual development in voles (Microtus)

Material Information

Title:
Studies of general and sexual development in voles (Microtus)
Creator:
Salo, Allen L., 1963- ( Dissertant )
Dewsbury, Donald A. ( Thesis advisor )
Branch, Marc N. ( Reviewer )
Brockmann, H. Jane ( Reviewer )
Kaufmann, John H. ( Reviewer )
Van Hartesveldt, Carol ( Reviewer )
Place of Publication:
Florida
Publisher:
University of Florida
Publication Date:
Copyright Date:
1992
Language:
English
Physical Description:
viii, 234 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Bedding ( jstor )
Body weight ( jstor )
Female animals ( jstor )
Mating behavior ( jstor )
Meadows ( jstor )
Prairies ( jstor )
Puberty ( jstor )
Species ( jstor )
Urine ( jstor )
Voles ( jstor )
Dissertations, Academic -- Psychology -- UF
Psychology thesis Ph. D
Voles -- Development ( lcsh )
Voles -- Reproduction ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )
theses ( marcgt )

Notes

Abstract:
Three experiments were conducted to investigate how patterns of general and sexual development might be correlated with the formation of social and mating systems among four species of voles ( Microtus ) . Species included pine voles ( Microtus pinetorum ) , prairie voles ( M. ochroqaster ) , meadow voles ( M. pennsylvanicus ) , and montane voles ( M. montanus ) . In Experiment 1, general and sexual development were monitored as voles were exposed to pheromones contained in the soiled bedding from family groups, adult males, or adult females. Few significant effects were found to be due to the treatment. Male pine voles exposed to family or male bedding were significantly heavier than those exposed to clean or female bedding. The uteri of female montane voles exposed to clean or male bedding were heavier than the uteri of those exposed to family or female bedding. In Experiment 2, the olfactory preferences of voles were measured when they were exposed to male and female bedding on weeks 4, 7, and 10 after birth. Few preferences were shown for either bedding type by any of the species. Female prairie voles and meadow voles revealed a significant preference for male versus female bedding. Both sexes of all species differed little in the total duration they remained near the female stimulus. Male meadow voles remained near the female stimulus significantly less than the males of the other species on week 10. Female montane voles remained near the male stimulus significantly longer on weeks 4 and 7 than did females of the other species. In Experiment 3, the influence of the fathers' presence and absence was studied during the rearing of the breeding pairs' first two litters. Pine voles produced their second litter considerably earlier if the male had been present during the rearing of the first litter rather than being absent. Pine voles weaned heavier offspring in the second litter than in the first, when the male had been present for the rearing of the first litter. Montane voles produced litters that were male-biased in sex ratio across both litters, if the male had been present during the rearing of the first litter. Results are discussed and interpreted from known differences in their contrasting social and mating systems.
Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 195-209).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Allen L. Salo.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
001867926 ( alephbibnum )
28997619 ( oclc )
AJU2442 ( notis )

Downloads

This item has the following downloads:


Full Text












STUDIES OF GENERAL AND SEXUAL DEVELOPMENT
IN VOLES (MICROTUS)
















By

ALLEN L. SALO




















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1992















This work is dedicated in remembrance of my father,

Raymond John Salo, who conducted comparative research on

two species of rabbits at the University of Massachusetts.















ACKNOWLEDGEMENTS


I would like to sincerely thank a number of people who

helped me complete the Ph.D. requirements. Although I am

not able to list all individuals who helped me in this

regard, a few should be named. First, I would like to thank

members of my family who supported my sometimes arduous trek

through academe. Most importantly my mother, Sonja Salo,

supported me spiritually as well as financially when the

going became difficult. My brother Mark Salo, his wife

Laurie, and children, were also constant sources of

inspiration. Other significant individuals include past and

present fellow graduate students: Florence Caputo, Jo

Manning, John Pierce, Joan Scheffer, Steve Taylor, Betty

Inglett, Sue Halsell, Nick Mills, and Kerry Sheehan.

Faculty members who have most strongly supported my

endeavors include Dr. Donald Dewsbury, Dr. H. Jane

Brockmann, and Dr. Jeffrey French. Certainly many other

members have impacted my life in various, positive ways.

Finally, I would like to thank members of the staff at the

University of Florida who helped in one way or another;

these individuals include: Theodore Fryer, Isaiah

Washington, Chris Wilcox, and Jeanene Griffin.















TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS... ................................... 11iii

ABSTRACT..... ......... ................................. vii

CHAPTERS

1 GENERAL INTRODUCTION ........................ ...... 1

Statement of the Problem ......................... 1
Plan of the Dissertation......................... 4
Definitional Considerations...................... 5
A Review of Mating Systems ....................... 6
Puberty Modulation in House Mice
(Mus musculus)................................. 7
The Social Biology of Voles (Microtus)............ 16
Puberty Modulation in Voles (Microtus)........... 19
Principles Underlying Puberty Modulation
in House Mice and Voles........................ 27
Problems with Previous Studies of Puberty
Modulation in Microtus......................... 38

2 GENERAL METHODS: SUBJECTS, HOUSING, AND
APPARATUS...................................... 41

3 EFFECT OF OLFACTORY CUES UPON PUBERTY
MODULATION IN FOUR SPECIES OF VOLES
(EXPERIMENT 1).................................. 43

Rationale. ....................................... 43
Method. .......................................... 45
Subjects. ...................................... 45
Procedure ......................................... 46
Statistical Analysis........................... 49
Results .......................................... 50
Body Weight .................................... 51
Anogenital Distance............................ 51
Adrenal Weight................................. 52
Testes Weight.................................. 54
Seminal Vesicle Weight........................ 54
Ovarian Weight. ............................... 54
Uterine Weight. ............................... 54
Status of Vaginal Perforation................. 56
Delay until Vaginal Perforation............... 57
Vaginal Smears. ............................... 58









Discussion .............................. ...... 65
Body Weight..................................... 65
Anogenital Distance............................ 67
Adrenal Weight................................. 69
Testes Weight.................................. 70
Seminal Vesicle Weight.................. ...... 71
Ovarian Weight................................. 72
Uterine Weight. ................................ 73
Characteristics of Vaginal Perforation........ 75
Vaginal Cytology............................... 77
Conclusions (Experiment 1)...................... 80

4 BEHAVIORAL RESPONSES OF VOLES (MICROTUS)
TO PUBERTY MODULATING STIMULI
(EXPERIMENT 2).................................. 116

Rationale........................................... 116
Method.............................................. 121
Subjects...................................... 121
Procedure...................................... 121
Statistical Analysis........................... 123
Results ......................................... 124
Body Weights. ................................. 229
Vaginal Smears................................. 231
Preference Tests.............................. 124
Differences in the Time Near the Stimuli
Across Weeks.................................. 126
Between-Species Comparisons................... 127
Duration within Center of Cage:
Within-Species Analyses..................... 130
Duration within Center of Cage:
Between-Species Analyses .................... 130
Discussion.......................................... 132
Olfactory Preference Tests .................... 132
Between-Species Comparisons................... 134
Duration within Center of Cage................ 136
Conclusions (Experiment 2)...................... 137

5 EFFECT OF SIRE PRESENCE OR ABSENCE ON
DEVELOPMENT OF OFFSPRING (EXPERIMENT 3)....... 145

Rationale........................................... 145
Method.............................................. 151
Subjects. ..................................... 151
Procedure...................................... 151
Statistical Analysis.......................... 152
Results ......................................... 153
Delay to Produce Litters...................... 153
Number of Offspring Born....................... 155
Age 50% of Offspring Opened Eyes............... 156
Number of Offspring Weaned.................... 157
Sex Ratio of Offspring Weaned................. 157
Body Weight of Offspring Weaned............... 158









Discussion. ..................................... 161
Delay to Produce Litters....................... 161
Number of Offspring Born....................... 163
Age 50% of Offspring Opened Eyes.............. 164
Number of Offspring Weaned.................... 165
Sex Ratio of Offspring Weaned................. 166
Body Weight of Offspring Weaned............... 167
A Reanalysis of the Evolution and
Expression of Male Parental Behavior......... 169

6 GENERAL OVERVIEW AND DISCUSSION................. 183

Overview of Experimental Results................ 183
An Attempt at a Synthesis........................ 183
Routes of Future Study........................... 190

REFERENCES............................................ 195

APPENDICES

A BODY WEIGHTS OF MICROTUS (EXPERIMENT 1).......... 210

B ANOGENITAL DISTANCES (EXPERIMENT 2).............. 215

C ANALYSES OF NUCLEATED CELLS AND
LEUKOCYTES FROM VAGINAL SMEARS OF
MICROTUS (EXPERIMENT 1)....................... 218

D BODY WEIGHTS OF MICROTUS (EXPERIMENT 2).......... 229

E ANALYSES OF CELLS FROM VAGINAL SMEARS OF
MICROTUS (EXPERIMENT 2)....................... 231

BIOGRAPHICAL SKETCH................................... 234








Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

STUDIES OF GENERAL AND SEXUAL DEVELOPMENT
IN VOLES (MICROTUS)

By

Allen L. Salo

December, 1992

Chairman: Donald A. Dewsbury
Major Department: Psychology

Three experiments were conducted to investigate how

patterns of general and sexual development might be

correlated with the formation of social and mating systems

among four species of voles (Microtus). Species included

pine voles (Microtus pinetorum), prairie voles

(M. ochrogaster), meadow voles (M. pennsvlvanicus), and

montane voles (M. montanus). In Experiment 1, general and

sexual development were monitored as voles were exposed to

pheromones contained in the soiled bedding from family

groups, adult males, or adult females. Few significant

effects were found to be due to the treatment. Male pine

voles exposed to family or male bedding were significantly

heavier than those exposed to clean or female bedding. The

uteri of female montane voles exposed to clean or male

bedding were heavier than the uteri of those exposed to

family or female bedding.

In Experiment 2, the olfactory preferences of voles

were measured when they were exposed to male and female

bedding on weeks 4, 7, and 10 after birth. Few preferences








were shown for either bedding type by any of the species.

Female prairie voles and meadow voles revealed a significant

preference for male versus female bedding. Both sexes of

all species differed little in the total duration they

remained near the female stimulus. Male meadow voles

remained near the female stimulus significantly less than

the males of the other species on week 10. Female montane

voles remained near the male stimulus significantly longer

on weeks 4 and 7 than did females of the other species.

In Experiment 3, the influence of the fathers' presence

and absence was studied during the rearing of the breeding

pairs' first two litters. Pine voles produced their second

litter considerably earlier if the male had been present

during the rearing of the first litter rather than being

absent. Pine voles weaned heavier offspring in the second

litter than in the first, when the male had been present for

the rearing of the first litter. Montane voles produced

litters that were male-biased in sex ratio across both

litters, if the male had been present during the rearing of

the first litter.

Results are discussed and interpreted from known

differences in their contrasting social and mating systems.


viii














CHAPTER 1
GENERAL INTRODUCTION

Statement of the Problem

Several theories have been advanced within the last two

decades to account for the evolution of different social and

mating systems among birds and mammals (e.g., Emlen & Oring,

1977; Kleiman, 1977; Orians, 1969; Vehrencamp & Bradbury,

1984; Wittenberger 1979). However, not all of the selective

forces that underlie the evolution of different mating

systems have been identified (see Vehrencamp & Bradbury,

1984). Wittenberger (1979) proposed that because mating

behavior is affected by nearly all aspects of an organism's

behavioral adjustment to its environment, a theory of mating

systems must be integrated with several other types of

behavioral theories. For example, theoretical advances

concerning the evolution of territoriality, parental

behavior, and sociality must be meshed within a broader

theory explaining the evolution of mating systems.

One area of study that appears promising for providing

greater insights into the evolution of social and mating

systems in mammals is research on the regulation of sexual

maturation or puberty modulation. For example, although its

function remains obscure, delayed sexual maturation is often

present among mammalian species that are considered to be

monogamous (Kleiman, 1977; Dewsbury, 1981). Some have








suggested that delayed sexual maturation functions to

inhibit incestuous mating (McGuire & Getz, 1981), although

others have suggested it reduces susceptibility to predation

(Batzli et al. 1977). Investigating the causes and function

of changes in sexual maturation may lead to an increased

understanding of the evolution of different social and

mating systems.

A second area of investigation that may offer insight

into the formation of the different mating systems is the

evolution of male parental care. Paternal care is found

most often among mammalian species that form monogamous

mating systems and have relatively few offspring (Kleiman,

1977; Dewsbury, 1981). Although it is often assumed that

paternal behavior has a beneficial effect on the development

of young, the assumption needs to be validated (Wuensch,

1985). The results of several studies conducted to

investigate this hypothesis among species of rodents are

equivocal (Dewsbury, 1988). It is possible that there are

important links between sexual suppression and the evolution

of paternal behavior, because the two often occur jointly in

monogamous mating systems (Kleiman, 1977; Dewsbury, 1981).

Until recently, the study of developmental processes

has been largely divorced from studies of the evolution of

mating systems. Muller (1990) noted that ontogeny has been

treated as a type of "black box" in evolutionary theory.

Similarly, Stearns (1989) discussing developmental

processes, noted that the types and sources of phenotypic








variation have been given little consideration in

evolutionary theory. The following set of experiments was

designed to investigate possible causal links between

puberty modulation and paternal care and the resulting

differences in social and mating systems in mammals.

Species of the genus Microtus are ideal for studying

the functions of sexual maturation and paternal behavior for

three primary reasons. First, the species display

differences in social and mating system that appear to be

associated with puberty modulation and paternal care. For

example, the mating systems among Microtus range from

monogamy in prairie voles (M. ochrocaster) to promiscuity in

meadow voles (M. pennsylvanicus)(Wolff, 1985). Pine voles

(M. pinetorum) and prairie voles are considered to be

monogamous and appear to be sensitive to the actions of

pheromonal cues that affect sexual maturation (Getz &

Hofmann, 1986; FitzGerald & Madison, 1983; Carter & Getz,

1985).

A second reason for studying these processes in

Microtus is that by investigating differences in puberty

modulation and paternal behavior with closely related

species, we are most likely to identify the selective

pressures and mechanisms that have shaped these species

differences (Clutton-Brock & Harvey, 1984; King, 1970;

Dewsbury, 1990). A third reason is that many species of

Microtus can be bred and maintained within the laboratory

where it is possible to systematically vary and control








exposure to stimuli. Such identification and control of

stimuli is difficult or impossible with these species under

natural conditions. The control offered in the laboratory

appears to be a prerequisite to identify the necessary and

sufficient stimuli that influence sexual maturation among

species. Finally, although there is a considerable

literature on such topics as the taxonomy, zoogeography,

anatomy, and habitats, of various species of Microtus,

currently little is known about many developmental processes

such as prenatal development, causes of interspecific

variation in litter size, and causes of mortality (Tamarin,

1985; Nadeau, 1985). It is possible that through the

systematic investigation of such phenomena, relationships

will become evident between them and the selective pressures

within the environment which shaped them.

Plan of the Dissertation

Three studies are presented that were designed to

investigate the role of two developmental phenomena that

might have influenced the evolution and expression of the

different social and mating systems among four species of

voles (Microtus). The two developmental phenomena include

the changes in the timing of puberty (puberty modulation)

and influence of paternal presence or absence upon

developing offspring. Specifically, Experiment 1 was

designed to explore how exposure to naturally occurring

pheromones, contained within soiled bedding, influences the

timing of puberty in four species of voles. The species








included pine voles (M. pinetorum), prairie voles,

(M. ochrogaster), meadow voles (M. pennsylvanicus), and

montane voles (M. montanus). In Experiment 2, the

behavioral reactions of the four species were studied when

they were exposed to soiled bedding in an odor preference

task. The study was designed to determine if behavioral

preferences for male and female soiled bedding were evident

within each species and sex and whether the preferences

changed as a function of age. In Experiment 3, the effect

of an adult male's presence or absence on his developing

offspring was studied. Together the three experiments were

designed to explore how these phenomena were related to

development and how they might be functionally linked to the

expression of different social and mating systems.

Below I present some definitional considerations,

present a review of the literature on puberty modulation in

house mice (Mus musculus) and voles (Microtus), present a

review of the social biology of Microtus, and present the

results of the studies designed to investigate puberty

modulation and the effects of paternal presence in Microtus.

Finally, I attempt to synthesize the results of the three

studies with results from other studies to explain possible

selective pressures that might have formed and may maintain

the different social and mating systems among these species.

Definitional Considerations

There is no globally accepted definition of puberty.

According to Hasler (1975), puberty has usually been defined








as the age that animals produce viable gametes. However, in

practice, it is often difficult to define the limits of

puberty with precision (Bronson & Rissman, 1986). Common

measures used in studies with female house mice have

included the day of vaginal opening or the first day of an

estrous vaginal smear (Drickamer, 1986). It has proven

difficult to find a reliable and non-terminal marker to

assess puberty in male mice. The presence of viable

spermatozoa or the ability to induce pregnancy have been

used in some studies as an index of puberty among males

(e.g., Vandenbergh, 1971).

Despite debate over how one defines puberty or what one

uses as an index of puberty, by choosing a standard measure

it is possible to compare differences in the onset of

"puberty" as being either accelerated or delayed compared

with some reference group. Thus, the term "puberty

modulation" is useful when referring to the onset of puberty

as being either accelerated, delayed, or both, when

comparing individuals or groups.

A Review of Mating Systems

The mating system of a population can be regarded as

the ensemble of behaviors and physical adaptations specific

for mating that are available to a population (Vehrencamp &

Bradbury, 1984). The mating system of a population is an

emergent property that reflects the traits and propensities

of the individuals within the population (West-Eberhard,

1979). For example, characteristics of a monogamous mating








system among mammals include (1) the continual close

proximity of an adult heterosexual pair during and outside

periods of reproduction, (2) mating preferences, (3) the

absence of unrelated adult conspecifics from the breeding

pairs home range, and (4) breeding by only one adult pair in

a family group (Kleiman, 1977). Other common forms of

mating systems include polygyny, in which there is a

prolonged association and essentially exclusive mating

relationship between one male and two or more females at one

time, and promiscuity, in which there is no prolonged

association between the sexes and multiple matings by

members of at least one sex (Wittenberger, 1979). The goal

of research has been to identify the primary selective

pressures or foci that have created the behavior differences

associated with each mating system, with the use of

appropriate theory and field work (Vehrencamp & Bradbury,

1984).

Puberty Modulation in House Mice (Mus musculus)

Most investigations of the phenomenon of puberty

modulation have been conducted with house mice

(Mus musculus). Researchers have identified several factors

that can influence the timing of puberty in young female

mice. Many studies have used indices such as the day of

vaginal perforation or the detection of an estrous smear as

a reliable means of assessing puberty. Considerably less

research has been designed to investigate the factors that

affect puberty in male mice. However, research with both








sexes appears necessary to understand how differences in the

timing of puberty might affect the expression of social and

mating systems.

A number of factors have been shown to influence the

timing of puberty in female house mice. Broadly, these

factors have included genetic differences (Drickamer, 1981),

social conditions (see Drickamer, 1986 or Vandenbergh &

Coppola, 1986 for recent reviews), and non-social factors

such as temperature (Barnett & Coleman, 1959), photoperiod

(Drickamer, 1975a), and season (Kruczek & Gruca, 1990). The

current review and experiments have been based on the

investigation of social factors that have been shown to

influence puberty. Specifically, chemosignals or

pheromones, found in the urine of several muroid species,

have been shown to differentially accelerate or retard the

attainment of puberty in young individuals (Levin &

Johnston, 1986).

At least four types of urinary chemosignals can

influence puberty in female mice (Drickamer, 1986). These

signals include urine from (1) male mice, (2) pregnant or

lactating females, (3) females in estrus, and (4)

group-caged females. The first three signals accelerate the

timing of puberty relative to female mice housed similarly

but not exposed to the chemosignals. The last signal delays

puberty in females compared to those not exposed to the

signals. Below, I present the principal findings associated








with each of the four stimuli that affect puberty in female

house mice.

Puberty Acceleration Caused by Male Urine

Not all urine from male mice is effective in

accelerating the onset of puberty in females relative to

those treated similarly but not exposed to the male urine.

Dominant males release an acceleratory substance in their

urine whereas subordinate males do not (Drickamer, 1983a).

Similarly, prepubertal or castrated males do not release an

acceleratory substance in their urine (Vandenbergh, 1969;

Lombardi et al. 1976). Neither the specific type of housing

condition nor the degree of genetic relatedness appears to

mediate the effectiveness of the male pheromone directly.

Drickamer (1983a) found that increasing the density of males

does not alter the pattern of the chemical's release. In

general, the effectiveness of the male chemosignal in

causing pubertal acceleration in females is robust.

Exposing females to amounts of male urine as small as 0.0001

cc per day effectively advances puberty in mice (Drickamer,

1982b, 1984a).

Despite the chemical's apparent acceleratory strength,

such factors as the season or ratio of light-dark exposure

can affect the female and thus indirectly affect whether the

male's chemosignal advances puberty. The substance appears

most effective in producing acceleration when it is

collected from males during light onset and presented to

females at this same time period (Drickamer, 1982a). The








effectiveness of the male chemosignal appears to vary

because of changes in sensitivity of young female mice to

the substance and not due to differences in the substance

excreted by males (Drickamer, 1986). Both male and female

mice have unique patterns of urine deposition that could

influence the likelihood that puberty will be modulated in

juveniles (Drickamer, 1989a).

Certain social conditions can influence the

effectiveness of the acceleratory substance to induce

puberty in females. Vandenbergh (1967) showed that the

acceleratory effect is greater when males are in physical

contact with females; contact-stimulation appears necessary

for the effect to occur (Drickamer, 1974, 1975b; Bronson &

Maruniak, 1975). The active chemical in male mouse urine

appears to be relatively nonvolatile and under natural

conditions it can influence females for some time after the

male has deposited urine (Drickamer, 1986). The substance

in male urine has been found to be effective in accelerating

puberty within three days of exposure (Colby & Vandenbergh,

1974). Acceleration can be accomplished by exposing females

to either male urine for two hours per day or to an intact

adult male for one hour per day (Drickamer, 1983a). Thus,

these findings show that although the age at which female

house mice reach puberty can be reduced by the actual

presence of an adult male, his presence is not necessary for

the acceleration of puberty.








The male chemosignal appears to have its physiological

influence in females via the vomeronasal-accessory olfactory

system (Drickamer & Assmann, 1981). This system may be a

common pathway affecting puberty modulation in various

species. Ablation of the vomeronasal organ eliminates the

ability of both mice and voles to receive pheromonal signals

(Vandenbergh, 1988; Lepri & Wysocki, 1987). Other work

investigating the physiological processes of male-induced

puberty acceleration has been conducted but will not be

reviewed here (see Reiter, 1982; Carter et al. 1986 for

review).

Puberty Acceleration Associated with Pregnant and/or

Lactating Females

Drickamer and Hoover (1979) first documented that the

urine of pregnant or lactating females accelerates sexual

development in female house mice. It is believed that the

active chemosignal is the same that is produced by females

in both reproductive conditions, because a number of

experiments revealed no clear differences between the

chemosignals (Drickamer 1986). Three days of exposure to

urine collected from either pregnant or lactating females is

necessary for puberty acceleration to occur in females that

are less than 30 days old (Drickamer, 1984b). However, the

substance does not cause acceleration in females when it is

presented during the winter months. This lack of an

acceleration in puberty appears due to differences in the

sensitivity of young females to the cues that correspond to








changes in the season (Drickamer, 1986). The acceleration

of puberty does not appear to differ as a function of

kinship among females (Drickamer, 1984c).

Under certain conditions, urine from reproductively

active females does not cause puberty acceleration in

females. For instance, when reproductively active females

are caged in groups or housed with non-reproductively active

females, their urine becomes ineffective in causing either

acceleration or delay in recipient females (Drickamer,

1983b). If young female mice are exposed to urine from both

reproductively active females and urine from group-caged,

non-reproductively active females, the exposed females are

delayed in reaching puberty (Drickamer, 1982c; 1986).

The chemosignals associated with reproductively active

females appear to be relatively volatile and are effective

in causing acceleration when presented in volumes slightly

larger than those of the male signal (0.03 cc of urine per

day; Drickamer & Hoover, 1979; Drickamer 1982b, 1983c).

One difference between the acceleratory chemosignals of

males and those of reproductively active females is that the

chemosignals from females are effective in causing

acceleration regardless of when they are collected or what

time of the day they are presented to females. As noted

previously, the urine from males is most effective in

accelerating puberty when it is collected and presented to

females during the onset of the light portion of the

light-dark cycle (Drickamer, 1982a).








Puberty Acceleration Caused by Females in Estrus

Less is known about how the particular phase of the

estrous cycle influences the chemosignals of the female

donors that can influence puberty in other females.

However, urine from singly-caged female mice in estrus

decreases the latency until puberty is reached in young

female mice. This effect is not seen when urine from

single-caged diestrous females is presented to prepubertal

females (Drickamer 1982c, 1984c).

The chemosignal in urine produced by estrous female

mice is effective in advancing puberty in about 3 days and

is effective when presented in small quantities such as

0.001 cc of urine per day (Drickamer, 1986). No differences

in acceleratory effects were found when urine was collected

from related or unrelated estrous females (Drickamer,

1984c).

Puberty Inhibition Caused by Group-Housed Females

Only one type of endogenously produced chemosignal has

been found to inhibit or delay puberty reliably in female

mice. Group-housed females produce a substance in their

urine that delays puberty in other females. The time that

estrous smears are detected in young female mice is delayed

when they have been exposed to either soiled bedding of

group-housed females or have had their nares painted with

urine from group-housed females (Drickamer, 1982c).

Females exposed to the urine of group-housed females

will begin to produce the inhibitory substance themselves.








The delay signal's effect occurs later than the effect of

any acceleratory chemosignal. Specifically the delay signal

requires four to seven days of exposure for inhibition to

occur (Drickamer, 1977). Season also influences the degree

of sensitivity young females have to this signal. Females

are not delayed in reaching puberty when exposed to the

substance during the summer, although during other seasons

they are delayed in its presence (Drickamer, 1986).

Puberty Modulation in the Natural Environment

Although the phenomenon of puberty suppression or

advancement is clearly established for house mice in the

laboratory, it is appropriate to question whether similar

processes occur within wild populations. A few studies have

provided evidence that both puberty acceleration and

suppression take place under natural conditions.

Massey and Vandenbergh (1980) conducted a series of

experiments with populations of wild house mice enclosed

within highway cloverleaf sections over a 2-year period.

The urine from the females of two populations was collected

via live-trapping during the spring, when population

densities were low. This urine did not have an effect on

the age of first vaginal estrus of laboratory-housed

females. However, urine collected from the second

population in December, when the population was crowded,

significantly delayed the first estrus in females when

compared with control females. Thus, the researchers








attributed the delay of puberty as being due to the

increased density in the second population.

Urine collected from wild male mice of the highway

"island" populations accelerated puberty in laboratory

females by an average of seven days as measured by age at

first estrus. Puberty acceleration occurred in response to

the male urine despite changes in the season or population

density at the time the urine was collected (Massey &

Vandenbergh, 1981). Thus, it appears that male chemosignals

and subsequent acceleration may be a more prevalent

influence upon puberty in females than the delay chemosignal

produced by female mice under crowded conditions.

In another set of experiments, acute population

explosions were created by Vandenbergh and Coppola (1986)

who introduced 40 second- or third-generation wild female

house mice onto highway islands. This procedure allowed for

a more critical test of the causal relationship between

population density and the release of delay chemosignals.

Urine samples were collected from females on each island

during monthly intervals. Urine samples collected three

weeks after the addition of the females onto the island

populations caused an average delay in first vaginal estrus

of 5.3 days in laboratory-reared females. Thus, this study

provided additional evidence that the puberty-delay

pheromone can be produced by females in response to acute

increases in female density under natural circumstances

(Vandenbergh & Coppola, 1986).








The Social Biology of Voles (Microtus)

A considerable amount of information exists on the

social and mating systems of several species of voles

(Microtus; see Wolff, 1985; Keller, 1985 for reviews). The

social behavior of Microtus is both complex and variable

(Wolff, 1985). Most species are active both during the day

and night, and make use of above-ground runways as well as

burrows. However, there are substantial differences in

patterns of territoriality and of mating system. Below are

listed some of the primary social differences among four

species of Microtus, which are the four species studied in

the following experiments. Broadly, their differences

portray the degree of diversity within the genus Microtus.

Montane voles. Montane voles (M. montanus) are

considered to commonly form a polygynous mating system,

although facultative monogamy may occur at low densities

(Jannett 1980; 1982). Both males and females typically

defend exclusive territories against other same-sexed

individuals (Jannett, 1982). The territories of males

generally overlap the territories of one or more females.

Thus, female montane voles appear typically to mate with

only one familiar male, although no pair bond is formed

(Wolff, 1985). Montane voles become more social and form

aggregations during the winter months (Madison, 1984). The

adult males of the species are typically larger than the

females (males 35% larger; Dewsbury et al. 1980).








Meadow voles. The mating and social system of meadow

voles (M. pennsylvanicus) differ in some respects from

montane voles, and many data have been gathered on their

social behavior (e.g., Madison, 1980, 1984; Webster &

Brooks, 1981). Meadow voles appear to form a promiscuous

mating system in which reproductively active females defend

territories. In contrast to montane voles, the home ranges

of males may overlap those of several other males as well as

those of females (Wolff, 1985). Cases of one female mating

with several males have been reported (e.g., Webster &

Brooks, 1981). Meadow voles form aggregations during the

late fall and winter months as do montane voles (Madison,

1984). The degree of sexual dimorphism in body weight is

also relatively high, with adult males being larger than

females (males 22% larger; Dewsbury, et al. 1980).

Prairie voles. In contrast to montane voles and meadow

voles, prairie voles (M. ochrogaster) appear to be more

social and form a monogamous mating system; a communal

nesting group appears to be the basic year-round social unit

(Getz et al. 1990). Evidence from the laboratory also

suggests monogamy. Both the breeding male and female engage

in extensive parental care of offspring (Getz & Carter,

1980), and male prairie voles failed to demonstrate the

Coolidge effect when presented with a new estrous female

after reaching sexual satiety (Gray & Dewsbury, 1973). Both

males and females appear to be territorial and are

aggressive to animals of the opposite sex. Both sexes of








breeding pairs defend a common home range that is

approximately the same size (Gaulin & FitzGerald, 1988).

The degree of sexual dimorphism in body weight is reduced in

comparison to the dimorphism in montane voles and meadow

voles, although the male is somewhat larger than the female

(males 17% larger; Dewsbury et al. 1980).

Pine voles. Pine voles (M. pinetorum) also appear to

be a highly social species and appear to live in communal

groups (FitzGerald & Madison, 1983; Schadler, 1990). The

mating system of pine voles appears to be monogamous,

although cooperative polyandry has been suggested as an

alternative system (FitzGerald & Madison, 1983). However,

more recent evidence suggests that breeding is commonly

restricted to the founding parents of a communal group

(Schadler, 1990). The existence of a monogamous mating

system had also been predicted to be most likely for pine

voles among eight species of Microtus reviewed by Dewsbury

(1981). Other evidence of monogamy includes male

participation in the rearing of young, including the

retrieval and brooding of infants (Shadler, 1990; Oliveras &

Novak, 1986).

In contrast to many other species of voles, pine voles

are almost entirely fossorial, and come to the surface only

occasionally to feed (Wolff, 1985). Pine voles also appear

to be territorial. FitzGerald and Madison (1983) found that

each family had a discrete non-overlapping territory from

other family groups. Differences in body weight between the








sexes appear minimal in pine voles (males 2% smaller;

Dewsbury, 1990).

Puberty Modulation in Voles (Microtus)

Although several researchers have clearly demonstrated

that puberty delay and acceleration occur in a few species

of Microtus, such as prairie voles (M. ochrogaster; Carter

et al. 1986) and California voles (M. californicus; Batzli

et al. 1977), others have made weaker claims that such

processes do or do not occur in other species of Microtus.

For example, Jannett (1978) referred to field evidence that

suppression occurs in montane voles (M. montanus) and Batzli

et al. (1977) claimed that suppression does not normally

occur in meadow voles (M. pennsylvanicus). Below, I review

the primary literature regarding the four species of

Microtus that were used in the later series of experiments.

Prairie voles. Most research concerning puberty

modulation in Microtus has been conducted with prairie

voles, where the presence of puberty suppression and

acceleration are strongly supported (see Carter et al, 1986

for review). In an early study that indicated puberty delay

occurs in prairie voles, Hasler & Nalbandov (1974) examined

pairs of weanling females caged with males with different

characteristics. Weanling females that were caged with a

littermate male had significantly longer latencies until

vaginal opening and production of a first litter than

compared to females that were caged with either a

non-littermate male or an adult male. The range of days








until vaginal opening occurred among these groups was

substantial. Females kept with littermate males became

perforate 30 days after pairing, on average, compared to

females paired with nonlittermate males that became

perforate eight days after pairing.

Batzli et al. (1977) found evidence that both sexes of

prairie voles could be suppressed developmentally, as

assessed by differences in body weight, when they were

housed with littermates. During this period, females

remained vaginally imperforate and males had abdominal

testes. After siblings were paired with unfamiliar adults

of the opposite sex, rapid increases in body weight and

subsequent reproduction occurred.

The stimuli necessary for the acceleration of puberty

in prairie voles have been studied. Young female prairie

voles, when exposed to a sexually experienced male for a 1 h

period, showed significant increases in uterine weight as

soon as 2 days after exposure (Carter et al. 1980). Drops

of male urine applied to the upper lip of young females

caused significant increases in uterine weight compared with

females caged alone, with a female sibling, a castrated

male, or exposed to the urine of a castrated male.

In another series of experiments, kinship per se was

shown not to be a limiting factor for the advancement of

puberty in prairie voles (Carter et al. 1980). When urine

from male siblings was applied to the nares of females for 6

consecutive days, it led to significantly heavier uterine








weight in females, compared to females housed with either a

sibling male or with a sibling male and treated with water

placed on the nares.

Together, the results of studies with prairie voles

suggest that a chemical agent or pheromone present in male

urine can be passed from male to female by direct contact.

From behavioral observations, Carter et al. (1980) suggested

that the active male pheromone was transmitted by

naso-genital contact between the sexes. However, sibling

prairie voles rarely engage in naso-genital investigation

and thus the lack of investigation between siblings may

function as a barrier to reproductive activation and

incestuous matings (Carter et al. 1986).

Although it is not clear how siblings inhibit

reproduction in other siblings, young female prairie voles

appear to be a primary source of reproductive suppression

for other male and female siblings. Getz et al. (1983)

found that for 15 suppressed litters, all but 1 of 32

females (or 97%) reared with sibling females were

suppressed, while only 3 of 15 (20%) reared with sibling

males were suppressed. A similar finding was apparent in

males; 10 of 11 males (91%) reared with sibling females were

suppressed and only 3 of 9 males (33%) raised with sibling

males were suppressed. The results suggest that for prairie

voles, the sex of siblings can have a substantial effect

upon the timing of puberty.








Additional work with prairie voles has revealed that

the suppressive effect of female presence appears largely

due to chemosignals present in their urine (Getz et al.

1983). In this study, virgin prairie voles were first

reproductively stimulated by exposing them to an unfamiliar

sexually experienced male for a one-hour period, then the

females were housed in a variety of conditions. The uteri

were weighed 48 hours later. The uterine weights of females

that had urine from either a female sibling, a virgin

non-sibling, or a pregnant female, placed on their nares did

not differ significantly from the uterine weights of

non-stimulated females. The only significant increase in

uterine weight was in the group of females that were

stimulated by the male and then maintained alone. Thus, the

results indicate that urine from female prairie voles is

effective in counteracting or suppressing puberty in female

prairie voles.

Although it is not known how the chemosignal is

transmitted between females, activation by the non-volatile

male acceleratory pheromone appears to be caused by mutual

anogenital investigation between the sexes (Carter et al.

1980). Hofmann and Getz (1988) found that virgin prairie

voles that were exposed frequently to unfamiliar males could

"override" the reproductive suppression typically

experienced by females that remained within a family group.

Thus, it seems plausible that under natural conditions where

density is dramatically increased, such as during the growth








phase of a population cycle, normal suppression can be

counteracted by male presence. In support of this argument,

Getz and Hofmann (1986) found that of free-ranging prairie

vole females that remained at the natal nest at low

population density, 18% were reproductively active, while

77% became reproductively active at high density. All

females that dispersed from the natal nest were found to

became reproductively active.

Pine voles. Both puberty delay and acceleration appear

to occur in pine voles, M. pinetorum (Schadler, 1983; Lepri

& Vandenbergh, 1986). Schadler (1983) presented evidence

that male siblings are an important source of reproductive

inhibition for female siblings. Mating was inhibited

between a female and an unrelated male, while a male sibling

was sequestered behind a wire mesh barrier. Other evidence

of puberty delay was shown by Lepri and Vandenbergh (1986)

when they placed young female pine voles with adult males

into two type of cages: one with clean bedding or another

with bedding that had been soiled for two weeks by the

female's family. Forty-eight hours later, the females'

uteri and ovaries were removed and weighed. The uteri and

ovaries of those placed in the clean cage were significantly

heavier than those placed in the cage soiled by the family.

In another experiment, Lepri and Vandenbergh (1986)

demonstrated puberty acceleration in pine voles by exposing

4-week-old females to feces and urine from animals in one of

the following groups: (1) intact males, (2) castrated males,








(3) singly-housed females, (4) group-housed females, or (5)

control (clean cage). When uterine and ovarian weights were

compared, results indicated significant increases in both

organs from females exposed to the stimuli from the males of

either groups compared those from the control females. No

significant differences were found when the organs of the

control group were compared to those in the two groups that

received female stimuli.

Meadow voles. Although there is evidence that pubertal

acceleration occurs in female meadow voles (Baddaloo &

Clulow, 1981), the evidence is equivocal for any type of

puberty delay. Pasley & McKinney (1973) presented evidence

that females caged in groups of eight had smaller ovaries

and uteri than singly-housed females. However, it is

possible that this type of sexual suppression was due to

overcrowding and stress and not due to pheromones. Batzli

et al. (1977) reported they did not find suppression to

occur in five of six litter-housed groups of meadow voles.

Additional data are needed to determine if sexual

suppression occurs in meadow voles as a result of pheromones

and not due to overcrowding and stress.

Montane voles. Fewer reports of pubertal acceleration

or delay can be found scattered in laboratory and field

reports for other species of Microtus. Field evidence

suggests that pubertal delay occurs in montane voles

(Jannett, 1978), although a critical test of this phenomenon

has not been reported. Recent laboratory evidence has shown








that puberty acceleration occurs in female montane voles

when exposed to males. Sawrey and Dewsbury (1991) found

shorter latencies until vaginal perforation and first

cornified estrouss) smears in females housed across a

wire-mesh barrier from males, as compared with females

without a male present. Controlled studies are needed to

determine if puberty delay occurs in montane voles when they

are exposed to pheromones.

It is possible that not all species of Microtus show

both delay and acceleration of puberty, or at least are not

affected to the same degree by comparable stimuli. Evidence

suggests that suppression does not occur, or occurs only

weakly, in female tundra voles (M. oeconomus) when they are

housed with male siblings (Facemire & Batzli, 1983). Batzli

et al. (1977) reported morphological differences between

California voles and prairie voles. California voles that

were housed with littermates grew more slowly than controls,

although females became perforate and males became scrotal

at about the same time in the littermate group as did the

controls (30 days). This pattern differed from prairie

voles in which males typically developed abdominal testes

and females remained vaginally imperforate when littermates

were held together. In both species, however, reproduction

by littermates was usually delayed until they were paired

with unfamiliar animals. Thus, there may be a variety of

physiological and behavioral differences related to puberty

modulation among various species of Microtus. The








illumination of the similarities and differences in patterns

of puberty modulation may prove instrumental in

understanding the dynamics of their social and mating

systems.

Puberty Modulation in Male Voles (Microtus)

Considerably less research has been conducted on the

various factors that can influence puberty modulation in

male Microtus, although cues from males may be critical for

influencing puberty in other siblings among some species.

Evidence for California voles suggests that male siblings

influence the sexual development of one another. Batzli et

al. (1977) found that the growth rates of males were

suppressed when they were paired with another male sibling,

although growth rates were not affected when they were

paired with a female, a male non-sibling, or with a female

sibling.

Other evidence suggests that the odors from family

members of California voles suppress male pubertal

development. In this species, the odors from mature males

or sires did not appear either to accelerate or inhibit the

sexual development of males (Rissman et al. 1984). However,

male California voles that were reared in the presence of

family bedding material and subsequently paired with another

non-relative female failed to cause increases in uterine

weight within four days (Rissman & Johnston, 1985). They

suggested that the lack of stimulation in the females was

due to low levels of circulating androgen in the males.








Thus, odors associated with a reproductively active female

might be one source of pubertal suppression in male

California voles.

A few anecdotal accounts of puberty modulation among

males of other species of Microtus exist. For example,

juvenile male common voles (M. arvalis) have been reported

to show reduced testicular development when housed near

cages with crowded adults (Lecyk, 1967). In summary, it

appears that although males of various species of Microtus

may be reproductively modulated by specific stimuli, the

limited data make conclusions premature. There is a paucity

of information regarding pubertal modulation in male voles,

and for males of other rodent species as well. It seems

worthwhile to correct this bias with future studies.

Principles Underlying Puberty Modulation in House Mice

and Voles

There has been little effort to integrate the large

body of literature into an effective theoretical framework

to study the relationships between primer pheromones, age at

first reproduction, and demographics (Vandenbergh & Coppola,

1986). Many of the existing hypotheses regarding puberty

modulation are based upon studies of house mice (Mus

musculus) because most of the research has been conducted

with this species. Vandenbergh and Coppola (1986) suggested

that some inferences may be applied to other species, such

as voles. However, they caution that given the differences

in the reproductive biology among genera, it seems unlikely








that the selective pressures influencing puberty modulation

will be identical across them. Observed similarities among

genera or species must be interpreted carefully when

attempting to formulate general principles.

Below I list some of the most prominent frameworks that

appear useful for integrating our understanding of the

causes and functions of the pheromonal effects upon puberty

modulation found among species such as house mice and voles.

A Life-History Theory of Puberty Modulating Pheromones

One of the broadest theoretical frameworks to

investigate and interpret patterns of puberty modulation is

the life history theory proposed by Vandenbergh & Coppola

(1986). According to this approach, life history tactics

are evolved sets of coadapted traits designed to solve

particular ecological problems (Stearns, 1976). Whether

animals typically reach sexual maturity relatively early or

late, all are viewed to have been selected to maximize

lifetime reproductive success.

Selective factors favoring puberty acceleration.

Life-history theory predicts that for increasing

populations, or those with large fluctuations or repeated

episodes of colonization, early maturity will be favored

(Vandenbergh & Coppola, 1986). Such characteristics

describe feral house mouse populations. In general, it is

believed that selection will favor early and total

investment by individuals to produce the maximum number of

young whenever the environment is highly favorable for








reproduction. They note that previous theories (e.g., Cole,

1954; Lewontin, 1965) have indicated that in a rapidly

expanding population, the age at first reproduction should

be driven to the physiological minimum by natural selection.

Species that possess a combination of life-history traits

including an early age of first reproduction, the production

of many young, and semelparity have been classified as being

"r-selected" (MacArthur & Wilson, 1967; Pianka, 1970). This

combination of traits is often present in species that

experience rapid population growth in favorable

environments.

Several factors are thought to favor early

reproduction (Vandenbergh & Coppola, 1986). First, early

reproduction will be favored as reproductive costs decrease

(Schaffer & Elson, 1975). Reproductive cost is the

deleterious effect of present reproduction on future

survival and/or fecundity. Early reproduction will also be

favored when the reproductive value of individuals decrease

as they grow older (Gadgil & Bossert, 1970). These patterns

suggest that early reproduction will be favored when it does

not negatively influence a female's ability to reproduce

later in life and thus lower lifetime reproductive success.

Delayed reproduction (puberty) in house mice. The

function of delayed reproduction or puberty is more obscure

than the function of puberty acceleration (Vandenbergh &

Coppola, 1986). They suggest that it is difficult to

explain why opportunistic species, such as house mice,








benefit from delaying puberty and reproduction. However,

whatever the functions) of puberty delay, it has been found

in several mammalian species (Vandenbergh & Coppola, 1986).

There appear to be general conditions that favor

delayed reproduction. Delayed reproduction could evolve if

individuals gained fecundity or produced better quality

offspring (Vandenbergh & Coppola, 1986). In general, it is

believed that many of the demographic or environmental

factors that favor delayed reproduction are the opposite of

those favoring early reproduction. Some have suggested that

delayed reproduction would be selected in stable populations

at or near the carrying capacity of the environment (Cole,

1954; Lewontin, 1965) or in declining populations (Hamilton,

1966; Mertz, 1971). Selection in saturated environments,

which favors the ability to compete and avoid predators, has

been referred to as "K-selection" (MacArthur & Wilson, 1967;

Pianka, 1970). Traits correlated with this type of

selection often include late maturity, the production of few

and large offspring, long life spans, and extended parental

care.

There are a few causal factors that are thought to

underlie delayed reproduction (Vandenbergh & Coppola, 1986).

For example, as the reproductive costs increase, in terms of

adult mortality, and as the reproductive value an individual

can accrue by not reproducing increases with age, delayed

reproduction will be favored. If reproductive success is









contingent upon age, size, or social status, delayed

reproduction is also favored (Geist, 1971).

Despite the general conditions that have been proposed

to predict under what circumstances early or late puberty

will occur, the current state of life-history theory allows

us to draw only rather vague and imprecise conclusions about

the timing of reproductive maturity (Vandenbergh & Coppola,

1986). Despite this drawback, life-history theory provides

a broad theoretical background from which to interpret new

information. In addition to the previous conditions which

favor puberty acceleration or delay, there are a few

additional principles that may guide our understanding of

these processes. Below, I summarize two of the most

prominent themes which have emerged.

Puberty Modulation (Mutualism (Cooperation) versus Conflict)

One assumption that underlies most explanations for the

evolution of puberty modulation is that puberty modulation

is either of mutual benefit to the sender and receiver of

pheromones or that the sender and receiver of the pheromones

are in some form of conflict.

Examples of puberty modulation in the context of

cooperation. Bronson (1979) proposed that adult male mice

and adult or prepubertal females both gain in fitness

through mutual stimulation of reproductive acceleration via

pheromonal transfer. In this hypothesis, male urinary

pheromones cause the release of LH in recipient females that

speeds the attainment of puberty and ovulation.








Reciprocally, female pheromones cause increases in LH in

males which ultimately leads to increases in testosterone

and additional pheromone synthesis in males. Such a system

could greatly enhance the speed at which males and females

could reproduce, thus benefiting both sexes of breeding

pairs.

A second proposed benefit of mutualism was raised by

Vandenbergh and Coppola (1986) who proposed that the

detection of the puberty-delay and acceleratory pheromones

produced by adult female house mice provides information to

young females about the general quality of the environment

for reproduction. Presumably under crowded conditions,

adult females release puberty delay pheromones that benefit

both the sender and receiver to postpone reproduction. In

more favorable environments for reproduction, the release of

pheromones that accelerate puberty by females would hasten

the speed at which puberty and reproduction would occur

among female offspring. Presumably, the young females,

their mates, and kin would benefit from more optimally timed

reproduction.

A third example where mutual benefits may occur between

the pheromone senders and receivers is between parents and

offspring in some cooperatively breeding species (Emlen,

1984). Cooperative breeding refers to any situation when

more than two individuals provide care in the rearing of

young (Emlen, 1984). In this example, puberty delay and








inbreeding avoidance may have been co-evolved traits in

species where cooperative breeding occurs.

It appears that certain ecological constraints, such as

a saturated habitat, favor the retention of subadult

individuals within a family group, at least until conditions

enable independent reproduction. Emlen (1984) suggested it

is possible that when the probability of successful

dispersal and independent breeding are low, average fitness

may be increased by remaining in a group until group size

reaches some optimum. Other benefits that may occur in

cooperatively breeding groups include: (1) benefits to

helpers that gain breeding experience; (2) inheritance of

the parental territory; (3) dispersion in groups when

competition for reproductive vacancies is strong; and (4)

some form of reciprocity may take place, such as when one

individual forgoes breeding and helps another rear offspring

with likelihood of return aid at a later point in time; and

(5) increased inclusive fitness through aiding in the

rearing of close relatives (Emlen, 1984). Thus, both the

direct and indirect components of inclusive fitness might be

increased via reproductive suppression when accompanied by

helping behavior.

A final proposal for mutualistic benefit of

reproductive suppression, or the lack thereof, in voles has

been proposed by Christian (1970). Christian (1970)

emphasized the selective pressures exerted by the habitat

where a particular species evolved. Species such as meadow









voles, which evolved in habitats that were patchy and

ephemeral, such as moist meadows, evolved mechanisms that

aided in the colonization of newly created habitats. Thus,

mechanisms that led to successful dispersal were adaptive

for all members. He proposed that a density-dependent

endocrine response that led to increased aggression with

subsequent dispersal would serve this function. Hence, a

lack of sexual suppression among siblings would be expected

in species that evolved under these circumstances.

In contrast, species such as prairie voles that evolved

in the continuous and extensive habitats of the great plains

of central North America were believed to have evolved

mechanisms that enabled a greater level of social tolerance

and be more sensitive to reproductive inhibition (Christian,

1970). It is plausible that mechanisms to inhibit

incestuous mating among family members might be most evident

in highly social species.

Puberty modulation in situations of conflict.

Different researchers have proposed instances where there

appear to be conflicts of interest between pheromone senders

and receivers. Bronson (1979) proposed there was an

antagonism between female house mice to inhibit puberty and

reproduction. Such antagonism between females may be

accomplished either directly through aggression or through

the production of delay pheromones.

Wasser and Barash (1983) proposed a similar hypothesis.

They stressed the role of female-female competition to








reproduce as a key driving force to explain the high rates

of general reproductive failure among the females of many

species of mammals. The antagonistic pheromonal influence

of reproduction between female prairie voles conforms to

their theory (Getz et al. 1983)

Competition among the males of some species via

pheromones may also be present in some species. Although

Bronson (1979) noted the high rate of aggression observed

between territorial male house mice and other males, he did

not discuss pheromonal interactions among them. Vandenbergh

(1971) found that adult male house mice had an inhibitory

effect on the reproductive development in young males, while

the presence of adult females accelerated their development.

The possibility that males may inhibit puberty and

reproduction of other males in other species through

pheromonal communication should remain open for continued

investigation.

Socially-Dependent Versus Socially-Independent Systems

Another framework within which to view and investigate

the reproductive influences of pheromones on the timing of

puberty can be called the socially-dependent versus

socially-independent dichotomy. The essence of this

framework is that some species of Microtus appear to be more

dependent upon the direct exposure of other animals in order

to become reproductively mature and active. Taylor

(1990/1991) characterized two types of estrus induction

patterns among females of different species of Microtus. He









provided evidence that the highly social prairie voles,

could be characterized as being male-dependent. Females

must typically be exposed to stimuli from males for a

relatively long period of time before estrus is attained.

In contrast, other less social species, such as meadow voles

and montane voles, were referred to as being

male-independent (Taylor 1990/1991). Species of this type

do not attain estrus totally independent of male

stimulation, but rather are much less dependent on direct

male contact than are male-dependent species.

Puberty Modulation as an Artifact

An alternative viewpoint of the phenomena of puberty

modulation is that the effects are some form of laboratory

artifact. Examples of puberty modulation must be evaluated

for the possibility that differences in the timing of

puberty could be the result of close confinement with other

animals, unnaturally high densities, or a product of

artificial selection (see Bronson, 1979; Vandenbergh and

Coppola, 1986).

Bronson (1979) suggested that while the mechanisms

underlying the phenomena in question may be real, it is

possible that the mechanisms evolved to serve one adaptive

purpose in the wild, yet find expression in other ways in

the laboratory environment. Vandenbergh and Coppola (1986)

have argued against the proposal that puberty modulation is

some form of laboratory artifact in house mice. First, they

suggested that the delay of puberty in female mice which








results from a specific stimulus such as a urinary cue

suggests an evolved signalling function that must have some

adaptive value. A second reason is that the general social

context in which puberty delay occurs in the laboratory is

known to occur in the field. Descriptions of the social

organization of commensal mice generally fit the conditions

necessary for pubertal suppression to occur. The last line

of evidence comes from the field studies conducted with wild

house mice confined to clover-leaf highway populations

(Massey & Vandenbergh 1980; 1981, see prior section). These

studies show that at least under some conditions, the

production of delay and acceleratory pheromones can occur in

animals from free-ranging environments. Nevertheless,

additional studies and further demonstrations that puberty

modulation occurs among mice and other species such as voles

in the field or under semi-natural conditions seem

warranted.

Summary of Principles of Puberty Modulation in House Mice

and Voles

From the primary findings and related principles

presented for puberty modulation in species such as house

mice and voles, it is clear there is no all-encompassing

theory that leads to clear predictions when puberty should

and should not occur, or to the specific stimuli that should

cause modulation, and the specific functions) it serves.

However, the emerging picture is that a variety of stimuli

are capable of influencing puberty and multiple functions









might be served by puberty modulation among various species.

Vandenbergh and Coppola (1986) suggested that discovering

the role of priming pheromones in the interactive process of

life-history and behavioral adaptations will require the

"melding" of empirical and theoretical points of view.

Problems with Previous Studies of Puberty Modulation

with Voles

Scattered reports indicate that puberty modulation

occurs in a variety of species of Microtus. Some species

have been studied in considerable detail (e.g., prairie

voles, see Carter et al. 1986). However, several sources of

confusion have hampered our ability to explore possible

relationships between puberty modulation and the expressed

differences in the social and mating systems among Microtus.

First, the lack of common procedures and measures used by

various researchers appear to be significant, although

reasonably easy, problems to correct. A variety of measures

and procedures have been used to assess the timing of

puberty, ranging from the measurement of body weight to

analyses of hormone binding sites within the brain. Such

diversity of procedures make meaningful comparisons among

species difficult. For example, Dewsbury (1981) attempted

to evaluate the usefulness of several proposed correlates of

monogamy to predict its presence in several muroid rodents.

However, he deemed it was not possible to compare

meaningfully the existing reports for patterns of sexual

maturation across several species. Thus, with such









variation evident for sexual maturation among muroid

rodents, future studies with identical procedures appear

necessary to make useful comparisons among species.

A second problem is the lack of proper controls that

are necessary to identify the effective stimuli influencing

puberty among species. Cues such as the presence or absence

of an adult male or sire, the presence and number of

opposite-sexed siblings, and the presence or absence of a

reproductively active female have all been found to be

effective stimuli that can influence the timing of puberty

in one or more muroid species. Unfortunately, in many

studies such cues are not systematically controlled or

reported. The use of soiled bedding that is transferred

from one cage to another provides one means of critically

testing the assumption that pheromones produced by one or

more individuals cause puberty modulation in others. Such a

procedure controls for the potential influence that

nonolfactory stimuli, such as visual or auditory, could have

upon puberty modulation.

Finally, the database, which is a prerequisite to

determine the general patterns or principles of puberty

modulation, is still relatively small and, in some cases,

misrepresentative of the genus Microtus. For example,

although only a few studies have provided evidence of

puberty modulation in some species of Microtus such as

montane voles (e.g., Sawrey & Dewsbury, 1991), substantial

data exist for other species, such as prairie voles (Carter









et al 1986; Getz et al 1983). In addition, most studies on

Microtus, like those for mice (Mus), are female-biased.

Most studies have been designed to determine how olfactory

and social cues affect puberty in females rather than males

of various species. Ideally, the factors influencing the

onset of puberty would be studied concurrently for both

sexes. Such patterns should be viewed for the possibility

of significant interactions occurring between the sexes of a

given species.

Thus, whereas the present list of problems encountered

with studies of puberty modulation in Microtus is not

exhaustive, it outlines significant sources of ambiguity and

deficiencies in prior studies. Fortunately, with standard

procedures and proper controls used with a number of

species, a greater understanding of how puberty modulation

can impact the formation of contrasting social and mating

systems among Microtus might be possible.














CHAPTER 2
GENERAL METHODS: SUBJECTS, HOUSING, AND APPARATUS

Four species of voles (Microtus) were studied in each

of three experiments. Species studied included pine voles

(Microtus pinetorum), prairie voles (M. ochrogaster), meadow

voles (M. pennsylvanicus), and montane voles (M. montanus).

Breeding colonies of each species were maintained in the

Psychology Building at the University of Florida. These

facilities were accredited by the American Association for

Accreditation of Laboratory Animal Care (A.A.A.L.A.C.).

All subjects were born to breeding pairs that were

laboratory-reared and that had been derived from wild

populations within the United States. Efforts were made to

maintain genetic diversity while guarding against possible

inbreeding in all species. All species were kept in

separate colonies that were housed in windowless and

air-conditioned rooms that were maintained on a reversed

16:8 light-dark photoperiod with light onset at 2000 hr.

All colonies contained animals of both sexes throughout the

experiments on a combination diet of Rabbit Chow (Purina

Mills) and Laboratory Rodent chow #5001 (Purina Mills) with

water available ad libitum. In an effort to promote

continuous breeding and a sufficient supply of subjects, a

handful of lettuce was given to breeding pairs on a weekly

basis. Breeding animals used in Experiment 3 similarly
41






42


received these weekly supplements, although no subjects used

in Experiments 1 and 2 received lettuce supplements during

the duration of their test phase.

All subjects were maintained in large 48 X 27 X 13 cm

polycarbonate cages unless they were housed individually, in

which case they were housed in 29 X 19 X 13 cm polycarbonate

cages. All subjects used in the studies were used in only

one experiment.















CHAPTER 3
EFFECT OF OLFACTORY CUES UPON PUBERTY IN FOUR SPECIES OF
VOLES (EXPERIMENT 1)

Rationale

Experiment 1 was designed to determine if puberty delay

and acceleration occur solely as a function of the presence

of olfactory cues (i.e., pheromones) that are contained in

the soiled bedding in each of four species of voles

Microtus. In several studies where puberty modulation has

been reported, researchers have either housed animals

directly with or across from other animals (e.g., caging

juvenile females with, or across a wire-mesh screen from,

adult males). Although olfactory cues and pheromones are

believed to be key stimuli that produce changes in the

timing of puberty, many procedures have not been designed to

exclude the possibility that other stimuli critically affect

differences in the timing of puberty (e.g., visual,

auditory, or somatosensory cues).

The use of soiled bedding, which is transferred between

donor and recipient animals, has been shown to be one

effective method to expose animals to pheromones while

controlling for exposure to other non-olfactory stimuli that

could influence sexual maturation. Soiled bedding has been

used successfully to reveal changes in reproductive

development among house mice (Drickamer, 1982c), pine voles








(Lepri & Vandenbergh, 1986), and California voles (Rissmann

et al. 1984; Rissmann & Johnston, 1985).

Measures and Predictions

Specific changes in the physiology of male voles that

would indicate sexual suppression would include reproductive

organs that weighed less than the organs of subjects exposed

to the clean bedding. In addition, the anogenital distance

among males considered suppressed would be smaller than the

anogenital distance in those exposed to clean bedding.

Measures that would reflect sexual suppression in female

voles, would include delayed latency until vaginal

perforation and smaller percentages of cornified cells in

the vaginal smears compared to those exposed to clean

bedding. Among several species of voles, higher proportions

of cornified cells are associated with a higher incidence of

estrus and sexual receptivity (Sawrey & Dewsbury, 1985;

Taylor et al. in press). In contrast to indices of sexual

suppression, indices of sexual advancement or acceleration

would include the opposite changes to those listed above for

sexual suppression.

Because of prior evidence that the highly social female

pine voles and prairie voles typically require direct male

contact for full reproductive activation (e.g., Carter et

al. 1987; Schadler & Butterstein, 1979), I predicted these

species would not be affected, or only minimally, when

exposed to the soiled bedding material. These species may

be considered socially dependent for reproductive activation








(i.e., socially-dependent or "male-dependent" species, see

Taylor, 1990/1991).

In contrast, the females of the less social species

(i.e., socially-independent species), meadow voles and

montane voles, have been shown to become sexually receptive

with little or no previous direct exposure to males or to

show marked changes in reproductive physiology through

exposure to male urine alone (Taylor, 1990/1991; Baddaloo &

Clulow, 1981). Field evidence has suggested that both male

and female montane voles are sexually suppressed when they

remain in natal family groups under high densities (Jannett,

1981). Thus, meadow voles and montane voles were predicted

to show marked responses to the presence of the soiled

bedding material. specifically, they were predicted to be

sexually suppressed when exposed to odors from the natal

family or from adults of the same sex, whereas they would be

sexually accelerated when exposed to odors from unfamiliar

opposite-sexed individuals. Young animals not exposed to

odors from others (e.g., clean bedding) were expected to

show intermediate rates of growth relative to the rates of

those in the other conditions.

Method

Subjects

A total of 514 animals were used in Experiment 1, 256

were male and 258 were female. All were born to existing

laboratory stock maintained at the University of Florida.

The four species included pine voles (Microtus pinetorum),








prairie voles (M. ochrogaster), meadow voles

(M. pennsylvanicus), and montane voles (M. montanus). All

measures were recorded between September 1990 through

November 1991.

Procedure

Animals from each of the four species of Microtus were

weighed and individually caged in 23 X 19 X 13 cm

polycarbonate cages at three weeks of age (day 21-22).

Individuals were assigned randomly (via a random number

table) to one of four conditions: (1) "Family" subjects

received transfers of soiled bedding from their family group

(groups containing an adult male, female, and subsequent

offspring) every other day; (2) "Male" subjects received

soiled bedding from a pooled sample that was derived from

the cages of five unfamiliar adult males (see additional

details below); (3) "Female" subjects received soiled

bedding from a pooled sample of five unfamiliar adult

females; and (4) "Control" subjects received transfers of

clean wood-chip bedding that had been placed in a vole-free

colony room and exposed to air as the bedding within the

cages of animals had been exposed.

Subsequent offspring born to the breeding females that

supplied the bedding in the Family condition were not

reduced in number and were removed from the group at three

weeks of age (21-22 days old). Additional litters born into

the family groups were treated in the same manner. No more









than two animals of the same sex and derived from the same

litter were used in the same condition.

All subjects were kept in homospecific colony rooms and

maintained as outlined in the general procedures. Animals

in each of the experimental conditions were housed in small

cages and maintained on separate shelves in each of the

colony rooms in order to minimize the exchange of olfactory

cues among subjects in the various conditions.

Bedding transfers and maintenance. Bedding transfers

occurred on the first day of placement into the experiment

(week 3) and then on every other day throughout the 6-week

test period. Bedding samples were collected and distributed

into recipient cages during the first four hours of light

onset (2000-2400 h). Pooled bedding came from the cages of

five unfamiliar, individually-housed adult animals of the

appropriate sex that did not have their bedding removed for

four days. Bedding samples were mixed and transferred into

the cages of subjects in 200 cc volumes with the use of

plastic cups that were washed with mild detergent after each

use. All donor cages were cleaned weekly. Cleaning

involved the replacement of all soiled bedding with clean

bedding, except for the retention of 800 cc of soiled

bedding within each family cage and 400 cc in each

individually-housed animal to maintain some common olfactory

cues available to the residents. Subjects that received

transferred bedding had 200 cc of bedding removed from their

cages on the day of exchange and had 200 cc of the









appropriate type of bedding added to replace the lost

volume.

Pooled bedding was used in the "Male" and "Female"

conditions to provide a more uniform stimulus to animals in

these conditions. It seemed plausible that some of the

bedding-donors would produce soiled bedding that differed in

stimulus quality from others. For example, females may have

excreted different amounts of metabolites in their urine as

a result of fluctuating hormone levels. Previous

researchers have used similar pooling methods in behavioral

preference tests with house mice (e.g., Coppola & O'Connell,

1988; Drickamer 1989b).

Physiological and morphological measures. Body weights

were recorded for subjects when they were weaned (day 21-22)

and at weekly intervals until they were 56 days of age (9

weeks). By this time, individuals from each species would

be considered to have reached adult status under standard

laboratory conditions, with the possible exception of pine

voles. Lepri and Vandenbergh (1986) found that the median

age for male pine voles to sire a litter was 57 days (North

Carolina population), whereas the median age for first

conception among the females was 50 days.

Other measures of puberty included the day of vaginal

opening for females and the anogenital distance of males at

weekly intervals. Vaginal smears were taken daily from each

female beginning on the day they first became perforate.

All other subjects, including males, were handled daily in a








similar fashion as were females, as a control procedure for

the effects of handling. During all handling procedures,

subjects were held with individually-assigned vinyl gloves

in order to prevent the transference of odors among

subjects.

At the end of the 9-week period, subjects were

euthanized and selected organs were removed and weighed to

the nearest 0.1 mg. Organs weighed included the uterus,

ovaries, and adrenal glands of females and the seminal

vesicles, testes, and adrenal glands of males. All

dissections and weighing of organs were conducted by the

author who used uniform procedures. In all cases, efforts

were made by the author to remain blind to the condition of

the subjects prior to the dissection procedure. This was

accomplished by dissecting animals in small groups if

possible. During this procedure, the identification cards

of subjects were placed upside-down and beneath the

individual trays that held the removed organs from each

subject. The organs were covered with moist paper towels in

a uniform fashion. The author would then scramble the

position of the dissection trays just prior to the weighing

procedure. Thus, in most cases, the author was blind to the

experimental condition of the animals during the final

cleaning and weighing procedure.

Statistical Analysis

Analysis of variance procedures (ANOVA's) were used to

assess the effects of treatment on subjects. All data were








transcribed to computer spreadsheets and analyzed with the

CSS: Statistica software program (StatSoft, Inc). Data were

analyzed independently for each species and sex, because of

occasional instances of heterogeneity of variances between

the species. Analyses consisted of either one- or two-way

ANOVA's with the experimental condition comprising the

primary between-subjects factor. Some measures, such as

body weight, were recorded over the course of the study and

were analyzed with the additional repeated-measure factor of

weeks. Paired organ weights were analyzed with the

repeated-measure factor of body location (left versus

right), because of known physical asymmetries among some

paired organs (e.g., Pinter, 1968) or other possible

functional asymmetries (e.g., Clark & Galef, 1990). The

alpha level was held at .05 in all comparisons, and all

comparisons were based on a two-tailed probability.

Results

Results of each of the measures are reported below for

the effect of the experimental conditions separately for

each species. The statistical values of ANOVA's are

reported directly in the text or in specified tables,

although the exact probability values of post-hoc

comparisons (Neuman-Keuls tests) are not reported in order

to streamline the text. Only post-hoc comparisons that were

significantly different (p's < .05) are discussed in the

text, or are clearly specified as not being significant when

discussed.








Body Weight

Generally, the experimental condition (bedding type)

did not substantially influence body weight among the

species. However, male pine voles were significantly

affected by the condition (see below).

Because only the factor of week was statistically

significant among the other species, simply reflecting

increases in body weight across age, analyses for each

species are located in Table 3-1 (see Appendix A for means

at each week).

Pine voles: males. Post-hoc analyses revealed that

male pine voles in the Family condition (22.71 + .53,

N = 16) and Male condition (22.96 + 1.05, N = 14) weighed

significantly more than those reared in either the Control

condition (21.24 .53, N = 17) or Female condition

(20.64 + .65, N = 16) at week 9 (interaction of condition

and week, F(18, 348) = 1.98, p = .010)(see Figure 3-1). The

main effect of week was significant, F(6, 348) = 600.87,

p = < .001), indicating larger body weights across weeks,

although the main effect of condition was not significant,

F(3, 58) = 1.68, e = .181 (see Appendix A for means at each

week).

Anogenital Distance

The anogenital distances of the male prairie voles,

meadow voles, and pine voles were not substantially

influenced by the condition. However, there was some








indication that anogenital distances among male montane

voles were influenced by the condition (see below).

Because only the factor of week was statistically

significant in the other species, simply reflecting

increases in anogenital distance with increasing age,

analyses for all other species are located in Table 3-2 (see

Appendix B for means at each week).

Montane voles. Although a significant interaction of

anogenital distance by week of measurement was found,

F(18, 360) = 1.84, p = .019, post-hoc comparisons did not

reveal significant differences among any of the groups (see

Table 3-2 for means at week 9 and ANOVA results; complete

means for each species and week are located in Appendix B).

The mean anogenital distance of montane voles in the Control

condition (M = 13.44 .44 mm, N = 16) approached being

significantly larger than compared to that in the Family

condition (M = 12.25 + .31 mm, N = 16 (p = .09). The mean

anogenital distances in the other conditions were

intermediate in value to those in the Control and Family

conditions. The main effect of week was significant,

F(6, 360) = 173.00, p < .001), although the main effect of

condition was not, F(3, 60) = 0.33, p = .798).

Adrenal Weight

Adrenal weights were minimally affected by the

condition among all species and sexes, with the exclusion of

male pine voles (see below). The mean adrenal weights and









analyses for all species, when uncorrected and corrected for

differences in body weight, are located in Table 3-3.

Pine voles: males. The condition did not significantly

influence the adrenal weights of male pine voles, when

uncorrected for differences in body weight, F(3, 59) = 1.26,

E = .296 (see Table 3-3 for means and ANOVA results).

However, the analysis of adrenal weights, corrected for

differences in body weight, revealed significant differences

among the conditions (see Figure 3-2). Post-hoc comparisons

revealed that the adrenals of males in the Female condition

(M = 24.78 + 1.17, N = 16) were significantly heavier than

in the Control condition (M = 21.60 1.11, N = 17) or

Family condition (M = 20.06 + 1.35; N = 16)(main effect of

condition, F(3, 59) = 3.32, p = .026). The adrenals from

males in the Male condition were similar in weight to those

from the Control and Family conditions but failed to differ

significantly from those in the Female condition (M = 21.02

+ 1.09, N = 14).

The adrenals were heavier on the left side of the body

than on the right, when corrected for body weight (left

adrenal: M = 22.49 + .64; right adrenal M = 21.28 + .62),

F(l, 59) = 7.58, p = .008. The interaction of condition and

body location was not statistically significant, F(3, 59) =

0.14, p = .929.

Among all species and sexes, the left adrenals were

significantly heavier than those on the right side of the

body (See Table 3-3 for complete means and analyses).









Testes Weight

Testes Weights were not significantly affected by the

condition within any of the species, whether they were or

were not corrected for differences in body weight. Complete

means and analyses are located in Table 3-4. The only

significant effects were attributed to small asymmetries in

the left and right testis weights within meadow voles and

montane voles (see means and effect of position in Table

3-4).

Seminal Vesicle Weight

The experimental conditions did not significantly

influence the seminal vesicle weight within any species,

whether they were or were not corrected for differences in

body weight (see complete means and analyses in Table 3-5.

Ovarian Weight

The experimental conditions did not significantly

influence the ovarian weight of any species, whether they

were or were not corrected for differences in body weight

(see means and analyses in Table 3-6).

Uterine Weight

The experimental condition did not significantly

influence the uterine weights among the pine voles, prairie

voles, or meadow voles (see means and analyses for all

species in Table 3-7). However, significant differences

were evident in the uterine weights of the montane voles as

a result of the condition.








Montane voles. The condition significantly influenced

the uterine weights of montane voles (see Figure 3-3). When

uterine weights were compared, without correcting for

differences in body weight, the uteri of females in the

Control condition (M = 27.46 + 1.92) weighed significantly

more than the uteri found in the Family condition (M = 18.56

+ 1.92) and in the Female condition (M = 16.52 + 2.22) (main

effect of condition, F(3, 60) = 4.55, p = .006). The mean

uterine weight among those in the Male condition was

intermediate to those in the other conditions (M = 24.35

+ 2.97) and was not significantly different from them. This

result is the first example, within a controlled

environment, of sexual suppression in female montane voles

that are exposed to the bedding from a family group or from

adult females.

When the analysis of uterine weight was corrected for

differences in body weight among montane voles, the results

were similar to those of the unadjusted analysis (main

effect of condition, F(3, 60) = 4.61, p = .006 (see Figure

3-3). The adjusted uterine weights were significantly

greater in the Control condition (M = 97.71 + 10.65) than in

the Family condition (M = 61.02 + 4.42) or Female condition

(M = 58.92 + 5.74). However, in addition, the females of

the Male condition (M = 91.21 + 14.45) had significantly

heavier uteri than those in the Family condition.









Status of Vaginal Perforation

The total numbers and percentages of females that

became vaginally perforate during the course of the study

are summarized in Table 3-8 for all species. Clear

differences were evident among the species in their modal

pattern of vaginal perforation. Pine voles displayed the

most atypical pattern of vaginal perforation with respect to

the other species. None of the 66 pine voles were perforate

at the beginning of the study (week 3) and only 2 of the 66

(3.0%) became perforate during the 9-week study. This

slight shift in the frequency of pine voles that became

perforate was not significant between weeks 3 and 9 (McNemar

chi-square: X2 = .50, p < .479). Nearly all of the females

of the other species became perforate at some time during

the study (see Table 3-8).

The percentages of females that were perforate at the

beginning of the test were different among the four species,

X2(3, N = 258) = 70.08, p < .001. Post-hoc comparisons

(chi-square) revealed that all species differed

significantly from pine voles in the proportions of females

that were perforate on week 3. However, significantly fewer

prairie voles were perforate at week 3 than were meadow

voles, (X2(1, N = 167) = 18.36, p < .001, and montane voles,

X2(1, N = 161) = 12.18, p < .001. Meadow voles and montane

voles did not differ significantly on this measure,

X2(3, N = 186) = .91, p < .341.









Delay until Vaginal Perforation

Statistical analyses of the delays of vaginal

perforation were problematic, because properly they would be

limited to those females that were not perforate on the

first day of the study (week 3). However, the majority of

all meadow voles (34 of 62 or 55%) and many montane voles

(26 of 64 or 41%) were perforate on the initial day of the

study and would be excluded from the analysis. In contrast,

only 5 of 66 (8%) prairie voles and none of the 66 pine

voles were perforate on the first day of the study. The

total number of subjects and percentages of subjects for

which vaginal smears were obtained are summarized in Tables

3-9 and 3-10 for each species and by condition.

The species differed appreciably in the time at which

they became vaginally perforate. Figure 3-4 displays the

cumulative percentages of females that became perforate

throughout the study as a function of species. Nearly 80%

of the meadow voles became perforate within the first 4

weeks of age, while prairie voles reached the 80% mark

nearly a week later (Day 33). Over 50% of the montane voles

were perforate by day 23, but the cumulative percentage of

all montane voles did not reach the 80% criterion until day

41. Only two pine voles became perforate, the first at 44

days of age and the second on last day of the study (Day

63).

The mean ages (in days) that vaginal perforation

occurred for subjects in each species and condition, that









were imperforate on the first day of the study, are shown in

Figure 3-5. Sample sizes were not uniform across all

species and conditions because of the different patterns of

typical vaginal opening. Total numbers of subjects ranged

from 14-16 for prairie voles, 5-6 for meadow voles, and 8-10

for montane voles among the conditions.

The condition was not found to significantly influence

perforation latency among the prairie voles, (Kruskal-Wallis

ANOVA, prairie voles: H (3, N = 59) = 4.34, p = .23) or

montane voles: H (3, N = 36) = 5.74, p = .12). Meadow voles

were not analyzed statistically because of the small number

of subjects amenable to this analysis (N = 22).

Vaginal Smears

Three types of cells were classified and counted from

the vaginal smears, following the method of Taylor

(1990/1991). The cell types analyzed included cornified

cells, nucleated cells, and leukocytes. Frequencies of each

type were converted to percentages of the total number of

cells in each smear, because of the variability in the total

number of cells obtained per smear. Similar conversions of

cell frequencies have been used previously in studies with

laboratory rats (e.g., McClintock, 1983, 1984) and with

voles (e.g., Sawrey, 1989/1990; Shapiro & Dewsbury, 1990).

Data from female pine voles were excluded from the analysis

because only two became perforate during the study.

All species produced vaginal smears that were typically

dominated by either cornified cells or leukocytes, with









appreciably fewer nucleated cells. These characteristics of

smears were typical of those found previously for several

species of voles (e.g., Sawrey & Dewsbury, 1985; Taylor,

1990/1991). Nucleated cells were relatively few in number

and relatively constant in proportion throughout the

duration of the study. In contrast, cornified cells and

leukocytes were typically in reverse proportion to one

another when changes in the percentages of cells were

detected.

Prior to conducting statistical analyses, all cell

percentages were pooled for each subject by two-day block

intervals. This procedure generally reduced daily

variability in the cell percentages across successive days.

Prior to pooling, graphs were made of the data as a function

of each day; no indications of cyclical fluctuations were

evident.

Effect of condition: within-species comparisons of cell

types. Two types of analysis were conducted for

within-species comparisons of cell types. The first

analysis consisted of comparing the proportions of each cell

type as a function of the experimental conditions for each

two-day block (Kruskal-Wallis ANOVA by Ranks tests). Thus,

a total of 21 two-day blocks (representing Days 21-22

through 61-62) were analyzed sequentially to determine if

the condition influenced the percentages of cell types. In

the case of a significant result, post-hoc comparisons were









made to identify which conditions differed significantly

(Mann-Whitney U Tests).

In the second analysis, repeated-measures analyses were

used to assess whether significant changes in the

percentages of each cell type occurred as a function of age.

These analyses were done separately for each condition and

species. Cell percentages were analyzed on alternate blocks

of days representing the ages of 33-34 days through 61-62

days (Friedman ANOVA by ranks tests). This reduced sample

of days was necessary due to fewer smears available during

the youngest ages and thus the necessary exclusion of data

from repeated-measures analysis (see Table 3-9). Thus, a

series of eight means was compared in each repeated-measures

analysis. Post-hoc comparisons (Wilcoxon Matched Pairs

Tests) were used to identify where significant changes

occurred across the blocks of days when the Friedman ANOVA

was found to be statistically significant (p's < .05).

Prairie voles: cornified cells: effect of condition.

The percentages of cornified cells from prairie voles are

shown as a function of experimental condition in Figure 3-6

(each mean plotted represents the pooled data of individuals

by two-day block intervals). It was evident that relatively

little variation occurred in the cell percentages as a

function of the condition among prairie voles. A

statistically significant change was identified for only one

of the two-day blocks (Day 27-28: H(3, N = 35) = 8.90,

p = .031). During this block, subjects in the Female









condition had significantly more cornified cells than those

in the Male condition (M's = 27.5% versus 17.9%; total

numbers of subjects can be determined from Table 3-9).

Prairie voles: effect of age. Relatively small changes

in the percentages of cornified cells occurred as a function

of age. Repeated-measures analyses failed to detect any

significant change in the cell percentages for subjects in

any of the four conditions.

Additional comparisons of the percentages of nucleated

cells and leukocytes have been placed in Appendix C. In

general, the proportions of nucleated cells varied little as

a function of the condition or age for all species. Thus,

the proportions of leukocytes often varied inversely to the

proportions of cornified cells and are somewhat redundant

for purposes of analysis.

Meadow voles: cornified cells: effect of condition.

The mean percentages of cornified cells as a function of the

experimental condition are shown in Figure 3-7. Despite

somewhat greater differences among the cell means than those

found in prairie voles, and relatively large sample sizes

(range of total N's = 47-56), no significant differences

were found within any of the blocks of days as a function of

condition. The lack of significant differences also held

true for all comparisons of nucleated cells and leukocytes

(Appendix C).

Meadow voles: effect of age. Repeated measures

analyses revealed significant changes in the percentages of









cornified cells within all four conditions of meadow voles.

There were general increases in the percentages of cornified

cells for subjects in all conditions, although the relative

percentages of cornified cells for the Control condition

were typically the smallest among those from all conditions

(see Figure 3-7; Control condition: X2(7, N = 11) = 31.39,

p < .001). Significant differences occurred among the

following blocks of days for subjects in the Control

condition (Note: differences indicated below are in the

direction of later days being significantly higher in cell

percentages than earlier days except where noted "*"): Days

33-34 versus Days 49-50 through Days 61-62; Days 37-38

versus Days 41-52, 49-50 through Days 61-62; Days 45-46

versus Days 53-54 through Days 61-62; Days 49-50 versus Days

57-58 through Days 61-62.

Significant changes in the percentages of cornified

cells for subjects in the Family condition were evident,

X2(7, N = 11) = 29.85, p < .001), and were as follows: Days

33-34 versus Days 45-46 through Days 57-58; Days 37-38

versus Days 45-46 through Days 61-62; Days 41-42 versus Days

45-46 through Days 61-62.

Significant changes in the percentages of cornified

cells for subjects in the Male condition were present,

X2(7, N = 12) = 36.97, p < .001), were as follows: Days

33-34 versus Days 41-42, Days 49-50 through Days 61-62; Days

37-38 versus Days 41-42, Days 49-50 through Days 61-62; Days









41-42 versus Days 53-54 through Days 61-62; Days 45-46

versus Days 49-50 through Days 61-62.

Significant changes in the percentages of cornified

cells for subjects in the Female condition were present,

X2(7, N = 13) = 52.08, p < .001), were as follows: Days

33-34 versus Days 41-42 through Days 61-62; Days 37-38

versus Days 45-46 through Days 61-62; Days 41-42 versus Days

45-46 through Days 61-62; Days 45-46 versus Days 53-54

through Days 61-62; Days 49-50 versus 57-58.

Montane voles: cornified cells. The mean percentages

of cornified cells among montane voles in the various

conditions varied more as a function of condition than the

percentages observed in prairie voles and meadow voles

(compare Figures 3-8 for montane voles to Figures 3-6, and

3-7). Typically, there were gradual increases in the

percentages of cornified cells throughout the study for

montane voles.

Montane voles: effect of condition. Statistical

differences were found among subjects in the various

conditions on blocks of days including Days 53-54, 57-58,

and 61-62. Among the first and last of these three blocks,

the subjects in the Male and Control conditions had

significantly greater mean percentages of cornified cells

than those in the Female and Family conditions. The

intermediate block (Days 57-58) also differed in the same

manner, except the difference between the Male and Family









condition failed to reach a significant level (see Figure

3-8 for clarification).

Montane voles: effect of age. Repeated measures

analysis revealed significant shifts in the percentages of

cornified cells within each of the four conditions. Montane

voles in the Control condition had significant shifts in the

proportions of cornified cells between the following blocks

of days: Days 33-34 versus 41-42, 49-50 through 61-62; Days

37-38 versus 57-58 through 61-62; Days 41-42 versus 53-54

through 61-62; Days 45-46 versus 49-50 through 61-62; Days

49-50 versus 57-58 through 61-62.

Significant shifts in the percentages of cornified

cells were found for those in the Family condition among the

following blocks of days: Days 33-34 versus 49-50, 57-58

through 61-62; Days 37-38 versus 57-58 through 61-62; Days

41-42 versus 49-50 through 61-62; Days 45-46 versus 61-62;

Days 49-50 versus 61-62; Days 53-54 versus 57-58.

Changes among blocks of days for subjects of the Male

condition were statistically significant among the following

blocks of days: Days 33-34 versus 53-54 through 61-62; Days

37-38 versus 53-54 through 61-62; Days 41-42 versus 49-50

through 61-62; Days 45-46 versus 53-54 through 61-62; and

Days 49-50 versus 53-54, and 61-62.

Finally, significant shifts in the percentages of

cornified cells were identified for subjects in the Female

condition among the following blocks of days: Days 33-34

versus 41-42 through 49-50, and 57-58 through 61-62; Days


__









37-38 versus 49-50 through 61-62; Days 45-46 versus 61-62;

Days 49-50 versus 61-62; Days 53-54 versus 61-62; and Days

57-58 versus 61-62.

Discussion

Prairie voles and pine voles were expected to show

fundamentally different patterns of response from those of

the meadow voles and montane voles. I predicted that the

less social meadow voles and montane voles would show

pronounced sexual advancement when exposed to the bedding of

unfamiliar opposite-sexed individuals, and show signs of

sexual suppression when exposed to same-sex odors. Animals

exposed to clean bedding (Control condition) were expected

to have intermediate indices of sexual development compared

to those in the other conditions. In contrast, I believed

there would be little or no differences in sexual

development within the prairie voles or pine voles.

The experimental results of the exposure to the soiled

bedding on general and sexual development are discussed

below for each of the measures.

Body Weight

Male pine voles were the only group that differed

significantly in body weight as a function of the condition.

Males in the Family and Male conditions weighed more than

those in the Control and Female conditions (Figure 3-1).

These results, if repeatable, are contrary to predictions.

The results suggest that the olfactory cues associated with

adult males or family groups cause general increases in the









rate of growth of males exposed to them, when compared to

patterns of growth of males exposed to clean bedding or to

odors from adult females. Unfortunately, it is not possible

to compare these results with those of other studies, as

pine voles have not been investigated in any similar manner.

However, general characteristics of natural populations

of pine voles suggest that male pine voles are not sexually

suppressed when living within family groups and often are in

the presence of at least one other scrotal male (FitzGerald

& Madison, 1983). Thus, limited information on the

free-ranging behavior of pine voles is at least compatible

with the notion that the exposure of males to the odors from

family groups and/or other males in some manner contributes

to increased body weight. One possible mechanism by which

males might experience increased body weight is simply

through increased food intake. Such a mechanism has been

identified that led to increased weight gain in male musk

shrews (Suncus murinus) when exposed to females (Wayne &

Rissman, 1990).

It is not known why pine voles were the only species to

experience differences in body weight as a function of the

condition. Other studies have shown that the body weights

of other species of Microtus can be affected by exposure to

pheromonal cues. Female meadow voles have been shown to

have significantly greater increases in body weight,

compared to controls, when the urine of mature males

trickled directly into the cages of females via small tubes








(Baddaloo & Clulow, 1981). Other research has provided

evidence that male and female prairie voles could differ in

body weight, by nearly 10 g respectively, as a function of

whether they had separate air supplies or shared a common

air supply within the colony (Batzli et al. 1977).

Together, the results of previous studies suggest that

differences in body weight might have been detected within

some of the other species, had the procedures been

different. The use of a more continuous supply of urine and

pheromones or a separate supply of air for each subject may

have led to detectable differences as a function of the

condition. However, given the generally small differences

in body weights among the male pine voles and number of

comparisons made, one must also consider the results were

due to chance alone and not due to pheromonal differences.

Additional study seems warranted to clarify whether these

patterns of weight gain among male pine voles can be

replicated, what mechanism accounts for differences in body

weight, and what functional differences they may be

producing.

Anogenital Distance

The anogenital distances among montane voles might have

differed as a function of the conditions, but the results

are ambiguous. Whereas the overall F statistic was

statistically significant, post-hoc comparisons failed to

detect significant differences among the groups. Males in

the Control condition had a larger mean anogenital distance








than those in the Family condition, although it was not

significantly different (E = .09; Neuman-Keuls test). This

difference is in the hypothesized direction, but the

absolute difference between the groups limits making any

firm conclusion.

In an attempt to determine statistically whether the

differences in anogenital distances among montane voles were

systematic effects caused by the conditions, an analysis of

covariance was conducted, using body weight at week 3 as a

covariate. Results were similar to those of the primary

analysis, i.e., the interaction of anogenital distance and

week of measurement was statistically significant, F(18,

360) = 1.84, p = 0.019, although no post-hoc comparisons

were significantly different. Thus, the results of the

anogenital distance comparisons are only suggestive of

differences produced by exposure to the different bedding.

Unfortunately, no similar studies have been conducted with

montane voles, or other species of Microtus, in order to

make any comparisons between studies.

It is not known what functional difference, if any, a

larger anogenital distance would reflect. Presumably, the

measure reflects, in part, the relative size of the testes

which lie proximally between the penis and the anus. In the

present study, the correlation between testes weight and

anogenital distance was relatively small among all montane

voles (N = 64; rg = .31), although it was slightly higher in

the Control males alone (N = 16; rs = .37). Thus,








differences in testis weight do not lend much support for

the possibility that the differences in anogenital distance

largely reflect differences in testis weight. The analysis

of the testes weights of the montane voles did not reveal

any significant influence of the conditions (see below).

Additional study would seem necessary to establish whether

the differences in anogenital distance in montane voles are

repeatable outcomes of exposure to these substrates.

Adrenal Weight

The adrenals differed systematically in weight as a

function of the treatment among male pine voles. The

adrenals, when corrected for differences in body weight,

from males in the Female condition were significantly

heavier than those in the Control and Family conditions.

Because males of the Female condition were significantly

lighter in body weight than those in the Family and Male

conditions, it is possible there is a causal relationship

between increased adrenal activity and decreased body

weight. Prior studies with male meadow voles have shown

that subcutaneous injections of ACTH produced significant

reductions in body weight and concurrent increases in

adrenal weight than among control males (Pasley & Christian,

1971).

It is not known what differences there are in bedding

soiled from male and female pine voles. Recent chemical

analysis of volatile compounds in male and female pine vole

urine did not reveal any qualitative difference in the








urinary profiles between them (Boyer et al. 1989). However,

female urine contained three volatiles in higher

concentrations than in male urine, whereas male urine

contained one compound present in higher concentration than

in female urine. Boyer et al. (1989) suggested there may be

differences in the nonvolatile urinary fraction that might

be investigated. Among house mice, the acceleratory

pheromone produced by male house mice appears to be mediated

by a nonvolatile urinary fraction contained in the urine

(Vandenbergh et al. 1975; 1976).

Among all species, the adrenals from the left side of

the body were significantly heavier than those located on

the right. This difference in asymmetry may be common to

all species of Microtus. Pinter (1968) reported that the

adrenals in montane voles were larger on the left than on

the right side of the body, regardless of the sex, age, or

regime of diet or photoperiod. A similar asymmetry in

adrenal weight had been reported for California voles

(M. californicus)(Mullen, 1960). The specific functional

difference of such asymmetry, if any, is not known.

Testes Weight

Comparisons of the testes weights did not indicate any

effect of the experimental condition among any of the

species. It is possible that changes in testicular weight

among male Microtus are relatively unaffected by exposure to

different pheromones, although exposure to stressful

environments may produce substantial changes. Lecyk (1967)








provided evidence that male common voles (M. arvalis), that

were housed adjacent to cages with crowded and sexually

active voles, had lighter testes than control males. Other

indirect evidence that testicular weight can change as a

function of stress was provided by Pasley and Christian

(1971). Male meadow voles, injected with ACTH across a

series of days, had significant decreases in testes weights

when compared to noninjected controls.

It is possible that testis weight is a relatively

insensitive measure of male reproductive activity, at least

when investigating the influence of pheromones or other

olfactory stimuli on reproductive maturation. More precise

measures of reproductive activity might reveal systematic

differences among voles exposed to different pheromones.

Measures such as the level of circulating androgen or

characterization and counts of spermatozoa may be

appropriate. Research with California voles has shown that

titers of androgens were found to be significantly lower in

males exposed to bedding from their mothers than in those

exposed to bedding from unrelated males; testes weights were

not found to differ among these groups (Rissman et al.

1984).

Seminal Vesicle Weight

The experimental condition did not have a noticeable

effect on the seminal vesicle weight among any of the

species. Why no significant differences were found is

puzzling. Prior studies with California voles have shown








that differences in seminal vesicle weight can be found when

males are exposed to different types of soiled bedding. The

seminal vesicles of California voles that were exposed to

bedding soiled by their own mother, their family group

(including the sire), or an unrelated mother were

significantly smaller than those of males reared in clean

bedding (Rissman et al. 1984; Rissman & Johnston, 1985). In

contrast, the presentation of soiled bedding from unfamiliar

adult males, or bedding from the family that had been

supplemented with bedding from the separated father,

produced seminal vesicles of similar weight in both groups.

Thus, the data from California voles suggest that the

development of the seminal vesicles can be delayed in young

males when they are exposed to odors from their mothers or

from unrelated mothers, but not delayed by exposure to odors

from unrelated adult males or from their fathers (Rissman et

al. 1984).

Ovarian Weight

The experimental condition did not have a noticeable

effect on the ovarian weight in any species, whether they

were corrected or uncorrected for differences in body

weight. However, previous studies suggest that the ovarian

weight among some species of Microtus can be affected by

exposure to pheromonal stimuli. Baddaloo and Clulow (1981)

found that female meadow voles that were exposed to male

urine, that had been designed to drip into the adjoining








(empty) half of divided cages, had significantly heavier

ovaries than compared to those of control females.

Female pine voles have also been shown to have

increases in ovarian development when exposed to pheromonal

stimuli. Lepri and Vandenbergh (1986) placed individual

females directly below the cages of animals of different

stimulus qualities. This procedure enabled feces and urine

from the animals) in the top cages to fall into the cages

housing the females. Results indicated that the ovaries of

the females exposed to the excreta from either individually

housed intact males or castrated males were significantly

heavier those of females exposed to an empty cage or to the

excreta from singly-housed females or group-housed females.

Uterine Weight

By at least one standard, the uterus is considered to

be the best and most reliable bioassay for circulating

estrogens in female house mice (Bronson & Stetson, 1973).

Female montane voles in the Control condition were found to

have significantly heavier uteri than those in the Family

and Female conditions (Figure 3-3). In addition, the uteri

from those in the Control and Male conditions were also

significantly heavier than those in the Family and Female

conditions when the uteri were corrected for differences in

body weight. These results are the first to demonstrate

sexual suppression of female montane voles when they are

exposed to adult female bedding or family bedding under

controlled conditions. Previous field evidence had only








suggested such a phenomenon (Jannett, 1978). The results

are in general agreement with predictions and complement the

recent finding of sexual acceleration in female montane

voles when they were exposed to male stimuli (Sawrey &

Dewsbury, 1991). However, if the responses of uterine

weight to the pheromones conform to the distinction of

socially-dependent versus independent dichotomy, it is not

clear why meadow voles also did not show differences in

uterine weight when they were exposed to the different

bedding types.

It is possible that many species of Microtus, including

meadow voles, may require both direct exposure to males and

their pheromones for complete reproductive activation. This

type of synergy has been demonstrated in house mice

(Drickamer, 1974; Bronson & Maruniak, 1975), montane voles

(Sawrey & Dewsbury, 1991), prairie voles (Carter et al.

1987), and pine voles (Lepri & Vandenbergh, 1986), although

the specific causes for it are not yet understood. For

example, among female prairie voles, direct contact with

male urine or housing them in male-soiled cages results in

increased uterine weights, but typically does not elicit

behavioral estrus (Carter et al. 1987). Thus, it is

possible that female meadow voles may require more direct

male stimulation in comparison to montane voles. Evidence

found by Taylor (1990/1991) supports this hypothesis, as

nulliparous female meadow voles showed extremely low

copulation rates after being kept behind a wire-mesh barrier









for up to seven days with only 1 h of daily direct

interactions with males.

Characteristics of Vaginal Perforation

Although none of the species differed significantly in

the latency to become vaginally perforate as a function of

exposure to soiled bedding, clear species differences were

evident in the latencies of the species to become perforate

(Figure 3-4). Meadow voles and montane voles became

perforate either before or shortly after weaning (Day 21).

In contrast, prairie voles were somewhat delayed in becoming

perforate. Pine voles were even more extreme, with only two

of the animals becoming perforate during the course of the

study.

The few pine voles that become vaginally perforate

during the course of the study do not seem uncharacteristic

from personal observations and from other reports (e.g.,

Schadler & Butterstein, 1979). It appears that most female

pine voles, whether housed in isolation or with siblings, do

not become perforate until they are placed with an

unfamiliar male or reach a considerably advanced age. In

pilot work for the study presented, 10 adult female pine

voles were examined for vaginal perforation. Whether pine

voles were housed singly, or with other siblings, none was

perforate despite ranging in age between 90 to 120 days.

Earlier work also supports the finding of substantial

delays in reproductive activity among pine voles. Schadler

and Butterstein (1979) found that among female pine voles








that had been paired for breeding with fertility tested

males, the mean age of first conception for the females was

105 days. Thus, it appears that direct and relatively long

exposure to males may be a normal prerequisite for vaginal

perforation and reproductive activation in pine voles, even

if they have reached a considerably advanced age.

Substantial delays among female prairie voles to become

vaginally perforate have also been documented. Richmond &

Conaway (1969b) found that 98% of more than 200 individually

housed females, or those in compatible female groups,

remained in a state of persistent anestrus with imperforate

vaginas between 3 to 5 weeks of age.

Within-species analysis of the perforation latencies

among prairie voles did not suggest they are strongly

affected by the olfactory cues. The overall comparison test

was not statistically significant, despite there being a

relatively large number of imperforate female prairie voles

at the beginning of the study (N's = 14-16 per condition).

Female prairie voles in the Male condition became perforate,

on average, earlier than those in the Control condition (M's

= 33.0 versus 27.5 days)(Figure 3-5).

One of the difficulties with analyzing the conditions'

effect upon perforation latencies among meadow voles and

montane voles was that many females were vaginally perforate

at the beginning of the test and thus excluded from the

formal analysis. Employing an earlier day of weaning did

not seem plausible for all species, although it was








attempted in pilot work. Weaning pine voles at an earlier

age than 21 days often resulted in their deaths within a day

or two following separation. In retrospect, it would seem

possible to wean each species at different ages that

corresponded more closely to their species-typical age for

weaning. Indices such as the cessation of nursing or the

timing of the first day to ingest solid food is typically

earlier for meadow voles and montane voles than they are for

pine voles and prairie voles (McGuire & Novak, 1984; 1986).

By weaning each species at an age that appears most

appropriate for each species, it would be possible to test

more readily how stimuli affected certain reproductive

responses, at least within each species.

Vaginal Cytology

It was expected that the less social species would show

higher proportions of cornified cells when the females were

exposed to male bedding than would females exposed to

bedding in the other conditions. Those in the Family and

Female conditions were predicted to have fewer cornified

cells than those in the Control condition. The patterns of

cornified cells for all species that became vaginally

perforate are shown in Figure 3-6 through Figure 3-8.

The percentages of cornified cells for prairie voles

among the various conditions were similar, and statistically

only one of the two-day blocks revealed a significant

difference among conditions. It seems plausible that this

one significant difference was spurious, because neither the









few preceding nor the few following blocks of days was

statistically significant. In addition, the total number of

multiple-comparisons tests was substantially high. This in

turn would inflate the likelihood of detecting at least one

statistically significant difference among the days

compared. As predicted, female prairie voles revealed

little response to the different bedding types.

Although variation in the cell percentages for meadow

voles suggested an influence of condition, none were

detected statistically (Figure 3-7). Inspection of the

percentages of cornified cells suggests a trend of more

cornified cells in females exposed to adult female odors

than those receiving no odors (Control). A priori, the

direction of this relationship was not expected but suggests

further study because it is possible that meadow voles

experience sexual acceleration when exposed to odors or

pheromones of other adult females. Among house mice, urine

from pregnant and lactating females accelerates sexual

development in female mice (Drickamer & Hoover, 1979), as

does the urine from female mice that are in estrus

(Drickamer 1982c). It seems plausible that among meadow

voles, pheromones contained in the urine of other meadow

voles that are of either sex may cause sexual acceleration

when compared to those that are not exposed to any

pheromonal source, at least under certain conditions.

Results from the analysis of vaginal smears of meadow voles








by Baddaloo and Clulow (1981) are in general agreement with

this prediction.

There is reasonable evidence that the conditions

affected the proportions of cornified cells of the montane

voles, at least during the last 10 days of the study. The

statistical results revealed systematic differences in the

proportions of cells as a function of condition during the

later portion of the study (Figure 3-8). Specifically,

during three of the last five two-day blocks of smears, the

females in the Male and Control conditions had substantially

more cornified cells than those in the Family and Female

conditions. Although there is the possibility that these

relatively few significant results were due to chance,

because of the relatively large number of comparison tests

(21 for each species and condition representing Days 21

through 62), the distribution of the significant effects

suggests a non-random pattern.

The differences among the uterine weights of the

montane voles add additional support for the hypothesis that

the percentages of cornified cells of montane voles reflect

systematic differences as a function of the condition. The

corrected uterine weights were significantly heavier from

females in the Male and Control conditions compared to those

from the Family and Female conditions. Thus, the

percentages of cornified cells appear to reflect these

differences; larger uteri were associated with higher

percentages of cornified cells among the montane voles.








Given the present results, the montane voles appeared

to be the species that was most affected by the exposure to

the pheromones. The results of the current study and of

earlier studies suggest that, even with similar levels of

male exposure, prairie voles and pine voles take appreciably

longer to become sexually receptive than montane voles and

meadow voles, and typically require direct exposure to males

for complete reproductive activation to occur (Carter et al.

1987; Lepri & Vandenbergh, 1986). Thus, results of the

vaginal smear data provide limited support for the

socially-dependent and socially-independent dichotomy. The

less social species, meadow voles and montane voles, appear

to be more sensitive to conspecific pheromonal cues than are

the more social species. Additional study of these species,

using common procedures, may be useful to determine if these

results are valid, and to shed light on possible functional

differences between the species.

Conclusions (Experiment 1)

A major theme that broadly characterizes the results of

Experiment 1 was the absence of an effect of exposure to the

different bedding types among the species. These results

were surprising, considering that a number of studies have

shown that the reproductive physiology of several species of

voles can be influenced by exposure to the same types of

olfactory stimuli or pheromones that were used in the

present experiment (see Puberty Modulation in Voles).









A few studies have shown specifically that the transfer

of chemosignals contained within soiled bedding can

effectively produce changes in reproductive activity in

house mice (Drickamer, 1982c) and in some species of voles.

Lepri and Vandenbergh (1986) showed that exposure to

chemosignals contained within the soiled bedding of family

groups caused reproductive suppression among female pine

voles. Female pine voles were paired with unfamiliar males

and then housed in cages with either clean or family-soiled

bedding for 48 h. Females that were housed on clean bedding

had significantly heavier uteri and ovaries than females

housed on family-soiled bedding.

Soiled bedding has also been shown to affect

reproductive development among male California voles

(Rissman et al. 1984). Males that were reared from weaning

in bedding from their families had significantly lighter

seminal vesicles when they were 45, 55, and 75 days of age

than those reared in clean bedding (Rissman et al. 1984).

Androgen levels were significantly higher among the males

reared in clean bedding versus those reared on the family

bedding on day 45. Together, these studies show that the

method of exposing animals to soiled bedding can be an

effective means to reveal specific changes in reproductive

physiology, at least among some species of muroid rodents.

Several reasons can be proposed as possible

explanations for the lack of substantial differences in most

reproductive measures as a result of the different








conditions. It is possible that pheromones may be minimally

involved, or their activity overestimated, with regulating

the onset of reproduction in all or most species of

Microtus. It is possible that pheromones are substantially

more effective in regulating reproduction in other species,

such as in house mice (e.g., Bronson, 1979; Vandenbergh &

Coppola, 1986). Clearly more comparative research is needed

to test this possibility. Seminatural studies may be an

appropriate method in which to run biologically relevant

experiments, while still enabling the close monitoring of

reproductive development in individuals exposed to

pheromonal sources.

If pheromones are critically involved with influencing

the timing of reproduction among Microtus, there are at

least three methodological reasons that could account for

the relatively few differences found. First, it is possible

that the pervasive odors within the colony rooms effectively

eliminated many of the differences that might have been

found had animals been reared in environments that

controlled for all or most of the extraneous olfactory

stimuli. All animals were reared and tested in colony rooms

where the odors of other conspecifics animals were housed

that varied in sex, age, and reproductive status, although

the subjects had been on separate shelving for each

condition. Some support for this hypothesis is found in the

work of Batzli et al. (1977), who controlled the air supply

to individual animals. Both male and female prairie voles









had substantially elevated gains in body weight across a

three month period as compared to animals that were treated

similarly but shared a common air supply with other prairie

voles. In addition, Sawrey (1989/1990) found differences in

the patterns of vaginal smears of female montane voles that

had been housed either with six other females in a separate

room or had been housed in a larger colony with animals of

mixed sex and age. Vaginal smears revealed that the six

separately housed females had smaller percentages of

cornified cells than smears from females in the large colony

room (11.9% and 29.8% respectively).

A second possible reason for the relatively few

differences found among the conditions is the possibility

that the daily handling procedure was stressful enough to

effectively obscure or eliminate systematic differences.

There is indirect evidence that suggests such a possibility.

Olsen & Seabloom (1973) found that the event of captivity

caused elevated and prolonged secretion of corticosterone in

wild-caught meadow voles. It is known that increased

secretion of adrenal corticoids, androgens, progesterone,

and other steroids are associated with an inhibition of

reproduction (e.g., Christian, 1975). Thus, the daily

handling procedure may have been a significant source of

stress that might have reduced differences in morphological

and physiological growth between animals in the different

conditions.









A third possible reason for the relatively few

differences among the conditions stems from the experimental

design. It is possible that larger differences would have

been found among the conditions had the soiled bedding and

pheromones been presented to the subjects more frequently

than every other day. Some researches have used procedures

whereby urine was channeled directly into the cages of

subjects via tubes or they have placed cages with wire

bottoms and urine donors directly over the subject's cage

(e.g., Baddaloo & Clulow, 1981; Lepri & Vandenbergh, 1986);

thus, a near-continuous supply of pheromonal odors had been

available. Some of the information gathered in house mice

suggests that some pheromones are relatively volatile and

therefore short-lived in functional activity, while others

appear to be more stable and active over a few days. The

actual chemical composition and properties of the pheromones

from voles remain largely unknown.

Future research that is designed to identify the

effects of exposure to pheromones may be based on the

general design of this study. However, three improvements

in the design could be based on the three possible sources

of largely negative results listed above. Ideally, subjects

would have individual supplies of air, be minimally stressed

through extraneously handling or exposure to other animals,

and have a more continuous exposure to pheromonal sources.

The apparatus of Baddaloo & Clulow (1981) represents a

modification in procedure that might detect physiological





85


responses from subjects. Despite the few significant

results found in the present study, the results of many

other studies suggest that the comparative study of the

functions of pheromones among species of Microtus, appears

to be a productive area for further advances in the

elaboration of their evolved reproductive and social

patterns.






86









S-H 1 n r 0
OcH O 2 0O 0 OO m Oa




H HO 0 m 0 00



0r-o r om CH O C





0C C o O O
04 (d 0 0 IV IN
. ') o H H







0 0 00 NN 0 0 oNo









0 0 -0 OOJ 0 00 0
.o -- Co C Co co




S 0) 00 )-i r1 r04
< >4








r 0 0 0 00)V
4J II I a) I IS4 a) a qI ro II a)I


4J)
mc X
*H H
to
60
a o .
*) 0 - -C -
(I Ho cI Ioo c o ci









D OC;H OCH 0HH 0H6
-UH U-H UcH U*H






*A Q) c t
S* 0
t3 onOo N Ho H co 00nH


a 0 CIN- o -N 0 Q-- N 0


0 u . Z* . u . i^ 0 Z a)
HN 0 H H Q)


>0 HH C c Oi**NO O ** O *
VV 0 H H NC U) (a M) H IC




H H -U IFU o
04 CC 0 0 II 0) 0 0I H II | c 1 C
4 41 r, U) *10 0 l 0 .o 0

1 0 H N 1-C M- l
m u 0 0 0) V
O en s z\ c i f g i e i s n*















-K
-K

OHO
m0 00

V

moo


0



o o








0
C -H
OH U
4- t(










-rlr l p
C () 4-)








n c
ri n

r-I






*cc
enH(rH
cc-


-K

CO H LO
rO -


V


n co LO

O H

n


-K








OON
C) 0 F
enO n





0 CM


in


0 0
CO CO 0

-C 0 0 0

4 O MO
00 00


QC C
C -H C -H
0 W-z 0 4Qr





U H U H


00
N




enV


. c
cO H H



en
(N-
i-l 0
* c
09 N


S ,-- N 0 -U
N c .NH 0
00 n 3 0 r H)
0 o -[ *H *o C) *. .- 0 >
a) HHH 0 OHH H NHH




m co3 m Ino (L o0 a)
*0 r-- 0r rc

O NH O OH tH > NHH >
> n- > n 0- n

o o C) Cl
'0 '0 4 m
(0 II II (0 II II C II II C
ra a4 o o
C C) 0 0
z zi ZI z 'dl ZI 0 'l 'l '0


a-K

O D
co o Ln

V

Ome









on M
en CM o


(N








0rl
0 04






e4 co


.-H O
C




U H













N V
- 4-1




















*c z
U OH











N--
O 'l r-1













NHH


NO



II II
. HHi
o(N 1r-

nI II


CM n


<

S*o
C.

a,V




CP4
-M
II





0
4 .

to
v







fo
S0



0*
0
rdl




in
SIo



g II


'0 >"
C H



-rl
4
'0 0 -H



=1 0


I
I 0

n U
ma


0C
-H
oa a

H C-






88






4-P

-0 4c
O ( NO r- 00

x v v v v

So 0 r nH o
U c) o03 r Nlio nOco
H 0 [ 0
> ) ( (N r 0
l *H H H 0C rH
(00

coco rt oo oo



ST-H nom msco mnco m om
r l 0 r- rl O mk rl

0 c C C C
s | o o o o
. C *r C *H C -H C -I
'c O'C Ofl OW) 0'f
4) 0 4- 0 0 4J O 0 4-O

U) O 'C) 0 --H -4 -
0 -1 H O O O-i ,i
C H 4-1 Qi ) H3 Q 0 )
u 4- Z) 4) 4-J r. ( -) r Q) 4C)
( 0 Q. 0 C 0 0) 0 )






SLO 0 N0 0 4

Q 00 Iq 0N 9 pi
) US H U-H U H UH


E no





I a L oI I
o N) N-m -1[ -O
gH . o 0
-- H
U ,r rHu r1 -4

4-) 0H
(d *
) C) -~ V 4J0



-H H H (N 0) L) H O
S 0 ) -H N m (n C)to L m' 0
- a) . o H . H H 0Q




r 0 H H -- HC 00- -H N
C)i


Ci) Q) 0 0 -/ 0N v
-O-' C )-o ac cli H c o i t




H *. a *LO .* *0 . 0 0 *

c 0 n v
(H C)0 C) o H
a C) s "- H-- H* H'-' 0





C 4 XI Z, 0
SE- C rl N l O rl > E r l 4 -x
I O C!) --- C) >- H- C rH- 3u
rl C) ** ** 4

c 11 11 N( H ( n C) 11 H1 c 11 11 11
-rl 3) 0
C) r[lI C) gIl Zl Zl C EI 2I d)i


















z C0 0

NOO0
N 0 0)


*00 m
000
k0 CO 0


HN CO --


-I rl -i-l


+1 +1

-1
01 0H 01 n o' 'O




0
g1 o
C rHl
0 4 \
C4 -H 4J 0w


U~H -- 4 -
'0-HO) HH 4
0 0 4- r-l H '
0 0H C < )










)O H
4
C0



SH 0)






aN ONr4 O







* o N O



Q-l



00 m


Sr(
0
O- i
















II HII i
xi rq r; i **2
si 1* ai si


Nino 'lN U
m 'X} *4)


)
e r U

+1+1

S000 zr N 0

NH
S- (NN C
H rim -Hr
II II


l0 4 C4


E D -I 4 HJ ,
C -H I-l 0 0
-H 0 ) H -- H


Co-m 4J r -
oo4 HH V
0 0C <
U a4
0
(4-

1'
0)
4J


noi o 0n
*. t *
i H 't O
N- N U




OO OnH 0


c0
HH OH 0

4


U'







CO I
NO 0 01 a
* - o -

OHH OH 0
CM d r H -





0)


II II Oo


10 H H -Q)
4-P


zwl i- ix o f


-NN
'00cc
co 0 00


*oo



000









nHm







0 0

4-4
00 00



NOO




00
C H









UN
0 -1











H H





NM
4, i




o cH
r<-i















*~ i


00
N r


+1+1
00 10
CO N

m

II II


4-




H H
-











OH'
H n






NN
n N







ON
rlv


U0N rHN




4J ..

..--I
0) -HI l
g| i-l 21 Sl0



















NOm
HOm





* 1 (N




+1+1 >1
'0
Hco 0


NO 0
NN --4

II I 0)


0 4J ~0

0 r t,- -
OC 0 -0 0
- O4 O 0-4 44


-000 4J -i
Hooc r0

UOIH O
0 0 r. *


44









A 0 0



00 HH H)
N Nr C



0

ON 0IN 4-

oH 0 0r H
1)



O HO CI















4-) 1-4
4J - r
11 4-411 110 (d


a)- ,'H 0
(M~Z (M P 0 PIr


O H0
000

V


o0 ,0 o H
4-ro X.m


N r--1 4 t -
H HH

4 ( '0
rl rlrl

- a0

0 +1+1

000 ri r0



IO II II


0 0
0 41 CU)
C -, H 4 .c u)
00C 4- -*H
- I0 0-'-I 4-



N00 N--I U
Uz H ) co 1






NHl +rlO H

o o < <





Ncl 0N 0
mu H M +


















UN N0 C)


0

*Q
.,4
4- 4

]I II I b (U
a) --H H
ZI --4 %


Cl H r-
OO
000
ooo

v


.zr 0 -N



rl




000







0
O
0 -H
00 r4-)
-r-l 0 U
4-) M0

















L 0
o H






) r-l








Ll H
M(N








r H






00

r 0
0 H H0
H1rl -


+I +I
+1+1




HH--
o r-




II II


4-)

4-4 M
a)l l
H C


< <











in






H 0
Cl












rio
Hn










rH-






*D
H-v


0d 0 :E
( -H
4-J-





O ,-
U 0


C =
*--4



S4-
0

O
I 0
n U

0)


41

14-411 11

21 I i aI o1















4'
-4

NOO
o**

V


Mr-1N co
(00




-HN
NNO

H 0 0












C
0
O
C -,
o C-1)
-r, OU


SOO

UPAH









0 0 c
\oN





o ^






HD 1)
ON
o N
u, iH




n 0
SCM
* *co
in H


-c

-H







1 1 a
0

l
q C










4-4
* 0





4 -H

0 -H 0'




-- 4
OO





01


'0

0n a
O 0



t e n


-O


Mm H




* ro

4)
0

In e
fcH




-
C -




4-1 k
^ --


41
II 1- II

einl zI


H c O c 0'

cO 00c -H

+1+1
co '0
NNN cc C 0
000


II II
en

C )
0 4- C"

0 4-) 4-4 i
-HOC)U e
'H U 0 HH 0-











4-N NN en
c-c 4-) 4 $-4
1m -i ) H -l










0 0 r- 0
NHH OH 0H










N- N- C)
C M







HH OH-1 0




N '- N U
* o * i


4-'


Ico r r-II co
- 0 CM 1-l





Sre
CS
ccc NH a- (
*. .co *. en







NHH OH 4
N'i-- N'-' H
HO- '?O e


NOO
CO r-l CO
N 0 cc
in o co

v








NO

000


-e H e



O
OC
0


0 r 4-
-H 0 U

4-) en


UNH






no

m CH




Nm
ICIH

eM H
n1-1

m U3



N H

nm H





m *N
en H


H- 0


+1+1









4-J
H n
II II










41
4-1 0











H H
me c








HH
H cc









N





M N

n


rC C 0


g 4
S -Hl
*a >9
C OH
0 0 -H



c 4



I 1
,14-
I 0O
n U

0e
C-l


E-


4


l4qI C-l X
a, *i-
















-K


NO

V

4O) O inci







0
, ooo


0 1 r'- 1
0


S000
rl1


0 00

4-) 4- -H
- -H OU

o) .I-HOa

61 CD)4)




00








O




0 0 H
4)
-C
cc
-rl











O





u


mul D





0l o
rl-
-O Ur~
(l -


61
CN



+1+1 0




U)
I1 u1 C
U




n.,i
a
0
)C 6

1-H 4







0
H 0-

4 -1



















3)
0 -IH Q)







a)




























r- r--1
co O


N
N* O


0




-
N 4 )







r- 3
H 6







cc 0


j


-K
-K
-K
LOHH
c 0

V

t H co
HONH














C
0
C -H
O

0 r 4-1
-H 0 (
4-) "1 fO

000
O-H ()








N l
) 4-')










0 0 H






H








* LA
c H





00
. i


[I II [I tn 0 II l II
a1 rI r -1 1
zI qzI z:w z zi 4 Zi


-K
-K


0Wo






O
4)







+1 +1
ONU




4) o -- rl-
*l El l

W C
('(N >9









0
0



















0 N
II II
If C C
























-1 n o 0

4- (i) O
a6 -H -H ou















rN ml o H













- 4 0 0
61 -HL 4)





























C 4) 000
-*
01



ON 0 iON


0 N -





61 --l 0








'0 1O NNH





N\inA 61 Irl

rN 0 'D


a V3 4 )


Ln

00)
o*



+1+1

N r




II II

,i

4-)
rn cr
co cc



cl rl

COtC








0 -
SM
N



COWn


H



co
NN

H N








N


d 00
r1


6 -H



C C H
V 0 -

-rH
C



*O H
0 a


c C
I 0




Full Text
STUDIES OF GENERAL AND SEXUAL DEVELOPMENT
IN VOLES (MICROTUS)
By
ALLEN L. SALO
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1992

This work is dedicated in remembrance of my father,
Raymond John Salo, who conducted comparative research on
two species of rabbits at the University of Massachusetts.

ACKNOWLEDGEMENTS
I would like to sincerely thank a number of people who
helped me complete the Ph.D. requirements. Although I am
not able to list all individuals who helped me in this
regard, a few should be named. First, I would like to thank
members of my family who supported my sometimes arduous trek
through academe. Most importantly my mother, Sonja Salo,
supported me spiritually as well as financially when the
going became difficult. My brother Mark Salo, his wife
Laurie, and children, were also constant sources of
inspiration. Other significant individuals include past and
present fellow graduate students: Florence Caputo, Jo
Manning, John Pierce, Joan Scheffer, Steve Taylor, Betty
Inglett, Sue Halsell, Nick Mills, and Kerry Sheehan.
Faculty members who have most strongly supported my
endeavors include Dr. Donald Dewsbury, Dr. H. Jane
Brockmann, and Dr. Jeffrey French. Certainly many other
members have impacted my life in various, positive ways.
Finally, I would like to thank members of the staff at the
University of Florida who helped in one way or another;
these individuals include: Theodore Fryer, Isaiah
Washington, Chris Wilcox, and Jeanene Griffin.

TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS iii
ABSTRACT vii
CHAPTERS
1GENERAL INTRODUCTION 1
Statement of the Problem 1
Plan of the Dissertation 4
Definitional Considerations 5
A Review of Mating Systems 6
Puberty Modulation in House Mice
fMus musculusl 7
The Social Biology of Voles (Microtus) 16
Puberty Modulation in Voles fMicrotus) 19
Principles Underlying Puberty Modulation
in House Mice and Voles 27
Problems with Previous Studies of Puberty
Modulation in Microtus 38
2 GENERAL METHODS: SUBJECTS, HOUSING, AND
APPARATUS 41
3 EFFECT OF OLFACTORY CUES UPON PUBERTY
MODULATION IN FOUR SPECIES OF VOLES
(EXPERIMENT 1) 43
Rationale 43
Method 45
Subjects 45
Procedure 4 6
Statistical Analysis 49
Results 50
Body Weight 51
Anogenital Distance 51
Adrenal Weight 52
Testes Weight 54
Seminal Vesicle Weight 54
Ovarian Weight 54
Uterine Weight 54
Status of Vaginal Perforation 56
Delay until Vaginal Perforation 57
Vaginal Smears 58
IV

Discussion 65
Body Weight 65
Anogenital Distance 67
Adrenal Weight 69
Testes Weight 7 0
Seminal Vesicle Weight 71
Ovarian Weight 72
Uterine Weight 73
Characteristics of Vaginal Perforation 75
Vaginal Cytology 77
Conclusions (Experiment 1) 80
4 BEHAVIORAL RESPONSES OF VOLES (MICROTUS)
TO PUBERTY MODULATING STIMULI
(EXPERIMENT 2) 116
Rationale 116
Method 121
Subjects 121
Procedure 121
Statistical Analysis 123
Results 124
Body Weights 229
Vaginal Smears 231
Preference Tests 124
Differences in the Time Near the Stimuli
Across Weeks 126
Between-Species Comparisons 127
Duration within Center of Cage:
Within-Species Analyses 130
Duration within Center of Cage:
Between-Species Analyses 130
Discussion 132
Olfactory Preference Tests 132
Between-Species Comparisons 134
Duration within Center of Cage 136
Conclusions (Experiment 2) 137
5 EFFECT OF SIRE PRESENCE OR ABSENCE ON
DEVELOPMENT OF OFFSPRING (EXPERIMENT 3) 145
Rationale 145
Method 151
Subjects 151
Procedure 151
Statistical Analysis 152
Results 153
Delay to Produce Litters 153
Number of Offspring Born 155
Age 50% of Offspring Opened Eyes 156
Number of Offspring Weaned 157
Sex Ratio of Offspring Weaned 157
Body Weight of Offspring Weaned 158
v

Discussion 161
Delay to Produce Litters 161
Number of Offspring Born 163
Age 50% of Offspring Opened Eyes 164
Number of Offspring Weaned 165
Sex Ratio of Offspring Weaned 166
Body Weight of Offspring Weaned 167
A Reanalysis of the Evolution and
Expression of Male Parental Behavior 169
6 GENERAL OVERVIEW AND DISCUSSION 183
Overview of Experimental Results 18 3
An Attempt at a Synthesis 18 3
Routes of Future Study 19 0
REFERENCES 19 5
APPENDICES
A BODY WEIGHTS OF MICROTUS (EXPERIMENT 1) 210
B ANOGENITAL DISTANCES (EXPERIMENT 2) 215
C ANALYSES OF NUCLEATED CELLS AND
LEUKOCYTES FROM VAGINAL SMEARS OF
MICROTUS (EXPERIMENT 1) 218
D BODY WEIGHTS OF MICROTUS (EXPERIMENT 2) 229
E ANALYSES OF CELLS FROM VAGINAL SMEARS OF
MICROTUS (EXPERIMENT 2) 231
BIOGRAPHICAL SKETCH 234
VI

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
STUDIES OF GENERAL AND SEXUAL DEVELOPMENT
IN VOLES ÍMICROTUS)
By
Allen L. Salo
December, 1992
Chairman: Donald A. Dewsbury
Major Department: Psychology
Three experiments were conducted to investigate how
patterns of general and sexual development might be
correlated with the formation of social and mating systems
among four species of voles (Micrqtus). Species included
pine voles fMicrotus pinetorunU , prairie voles
(M. ochroaaster^, meadow voles (M. pennsvlvanicus^, and
montane voles fM. montanuslâ–  In Experiment 1, general and
sexual development were monitored as voles were exposed to
pheromones contained in the soiled bedding from family
groups, adult males, or adult females. Few significant
effects were found to be due to the treatment. Male pine
voles exposed to family or male bedding were significantly
heavier than those exposed to clean or female bedding. The
uteri of female montane voles exposed to clean or male
bedding were heavier than the uteri of those exposed to
family or female bedding.
In Experiment 2, the olfactory preferences of voles
were measured when they were exposed to male and female
bedding on weeks 4, 7, and 10 after birth. Few preferences
Vll

were shown for either bedding type by any of the species.
Female prairie voles and meadow voles revealed a significant
preference for male versus female bedding. Both sexes of
all species differed little in the total duration they
remained near the female stimulus. Male meadow voles
remained near the female stimulus significantly less than
the males of the other species on week 10. Female montane
voles remained near the male stimulus significantly longer
on weeks 4 and 7 than did females of the other species.
In Experiment 3, the influence of the fathers' presence
and absence was studied during the rearing of the breeding
pairs' first two litters. Pine voles produced their second
litter considerably earlier if the male had been present
during the rearing of the first litter rather than being
absent. Pine voles weaned heavier offspring in the second
litter than in the first, when the male had been present for
the rearing of the first litter. Montane voles produced
litters that were male-biased in sex ratio across both
litters, if the male had been present during the rearing of
the first litter.
Results are discussed and interpreted from known
differences in their contrasting social and mating systems.
viii

CHAPTER 1
GENERAL INTRODUCTION
Statement of the Problem
Several theories have been advanced within the last two
decades to account for the evolution of different social and
mating systems among birds and mammals (e.g., Emlen & Oring,
1977; Kleiman, 1977; Orians, 1969; Vehrencamp & Bradbury,
1984; Wittenberger 1979). However, not all of the selective
forces that underlie the evolution of different mating
systems have been identified (see Vehrencamp & Bradbury,
1984). Wittenberger (1979) proposed that because mating
behavior is affected by nearly all aspects of an organism's
behavioral adjustment to its environment, a theory of mating
systems must be integrated with several other types of
behavioral theories. For example, theoretical advances
concerning the evolution of territoriality, parental
behavior, and sociality must be meshed within a broader
theory explaining the evolution of mating systems.
One area of study that appears promising for providing
greater insights into the evolution of social and mating
systems in mammals is research on the regulation of sexual
maturation or puberty modulation. For example, although its
function remains obscure, delayed sexual maturation is often
present among mammalian species that are considered to be
monogamous (Kleiman, 1977; Dewsbury, 1981). Some have
1

2
suggested that delayed sexual maturation functions to
inhibit incestuous mating (McGuire & Getz, 1981), although
others have suggested it reduces susceptibility to predation
(Batzli et al. 1977). Investigating the causes and function
of changes in sexual maturation may lead to an increased
understanding of the evolution of different social and
mating systems.
A second area of investigation that may offer insight
into the formation of the different mating systems is the
evolution of male parental care. Paternal care is found
most often among mammalian species that form monogamous
mating systems and have relatively few offspring (Kleiman,
1977; Dewsbury, 1981). Although it is often assumed that
paternal behavior has a beneficial effect on the development
of young, the assumption needs to be validated (Wuensch,
1985). The results of several studies conducted to
investigate this hypothesis among species of rodents are
equivocal (Dewsbury, 1988). It is possible that there are
important links between sexual suppression and the evolution
of paternal behavior, because the two often occur jointly in
monogamous mating systems (Kleiman, 1977; Dewsbury, 1981).
Until recently, the study of developmental processes
has been largely divorced from studies of the evolution of
mating systems. Muller (1990) noted that ontogeny has been
treated as a type of "black box" in evolutionary theory.
Similarly, Stearns (1989) discussing developmental
processes, noted that the types and sources of phenotypic

3
variation have been given little consideration in
evolutionary theory. The following set of experiments was
designed to investigate possible causal links between
puberty modulation and paternal care and the resulting
differences in social and mating systems in mammals.
Species of the genus Microtus are ideal for studying
the functions of sexual maturation and paternal behavior for
three primary reasons. First, the species display
differences in social and mating system that appear to be
associated with puberty modulation and paternal care. For
example, the mating systems among Microtus range from
monogamy in prairie voles (M. ochroaaster^ to promiscuity in
meadow voles (M. pennsvlvanicusl(Wolff, 1985). Pine voles
ÍM. pinetorunU and prairie voles are considered to be
monogamous and appear to be sensitive to the actions of
pheromonal cues that affect sexual maturation (Getz &
Hofmann, 1986; FitzGerald & Madison, 1983; Carter & Getz,
1985).
A second reason for studying these processes in
Microtus is that by investigating differences in puberty
modulation and paternal behavior with closely related
species, we are most likely to identify the selective
pressures and mechanisms that have shaped these species
differences (Clutton-Brock & Harvey, 1984; King, 1970;
Dewsbury, 1990). A third reason is that many species of
Microtus can be bred and maintained within the laboratory
where it is possible to systematically vary and control

4
exposure to stimuli. Such identification and control of
stimuli is difficult or impossible with these species under
natural conditions. The control offered in the laboratory
appears to be a prerequisite to identify the necessary and
sufficient stimuli that influence sexual maturation among
species. Finally, although there is a considerable
literature on such topics as the taxonomy, zoogeography,
anatomy, and habitats, of various species of Microtus,
currently little is known about many developmental processes
such as prenatal development, causes of interspecific
variation in litter size, and causes of mortality (Tamarin,
1985; Nadeau, 1985). It is possible that through the
systematic investigation of such phenomena, relationships
will become evident between them and the selective pressures
within the environment which shaped them.
Plan of the Dissertation
Three studies are presented that were designed to
investigate the role of two developmental phenomena that
might have influenced the evolution and expression of the
different social and mating systems among four species of
voles (Microtus). The two developmental phenomena include
the changes in the timing of puberty (puberty modulation)
and influence of paternal presence or absence upon
developing offspring. Specifically, Experiment 1 was
designed to explore how exposure to naturally occurring
pheromones, contained within soiled bedding, influences the
timing of puberty in four species of voles. The species

5
included pine voles (M. pinetorum). prairie voles,
(M. ochrooasterl, meadow voles CM. pennsvlvanicus), and
montane voles (M. montanus). In Experiment 2, the
behavioral reactions of the four species were studied when
they were exposed to soiled bedding in an odor preference
task. The study was designed to determine if behavioral
preferences for male and female soiled bedding were evident
within each species and sex and whether the preferences
changed as a function of age. In Experiment 3, the effect
of an adult male's presence or absence on his developing
offspring was studied. Together the three experiments were
designed to explore how these phenomena were related to
development and how they might be functionally linked to the
expression of different social and mating systems.
Below I present some definitional considerations,
present a review of the literature on puberty modulation in
house mice (Mus musculus) and voles (Microbus), present a
review of the social biology of Microtus. and present the
results of the studies designed to investigate puberty
modulation and the effects of paternal presence in Microtus.
Finally, I attempt to synthesize the results of the three
studies with results from other studies to explain possible
selective pressures that might have formed and may maintain
the different social and mating systems among these species.
Definitional Considerations
There is no globally accepted definition of puberty.
According to Hasler (1975), puberty has usually been defined

6
as the age that animals produce viable gametes. However, in
practice, it is often difficult to define the limits of
puberty with precision (Bronson & Rissman, 1986). Common
measures used in studies with female house mice have
included the day of vaginal opening or the first day of an
estrous vaginal smear (Drickamer, 1986). It has proven
difficult to find a reliable and non-terminal marker to
assess puberty in male mice. The presence of viable
spermatozoa or the ability to induce pregnancy have been
used in some studies as an index of puberty among males
(e.g., Vandenbergh, 1971).
Despite debate over how one defines puberty or what one
uses as an index of puberty, by chosing a standard measure
it is possible to compare differences in the onset of
"puberty" as being either accelerated or delayed compared
with some reference group. Thus, the term "puberty
modulation" is useful when referring to the onset of puberty
as being either accelerated, delayed, or both, when
comparing individuals or groups.
A Review of Mating Systems
The mating system of a population can be regarded as
the ensemble of behaviors and physical adaptations specific
for mating that are available to a population (Vehrencamp &
Bradbury, 1984). The mating system of a population is an
emergent property that reflects the traits and propensities
of the individuals within the population (West-Eberhard,
1979). For example, characteristics of a monogamous mating

7
system among mammals include (1) the continual close
proximity of an adult heterosexual pair during and outside
periods of reproduction, (2) mating preferences, (3) the
absence of unrelated adult conspecifics from the breeding
pairs home range, and (4) breeding by only one adult pair in
a family group (Kleiman, 1977). Other common forms of
mating systems include polygyny, in which there is a
prolonged association and essentially exclusive mating
relationship between one male and two or more females at one
time, and promiscuity, in which there is no prolonged
association between the sexes and multiple matings by
members of at least one sex (Wittenberger, 1979). The goal
of research has been to identify the primary selective
pressures or foci that have created the behavior differences
associated with each mating system, with the use of
appropriate theory and field work (Vehrencamp & Bradbury,
1984) .
Puberty Modulation in House Mice (Mus musculus)
Most investigations of the phenomenon of puberty
modulation have been conducted with house mice
(Mus musculus). Researchers have identified several factors
that can influence the timing of puberty in young female
mice. Many studies have used indices such as the day of
vaginal perforation or the detection of an estrous smear as
a reliable means of assessing puberty. Considerably less
research has been designed to investigate the factors that
affect puberty in male mice. However, research with both

8
sexes appears necessary to understand how differences in the
timing of puberty might affect the expression of social and
mating systems.
A number of factors have been shown to influence the
timing of puberty in female house mice. Broadly, these
factors have included genetic differences (Drickamer, 1981),
social conditions (see Drickamer, 1986 or Vandenbergh &
Coppola, 1986 for recent reviews), and non-social factors
such as temperature (Barnett & Coleman, 1959), photoperiod
(Drickamer, 1975a), and season (Kruczek & Gruca, 1990). The
current review and experiments have been based on the
investigation of social factors that have been shown to
influence puberty. Specifically, chemosignals or
pheromones, found in the urine of several muroid species,
have been shown to differentially accelerate or retard the
attainment of puberty in young individuals (Levin &
Johnston, 1986) .
At least four types of urinary chemosignals can
influence puberty in female mice (Drickamer, 1986). These
signals include urine from (1) male mice, (2) pregnant or
lactating females, (3) females in estrus, and (4)
group-caged females. The first three signals accelerate the
timing of puberty relative to female mice housed similarly
but not exposed to the chemosignals. The last signal delays
puberty in females compared to those not exposed to the
signals. Below, I present the principal findings associated

with each of the four stimuli that affect puberty in female
house mice.
9
Puberty Acceleration Caused bv Male urine
Not all urine from male mice is effective in
accelerating the onset of puberty in females relative to
those treated similarly but not exposed to the male urine.
Dominant males release an acceleratory substance in their
urine whereas subordinate males do not (Drickamer, 1983a).
Similarly, prepubertal or castrated males do not release an
acceleratory substance in their urine (Vandenbergh, 1969;
Lombardi et al. 1976). Neither the specific type of housing
condition nor the degree of genetic relatedness appears to
mediate the effectiveness of the male pheromone directly.
Drickamer (1983a) found that increasing the density of males
does not alter the pattern of the chemical's release. In
general, the effectiveness of the male chemosignal in
causing pubertal acceleration in females is robust.
Exposing females to amounts of male urine as small as 0.0001
cc per day effectively advances puberty in mice (Drickamer,
1982b, 1984a) .
Despite the chemical's apparent acceleratory strength,
such factors as the season or ratio of light-dark exposure
can affect the female and thus indirectly affect whether the
male's chemosignal advances puberty. The substance appears
most effective in producing acceleration when it is
collected from males during light onset and presented to
females at this same time period (Drickamer, 1982a). The

10
effectiveness of the male chemosignal appears to vary
because of changes in sensitivity of young female mice to
the substance and not due to differences in the substance
excreted by males (Drickamer, 1986). Both male and female
mice have unique patterns of urine deposition that could
influence the likelihood that puberty will be modulated in
juveniles (Drickamer, 1989a).
Certain social conditions can influence the
effectiveness of the acceleratory substance to induce
puberty in females. Vandenbergh (1967) showed that the
acceleratory effect is greater when males are in physical
contact with females; contact-stimulation appears necessary
for the effect to occur (Drickamer, 1974, 1975b; Bronson &
Maruniak, 1975). The active chemical in male mouse urine
appears to be relatively nonvolatile and under natural
conditions it can influence females for some time after the
male has deposited urine (Drickamer, 1986). The substance
in male urine has been found to be effective in accelerating
puberty within three days of exposure (Colby & Vandenbergh,
1974). Acceleration can be accomplished by exposing females
to either male urine for two hours per day or to an intact
adult male for one hour per day (Drickamer, 1983a). Thus,
these findings show that although the age at which female
house mice reach puberty can be reduced by the actual
presence of an adult male, his presence is not necessary for
the acceleration of puberty.

11
The male chemosignal appears to have its physiological
influence in females via the vomeronasal-accessory olfactory
system (Drickamer & Assmann, 1981). This system may be a
common pathway affecting puberty modulation in various
species. Ablation of the vomeronasal organ eliminates the
ability of both mice and voles to receive pheromonal signals
(Vandenbergh, 1988; Lepri & Wysocki, 1987). Other work
investigating the physiological processes of male-induced
puberty acceleration has been conducted but will not be
reviewed here (see Reiter, 1982; Carter et al. 1986 for
review).
Puberty Acceleration Associated with Pregnant and/or
Lactatina Females
Drickamer and Hoover (1979) first documented that the
urine of pregnant or lactating females accelerates sexual
development in female house mice. It is believed that the
active chemosignal is the same that is produced by females
in both reproductive conditions, because a number of
experiments revealed no clear differences between the
chemosignals (Drickamer 1986). Three days of exposure to
urine collected from either pregnant or lactating females is
necessary for puberty acceleration to occur in females that
are less than 30 days old (Drickamer, 1984b). However, the
substance does not cause acceleration in females when it is
presented during the winter months. This lack of an
acceleration in puberty appears due to differences in the
sensitivity of young females to the cues that correspond to

12
changes in the season (Drickamer, 1986) . The acceleration
of puberty does not appear to differ as a function of
kinship among females (Drickamer, 1984c).
Under certain conditions, urine from reproductively
active females does not cause puberty acceleration in
females. For instance, when reproductively active females
are caged in groups or housed with non-reproductively active
females, their urine becomes ineffective in causing either
acceleration or delay in recipient females (Drickamer,
1983b). If young female mice are exposed to urine from both
reproductively active females and urine from group-caged,
non-reproductively active females, the exposed females are
delayed in reaching puberty (Drickamer, 1982c; 1986).
The chemosignals associated with reproductively active
females appear to be relatively volatile and are effective
in causing acceleration when presented in volumes slightly
larger than those of the male signal (0.03 cc of urine per
day; Drickamer & Hoover, 1979; Drickamer 1982b, 1983c).
One difference between the acceleratory chemosignals of
males and those of reproductively active females is that the
chemosignals from females are effective in causing
acceleration regardless of when they are collected or what
time of the day they are presented to females. As noted
previously, the urine from males is most effective in
accelerating puberty when it is collected and presented to
females during the onset of the light portion of the
light-dark cycle (Drickamer, 1982a).

13
Puberty Acceleration Caused by Females in Eatrus
Less is known about how the particular phase of the
estrous cycle influences the chemosignals of the female
donors that can influence puberty in other females.
However, urine from singly-caged female mice in estrus
decreases the latency until puberty is reached in young
female mice. This effect is not seen when urine from
single-caged diestrous females is presented to prepubertal
females (Drickamer 1982c, 1984c).
The chemosignal in urine produced by estrous female
mice is effective in advancing puberty in about 3 days and
is effective when presented in small quantities such as
0.001 cc of urine per day (Drickamer, 1986). No differences
in acceleratory effects were found when urine was collected
from related or unrelated estrous females (Drickamer,
1984c).
Puberty Inhibition Caused by Group-Housed Females
Only one type of endogenously produced chemosignal has
been found to inhibit or delay puberty reliably in female
mice. Group-housed females produce a substance in their
urine that delays puberty in other females. The time that
estrous smears are detected in young female mice is delayed
when they have been exposed to either soiled bedding of
group-housed females or have had their nares painted with
urine from group-housed females (Drickamer, 1982c).
Females exposed to the urine of group-housed females
will begin to produce the inhibitory substance themselves.

14
The delay signal's effect occurs later than the effect of
any acceleratory chemosignal. Specifically the delay signal
requires four to seven days of exposure for inhibition to
occur (Drickamer, 1977). Season also influences the degree
of sensitivity young females have to this signal. Females
are not delayed in reaching puberty when exposed to the
substance during the summer, although during other seasons
they are delayed in its presence (Drickamer, 1986).
Puberty Modulation in the Natural Environment
Although the phenomenon of puberty suppression or
advancement is clearly established for house mice in the
laboratory, it is appropriate to question whether similar
processes occur within wild populations. A few studies have
provided evidence that both puberty acceleration and
suppression take place under natural conditions.
Massey and Vandenbergh (1980) conducted a series of
experiments with populations of wild house mice enclosed
within highway cloverleaf sections over a 2-year period.
The urine from the females of two populations was collected
via live-trapping during the spring, when population
densities were low. This urine did not have an effect on
the age of first vaginal estrus of laboratory-housed
females. However, urine collected from the second
population in December, when the population was crowded,
significantly delayed the first estrus in females when
compared with control females. Thus, the researchers

attributed the delay of puberty as being due to the
increased density in the second population.
15
Urine collected from wild male mice of the highway
"island" populations accelerated puberty in laboratory
females by an average of seven days as measured by age at
first estrus. Puberty acceleration occurred in response to
the male urine despite changes in the season or population
density at the time the urine was collected (Massey &
Vandenbergh, 1981). Thus, it appears that male chemosignals
and subsequent acceleration may be a more prevalent
influence upon puberty in females than the delay chemosignal
produced by female mice under crowded conditions.
In another set of experiments, acute population
explosions were created by Vandenbergh and Coppola (1986)
who introduced 40 second- or third-generation wild female
house mice onto highway islands. This procedure allowed for
a more critical test of the causal relationship between
population density and the release of delay chemosignals.
Urine samples were collected from females on each island
during monthly intervals. Urine samples collected three
weeks after the addition of the females onto the island
populations caused an average delay in first vaginal estrus
of 5.3 days in laboratory-reared females. Thus, this study
provided additional evidence that the puberty-delay
pheromone can be produced by females in response to acute
increases in female density under natural circumstances
(Vandenbergh & Coppola, 1986).

16
The Social Biology of Voles (Microtus)
A considerable amount of information exists on the
social and mating systems of several species of voles
(Microtus; see Wolff, 1985; Keller, 1985 for reviews). The
social behavior of Microtus is both complex and variable
(Wolff, 1985). Most species are active both during the day
and night, and make use of above-ground runways as well as
burrows. However, there are substantial differences in
patterns of territoriality and of mating system. Below are
listed some of the primary social differences among four
species of Microtus. which are the four species studied in
the following experiments. Broadly, their differences
portray the degree of diversity within the genus Microtus.
Montane voles. Montane voles (M. aontanus) are
considered to commonly form a polygynous mating system,
although facultative monogamy may occur at low densities
(Jannett 1980; 1982). Both males and females typically
defend exclusive territories against other same-sexed
individuals (Jannett, 1982) . The territories of males
generally overlap the territories of one or more females.
Thus, female montane voles appear typically to mate with
only one familiar male, although no pair bond is formed
(Wolff, 1985). Montane voles become more social and form
aggregations during the winter months (Madison, 1984). The
adult males of the species are typically larger than the
females (males 35% larger; Dewsbury et al. 1980).

17
Meadow voles. The mating and social system of meadow
voles I'M, pennsvlvanicusl differ in some respects from
montane voles, and many data have been gathered on their
social behavior (e.g., Madison, 1980, 1984; Webster &
Brooks, 1981). Meadow voles appear to form a promiscuous
mating system in which reproductively active females defend
territories. In contrast to montane voles, the home ranges
of males may overlap those of several other males as well as
those of females (Wolff, 1985). Cases of one female mating
with several males have been reported (e.g., Webster &
Brooks, 1981). Meadow voles form aggregations during the
late fall and winter months as do montane voles (Madison,
1984). The degree of sexual dimorphism in body weight is
also relatively high, with adult males being larger than
females (males 22% larger; Dewsbury, et al. 1980).
Prairie voles. In contrast to montane voles and meadow
voles, prairie voles (M. ochroaasteri appear to be more
social and form a monogamous mating system; a communal
nesting group appears to be the basic year-round social unit
(Getz et al. 1990). Evidence from the laboratory also
suggests monogamy. Both the breeding male and female engage
in extensive parental care of offspring (Getz & Carter,
1980), and male prairie voles failed to demonstrate the
Coolidge effect when presented with a new estrous female
after reaching sexual satiety (Gray & Dewsbury, 1973). Both
males and females appear to be territorial and are
aggressive to animals of the opposite sex. Both sexes of

18
breeding pairs defend a common home range that is
approximately the same size (Gaulin & FitzGerald, 1988).
The degree of sexual dimorphism in body weight is reduced in
comparison to the dimorphism in montane voles and meadow
voles, although the male is somewhat larger than the female
(males 17% larger; Dewsbury et al. 1980).
Pine voles. Pine voles (M. pinetorund also appear to
be a highly social species and appear to live in communal
groups (FitzGerald & Madison, 1983; Schadler, 1990). The
mating system of pine voles appears to be monogamous,
although cooperative polyandry has been suggested as an
alternative system (FitzGerald & Madison, 1983). However,
more recent evidence suggests that breeding is commonly
restricted to the founding parents of a communal group
(Schadler, 1990) . The existence of a monogamous mating
system had also been predicted to be most likely for pine
voles among eight species of Microtus reviewed by Dewsbury
(1981). Other evidence of monogamy includes male
participation in the rearing of young, including the
retrieval and brooding of infants (Shadier, 1990; Oliveras &
Novak, 1986).
In contrast to many other species of voles, pine voles
are almost entirely fossorial, and come to the surface only
occasionally to feed (Wolff, 1985). Pine voles also appear
to be territorial. FitzGerald and Madison (1983) found that
each family had a discrete non-overlapping territory from
other family groups. Differences in body weight between the

19
sexes appear minimal in pine voles (males 2% smaller;
Dewsbury, 1990).
Puberty Modulation in Voles (Microtus)
Although several researchers have clearly demonstrated
that puberty delay and acceleration occur in a few species
of Microtus, such as prairie voles (M. ochrogaster; Carter
et al. 1986) and California voles (M. californicus; Batzli
et al. 1977), others have made weaker claims that such
processes do or do not occur in other species of Microtus.
For example, Jannett (1978) referred to field evidence that
suppression occurs in montane voles (M. montanusl and Batzli
et al. (1977) claimed that suppression does not normally
occur in meadow voles (M. pennsvlvanicus). Below, I review
the primary literature regarding the four species of
Microtus that were used in the later series of experiments.
Prairie voles. Most research concerning puberty
modulation in Microtus has been conducted with prairie
voles, where the presence of puberty suppression and
acceleration are strongly supported (see Carter et al, 1986
for review). In an early study that indicated puberty delay
occurs in prairie voles, Hasler & Nalbandov (1974) examined
pairs of weanling females caged with males with different
characteristics. Weanling females that were caged with a
littermate male had significantly longer latencies until
vaginal opening and production of a first litter than
compared to females that were caged with either a
non-1ittermate male or an adult male. The range of days

20
until vaginal opening occurred among these groups was
substantial. Females kept with littermate males became
perforate 30 days after pairing, on average, compared to
females paired with nonlittermate males that became
perforate eight days after pairing.
Batzli et al. (1977) found evidence that both sexes of
prairie voles could be suppressed developmentally, as
assessed by differences in body weight, when they were
housed with littermates. During this period, females
remained vaginally imperforate and males had abdominal
testes. After siblings were paired with unfamiliar adults
of the opposite sex, rapid increases in body weight and
subsequent reproduction occurred.
The stimuli necessary for the acceleration of puberty
in prairie voles have been studied. Young female prairie
voles, when exposed to a sexually experienced male for a 1 h
period, showed significant increases in uterine weight as
soon as 2 days after exposure (Carter et al. 1980). Drops
of male urine applied to the upper lip of young females
caused significant increases in uterine weight compared with
females caged alone, with a female sibling, a castrated
male, or exposed to the urine of a castrated male.
In another series of experiments, kinship per se was
shown not to be a limiting factor for the advancement of
puberty in prairie voles (Carter et al. 1980). When urine
from male siblings was applied to the nares of females for 6
consecutive days, it led to significantly heavier uterine

21
weight in females, compared to females housed with either a
sibling male or with a sibling male and treated with water
placed on the nares.
Together, the results of studies with prairie voles
suggest that a chemical agent or pheromone present in male
urine can be passed from male to female by direct contact.
From behavioral observations, Carter et al. (1980) suggested
that the active male pheromone was transmitted by
naso-genital contact between the sexes. However, sibling
prairie voles rarely engage in naso-genital investigation
and thus the lack of investigation between siblings may
function as a barrier to reproductive activation and
incestuous matings (Carter et al. 1986).
Although it is not clear how siblings inhibit
reproduction in other siblings, young female prairie voles
appear to be a primary source of reproductive suppression
for other male and female siblings. Getz et al. (1983)
found that for 15 suppressed litters, all but 1 of 32
females (or 97%) reared with sibling females were
suppressed, while only 3 of 15 (20%) reared with sibling
males were suppressed. A similar finding was apparent in
males; 10 of 11 males (91%) reared with sibling females were
suppressed and only 3 of 9 males (33%) raised with sibling
males were suppressed. The results suggest that for prairie
voles, the sex of siblings can have a substantial effect
upon the timing of puberty.

22
Additional work with prairie voles has revealed that
the suppressive effect of female presence appears largely
due to chemosignals present in their urine (Getz et al.
1983). In this study, virgin prairie voles were first
reproductively stimulated by exposing them to an unfamiliar
sexually experienced male for a one-hour period, then the
females were housed in a variety of conditions. The uteri
were weighed 48 hours later. The uterine weights of females
that had urine from either a female sibling, a virgin
non-sibling, or a pregnant female, placed on their nares did
not differ significantly from the uterine weights of
non-stimulated females. The only significant increase in
uterine weight was in the group of females that were
stimulated by the male and then maintained alone. Thus, the
results indicate that urine from female prairie voles is
effective in counteracting or suppressing puberty in female
prairie voles.
Although it is not known how the chemosignal is
transmitted between females, activation by the non-volatile
male acceleratory pheromone appears to be caused by mutual
anogenital investigation between the sexes (Carter et al.
1980). Hofmann and Getz (1988) found that virgin prairie
voles that were exposed frequently to unfamiliar males could
"override" the reproductive suppression typically
experienced by females that remained within a family group.
Thus, it seems plausible that under natural conditions where
density is dramatically increased, such as during the growth

23
phase of a population cycle, normal suppression can be
counteracted by male presence. In support of this argument,
Getz and Hofmann (1986) found that of free-ranging prairie
vole females that remained at the natal nest at low
population density, 18% were reproductively active, while
77% became reproductively active at high density. All
females that dispersed from the natal nest were found to
became reproductively active.
Pine voles. Both puberty delay and acceleration appear
to occur in pine voles, M. pinetorum (Schadler, 1983; Lepri
& Vandenbergh, 1986). Schadler (1983) presented evidence
that male siblings are an important source of reproductive
inhibition for female siblings. Mating was inhibited
between a female and an unrelated male, while a male sibling
was sequestered behind a wire mesh barrier. Other evidence
of puberty delay was shown by Lepri and Vandenbergh (1986)
when they placed young female pine voles with adult males
into two type of cages; one with clean bedding or another
with bedding that had been soiled for two weeks by the
female's family. Forty-eight hours later, the females'
uteri and ovaries were removed and weighed. The uteri and
ovaries of those placed in the clean cage were significantly
heavier than those placed in the cage soiled by the family.
In another experiment, Lepri and Vandenbergh (1986)
demonstrated puberty acceleration in pine voles by exposing
4-week-old females to feces and urine from animals in one of
the following groups: (1) intact males, (2) castrated males,

24
(3) singly-housed females, (4) group-housed females, or (5)
control (clean cage). When uterine and ovarian weights were
compared, results indicated significant increases in both
organs from females exposed to the stimuli from the males of
either groups compared those from the control females. No
significant differences were found when the organs of the
control group were compared to those in the two groups that
received female stimuli.
Meadow voles. Although there is evidence that pubertal
acceleration occurs in female meadow voles (Baddaloo &
Clulow, 1981) , the evidence is equivocal for any type of
puberty delay. Pasley & McKinney (1973) presented evidence
that females caged in groups of eight had smaller ovaries
and uteri than singly-housed females. However, it is
possible that this type of sexual suppression was due to
overcrowding and stress and not due to pheromones. Batzli
et al. (1977) reported they did not find suppression to
occur in five of six litter-housed groups of meadow voles.
Additional data are needed to determine if sexual
suppression occurs in meadow voles as a result of pheromones
and not due to overcrowding and stress.
Montane voles. Fewer reports of pubertal acceleration
or delay can be found scattered in laboratory and field
reports for other species of Microtus. Field evidence
suggests that pubertal delay occurs in montane voles
(Jannett, 1978), although a critical test of this phenomenon
has not been reported. Recent laboratory evidence has shown

25
that puberty acceleration occurs in female montane voles
when exposed to males. Sawrey and Dewsbury (1991) found
shorter latencies until vaginal perforation and first
cornified (estrous) smears in females housed across a
wire-mesh barrier from males, as compared with females
without a male present. Controlled studies are needed to
determine if puberty delay occurs in montane voles when they
are exposed to pheromones.
It is possible that not all species of Microtus show
both delay and acceleration of puberty, or at least are not
affected to the same degree by comparable stimuli. Evidence
suggests that suppression does not occur, or occurs only
weakly, in female tundra voles fM. oeconomus) when they are
housed with male siblings (Facemire & Batzli, 1983). Batzli
et al. (1977) reported morphological differences between
California voles and prairie voles. California voles that
were housed with littermates grew more slowly than controls,
although females became perforate and males became scrotal
at about the same time in the littermate group as did the
controls (30 days). This pattern differed from prairie
voles in which males typically developed abdominal testes
and females remained vaginally imperforate when littermates
were held together. In both species, however, reproduction
by littermates was usually delayed until they were paired
with unfamiliar animals. Thus, there may be a variety of
physiological and behavioral differences related to puberty
modulation among various species of Microtus. The

26
illumination of the similarities and differences in patterns
of puberty modulation may prove instrumental in
understanding the dynamics of their social and mating
systems.
Puberty Modulation in Male Voles (Microtus)
Considerably less research has been conducted on the
various factors that can influence puberty modulation in
male Microtus. although cues from males may be critical for
influencing puberty in other siblings among some species.
Evidence for California voles suggests that male siblings
influence the sexual development of one another. Batzli et
al. (1977) found that the growth rates of males were
suppressed when they were paired with another male sibling,
although growth rates were not affected when they were
paired with a female, a male non-sibling, or with a female
sibling.
Other evidence suggests that the odors from family
members of California voles suppress male pubertal
development. In this species, the odors from mature males
or sires did not appear either to accelerate or inhibit the
sexual development of males (Rissman et al. 1984). However,
male California voles that were reared in the presence of
family bedding material and subsequently paired with another
non-relative female failed to cause increases in uterine
weight within four days (Rissman & Johnston, 1985). They
suggested that the lack of stimulation in the females was
due to low levels of circulating androgen in the males.

27
Thus, odors associated with a reproductively active female
might be one source of pubertal suppression in male
California voles.
A few anecdotal accounts of puberty modulation among
males of other species of Microtus exist. For example,
juvenile male common voles CM. arvalis) have been reported
to show reduced testicular development when housed near
cages with crowded adults (Lecyk, 1967). In summary, it
appears that although males of various species of Microtus
may be reproductively modulated by specific stimuli, the
limited data make conclusions premature. There is a paucity
of information regarding pubertal modulation in male voles,
and for males of other rodent species as well. It seems
worthwhile to correct this bias with future studies.
Principles Underlying Puberty Modulation in House Mice
and Voles
There has been little effort to integrate the large
body of literature into an effective theoretical framework
to study the relationships between primer pheromones, age at
first reproduction, and demographics (Vandenbergh & Coppola,
1986). Many of the existing hypotheses regarding puberty
modulation are based upon studies of house mice (Mus
musculus) because most of the research has been conducted
with this species. Vandenbergh and Coppola (1986) suggested
that some inferences may be applied to other species, such
as voles. However, they caution that given the differences
in the reproductive biology among genera, it seems unlikely

28
that the selective pressures influencing puberty modulation
will be identical across them. Observed similarities among
genera or species must be interpreted carefully when
attempting to formulate general principles.
Below I list some of the most prominent frameworks that
appear useful for integrating our understanding of the
causes and functions of the pheromonal effects upon puberty
modulation found among species such as house mice and voles.
A Life-History Theory of Puberty Modulating Pheromones
One of the broadest theoretical frameworks to
investigate and interpret patterns of puberty modulation is
the life history theory proposed by Vandenbergh & Coppola
(1986). According to this approach, life history tactics
are evolved sets of coadapted traits designed to solve
particular ecological problems (Stearns, 1976). Whether
animals typically reach sexual maturity relatively early or
late, all are viewed to have been selected to maximize
lifetime reproductive success.
Selective factors favoring puberty acceleration.
Life-history theory predicts that for increasing
populations, or those with large fluctuations or repeated
episodes of colonization, early maturity will be favored
(Vandenbergh & Coppola, 1986). Such characteristics
describe feral house mouse populations. In general, it is
believed that selection will favor early and total
investment by individuals to produce the maximum number of
young whenever the environment is highly favorable for

29
reproduction. They note that previous theories (e.g., Cole,
1954; Lewontin, 1965) have indicated that in a rapidly
expanding population, the age at first reproduction should
be driven to the physiological minimum by natural selection.
Species that possess a combination of life-history traits
including an early age of first reproduction, the production
of many young, and semelparity have been classified as being
"r-selected" (MacArthur & Wilson, 1967; Pianka, 1970). This
combination of traits is often present in species that
experience rapid population growth in favorable
environments.
Several factors are thought to favor early
reproduction (Vandenbergh & Coppola, 1986). First, early
reproduction will be favored as reproductive costs decrease
(Schaffer & Elson, 1975). Reproductive cost is the
deleterious effect of present reproduction on future
survival and/or fecundity. Early reproduction will also be
favored when the reproductive value of individuals decrease
as they grow older (Gadgil & Bossert, 1970). These patterns
suggest that early reproduction will be favored when it does
not negatively influence a female's ability to reproduce
later in life and thus lower lifetime reproductive success.
Delayed reproduction (nubertv) in house mice. The
function of delayed reproduction or puberty is more obscure
than the function of puberty acceleration (Vandenbergh &
Coppola, 1986). They suggest that it is difficult to
explain why opportunistic species, such as house mice,

30
benefit from delaying puberty and reproduction. However,
whatever the function(s) of puberty delay, it has been found
in several mammalian species (Vandenbergh & Coppola, 1986).
There appear to be general conditions that favor
delayed reproduction. Delayed reproduction could evolve if
individuals gained fecundity or produced better guality
offspring (Vandenbergh & Coppola, 1986). In general, it is
believed that many of the demographic or environmental
factors that favor delayed reproduction are the opposite of
those favoring early reproduction. Some have suggested that
delayed reproduction would be selected in stable populations
at or near the carrying capacity of the environment (Cole,
1954; Lewontin, 1965) or in declining populations (Hamilton,
1966; Mertz, 1971). Selection in saturated environments,
which favors the ability to compete and avoid predators, has
been referred to as "K-selection" (MacArthur & Wilson, 1967;
Pianka, 1970). Traits correlated with this type of
selection often include late maturity, the production of few
and large offspring, long life spans, and extended parental
care.
There are a few causal factors that are thought to
underlie delayed reproduction (Vandenbergh & Coppola, 1986).
For example, as the reproductive costs increase, in terms of
adult mortality, and as the reproductive value an individual
can accrue by not reproducing increases with age, delayed
reproduction will be favored. If reproductive success is

contingent upon age, size, or social status, delayed
reproduction is also favored (Geist, 1971).
31
Despite the general conditions that have been proposed
to predict under what circumstances early or late puberty
will occur, the current state of life-history theory allows
us to draw only rather vague and imprecise conclusions about
the timing of reproductive maturity (Vandenbergh & Coppola,
1986) . Despite this drawback, life-history theory provides
a broad theoretical background from which to interpret new
information. In addition to the previous conditions which
favor puberty acceleration or delay, there are a few
additional principles that may guide our understanding of
these processes. Below, I summarize two of the most
prominent themes which have emerged.
Puberty Modulation (Mutualism (Cooperation! versus Conflict!
One assumption that underlies most explanations for the
evolution of puberty modulation is that puberty modulation
is either of mutual benefit to the sender and receiver of
pheromones or that the sender and receiver of the pheromones
are in some form of conflict.
Examples of puberty modulation in the context of
cooperation. Bronson (1979) proposed that adult male mice
and adult or prepubertal females both gain in fitness
through mutual stimulation of reproductive acceleration via
pheromonal transfer. In this hypothesis, male urinary
pheromones cause the release of LH in recipient females that
speeds the attainment of puberty and ovulation.

Reciprocally, female pheromones cause increases in LH in
males which ultimately leads to increases in testosterone
32
and additional pheromone synthesis in males. Such a system
could greatly enhance the speed at which males and females
could reproduce, thus benefiting both sexes of breeding
pairs.
A second proposed benefit of mutualism was raised by
Vandenbergh and Coppola (1986) who proposed that the
detection of the puberty-delay and acceleratory pheromones
produced by adult female house mice provides information to
young females about the general quality of the environment
for reproduction. Presumably under crowded conditions,
adult females release puberty delay pheromones that benefit
both the sender and receiver to postpone reproduction. In
more favorable environments for reproduction, the release of
pheromones that accelerate puberty by females would hasten
the speed at which puberty and reproduction would occur
among female offspring. Presumably, the young females,
their mates, and kin would benefit from more optimally timed
reproduction.
A third example where mutual benefits may occur between
the pheromone senders and receivers is between parents and
offspring in some cooperatively breeding species (Emlen,
1984). Cooperative breeding refers to any situation when
more than two individuals provide care in the rearing of
young (Emlen, 1984). In this example, puberty delay and

inbreeding avoidance may have been co-evolved traits in
species where cooperative breeding occurs.
33
It appears that certain ecological constraints, such as
a saturated habitat, favor the retention of subadult
individuals within a family group, at least until conditions
enable independent reproduction. Emlen (1984) suggested it
is possible that when the probability of successful
dispersal and independent breeding are low, average fitness
may be increased by remaining in a group until group size
reaches some optimum. Other benefits that may occur in
cooperatively breeding groups include: (1) benefits to
helpers that gain breeding experience; (2) inheritance of
the parental territory; (3) dispersion in groups when
competition for reproductive vacancies is strong; and (4)
some form of reciprocity may take place, such as when one
individual forgoes breeding and helps another rear offspring
with likelihood of return aid at a later point in time; and
(5) increased inclusive fitness through aiding in the
rearing of close relatives (Emlen, 1984). Thus, both the
direct and indirect components of inclusive fitness might be
increased via reproductive suppression when accompanied by
helping behavior.
A final proposal for mutualistic benefit of
reproductive suppression, or the lack thereof, in voles has
been proposed by Christian (1970). Christian (1970)
emphasized the selective pressures exerted by the habitat
where a particular species evolved. Species such as meadow

34
voles, which evolved in habitats that were patchy and
ephemeral, such as moist meadows, evolved mechanisms that
aided in the colonization of newly created habitats. Thus,
mechanisms that led to successful dispersal were adaptive
for all members. He proposed that a density-dependent
endocrine response that led to increased aggression with
subsequent dispersal would serve this function. Hence, a
lack of sexual suppression among siblings would be expected
in species that evolved under these circumstances.
In contrast, species such as prairie voles that evolved
in the continuous and extensive habitats of the great plains
of central North America were believed to have evolved
mechanisms that enabled a greater level of social tolerance
and be more sensitive to reproductive inhibition (Christian,
1970). It is plausible that mechanisms to inhibit
incestuous mating among family members might be most evident
in highly social species.
Puberty modulation in situations of conflict.
Different researchers have proposed instances where there
appear to be conflicts of interest between pheromone senders
and receivers. Bronson (1979) proposed there was an
antagonism between female house mice to inhibit puberty and
reproduction. Such antagonism between females may be
accomplished either directly through aggression or through
the production of delay pheromones.
Wasser and Barash (1983) proposed a similar hypothesis.
They stressed the role of female-female competition to

35
reproduce as a key driving force to explain the high rates
of general reproductive failure among the females of many
species of mammals. The antagonistic pheromonal influence
of reproduction between female prairie voles conforms to
their theory (Getz et al. 1983)
Competition among the males of some species via
pheromones may also be present in some species. Although
Bronson (1979) noted the high rate of aggression observed
between territorial male house mice and other males, he did
not discuss pheromonal interactions among them. Vandenbergh
(1971) found that adult male house mice had an inhibitory
effect on the reproductive development in young males, while
the presence of adult females accelerated their development.
The possibility that males may inhibit puberty and
reproduction of other males in other species through
pheromonal communication should remain open for continued
investigation.
Sociallv-Dependent Versus Sociallv-Independent Systems
Another framework within which to view and investigate
the reproductive influences of pheromones on the timing of
puberty can be called the socially-dependent versus
socially-independent dichotomy. The essence of this
framework is that some species of Microtus appear to be more
dependent upon the direct exposure of other animals in order
to become reproductively mature and active. Taylor
(1990/1991) characterized two types of estrus induction
patterns among females of different species of Microtus. He

36
provided evidence that the highly social prairie voles,
could be characterized as being male-dependent. Females
must typically be exposed to stimuli from males for a
relatively long period of time before estrus is attained.
In contrast, other less social species, such as meadow voles
and montane voles, were referred to as being
male-independent (Taylor 1990/1991). Species of this type
do not attain estrus totally independent of male
stimulation, but rather are much less dependent on direct
male contact than are male-dependent species.
Puberty Modulation as an Artifact
An alternative viewpoint of the phenomena of puberty
modulation is that the effects are some form of laboratory
artifact. Examples of puberty modulation must be evaluated
for the possibility that differences in the timing of
puberty could be the result of close confinement with other
animals, unnaturally high densities, or a product of
artificial selection (see Bronson, 1979; Vandenbergh and
Coppola, 1986).
Bronson (1979) suggested that while the mechanisms
underlying the phenomena in question may be real, it is
possible that the mechanisms evolved to serve one adaptive
purpose in the wild, yet find expression in other ways in
the laboratory environment. Vandenbergh and Coppola (1986)
have argued against the proposal that puberty modulation is
some form of laboratory artifact in house mice. First, they
suggested that the delay of puberty in female mice which

37
results from a specific stimulus such as a urinary cue
suggests an evolved signalling function that must have some
adaptive value. A second reason is that the general social
context in which puberty delay occurs in the laboratory is
known to occur in the field. Descriptions of the social
organization of commensal mice generally fit the conditions
necessary for pubertal suppression to occur. The last line
of evidence comes from the field studies conducted with wild
house mice confined to clover-leaf highway populations
(Massey & Vandenbergh 1980; 1981, see prior section). These
studies show that at least under some conditions, the
production of delay and acceleratory pheromones can occur in
animals from free-ranging environments. Nevertheless,
additional studies and further demonstrations that puberty
modulation occurs among mice and other species such as voles
in the field or under semi-natural conditions seem
warranted.
Summary of Principles of Puberty Modulation in House Mice
and Voles
From the primary findings and related principles
presented for puberty modulation in species such as house
mice and voles, it is clear there is no all-encompassing
theory that leads to clear predictions when puberty should
and should not occur, or to the specific stimuli that should
cause modulation, and the specific function(s) it serves.
However, the emerging picture is that a variety of stimuli
are capable of influencing puberty and multiple functions

38
might be served by puberty modulation among various species.
Vandenbergh and Coppola (1986) suggested that discovering
the role of priming pheromones in the interactive process of
life-history and behavioral adaptations will require the
"melding" of empirical and theoretical points of view.
Problems with Previous Studies of Puberty Modulation
with Voles
Scattered reports indicate that puberty modulation
occurs in a variety of species of Microtusâ–  Some species
have been studied in considerable detail (e.g., prairie
voles, see Carter et al. 1986). However, several sources of
confusion have hampered our ability to explore possible
relationships between puberty modulation and the expressed
differences in the social and mating systems among Microtus.
First, the lack of common procedures and measures used by
various researchers appear to be significant, although
reasonably easy, problems to correct. A variety of measures
and procedures have been used to assess the timing of
puberty, ranging from the measurement of body weight to
analyses of hormone binding sites within the brain. Such
diversity of procedures make meaningful comparisons among
species difficult. For example, Dewsbury (1981) attempted
to evaluate the usefulness of several proposed correlates of
monogamy to predict its presence in several muroid rodents.
However, he deemed it was not possible to compare
meaningfully the existing reports for patterns of sexual
maturation across several species. Thus, with such

39
variation evident for sexual maturation among muroid
rodents, future studies with identical procedures appear
necessary to make useful comparisons among species.
A second problem is the lack of proper controls that
are necessary to identify the effective stimuli influencing
puberty among species. Cues such as the presence or absence
of an adult male or sire, the presence and number of
opposite-sexed siblings, and the presence or absence of a
reproductively active female have all been found to be
effective stimuli that can influence the timing of puberty
in one or more muroid species. Unfortunately, in many
studies such cues are not systematically controlled or
reported. The use of soiled bedding that is transferred
from one cage to another provides one means of critically
testing the assumption that pheromones produced by one or
more individuals cause puberty modulation in others. Such a
procedure controls for the potential influence that
nonolfactory stimuli, such as visual or auditory, could have
upon puberty modulation.
Finally, the database, which is a prerequisite to
determine the general patterns or principles of puberty
modulation, is still relatively small and, in some cases,
misrepresentative of the genus Microtusâ–  For example,
although only a few studies have provided evidence of
puberty modulation in some species of Microtus such as
montane voles (e.g., Sawrey & Dewsbury, 1991), substantial
data exist for other species, such as prairie voles (Carter

40
et al 1986; Getz et al 1983). In addition, most studies on
Microtus. like those for mice (Mus), are female-biased.
Most studies have been designed to determine how olfactory
and social cues affect puberty in females rather than males
of various species. Ideally, the factors influencing the
onset of puberty would be studied concurrently for both
sexes. Such patterns should be viewed for the possibility
of significant interactions occurring between the sexes of a
given species.
Thus, whereas the present list of problems encountered
with studies of puberty modulation in Microtus is not
exhaustive, it outlines significant sources of ambiguity and
deficiencies in prior studies. Fortunately, with standard
procedures and proper controls used with a number of
species, a greater understanding of how puberty modulation
can impact the formation of contrasting social and mating
systems among Microtus might be possible.

CHAPTER 2
GENERAL METHODS: SUBJECTS, HOUSING, AND APPARATUS
Four species of voles (Microbus) were studied in each
of three experiments. Species studied included pine voles
iM-kurotus pinetoruirO . prairie voles ÍM. ochrogaster1 . meadow
voles fM. oennsvlvanicusl, and montane voles (M. montanus).
Breeding colonies of each species were maintained in the
Psychology Building at the University of Florida. These
facilities were accredited by the American Association for
Accreditation of Laboratory Animal Care (A.A.A.L.A.C.).
All subjects were born to breeding pairs that were
laboratory-reared and that had been derived from wild
populations within the United States. Efforts were made to
maintain genetic diversity while guarding against possible
inbreeding in all species. All species were kept in
separate colonies that were housed in windowless and
air-conditioned rooms that were maintained on a reversed
16:8 light-dark photoperiod with light onset at 2000 hr.
All colonies contained animals of both sexes throughout the
experiments on a combination diet of Rabbit Chow (Purina
Mills) and Laboratory Rodent chow #5001 (Purina Mills) with
water available ad libitum. In an effort to promote
continuous breeding and a sufficient supply of subjects, a
handful of lettuce was given to breeding pairs on a weekly
basis. Breeding animals used in Experiment 3 similarly
41

42
received these weekly supplements, although no subjects used
in Experiments 1 and 2 received lettuce supplements during
the duration of their test phase.
All subjects were maintained in large 48 X 27 X 13 cm
polycarbonate cages unless they were housed individually, in
which case they were housed in 29 X 19 X 13 cm polycarbonate
cages. All subjects used in the studies were used in only
one experiment.

CHAPTER 3
EFFECT OF OLFACTORY CUES UPON PUBERTY IN FOUR SPECIES OF
VOLES (EXPERIMENT 1)
Rationale
Experiment 1 was designed to determine if puberty delay
and acceleration occur solely as a function of the presence
of olfactory cues (i.e., pheromones) that are contained in
the soiled bedding in each of four species of voles
Microtus. In several studies where puberty modulation has
been reported, researchers have either housed animals
directly with or across from other animals (e.g., caging
juvenile females with, or across a wire-mesh screen from,
adult males). Although olfactory cues and pheromones are
believed to be key stimuli that produce changes in the
timing of puberty, many procedures have not been designed to
exclude the possibility that other stimuli critically affect
differences in the timing of puberty (e.g., visual,
auditory, or somatosensory cues).
The use of soiled bedding, which is transferred between
donor and recipient animals, has been shown to be one
effective method to expose animals to pheromones while
controlling for exposure to other non-olfactory stimuli that
could influence sexual maturation. Soiled bedding has been
used successfully to reveal changes in reproductive
development among house mice (Drickamer, 1982c), pine voles
43

(Lepri & Vandenbergh, 1986), and California voles (Rissmann
et al. 1984; Rissmann & Johnston, 1985).
Measures and Predictions
44
Specific changes in the physiology of male voles that
would indicate sexual suppression would include reproductive
organs that weighed less than the organs of subjects exposed
to the clean bedding. In addition, the anogenital distance
among males considered suppressed would be smaller than the
anogenital distance in those exposed to clean bedding.
Measures that would reflect sexual suppression in female
voles, would include delayed latency until vaginal
perforation and smaller percentages of cornified cells in
the vaginal smears compared to those exposed to clean
bedding. Among several species of voles, higher proportions
of cornified cells are associated with a higher incidence of
estrus and sexual receptivity (Sawrey & Dewsbury, 1985;
Taylor et al. in press). In contrast to indices of sexual
suppression, indices of sexual advancement or acceleration
would include the opposite changes to those listed above for
sexual suppression.
Because of prior evidence that the highly social female
pine voles and prairie voles typically require direct male
contact for full reproductive activation (e.g., Carter et
al. 1987; Schadler & Butterstein, 1979), I predicted these
species would not be affected, or only minimally, when
exposed to the soiled bedding material. These species may
be considered socially dependent for reproductive activation

(i.e., socially-dependent or "male-dependent" species, see
Taylor, 1990/1991).
45
In contrast, the females of the less social species
(i.e., socially-independent species), meadow voles and
montane voles, have been shown to become sexually receptive
with little or no previous direct exposure to males or to
show marked changes in reproductive physiology through
exposure to male urine alone (Taylor, 1990/1991; Baddaloo &
Clulow, 1981). Field evidence has suggested that both male
and female montane voles are sexually suppressed when they
remain in natal family groups under high densities (Jannett,
1981). Thus, meadow voles and montane voles were predicted
to show marked responses to the presence of the soiled
bedding material. Specifically, they were predicted to be
sexually suppressed when exposed to odors from the natal
family or from adults of the same sex, whereas they would be
sexually accelerated when exposed to odors from unfamiliar
opposite-sexed individuals. Young animals not exposed to
odors from others (e.g., clean bedding) were expected to
show intermediate rates of growth relative to the rates of
those in the other conditions.
Method
Subjects
A total of 514 animals were used in Experiment 1, 256
were male and 258 were female. All were born to existing
laboratory stock maintained at the University of Florida.
The four species included pine voles (Microtus pinetorum),

46
prairie voles CM. ochroqasterl, meadow voles
iM. pennsvlvanicus), and montane voles ÍM. montanus). All
measures were recorded between September 1990 through
November 1991.
Procedure
Animals from each of the four species of Microtus were
weighed and individually caged in 23 X 19 X 13 cm
polycarbonate cages at three weeks of age (day 21-22).
Individuals were assigned randomly (via a random number
table) to one of four conditions: (1) "Family" subjects
received transfers of soiled bedding from their family group
(groups containing an adult male, female, and subsequent
offspring) every other day; (2) "Male" subjects received
soiled bedding from a pooled sample that was derived from
the cages of five unfamiliar adult males (see additional
details below); (3) "Female" subjects received soiled
bedding from a pooled sample of five unfamiliar adult
females; and (4) "Control" subjects received transfers of
clean wood-chip bedding that had been placed in a vole-free
colony room and exposed to air as the bedding within the
cages of animals had been exposed.
Subsequent offspring born to the breeding females that
supplied the bedding in the Family condition were not
reduced in number and were removed from the group at three
weeks of age (21-22 days old). Additional litters born into
the family groups were treated in the same manner. No more

47
than two animals of the same sex and derived from the same
litter were used in the same condition.
All subjects were kept in homospecific colony rooms and
maintained as outlined in the general procedures. Animals
in each of the experimental conditions were housed in small
cages and maintained on separate shelves in each of the
colony rooms in order to minimize the exchange of olfactory
cues among subjects in the various conditions.
Bedding transfers and maintenance. Bedding transfers
occurred on the first day of placement into the experiment
(week 3) and then on every other day throughout the 6-week
test period. Bedding samples were collected and distributed
into recipient cages during the first four hours of light
onset (2000-2400 h). Pooled bedding came from the cages of
five unfamiliar, individually-housed adult animals of the
appropriate sex that did not have their bedding removed for
four days. Bedding samples were mixed and transferred into
the cages of subjects in 200 cc volumes with the use of
plastic cups that were washed with mild detergent after each
use. All donor cages were cleaned weekly. Cleaning
involved the replacement of all soiled bedding with clean
bedding, except for the retention of 800 cc of soiled
bedding within each family cage and 400 cc in each
individually-housed animal to maintain some common olfactory
cues available to the residents. Subjects that received
transferred bedding had 200 cc of bedding removed from their
cages on the day of exchange and had 200 cc of the

48
appropriate type of bedding added to replace the lost
volume.
Pooled bedding was used in the "Male" and "Female"
conditions to provide a more uniform stimulus to animals in
these conditions. It seemed plausible that some of the
bedding-donors would produce soiled bedding that differed in
stimulus guality from others. For example, females may have
excreted different amounts of metabolites in their urine as
a result of fluctuating hormone levels. Previous
researchers have used similar pooling methods in behavioral
preference tests with house mice (e.g., Coppola & O'Connell,
1988; Drickamer 1989b).
Physiological and morphological measures. Body weights
were recorded for subjects when they were weaned (day 21-22)
and at weekly intervals until they were 56 days of age (9
weeks). By this time, individuals from each species would
be considered to have reached adult status under standard
laboratory conditions, with the possible exception of pine
voles. Lepri and Vandenbergh (1986) found that the median
age for male pine voles to sire a litter was 57 days (North
Carolina population), whereas the median age for first
conception among the females was 50 days.
Other measures of puberty included the day of vaginal
opening for females and the anogenital distance of males at
weekly intervals. Vaginal smears were taken daily from each
female beginning on the day they first became perforate.
All other subjects, including males, were handled daily in a

49
similar fashion as were females, as a control procedure for
the effects of handling. During all handling procedures,
subjects were held with individually-assigned vinyl gloves
in order to prevent the transference of odors among
subj ects.
At the end of the 9-week period, subjects were
euthanized and selected organs were removed and weighed to
the nearest 0.1 mg. Organs weighed included the uterus,
ovaries, and adrenal glands of females and the seminal
vesicles, testes, and adrenal glands of males. All
dissections and weighing of organs were conducted by the
author who used uniform procedures. In all cases, efforts
were made by the author to remain blind to the condition of
the subjects prior to the dissection procedure. This was
accomplished by dissecting animals in small groups if
possible. During this procedure, the identification cards
of subjects were placed upside-down and beneath the
individual trays that held the removed organs from each
subject. The organs were covered with moist paper towels in
a uniform fashion. The author would then scramble the
position of the dissection trays just prior to the weighing
procedure. Thus, in most cases, the author was blind to the
experimental condition of the animals during the final
cleaning and weighing procedure.
Statistical Analysis
Analysis of variance procedures (ANOVA's) were used to
assess the effects of treatment on subjects. All data were

50
transcribed to computer spreadsheets and analyzed with the
CSS: Statistica software program (StatSoft, Inc). Data were
analyzed independently for each species and sex, because of
occasional instances of heterogeneity of variances between
the species. Analyses consisted of either one- or two-way
ANOVA's with the experimental condition comprising the
primary between-subjects factor. Some measures, such as
body weight, were recorded over the course of the study and
were analyzed with the additional repeated-measure factor of
weeks. Paired organ weights were analyzed with the
repeated-measure factor of body location (left versus
right), because of known physical asymmetries among some
paired organs (e.g., Pinter, 1968) or other possible
functional asymmetries (e.g., Clark & Galef, 1990). The
alpha level was held at .05 in all comparisons, and all
comparisons were based on a two-tailed probability.
Results
Results of each of the measures are reported below for
the effect of the experimental conditions separately for
each species. The statistical values of ANOVA's are
reported directly in the text or in specified tables,
although the exact probability values of post-hoc
comparisons (Neuman-Keuls tests) are not reported in order
to streamline the text. Only post-hoc comparisons that were
significantly different (p's < .05) are discussed in the
text, or are clearly specified as not being significant when
discussed.

51
Body Weight
Generally, the experimental condition (bedding type)
did not substantially influence body weight among the
species. However, male pine voles were significantly
affected by the condition (see below).
Because only the factor of week was statistically
significant among the other species, simply reflecting
increases in body weight across age, analyses for each
species are located in Table 3-1 (see Appendix A for means
at each week).
Pine voles: males. Post-hoc analyses revealed that
male pine voles in the Family condition (22.71 + .53,
N = 16) and Male condition (22.96 + 1.05, N = 14) weighed
significantly more than those reared in either the Control
condition (21.24 + .53, N = 17) or Female condition
(20.64 + .65, N = 16) at week 9 (interaction of condition
and week, F(18, 348) = 1.98, p = .010)(see Figure 3-1). The
main effect of week was significant, F(6, 348) = 600.87,
p = < .001), indicating larger body weights across weeks,
although the main effect of condition was not significant,
F (3 , 58) = 1.68, p = .181 (see Appendix A for means at each
week).
Anogenital Distance
The anogenital distances of the male prairie voles,
meadow voles, and pine voles were not substantially
influenced by the condition. However, there was some

52
indication that anogenital distances among male montane
voles were influenced by the condition (see below).
Because only the factor of week was statistically
significant in the other species, simply reflecting
increases in anogenital distance with increasing age,
analyses for all other species are located in Table 3-2 (see
Appendix B for means at each week).
Montane voles. Although a significant interaction of
anogenital distance by week of measurement was found,
F(18, 360) = 1.84, p = .019, post-hoc comparisons did not
reveal significant differences among any of the groups (see
Table 3-2 for means at week 9 and ANOVA results; complete
means for each species and week are located in Appendix B).
The mean anogenital distance of montane voles in the Control
condition (M = 13.44 + .44 mm, N = 16) approached being
significantly larger than compared to that in the Family
condition (M = 12.25 + .31 mm, N = 16 (p = .09). The mean
anogenital distances in the other conditions were
intermediate in value to those in the Control and Family
conditions. The main effect of week was significant,
F(6, 360) = 173.00, p < .001), although the main effect of
condition was not, F(3, 60) = 0.33, p = .798).
Adrenal Weight
Adrenal weights were minimally affected by the
condition among all species and sexes, with the exclusion of
male pine voles (see below). The mean adrenal weights and

53
analyses for all species, when uncorrected and corrected for
differences in body weight, are located in Table 3-3.
Pine voles: males. The condition did not significantly
influence the adrenal weights of male pine voles, when
uncorrected for differences in body weight, F(3, 59) = 1.26,
p = .296 (see Table 3-3 for means and ANOVA results).
However, the analysis of adrenal weights, corrected for
differences in body weight, revealed significant differences
among the conditions (see Figure 3-2). Post-hoc comparisons
revealed that the adrenals of males in the Female condition
(M = 24.78 + 1.17, N = 16) were significantly heavier than
in the Control condition (M = 21.60 + 1.11, N = 17) or
Family condition (M = 20.06 + 1.35; N = 16)(main effect of
condition, F(3, 59) = 3.32, p = .026). The adrenals from
males in the Male condition were similar in weight to those
from the Control and Family conditions but failed to differ
significantly from those in the Female condition (M = 21.02
± 1.09, N = 14).
The adrenals were heavier on the left side of the body
than on the right, when corrected for body weight (left
adrenal: M = 22.49 + .64; right adrenal M = 21.28 + .62),
F(1, 59) = 7.58, p = .008. The interaction of condition and
body location was not statistically significant, F(3, 59) =
0.14, p = .929.
Among all species and sexes, the left adrenals were
significantly heavier than those on the right side of the
body (See Table 3-3 for complete means and analyses).

54
Testes Weight
Testes Weights were not significantly affected by the
condition within any of the species, whether they were or
were not corrected for differences in body weight. Complete
means and analyses are located in Table 3-4. The only
significant effects were attributed to small asymmetries in
the left and right testis weights within meadow voles and
montane voles (see means and effect of position in Table
3-4) .
Seminal Vesicle Weight
The experimental conditions did not significantly
influence the seminal vesicle weight within any species,
whether they were or were not corrected for differences in
body weight (see complete means and analyses in Table 3-5.
Ovarian Weight
The experimental conditions did not significantly
influence the ovarian weight of any species, whether they
were or were not corrected for differences in body weight
(see means and analyses in Table 3-6).
Uterine Weight
The experimental condition did not significantly
influence the uterine weights among the pine voles, prairie
voles, or meadow voles (see means and analyses for all
species in Table 3-7). However, significant differences
were evident in the uterine weights of the montane voles as
a result of the condition.

55
Montane voles. The condition significantly influenced
the uterine weights of montane voles (see Figure 3-3). When
uterine weights were compared, without correcting for
differences in body weight, the uteri of females in the
Control condition (M = 27.46 + 1.92) weighed significantly
more than the uteri found in the Family condition (M = 18.56
+ 1.92) and in the Female condition (M = 16.52 + 2.22) (main
effect of condition, F(3, 60) = 4.55, p = .006). The mean
uterine weight among those in the Male condition was
intermediate to those in the other conditions (M = 24.35
+ 2.97) and was not significantly different from them. This
result is the first example, within a controlled
environment, of sexual suppression in female montane voles
that are exposed to the bedding from a family group or from
adult females.
When the analysis of uterine weight was corrected for
differences in body weight among montane voles, the results
were similar to those of the unadjusted analysis (main
effect of condition, F(3, 60) = 4.61, p = .006 (see Figure
3-3). The adjusted uterine weights were significantly
greater in the Control condition (M = 97.71 + 10.65) than in
the Family condition (M = 61.02 + 4.42) or Female condition
(M = 58.92 + 5.74). However, in addition, the females of
the Male condition (M = 91.21 + 14.45) had significantly
heavier uteri than those in the Family condition.

56
Status of Vaginal Perforation
The total numbers and percentages of females that
became vaginally perforate during the course of the study
are summarized in Table 3-8 for all species. Clear
differences were evident among the species in their modal
pattern of vaginal perforation. Pine voles displayed the
most atypical pattern of vaginal perforation with respect to
the other species. None of the 66 pine voles were perforate
at the beginning of the study (week 3) and only 2 of the 66
(3.0%) became perforate during the 9-week study. This
slight shift in the frequency of pine voles that became
perforate was not significant between weeks 3 and 9 (McNemar
chi-square: X2 = .50, p < .479). Nearly all of the females
of the other species became perforate at some time during
the study (see Table 3-8).
The percentages of females that were perforate at the
beginning of the test were different among the four species,
X2(3, N = 258) = 70.08, p < .001. Post-hoc comparisons
(chi-square) revealed that all species differed
significantly from pine voles in the proportions of females
that were perforate on week 3. However, significantly fewer
prairie voles were perforate at week 3 than were meadow
voles, (X2(1, N = 167) = 18.36, p < .001, and montane voles,
X2(l, N = 161) = 12.18, p < .001. Meadow voles and montane
voles did not differ significantly on this measure,
X2(3, N = 186) = .91, p < .341.

57
Delay until Vaginal Perforation
Statistical analyses of the delays of vaginal
perforation were problematic, because properly they would be
limited to those females that were not perforate on the
first day of the study (week 3). However, the majority of
all meadow voles (34 of 62 or 55%) and many montane voles
(26 of 64 or 41%) were perforate on the initial day of the
study and would be excluded from the analysis. In contrast,
only 5 of 66 (8%) prairie voles and none of the 66 pine
voles were perforate on the first day of the study. The
total number of subjects and percentages of subjects for
which vaginal smears were obtained are summarized in Tables
3-9 and 3-10 for each species and by condition.
The species differed appreciably in the time at which
they became vaginally perforate. Figure 3-4 displays the
cumulative percentages of females that became perforate
throughout the study as a function of species. Nearly 80%
of the meadow voles became perforate within the first 4
weeks of age, while prairie voles reached the 80% mark
nearly a week later (Day 33). Over 50% of the montane voles
were perforate by day 23, but the cumulative percentage of
all montane voles did not reach the 80% criterion until day
41. Only two pine voles became perforate, the first at 44
days of age and the second on last day of the study (Day
63) .
The mean ages (in days) that vaginal perforation
occurred for subjects in each species and condition, that

58
were imperforate on the first day of the study, are shown in
Figure 3-5. Sample sizes were not uniform across all
species and conditions because of the different patterns of
typical vaginal opening. Total numbers of subjects ranged
from 14-16 for prairie voles, 5-6 for meadow voles, and 8-10
for montane voles among the conditions.
The condition was not found to significantly influence
perforation latency among the prairie voles, (Kruskal-Wallis
ANOVA, prairie voles: H (3, N = 59) = 4.34, p = .23) or
montane voles: H (3, N = 36) = 5.74, p = .12). Meadow voles
were not analyzed statistically because of the small number
of subjects amenable to this analysis (N = 22).
Vaginal Smears
Three types of cells were classified and counted from
the vaginal smears, following the method of Taylor
(1990/1991). The cell types analyzed included cornified
cells, nucleated cells, and leukocytes. Frequencies of each
type were converted to percentages of the total number of
cells in each smear, because of the variability in the total
number of cells obtained per smear. Similar conversions of
cell frequencies have been used previously in studies with
laboratory rats (e.g., McClintock, 1983, 1984) and with
voles (e.g., Sawrey, 1989/1990; Shapiro & Dewsbury, 1990).
Data from female pine voles were excluded from the analysis
because only two became perforate during the study.
All species produced vaginal smears that were typically
dominated by either cornified cells or leukocytes, with

59
appreciably fewer nucleated cells. These characteristics of
smears were typical of those found previously for several
species of voles (e.g., Sawrey & Dewsbury, 1985; Taylor,
1990/1991). Nucleated cells were relatively few in number
and relatively constant in proportion throughout the
duration of the study. In contrast, cornified cells and
leukocytes were typically in reverse proportion to one
another when changes in the percentages of cells were
detected.
Prior to conducting statistical analyses, all cell
percentages were pooled for each subject by two-day block
intervals. This procedure generally reduced daily
variability in the cell percentages across successive days.
Prior to pooling, graphs were made of the data as a function
of each day; no indications of cyclical fluctuations were
evident.
Effect of condition: within-species comparisons of cell
types. Two types of analysis were conducted for
within-species comparisons of cell types. The first
analysis consisted of comparing the proportions of each cell
type as a function of the experimental conditions for each
two-day block (Kruskal-Wallis ANOVA by Ranks tests). Thus,
a total of 21 two-day blocks (representing Days 21-22
through 61-62) were analyzed sequentially to determine if
the condition influenced the percentages of cell types. In
the case of a significant result, post-hoc comparisons were

60
made to identify which conditions differed significantly
(Mann-Whitney U Tests).
In the second analysis, repeated-measures analyses were
used to assess whether significant changes in the
percentages of each cell type occurred as a function of age.
These analyses were done separately for each condition and
species. Cell percentages were analyzed on alternate blocks
of days representing the ages of 33-34 days through 61-62
days (Friedman ANOVA by ranks tests). This reduced sample
of days was necessary due to fewer smears available during
the youngest ages and thus the necessary exclusion of data
from repeated-measures analysis (see Table 3-9). Thus, a
series of eight means was compared in each repeated-measures
analysis. Post-hoc comparisons (Wilcoxon Matched Pairs
Tests) were used to identify where significant changes
occurred across the blocks of days when the Friedman ANOVA
was found to be statistically significant (p's < .05).
Prairie voles: cornified cells: effect of condition.
The percentages of cornified cells from prairie voles are
shown as a function of experimental condition in Figure 3-6
(each mean plotted represents the pooled data of individuals
by two-day block intervals). It was evident that relatively
little variation occurred in the cell percentages as a
function of the condition among prairie voles. A
statistically significant change was identified for only one
of the two-day blocks (Day 27-28: H(3, N = 35) = 8.90,
p = .031). During this block, subjects in the Female

61
condition had significantly more cornified cells than those
in the Male condition (M's = 27.5% versus 17.9%; total
numbers of subjects can be determined from Table 3-9).
Prairie voles: effect of age. Relatively small changes
in the percentages of cornified cells occurred as a function
of age. Repeated-measures analyses failed to detect any
significant change in the cell percentages for subjects in
any of the four conditions.
Additional comparisons of the percentages of nucleated
cells and leukocytes have been placed in Appendix C. In
general, the proportions of nucleated cells varied little as
a function of the condition or age for all species. Thus,
the proportions of leukocytes often varied inversely to the
proportions of cornified cells and are somewhat redundant
for purposes of analysis.
Meadow voles: cornified cells: effect of condition.
The mean percentages of cornified cells as a function of the
experimental condition are shown in Figure 3-7. Despite
somewhat greater differences among the cell means than those
found in prairie voles, and relatively large sample sizes
(range of total N's = 47-56), no significant differences
were found within any of the blocks of days as a function of
condition. The lack of significant differences also held
true for all comparisons of nucleated cells and leukocytes
(Appendix C).
Meadow voles: effect of age. Repeated measures
analyses revealed significant changes in the percentages of

62
cornified cells within all four conditions of meadow voles.
There were general increases in the percentages of cornified
cells for subjects in all conditions, although the relative
percentages of cornified cells for the Control condition
were typically the smallest among those from all conditions
(see Figure 3-7; Control condition: X2(7, N = 11) = 31.39,
p < .001). Significant differences occurred among the
following blocks of days for subjects in the Control
condition (Note: differences indicated below are in the
direction of later days being significantly higher in cell
* A _
percentages than earlier days except where noted " "): Days
33-34 versus Days 49-50 through Days 61-62; Days 37-38
versus Days 41-52, 49-50 through Days 61-62; Days 45-46
versus Days 53-54 through Days 61-62; Days 49-50 versus Days
57-58 through Days 61-62.
Significant changes in the percentages of cornified
cells for subjects in the Family condition were evident,
X2(7, N = 11) = 29.85, p < .001), and were as follows: Days
33-34 versus Days 45-46 through Days 57-58; Days 37-38
versus Days 45-46 through Days 61-62; Days 41-42 versus Days
45-46 through Days 61-62.
Significant changes in the percentages of cornified
cells for subjects in the Male condition were present,
X2(7, N = 12) = 36.97, p < .001), were as follows: Days
33-34 versus Days 41-42, Days 49-50 through Days 61-62; Days
37-38 versus Days 41-42, Days 49-50 through Days 61-62; Days

41-42 versus Days 53-54 through Days 61-62; Days 45-46
versus Days 49-50 through Days 61-62.
63
Significant changes in the percentages of cornified
cells for subjects in the Female condition were present,
X2(7, N = 13) = 52.08, p < .001), were as follows: Days
33-34 versus Days 41-42 through Days 61-62; Days 37-38
versus Days 45-46 through Days 61-62; Days 41-42 versus Days
45-46 through Days 61-62; Days 45-46 versus Days 53-54
through Days 61-62; Days 49-50 versus 57-58.
Montane voles: cornified cells. The mean percentages
of cornified cells among montane voles in the various
conditions varied more as a function of condition than the
percentages observed in prairie voles and meadow voles
(compare Figures 3-8 for montane voles to Figures 3-6, and
3-7). Typically, there were gradual increases in the
percentages of cornified cells throughout the study for
montane voles.
Montane voles: effect of condition. Statistical
differences were found among subjects in the various
conditions on blocks of days including Days 53-54, 57-58,
and 61-62. Among the first and last of these three blocks,
the subjects in the Male and Control conditions had
significantly greater mean percentages of cornified cells
than those in the Female and Family conditions. The
intermediate block (Days 57-58) also differed in the same
manner, except the difference between the Male and Family

condition failed to reach a significant level (see Figure
3-8 for clarification).
64
Montane voles: effect of age. Repeated measures
analysis revealed significant shifts in the percentages of
cornified cells within each of the four conditions. Montane
voles in the Control condition had significant shifts in the
proportions of cornified cells between the following blocks
of days: Days 33-34 versus 41-42, 49-50 through 61-62; Days
37-38 versus 57-58 through 61-62; Days 41-42 versus 53-54
through 61-62; Days 45-46 versus 49-50 through 61-62; Days
49-50 versus 57-58 through 61-62.
Significant shifts in the percentages of cornified
cells were found for those in the Family condition among the
following blocks of days: Days 33-34 versus 49-50, 57-58
through 61-62; Days 37-38 versus 57-58 through 61-62; Days
41-42 versus 49-50 through 61-62; Days 45-46 versus 61-62;
Days 49-50 versus 61-62; Days 53-54 versus 57-58.
Changes among blocks of days for subjects of the Male
condition were statistically significant among the following
blocks of days: Days 33-34 versus 53-54 through 61-62; Days
37-38 versus 53-54 through 61-62; Days 41-42 versus 49-50
through 61-62; Days 45-46 versus 53-54 through 61-62; and
Days 49-50 versus 53-54, and 61-62.
Finally, significant shifts in the percentages of
cornified cells were identified for subjects in the Female
condition among the following blocks of days: Days 33-34
versus 41-42 through 49-50, and 57-58 through 61-62; Days

65
37-38 versus 49-50 through 61-62; Days 45-46 versus 61-62;
Days 49-50 versus 61-62; Days 53-54 versus 61-62; and Days
57-58 versus 61-62.
Discussion
Prairie voles and pine voles were expected to show
fundamentally different patterns of response from those of
the meadow voles and montane voles. I predicted that the
less social meadow voles and montane voles would show
pronounced sexual advancement when exposed to the bedding of
unfamiliar opposite-sexed individuals, and show signs of
sexual suppression when exposed to same-sex odors. Animals
exposed to clean bedding (Control condition) were expected
to have intermediate indices of sexual development compared
to those in the other conditions. In contrast, I believed
there would be little or no differences in sexual
development within the prairie voles or pine voles.
The experimental results of the exposure to the soiled
bedding on general and sexual development are discussed
below for each of the measures.
Body Weight
Male pine voles were the only group that differed
significantly in body weight as a function of the condition.
Males in the Family and Male conditions weighed more than
those in the Control and Female conditions (Figure 3-1).
These results, if repeatable, are contrary to predictions.
The results suggest that the olfactory cues associated with
adult males or family groups cause general increases in the

66
rate of growth of males exposed to them, when compared to
patterns of growth of males exposed to clean bedding or to
odors from adult females. Unfortunately, it is not possible
to compare these results with those of other studies, as
pine voles have not been investigated in any similar manner.
However, general characteristics of natural populations
of pine voles suggest that male pine voles are not sexually
suppressed when living within family groups and often are in
the presence of at least one other scrotal male (FitzGerald
& Madison, 1983). Thus, limited information on the
free-ranging behavior of pine voles is at least compatible
with the notion that the exposure of males to the odors from
family groups and/or other males in some manner contributes
to increased body weight. One possible mechanism by which
males might experience increased body weight is simply
through increased food intake. Such a mechanism has been
identified that led to increased weight gain in male musk
shrews (Suncus murinus) when exposed to females (Wayne &
Rissman, 1990).
It is not known why pine voles were the only species to
experience differences in body weight as a function of the
condition. Other studies have shown that the body weights
of other species of Microtus can be affected by exposure to
pheromonal cues. Female meadow voles have been shown to
have significantly greater increases in body weight,
compared to controls, when the urine of mature males
trickled directly into the cages of females via small tubes

67
(Baddaloo & Clulow, 1981). Other research has provided
evidence that male and female prairie voles could differ in
body weight, by nearly 10 g respectively, as a function of
whether they had separate air supplies or shared a common
air supply within the colony (Batzli et al. 1977).
Together, the results of previous studies suggest that
differences in body weight might have been detected within
some of the other species, had the procedures been
different. The use of a more continuous supply of urine and
pheromones or a separate supply of air for each subject may
have led to detectable differences as a function of the
condition. However, given the generally small differences
in body weights among the male pine voles and number of
comparisons made, one must also consider the results were
due to chance alone and not due to pheromonal differences.
Additional study seems warranted to clarify whether these
patterns of weight gain among male pine voles can be
replicated, what mechanism accounts for differences in body
weight, and what functional differences they may be
producing.
Anogenital Distance
The anogenital distances among montane voles might have
differed as a function of the conditions, but the results
are ambiguous. Whereas the overall F statistic was
statistically significant, post-hoc comparisons failed to
detect significant differences among the groups. Males in
the Control condition had a larger mean anogenital distance

68
than those in the Family condition, although it was not
significantly different (g = .09; Neuman-Keuls test). This
difference is in the hypothesized direction, but the
absolute difference between the groups limits making any
firm conclusion.
In an attempt to determine statistically whether the
differences in anogenital distances among montane voles were
systematic effects caused by the conditions, an analysis of
covariance was conducted, using body weight at week 3 as a
covariate. Results were similar to those of the primary
analysis, i.e., the interaction of anogenital distance and
week of measurement was statistically significant, F(18,
360) = 1.84, p = 0.019, although no post-hoc comparisons
were significantly different. Thus, the results of the
anogenital distance comparisons are only suggestive of
differences produced by exposure to the different bedding.
Unfortunately, no similar studies have been conducted with
montane voles, or other species of Microtus. in order to
make any comparisons between studies.
It is not known what functional difference, if any, a
larger anogenital distance would reflect. Presumably, the
measure reflects, in part, the relative size of the testes
which lie proximally between the penis and the anus. In the
present study, the correlation between testes weight and
anogenital distance was relatively small among all montane
voles (N = 64; rs = .31), although it was slightly higher in
the Control males alone (N = 16; rs =
.37). Thus,

69
differences in testis weight do not lend much support for
the possibility that the differences in anogenital distance
largely reflect differences in testis weight. The analysis
of the testes weights of the montane voles did not reveal
any significant influence of the conditions (see below).
Additional study would seem necessary to establish whether
the differences in anogenital distance in montane voles are
repeatable outcomes of exposure to these substrates.
Adrenal Weight
The adrenals differed systematically in weight as a
function of the treatment among male pine voles. The
adrenals, when corrected for differences in body weight,
from males in the Female condition were significantly
heavier than those in the Control and Family conditions.
Because males of the Female condition were significantly
lighter in body weight than those in the Family and Male
conditions, it is possible there is a causal relationship
between increased adrenal activity and decreased body
weight. Prior studies with male meadow voles have shown
that subcutaneous injections of ACTH produced significant
reductions in body weight and concurrent increases in
adrenal weight than among control males (Pasley & Christian,
1971).
It is not known what differences there are in bedding
soiled from male and female pine voles. Recent chemical
analysis of volatile compounds in male and female pine vole
urine did not reveal any qualitative difference in the

70
urinary profiles between them (Boyer et al. 1989). However,
female urine contained three volatiles in higher
concentrations than in male urine, whereas male urine
contained one compound present in higher concentration than
in female urine. Boyer et al. (1989) suggested there may be
differences in the nonvolatile urinary fraction that might
be investigated. Among house mice, the acceleratory
pheromone produced by male house mice appears to be mediated
by a nonvolatile urinary fraction contained in the urine
(Vandenbergh et al. 1975; 1976).
Among all species, the adrenals from the left side of
the body were significantly heavier than those located on
the right. This difference in asymmetry may be common to
all species of Microtus. Pinter (1968) reported that the
adrenals in montane voles were larger on the left than on
the right side of the body, regardless of the sex, age, or
regime of diet or photoperiod. A similar asymmetry in
adrenal weight had been reported for California voles
(M. californicus)(Mullen, 1960). The specific functional
difference of such asymmetry, if any, is not known.
Testes Weight
Comparisons of the testes weights did not indicate any
effect of the experimental condition among any of the
species. It is possible that changes in testicular weight
among male Microtus are relatively unaffected by exposure to
different pheromones, although exposure to stressful
environments may produce substantial changes. Lecyk (1967)

71
provided evidence that male common voles fM. arvalisl, that
were housed adjacent to cages with crowded and sexually
active voles, had lighter testes than control males. Other
indirect evidence that testicular weight can change as a
function of stress was provided by Pasley and Christian
(1971). Male meadow voles, injected with ACTH across a
series of days, had significant decreases in testes weights
when compared to noninjected controls.
It is possible that testis weight is a relatively
insensitive measure of male reproductive activity, at least
when investigating the influence of pheromones or other
olfactory stimuli on reproductive maturation. More precise
measures of reproductive activity might reveal systematic
differences among voles exposed to different pheromones.
Measures such as the level of circulating androgen or
characterization and counts of spermatozoa may be
appropriate. Research with California voles has shown that
titers of androgens were found to be significantly lower in
males exposed to bedding from their mothers than in those
exposed to bedding from unrelated males; testes weights were
not found to differ among these groups (Rissman et al.
1984) .
Seminal Vesicle Weight
The experimental condition did not have a noticeable
effect on the seminal vesicle weight among any of the
species. Why no significant differences were found is
puzzling. Prior studies with California voles have shown

72
that differences in seminal vesicle weight can be found when
males are exposed to different types of soiled bedding. The
seminal vesicles of California voles that were exposed to
bedding soiled by their own mother, their family group
(including the sire), or an unrelated mother were
significantly smaller than those of males reared in clean
bedding (Rissman et al. 1984; Rissman & Johnston, 1985). In
contrast, the presentation of soiled bedding from unfamiliar
adult males, or bedding from the family that had been
supplemented with bedding from the separated father,
produced seminal vesicles of similar weight in both groups.
Thus, the data from California voles suggest that the
development of the seminal vesicles can be delayed in young
males when they are exposed to odors from their mothers or
from unrelated mothers, but not delayed by exposure to odors
from unrelated adult males or from their fathers (Rissman et
al. 1984).
Ovarian Weight
The experimental condition did not have a noticeable
effect on the ovarian weight in any species, whether they
were corrected or uncorrected for differences in body
weight. However, previous studies suggest that the ovarian
weight among some species of Microtus can be affected by
exposure to pheromonal stimuli. Baddaloo and Clulow (1981)
found that female meadow voles that were exposed to male
urine, that had been designed to drip into the adjoining

73
(empty) half of divided cages, had significantly heavier
ovaries than compared to those of control females.
Female pine voles have also been shown to have
increases in ovarian development when exposed to pheromonal
stimuli. Lepri and Vandenbergh (1986) placed individual
females directly below the cages of animals of different
stimulus qualities. This procedure enabled feces and urine
from the animal(s) in the top cages to fall into the cages
housing the females. Results indicated that the ovaries of
the females exposed to the excreta from either individually
housed intact males or castrated males were significantly
heavier those of females exposed to an empty cage or to the
excreta from singly-housed females or group-housed females.
Uterine Weight
By at least one standard, the uterus is considered to
be the best and most reliable bioassay for circulating
estrogens in female house mice (Bronson & Stetson, 1973).
Female montane voles in the Control condition were found to
have significantly heavier uteri than those in the Family
and Female conditions (Figure 3-3). In addition, the uteri
from those in the Control and Male conditions were also
significantly heavier than those in the Family and Female
conditions when the uteri were corrected for differences in
body weight. These results are the first to demonstrate
sexual suppression of female montane voles when they are
exposed to adult female bedding or family bedding under
controlled conditions. Previous field evidence had only

74
suggested such a phenomenon (Jannett, 1978). The results
are in general agreement with predictions and complement the
recent finding of sexual acceleration in female montane
voles when they were exposed to male stimuli (Sawrey &
Dewsbury, 1991). However, if the responses of uterine
weight to the pheromones conform to the distinction of
socially-dependent versus independent dichotomy, it is not
clear why meadow voles also did not show differences in
uterine weight when they were exposed to the different
bedding types.
It is possible that many species of Microbus, including
meadow voles, may require both direct exposure to males and
their pheromones for complete reproductive activation. This
type of synergy has been demonstrated in house mice
(Drickamer, 1974; Bronson & Maruniak, 1975), montane voles
(Sawrey & Dewsbury, 1991), prairie voles (Carter et al.
1987), and pine voles (Lepri & Vandenbergh, 1986), although
the specific causes for it are not yet understood. For
example, among female prairie voles, direct contact with
male urine or housing them in male-soiled cages results in
increased uterine weights, but typically does not elicit
behavioral estrus (Carter et al. 1987). Thus, it is
possible that female meadow voles may require more direct
male stimulation in comparison to montane voles. Evidence
found by Taylor (1990/1991) supports this hypothesis, as
nulliparous female meadow voles showed extremely low
copulation rates after being kept behind a wire-mesh barrier

for up to seven days with only 1 h of daily direct
interactions with males.
75
Characteristics of Vaginal Perforation
Although none of the species differed significantly in
the latency to become vaginally perforate as a function of
exposure to soiled bedding, clear species differences were
evident in the latencies of the species to become perforate
(Figure 3-4). Meadow voles and montane voles became
perforate either before or shortly after weaning (Day 21).
In contrast, prairie voles were somewhat delayed in becoming
perforate. Pine voles were even more extreme, with only two
of the animals becoming perforate during the course of the
study.
The few pine voles that become vaginally perforate
during the course of the study do not seem uncharacteristic
from personal observations and from other reports (e.g.,
Schadler & Butterstein, 1979). It appears that most female
pine voles, whether housed in isolation or with siblings, do
not become perforate until they are placed with an
unfamiliar male or reach a considerably advanced age. In
pilot work for the study presented, 10 adult female pine
voles were examined for vaginal perforation. Whether pine
voles were housed singly, or with other siblings, none was
perforate despite ranging in age between 90 to 120 days.
Earlier work also supports the finding of substantial
delays in reproductive activity among pine voles. Schadler
and Butterstein (1979) found that among female pine voles

76
that had been paired for breeding with fertility tested
males, the mean age of first conception for the females was
105 days. Thus, it appears that direct and relatively long
exposure to males may be a normal prerequisite for vaginal
perforation and reproductive activation in pine voles, even
if they have reached a considerably advanced age.
Substantial delays among female prairie voles to become
vaginally perforate have also been documented. Richmond &
Conaway (1969b) found that 98% of more than 200 individually
housed females, or those in compatible female groups,
remained in a state of persistent anestrus with imperforate
vaginas between 3 to 5 weeks of age.
Within-species analysis of the perforation latencies
among prairie voles did not suggest they are strongly
affected by the olfactory cues. The overall comparison test
was not statistically significant, despite there being a
relatively large number of imperforate female prairie voles
at the beginning of the study (N's = 14-16 per condition).
Female prairie voles in the Male condition became perforate,
on average, earlier than those in the Control condition (M's
= 33.0 versus 27.5 days)(Figure 3-5).
One of the difficulties with analyzing the conditions'
effect upon perforation latencies among meadow voles and
montane voles was that many females were vaginally perforate
at the beginning of the test and thus excluded from the
formal analysis. Employing an earlier day of weaning did
not seem plausible for all species, although it was

77
attempted in pilot work. Weaning pine voles at an earlier
age than 21 days often resulted in their deaths within a day
or two following separation. In retrospect, it would seem
possible to wean each species at different ages that
corresponded more closely to their species-typical age for
weaning. Indices such as the cessation of nursing or the
timing of the first day to ingest solid food is typically
earlier for meadow voles and montane voles than they are for
pine voles and prairie voles (McGuire & Novak, 1984; 1986).
By weaning each species at an age that appears most
appropriate for each species, it would be possible to test
more readily how stimuli affected certain reproductive
responses, at least within each species.
Vaginal Cytology
It was expected that the less social species would show
higher proportions of cornified cells when the females were
exposed to male bedding than would females exposed to
bedding in the other conditions. Those in the Family and
Female conditions were predicted to have fewer cornified
cells than those in the Control condition. The patterns of
cornified cells for all species that became vaginally
perforate are shown in Figure 3-6 through Figure 3-8.
The percentages of cornified cells for prairie voles
among the various conditions were similar, and statistically
only one of the two-day blocks revealed a significant
difference among conditions. It seems plausible that this
one significant difference was spurious, because neither the

78
few preceding nor the few following blocks of days was
statistically significant. In addition, the total number of
multiple-comparisons tests was substantially high. This in
turn would inflate the likelihood of detecting at least one
statistically significant difference among the days
compared. As predicted, female prairie voles revealed
little response to the different bedding types.
Although variation in the cell percentages for meadow
voles suggested an influence of condition, none were
detected statistically (Figure 3-7). Inspection of the
percentages of cornified cells suggests a trend of more
cornified cells in females exposed to adult female odors
than those receiving no odors (Control). A priori. the
direction of this relationship was not expected but suggests
further study because it is possible that meadow voles
experience sexual acceleration when exposed to odors or
pheromones of other adult females. Among house mice, urine
from pregnant and lactating females accelerates sexual
development in female mice (Drickamer & Hoover, 1979), as
does the urine from female mice that are in estrus
(Drickamer 1982c). It seems plausible that among meadow
voles, pheromones contained in the urine of other meadow
voles that are of either sex may cause sexual acceleration
when compared to those that are not exposed to any
pheromonal source, at least under certain conditions.
Results from the analysis of vaginal smears of meadow voles

by Baddaloo and Clulow (1981) are in general agreement with
this prediction.
79
There is reasonable evidence that the conditions
affected the proportions of cornified cells of the montane
voles, at least during the last 10 days of the study. The
statistical results revealed systematic differences in the
proportions of cells as a function of condition during the
later portion of the study (Figure 3-8). Specifically,
during three of the last five two-day blocks of smears, the
females in the Male and Control conditions had substantially
more cornified cells than those in the Family and Female
conditions. Although there is the possibility that these
relatively few significant results were due to chance,
because of the relatively large number of comparison tests
(21 for each species and condition representing Days 21
through 62), the distribution of the significant effects
suggests a non-random pattern.
The differences among the uterine weights of the
montane voles add additional support for the hypothesis that
the percentages of cornified cells of montane voles reflect
systematic differences as a function of the condition. The
corrected uterine weights were significantly heavier from
females in the Male and Control conditions compared to those
from the Family and Female conditions. Thus, the
percentages of cornified cells appear to reflect these
differences; larger uteri were associated with higher
percentages of cornified cells among the montane voles.

Given the present results, the montane voles appeared
to be the species that was most affected by the exposure to
80
the pheromones. The results of the current study and of
earlier studies suggest that, even with similar levels of
male exposure, prairie voles and pine voles take appreciably
longer to become sexually receptive than montane voles and
meadow voles, and typically require direct exposure to males
for complete reproductive activation to occur (Carter et al.
1987; Lepri & Vandenbergh, 1986). Thus, results of the
vaginal smear data provide limited support for the
socially-dependent and socially-independent dichotomy. The
less social species, meadow voles and montane voles, appear
to be more sensitive to conspecific pheromonal cues than are
the more social species. Additional study of these species,
using common procedures, may be useful to determine if these
results are valid, and to shed light on possible functional
differences between the species.
Conclusions (Experiment 1)
A major theme that broadly characterizes the results of
Experiment 1 was the absence of an effect of exposure to the
different bedding types among the species. These results
were surprising, considering that a number of studies have
shown that the reproductive physiology of several species of
voles can be influenced by exposure to the same types of
olfactory stimuli or pheromones that were used in the
present experiment (see Puberty Modulation in Volesi .

81
A few studies have shown specifically that the transfer
of chemosignals contained within soiled bedding can
effectively produce changes in reproductive activity in
house mice (Drickamer, 1982c) and in some species of voles.
Lepri and Vandenbergh (1986) showed that exposure to
chemosignals contained within the soiled bedding of family
groups caused reproductive suppression among female pine
voles. Female pine voles were paired with unfamiliar males
and then housed in cages with either clean or family-soiled
bedding for 48 h. Females that were housed on clean bedding
had significantly heavier uteri and ovaries than females
housed on family-soiled bedding.
Soiled bedding has also been shown to affect
reproductive development among male California voles
(Rissman et al. 1984). Males that were reared from weaning
in bedding from their families had significantly lighter
seminal vesicles when they were 45, 55, and 75 days of age
than those reared in clean bedding (Rissman et al. 1984).
Androgen levels were significantly higher among the males
reared in clean bedding versus those reared on the family
bedding on day 45. Together, these studies show that the
method of exposing animals to soiled bedding can be an
effective means to reveal specific changes in reproductive
physiology, at least among some species of muroid rodents.
Several reasons can be proposed as possible
explanations for the lack of substantial differences in most
reproductive measures as a result of the different

82
conditions. It is possible that pheromones may be minimally
involved, or their activity overestimated, with regulating
the onset of reproduction in all or most species of
Microtus. It is possible that pheromones are substantially
more effective in regulating reproduction in other species,
such as in house mice (e.g., Bronson, 1979; Vandenbergh &
Coppola, 1986). Clearly more comparative research is needed
to test this possibility. Seminatural studies may be an
appropriate method in which to run biologically relevant
experiments, while still enabling the close monitoring of
reproductive development in individuals exposed to
pheromonal sources.
If pheromones are critically involved with influencing
the timing of reproduction among Microtus. there are at
least three methodological reasons that could account for
the relatively few differences found. First, it is possible
that the pervasive odors within the colony rooms effectively
eliminated many of the differences that might have been
found had animals been reared in environments that
controlled for all or most of the extraneous olfactory
stimuli. All animals were reared and tested in colony rooms
where the odors of other conspecifics animals were housed
that varied in sex, age, and reproductive status, although
the subjects had been on separate shelving for each
condition. Some support for this hypothesis is found in the
work of Batzli et al. (1977), who controlled the air supply
to individual animals. Both male and female prairie voles

83
had substantially elevated gains in body weight across a
three month period as compared to animals that were treated
similarly but shared a common air supply with other prairie
voles. In addition, Sawrey (1989/1990) found differences in
the patterns of vaginal smears of female montane voles that
had been housed either with six other females in a separate
room or had been housed in a larger colony with animals of
mixed sex and age. Vaginal smears revealed that the six
separately housed females had smaller percentages of
cornified cells than smears from females in the large colony
room (11.9% and 29.8% respectively).
A second possible reason for the relatively few
differences found among the conditions is the possibility
that the daily handling procedure was stressful enough to
effectively obscure or eliminate systematic differences.
There is indirect evidence that suggests such a possibility.
Olsen & Seabloom (1973) found that the event of captivity
caused elevated and prolonged secretion of corticosterone in
wild-caught meadow voles. It is known that increased
secretion of adrenal corticoids, androgens, progesterone,
and other steroids are associated with an inhibition of
reproduction (e.g., Christian, 1975). Thus, the daily
handling procedure may have been a significant source of
stress that might have reduced differences in morphological
and physiological growth between animals in the different
conditions.

84
A third possible reason for the relatively few
differences among the conditions steins from the experimental
design. It is possible that larger differences would have
been found among the conditions had the soiled bedding and
pheromones been presented to the subjects more frequently
than every other day. Some researches have used procedures
whereby urine was channeled directly into the cages of
subjects via tubes or they have placed cages with wire
bottoms and urine donors directly over the subject's cage
(e.g., Baddaloo & Clulow, 1981; Lepri & Vandenbergh, 1986);
thus, a near-continuous supply of pheromonal odors had been
available. Some of the information gathered in house mice
suggests that some pheromones are relatively volatile and
therefore short-lived in functional activity, while others
appear to be more stable and active over a few days. The
actual chemical composition and properties of the pheromones
from voles remain largely unknown.
Future research that is designed to identify the
effects of exposure to pheromones may be based on the
general design of this study. However, three improvements
in the design could be based on the three possible sources
of largely negative results listed above. Ideally, subjects
would have individual supplies of air, be minimally stressed
through extraneously handling or exposure to other animals,
and have a more continuous exposure to pheromonal sources.
The apparatus of Baddaloo & Clulow (1981) represents a
modification in procedure that might detect physiological

85
responses from subjects. Despite the few significant
results found in the present study, the results of many
other studies suggest that the comparative study of the
functions of pheromones among species of Microtus. appears
to be a productive area for further advances in the
elaboration of their evolved reproductive and social
patterns.

Table 3-1. Mean Body Weights (in g + S.E.) at Week 9 and Analysis (Experiment 1).
Condition Analysis of Variance
Control Family Male Female Factor df F p
Pine Voles: Males
M =
21.24
22.71
22.96
20.64
Condition
3 ,
58
1.68
. 181
(0.53)
(0.88)
(1.05)
(0.65)
Weeka
6,
348
600.87
<â– 001
N =
17
16
14
16
Interaction
18,
348
1.98
. 010
Pine
Voles: Females
M =
18.87
21.40
20.88
20.45
Condition
3,
62
1.62
. 192
(0.64)
(1.03)
(0.73)
(0.80)
Week
6,
372
466.46
< . 001
N =
16
16
18
16
Interaction
18,
372
0.51
.953
Prairie Voles:
Males
M's =
= 31.94
32.42
31.74
30.90
Condition
3,
60
0.57
. 631
(1.41)
(1.28)
(1.46)
(1.35)
Week
6,
360
346.39
< . 001
N =
16
16
16
16
Interaction
18,
360
.35
. 994
Prairie Voles:
Females
M =
26.65
26.58
27.91
26.78
Condition
3,
62
0.29
. 830
(1.03)
(0.74)
(1.31)
( -93)
Week
6,
372
247.61
<•001
N =
18
16
16
16
Interaction
18,
372
0.36
.993
Means for all weeks are located in Appendix A.
= p < 0.05; ** = p < 0.01; *** = p < 0.001.
a
*

Table 3-1—continued. Mean Body Weights (in g + S.E.) at Week 9 and Analysis (Experiment
l) •
Condition Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Meadow
Voles:
Males
M =
39.18
41.64
43.49
43.13
Condition
3,
60
1.02
.390
(2.14)
(1.97)
(1.28)
(1.58)
Week3
6,
360
589.35
<•001
N =
16
16
16
16
Interaction
18,
360
1.55
.069
Meadow
Voles:
Females
M =
30.95
30.65
32.02
32.79
Condition
3,
58
0.31
.818
(1.30)
(1.34)
(1.14)
(1.55)
Week
6,
348
302.85
<•001
M =
16
15
16
15
Interaction
18,
348
1.53
. 075
Montane Voles:
: Males
M =
37.70
37.22
39.12
38.64
Condition
3,
60
0.10
.954
(1.67)
d-51)
(2.04)
(1.35)
Week
6,
360
537.85
<â– 001
N =
16
16
16
16
Interaction
18,
360
0.92
. 545
Montane Voles:
: Females
M =
28.92
30.16
28.37
27.67
Condition
3,
59
0.30
.818
(1.43)
(1.71)
(1.43)
(1.44)
Week
6,
354
284.25
< .001
N =
16
16
15
16
Interaction
18,
354
0.93
.532
Means for all weeks are located in Appendix A.
= E < 0.05; * = p < 0.01; *** = p < 0.001.
a
*

Table 3-2. Mean Anogenital Distance (in mm + S.E.) at Week 9 and Analysis (Experiment
l) •
Condition Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Pine
Voles: Males
M =
7.71
7.70
7.96
7.75
Condition
3,
58
1.07
.365
( .22)
( .26)
( .25)
( .24)
Weeka
6,
348
140.43
_ _ „ * -k -k
< . 001
N =
17
15
14
16
Interaction
18,
348
1.17
. 282
Prairie Voles;
; Males
M =
12.20
12.31
12.28
11.25
Condition
3,
59
1.36
. 262
( -38)
( .32)
( -35)
( -36)
Week
6,
354
125.71
<.001***
N =
15
16
16
16
Interaction
18,
354
.58
. 907
Meadow Voles:
Males
M =
18.22
18.09
19.19
18.72
Condition
3 ,
60
0.41
.746
( .61)
( -50)
( -55)
( .74)
Week
6;
360
241.53
<.001***
N =
16
16
16
16
Interaction
18,
360
0.95
. 517
Montane Voles;
: Males
M =
13.44
12.25
13.09
12.66
Condition
3 ,
60
. 33
. 798
( .54)
( -31)
( .48)
( .36)
Week
6,
360
173.00
<•001***
N =
16
16
16
16
Interaction
18,
360
1.84
.019*
Means for all weeks are located in Appendix B.
= £ < 0.05; **=£>< 0.01; *** = p < 0.001.
a
*

Table
3-3. Mean Adrenal
Weights
(± S.E.)
and
Analysis (Experiment
1) .
Condition
Analysis of
Variance
Control
Family
Male
Female
Factor df
F
Pine
Voles: Males (mg tissue, uncorrected
for
differences in body
weight)
M =
4.68
4.56
4.89
5.22
Condition 3, 59
1.26
.296
Left
( -18)
( -36)
( -21)
( -17)
Position 1, 59
7.58
.008**
N =
17
16
14
16
Interaction 3, 59
0.17
.910
M =
4.43
4.42
4.59
4.91
(All Left M = 4.83
+ .12)
Right
( -25)
( -26)
( .24)
( .24)
(All Right M = 4.59
± -12)
Pine
Voles: Males (corrected for
body weight;
mg tissue/100 g body
weight)
M =
22.25
20.42
21.69
25.52
Condition 3, 59
3.32
. 026*
Left
(1.03)
(1.62)
(1.06)
( -99)
Position 1, 59
7.58
.008**
N =
17
16
14
16
Interaction 3, 59
0.14
. 929
M_ =
20.96
19.70
20.36
24.04
(All Left M = 22.49
+ .64)
Right
(1.20)
(1.09)
(1.12)
(1.36)
(All Right M = 21.28
± -62)
Pine
Voles: Females (mg
tissue, i
uncorrected for differences in body weight)
M =
4.61
4.24
4.68
4.35
Condition 3, 62
0.48
. 696
Left
( -25)
( -31)
( .22)
( -26)
Position 1, 62
9.73
.002***
N =
16
15
18
16
Interaction 3, 62
0.27
.847
M =
4.14
4.01
4.32
4.10
(All Left M = 5.86
+ .20)
Right
( .29)
( -25)
( .22)
( -31)
(All Right M = 5.26
± • 19)
= e < 0.05;
= e < 0.01;
* ★ *
= £ < 0 - 001.
oo
vo

Table 3-3—continued. Mean Adrenal Weights (+ S.E.)and Analysis (Experiment 1).
Condition
Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Pine Voles: Females (corrected for body weight; mg tissue/100 g body
weight)
M =
24 .66
20.03
22.62
21.50
Condition
3, 62
1.98
. 126
Left
(1.47)
(1.28)
(1.02)
(1.29)
Position
1, 62
8.86
.004**
N =
16
15
18
16
Interaction
3, 62
0.28
.834
M =
22.19
18.97
21.10
20.10
(All Left M
= 22.21 +
. 65)
Right
(1.66)
( -99)
(1.14)
(1.35)
(All Right M
= 20.58 +
. 65)
Prairie Voles:
Males (mg
tissue,
uncorrected
for differences in body
weight)
M =
6.06
5.22
4.89
5.22
Condition
3 , 60
2.62
. 058
Left
( .42)
( -24)
( -19)
( -31)
Position
1, 60
121.01 <.
* * *
. 001
N =
16
16
16
16
Interaction
3, 60
2.36
. 079
M_ =
5.19
4.27
4.34
4.65
(All Left M
= 5.34 + .
.16)
Right
( .38)
( -16)
( .12)
( -30)
(All Right M
= 4.61 + .
.14)
Prairie Voles:
Males (corrected
for body
weight; mg tissue/100 g body weight)
M =
19.10
16.32
15.83
16.97
Condition
3 , 60
2.44
. 073
Left
(1.09)
( .79)
( .92)
( .75)
Position
1, 60
135.70 <.
_ _ _ * * *
. 001
N =
16
16
16
16
Interaction
3 , 60
2.44
. 073
M =
16.32
13.48
14.15
15.15
(All Left M
= 17.05 +
.46)
Right
( -94)
( .69)
( .82)
( -81)
(All Right M
= 14.77 +
.42)
= E
< 0.05;
= p < 0 ,
.01;
= £ < 0
.001.

Table 3-3—continued. Mean Adrenal Weights (+ S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control Family Male Female Factor df Z E
Prairie
Voles
: Females
(mg tissue,
, uncorrected for differences in body weight)
M =
5.33
6.01
6.05
6.12
Condition
3, 62
1.21
.312
Left
( -29)
( -29)
( -42)
( -56)
Position
1, 62
83.22
<.001***
N =
18
16
16
16
Interaction
3, 62
0.98
.407
M =
4.61
5.36
5.62
5.54
(All Left M
= 5.86 +
.20)
Right
( -28)
( -32)
( -42)
( -49)
(All Right M
= 5.26 ±
•19)
Prairie
Voles
: Females
(corrected
for body
weight; mg tissue/100 g body weight)
M =
20.18
22.68
21.98
22.87
Condition
3, 62
1.18
.323
Left
( -92)
(1.04)
(1.59)
(1.72)
Position
1, 62
86.82
„ _ * * *
<•001
N =
18
16
16
16
Interaction
3, 62
0.98
.403
M =
17.40
20.21
20.33
20.72
(All Left M
= 21.88 +
. 66)
Right
( -93)
(1.18)
(1.48)
(1.52)
(All Right M
= 19.60 +
. 65)
Meadow
Voles:
Males (mg
tissue, uncorrected for differences
in body
weight)
M =
3.59
3.25
3.52
3.53
Condition
3, 60
0.74
. 528
Left
( -29)
( -16)
( .15)
( -20)
Position
1, 60
17.62
<.001***
N =
17
16
16
16
Interaction
3, 60
0.24
.862
M =
3.30
2.91
3.17
3.03
(All Left M
= 3.47 +
•11)
Right
( -21)
( -18)
( .16)
( -16)
(All Right M
= 3.11 +
. 09)
= p < 0.05;
* ;k
= e < 0.01;
***
= E < 0.001.

Table 3-3—continued. Mean Adrenal Weights (+ S.E.) and Analysis (Experiment 1).
Condition Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Meadow
Voles:
Males (corrected for body weight; mg tissue/100 g body weight)
M =
9.30
7.97
8.16
8.38
Condition
3 , 60
1.46
.233
Left
( -83)
( -45)
( .37)
( -58)
Position
1, 60
14.45
_ _ „ * * *
< . 001
N =
17
16
16
16
Interaction
3 , 60
0.23
. 874
M_ =
8.71
7.18
7.39
7.29
(All Left M
= 8.47
+ .30)
Right
( -77)
( -49)
( .45)
( -46)
(All Right M
= 7.67
± -29)
Meadow
Voles:
Females (mg tissue,
uncorrected for differences in body weight)
M =
7.94
8.15
8.12
8.27
Condition
3 , 58
0.14
.935...
Left
( -64)
( -74)
( -67)
( .43)
Position
1, 58
34.71
< . 001
N =
16
15
16
15
Interaction
3 , 58
1.18
.321
M_ =
7.24
6.34
6.37
6.96
(All Left M
= 8.12
+ .31)
Right
( -55)
( .62)
( .59)
( -52)
(All Right M
= 6.73
± -28)
Meadow
Voles:
Females (corrected
for body weight; mg tissue/100 g
body weight)
M =
26.30
27.44
25.53
25.66
Condition
3 , 58
0.29
. 829
Left
(2.13)
(2.80)
(2.01)
(1.42)
Position
1, 58
32.11
<.001***
N =
16
15
16
15
Interaction
3 , 58
1.12
.345
M =
23.81
21.20
19.88
21.74
(All Left M
= 26.22
+ 1.05)
Right
(1.76)
(2.20)
(1.75)
(1.85)
(All Right M
= 21.67
± -94)
* = E < 0.05; ** = e < 0.01; *** = p < 0.001.
UD
CO

Table 3-3—continued. Mean Adrenal Weights (+ S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control
Family
Male
Female
Factor
df
F
Montane
Voles:
Males (mg tissue,
uncorrected for differences in body weight)
M =
2.91
3.01
2.81
2.74
Condition
3,
59
0.33
.799
Left
( -19)
( -19)
( -17)
( -16)
Position
1,
59
42.77
<•001
N =
16
16
16
16
Interaction
3,
59
0.46
.709
M_ =
2.56
2.57
2.53
2.42
(All Left M
= 2.
87
+ .08)
Right
( -14)
( -16)
( .11)
( -17)
(All Right M
= 2.
52
± -07)
Montane
Voles:
Males (corrected
for body weight; mg tissue/100
g body weight)
M =
7.85
8.32
7.47
7.27
Condition
3,
59
0.53
. 665
Left
( -59)
( -68)
( -61)
( -44)
Position
1,
59
41.33
< . 001
N =
16
16
16
16
Interaction
3 ,
59
0.40
.749
M_ =
7.00
7.14
6.70
6.42
(All Left M
= 7.
73
+ .29)
Right
( -51)
( -62)
( -43)
( -43)
(All Right M
= 6.
82
+ -25)
Montane
Voles:
Females
(mg tissue, uncorrected for differences
in
body weight)
M =
6.49
6.26
6.02
5.67
Condition
3 ,
60
0.26
. 850
Left
( -51)
( -48)
( -41)
( .33)
Position
1,
60
9.99
. 002
N =
16
17
15
16
Interaction
3 ,
60
1.08
. 360
M =
5.86
5.56
5.55
5.64
(All Left M
= 6.
11
+ .22)
Right
( -48)
( -45)
( -36)
( .52)
(All Right M
= 5.
65
± -22)
* = E < 0.05; ** = e < 0.01; *** = E < 0.001.
U>

Table 3-3—continued. Mean Adrenal Weights (+ S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control
Family
Male
Female
Factor
df
F 2
Montane Voles:
Females
(corrected
for body
weight; mg tissue/100 g
body weight)
M =
23.55
20.96
21.92
21.05
Condition
3, 60
0.39 .758
Left
(2.56)
(1.22)
(1.74)
(1.27)
Position
1, 60
7.97 .006**
N =
16
17
15
16
Interaction
3, 60
1.06 .370
M_ =
21.11
18.51
20.25
21.11
(All Left M
= 21.85
+
.87)
Right
(2.25)
(1.09)
(1.61)
(2.22)
(All Right M
= 20.22
+
.91)
* *
= E
< 0.05;
= E <
0.01;
= p < 0.001.
ID
►£»

Table 3-4. Mean Testes Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Pine '
Voles: (mg
tissue,
uncorrected for differences in body
weight)
M =
20.09
19.87
22.34
22.04
Condition
3, 59
0.75
. 526
Left
(1.45)
(1.84)
(2.10)
(1.55)
Position
1, 59
0.003
.956
N =
17
16
14
16
Interaction
3, 59
0.41
.747
M =
19.91
19.46
22.34
22.09
Right
(1.45)
(1.84)
(1.98)
(1.39)
Pine ’
Voles: (corrected
for body
weight; mg
tissue/100 g body weight)
M =
95.77
88.14
99.89
108.11
Condition
3, 59
1.38
.255
Left
(7.45)
(7.53)
(9.40)
(8.28)
Position
1, 59
0.02
. 873
N =
17
16
14
16
Interaction
3, 59
0.48
. 693
M =
94.87
86.81
99.87
111.28
Right
(7.43)
(7.92)
(9.13)
(7.83)
Prairie Voles:
(mg tissue, uncorrected for
differences in body weight)
M =
147.26
140.01
162.18
139.50
Condition
3, 60
0.95
. 421
Left
(14.70)
( 8.09)
(10.98)
(10.17)
Position
1, 60
0.01
. 891
N =
16
16
16
16
Interaction
3, 60
0.21
.882
M =
145.81
140.46
164.64
139.01
Right
(15.23)
( 8.29)
(12.57)
(10.55)
* = p < 0.05; ** = p < 0.01; *** = e < 0.001.
10
U1

Table 3-4—continued. Mean Testes Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control Family Male Female Factor df F p
Prairie Voles: (corrected for body weight: mg tissue/100 g body weight)
M =
467.77
429.71
505.93
447.20
Condition
3, 60
1.68
. 179
Left
(42.22)
(16.54)
(25.33)
(20.55)
Position
1, 60
0.06
.801
N =
16
16
16
16
Interaction
3, 60
0.29
. 831
M =
462.69
431.16
517.01
446.19
Right
(42.80)
(17.19)
(22.40)
(19.95)
Meadow Voles:
(mg tissue, uncorrected for
differences in body weight)
M =
503.48
447.69
490.08
529.88
Condition
3, 61
0.85
.471
Left
(31.13)
(39.38)
(36.09)
(34.76)
Position
1, 61
13.87 <.
. 001
N =
17
16
16
16
Interaction
3 , 61
1.93
. 132
M =
505.27
466.74
496.57
545.57
(All Left M
= 492.95 ±
17.63)
Right
(32.73)
(42.09)
(38.06)
(31.84)
(All Right M
= 503.56 +
18.07)
Meadow Voles:
(corrected for body weight;
mg tissue/100 g body weight)
M =
1278.29
1109.82
1113.91
1238.01
Condition
3, 61
1.17
. 326
Left
(54.28)
(93.86)
(72.00)
(79.97)
Position
1, 61
13.39 <,
. 001
N =
17
16
16
16
Interaction
3, 61
2.13
. 105
M =
1282.24
1158.07
1127.58
1274.68
(All Left M
= 1186.44 ±
38.16)
Right
(56.28)
(99.57)
(74.95)
(72.13)
(All Right M
= 1211.74 +
38.43)
*
* *
= p < 0.05;
= p < 0.01;
** *
= p < 0.001.
KD
G\

Table 3-4—continued. Mean Testes Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Montane Voles:
(mg tissue, uncorrected for
differences in body weight)
M =
165.79
155.31
154.27
147.48
Condition
3, 60
1.06
.371
Left
( 8.08)
( 8.06)
( 8.11)
( 5.03)
Position
1, 60
4.15
. 045*
N =
16
16
16
16
Interaction
3, 60
0.13
. 938
M_ =
163.57
150.96
151.02
145.46
(All Left M
= 155.71 +
3.72)
Right
( 7.50)
( 8.33)
( 9-14)
( 4.54)
(All Right M
= 152.75 +
3.79)
Montane Voles:
(corrected for body weight;
mg
tissue/100 g
body weight)
M =
451.39
427.33
403.81
390.55
Condition
3, 60
1.31
.276
Left
(26.79)
(25.53)
(22.94)
(21.05)
Position
1, 60
6.76
. Oil*
N =
16
16
16
16
Interaction
3, 60
0.12
. 944
M_ =
445.17
415.73
394.03
383.03
(All Left M
= 418.27 +
12.15)
Right
(25.34)
(25.82)
(24.24)
(17.38)
(All Right M
= 409.49 +
11.82)
k
= E
< 0.05;
* k
= E <
0.01; ***
= p < 0.001.
«3

Table 3-5. Mean Seminal Vesicle Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control Family Male Female Factor df F
Pine Voles: (mg tissue, uncorrected for differences in body weight)
M = 25.62 23.98 29.06 27.09 Condition 3, 59 0.26
(5.61) (2.82) (3.60) (3.57)
N = 17 16 14 16
Pine Voles: (corrected for body weight; mg tissue/100 g body weight)
M = 115.97 104.57 127.60 133.30 Condition 3, 59 0.54
(22.03) (10.76) (15.09) (18.09)
Prairie Voles: (mg tissue, uncorrected for differences in body weight)
M = 211.02 187.22 193.06 161.76 Condition 3, 60 0.97
(23.62) (16.16) (23.72) (17.66)
N = 16 16 16 16
Prairie Voles: (corrected for body weight; mg tissue/100 g body weight)
M = 653.62 569.78 599.05 527.27 Condition 3, 60 0.90
(65.26) (30.48) (60.36) (59.74)
* = E < 0.05; ** = p < 0.01; *** = E < 0.001.
2
. 853
. 654
.408
.442
CO
CD

Table 3-5—continued. Mean Seminal Vesicles (in mg + S.E.) and Analysis (Experiment 1)
Condition Analysis of Variance
Control
Family Male Female Factor df F e
Meadow Voles:
(mg tissue, uncorrected for differences in body weight)
M = 269.46
(28.96)
N = 17
281.26 317.44 362.02 Condition 3, 61 1.81 .153
(32.58) (32.18) (30.07)
16 16 16
Meadow Voles:
(corrected for body weight; mg tissue/100 g body weight)
M = 695.37
(64.53)
675.15 730.41 839.85 Condition 3, 61 1.14 .336
(68.21) (72.77) (67.33)
Montane Voles:
: (mg tissue, uncorrected for differences in body weight)
M = 160.32
(10.82)
N = 16
142.56 175.96 177.37 Condition 3, 60 0.83 .480
(17.28) (22.98) (18.07)
16 16 16
Montane Voles:
: (corrected for body weight; mg tissue/100 g body weight)
M = 433.63
(32.00)
370.09 435.75 459.97 Condition 3, 60 0.94 .422
(36.41) (41.67) (46.20)
* = 2 < 0.05;
** _ „ _ ***
= E < 0.01; = E < 0.001.

Table 3-6. Mean Ovarian Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Pine
Voles: (mg
tissue,
uncorrected for differences in body weight)
M =
1.96
1.88
2.40
2.21
Condition
3, 61
1.56
.207
Left
( -12)
( -14)
(
.18)
( .14)
Position
1, 61
0.007
.933
N =
16
15
18
16
Interaction
3, 61
2.15
. 102
M =
2.12
1.91
2.11
2.28
Right
( -16)
( .20)
(
.13)
( .14)
Pine
Voles: (corrected
for body
weight; mg
tissue/100 g body weight)
M =
10.48
9.09
11.47
10.84
Condition
3, 61
1.92
. 134
Left
( -65)
( -65)
(
.80)
( .60)
Position
1, 61
0.001
. 996
N =
16
15
18
16
Interaction
3, 61
1.75
. 165
M =
11.31
9.24
10.16
11.16
Right
( -81)
( -91)
(
. 64)
( .52)
Prairie Voles:
(mg tissue
, uncorrected for
differences in body weight)
M =
2.43
2.96
2.71
2.59
Condition
3, 62
1.86
. 144
Left
( .18)
( -26)
(
.25)
( -15)
Position
1, 62
0.09
.757
N =
18
16
16
16
Interaction
3, 62
0.39
.755
M =
2.52
3.11
2.66
2.51
Right
( .16)
( -22)
(
. 22)
( .15)
* = E < 0.05; ** = p < 0.01; *** = p < 0.001.
100

Table 3-6—continued. Mean Ovarian Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control Family Male Female Factor df F
Prairie Voles: (corrected for body weight; mg tissue/100 g body weight)
M =
9.27
11.26
9.65
9.94
Condition
3,
62
2.12
Left
( -57)
( 1.00)
( -73)
( .76)
Position
1,
62
0.08
N =
18
16
16
16
Interaction
3 ,
62
0.26
M =
9.28
11.83
9.66
9.72
Right
( -47)
( -88)
( .83)
( .81)
Meadow
Voles:
(mg tissue
, uncorrected for
differences in body
weight)
M =
5.37
5.67
6.76
6.67
Condition
3 ,
58
1.14
Left
( -45)
( -59)
( -62)
( .33)
Position
1,
58
0.69
N =
16
15
16
15
Interaction
3 ,
58
1.81
M =
5.92
6.29
6.29
6.62
Right
( .44)
( -46)
( .52)
( .30)
Meadow
Voles:
(corrected
for body
weight;
mg
tissue/100 g
body weight)
M =
17.53
18.85
21.10
20.86
Condition
3,
58
0.55
Left
( 1.38)
( 1.93)
( 1-81)
( 1.22)
Position
1,
58
0.85
N =
16
15
16
15
Interaction
3 ,
58
1.71
M =
19.46
20.56
19.75
20.86
Right
( 1.41)
( 1.28)
( 1.53)
( 1.28)
*
= £
< 0.05;
•k-k
= p < 0
.01; ***
= p < 0.
001.
E
. 106
. 774
. 850
.338
.409
. 155
. 647
.360
. 174
101

Table 3-6—continued. Mean Ovarian Weights (in mg + S.E.) and Analysis (Experiment 1).
Condition Analysis of Variance
Control Family Male Female Factor df F p
Montane Voles: (mg tissue, uncorrected for differences in body weight)
M =
2.89
2.89
2.94
2.24
Condition
3,
60
1.46
.233
Left
( -26)
( .21)
( .18) (
.16)
Position
1,
60
0.77
.381
N =
16
17
15
16
Interaction
3,
60
1.96
. 129
M =
3.11
2.85
2.70
2.65
Right
( -24)
( .23)
( -18) (
.26)
Montane Voles:
(corrected for body
weight;
mg tissue/100 g
body
' weight)
M =
10.05
9.75
10.46
8.19
Condition
3,
60
1.64
. 189
Left
( -83)
( .62)
( -58) (
.45)
Position
1,
60
0.73
. 393
N =
16
17
15
16
Interaction
3,
60
2.22
. 094
M =
11.11
9.53
9.54
9.49
Right
( -97)
( .53)
( -52) (
. 66)
* = p < 0.05; ** = p < 0.01; *** = p < 0.001.
102

Table 3-7. Mean Uterine Weight (in mg + S.E.) and Analysis (Experiment 1).
Control
Condition Analysis of Variance
Family Male Female Factor df EE
Pine Voles: (mg tissue, uncorrected for differences in body weight)
M = 7.33 7.20 7.34 7.00 Condition 3, 61 0.05 .981
( .86) ( .58) ( .60) ( .56)
N = 16 15 18 16
Pine Voles: (corrected for body weight; mg tissue/100 g body weight)
M = 38.67
( 3.60)
34.72 34.51 33.81 Condition 3, 61 0.69 .558
( 2.72) ( 2.04) ( 1.88)
Prairie Voles:
(mg tissue, uncorrected for differences in body weight)
M = 11.19
( -94)
N = 18
13.59 13.92 12.16 Condition 3, 62 1.48 .228
( 1.03) ( 1.44) ( .71)
16 16 16
Prairie Voles:
(corrected for body weight; mg tissue/100 g body weight)
M = 42.04
( 3.18)
51.52 49.70 45.98 Condition 3, 62 1.38 .256
( 3.90) ( 4.39) ( 2.94)
* = E < 0.05;
** = E < 0.01; *** = e < 0.001.
103

Table 3-7—continued. Mean Uterine Weight (in mg + S.E.) and Analysis (Experiment 1).
Condition
Analysis of Variance
Control
Family
Male
Female
Factor
df
F
E
Meadow
Voles:
(mg tissue
, uncorrected for
differences in body weight)
M =
22.57
25.17
29.49
25.04
Condition
3, 57
0.38
.767
( 3.83)
( 5.11)
( 6.05)
( 3.32)
N =
16
15
16
15
Meadow
Voles:
(corrected
for body weight;
mg tissue/100 g
body weight)
M =
72.17
82.88
89.91
76.12
Condition
3 , 57
0.29
.831
(11.39)
(17.65)
(17.92)
( 9.49)
Montane Voles:
(mg tissue, uncorrected for differences in
body weight)
M =
27.46
18.56
24.35
16.52
Condition
3, 60
4.55
.006**
( 2.39)
( 1.92)
( 2.97)
( 2.22)
N =
16
17
15
16
Montane Voles:
(corrected for body weight
; mg tissue/100 g body weight)
M =
97.71
61.02
91.21
58.92
Condition
3, 60
4.61
.005**
(10.65)
( 4.42)
(14.45)
( 5.74)
= E
< 0.05;
= p < 0
. 01;
= 2 < 0.
001.
104

Table 3-8. Numbers and Percentages of Vaginally Perforate Female Microtus (Experiment 2).
Species
Total Number
Pine
Prairie
Meadow
Montane
of Females
in Study
66
66
62
64
Number of
Females
0
5
34
26
Perforate
Week 3
(0.0%)
(7.6%)
(54.8%)
(40.6%)
Number of
2
64
56
62
Females
perforate by
Week 9
(3.0%)
(97.0%)
(90.3%)
(96.9%)

Table 3-9. Species and Number of Subjects that were Vaginally Perforate on a given Day:
(Experiment 1).
Age in Days
21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61
Condition Prairie Voles
Control
N =
3
4
7
7
7
10
13
13
14
14
15
15
15
16
16
17
17
17
17
17
17
Family
N =
2
4
8
8
9
10
10
12
13
15
15
15
15
15
15
15
15
15
15
15
15
Male
N =
3
4
10
11
14
15
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
Female
N =
3
5
8
9
12
13
15
15
16
16
16
16
16
16
16
16
16
16
16
16
16
Total
N =
11
17
33
35
42
48
54
56
59
61
62
62
62
63
63
64
64
64
64
64
64
Meadow Voles
Control
N =
10
11
11
11
11
11
11
11
12
12
12
12
12
12
12
13
13
13
13
14
14
Family
N =
7
10
10
10
11
11
11
11
12
12
12
12
12
12
13
13
13
13
13
13
13
Male
N =
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
13
13
13
15
15
15
Female
N =
9
12
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
14
Total
N =
38
45
46
46
47
47
47
47
49
49
50
50
50
50
52
53
53
53
55
56
56
Montane Voles
Control
N =
7
9
10
10
10
11
14
15
15
15
15
16
16
16
16
16
16
16
16
16
16
Family
N =
7
7
7
8
8
8
11
11
12
12
12
13
14
15
15
15
16
16
17
17
17
Male
N =
8
10
10
11
11
11
12
12
12
12
14
14
14
14
15
15
15
15
15
15
15
Female
N =
6
9
9
9
9
9
9
9
9
10
9
11
13
13
14
13
14
14
14
14
14
Total
N =
28
35
36
38
38
39
46
47
48
49
50
54
57
58
60
59
61
61
62
62
62
106

Table 3-10. Species and Percentages of Subjects that were Vaginally Perforate on a given
Day (Experiment 1).
Age in Days
21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61
Condition Prairie Voles
Control%
18
24
41
41
59
76
82
82
88
88
88
94
94
100
100
100
100
100
100
100
100
Family
~o
13
27
53
53
60
67
67
80
87
100
100
100
100
100
100
100
100
100
100
100
100
Male
%
19
25
63
69
88
94
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Female
o.
19
31
50
56
75
81
94
94
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
o.
17
27
52
55
66
75
84
88
92
95
97
97
97
98
98
100
100
100
100
100
100
Meadow Voles
Control
%
71
79
79
79
79
79
79
79
86
86
86
86
86
86
86
93
93
93
93
100
100
Family
%
54
77
77
77
85
85
85
85
92
92
92
92
92
92
100
100
100
100
100
100
100
Male
o.
80
80
80
80
80
80
80
80
80
80
80
80
80
80
87
87
87
87
100
100
100
Female
o.
64
86
93
93
93
93
93
93
93
93
100
100
100
100
100
100
100
100
100
100
100
Total
%
68
80
82
82
84
84
84
84
88
88
89
89
89
89
93
95
95
95
98
100
100
Montane Voles
Control
44
56
63
63
63
69
88
94
94
94
94
100
100
100
100
100
100
100
100
100
100
Family
%
41
41
41
47
47
47
65
65
71
71
71
76
82
88
88
88
94
94
100
100
100
Male
2-
53
67
67
73
73
73
80
80
80
93
93
93
93
100
100
100
100
100
100
100
100
Female
%
43
64
64
64
64
64
64
64
64
71
64
79
93
100
93
100
100
100
100
100
100
Total
%
45
56
58
61
61
63
74
76
77
79
81
87
92
94
97
95
98
98
100
100
100
107

108
Condition
Figure 3-1. Mean body weights (in g + standard error) of
male pine voles in each condition at 9 weeks of age (see
text for explanation of conditions). Columns with different
letters are significantly different (p's < .05).

109
Condition
Figure 3-2. Mean Adrenal weights (corrected for differences
in body weight; mg/100 g body weight + standard error) of
male pine voles in each condition at 9 weeks of age (see
text for explanation of conditions). Columns with different
letters are significantly different (p's < .05).

110
Condition
Figure 3-3. Mean Uterine weights ( + standard error) of
montane voles in each condition at 9 weeks of age (see text
for explanation of conditions). a) uterine weight
(uncorrected for differences in body weight) in mg; b)
uterine weight corrected for differences in body weight
(mg/100 g body weight). Columns with different letters are
significantly different (p's < .05).

Ill
Age in Days
Figure 3-4. Cumulative percentages of females that became
vaginally perforate during Experiment 1 at given ages (in
days). Prairie: prairie voles; Meadow: meadow voles;
Montane: montane voles.

112
Figure 3-5. Mean age of vaginal perforation among Microtus
(in days of age + standard error) as a function of
experimental condition. No significant differences were
evident within each species (see text for explanation of
experimental conditions).

113
Figure 3-6. Mean percentages of cornified cells found in
vaginal smears of prairie voles as a function of
experimental condition and age (data are shown by two-day
block intervals. represents a significant differences
(p's < .05) among conditions on given day (see text for
explanation of conditions and significant differences).

114
Age in Days
Figure 3-7. Mean percentages of cornified cells found in
vaginal smears of meadow voles as a function of experimental
condition and age (data are shown by two-day block
intervals; see text for explanation of conditions).

Age in Days
Figure 3-8. Mean percentages of cornified cells found in
vaginal smears of montane voles as a function of
experimental condition and age (data are shown by two-day
block intervals. represents significant differences (j>'
< .05) among conditions on given day (see text for
explanation of conditions and significant differences).

CHAPTER 4
BEHAVIORAL RESPONSES OF VOLES (MICROTUS) TO
PUBERTY MODULATING STIMULI (EXPERIMENT 2)
Rationale
Although a considerable amount is known about the
physiological responses of house mice and voles to
pheromones that produce puberty modulation, very little is
known about how these species respond behaviorally to their
presence (see Coppola and O'Connell, 1988; Drickamer,
1989b). Knowledge about how individuals might behaviorally
diminish or enhance the effects of pheromones found in the
environment appear critical to understand the process of
puberty modulation and possibly the formation of social and
mating systems. Vandenbergh and Coppola (1986) suggested
that we determine whether young animals will investigate or
avoid urine that can influence puberty. It is necessary to
determine whether behavior directed toward such stimuli can
be influenced by the social environment or other relevant
factors. For example, do female mice investigate the urine
of males if more dominant females are present? Similarly,
do fluctuations in resources, such as food availability,
influence behavior and presumably the timing of puberty and
reproduction?
Suggestive, although limited, data have been gathered
for behavioral responses of mice to various cues. Female
116

117
mice will selectively avoid or approach puberty-modulating
chemosignals depending upon their age and reproductive
status. Drickamer (1989b) found that prepubertal female
mice typically avoided odors of adult males, whereas
postpubertal or adult females were preferentially attracted
to them. Recording the behavioral responses of young male
mice or voles to similar cues would be of interest, although
no controlled studies have been reported.
In a related study, Drickamer (1988) provided evidence
that early puberty and early reproduction were associated
with a shorter life span for female mice compared with those
that bred at later ages. Thus, it seems reasonable to
assume that females will selectively avoid odors that cause
them to attain puberty before their body systems can
withstand the energetic demands of reproduction and hence
limit their lifetime reproductive success. It can be
predicted that young males would seek cues associated with
females at a very early age, unless social factors inhibited
them. Among house mice, aggression by territorial males
toward other males is a pervasive phenomenon (Bronson &
Coquelin, 1980).
Limited evidence suggests that investigation or
avoidance of olfactory and pheromonal cues occurs among some
species of voles, and may be influenced by the state of
sexual maturity. Sawrey (1989/1990) conducted odor
preference tests with adult male and female montane voles
for male-soiled and female-soiled bedding. Females

118
preferred male-soiled bedding compared to female-soiled
bedding. Somewhat surprisingly, male montane voles also
preferred male-soiled bedding. Sawrey noted that although
the preference of females for male bedding had been found in
a variety of species, the result of the males' preference
was unexpected. He suggested the possibility that the
establishment of a territory exclusive of other males may be
a prerequisite for the acquisition of mates by an unmated
male. Additional data are needed to test this possibility.
Similar measures gathered for other species of Microtus, and
with subjects that varied in reproductive status, would be
useful to determine how behavior might be functionally
linked to puberty modulation and observed differences in
social and mating systems.
There is suggestive evidence that voles avoid common
sources of olfactory cues or pheromones in their natural
environment. Wolff (1980) found that non-reproductive
female taiga voles (M. xanthoqnathus) and subordinate males
did not use scat piles produced by dominant males, although
reproductively active females used them. It is possible to
interpret this pattern as one of avoidance of male urinary
cues, female cues, or both, by juveniles. Additional study
is necessary to determine whether similar behavioral
patterns occur in other species of Microtus and what
function they serve.
The following study addresses the need for additional
investigation of how behavior may modulate the timing of

119
puberty and how these behaviors may be related to the
expressed social and mating system. Few data are available
on such preferences; thus the experiment was largely
exploratory, with no firm indications that changes in
development would be associated with changes in olfactory
preferences. Similar procedures were used for each sex and
species to enable meaningful comparisons among them.
Measures and predictions. Because Experiments 1 and 2
were run concurrently, it was not possible to derive
predictions based on the results of Experiment 1. However,
I made general predictions that the behavioral patterns
would reflect some of the patterns seen in Experiment 1. I
predicted that those species that experienced substantial
reproductive change from exposure to the stimuli in
Experiment 1, would show related behavioral differences in
Experiment 2. Specifically, because it was believed that
the two less social species would show signs of puberty
acceleration when exposed to opposite-sex stimuli in
Experiment 1, I expected the meadow voles and montane voles
would prefer opposite-sex odors and would either show no
preference toward stimuli that caused puberty inhibition or
reveal an aversion to them.
In addition to these predictions on odor preference,
additional predictions were based on patterns of general
development. Providing that the social species were found
to be affected by the stimuli in Experiment 1, I believed
that the more social and generally slower developing

120
species, pine voles and prairie voles, would not be
attracted to odors of opposite-sexed adults or possibly be
attracted to odors from same-sexed adults when they were
very young (4 weeks of age). These predictions were based
on Drickamer's (1989b) finding that prepubertal female mice
typically avoided odors of adult males, while postpubertal
or adult females were preferentially attracted to them. I
believed that as the more social species reached maturity,
there would be a shift in their behavior to one of a
preference for opposite-sex odors.
In contrast to the highly social species, I believed
that the less social species, meadow voles and montane
voles, would display an earlier and sustained attraction for
odors of the opposite sex, providing they had been affected
by the stimuli in Experiment 1. These predictions were
based on the generally quicker rates of maturation among
meadow voles and montane voles compared to the other species
(McGuire & Novak, 1984, 1986; Nadeau, 1985), and the
demonstration of olfactory preferences for opposite-sex
odors by meadow voles during periods of reproductive
activation during the summer months (Ferkin & Seamon, 1987).
Two behavioral measures were recorded to investigate
how behavior might be associated with changes in development
and expressed differences in social and mating system.
First, measures of odor preference were recorded by
measuring the total duration subjects remained within 1 cm
of each stimulus. Second, the total durations subjects

remained in the center (neutral) area of the test cage were
recorded. The second measure was recorded to reveal
121
possible differences in general aversion to both stimuli
among the subjects. It was possible that subjects might not
display olfactory preferences for one stimulus versus
another, although differences could exist within or between
species in general aversion to the stimuli that might be
revealed by this measure.
Method
Subjects
A total of 129 animals, 15 to 18 males and 15 to 18
females from each of four species of voles served as
subjects. Species included pine voles fMicrotus pinetorund .
prairie voles (M. ochroqasterl, meadow voles
(M. pennsvlvanicus), and montane voles (M. montanusl.
All behavioral measures were taken between January 1991 and
November 1991.
Procedure
Animals were selected randomly, via a random number
table, from litters, providing that at least two of the
animals were of each sex. This criterion served to ensure
that all subjects had previously been exposed to siblings of
both sexes prior to testing. All subjects were weaned and
individually caged at three weeks of age (21-22 days of
age). No more than one male and female from the same litter
were used in the experiment. All subjects were housed and
maintained in small individual cages as previously

described. Cages were cleaned weekly by animal caretaking
staff.
122
Odor preference tests. The olfactory preferences of
each subject were tested in a repeated-measures design when
they were 4, 7, and 10 weeks of age. Each test session
consisted of recording selected behavioral patterns of
individuals when they were presented with a simultaneous
preference (choice) task that consisted of pooled bedding
from unfamiliar adult males and adult females. Stimulus
bedding material was pooled separately for both sexes that
came from the cages of five adult males and five females
that had not had their bedding changed for one week. Soiled
bedding was collected in 200 cc samples from each cage and
was pooled and mixed thoroughly. Small samples from the
pooled bedding of each sex (approximately 20 cc) were placed
in two jars (4.5 cm in diameter and 8.5 cm length) that were
covered with a fine-mesh, concave screen that enabled
subjects to inspect the stimulus bedding. The two jars were
inserted into both ends of a large 48 X 27 X 13 cm clear
plastic cage that was modified to hold the jars securely.
The length of the cage was marked into three equal areas.
One side contained the jar of the male bedding, the other
side contained the jar with the female bedding. The third
(center) area was free of any soiled bedding. The left and
right placements of the male and female stimuli were
randomized throughout the experiment. The apparatus had
been used previously and thus validated to detect estrous

123
and diestrous preferences in male prairie voles (Taylor and
Dewsbury, 1988), and to reveal olfactory preferences among
adult montane voles (Sawrey, 1989/1990).
Each olfactory test was 10 min in duration, with the
onset of the 10-min behavioral period being initiated when
the subjects were observed to place their nares within 1 cm
of one of the stimulus jars. Behavioral measures were
recorded on a portable computer that was transported into
the colony room of each species prior to testing. Subjects
were tested within the dark portion of the photoperiod (1200
h - 2000 h) under the illumination of two 25-watt red light
bulbs. Behavioral measures included the total duration the
subject's nares were within 1 cm of each stimulus jar and
total duration subjects occupied the center portion of the
test cage.
Body weight and reproductive status. Body weights were
recorded for subjects at the end of each behavioral test.
In addition, a vaginal smear was taken from female subjects
that were perforate after each behavioral test. Vaginal
smears were scored under blind conditions by the author to
assess the percentages of cells in each smear.
Statistical Analysis
Results of the odor preference tests were analyzed
independently for each species and sex by means of two-way
analysis of variances (ANOVA) with the bedding type (male or
female) and week of testing comprising repeated-measure
factors. Student Neuman-Keuls tests were conducted for

124
comparisons of means where the F value for the main effect
or interaction was statistically significant. Alpha was
held at .05 in all comparisons, and all were based on a
two-tailed distribution. Comparisons of cell types of the
vaginal smears were compared using nonparametric analyses
after the cell frequencies were converted to percentages of
the total cell number in each smear.
Results
Preference Tests
Few significant differences were found among the
within-species comparisons of preferences for male or female
bedding. Only female prairie voles and female meadow voles
displayed a preference for male versus female bedding,
although male pine voles, male meadow voles, female meadow
voles, and male montane voles showed significant differences
in the amount of time they remained near the stimuli (see
below). Means and analyses for these within-species
comparisons are summarized in Table 4-1 for males and Table
4-2 for females of all species.
Prairie voles: females. Female prairie voles displayed
a significant preference for the male bedding compared to
the female bedding, but only when viewed across the three
test sessions, (main effect of stimulus, F(l, 15) = 8.59,
p = .010). Post-hoc comparisons for each week did not
reveal any significant preferences for the male bedding.
Across the three test sessions, females remained within 1 cm
of the male stimulus for 50.34 + 11.19 s, on average, while

125
remaining near the female stimulus for 33.60 + 7.02 s. They
did not differ significantly in the amount of time they
remained near the male and female bedding across the study,
(main effect of week, F(2, 30) = 0.59, p = .560). The
interaction of bedding type and week of test was not
statistically significant, F(2, 30) = 0.15, p = .860 (see
complete means in Table 4-2).
Meadow voles: females. Female meadow voles displayed a
significant preference for the male bedding versus the
female bedding, when viewed across the three test sessions
(main effect of stimulus, F(l, 17) = 12.45, p = .002)(see
complete means in Table 4-2). Across the three test
sessions, female meadow voles remained preferentially within
1 cm of the male stimulus for 62.80 + 10.78 s, on average,
while near the female stimulus for 33.05 + 7.47 s. Post-hoc
comparisons revealed that females remained significantly
longer near the male bedding than near the female bedding
during week 7 (M's = 79.15 + 13.13 s versus 37.45 + 9.57 s).
The females remained near both stimuli (collectively)
significantly more during week 7 (M = 58.30 + 11.35 s), than
during week 10 (week 10, M = 38.05 ± 7.08 s)(main effect of
week F(2, 34) = 3.73, p = .034). The mean duration that
females were near the stimuli during week 4 was intermediate
in value to those recorded during the other weeks (M = 47.42
+ 8.96 s). The interaction of bedding type and week of test
was not statistically significant, F(2, 34) = 0.84, p =
.438.

126
Differences In the Time Near the Stimuli Across Weeks
Pine voles: males. Male pine voles did not display a
significant preference for either type of bedding (main
effect of stimulus, F(l, 14) = 0.009, p = .923)(see means in
Table 4-1) . However, they remained near to the stimuli,
collectively, for a longer total duration during week 4 (M =
73.40 + 11.24 s) than during week 10 (M = 40.54 + 12.57
s)(main effect of week, F(2,28) = 5.98, p = .006). The
total duration they remained near the stimuli during week 7
was intermediate in duration to those of the other two weeks
(M = 59.37 + 12.57 s). The interaction of bedding type and
week of test was not statistically significant, F(2, 28) =
1.55, p = .227.
Meadow voles: males. Male meadow voles did not display
a significant preference for either type of bedding (main
effect of stimulus, F(l, 17) = 0.98, p = . 334) (see complete
means in Table 4-1). However, their total time near the
stimuli decreased significantly each week (week 4: M = 62.98
+ 8.07 s; week 7: M = 44.40 + 9.06 s; and week 10: M = 26.08
+ 7.25 s)(main effect of week, F(2, 34) = 17.25, p < .001).
The interaction of bedding type and week of test was not
statistically significant, F(2, 34) = 2.40, p = .105.
Montane voles: males. Male montane voles did not
display a significant preference for either type of bedding,
(main effect of stimulus, F(l, 15) = 0.64, p = .434 (see
complete means in Table 4-1). However, they remained near
the stimuli significantly longer during week 4 (M = 113.67 +

127
17.69 s), than during week 7 (M = 73.63 + 16.11 s) or week
10 (M = 72.93 + 18.42 s) producing a main effect of week,
F (2, 30) = 7.93, p = .001. The interaction of bedding type
and week of test was not statistically significant, F(2, 30)
= 0.92, p = .40.
Between-Soecies Comparisons
Between-species comparisons were conducted with two
types of analysis. First, the preference ratios of male to
female bedding were compared across each species
independently for each sex and for each week of study (Table
4-1 and 4-2). Second, the total durations that the nares of
the subjects remained within 1 cm of each stimulus were
compared across each species and independently for each sex
and week of study (Table 4-1 and 4-2). Nonparametric
analyses (Kruskal-Wallis ANOVA by ranks tests and subsequent
Mann-Whitney U tests) were used in both analyses because of
heterogeneity of variances between species (see results
below).
Preference ratios of males for male versus female
bedding. No significant differences were found when the
preference ratios of male to female bedding were compared
among the males of all species for each week: week 4,
H(3, N = 64) = 2.08, p = .554; week 7, H(3, N = 64) = 2.29,
p = .513; week 10, H(3, N = 64) = 1.78, p = .617 (see Table
4-1 for means).
Preference ratios of females for male versus female
bedding. No significant differences were found when the

128
ratios of male to female preferences were compared among the
females of all species for each week: week 4, H(3, N = 67) =
0.38, p = .944; week 7, H(3, N = 67) = 2.17, p = .536; week
10, H(3, N = 67) = 2.88, p = .410 (see Table 4-2 for means).
Comparisons of the total durations males remained
within 1 cm of the stimuli. The males of all species
differed in the total duration their nares were within 1 cm
of the male stimulus during each of the three tests.
However, the males differed significantly in the total
duration they remained near the female stimulus only during
week 10 (see Table 4-1 for means).
During week 4, the male montane voles remained near the
male stimulus significantly longer than the pine voles (U =
45.0, p = .003), prairie voles (U = 56.0, p = .011), and
meadow voles (U = 56.0, p = .002), H(3, N = 64) = 12.44,
p = .006. None of the species differed in the total time
they remained near the female stimulus on week 4, H(3, N =
64) = 4.04, p = .256.
During week 7, male prairie voles remained near the
male stimulus significantly more than either the pine voles
(U = 62.0, p = .036) or the meadow voles (U = 62.0,
p = .008), H(3, N = 64) = 8.59, p = .035. None of the
species differed in total time they remained near the female
stimulus, H(3, N = 64) = 1.46, p = .690.
During week 10, male meadow voles remained near the
male stimulus significantly less than the prairie voles
(U = 72.0, p = .022) and the montane voles (U = 57.0,

129
p = .002), H (3, N = 64) = 11.32, £ = .010. Male meadow
voles also remained near the female stimulus significantly
less than the pine voles (U = 71.0, p = .020), prairie voles
(U = 56.0, p = .004), and montane voles (U = 76.0,
p = .018), H(3, N = 64) = 10.71, p = .013.
Comparisons of the total duration females remained
within 1 cm of the stimuli. The females of the species
revealed fewer species differences in the total duration
they remained within 1 cm of the stimuli than did the males.
Females were found to differ significantly in their total
durations near the male stimulus during weeks 4 and 7. No
species differences were found in the total durations
females remained near the female stimulus during any of the
tests (see Table 4-2).
During week 4, the female montane voles remained near
the male stimulus significantly longer than the pine voles
(U = 79.0, p = .042), prairie voles (U = 65.0, p = .006),
and meadow voles (U = 90.0, p = .022), H(3, N = 67) = 9.28.
None of the species differed significantly in their total
time near the female stimulus, H(3, N = 67) = 7.68,
p = .053.
During week 7, the female montane voles remained near
the male stimulus significantly longer than the pine voles
(U = 43.0, p < .001) and prairie voles (U = 65.0, p = .006),
H(3, N = 67) = 15.32, p = .001. Additionally, female meadow
voles remained near the male stimulus longer than did the
pine voles (U = 72.0, p = .022). None of the species

130
differed in their total time near the female stimulus,
H (3 , N = 67) = 2.15, p = .540.
During week 10, none of the females of the species
differed in the duration they were near the male stimulus,
H(3, N = 67) = 6.03, p = .110, or the female stimulus,
H (3, N = 67) = 4.16, p = .244 .
Duration within Center of Cage: Within-Species Analyses
Within-species analyses of the total duration that
subjects remained within the center (neutral) of the test
cage revealed little systematic variation (see Table 4-3 for
means and analyses of males and Table 4-4 for females).
Comparisons revealed that male meadow voles remained in the
center of the cage for a longer duration during weeks 4 and
7 than on week 10 (see below). The remaining sex and
species differences were not significantly affected by age
for this measure.
Meadow voles. Male meadow voles remained in the center
of the test cage significantly longer during week 4
(M = 94.88 + 11.92) and week 7 (M = 88.07 + 9.73) than
during week 10 (M = 57.50 + 8.29), F(2, 34) = 7.03,
p = .002.
Duration within Center of Cage: Between-Species Analyses
Between-species comparisons revealed significant
differences among the total duration males remained in the
center of the test cage, although they were apparent only
during week 10 (see Table 4-3 for means).

During week 10, male meadow voles remained in the
center region significantly less than did the pine voles
(U = 56.0, p = .004) and prairie voles (U = 59.0, p = .006)
H(3, N = 64) = 11.09, p = .011. No differences were found
among the durations males were in the center of the cage
during week 4, H(3, N = 64) = 5.46, p = .140, or week 7,
H(3, N = 64) = 1.06, p = .785.
There was more variation among females of the species
to remain in the center of the cage than were found among
the males (see means in Table 4-4). Species differences
among females were found during each of the three test
sessions. During week 4, female meadow voles remained in
the center of the cage significantly less than the female
pine voles (U = 66.0, p = .012) and prairie voles (U = 80.0
P = .027), H(3, N = 67) = 7.85, p = .049.
During week 7, female meadow voles remained in the
center of the cage significantly less than did the female
pine voles (U = 53.0, p = .003) and montane voles (U = 99.0
p = .046), H(3, N = 67) = 10.18, p = .017. During week 10,
meadow voles remained in the center of the cage
significantly less than did the pine voles (U = 41.0,
p < .001), H(3, N = 67) = 10.63, p = .013.
Body Weights and Vaginal Smears
Because the analyses of body weight and vaginal
cytology were secondary analyses and not the primary focus
of the present study, the results of these analyses are
located in Appendix D and Appendix E.

132
Discussion
Olfactory Preference Tests
Surprisingly few significant differences were found for
olfactory preferences for male or female odors within each
species at any of the weeks of testing. Female prairie
voles and meadow voles revealed significant preferences to
be near the male bedding when their data were pooled across
the test sessions for each species. Female meadow voles
revealed the highest level of attraction for the male
bedding during the second test (week 7). Whether the high
level of attraction by meadow voles for male bedding on week
7 is associated with a general peak of attractivity to male
stimuli during the maturation period is not clear. In
partial support of this possibility, female montane voles
also showed a peak in attraction to the male stimulus during
week 7. Thus, because the developmental rates of these two
species appear to be similar (McGuire & Novak, 1984, 1986;
Nadeau, 1985), the data suggest that some form of peak of
male attraction occurs somewhere near week 7 of age for
meadow voles and montane voles. Additional study is needed
to support or refute this hypothesis.
In contrast, there was little variation in the
durations female prairie voles investigated each stimulus
across each week. Thus, there was no indication that
prairie voles showed a gradual increase in the amount of
time they remained near the male stimulus as they matured.
Such a developmental pattern had been shown by female house

133
mice as they matured (Coppola & O'Connell, 1988; Drickamer,
1989b).
There is only limited comparative information available
concerning olfactory preferences for male or female odors
among species of Microtus. Ferkin and Seamon (1987) found
that season influenced odor preferences in free-ranging
adult female meadow voles. During the spring-summer
breeding season, with long hours of daylight, both wild
males and females preferred odors of opposite- versus
same-sex conspecifics. However, during the late fall and
winter period, which is typically a period of reproductive
quiescence, females preferred female odors, whereas males
did not show a preference. In other research, the
preference of female meadow voles for male odors during the
summer months was shown to be estrogen dependent;
ovariectomy eliminated the preference, whereas estradiol
reinstated the preference (Ferkin & Zucker, 1991). Thus,
these studies suggest that as estrogen levels rise among
female meadow voles, a preference for male versus female
odors emerges. A preference for male odors may also develop
as estrogen levels rise during the normal maturational
process. Concurrent studies of changes in hormone levels
and odor preferences across the early developmental period
would clarify this possibility.
Although the males of each species did not display any
olfactory preference, pine voles, meadow voles, and montane
voles remained near the stimuli for different durations. In

134
all cases, longer durations of investigation were found in
the earlier rather than later test sessions. These findings
are counter-intuitive to the prediction that greater
preferences would be shown as the males developed into
adults. However, it is possible that some of the decreases
in durations that the males remained near the stimuli were
due to repeated exposure across the course of the study. I
anticipated that such decreases would be minimal given the
few tests imposed and the relatively long three-week
interval between each test. However, because
repeated-measures tests were used, it is not possible to
rule out the possibility that repeated testing affected the
general decline in the durations of investigation by these
three species. A between-subjects design could provide
evidence whether the gradual decline of investigation was
due primarily to repeated testing, or whether the decline
was due to changes associated with maturation.
Between-Species Comparisons
Preference ratios of males and females for soiled
bedding. Comparisons of the ratios of male to female
preferences between species revealed no significant
differences among any of the species during any of the test
sessions for each sex. However, such analyses eliminate
differences in the actual amount of time that the species
remained preferentially near one stimulus versus the other.
Comparisons of the total durations males were near the
male and female stimuli. Few systematic differences are

135
discernable when comparing the total durations males were
near the male stimulus. Male montane voles revealed high
levels of attraction to the male stimulus compared to the
other species during week 4, although the level dropped by
nearly 50% by week 7. Previously, Sawrey (1989/1990) found
that adult male montane voles preferred to be on the cage
side containing male versus female bedding. Male meadow
voles revealed relatively low levels of investigation on
weeks 7 and 10 compared to the other species.
Male meadow voles were the only group to differ
significantly in the duration they remained near the female
stimulus. Male meadow voles displayed only a small interest
in the female stimulus compared to the other species at week
10 (Table 4-1). However, this difference does not appear to
be due solely to the nature of the stimulus, because they
also displayed a gradual decrease in their interest to both
stimuli across the study.
Comparisons of the total durations females were near
the male and female stimuli. Systematic differences were
more clear among the females (Table 4-2). Female montane
voles remained near the male stimulus significantly longer
than any of the other species during weeks 4 and 7, although
no systematic differences were found among the durations
females remained near the female stimulus. Because the
uteri of montane voles were found to be significantly
heavier among those exposed to male bedding in experiment 1,
I had anticipated there would be a corresponding preference

136
to male bedding in Experiment 2. However, such a
relationship was not found to be statistically significant.
In a previous study, Sawrey (1989/1990) found that
adult female montane voles preferred to remain on the cage
side containing male versus female soiled bedding. Although
no significant preferences were found in the present study
for female montane voles to remain near the male versus the
female bedding, they did remain near the male cue on average
longer than with the female stimulus across each week (Table
4-2). Together, the data from the present study and that of
Sawrey's (1989/1990) suggest that female montane voles
display preferences for male bedding under some conditions.
It is possible that because male montane voles are
territorial, females may be preferentially attracted to male
odors if they are isolated from other animals and are in
search of mates.
Duration with Center of Cage
Within-species comparisons of the total duration
subjects remained in the center of the test cage revealed
only one significant difference among the males of the
species. Females of each species were not found to differ
significantly on this measure. Male meadow voles remained
in the center of the cage significantly longer during weeks
4 and 7 than week 10 (Table 4-3). Why this pattern would
occur only among male meadow voles is not clear. The
response pattern does not appear to be one of increased
attraction to the stimuli, because there were concurrent and

137
significant decreases in the total duration that male meadow
voles remained within 1 cm of the stimuli across each week
(Table 4-1). Behavioral differences associated with any
common pattern of development among the four species of
Microtus also do not appear to explain the gradual decrease.
None of the males of the other species revealed an
indication of a gradual decrease in the duration they
remained in the center of the cage with increasing age
(Table 4-3).
Between-species comparisons of the duration subjects
remained in the center of the cage revealed most differences
among the females of the species. Male meadow voles
remained in the center of the cage significantly less than
did the pine voles or prairie voles on week 10 (Table 4-3).
Female meadow voles remained in the center of the test cage
significantly less than the pine voles and prairie voles
during week 4, significantly less than the pine voles and
montane voles on week 7, and significantly less than the
pine voles on week 10 (Table 4-4). Thus, the pattern that
emerges is that female meadow voles typically remained in
the center of the cage for durations that were considerably
less than those of the other species. During this time,
female meadow voles showed high interest in male bedding
across each week.
Conclusions (Experiment 2)
This exploratory study of olfactory preferences for
adult female and male odors among young Microtus revealed

138
that few significant preferences were displayed by each
species. Female prairie voles and female meadow voles
revealed a small preference for the male versus the female
bedding across the weeks of study. However, these few
results do not enable one to make a clear evaluation of
whether preferences toward the bedding are associated with
differences in social or mating systems.
Two explanations seem most plausible to explain the
general paucity of results. First, olfactory preferences
and other behavioral differences that are influenced by
olfaction may not be displayed by relatively young Microtus
(those between 28 and 70 days of age). Second, this lack of
preference could be due to relatively little experience with
these stimuli. Richmond & Stehn (1976) suggested that prior
experience or learning may significantly affect whether
olfactory preferences will be displayed among some species
of Microtus.
A few studies have shown that olfactory preferences
occur among the adults of some species of Microtus.
Olfactory preferences have been reported among adult male
and female montane voles (ages 60-185 days)(Sawrey,
1989/1990), adult female meadow voles (ages 50-90
days)(Ferkin & Zucker, 1991), adult male meadow voles (ages
70-120 days of age)(Ferkin & Gorman, 1992), and adult
prairie voles (specific ages not reported)(Taylor, 1988).
Thus, it is possible that preferences might be detected
among the species studied if they were tested at slightly

139
older ages or given more extensive exposure to the stimuli.
The results of the present study do not suggest any clear
indication that preferences were displayed during the last
test session, although all species and sexes displayed a
greater attraction to the male versus the female bedding on
week 10 (see Tables 4-1 and 4-2). Additional study with
individuals that ranged in age more widely could provide
evidence whether and when preferences are shown by various
species of Microtus.
An alternative explanation for the relatively few
olfactory preferences observed among the species is that the
subjects of all or some of the species may have preferred
one stimulus compared to the other, but the preferences were
not expressed because the apparatus failed to elicit them.
Taylor and Dewsbury (1988) found that adult male prairie
voles did not prefer bedding that had been soiled by females
in estrus compared to bedding from diestrous females,
providing the bedding was presented in jars. However,
preferences were displayed by the males when females were
tethered or placed within small cages. Thus, it is possible
that slight modifications in the procedure might reveal
preferences among some of the species at these ages.
Clearly, additional behavioral studies are needed to
distinguish among the possibilities described above. The
use of different methods to determine olfactory preferences
may clarify whether which, if any, of these species display
olfactory preferences for male versus soiled bedding at

140
young ages. In addition, such study may enable us to
determine whether functional links exist between olfactory
preferences and expressed social and mating systems.

Table 4-1. Mean Durations of Males (in s + S.E.) within 1 cm of Bedding (Experiment 2).
Week: Week 4
Odor: Male Famale
Week 7
Male Female
Week 10
Male Female
Analysis of
Factor df
Variance
F
E
Pine Voles: Males (N
= 15)
M = 70.66 76.14
49.96 68.78
51.21
29.88
Odor
1,
14
0.009
. 923
(10.38) (12.10)
(10.39) (14.75)
(11.11)
( 4.26)
Week
2,
28
5.98
.006**
Inter-
2,
28
1.55
. 227
% M:Fa = 47.76
46.27
57 .
. 47
action
Prairie Voles: Males
(N = 15)
M = 74.06 64.90
94.08 65.20
66.71
56.03
Odor
1,
14
4.08
. 062
(13.70) (13.06)
(14.65) (12.74)
(13.36)
(12.41)
Week
2,
28
1.17
. 323
Inter-
2,
28
0.46
. 630
% M:F = 52.79
57.82
55.
. 19
action
Meadow Voles: Males
(N = 18)
M = 71.29 54.68
40.87 47.93
34.84
17.32
Odor
1,
17
0.98
.334
( 7.96) ( 8.18)
( 6.42) (11.71)
(10.17)
( 4.33)
Week
2,
34
17.25
<•001***
Inter-
2,
34
2.40
. 105
% M:F = 56.68
47.93
63 .
. 22
action
Montane Voles: Males
(N = 16)
M = 134.79 92.56
70.89 76.37
75.89
69.97
Odor
1,
15
0.64
.434
(19.49) (15.90)
(14.56) (17.65)
(12.86)
(23.99)
Week
2,
30
7.93
.001***
Inter-
2,
30
0.92
.400
% M:Fa = 59.95
52.94
62.
,28
action
a% M:F Represents Mean percentage of time near male stimulus versus female stimulus.
* = p < 0.05; ** = p < 0.01; *** = p < 0.001.
141

Table 4-2. Mean Durations of Females (in s + S.E.) within 1 cm of Bedding (Experiment 2)
Week: Week 4
Week 7
Week 10
Analysis
of
Variance
Odor: Male
Famale
Male
Female
Male
Female
Factor
df
F
n
Pine Voles:
Females
(N = 15)
M = 70.34
57.57
36.49
79.98
57.93
45.71
Odor
1,
14
0.19
. 669
(16.76)
(16.32)
( 7.03)
(23.41)
(16.27)
(14.45)
Week
2,
28
1.02
.373
Inter-
2,
28
2.85
. 074
% M:F = 53
. 66
51.
72
51
.48
action
Prairie Voles: Females (N =
16)
M = 53.26
40.41
49.61
29.70
48.17
30.69
Odor
1,
15
8.59
.010**
(13.89)
( 9.33)
( 9.31)
( 5.67)
(10.40)
( 6.06)
Week
2,
30
0.59
.560
Inter-
2,
30
0.15
.860
% M:Fa = 54
. 00
60.
06
60
.32
action
Meadow Voles: Females (N = 18)
M = 57.59
37.25
79.16
37.45
51.66
24.45
Odor
1,
17
12.45
.002**
(10.21)
( 7.72)
(13.13)
( 9.57)
( 9.03)
( 5.12)
Week
2,
34
3.73
. 034*
Inter-
2,
34
0.84
.438
% M:F = 56
.70
65.
21
69
.07
action
Montane Voles: Females (N =
15)
M = 120.72
85.73
132.77
59.38
91.83
74.95
Odor
1,
17
2.77
. 114
(20.76)
(14.64)
(24.07)
(15.20)
(18.45)
(22.05)
Week
2,
34
1.52
.232
Inter-
2,
34
1.78
.183
% M:F = 57.
15
69.93
61.'
75
action
a% M:F Represents Mean percentage of time near male stimulus versus female stimulus.
* = p < 0.05; ** = p < 0.01; *** = p < 0.001.
142

Table 4-3. Mean
Week: Week 4
Durations
Week 7
Males were
Week 10
(in sec + S.E.) in Center of Cage (Experiment 2).
Analysis of Variance
Factor df F p
Pine Voles: Males (N = 15)
M =
116.05
(16.29)
97.48
( 9.67)
101.10
(11.05)
Week
2,
28
0.82
.448
Prairie
Voles:
Males (N =
15
)
M =
99.58
(14.95)
102.85
(16.00)
111.65
(14.60)
Week
2,
28
0.31
.735
Meadow '
Voles: Males (N =
18)
M =
94.88
(11.92)
88.07
( 9.73)
57.50
( 8.29)
Week
2,
34
7.03
. 002
Montane
Voles:
Males (N =
16
)
M =
68.30
( 8.26)
90.24
(15.54)
83.72
(14.27)
Week
2,
30
1.17
.323
* = p < 0.05; ** = p < 0.01; *** = p < 0.001.
143

Table 4-4. Mean Durations Females (in sec + S.E.) were in Center of Cage (Experiment 2).
Week: Week 4 Week 7 Week 10 Analysis of Variance
Factor df F p
Pine Voles: Females (N = 15)
M =
125.85 147.85
(15.58) (20.05)
128.09
(16.20)
Week
2,
28
0.73
.488
Prairie
Voles: Females (N
= 16)
M =
144.52 113.28
(29.34) (18.18)
105.64
(20.55)
Week
2,
30
1.25
.299
Meadow Voles: Females (N =
= 18)
M =
73.69 72.29
(12.07) (11.92)
61.79
(10.59)
Week
2,
34
0.48
. 621
Montane
Voles: Females (N
= 18)
M =
107.42 107.80
(12.77) (15.40)
96.01
(16.82)
Week
2,
34
0.41
. 664
* = p < 0.05; ** = p < 0.01; *** = p < 0.001.
144

CHAPTER 5
EFFECT OF SIRE PRESENCE OR ABSENCE ON DEVELOPMENT OF
OFFSPRING (EXPERIMENT 3)
Rationale
One of the greatest differences between the sexes of
mammalian species is the variation in their involvement in
parental care (Clutton-Brock, 1991). Many species of
monogamous mammals are comprised of breeding pairs where the
adult male aids the female, either directly or indirectly,
with the rearing of the offspring (Kleiman, 1977; Kleiman &
Malcolm, 1981). However, the reasons why some species form
monogamous mating systems, with extensive male parental
care, while others do not, are still unknown (Dewsbury,
1987). Thus, a complete understanding of the formation of
mating systems among mammals must incorporate an
understanding of the evolution of parental behavior
(Wittenberger, 1979).
One difficulty with understanding the evolution of male
parental care in mammals is that males appear to reduce
their reproductive potential by mating with only a single
female, rather than seek out additional mates (Kleiman,
1977). Seeking out additional mates would appear to be the
most advantageous for males, unless some other constraint,
such as the spacing of females, limited a male's access to
them (Emlen & Oring, 1977). However, under circumstances
145

146
where a male cannot monopolize a disproportionate share of
mates, it may be to his advantage to remain with one female
and to act paternally (Emlen & Oring, 1977).
Monogamy and parental care may be favored whenever more
than a single individual is needed to rear the offspring
(e.g., Kleiman, 1977). There is evidence that among certain
species of primates, male parental care is associated with
relatively high neonate:mother weight ratios (Kleiman,
1977). Thus, some species appear to have evolved a strategy
of producing large offspring, or many offspring per litter,
that a solitary female cannot normally rear alone (Kleiman,
1977) . However, many researchers have simply assumed that
paternal care has beneficial effects for the offspring;
critical tests that the presence of adult males has such an
effect are still needed (Wuensch, 1985).
Parental care in Microtus. Both male and female pine
voles and prairie voles interact extensively with their
offspring (Oliveras & Novak, 1986; Wilson, 1982), whereas in
meadow voles and montane voles, the females usually spend
substantially larger amounts of time interacting with the
offspring than do males (Hartung & Dewsbury, 1979; Oliveras
& Novak, 1986; Wang & Novak, 1992).
The consequences of parental care among Microtus have
been studied to a limited extent. Wilson (1982) studied
male and female meadow voles and prairie voles under
laboratory conditions. Male prairie voles were more likely
to huddle with the young, while females were absent from the

147
nest, than were male meadow voles. However, Wilson's
results did not provide direct evidence that the father's
presence contributed to the growth and survival of the
offspring. Wilson (1982) suggested that the amount of body
contact commonly observed among the members in a family
group under captive conditions might produce major species
differences in social structure while under natural
conditions.
A preliminary study that addressed the question of
whether, or how, a sires' presence may directly influence
his offspring was conducted with prairie voles by Pierce and
Dewsbury (1989). Their design included monitoring the
development of the first and second litters of inexperienced
breeding pairs. Adult males were run in a counterbalanced
order so they were present either throughout the rearing of
the first litter or second litter, but not both. The
results indicated that the number of pups weaned, sex ratio,
and body weights of surviving offspring at weaning were not
significantly affected by male presence. However, females
produced more offspring in the second litter when the sire
had been present during the first litter. These results
suggest that some measures may be useful for detecting the
influence of male presence in Microtus. and that male
presence may be especially beneficial for new breeding
pairs.
In a related study, Storey and Snow (1987) found that
among meadow voles, the presence of either the sire or

148
another adult male led to significant increases in the body
weight of offspring compared to the weight of offspring
reared alone by the female. However, in a recent
investigation conducted in a seminatural environment, Wang &
Novak (1992) found evidence that the presence of an adult
meadow vole father, juvenile offspring, or both, exerted
primarily a negative influence on the growth and development
of litters. Litters that were reared only in the presence
of the mother developed the most rapidly. Mother-reared
offspring experienced the earliest fur and eye opening of
offspring from either condition. In contrast, the same
researchers found a generally positive influence of the
adult father, additional juveniles, or both, when they were
present with female prairie voles to rear litters. Their
data revealed that prairie vole fathers contributed
extensively to the rearing of the offspring by remaining in
the natal nest and exhibiting parental care. Offspring that
were reared with both parents, or with both parents and
juveniles present, ate solid food earlier and moved out of
the natal nest earlier than offspring reared without the
father present. Thus, the results of the research by Wang &
Novak (1992) showed virtually opposite patterns of pup
development between meadow voles and prairie voles.
Measures of offspring development were generally negative
among meadow vole litters when additional family members
were present, whereas measures of offspring development were

149
generally positive with additional family members present
for the rearing of prairie vole litters.
Together the studies by Pierce and Dewsbury (1989),
Storey and Snow (1987), and Wang and Novak (1992), suggest
that there are positive effects upon the development of
prairie vole offspring when adult sires, or other family
members, are present during the rearing of litters.
However, the results are eguivocal whether paternal
influence is positive or negative upon developing meadow
vole offspring (Storey & Snow, 1987; Wang & Novak, 1992).
It is possible that the differences found between the
studies reviewed above are a function of procedural
differences or other differences between the laboratories.
Nevertheless, it is worthwhile to critically test whether
male presence can influence the development of offspring
among several species of Microtus that differ in social
organization.
Although an adult sire may often be present for prairie
voles or pine voles under field conditions, a sire or other
adult male may be present for several species of Microtus at
particular times of the year, such as during the late fall
and winter months. It has been documented that winter
breeding occasionally occurs in meadow voles (Madison, 1984)
and montane voles (Jannett, 1984). Conceivably, any
stimulus that has been available reliably throughout the
evolution of a species' social and mating system could

150
function to influence general development and other social
behaviors.
The following study was designed to assess the
influence that an adult male (father) might have on the
development of his offspring, among four species of voles
(Microbus). Several measures were recorded in an attempt to
determine which, if any, could detect differences in the
effect of male presence or absence.
Measures and predictions. I predicted that male
presence in the more social species, pine voles and prairie
voles, would result in positive benefits upon their
offspring's development. I predicted that, in comparison to
litters reared alone by the females, male-present litters
would weigh more at weaning, the day of eye-opening would be
earlier, there would be a shorter inter-birth interval
between the first and second litters, and there would be
more offspring in second litters. In contrast to the
assumptions for the more social species, I predicted there
would be little or no positive influence upon the measures
listed above for litters born to meadow voles or montane
voles. Meadow voles and montane voles are not typically
reared in the immediate presence of an adult male, and have
a generally more rapid rate of growth than pine voles or
prairie voles (McGuire & Novak, 1984, 1986; Nadeau, 1985).

151
Method
Subjects
A total of 364 litters born to the four species of
voles were observed for differences in offspring development
as a function of male presence. Breeding pairs consisted
initially of sexually inexperienced adult males and females.
A total of 49 pairs of pine voles CM. binetorunü , 38 pairs
of prairie voles ÍM. ochroqasterl, 61 pairs of meadow voles
fM. oennsvlvanicusi, and 34 pairs of montane voles
ÍM. montanus), were monitored during the development of
their first and second litters. Data were gathered from
newly established breeding pairs between September 1989 and
June 1992.
Procedure
Each litter born to the breeding pairs was assigned
randomly into one of two conditions. In the male-present
condition (Together-Alone order), the first litter produced
by a breeding pair developed in the presence of the dam and
sire throughout the litter's first three weeks of
development. The male was then removed from the female one
day after the birth of the second litter. Removal of the
male one day after parturition provided an opportunity for
postpartum mating and subsequent pregnancies in both
conditions. In the male-absent condition (Alone-Together
order), the sire was removed one day after the birth of the
first litter and was replaced on the day of weaning of the
first litter to be present during the rearing of the second

152
litter. Thus, breeding pairs were treated in a
counter-balanced design by removing or retaining the sire
during the rearing of the first and second litters (see
Figure 5-1 for graph of experimental design).
The number of offspring born in each litter was counted
on the day of birth and the number that survived to day 21
(weaning) was recorded. The sexes of the surviving
offspring were determined at weaning. Each offspring was
weighed to the nearest 0.01 g on a pan-balance scale at
weaning. In addition, the day that eye-opening occurred for
the majority of offspring in a given litter (50%) was
recorded.
Statistical Analysis
All measures of reproductive performance were analyzed
with Analysis of Variance (ANOVA) techniques and Analysis of
Covariance (ANCOVA) techniques where indicated. The results
from each species were analyzed independently because of
instances of heterogeneity of variances between species.
The between-subject factor was the condition (order of male
presence) and the repeated measure factor was the litter
number (first and second). Post-hoc analyses were conducted
with Neuman-Keuls post-hoc comparison tests. The
probability values for the post-hoc tests are not presented
in the text in order to streamline the section. The alpha
level was held at .05 in all comparisons, and all
comparisons were based on a two-tailed probability
distribution.

153
Results
Delay to Produce Litters
The order of male presence (condition) significantly
influenced the time required for pine voles to produce their
second litters. The mean number of days that passed before
breeding pairs of all species produced their first and
second litters are found in Table 5-1 with the analyses.
The delay of the birth of the first litter was determined by
counting the number of days that passed from the initial
date of pairing to the birth of the first litter. The delay
until the birth of the second litter was determined by
counting the number of days from the day of birth of the
first litter until the birth of the second litter.
Pine voles. Pine vole pairs in both conditions gave
birth to their first litters after approximately the same
number of days following pairing (Together-Alone order:
M = 59.87 + 6.85 days, N = 24; Alone-Together order:
M = 51.92 + 6.64 days, N = 25). However, pairs in the
Together-Alone order gave birth to their second litter
significantly sooner than those in the Alone-Together order
(M = 32.29 + 2.01 days after the first litter and M = 53.88
+4.98 respectively). A significant interaction of
condition and litter number was present, F(l, 47) = 8.71,
E < .005) .
Figure 5-2 displays the frequency distribution of the
number of second litters born to breeding pairs of pine
voles. Frequencies of litters are arranged by the number of

154
days that elapsed between the births of the first and second
litters (see Figure 5-1). The frequency distribution
reveals that most breeding pairs in the Together-Alone order
(18 of 24 or 75%) gave birth to their second litter within
34 days after birth of the first litter. This time span
represents pairs producing the second litter within 10 days
of a normal gestation length of 24 days (Nadeau, 1985) and
suggests that most of the Together-Alone pairs had
successful fertilization within the post-partum estrus
following the birth of the first litter.
In contrast to pairs of the Together-Alone condition,
only 4 of 25 (16%) pairs in the Alone-Together condition
gave birth to their second litters within 34 days of
producing the first litter. Nearly half of these pairs (12
of 15 or 48%) gave birth to their second litter within 45-54
days after the birth of the first litter. Thus, this delay
until birth of the second litter suggests that most of the
breeding pairs in the Alone-Together condition aborted or
skipped a pregnancy after the male was removed one day
following the parturition of the first litter.
Because the distributions of the delay time for the
pine voles to produce their first and second litters were
not normally distributed in either condition, the data were
compared with nonparametric analyses. Briefly, the results
of the nonparametric analyses reflected those of the
analysis of variance. No differences were found between the
breeders of the conditions to produce their first litter

155
(Mann-Whitney U test: U = 248.00, 2 = .298), although pairs
in the Together-Alone condition delivered their second
litters significantly sooner than those in the Alone-
Together order (U = 91.50, p < .001).
Additional ANOVA results indicated the main effect of
litter number was statistically significant, F(l, 47) =
6.55, p = .013, although the main effect of condition was
not, F (1, 47) = 1.31, p = .256.
Number of Offspring Born
The condition did not significantly affect the number
of offspring born among any of the species for either
litter, although prairie voles and montane voles had larger
second litters compared to their first litters (see below).
The mean number of offspring born to each species and the
analyses are located in Table 5-2.
Prairie voles. The number of offspring born to prairie
voles in their first and second litters was not
significantly influenced by the condition, F(l, 36) = 0.57,
p = .454 (see means Table 5-2). However, fewer offspring
were born to prairie voles in the first litter (M = 3.29 +
.18, N = 38), than in the second litter (M = 3.81 + .24,
N = 38), which produced a significant main effect of litter
number, F(l, 36) = 4.25, p = .046. The interaction of the
condition and litter number was not significant, F(l, 36) =
0.51, p = .478.
Montane voles. The number of offspring born to montane
voles was not significantly influenced by the condition,

156
F (1, 32) = 1.42, g = .241 (see means in Table 5-2).
However, the average litter size was smaller in the first
litter (M = 3.88 + .26, N = 34) than in the second litter
(M = 4.70 + .24, N = 34), producing a main effect of litter
number, F(l, 32) = 5.42, p = .026. The interaction of
condition by litter number was not statistically
significant, F(l, 32) = 0.78, p = .383.
Age 50% of Offspring Opened Eves
The condition did not significantly affect the age when
any of the species' offspring opened their eyes, whether the
results were analyzed with or without adjusting for the
number of offspring born to each litter (see means and
analyses in Table 5-3). However, the eyes of the second
litters produced by meadow voles opened significantly
earlier than those of the first litter.
Meadow voles. The day that meadow voles opened their
eyes was not significantly influenced by the condition,
F(l, 30) = 1.63, p = .688 (see means Table 5-3). However,
offspring from the second litter (M = 8.03 + .11, N = 32)
opened their eyes significantly earlier than those in the
first litter (M = 8.41 + .15, N = 32)(main effect of litter
number, F(l, 30) = 7.56, p = .009). The interaction of
condition by litter number was not statistically
significant, F(l, 30) = 0.57, p = .452.
An analysis of covariance (ANCOVA), using the number
born in each litter as a covariate, produced results similar
to those of the primary analysis (main effect of condition,

157
1(1, 29) = 0.15, ¡a = .699, litter number, F(l, 29) = 7.82,
2 = .009; and interaction of condition and litter number,
F(l, 29) = 0.44, 2 = .510.
Number of Offspring Weaned
The number of offspring surviving from the day of birth
until weaning was uniformly high for all species and for
both litters; typically this was 80% or greater. The
condition did not significantly affect the number of
offspring weaned among any of the species, whether or not
the results were analyzed with or without a covariate of the
number of offspring born to each litter (see means and
analyses in Table 5-4).
Sex Ratio of Offspring Weaned
The condition produced a significant main effect on the
sex ratio among the montane voles. Significantly more males
were born to pairs in the Together-Alone order; no
significant interaction of condition by litter number was
evident (see below). Mean sex ratios and analyses are
located in Table 5-5.
Montane voles. The sex ratio of montane voles that
were weaned were significantly influenced by the condition,
(main effect of condition, F(l, 26) = 4.73, p = .038).
Across both litters, montane vole pairs in the
Together-Alone condition produced a greater proportion of
males than females compared to those in the Alone-Together
condition (M's = 61.77% + 6.09% versus 47.18% + 7.01%)
although none of the pairwise comparisons among means by

158
each condition and litter were statistically different (see
Figure 5-3 and Table 5-6 for means). Neither the main
effect of litter number, F(l, 26) = 2.07, 2 = .161, nor the
interaction of condition by litter number, F(l, 26) = 0.309,
2 = .582, was statistically significant.
An analysis of covariance (ANCOVA), using the number
weaned in each litter as a covariate, produced results that
were similar to those of the primary analysis (main effect
of condition, F(l, 25) = 4.38, 2 = -046; main effect of
litter number, F(l, 25) = 1.20, 2 = .283; and interaction of
condition and litter number, F(l, 25) = 0.32, 2 = .572).
None of the pairwise comparisons among individual means were
statistically significant.
Body Weight of Offspring Weaned
The condition significantly influenced the mean body
weight of pine vole offspring (see below). The analysis
also indicated an influence of litter number. The mean body
weights were significantly heavier in the second litters
produced by the prairie voles and montane voles when they
were corrected for the number of offspring weaned (see
below). The mean individual body weights of offspring at
weaning and their analyses are located in Table 5-6 for each
species.
Pine voles. The mean body weights of individual pine
voles that were weaned (Day 21) were significantly
influenced by condition (interaction of condition and litter
number, F(l, 26) = 6.89, p = .014). Post-hoc comparisons

159
revealed that offspring weights were similar during the
rearing of the first litter in both conditions and not
statistically different (Together-Alone: M = 11.85 + .65,
N = 16; Alone-Together: M = 13.05 + .79, N = 12)(see Figure
5-4). However, offspring produced by pairs in the
Together-Alone condition were significantly heavier in the
second litter (M = 13.68 + .63, N = 16) than in the first
litter. Pairs in the Alone-Together condition weaned
offspring during the second litter that were smaller than in
the first litter (M = 12.38 + .85, N = 12), but were not
statistically different between litters.
Neither the main effect of condition, F(l, 26) = 0.003,
p = .956, nor the main effect of litter number, F(l, 26) =
1.48 p = .234, was statistically significant.
An analysis of covariance (ANCOVA), using the number of
offspring weaned in each litter as a covariate, produced
results similar to those of the primary analysis (main
effect
of
male presence,
Z(l,
25)
= 0.11,
E =
.738 ;
main
effect
of
litter number,
Z(l,
25)
= 2.95,
E =
.097;
and
interaction of condition and litter number, F(l, 25) = 9.22,
p = .005. Post-hoc analyses revealed that only the second
litters of pairs in the Together-alone order were
significantly heavier than those produced in the first
litter.
Prairie voles. Although the standard analysis of
variance failed to detect a significant effect of condition,
litter number, or interaction, the analysis of covariance

160
revealed a main effect of litter number, (see Table 5-6 for
means and analyses).
An analysis of covariance (ANCOVA), using the number
weaned in each litter as a covariate, produced results
similar to those of the primary analysis (main effect of
condition, F(l, 29) = 0.26, p = .613). However, a main
effect of litter number was detected, F(l, 29) = 6.64,
p = .015. The mean weight of prairie voles weaned during
the second litter were significantly heavier than those in
the first litter, despite the litter size increasing
slightly between the first and second litters (first litter
mean offspring weight: M = 18.68 + .74; second litter:
M = 19.08 + .63). The interaction of condition and litter
number was not statistically significant, F(l, 29) = 0.48,
p = .491.
Montane voles. Although the standard analysis of
variance failed to detect a significant effect of condition,
litter number, or interaction, the analysis of covariance
revealed a main effect of litter number, (see Table 5-6 for
means and analyses). An analysis of covariance, using the
number weaned in each litter as a covariate, failed to
detect a significant main effect of condition, F(l, 25)
= 0.27, p = .273. However, the mean weights of offspring
weaned during the second litter were significantly heavier
than those in the first litter, despite an increase in the
litter size between the first and second litters (first
litter mean offspring weight: M = 15.38 + .62; second

161
litter: M = 16.10 + .59)(main effect of litter number,
F(l, 25) = 5.19, g = .031). The interaction of condition
and litter number was not statistically significant,
F (1, 25) = 2.45, £ = .129.
Discussion
Although most of the results suggested little positive
influence of male presence among the species, two results
suggest an influence of male presence on offspring
development among pine voles. The first indicates that male
presence can have a substantial effect on the amount of time
required before pine voles produce litters. The second
result suggests that the average body weight of pine vole
offspring is greater during the second litter if the male is
present during the rearing of the first litter (see later
section on body weight of individual offspring). Together
these results suggest that continued male presence can have
a positive influence on offspring development and
survivorship and hence may influence the formation of the
social and mating system in pine voles (see discussion
below).
Delay to Produce Litters
Pine voles that were run in the Together-Alone
condition produced their second litter, on average, 32 days
after having the first litter (see Figures 5-1 and 5-2).
However, if the males had been absent during the first
litter (Alone-Together condition), the pairs typically
produced their second litter 54 days after having the first

162
litter. This result is suggestive evidence that the
continued presence of the adult sire is critical for the
retention of a second pregnancy in pine voles, following
post-partum mating, and may be critical for the first litter
to be produced as well. In effect, male pine voles may have
to remain in close proximity to a given female, even after
mating, in order to successfully reproduce. This constraint
could predispose males to pair monogamously with females and
could lead to the evolution of paternal care, as has been
documented in this species (McGuire & Novak, 1984; Schadler,
1990).
Studies have suggested that female pine voles are
sensitive to changes in the social environment and will not
retain pregnancies or will fail to rear existing offspring
under disrupted conditions. It was reported that a
substantial amount of time was necessary for new breeding
pairs to become "acclimatized" to laboratory conditions
before successful reproduction began (Kirkpatrick &
Valentine, 1970). Schadler (1982) found that female pine
voles that were exposed to strange males experienced
abortions at most stages of pregnancy, even those that were
lactating. In another study, the placement of an unfamiliar
male with a female and a 4-day old litter led to a
substantial reduction in the number of offspring that
survived when compared to the survivorship of offspring if
the female was left alone (Schadler, 1985). The loss of
offspring occurred despite the lack of obvious wounding of

163
the offspring or fighting between the female and the
unfamiliar male.
There is some evidence that prairie voles may also be
relatively sensitive to changes in the environment during
reproduction. Nadeau (1985) reported that among five
species compared for levels of prenatal mortality (percent
of ova lost), prairie voles had the highest rate whereas
montane voles had the lowest. Pine voles were not compared
in the analysis. The apparent reproductive sensitivity of
pine voles, and perhaps prairie voles, to social and
environmental disruptions can be contrasted with, what
appear to be, lower sensitivities for disruption in some
other species of Microtus. For example, in a design similar
to that used by Schadler (1985) with pine voles, Storey and
Snow (1987) placed unfamiliar male meadow voles with
pregnant females and assessed the males' influence on litter
development. Results indicated comparable survivorship and
average weight gain in offspring reared either with an
unfamiliar male (non-sire) or with a sire. Both groups
produced more offspring that survived and were heavier
compared to offspring reared by the females alone.
Number of Offspring Born
The condition (order of male presence) did not have a
noticeable influence on the number of offspring produced in
either litter among any of the species. However, prairie
voles and montane voles had significantly larger second
litters. This result appears to be a common one found among

164
several species of Microtus. at least for their first few
litters. For example, litter size has been shown to
increase with age and parity in montane voles (Negus &
Pinter, 1965), but not in meadow voles (Keller & Krebs,
1970). There is some support that litter size increases
across the first few litters in prairie voles, but then
declines (Richmond & Conaway, 1969a). The results of the
present study coincide with the results of these earlier
studies.
Interestingly, litter size may be one of the attributes
that responds most rapidly, and perhaps most often, to
selection pressures (Schaffer & Tamarin, 1973). The only
measure that was found to be statistically significant, in
the study by Pierce and Dewsbury (1989) of male influence
with prairie voles, was litter size. More offspring were
produced in the second litter when the male had been present
with the female for the rearing of the first litter.
However, these results were not replicated in the present
study with prairie voles.
Age 50% of Offspring Opened Eves
The males' presence or absence did not significantly
influence the time of eye opening among any of the species,
although meadow vole offspring opened their eyes
significantly earlier in the second litter than the first.
This result might have occurred because there were more
offspring born in the second litters, but some of them died
before weaning. It is possible that the surviving offspring

165
were able to share resources, such as food and possibly body
heat from the female, with fewer siblings, which in turn led
to an earlier day of eye opening among the survivors.
Previous study of prairie voles has shown that individual
offspring weights are inversely dependent on litter size
(Richmond & Conaway, 1969a).
The ages when most offspring opened their eyes in the
present study are similar to values reported for eye-opening
in earlier studies for each species (McGuire & Novak; 1984,
1986; Dewsbury, 1990).
Number of Offspring Weaned
Although male presence or absence did not have a
significant effect on the number of offspring weaned for any
species in the present study, other research has shown
differential effects of male presence on the number of
offspring weaned. McGuire et al. (1992) found that when
adult male meadow voles (sires) remained with females to
rear litters, fewer offspring survived than when the males
were removed shortly after mating. In contrast, female
prairie voles were generally successful at rearing all
offspring produced in a litter, whether or not the male was
present during rearing. It is not clear why differences
were found between the McGuire et al. (1992) study and the
present study. One possibility is that the breeding pairs
in the McGuire et al. (1992) study had been exposed to each
other periodically only for three days prior to the mating,
gestation, and subsequent rearing of the first litter. In

166
the present study, breeding pairs remained in the same cage,
throughout the rearing of the first two litters.
Sex Ratio of Offspring Weaned
The male's presence or absence was found to
significantly affect the sex ratio in montane voles. Pairs
in the Together-Alone condition had a greater percentage of
male offspring than females, when the data were collapsed
across both litters, whether the analyses were or were not
corrected for differences in litter size (see Figure 5-3).
However, there were no significant differences in the sex
ratios when each litter was compared between each condition
(post-hoc analyses).
Interpretation of this male-biased sex ratio is not
straightforward. It is possible this difference was due to
chance, because the effect was small and the interaction
between the condition and litter number was not
statistically significant. If male presence or absence
affected the sex ratio in útero, we would expect larger
differences in the sex ratio during the second litter
because males of both conditions were present during the
gestation of the first litter. Although the absolute
differences in the sex ratios were larger in the second
litter, the differences between the sex ratios was not
statistically different during the second litter.
A few reports have indicated skews in the sex ratio of
montane voles, although they were in the opposite direction
found in the present study. Vaughan et al. (1973) reported

167
there were more female than male montane voles born within a
laboratory colony. A similar finding was reported by
Jannett (1981), where there were more older females than
males in five wild populations of montane voles (see Nadeau,
1985). In any case, most species of Microtus appear to have
litters that approximate a 1:1 sex-ratio (Nadeau, 1985).
Additional study appears necessary to determine if more
males than females typically are produced using these or
similar procedures.
Body Weight of Offspring Weaned
Male presence appeared to substantially affect the
individual offspring weights of pine voles. Mean weights
were greater in the second litter than they were in first
litter, when the offspring had been reared by pairs in the
Together-Alone condition (see Figure 5-4) . This trend
remained after correcting for the number of offspring weaned
in each litter. In contrast, the mean offspring weights of
pine voles decreased between the litters produced by pairs
in the Alone-Together condition. No significant differences
in offspring weight were found between the conditions, when
they were compared separately within the first and second
litters.
Two interpretations seem possible for the interaction
of condition and litter number or mean body weights among
pine voles. First, male pine voles may have little impact
during the rearing of the first litter, whether the male is
or is not present. However, if the male is present during

168
the second litter, when he had not been present for the
first, he somehow disrupts the female's ability to rear
large offspring. A second interpretation is that males have
a positive impact when they are present during the rearing
of the first litter, but it is not revealed clearly until
the second litter is produced and the male is absent.
Unfortunately, it is not possible to clearly
differentiate these possibilities with the present
experimental design. However, with the use of two other
groups, these different hypotheses could be tested. Two
additional groups, a group in which the male remains for
both the first and second litters (Together-Together
condition), and one in which he is present only for mating
(Alone-Alone condition), could determine among the
possibilities. Results for the first litter should not
change; in other words, the mean weight of all groups should
be similar and not statistically different during the
rearing of the first litter. However, during the rearing of
the second litter, the hypotheses could be tested. If
offspring weights of those in the Together-Together
condition were equal to or greater than offspring weights
during the second litter of those run in the Together-Alone
condition, this would suggest a positive influence of the
male. If the offspring weights were significantly less than
those in the Together-Alone condition, then it would suggest
a disruptive effect of male presence.

169
A Reanalvsis of Che Evolution and Expression oí Mala
Parental Behavior
The results of Experiment 3 suggest that some measures,
recorded under certain conditions, can reveal positive or
negative effects of a male's presence or absence during the
rearing of pine vole litters. However, the results of many
studies with other rodent species have shown that the
effects are not always reliably detected (Dewsbury, 1985).
A number of factors that could contribute to this problem,
and possible solutions, are listed below.
First, in studies where the male is removed to
determine the effect of his absence, no effect may be found
because the remaining female may be capable of adjusting her
behavior to compensate for the loss of the other parent
(Wuensch, 1985). However, the results of the present study,
and of Dewsbury's (1988), suggest that the use of
repeated-measures tests, across two or more litters, might
be useful to detect the gradual effect of either the loss or
gain of a male's energy to the rearing of offspring.
Ideally, studies would measure lifetime reproductive success
of females with and without male presence, but studies of
shorter duration might be sufficient to demonstrate the loss
or gain of the male's contributions. For example,
monitoring the development of two or three litters could be
sufficient to demonstrate positive or negative effects of a
male's presence or absence. Similarly, if a female must
devote considerable effort to rear offspring without a male,

170
measures such as the time invested in eating or total
calories consumed should reflect the larger energy
expenditure by the female. In contrast, calories saved by a
male's presence may reflect less calories consumed by
females.
A second reason why the effects of male presence may
not be detected readily is that males may normally
contribute to the welfare of the offspring, but it is simply
not observed under typical conditions of the laboratory
environment (Dewsbury, 1985). For example, there is
evidence that male defense of a litter may be more common
than thought among some species of Microtus. Shrews and Blarina) appear to be common predators of Microtus
offspring (Pearson, 1985), but might be deterred under some
circumstances. Getz et al. (1992) revealed that both sexes
of prairie voles displayed aggressive behavior toward and
successfully defended nestlings from short-tailed shrews
(Blarina brevicauda^. In contrast, female meadow voles did
not behave aggressively toward the shrews or otherwise
protect the nestlings. Indirect evidence suggests that male
pine voles may also successfully defend their offspring
against some forms of predators, such as shrews. Results of
paired encounters between pine voles and meadow voles
revealed that male pine voles were more aggressive than male
meadow voles, and in some cases were more dominant (Cranford
& Derting, 1983; Novak & Getz, 1969).

171
Together the results of the present study and of others
suggest that for pine voles, the continued presence of a
male may make the difference between the retention of a
gestating litter and the survival of his current offspring.
Within the present experiment, male presence appeared to be
most critical for successful reproduction in pine voles, but
given the undoubtedly more difficult conditions in the
field, the retention of a male may be critical for
successful reproduction in other species of Microtus as
well. For example, relatively little is known about the
social dynamics that occur during winter breeding, but it
has been reported to occur among several species of
Microtus. including meadow voles, prairie voles, California
voles CM. californicus). and Townsend's voles
(M. townsendii)(see Jannett, 1984). It may be that during
harsh times as these, males contribute substantially to the
success of their offspring.
The emerging view is that male paternal behavior may be
expressed by many, if not all species of Microtus, at least
under particular conditions (see Hartung & Dewsbury, 1979).
Dewsbury (1985) noted that although stable species
differences exist in the levels of paternal behavior
expressed among rodents, the behavioral patterns appear to
be similar in form but are expressed under varying
thresholds and conditions. The next critical step appears
to be to identify what conditions elicit paternal care,
determine what the thresholds are, and how the benefits and

172
costs of paternal behavior are traded among individuals
forming the social groups where it is expressed. It seems
plausible that once this information is known, we may be in
a better position to understand the selective pressures
shaping and maintaining given social and mating strategies
among Microtus.

Table 5-1
Mean Number of Days (+ S.E.) until Birth of Litters (Experiment 3)
Condition: Together-Alone Alone-Together
Litter: First Second First Second
Analysis of Variance
Factor df F p
Pine Voles:
M =
59.87
32.29
51.92
53.88
Condition
1,
47
1.31
.256
( 6.85)
( 2.01)
( 6.63)
( 4.98)
Litter
1,
47
6.55
.013
N =
24
24
25
25
Interaction
1,
47
8.71
_ „ **
. 004
Prairie
Voles:
M =
32.11
25.06
29.50
33.60
Condition
1,
36
0.45
. 502
( 7.35)
( 1-58)
( 2.95)
( 2.51)
Litter
1,
36
0.15
.700
N =
18
18
20
20
Interaction
1,
36
2.13
. 152
Meadow
Voles:
M =
27.93
28.30
31.90
27.68
Condition
1,
59
0.45
. 501
( 2.24)
( 2.35)
( 3.43)
( 1.86)
Litter
1,
59
0.54
.463
N =
30
30
31
31
Interaction
1,
59
0.77
.383
Montane
Voles:
M =
25.31
25.50
27.56
31.33
Condition
1,
32
2.42
. 129
( 1-41)
( 2.14)
( 2.70)
( 2.73)
Litter
1,
32
0.87
.356
N =
16
16
18
18
Interaction
1,
32
0.71
.403
*
= E
< 0.05;
= p < 0.01;
= E <
0.001.
173

Table 5-2. Mean Number of Offspring (+ S.E.) Born in Litters (Experiment 3)
Condition: Together-Alone Alone-Together
Litter: First Second First Second
Analysis of Variance
Factor df F p
Pine Voles:
M =
1.62
2.00
1.60
1.56
Condition
1,
47
3.36
.072
( -12}
( -13)
(
•11)
( -10)
Litter
1,
47
2.42
. 125
N =
24
24
25
25
Interaction
1,
47
3.72
. 059
Prairie
Voles:
M =
3.06
3.78
3.50
3.85
Condition
1,
36
0.57
. 454
( -27)
( -36)
(
.23)
( .33)
Litter
1,
36
4.25
.046*
N =
18
18
20
20
Interaction
1,
36
0.51
.478
Meadow
Voles:
M =
3.87
4.67
4.03
3.62
Condition
1,
60
1.55
.217
( -34)
( -41)
(
.29)
( -30)
Litter
1,
60
0.37
.542
N =
30
30
32
32
Interaction
1,
60
3.51
. 065
Montane
Voles:
M =
3.81
4.31
3.94
5.06
Condition
1,
32
1.42
. 241
( -28)
( -39)
(
.44)
( -33)
Litter
1,
32
5.42
. 026*
N =
16
16
18
18
Interaction
1,
32
0.78
.383
*
= E
< 0.05; ** =
H
O
O
V
Pi
***
“ E <
0.001.
174

Table 5-3. Mean Age (in Days + S.E.) when 50% or More of Offspring Opened Eyes
(Experiment 3).
Condi-: Together-Alone Alone-Together Analysis of Variance Analysis of Covariance
tion
Litter: First Second First Second Factor df F p df F p
Pine Voles: (Covariate: Number of Offspring Born)
M =
11.06
10.65
11.37
11.31
Condition
1,
31
2.17
. 149
1,
30
2.15
. 152
(.30)
(.23)
(.31)
(.31)
Litter
1,
31
0.97
.331
1,
30
1.09
.304
N =
17
17
16
16
Interaction
1,
31
0.52
. 472
1,
30
0.62
.434
Prairie
Voles:
M =
8.37
8.81
8.89
8.84
Condition
1,
33
1.65
.206
1,
32
0.82
. 370
( -24)
( -21)
( .10)
( -14)
Litter
1,
33
2.51
. 122
1,
32
1.42
. 241
N =
16
16
19
19
Interaction
1,
33
4.07
. 051
1,
32
3.63
. 065
Meadow Voles:
M =
8.40
8.13
8.41
7.94
Condition
1,
30
1.63
. 688
1,
29
0.15
. 699
( -23)
( -13)
( .19)
( -16)
Litter
1,
30
7.56
.009**
1,
29
7.82
.009**
N =
15
15
17
17
Interaction
1,
30
0.57
.452
1,
29
0.44
.510
Montane
Voles:
M =
10.77
10.77
10.50
10.79
Condition
1,
25
0.22
. 640
1,
24
0.29
.592
( .20)
( .23)
( .23)
( -19)
Litter
1,
25
1.06
.310
1,
24
0.003
.954
N =
13
13
14
14
Interaction
1,
25
1.06
.310
1,
24
0.98
.331
* = p < 0.05; ** = p < 0.01; *** = p < 0.001.
175

Table 5-4
Mean Number of Offspring (+ S.E.) Weaned in Litters (Experiment 3)
Condi-: Together-Alone Alone-Together
tion
Litter: First Second First Second
Analysis of Variance Analysis of Covariance
Factor df F
E df F p
Pine Voles: (Covariate: Number of Offspring Born)
M =
1.37
1.71
1.35
1.26
Condition
1,
45
2.51
. 119
1,
44
0.27
. 603
( -14)
( .18)
( -16)
( -14)
Litter
1,
45
0.52
.472
1,
44
0.008
. 925
N =
24
24
23
23
Interaction
1,
45
1.52
. 223
1,
44
0.27
. 603
Prairie
Voles:
M =
2.83
3.28
3.23
3.65
Condition
1,
33
0.94
.337
1,
32
1.01
.321
( -34)
( -40)
( -29)
( -34)
Litter
1,
33
2.19
. 148
1,
32
1.06
.310
N =
18
18
17
17
Interaction
1,
33
0.003
.955
1,
32
1.32
.258
Meadow
Voles:
M =
3.24
3.59
3.48
3.11
Condition
1,
54
0.09
.759
1,
53
0.37
. 541
( -30)
( -44)
( -33)
( -37)
Litter
1,
54
0.001
. 970
1,
53
1.14
.290
N =
29
29
27
27
Interaction
1,
54
1.07
. 305
1,
53
0.16
. 688
Montane
Voles:
M =
3.20
4.27
3.94
4.47
Condition
1,
30
0.98
.327
1,
29
0.05
. 821
( -40)
( -38)
( -42)
( -51)
Litter
1,
30
3.87
. 058
1,
29
0.21
. 645
N =
15
15
17
17
Interaction
1,
30
0.43
. 512
1,
29
2.09
. 158
* = P
< 0.05;
* * = P
< 0.01;
* * * = p
< 0.001.
176

Table 5-5. Mean Sex Ratio of Offspring (+ S.E.) Weaned in Litters (Experiment 3).
Condi-: Together-Alone Alone-Together Analysis of Variance Analysis of Covariance
tion
Litter: First Second First Second Factor df F p df F p
Pine Voles: (Covariate: Number of Offspring Weaned)
M =
66.66
50.00
68.18
77.27
Condition
1,
24
1.20
. 282
1,
23
1.47
.237
(11.61)(11.03)
(12.20)(10.36)
Litter
1,
24
0.14
.705
1,
23
0.31
. 577
N =
15
15
11
11
Interaction
1,
24
1.69
.205
1,
23
1.96
. 174
Prairie
Voles:
M =
44.78
56.02
44.48
47.29
Condition
1,
29
0.40
. 530
1,
28
0.45
. 503
(8.75)
(6.41)
(7.42)
(8.09)
Litter
1,
29
0.73
. 399
1,
28
0.14
. 708
N =
15
15
16
16
Interaction
1,
29
0.26
.611
1,
28
0.12
. 730
Meadow
Voles:
M =
56.04
52.08
58.69
58.19
Condition
1,
45
0.44
. 509
1,
44
0.42
. 516
(6.11)
(6.50)
(5.09)
(5.59)
Litter
1,
45
0.19
. 660
1,
44
0.07
.782
N =
24
24
23
23
Interaction
1,
45
0.11
. 733
1,
44
0.01
.906
Montane
Voles:
M =
55.12
68.43
44.23
50.14
Condition
1,
26
4.73
. 038*
1,
25
4.38
. 046
(6.99)
(5.20)
(8.50)
(5.53)
Litter
1,
26
2.07
. 161
1,
25
1.20
.283
N =
14
14
14
14
Interaction
1,
26
0.30
. 582
1,
25
0.32
. 572
*
* *
= E
< 0.05;
= E
< 0.01;
= E
< 0.001.
177

Table 5-6. Mean Individual Body Weight of Offspring (+ S.E.) Weaned (Experiment 3).
Analysis of Variance Analysis of Covariance
Factor df F p df F p
Condi-: Together-Alone Alone-Together
tion
Litter: First Second First Second
Pine Voles: (Covariate: Number of Offspring Weaned)
M =
11.85
13.68
13.05
12.38
Condition
1,
26
0.003
.956
1,
25
0.11
.738
( -65)
( -63)
( -79)
( .85)
Litter
1,
26
1.48
.234
1,
25
2.95
. 097
N =
16
16
12
12
Interaction
1,
26
6.89
. 014*
1,
25
9.22
. 005
Prairie
Voles:
M =
18.96
19.37
18.41
18.79
Condition
1,
30
0.43
.516
1,
29
0.26
.613
( -85)
( .55)
( -64)
( .71)
Litter
1,
30
0.67
.416
1,
29
6.64
. 015
N =
16
16
16
16
Interaction
1,
30
0.001
.966
1,
29
0.48
.491
Meadow
Voles:
M =
22.31
21.62
21.62
22.03
Condition
1,
45
0.02
.879
1,
44
0.01
.920
( .83)
( -57)
( -83)
( -71)
Litter
1,
45
0.08
.772
1,
44
0.005
.942
N =
24
24
23
23
Interaction
1,
45
1.34
.252
1,
44
0.71
.400
Montane
Voles:
M =
15.28
16.86
15.49
15.34
Condition
1,
26
0.91
.346
1,
25
1.25
.273
( -61)
( -67)
( -64)
( -51)
Litter
1,
26
1.79
. 192
1,
25
5.19
.031
N =
14
14
14
14
Interaction
1,
26
2.61
.118
X,
25
2.45
.129
*
= R
< 0.05;
* *
= 32
< 0.01;
** *
= R
< 0.001.
178

179
â–¡ Mala Absent
0 Male Present
Birth erf L1
L1 Removed
i
CTogrtfwr-Alone)
â–  (Aiona-Toqother)
Day 0
7
14
21
26
35 42
>
Figure 5-1. Graphical representation of the order and time
of male presence and absence during Experiment 3. Together-
Alone: males present during the rearing of the first litter;
Alone-Together: males absent during the rearing of the first
litter. LI refers to first litter. Cross-hatched area
represents variable amount of time before males were
removed; males were removed one day following the birth of
the second litter (see text for further explanation of
conditions).

180
l'Jumoe1' ‘ DovS 'fC’in Birth of First to do ond Litro;
Figure 5-2. Frequency distribution of the number of
breeding pairs of pine voles that produced the second litter
within a given number of days from the birth of the first
litter. Alone-together: males absent during the rearing of
the first litter; Together-Alone: males present during the
rearing of the first litter (see text for further
explanation of conditions).

181
Figure 5-3. Mean sex ratio (expressed as percentages) of
number of male to female offspring weaned among breeding
pairs of montane voles (+ standard error) as a function of
the litter number. Values above 50% represent male-biased
litters, those below 50% as female-biased. No significant
differences were found within each litter (see text for
explanation of conditions).

182
First Litter Second Utter
Figure 5-4. Mean individual offspring weights of pine voles
at weaning (in g + standard error) as a function of male
presence and litter number (see text for explanation of
conditions). Offspring produced by pairs in the Together-
Alone condition weighed more in the second litter than the
first. No significant differences were found between the
conditions within each litter.

CHAPTER 6
GENERAL OVERVIEW AND DISCUSSION
In this final chapter I attempt to synthesize the
results of the present studies, along with results from
other studies, to form some possible scenarios concerning
the functional and proximate means by which social and less
social mating systems form among Microtus. Finally, I
present an outline for routes of future study.
Overview of Experimental Results
Together the results of the present series of studies
provide only limited support for the proposal that studies
of puberty modulation and paternal influence may help to
elucidate the evolution and maintenance of social and mating
systems among species of Microtus. However, I suggest that
the results of previous studies, with various species of
Microtus. suggest the need for additional investigation into
the relationships among puberty modulation and the formation
of social and mating systems.
An Attempt at a Synthesis
No grand unifying theory is yet available that
integrates the effects of puberty modulation and paternal
behavior with differences in mating system among Microtus.
Wittenberger (1979) suggested that no simple hypothesis or
theory could be expected to explain the diversity of animal
mating systems, although he believed a reasonably
183

184
comprehensive body of theory could be devised from an
integration of relatively few general principles. Below I
sketch out the key findings of the effects of puberty
modulation and paternal behavior, from the present studies
and others, that suggest relationships among these
developmental processes and the formation of different
social and mating systems.
Possible Relationships between Puberty Modulation. Paternal
Behavior, and Social and Mating Systems
There is much evidence that shows puberty modulation
occurs in house mice and several species of voles, although
it has been difficult to compare meaningfully reported
differences among species to determine if species
differences exist and whether they vary systematically with
particular ecological factors or social traits and expressed
mating systems (Dewsbury, 1981). The lack of common
measures and differences between laboratories have been
major difficulties for this comparative analysis.
Summaries of the key social characteristics of house
mice and the four species of voles that have been documented
in response to social or pheromonal cues are shown in Tables
6-1 (females) and 6-2 (males). Although the methods and
results of each of the studies often differ in some
respects, at least two trends seem apparent. First, within
the genus Microbus, evidence of puberty delay among females,
and possibly males, is associated with a high degree of
paternal care and a monogamous mating system. Puberty delay

185
being defined as the delay of some marker of puberty, or
smaller weight of reproductive organs, compared to those in
a control condition. A second trend is that puberty
acceleration appears common among the females of all species
of Microtus, regardless of the pattern of parental care and
mating system of the species (puberty acceleration has been
treated here as opposite to those differences listed above
for puberty delay). Below I clarify each of these
relationships and discuss possible ramifications of them.
Puberty delay associated with paternal care and a
monogamous mating system. The relationship among sexual
suppression, parental care, and a monogamous mating system
is not a novel finding. Kleiman (1977) described these
relationships among several mammalian species that formed a
monogamous mating system. Why the relationship exists for
at least two species of Microtus. pine voles and prairie
voles, appears to be the most difficult question to answer.
There may be several reasons which predispose various
species to express these characteristics.
One of the preconditions that may favor the evolution
of these traits is when group living leads to the
acquisition of limited resources necessary for survival,
reproduction, or both (Kleiman, 1977). The large expansive
habitats of the great plains of central North America,
without the pressure for long dispersal, may have selected
for increased abilities to compete with others (Christian
1970). Among prairie voles, both sexes are territorial,

186
thus historically, breeding pairs may have had greater
success defending areas in pairs or in small extended family
groups. If the habitat became saturated, the evolution of
puberty delay may have enabled offspring to remain within
the family territory until dispersal, and may have
functioned as a means of incest avoidance (Carter & Getz,
1985) .
There are suggestions in the literature that offspring
that are sexually suppressed, remain within a family group,
and are exposed to an adult male, accrue benefits that
enable them to become more reproductively successful than
others that have not had the same experiences. Wang and
Novak (1992) suggested two factors that may lead to the
formation of extended families of prairie voles. First,
offspring that help with caretaking may cause increases in
the survival and quality of the young and thereby increase
their inclusive fitness (Hamilton, 1964). Second, the
offspring that help rear siblings may gain valuable
experience, such as parental care, that may enable them to
become successful breeders. Salo and French (1989) provided
evidence in Mongolian gerbils that juveniles that were
exposed to younger siblings, while in an extended family
group, became more successful at rearing their own offspring
compared to juveniles not exposed to young siblings.
Specifically, the experience seemed to be most beneficial
for male gerbils that had early exposure to younger
siblings, rather than females.

187
\
There is other indirect evidence that experience gained
within a social family group leads to enhanced reproductive
success. Jakubowski and Terkel (1982) found that when wild
house mice remained m a family group, along with their
parents and a subseguent litter, the males later displayed
paternal behavior and not typical pup-killing behavior.
Similar behaviors have been shown in rats. Rosenblatt
(1967) found that male laboratory rats exposed to young for
several days began to display paternal responsiveness.
Thus, the results of these two studies suggest that the
propensity to display paternal behavior can occur as a
result of exposure to a family group with young offspring.
Such exposure and behavior are likely to occur in species
such as prairie voles and pine voles (Getz & Carter, 1980;
Schadler, 1990; Wang & Novak, 1992).
It is possible that female mate choice may operate
among species that are highly social with paternal care.
Theoretically, females may increase their level of fitness
by selecting a mate with parental competence (Vehrencamp &
Bradbury, 1984). Although studies of mate choice for
parental competence have not been conducted in voles, they
have been done for assessing dominance (Shapiro & Dewsbury,
1986), familiarity (Newman & Halpin, 1982; Shapiro et al.
1986) and recency of past mating (Pierce & Dewsbury, 1991).
The results of all these studies have indicated that female
prairie voles preferred males that were more dominant,
familiar, or recently unmated. In contrast, female montane

188
voles did not display any systematic preference for males
that differed in these qualities. Thus, it seems plausible
that female prairie voles may prefer males with good
care-taking capacities.
It seems plausible that females of some species of
Microtus could indirectly choose mates with good care-taking
qualities. Because female prairie voles and pine voles
typically require longer durations of male exposure to
become sexually receptive than do montane voles and meadow
voles, males may have to remain in close proximity to a
single female in order to mate successfully with them
(Taylor et al. in press). This pattern of the male
remaining within the proximity of one or few females might
predispose males to behave paternally toward their progeny.
This sequence of events might lead to the formation of a
monogamous mating system.
In a related issue, the retention of pregnancies also
appears dependent on continued male presence for some
species of voles, including prairie voles (Richmond & Stehn,
1976) and montane voles (Berger & Negus, 1982; Taylor,
1990/1991). In Experiment 3, female pairmates of male pine
voles that were removed after the birth of the first litter,
produced a second litter significantly later than the female
pairmates of those that remained in the female's presence.
Thus, indirectly, the continued presence of a male may favor
the formation of a monogamous mating system and serve as a
means to predispose them to help with the rearing of the

189
offspring. It seems plausible that the evolution of sexual
suppression could occur from the continued presence of the
breeding male and possibly act as a means of incest
avoidance among family members (Carter & Getz, 1985).
If females chose males with good caretaking abilities,
and also that remain nearby to ensure estrus and pregnancy
maintenance, one can envision a process where the same
behaviors are expressed by their offspring. This process,
in turn, could lead to the evolution of larger scale
differences we see between species. Certainly these
scenarios are only speculations that need additional
support, but they are plausible routes of evolution for some
of the differences in social and mating patterns we see
today among Microtus.
Puberty acceleration appears common to all mating
systems (Microtus). Puberty acceleration appears to be a
ubiquitous phenomenon among the females of all four species
of Microtus. but has not been demonstrated among the males
(Table 6-1 and 6-2). However, it should not be ruled out
that puberty acceleration does not occur in males of some
species of voles. In Experiment 1, the body weights of male
pine voles receiving soiled bedding from the family group or
from adult males were heavier than those receiving clean
bedding or bedding from adult females. However, it is not
clear if changes in body weight alone should be considered a
form of sexual suppression. More precise measures of
reproductive activity among male Microtus may reveal cases

of acceleration (e.g., Rissman et al. 1984). Additional
investigation of puberty acceleration and suppression among
male Microtus should continue.
190
The apparent capacity for puberty acceleration among
the females of many species of Microtus may have assisted in
their large distribution over much of the Old and New Worlds
(Tamarin, 1985). A similar capacity among house mice may
have aided in their colonization. Bronson (1979) suggested
that the near-global distribution of house mice was likely
to have been aided substantially by their pheromonal cueing
system. Certainly, more critical research is needed, with
various species of Microtus. before we are able to form a
greater understanding of how pheromones, puberty modulation,
and the development of paternal behavior may intertwine.
The great diversity of different social and mating patterns
among Microtus offers a fertile resource for continued
investigation.
Routes of Future Study
Together the results of these studies suggest a number
of routes for future study, although with methodological
precautions; these are listed below.
(1) The study of a few additional species might provide
greater insight into the relationship between puberty
modulation and social expression. For example, additional
information gathered for field voles (M. aorestis) would
seem useful because of the large amount of research
previously done on the reproductive physiology of this

191
species (see Sawrey & Dewsbury, 1985). Information gathered
on taiga voles (M. xanthognathus) may also prove valuable
for comparative study. Taiga voles are considered
polygynous, but unlike other species of Microtus. only the
males appear to be territorial while the females are not
(Wolff, 1980).
(2) Future research should be designed to identify the
specific source, effective compounds, and normal means of
transfer that cause changes in reproductive maturation. For
example, one of the unique differences among the four
species studied is the presence of hip-glands on montane
voles. There is evidence that chemical cues associated with
these glands are behaviorally attractive to adult males,
thus they could cause changes in reproductive physiology as
well (Jannett, 1978) .
(3) Care must be taken to selectively control
environmental stimuli in experiments that are sometimes not
controlled in studies of puberty modulation. Several
reports indicate significant changes in behavioral and
reproductive activity as a function of changes in the
photoperiod and light intensity. For example, Geyer and
Rogers (1979) found that the rate of litter production of
pine voles exposed to high intensity light (75-200 lumens)
was nearly twice the rate under low light intensity (0-75
lumens). Similarly, Ferkin and Zucker (1991) have shown
that during the spring-summer breeding season, female meadow
voles prefer odors of males over those of females. However,

192
in the autumn-winter season of reproductive quiescence, this
preference is reversed.
It is only through the selected control of stimuli such
as these that we will be able to develop a clearer
understanding of how olfactory stimuli affect reproductive
processes, including behavior. Much experimentation,
ideally interlaced with carefully designed field
observations, remains to be done to clarify the relevance of
puberty modulation to the social and mating strategies of
Microtus.

193
Table 6-1. Comparison Table of Key Characteristics
Among Female Muroid Rodents.
Species
House
Mice
Pine
Voles
Prairie
Voles
Meadow
Voles
Montane
Voles
Social Deme
System Territory
Social'*'
Monogamy
Social2
Monogamy
Social3
Promiscuous
Asocial4
Polygynous
Asocial5
Territor¬
iality?
Males1
Males,2
Females
Males,3
Females
Females4
Males,5
Females
Extensive
Paternal
Care?
No1
Yes7
Yes7
No7
No9
Puberty
Delay?
Yes15
Yes17'18 Yes8'19
No?8'10
Yes?11,24
Puberty
Acceler-
tion via
Male Cues
Yes13
?
Yes18
Yes6
Yes12
Yes16
Puberty
Acceler-
tion via Yes14
Reproductively
Active Females?
No? 23
(Delay)
NO?19
(Delay)
?
9
Puberty
Acceler-
• . 1 5
tion via Yes
Females in
Estrus?
?
No?20
(Delay)
?
9
1 Bronson, (1979); 2 FitzGerald & Madison, (1983); 3 Getz &
Hofmann (1986); 4 Madison (1980); 5 Jannett (1980); 6 Carter
et al. (1980); 7 Oliveras & Novak, (1986); 8 Batzli et al.
(1977); 9 McGuire & Novak (1986); 10 Pasley & McKinney
(1973); 11 Jannett (1978); 12 Baddaloo & Clulow (1981); 13
Colby & Vandenbergh (1974); 14 Drickamer & Hoover (1979); 15
Drickamer (1982) ; 16 Sawrey & Dewsbury (1991) ; 17 Schadler
(1983); 18 Lepri & Vandenbergh, J. G. (1986); 19 Getz et al.
(1983); 20 Carter & Getz (1985); 21 McKinney & Desjardins
(1973); 22 Vandenbergh (1971); 23 Schadler (1990); 24 This
study (Experiment 1).

194
Table 6-2. Comparison Table of Key Characteristics Among
Male Muroid Rodents.
Species
House
Mice
Pine
Voles
Prairie
Voles
Meadow
Voles
Montane
Voles
Social Deme
System Territory
Social1
Monogamy
Social2
Monogamy
Social3
Promiscuous
Asocial4
Polygynous
Asocial5
Territor¬
iality?
Males1
Males,2
Females
Males,3
Females
Females4
Males,5
Females
Extensive
Paternal
Care?
No1
Yes7
Yes7
No7
No9
Puberty
Delay?
Yes21'
22 Yes?23
Yes?8
No?8
No?11
Puberty
Acceler- Yes22
tion via
Female Cues?
NO?24
NO?24
No?24
No?24
Puberty
Acceler-
tion via ?
Reproductively
Active Females?
9
9
9
9
Puberty
Acceler-
tion via
9
9
9
9
9
Females in
Estrus?
1 Bronson, (1979); 2 FitzGerald & Madison, (1983); 3 Getz &
Hofmann (1986); 4 Madison (1980); 5 Jannett (1980); 6 Carter
et al. (1980); 7 Oliveras & Novak, (1986); 8 Batzli et al.
(1977); 9 McGuire & Novak (1986); 10 Pasley & McKinney
(1973); 11 Jannett (1978); 12 Baddaloo & Clulow (1981); 13
Colby & Vandenbergh (1974); 14 Drickamer & Hoover (1979); 15
Drickamer (1982) ; 16 Sawrey & Dewsbury (1991) ; 17 Schadler
(1983); 18 Lepri & Vandenbergh, J. G. (1986); 19 Getz et al.
(1983); 20 Carter & Getz (1985); 21 McKinney & Desjardins
(1973); 22 Vandenbergh (1971); 23 Schadler (1990); 24 This
study (Experiment 1).

REFERENCES
Baddaloo, E. G. Y., & Clulow, F. V. (1981). Effects of the
male on growth, sexual maturation, and ovulation of
young female meadow voles, Microtus nennsvlvanicus.
Canadian Journal of Zoology. 59, 415-421.
Barnett, S. A., & Coleman, E. M. (1959). The effect of low
environmental temperature on the reproductive cycle of
female mice. Journal of Endocrinology. 19, 232-265.
Batzli, G. 0., Getz, L. L., & Hurley, S. S. (1977).
Suppression of growth and reproduction of microtine
rodents by social factors. Journal of Mammalogy. 58,
583-591.
Berger, P. J., & Negus, N. C. (1982). Stud male maintenance
of pregnancy in Microtus montanus. Journal of
Mammalogy. 63. 148-151.
Boyer, M. L., Jemiolo, B., Andreolini, F., Wiesler, D., and
Novotny, M. (1989). Urinary volatile profiles of pine
voles, Microtus pinetorum. and their endocrine
dependency. Journal of Chemical Ecology. 15, 649-662.
Bronson, F. H. (1979). The reproductive ecology of the house
mouse. Quarterly Review of Biology. 54. 265-299.
Bronson, F. H., & Coquelin, A. (1980). The modulation of
reproduction by priming pheromone in house mice:
speculations on adaptive function (pp. 243-265). In D.
Muller-Schwarze & R. M. Silverstein (Eds.), Chemical
signals: Vertebrates and aguatic invertebrates. New
York: Plenum Press.
Bronson, F. H., & Maruniak, J. A. (1975). Male-induced
puberty in female mice: Evidence for a synergistic
action of social cues. Biology of Reproduction. 13.
94-98.
Bronson, F. H., & Rissman, E. F. (1986). The biology of
puberty. Biological Reviews. 61, 157-195.
Bronson, F. H., & Stetson, M. H. (1973). Gonadotropin
release in prepubertal female mice following exposure:
a comparison with the adult cycle. Biology of
Reproduction. 9, 449-459.
195

196
Carter, C. S., & Getz, L. L. (1985). Social and hormonal
determinants of reproductive patterns in the prairie
vole. In R. Gilíes & J. Balthazart (Eds.),
Neurobioloqy: Current comparative approaches (pp.
18-36). New York: Springer-Verlag.
Carter, C. S., Getz, L. L., & Cohen-Parsons, M. (1986).
Relationships between social organization and
behavioral endocrinology in a monogamous mammal. In
Advances in the study of behavior. Vol. 16 (pp.
109-145). Orlando, FL: Academic Press.
Carter, C. S., Getz, L. L., Gavish, L., McDermott, J. L., &
Arnold, P. (1980). Male-related pheromones and the
activation of female reproduction in the prairie vole
(Microtus ochroaaster). Biology of Reproduction. 23.
1038-1045.
Carter, C. S., Witt, D. M., Schneider, J., Harris, Z. L., &
Volkening, D. (1987). Male stimuli are necessary for
female sexual behavior and uterine growth in prairie
voles CMicrotus ochrogaster). Hormones and Behavior.
21, 74-82.
Christian, J. J. (1970). Social subordination, population
density, and mammalian evolution. Science. 168. 84-90.
Christian, J.J. (1975). Hormonal control of population
growth. In B. E. Eleftheriou & R. L. Sprott (Eds.),
Hormonal correlates of behavior (pp. 205-274). New
York: Plenum Publishing Corporation.
Clark, M. M., & Galef, B. G., Jr. (1990). Sexual segregation
in the left and right horns of the gerbil uterus: "the
male embryo is usually on the right, the female on the
left" (Hippocrates) . Developmental Psychobiology. 23.,
29-37.
Clutton-Brock, T. H. (1991). The evolution of parental care.
Princeton, NJ: Princeton University Press.
Clutton-Brock, T. H., & Harvey, P. H. (1984). Comparative
approaches to investigating adaptation. In J. R. Krebs
& N. B. Davies (Eds.), Behavioural ecology: An
evolutionary approach (pp. 7-29). Sunderland, MA:
Sinauer Press.
Colby, D. R., & Vandenbergh, J. G. (1974). Regulatory
effects of urinary pheromones on puberty in the mouse.
Biology of Reproduction. 11, 268-279.
Cole, L. C. (1954). The population conseguences of
life-history phenomena. Quarterly Review of Biology.
29, 103-137.

197
Coppola, D. M., & O'Connell, R. J. (1988). Behavioral
responses of peripubertal female mice towards
puberty-accelerating and puberty-delaying chemical
signals. Chemical Senses. 13. 407-424.
Cranford, J. A., & Derting, T. L. (1983). Intra and
interspecific behavior of Microtus pennsvlvanicus and
Microtus pinetorum. Behavioral Ecology and
Sociobiologv. 13. 7-11.
Dewsbury, D. A. (1981). An exercise in the prediction of
monogamy in the field from laboratory data on 42
species of muroid rodents. The Biologist. 63. 138-162.
Dewsbury, D. A. (1985). Paternal behavior in rodents.
American Zoologist. 25. 841-852.
Dewsbury, D. A. (1987). The comparative psychology of
monogamy, The Nebraska Symposium on Motivation. 35,
1-50.
Dewsbury, D. A. (1988, June). Male role and related
processes in the development of deer mice (Peromvscus
maniculatus bairdi'i . Paper presented at the 68th Annual
Meeting of the American Society of Mammalogists.
Clemson University, Clemson, South Carolina.
Dewsbury, D. A. (1990). Individual attributes generate
contrasting degrees of sociality in voles. In R. H.
Tamarin (Ed.), Social systems and population cycles in
voles (pp. 1-9). Basel, Switzerland: Birkhauser Verlag.
Dewsbury, D. A., Baumgardner, D. J., Evans, R. L., &
Webster, D. G. (1980). Sexual dimorphism for body mass
in 13 taxa of muroid rodents under laboratory
condition. Journal of Mammalogy. 61, 146-149.
Drickamer, L. C. (1974). Contact stimulation, androgenized
females and accelerated sexual maturation in female
mice. Behavioral Biology. 12. 101-110.
Drickamer, L. C. (1975a). Daylength and sexual maturation in
female house mice. Developmental Psychobiology. 8,
561-570.
Drickamer, L. C. (1975b). Contact stimulation and
accelerated sexual maturation of female mice,
Behavioral Biology. 15. 113-115.
Drickamer, L. C. (1977). Delay of sexual maturation in
female house mice by exposure to grouped females or
urine from grouped females. Journal of Reproduction and
Fertility. 51. 77-81.

198
Drickamer, L. C. (1981). Acceleration and delay of sexual
maturation in female house mice previously selected for
early and late first vaginal oestrus. Journal of
Reproduction and Fertility. 63. 325-329.
Drickamer, L. C. (1982a). Acceleration and delay of sexual
maturation in female mice via chemosignals: circadian
rhythm effects. Biology of Reproduction. 27, 596-601.
Drickamer, L. C. (1982b). Acceleration and delay of first
vaginal oestrus in female mice by urinary chemosignals:
dose levels and mixing urine treatment sources. Animal
Behaviour. 30. 456-460.
Drickamer, L. C. (1982c). Delay and acceleration of puberty
in female mice by urinary chemosignals from other
females. Developmental Psychobiology. 15. 433-445.
Drickamer, L. C. (1983a). Male acceleration of puberty in
female mice (Mus musculus). Journal of Comparative
Psychologyâ–  97. 191-200.
Drickamer, L. C. (1983b). Chemosignal effects on puberty in
young female mice: Urine from pregnant and lactating
females. Developmental Psychobiology, 16, 207-217.
Drickamer, L. C. (1983c). Effect of period of grouping of
donors and duration of stimulus exposure on delay of
puberty in female mice by a urinary chemosignal from
grouped females. Journal of Reproduction and Fertility.
69, 723-727.
Drickamer, L. C. (1984a). Effects of very small doses of
urine on acceleration and delay of sexual maturation in
female house mice. Journal of Reproduction and
Fertility. 72, 475-477.
Drickamer, L. C. (1984b). Acceleration of puberty in female
mice by a urinary chemosignal from pregnant or
lactating females: Timing and duration of stimulation.
Developmental Psychobiology. 17. 451-455.
Drickamer, L. C. (1984c). Urinary chemosignals from mice
(Mus musculus): Acceleration and delay of puberty in
related and unrelated young females. Journal of
Comparative Psychology, 98, 414-420.
Drickamer, L. C. (1986). Puberty-influencing chemosignals in
house mice: Ecological and evolutionary considerations.
In D. Duvall, D. Muller-Schwarze, & R. M. Silverstein
(Eds.), Chemical signals in vertebrates, Vol 4, (pp.
441-455). Plenum Publishing.

199
Drickamer, L. C. (1988). Long-term effects of accelerated or
delayed sexual maturation on reproductive output in
wild female house mice (Mus musculus). Journal of
Reproduction and Fertility. 83. 439-445.
Drickamer, L. C. (1989a). Patterns of deposition of urine
containing chemosignals that affect puberty and
reproduction by wild stock male and female house mice
(Mus domesticus) . Journal of Chemical Ecology, 15,
1407-1421.
Drickamer, L. C. (1989b). Odor preferences of wild stock
female house mice (Mus domesticus) tested at three ages
using urine and other cues from conspecific males and
females. Journal of Chemical Ecology. 15, 1971-1987.
Drickamer, L. C., & Assmann, S. M. (1981). Acceleration and
delay of puberty in female house mice: Methods of
delivery of the urinary stimulus. Developmental
Psychobiology. 14. 487-497.
Drickamer, L. C., & Hoover, J. E. (1979). Effects of urine
from pregnant and lactating female house mice on sexual
maturation of juvenile females. Developmental
Psychobiology. 12. 545-551.
Emlen, S. T. (1984). Cooperative breeding in birds and
mammals. In J. R. Krebs & N. B. Davies (Eds.),
Behavioural ecology: An evolutionary approach (pp.
305-399). Sunderland, MA: Sinauer Press.
Emlen, S. T., & Oring, L. W. (1977). Ecology, sexual
selection and the evolution of mating systems. Science.
197 . 215-223.
Facemire, C. F., & Batzli, G. 0. (1983). Suppression of
growth and reproduction by social factors in microtine
rodents: Tests of two hypotheses. Journal of Mammalogy.
64, 152-156.
Ferkin, M. H., & Gorman, M. R. (1992). Photoperiod and
gonadal hormones influence odor preferences of the male
meadow vole, Microtus pennsvlvanicus. Physiology &
Behavior. 51. 1087-1091.
Ferkin, M. H., & Seamon, J. O. (1987). Odor preferences and
social behavior in meadow voles, Microtus
pennsvlvanicus: Seasonal differences. Canadian Journal
of Zoology. 65. 2931-2937.
Ferkin, M. H., & Zucker, I. (1991). Seasonal control of
odour preferences of meadow voles (Microtus
pennsvlvanicus) by photoperiod and ovarian hormones.
Journal of Reproduction and Fertility. 92. 433-441.

200
FitzGerald, R. W., & Madison, D. M. (1983). Social
organization of a free-ranging population of pine
voles, Microtus pinetorum. Behavioral Ecology &
Sociobiology. 13. 183-187.
Gadgil, M., & Bossert, W. (1970). Life history consequences
of natural selection, American Naturalist. 104. 1-24.
Gaulin, S. J., & FitzGerald, R. W. (1988). Home-range size
as a predictor of mating systems in Microtus. Journal
of Mammalogy. 69, 311-319.
Geist, V. (1971). Mountain sheep: A study in behavior and
evolution. Chicago IL: University of Chicago Press.
Getz, L. L., & Carter, C. S. (1980). Social organization in
Microtus ochroqaster populations. The Biologist, 60,
134-147.
Getz, L. L., Dluzen, D., & McDermott, J. L. (1983).
Suppression of reproductive maturation in
male-stimulated virgin female Microtus by a female
urinary chemosignal. Behavioural Processes. 8, 59-64.
Getz, L. L., & Hofmann, J. E. (1986). Social organization in
free-living prairie voles, Microtus ochroqaster.
Behavioral Ecology & Sociobiology. 18. 275-282.
Getz, L. L., Larson, C. M., & Lindstrom, K. A. (1992).
Blarina Brevicauda as a predator on nestling voles.
Journal of Mammalogy. 73., 591-596.
Getz, L. L., McGuire, B., Hofmann, J., Pizzuto, T., & Frase,
B. (1990). Social organization and mating system of the
prairie vole, Microtus ochroqaster. In R. H. Tamarin
(Ed.), Social systems and population cycles in voles
(pp. 69-80). Basel, Switzerland: Birkhauser Verlag.
Geyer, L. A., & Rogers, J. G. Jr. (1979). The influence of
light intensity on reproduction in pine voles, Microtus
pinetorum. Journal of Mammalogy. 60, 839-841.
Gray, G. D., & Dewsbury, D. A. (1973). A quantitative
description of copulatory behavior in prairie voles
(Microtus ochroqaster). Brain, Behavior, and Evolution.
43. 351-358.
Hamilton, W. D. (1964). The evolution of social behavior.
Journal of Theoretical Biology. 7, 1-52.
Hamilton, W. D. (1966). The moulding of senescence by
natural selection. Journal of Theoretical Biology. 12,
12-45.

201
Hartung, T. G. , & Dewsbury, D. A. (1979). Paternal behavior
in six species of muroid rodents. Behavioral and Neural
Biology. 26. 466-478.
Hasler, J. F. (1975). A review of reproduction and sexual
maturation in the microtine rodents. The Biologist. 57.
52-86.
Hasler, M. J., & Nalbandov, A. V. (1974). The effect of
Weanling and Adult males of sexual maturation in female
voles (Microtus ochrogaster). General and Comparative
Endocrinology. 23. 237-238.
Hofmann, J. E., & Getz, L. L. (1988). Multiple exposures to
adult males and reproductive activation of virgin
female Microtus ochrogaster. Behavioural Processes. 17,
57-61.
Jakubowski, M., & Terkel, J. (1982). Infanticide and
caretaking in nonlactating Mus musculus: Influence of
genotype, family group and sex. Animal Behaviour. 30.
1029-1035.
Jannett, F. J. Jr. (1978). The density-dependent formation
of extended maternal families of the montane vole,
Microtus montanus nanus. Behavioral Ecology &
Sociobiologv. 3., 245-263.
Jannett, F. J. Jr. (1980). Social dynamics of the montane
vole, Microtus montanus. as a paradigm. The Biologist.
62, 3-19.
Jannett, F. J. Jr. (1981). Sex ratios in high-density
population of the montane vole, Microtus montanus. and
the behavior of territorial males. Behavioral Ecology
and Sociobiologv. 8, 297-307.
Jannett, F. J. Jr. (1982). Nesting patterns of adult voles,
Microtus montanus. in field populations. Journal of
Mammalogy. 63, 495-498.
Jannett, F. J. Jr. (1984). Reproduction of the montane vole,
Microtus montanus. in subnivean populations. Special
Publication Carnegie Museum of Natural History. No. 10,
215-224 .
Keller, B. L. (1985). Reproductive patterns. In R. H.
Tamarin (Ed.), Biology of New World Microtus (pp.
725-778). Lawrence, KS: American Society of
Mammalogists.

202
Keller, B. L., & Krebs, C. J. (1970). Microtus population
biology. III. Reproductive changes in fluctuating
populations of M. ochroqaster and M. pennsvlvanicus in
southern Indiana. 1965-1967. Ecological Monographs. 40.
263-294.
King, J. A. (1970). Ecological psychology: An approach to
motivation. Nebraska Symposium on Motivation. 18, 1-33.
Lincoln: University of Nebraska Press.
Kirkpatrick, R. L., & Valentine, G. L. (1970). Reproduction
in captive pine voles, Microtus pinetorum. Journal of
Mammalogy. 51, 779-785.
Kleiman, D. G. (1977). Monogamy in mammals. Quarterly Review
of Biology. 52. 39-69.
Kleiman, D. G., & Malcolm, J. R. (1981). The evolution of
male parental investment in mammals. In D. J. Gubernick
and P. H. Klopfer (eds.), Parental care in mammals, pp.
347-387. New York: Plenum.
Kruczek, M., & Gruca, A. (1990). Seasonal variations in male
mice at the time of sexual maturation. Laboratory
Animals. 24. 36-39.
Lecyk, M. (1967). The influence of crowded population
stimuli on the development of reproductive organs in
the common vole. Acta Theriologica. 12, 177-179.
Lepri, J. J., & Vandenbergh, J. G. (1986). Puberty in pine
voles, Microtus pinetorum. and the influence of
chemosignals on female reproduction. Biology of
Reproduction. 34. 370-377.
Lepri, J. J., & Wysocki, C. J. (1987). Removal of the
vomeronasal organ disrupts the activation of
reproduction in female voles. Physiology St Behavior.
40. 349-355.
Levin, R. L., Si Johnston, R. E. (1986). Social mediation of
puberty: An adaptive female strategy. Behavioral and
Neural Biology. 46, 308-324.
Lewontin, R. C. (1965). Selection for colonizing ability. In
H. G. Baker S: G. L. Stebbins (Eds.) , The genetics of
colonizing species (pp. 79-94). New York: Academic
Press.
Lombardi, J. L. , Vandenbergh, J. G., St Whitsett, J. M.
(1976). Androgen control of the sexual maturation
pheromone in house mouse urine. Biology of
Reproductionâ–  15. 179-186.

203
MacArthur, R. H., & Wilson, E. 0. (1967). Theory of island
biogeography. Princeton, NJ: Princeton University
Press.
Mackintosh, J. H. (1970). Territory formation by laboratory
mice. Animal Behaviour. 18, 177-183.
Madison, D. M. (1980). An integrated view of the social
biology of Microtus pennsvlvanicus. The Biologist. 62,
20-33.
Madison, D. M. (1984). Group nesting and its ecological and
evolutionary significance in overwintering microtine
rodents. In: Special Publication Carnegie Museum of
Natural History. No. 10. 267-274.
Massey, A., & Vandenbergh, J. G. (1980). Puberty delay by a
urinary cue from female house mice in feral
populations. Science. 209â–  821-822.
Massey, A., & Vandenbergh, J. G. (1981). Puberty
acceleration by a urinary cue from male mice in feral
populations. Biology of Reproduction. 24. 523-527.
McClintock, M. (1983). Modulation of the estrous cycle by
pheromone from pregnant and lactating rats. Biology of
Reproduction. 28., 823-829.
McClintock, M. (1984). Estrous synchrony: Modulation of
ovarian cycle length by female pheromones. Physiology
and Behavior. 32., 701-705.
McGuire, B., & Novak, M. (1984). A comparison of maternal
behaviour in the meadow vole (Microtus pennsvlvanicus),
prairie vole (M. ochrogasterl and pine vole
(M. pinetorum). Animal Behaviour. 32. 1132-1141.
McGuire, B., & Novak, M. (1986). Parental care and its
relationship to social organization in the montane vole
(M. montanus^. Journal of Mammalogy. 67, 305-312.
McGuire, B., Russell, K. D., Mahoney, T., & Novak, M.
(1992). The effects of mate removal on pregnancy
success in prairie voles (Microtus ochrogaster) and
Meadow voles (Microtus pennsvlvanicus). Biology of
Reproduction. 47. 37-42.
McGuire, M. R., & Getz, L. L. (1981). Incest taboo between
sibling Microtus ochrogaster. Journal of Mammalogy. 62,
213-215.

204
Mertz, D. B. (1971). Life history phenomena in increasing
and decreasing populations. In G. P. Patil, E. C.,
Pielou, & W. E. Waters (Eds.), Statistical ecology (pp.
361-399). University Park, PA: Pennsylvania State
University Press.
Mullen, D. A. (1960). Adrenal weight changes in Microtus.
Journal of Mammalogy. 41. 129-130.
Muller, G. B. (1990). Developmental mechanisms at the origin
of morphological novelty: A side-effect hypothesis. In
M. H. Nitecki (Ed.), Evolutionary innovations (pp.
99-130). Chicago, IL: The University of Chicago Press.
Nadeau, J. H. (1985). Ontogeny. In R. H. Tamarin (Ed.),
Biology of new world Microtus (pp. 254-255). Lawrence,
KS: American Society of Mammalogists.
Negus, N. C., & Pinter, A. J. (1965). Litter sizes of
Microtus montanus in the laboratory. Journal of
Mammalogy. 46. 434-437.
Newman, K. S., & Halpin, Z. T. (1982). Individual odours and
mate recognition in the prairie vole, Microtus
ochroqaster. Animal Behaviour. 36. 1779-1787.
Novak, M. A., & Getz, L. L. (1969). Aggressive behavior of
meadow voles and pine voles. Journal of Mammalogy. 50,
637-639.
Oliveras, D., & Novak, M. (1986). A comparison of paternal
behaviour in the meadow vole Microtus pennsvlvanicus.
the pine vole M. pinetorum and the prairie vole
M. ochroqaster. Animal Behaviour. 34. 519-526.
Olsen, D. E., & Seabloom, R. W. (1973). Adrenocortical
response to captivity in Microtus pennsvlvanicus.
Journal of Mammalogy. 54, 779-781.
Orians, G. H. (1969). On the evolution of mating systems in
birds and mammals. American Naturalist. 103. 589-603.
Pasley, J. N., & Christian, J. J. (1971). Effects of ACTH on
voles (Microtus pennsvlvanicus) related to reproductive
function and renal disease. Proceedings of the Society
for Experimental and Biological Medicine. 137â–  268-272.
Pasley, J. N., & McKinney, T. D. (1973). Grouping and
ovulation in Microtus pennsvlvanicus. Journal of
Reproduction and Fertility. 34. 527-530.
Pearson, O. P. (1985). Predation. In R. H. Tamarin (Ed.),
Biology of new world Microtus (pp. 535-566). Lawrence,
KS: American Society of Mammalogists.

205
Pianka, E. R. (1970). On r- and K-selection. The American
Naturalist. 104. 592-597.
Pierce, J. D., & Dewsbury, D. A. (1989). Male role in the
survival and development of prairie voles. Paper
presented at the 1989 joint meeting of the Western
Psychological Association and the Rocky Mountain
Psychological Association.
Pierce, J. D., Jr., & Dewsbury, D. A. (1991). Female
preferences for unmated versus mated males in two
species of voles (Microtus ochrogaster and Microtus
montarms). Journal of Comparative Psychology. 105.
165-171.
Pinter, A. J. (1968). Effects of diet and light on growth,
maturation, and adrenal size of Microtus montanus.
American Journal of Physiology. 215. 461-466.
Reiter, E. 0. (1982). Neuroendocrine control mechanisms and
the onset of puberty. Annual Review of Physiology. 44.
595-613.
Richmond, M. E., & Conaway, C. H. (1969a). Management,
breeding and reproductive performance of the vole,
Microtus ochroqaster in a laboratory colony. Laboratory
Animal Care. 19, 80-87.
Richmond, M. E., & Conaway, C. H. (1969b). Induced ovulation
and oestrus in Microtus ochroqaster. Journal of
Reproduction and Fertility, Supplement. 6, 357-376.
Richmond, M., & Stehn, R. (1976). Olfaction and reproductive
behavior in microtine rodents. In R. L. Doty (Ed.),
Mammalian olfaction, reproductive processes, and
behavior (pp. 197-217). New York: Academic Press.
Rissman, E. F., & Johnston, R. E. (1985). Female
reproductive development is not activated by male
California voles exposed to family cues. Biology of
Reproduction. 32., 352-360.
Rissman, E. F., Sheffield, S. D., Kretzmann, M. B., Fortune,
J. E., & Johnston, R. E. (1984). Chemical cues from
families delay puberty in male California voles.
Biology of Reproduction. 31. 324-331.
Rosenblatt, J. S. (1967). Nonhormonal basis of maternal
behavior in the rat. Science. 156. 1512-1514.

206
Salo, A. L. , & French, J. A. (1989). Early experience,
reproductive success, and development of parental
behaviour in Mongolian gerbils, Animal Behaviour. 38.
693-702.
Sawrey, D. K. (1990). Laboratory studies on the influence of
males on reproductive activation in female montane
voles (Microtus montanus) (Doctoral dissertation,
University of Florida, 1989). Dissertation Abstracts
International. 51, 1541B.
Sawrey, D. K., & Dewsbury, D. A. (1985). Control of
ovulation, vaginal estrus, and behavioral receptivity
in voles (Microtus). Neuroscience and Biobehavioral
Reviews. 9, 563-571.
Sawrey, D. K., & Dewsbury, D. A. (1991). Males accelerate
reproductive development in female montane voles.
Journal of Mammalogy. 72., 343-346.
Schadler, M. H. (1982, March). Strange males block pregnancy
in lactating pine voles, Microtus pinetorum. and reduce
survival and growth of nursing young. In: Proceedings
of the sixth eastern pine and meadow vole symposium
(pp. 132-138). Harpers Ferry, WV.
Schadler, M. H. (1983). Male siblings inhibit reproductive
activity in female pine voles, Microtus pinetorum.
Biology of Reproduction. 28., 1137-1139.
Schadler, M. H. (1985). Strange males cause death or
suppression of growth in infant pine voles, Microtus
pinetorum. Journal of Mammalogy. 66, 387-390.
Schadler, M. H. (1990). Social organization and population
control in the pine vole, Microtus pinetorum. In R. H.
Tamarin (Ed.), Social systems and population cycles in
voles (pp. 121-130). Birkhauser Verlag: Basel,
Switzerland.
Schadler, M. H., & Butterstein, G. M. (1979). Reproduction
in the pine vole, Microtus pinetorum. Journal of
Mammalogy. 60, 841-844.
Schaffer, W. M., & Elson, P. F. (1975). The adaptive
significance of variation in life history among local
populations of Atlantic Salmon in North America.
Ecology. 56, 577-590.
Schaffer, W. M. , & Tamarin, R. H. (1973). Changing
reproductive rates and population cycles in lemmings
and voles. Evolution. 27., 111-124.

207
Shapiro, L. E., & Austin, D., Ward, S. E., & Dewsbury, D. A.
(1986). Familiarity and female mate choice in two
species of voles (Microtus ochrogaster and Microtus
montanus). Animal Behaviour. 34. 90-97.
Shapiro, L. E., & Dewsbury, D. A. (1986). Male dominance,
female choice and male copulatory behavior in two
species of voles (Microtus ochrogaster and Microtus
montanus). Behavioral Ecology and Sociobioloav. 18,
267-274.
Shapiro, L. E., & Dewsbury, D. A. (1990). Differences in
affiliative behavior, pair bonding, and vaginal
cytology in two species of voles (Microtus ochrogaster
and M. montanusi. Journal of Comparative Psychology.
104. 268-274.
Stearns, S. C. (1976). Life-history tactics: A review of
ideas. Quarterly Review of Biology. 51, 3-47.
Stearns, S. C. (1989). The evolutionary significance of
phenotypic plasticity. Bioscience. 39, 436-445.
Storey, A. E. & Snow, D. T. (1987). Male identity and
enclosure size affect paternal attendance of meadow
voles, Microtus pennsvlvanicus. Animal Behaviour. 35,
411-419.
Tamarin, R. H. (1985). Biology of new world Microtus.
Lawrence, KS: American Society of Mammalogists.
Taylor, S. A. (1991). Laboratory studies of estrus induction
and pregnancy maintenance in voles (Microtus) (Doctoral
dissertation, University of Florida, 1990).
Dissertation Abstracts International. 52, 1772B.
Taylor, S. A., & Dewsbury, D. A. (1988). Effects of
experience and available cues on estrous versus
diestrous preferences in male prairie voles, Microtus
ochrogaster. Physiology & Behavior. 42. 379-388.
Taylor, S. A., Salo, A. L. & Dewsbury, D. A. (in press).
Estrus induction in four species of voles (Microtus).
Journal of Comparative Psychology.
Vandenbergh, J. G. (1967). Effect of the presence of a male
on the sexual maturation of female mice. Endocrinology.
81. 345-349.
Vandenbergh, J. G. (1969). Male odor accelerates female
sexual maturation in mice. Endocrinology. 84. 658-660.

208
Vandenbergh, J. G. (1971). The influence of the social
environment on sexual maturation in male mice. Journal
of Reproduction and Fertility. 24. 383-390.
Vandenbergh, J. G. (1988). Social interactions and the
coordination of reproductive behavior in rodents and
nonhuman primates. Journal of the American Veterinary
Medical Association. 193. 1161-1164.
Vandenbergh, J. G., & Coppola, D. M. (1986). The physiology
and ecology of puberty modulation by primer pheromones.
In Advances in the study of behavior. Vol, 16 (pp.
71-107). Orlando, FL: Academic Press.
Vandenbergh, J. G., Finlayson, J. S., Dobrogosz, W. J.,
Dills, S. S., & Kost, T. A. (1976). Chromatographic
separation of puberty accelerating pheromone from male
mouse urine. Biology of Reproduction. 15. 260-265.
Vandenbergh, J. G., Whitsett, J. M., & Lombardi, J. H.
(1975). Partial isolation of a pheromone accelerating
puberty in female mice. Journal of Reproduction and
Fertility. 43, 515-523.
Vaughan, M. K., Vaughan, G. M., & Reiter, R. J. (1973).
Effect of ovariectomy and constant dark on the weight,
reproductive and certain other organs in the female
vole, Microtus montanus. Journal of Reproduction and
Fertility. 32. 9-14.
Vehrencamp, S. L. & Bradbury, J. W. (1984). Mating systems
and ecology. In J. R. Krebs & N. B. Davies (Eds.),
Behavioural ecology: An evolutionary approach (2nd
Ed.), (pp. 251-278). Sunderland, MA: Sinauer
Associates.
Wang, Z., & Novak, M. A. (1992). Influence of the social
environment on parental behavior and pup development of
meadow voles fMicrotus pennsvlvanicus) and prairie
voles (M. ochrogaster). Journal of Comparative
Psychology. 106. 163-171.
Wasser, S. K., & Barash, D. P. (1983). Reproductive
suppression among female mammals: Implications for
biomedicine and sexual selection theory. Quarterly
Review of Biology. 58, 513-538.
Wayne, N. L., & Rissman, E. F. (1990). Environmental
regulation of reproduction in an opportunistic breeder:
the musk shrew (Insectivora: Suncus murinus). In A.
Epple, C. G. Scanes & M. H. Stetson (Eds), Progress in
comparative endocrinology, (pp. 440-473). New York:
Wiley-Liss.

209
Webster, A. B., & Brooks, R. J. (1981). Social behavior of
Microtus pennsvlvanicus in relation to seasonal changes
in demography. Journal of Mammalogy. 62. 738-751.
West-Eberhard, M. J. (1979). Sexual selection, social
competition, and evolution. Proceedings of the
Philosophical Society of America. 123. 222-234.
Wilson, S. C. (1982). Parent-young contact in prairie and
meadow voles. Journal of Mammalogy. 63. 300-305.
Wittenberger, J. F. (1979). The evolution of mating systems
in birds and mammals. In P. Marler & J. G. Vandenbergh
(Eds.), Handbook of behavioral neurobiology. Vol. 3.
(pp. 271-349). New York: Plenum Press.
Wolff, J. O. (1980). Social organization of the taiga vole
(Microtus xanthoanathusl. The Biologist. 62, 34-45.
Wolff, J. O. (1985). Behavior. In R. H. Tamarin (Ed.),
Biology of new world Microtus (pp. 340-372). Lawrence,
KS: American Society of Mammalogists.
Wuensch, K. L. (1985). Effects of early paternal presence
upon nonhuman offsprings' development. American
Zoologist. 25. 911-923.

APPENDIX A
MEANS OF BODY WEIGHTS AND NUMBERS OF SUBJECTS
MICROTUS (EXPERIMENT 1)
The following tables contain the mean body weights of
subjects in Experiment 1. See text for accompanying
explanation and discussion.
210

Appendix A. Mean Body Weight (in g + S.E.) of Microtus (Experiment 3).
Condition N Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9
Pine Voles: Males
Control
17
11.26
( -46)
14.82
( -49)
17.81
( -54)
19.53
( -55)
20.17
( -56)
21.24
( -53)
21.24
( -53)
Family
16
11.49
( -59)
15.56
( -69)
18.65
( -69)
20.68
( -72)
21.50
( -79)
22.34
( -81)
22.71
( -88)
Male
14
12.61
( -62)
16.62
( -67)
19.22
( -86)
20.78
( -90)
21.94
( -96)
22.51
( i.oi)
22.96
( 1.05)
Female
16
11.77
( -60)
15.64
( -56)
17.63
( -59)
18.95
( -59)
19.76
( -62)
20.29
( -59)
20.64
( -65)
Prairie
Voles:
Males
Control
16
18.48
( -44)
24.66
( -75)
26.85
( .92)
28.37
( 1-04)
29.80
( 1-17)
30.69
( 1-29)
31.94
( 1-41)
Family
16
20.53
( -65)
25.49
( -78)
27.79
( -89)
29.34
( 1.07)
30.47
( 1.13)
31.44
( 1-17)
32.42
( 1.28)
Male
16
18.68
( -53)
24.45
( -79)
26.53
( 1.03)
28.04
( 1.26)
29.65
( 1-34)
31.00
( 1.41)
31.74
( 1-46)
Female
16
18.46
( -65)
24.04
( -80)
25.72
( .97)
27.17
( 1-07)
28.64
( 1-17)
29.77
( 1.28)
30.90
( 1.35)
211

Appendix A—continued.
Condition N Week 3
Mean Body Weight (in g + S.E.)
Week 4 Week 5 Week 6
of Microtus (Experiment 3).
Week 7 Week 8 Week 9
Meadow '
Voles:
Males
Control
16
21.23
26.93
29.43
( .94)
( 1-14)
( 1.25)
Family
16
21.89
27.26
29.95
( .80)
( 1.07)
( 1.25)
Male
16
23.16
28.73
31.50
( .55)
( -45)
( -67)
Female
16
20.35
27.07
30.68
( .58)
( .12)
( 1.02)
Montane
Voles:
: Males
Control
16
16.09
23.27
27.74
( .87)
( 1-05)
( 1-11)
Family
16
17.29
25.19
28.80
( .94)
( 1.03)
( 1.14)
Male
16
17.22
24.03
27.81
( .87)
( -88)
( 1.15)
Female
16
18.04
24.76
28.46
( .80)
( -77)
( 1-09)
32.
. 54
35 ,
.22
37 .
.83
39 .
. 18
( 1 â– 
. 52)
( 1 â– 
.74)
( 1.
.99)
( 2.
.14)
33 .
.42
36 .
. 89
39.
. 67
41.
. 64
( 1 •
.45)
( 1.
.66)
( 1.
.85)
( 1.
.97)
35.
.45
38 ,
. 99
41.
. 56
43 .
.49
( â– 
.89)
( â– 
. 94)
( 1.
.16)
( 1 â– 
.28)
34.
.44
37 .
.90
41.
. 09
43 .
, 13
( 1 •
.32)
( 1 â– 
, 37)
( 1.
. 50)
( 1.
.58)
31.65
( 1.28)
34.47
( 1.45)
36.54
( 1.63)
37.70
( 1-67)
31.29
( 1.40)
33.61
( 1-48)
35.52
( 1-50)
37.22
( 1-51)
31.39
( 1-46)
34.71
( 1.61)
37.21
( 1-83)
39.12
( 2.04)
32.21
( 1.30)
34.80
( 1.27)
36.65
( 1-32)
38.64
( 1-35)

Appendix A—continued.
Mean Body Weight (in g + S.E.) of Microtus (Experiment 3).
Condition N Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9
Pine Voles: Females
Control
16
10.
68
13 .
. 81
16,
.40
17 .
. 79
18.
. 38
18 .
. 66
18 .
,87
( •
55)
( â– 
,48)
(
. 48)
(
. 55)
( â– 
. 53)
( â– 
. 54)
( â– 
. 64)
Family
16
11.
89
15 .
,47
17 ,
.96
19 .
.35
20.
.31
21.
. 13
21.
.40
( •
62)
( •
. 55)
(
.74)
( •
.83)
( •
.92)
( •
.99)
( 1.
. 03)
Male
18
11.
85
15.
. 62
17
.90
19 ,
.26
20.
. 04
20.
. 59
20 .
, 88
( •
61)
( â– 
.62)
(
.67)
(
. 66)
( â– 
. 69)
( â– 
. 69)
( â– 
.73)
Female
16
11.
50
14 .
,95
17 ,
. 37
18 ,
.94
19 .
.83
20.
. 04
20,
.45
( •
77)
( â– 
.67)
(
.66)
( â– 
. 72)
( •
.76)
( •
.80)
( â– 
,80)
Prairie
Voles:
Females
Control
18
18 .
12
22 .
.07
23
. 13
23 ,
.95
25.
.01
25.
. 80
26.
. 65
( â– 
63)
( â– 
, 56)
(
. 66)
(
.79)
( â– 
.82)
( â– 
.99)
( 1.
. 03)
Family
16
17 .
46
21.
.91
23 ,
. 09
24 ,
. 14
24 .
.92
25.
.89
26.
. 58
( •
68)
( â– 
. 64)
( .
.61)
(
. 65)
( â– 
. 69)
( â– 
.71)
( â– 
.74)
Male
16
18.
11
22 .
.40
23 ,
.83
24 ,
. 89
26.
. 14
26.
. 96
27 .
.91
( •
57)
( â– 
.61)
(
.82)
( â– 
.98)
( 1.
.16)
( 1.
.31)
( 1.
.31)
Female
16
18 .
12
22 .
. 03
23 ,
.28
24 ,
. 32
25.
, 00
25.
.90
26 .
.78
( •
61)
( â– 
.57)
(
.63)
(
.75)
( â– 
.80)
( â– 
. 90)
( â– 
.93)
213

Appendix A—continued. Mean Body Weight (
Condition N Week 3 Week 4 Week 5
Meadow ’
Voles:
Females
Control
16
21.91
25.26
26.02
( -79)
( -82)
( -92)
Family
15
19.87
24.34
24.92
( -84)
( -90)
( 1-05)
Male
16
20.75
24.61
25.51
( -66)
( -66)
( -73)
Female
15
20.61
24.50
26.31
( -90)
( -98)
( 1.16)
Montane
Voles:
Females
Control
16
16.79
22.24
23.84
( -74)
( -78)
( -81)
Family
16
16.44
21.95
23.78
( -70)
( -90)
( 1-07)
Male
15
15.87
21.38
23.04
( -60)
( -71)
( -82)
Female
16
16.84
21.99
23.31
( -64)
( -65)
( -77)
g + S.E.) of Microtus (Experiment 3).
Week 6
Week 7
Week 8
Week 9
27.39
( 1.02)
29.06
( 1.06)
29.89
( 1-23)
30.95
( 1-30)
26.67
( 1-21)
28.28
( 1-24)
29.56
( 1.28)
30.65
( 1-34)
27.15
( -83)
29.17
( -90)
30.83
( 1.14)
32.02
( 1.14)
28.13
( 1-34)
29.99
( 1-43)
31.69
( 1-58)
32.79
( 1-55)
25.25
( 1.03)
26.65
C 1-19)
27.76
( 1-36)
28.92
( 1-43)
25.71
( 1.28)
27.61
( 1-44)
29.08
( 1.61)
30.16
( 1-71)
24.75
( 1.06)
26.13
( 1.21)
27.25
( 1-37)
28.37
( 1-43)
24.39
( -99)
25.53
( 1-23)
26.82
( 1-37)
27.67
( 1-44)
214

APPENDIX B
MEANS OF ANOGENITAL DISTANCE (IN MM) IN
MICROTUS (EXPERIMENT 2)
The following tables contain the mean anogenital
distances, standard errors of the mean, and numbers of males
of the four Microtus species studied in Experiment 1. See
text for further discussion of the results.
215

Appendix B. Mean Anogenital Distance (in mm + S.E.) of Male Microtus (Experiment 3).
Condition
N
Week 3
Week 4
Week 5
Week 6
Week 7
Week 8
Week 9
Pine Voles
Control
17
4.32
( -29)
5.21
( -26)
6.15
( -24)
6.68
( -27)
7.21
( -20)
7.56
( -32)
7.71
( -22)
Family
15
4.47
( .28)
5.50
( -21)
6.07
( -27)
6.23
( -30)
7.00
( -25)
7.40
( .26)
7.70
( -26)
Male
14
4.75
( -IV)
5.75
( -21)
6.75
( -32)
6.86
( -27)
7.00
( -26)
7.82
( -31)
7.96
( .25)
Female
16
4.62
( .22)
5.41
( -24)
5.66
( -24)
6.41
( -22)
6.94
( .28)
6.97
( -18)
7.75
( .24)
Prairie Voles
Control
15
7.53
( .34)
9.93
( -27)
11.03
( -38)
11.37
( -41)
11.67
( .39)
11.73
( -50)
12.20
( .38)
Family
16
8.28
( -32)
10.12
( -33)
10.97
( -43)
11.91
( -39)
11.66
( -40)
11.91
( -31)
12.31
( -32)
Male
16
8.12
( -34)
10.37
( -26)
11.06
( -31)
11.53
( .33)
11.60
( -27)
11.97
( -35)
12.28
( .35)
Female
16
7.75
( .22)
9.66
( -31)
10.53
( -31)
10.94
( -41)
11.31
( .37)
11.12
( -35)
11.25
( -36)

Appendix B—continued. Mean Anogenital Distance (in mm + S.E.) of Male Microtus
(Experiment 3).
Condition
N
Week 3
Week 4
Week 5
Week 6
Week 7
Week 8
Week 9
Meadow Voles
Control
16
10.94
12.81
14.97
16.97
17.19
18.09
18.22
( -49)
( .44)
( .47)
( -55)
( .72)
( .62)
( .61)
Family
16
10.47
12.94
15.16
15.81
16.81
17.69
18.09
( .39)
( -44)
( .59)
( -44)
( .53)
( .56)
( -50)
Male
16
11.16
13.09
15.78
16.53
17.62
17.91
19.19
( -37)
( .41)
( .49)
( .59)
( .51)
( .61)
( -55)
Female
16
10.65
13.09
14.72
17.16
18.12
18.16
18.72
( .47)
( -47)
( -58)
( .65)
( .64)
( .59)
( .74)
Montane Voles
Control
16
7.00
8.81
10.31
12.19
12.75
12.69
13.44
( .38)
( .52)
( .45)
( .53)
( -69)
( -37)
( .54)
Family
16
7.00
9.59
10.53
11.62
11.68
11.72
12.25
( .36)
( .45)
( .48)
( -38)
( .30)
( -29)
( .31)
Male
16
6.87
8.47
10.75
10.91
11.84
12.19
13.09
( -44)
( .47)
( .45)
( .44)
( -60)
( .48)
( -48)
Female
16
7.50
9.41
10.66
11.25
12.28
11.97
12.66
( .32)
( -45)
( .46)
( -55)
( .37)
( -37)
( -36)
217

APPENDIX C
ANALYSES OF NUCLEATED CELLS AND LEUKOCYTES FROM VAGINAL
SMEARS OF MICROTUS (EXPERIMENT 1)
Appendices c-1 through C-3 contain the number of
subjects and the mean percentages of nucleated, and
leukocytes for each species in Experiment 1. These analyses
of nucleated cells and leukocytes complement the results
reported in the primary text for cornified cells (see
Vaginal Smears of results section for methods).
Within-Snecies Comparisons of Cell Types
Prairie voles: nucleated cells: condition effects:
Few significant differences were found among the percentages
of nucleated cells from prairie voles as a function of
condition. On Days 43-44, the mean percentage of cells in
the Control condition (M = 3.8%) was significantly less than
those in the Family and Male conditions (M's = 9.3% and 7.8%
respectively), (Kruskal-Wallis ANOVA: H(3, N = 62) = 7.85,
p = .049). The number of nucleated cells in the Female
condition were intermediate in value to the others (M =
6.7%). A second significant effect was found for Days
51-52; the mean percentage of nucleated cells from the Male
condition (M = 11.5%) was significantly higher than those in
either the Control or Female conditions (M's = 6.3% and
5.0%), H( 3, N = 64) = 9.27, E = -026.
218

Age effects. In the repeated measures analyses, the
percentages of nucleated cells from prairie voles in the
219
Female condition shifted significantly across alternate
blocks of days, although not in any clear, systematic
fashion (Friedman ANOVA: X2(7, N = 15) = 16.21, p < .023).
Levels of nucleated cell were relatively high on days 33-34
(M = 8.5%), then decreased in level by Day 41-2 (M = 4.6%)
and then rose again to the highest level (M = 9.6%) on Days
49-50. (Note: differences below are indicated by direction
of later days being higher in cell percentages than earlier
days except where noted "*"). Other significantly different
blocks of days from those in the Female condition included:
Days 33-34 versus 37-38*, Days 37-38 versus Days 49-50; Days
41-42 vs. 49-50; and Days 49-50 versus Day 57-58*.
Changes in the percentages of nucleated cells across
days, among prairie voles in the other three conditions, did
not reach statistically significant levels.
Leukocytes: condition effects. Relatively small
changes in the percentages of leukocytes among prairie voles
resulted in only one significant shift among the blocks of
days. During Days 35-36, significantly more leukocytes were
found in the Control and Female conditions than in the
Family condition (M's = 78.2% and 83.3% versus 67.7%),
H(3, N = 55) = 9.72, p = .021.
Aae effects. Repeated measures analyses for leukocytes
mirrored those of cornified cells. No significant shifts
were found among any of the four experimental conditions.

220
Meadow voles: nucleated cells: condition effects.
Analyses of the percentages of nucleated cells among the
conditions did not reveal any statically significant
differences.
Age effects. Only meadow voles in the Female condition
displayed significant shifts in the percentages of nucleated
cells, X2(7, N = 13) = 17.42, p < .015. Typically there
were decreasing proportions of nucleated cells as the
females in Female condition grew older. Specific
significant differences in the percentages of nucleated
cells included: Days 37-38 versus 49-50* through 61-62*;
Days 41-42 versus 49-50*, 57-58 through 61-62*; and Days
45-46 versus 57-58* through 61-62*.
Leukocytes: condition effects. Comparisons of the
percentages of leukocytes by blocks of days among the
different conditions did not reveal any statically
significant differences.
Age effects. Significant changes in the percentages of
leukocytes were evident as a function of age. These changes
in percentages largely reflected, inversely, the differences
reported in the main text for the percentages of cornified
cells. Analyses revealed there were typically decreasing
proportions of leukocytes, in all conditions, across the
course of the study. All conditions had significant changes
in the percentage of leukocytes.
Control condition. Subjects in the Control conditioned
had the highest percentages of leukocytes during the first

221
blocks compared to latter days (Day 33-34 and 37-38; M's =
80.3% and 81.2%) (X2(7, N = 11) = 28.73, p < .001).
Specific differences included: Day 33-34 versus Days 49-50
through Days 61-62*; Days 37-38 versus Days 41-42 through
Days 61-62 ; Days 45-46 versus Days 53-54 through Days
61-62*.
Family condition. Percentages of leukocytes were higher
* • • 9
in the earlier days of observation than later, X (7, N = 11)
= 32.76, p < .001. Specific pairwise comparisons that were
significantly different include: Days 33-34 versus Days
45-46 through Days 61-62*; Days 37-38 versus Days 45-46
through Days 61-62*; Days 41-42 versus Days 45-46 through
Days 61-62*; Days 37-38 versus Days 41-42*.
Male condition. Percentages of leukocytes were higher
■ • 9
m the earlier days of observation than later, X (7, N = 12)
= 36.11, p < .001. Significant differences included: Days
33-34 versus Days 41-42 through Days 61-62*; Days 37-38
versus Days 49-50 through Days 61-62 ; Days 41-42 versus
Days 57-58 through Days 61-62 ; and Days 45-46 versus Days
53-54 through Days 61-62*.
Female condition. Percentages of leukocytes were higher
in the earlier days of observation than later, X2(7, N =
13)= 46.15, p < .001. Significant differences included:
Days 33-34 versus Days 45-46 through Days 61-62*; Days 37-38
versus Days 45-46 through Days 61-62*; Days 41-42 versus
Days 45-46 through Days 61-62 ; and Days 45-46 versus Days

222
53-54 through Days 61-62*; and Days 49-50 versus Days
57-58*.
Montane voles: nucleated cells: condition effects.
Variations in the percentages of nucleated cells among the
different conditions were minimal and resulted in only one
significant result among the groups on Days 61-62, H(3, N =
62) = 9.78, p = .020. On this block, the mean percentage of
nucleated cells was significantly greater from subjects in
the Female condition (M = 6.9%) compared to the percentages
of nucleated cells from those in either the Control (U =
50.0, p = .01) or Family conditions (U = 45.0, p = .003)(M's
= 4.6% and 4.0%).
Age effects. Repeated measures analysis revealed that
only one condition resulted in significant differences in
the percentages of nucleated cells. Subjects of the Family
condition had significant change in the proportions of
nucleated cells, X2(7, N = 11) = 14.21, p < .047. Post-hoc
comparisons indicated that all days, except Days 53-54, had
significantly more nucleated cells than on Day 61-62 (M =
3.3%). The cell percentages were relatively small among the
earlier blocks of days, (M's range: 5.0% to 8.7%), but were
significantly more than the last block of Days 61-62.
Leukocytes; condition effects: Three of the last five
two-day blocks of days were found to have significant
differences in the percentages of leukocytes among the
conditions (Days 53-54: H(3, N = 61) = 8.77, p = .032; Days

223
57-58: H(3, N = 62) = 8.47, e = .037; Days 61-62: H(3, N =
62) = 9.17, E = -027) .
Post-hoc comparisons of the cell Eercentages for Days
53-54 revealed that the Control condition had significantly
more leukocytes than those in the Female condition (M's =
61.1% and 41.0%; U = 64.0, e = .046), those in the Family
condition had significantly more cells than those in the
Male condition (M's = 56.3% versus 53.0%; U = 66.0, e =
.033), and those in the Male condition had more than those
in the Female condition (M's = 53.0% versus 41.0%; U = 55.0,
E = •029).
Analyses of Days 57-58 revealed only one significant
difference among the conditions, those in the Control
condition had significantly more leukocytes than those in
the Family condition (M's = 60.5% versus 53.9%; U = 79.0,
E = .040). Those in the Control and Female failed to be
significantly different (e = .051).
On days 61-62, two statistical differences were found.
Those in Control condition had a greater percentage of cells
than those in the Family condition (M's = 60.5% versus
57.7%; U = 72.0, e = .021), and those in the Family
condition had significantly more leukocytes than those in
the Male condition (M's = 57.7% versus 49.7%; U = 70.0,
E = .030). Those in the Control and Female failed to differ
significantly (e = .051).
Leukocytes: age effects: Comparisons of the percentages
of leukocytes among the different conditions revealed

224
significant differences within each group. Statistical
differences were found among blocks of days including Days
53-54, 57-58, and 61-62. These blocks of days were the same
blocks where differences were found among the percentages of
cornified cells (see main text).
Control condition. Percentages of leukocytes were
higher in the earlier days of observation than later,
X2(7, N = 14) = 20.48, p < .004. Specific differences
included: Days 33-34 versus Days 41-42 and Days 49-50
through Days 61-62*; Days 37-38 versus Days 57-58 through
Days 61-62*; Days 41-42 versus Days 57-58 through Days
61-62 ; and Days 45-46 versus Days 53-54 through Days
61-62*; and Days 49-50 versus Days 61-62*.
Family condition. Percentages of leukocytes were higher
in the earlier days of observation than later, X2(7, N = 11)
= 18.12, p < .011. Significantly different blocks of days
included: Days 33-34 versus Days 57-58 through Days 61-62*;
Days 41-42 versus Days 49-50 through Days 61-62*; and Days
53-54 versus Days 57-58*.
Male condition. Percentages of leukocytes were higher
in the earlier days of observation than later, X2(7, N = 12)
=23.14, p < .002. Significantly different blocks of days
included: Days 33-34 versus Days 49-50 through Days 61-62*;
Days 37-38 versus Days 53-54, Days 61-62*; Days 41-42 versus
Days 53-54, Days 61-62*; and Days 49-50 versus Days 53-54,
Days 61-62*.

225
Female condition. Percentages of leukocytes were higher
. . . o
m the earlier days of observation than later, X (7, N = 8)
= 20.96, g < .004. Days 33-34 versus Days 41-42 through
Days 49-50, Days 57-58 through Day 61-62*; Days 37-38 versus
Days 49-50 through Days 61-62*; Days 45-46 versus Days
61-62*; and Days 53-54 versus Days 61-62*; and Days 57-58
versus Days 61-62*.

226
Appendix C-l. Summary Table of Mean Percentages of
Nucleated Cells and Leukocytes for Prairie Voles
(Experiment 1).
Condition
Control Family Male Female
Days
Nuc
Leuk
N
NUC
Leuk
N
Nuc
Leuk
N
NUC
Leuk
N
21-
-22
13 .
, 1
72 .
. 0
3
10.
. 4
78.
, 0
2
13 .
. 2
72 .
.4
3
21.
, 1
64.
, 9
3
23-
-24
11.
, 4
64 .
, 5
4
11.
, 3
74 .
. 6
4
10.
. 1
66.
. 1
4
11.
, 1
62 .
, 7
5
25-
-26
9 .
, 6
74 .
. 0
7
11.
, 4
67 .
, 1
8
13 .
.8
68.
, 0
10
13 .
. 0
65 .
, 0
8
27-
-28
7 .
. 0
70 ,
. 3
7
13 ,
. 1
66.
. 1
8
9 .
, 8
72 .
, 3
11
10.
, 3
62 .
, 2
9
29-
-30
7 .
, 0
74.
. 6
7
8 .
, 7
73 .
, 9
9
10.
.4
70.
.9
14
7 .
. 4
72 .
, 1
12
31-
-32
4 .
, 9
77 .
.8
10
6.
. 3
80.
. 2
10
8.
. 1
73 .
.0
14
5 .
, 2
77 .
, 4
13
33-
-34
6.
,8
76.
. 2
13
6.
. 9
77.
.9
10
10 .
. 2
66.
. 6
14
8 .
, 5
77 .
, 0
15
35-
-36
6.
, 8
78 .
. 2
13
10 .
. 2
67.
. 7
12
7 .
.2
73 .
. 8
15
4 .
. 0
83 .
. 3
15
37-
-38
5.
, 3
81.
. 2
14
7 .
. 0
77.
. 6
13
9 ,
. 0
70.
.2
16
4 .
, 8
82
2
16
39-
-40
6.
, 9
79 .
.4
14
8 .
. 5
75.
, 7
13
6.
. 2
76.
, 2
16
6 .
. 3
79 .
, 7
16
41-
-42
6.
, 1
77 .
, 7
15
7 .
, 8
77.
.9
15
9 .
. 0
77 .
. 0
16
4 .
, 6
82 .
, 2
16
43-
-44
3 .
, 8
78 .
. 9
15
9 .
. 3
71.
. 6
15
7 ,
. 8
76.
.7
16
6.
, 7
77 .
, 4
16
45-
-46
6.
.4
74 .
, 5
15
7 .
. 7
78.
. 5
15
8 .
. 8
76.
.8
16
7 .
, 9
78 .
. 0
16
47-
-48
6.
, 1
78 .
. 6
16
6.
. 1
79.
. 1
15
9 ,
. 8
78 .
. 4
16
6.
, 5
80.
, 9
16
49-
-50
7 .
, 6
70.
, 8
16
8 .
, 8
76.
, 5
15
9 .
. 7
77 .
.2
16
9 .
, 6
77 .
, 0
16
51-
-52
6.
. 3
75.
. 4
17
7 .
. 3
77.
.9
15
11.
. 5
72 .
. 3
16
5 .
.0
82 .
, 3
16
53-
-54
5.
, 9
77 .
. 3
17
9 .
. 1
76.
. 1
15
8 .
. 3
80.
. 1
16
7 .
, 2
81.
, 3
16
55-
-56
7.
. 2
79 .
.5
17
7 ,
.8
75.
.4
15
10.
. 9
72 .
.4
16
6.
.9
82 .
.8
16
57-
-58
8.
. 6
74 .
. 7
17
11.
. 5
74 .
. 0
15
10.
. 0
78.
. 8
16
7 .
, 1
82 .
, 1
16
59-
-60
6.
,7
76.
.0
17
8.
. 3
75.
.9
15
11.
. 0
74 .
. 4
16
7 .
.7
78.
. 5
16
61-
-62
8.
.7
78 .
. 1
17
10.
. 8
77.
. 3
15
11.
. 1
75.
. 2
16
6.
. 1
79.
.6
16
Nuc: Nucleated Cells; Leuk; Leukocytes.

227
Appendix C-2. Summary Table of Mean Percentages of
Nucleated Cells and Leukocytes for Meadow Voles
(Experiment 1).
Condition
Control Family Male Female
Days
Nuc Leuk N
Nuc Leuk N
Nuc
Leuk
N
Nuc
Leuk
N
21-22
10.1
64.9
10
7.3
45.5
7
5.2
61.3
12
8.7
56.7
9
23-24
11.7
73.4
11
7.6
77.2
10
5.7
74.8
12
7.9
83.7
12
25-26
7.7
75.2
11
7.9
77.5
10
5.2
84.1
12
6.6
83.8
13
27-28
4.4
80.5
11
6.4
73.4
10
4.7
81.1
12
4.1
84.6
13
29-30
3.3
85.0
11
4.8
73.4
11
5.4
79.4
12
4.1
85.4
13
31-32
6.0
77.7
11
3.9
74.7
11
3.5
81.8
12
4.6
83.7
13
33-34
2.6
80.3
11
4.3
75.2
11
4.0
76.1
12
4.0
82.2
13
35-36
3.2
75.9
11
4.2
75.5
11
4.9
71.4
12
7.3
76.5
13
37-38
3.0
81.9
12
4.2
73.8
12
5.5
75.1
12
6.1
75.4
13
39-40
4.5
78.4
12
4.5
70.6
12
5.4
67.8
12
7.0
69.4
13
41-42
6.6
65.4
12
2.4
79.6
12
4.7
62.8
12
5.5
66.6
14
43-44
4.0
73.7
12
3.0
73.5
12
4.3
62.6
12
4.2
62.5
14
45-46
4.6
72.0
12
6.2
59.3
12
4.5
64.2
12
4.9
57.1
14
47-48
5.7
67.8
12
5.5
59.0
12
3.3
52.4
12
3.8
52.8
14
49-50
2.5
66.8
12
5.7
55.4
13
2.5
57.8
13
2.8
50.0
14
51-52
4.3
63.7
13
4.5
56.5
13
4.0
56.0
13
4.5
47.9
14
53-54
6.7
61.1
13
5.4
56.3
13
3.6
53.0
13
2.8
41.0
14
55-56
5.9
54.2
13
3.1
61.3
13
3.5
54.5
13
4.5
43.7
14
57-58
4.4
60.5
13
4.7
53.9
13
3.0
56.3
13
2.5
42.3
14
59-60
4.4
58.8
14
4.1
54.3
13
5.0
53.8
15
3.4
39.9
14
61-62
4.2
60.5
14
4.8
57.7
13
4.3
49.7
15
2.6
43.1
14
Nuc: Nucleated Cells; Leuk: Leukocytes.

228
Appendix C-3. Summary Table of Mean Percentages of
Nucleated Cells and Leukocytes for Montane Voles
(Experiment 1).
Condition
Control Family Male Female
Days
NUC
Leuk
N
NUC
Leuk
N
Nuc
Leuk
N
Nuc
Leuk
N
21
-22
16.
. 6
66.
. 3
7
16.
.6
68.
. 6
7
13 .
. 2
73 ,
. 9
8
13 .
.3
78 .
.7
6
23
-24
10.
.2
68.
. 4
9
10.
.2
78.
.8
7
12 ,
. 5
71,
. 6
10
11.
. 9
73 .
. 8
9
25
-26
9.
.8
74 .
.4
10
9.
. 6
72 .
.9
7
10.
. 7
72 ,
. 1
10
16.
.6
67 .
.4
9
27
-28
10.
. 7
71.
. 3
10
6 ,
. 5
80.
.7
8
11.
. 4
72 ,
. 2
11
8.
. 1
71.
. 1
9
29
-30
7 .
.9
58.
. 6
10
8.
.9
70.
. 0
8
13 .
. 8
68.
. 2
11
4 .
.9
79.
. 2
9
31
-32
4 .
. 3
57 .
.4
11
8.
. 3
68.
. 3
8
12 .
. 6
68.
. 0
11
5.
.8
77.
. 2
9
33
-34
7.
. 6
55.
. 5
14
8 ,
. 6
65.
. 1
11
10.
. 1
57 ,
. 6
12
4 .
.8
77.
.9
9
35
-36
6.
.2
57.
.8
15
10.
. 5
66 .
.7
11
5.
. 8
64 ,
. 7
12
4 .
. 1
80.
. 0
9
37
-38
8 .
.4
51.
. 1
15
8.
. 0
62 .
. 1
12
10.
. 0
50 ,
.7
12
8.
. 2
73 .
. 7
9
39
-40
7 .
. 4
53 .
, 9
15
7 .
.2
63 .
. 3
12
8 .
. 0
51,
. 4
12
9.
. 1
70.
. 0
10
41
-42
8 .
. 6
48.
. 2
15
6 ,
. 8
64 .
. 7
12
7 .
. 8
56,
. 2
14
8 .
.9
66.
.4
9
43
-4 4
10.
.2
46.
.4
16
9.
.4
56.
. 5
13
8 .
. 0
57.
.8
14
9.
. 6
68.
.4
11
45
-4 6
6.
. 1
52 .
.7
16
7 .
. 6
57.
.7
14
6.
. 1
54 .
. 0
14
6.
.7
63 .
. 5
13
47
-48
6.
. 4
43 .
. 6
16
7 .
.4
58.
.8
15
7 .
, 1
55.
.8
14
8 .
.4
66.
. 2
13
49
-50
6.
. 1
46.
. 4
16
8 ,
. 2
56.
. 1
15
10.
. 0
46,
. 8
15
9.
. 3
62 .
.4
14
51
-52
5.
.9
39 .
.7
16
6.
. 6
58 .
. 4
15
7 ,
. 7
49 ,
. 0
15
8 .
. 8
59 .
. 5
13
53
-54
5.
.2
43 .
.5
16
6.
.2
61.
.7
16
6 .
. 2
38.
. 8
15
7.
. 3
60.
. 6
14
55
-56
6.
.2
41.
.3
16
7 .
. 3
50.
.8
16
5,
. 4
51.
. 1
15
7.
. 6
60.
. 6
14
57
-58
5 .
.9
38 .
. 0
16
6 ,
. 3
55.
.9
17
5,
. 6
42 ,
. 7
15
9.
. 0
58.
. 6
14
59
-60
4 .
. 7
42 .
. 7
16
6 ,
. 0
55 .
. 6
17
4 ,
. 2
44 ,
. 8
15
9 .
. 1
60.
. 3
14
61
-62
4 .
. 6
35.
.8
16
4 ,
. 0
57 .
. 0
17
5.
. 6
33 .
. 7
15
6.
.9
50.
.7
14
Nuc: Nucleated Cells; Leuk: Leukocytes.

APPENDIX D
BODY WEIGHTS OF MICROTUS (EXPERIMENT 2)
Body Weight
Body weights for the four species that were measured in
Experiment 2 are located in the following table (Appendix
D). The body weights were not analyzed statistically
because analyses were conducted on body weights in
Experiment 1. The data are presented for complete
descriptive purposes.
Male meadow voles and montane voles showed large
increases in body weight across the ten weeks of study
(changes resulted in 57.0% and 76.7% increases of week 4
values respectively); increases in body weight for male pine
voles and prairie voles were less (40.6% and 39.7%
respectively). Percentage increases in body weight for
females of all species were more similar in value (36.0%,
25.3%, 35.2%, and 36.6% for pine voles, prairie voles,
meadow voles, and montane voles respectively).
Patterns for gain in body weight for females generally
mirrored those of males, except the degree of weight gain
across the study was not as great and there was less
variation among species. Pine voles were clearly the
lightest species in the study.
229

230
Appendix D. Mean Body Weights of Microtus (in g)
During Olfactory Preference Study (Experiment 2).
Species
Pine Prairie Meadow Montane
Males
Week Number
(N=15)
(N=14)
(N=18)
(N=16)
Week 4
14.53
25.42
27.73
23.09
(0.63)
(0.93)
(0.77)
(0.88)
Week 7
19.24
31.55
37.77
35.16
(0.54)
(1.61)
(1.37)
(1.34)
Week 10
20.43
35.52
43.62
40.80
(0.55)
(1.61)
(1.53)
(2.11)
Females
Week Number
(N=15)
(N=16)
(N=18)
(N=18)
Week 4
14.37
24.01
24.85
21.21
(0.63)
(0.93)
(0.74)
(0.59)
Week 7
18.74
27.81
30.64
27.07
(0.63)
(1.43)
(1.00)
(0.87)
Week 10
19.88
30.08
33.60
28.97
(0.66)
(1.67)
(1.10)
(1.01)
Columns represent mean values (+ S.E.) for pine voles
fMicrotus pinetorum), prairie voles (M. ochrogaster^, meadow
voles (M. pennsvlvanicus) and montane voles (M. montanus).

APPENDIX E
ANALYSES OF CELLS FOR VAGINAL SMEARS
OF MICROTUS (EXPERIMENT 2)
Cliaiáctiristus of Cells in Vaginal Smears
Mean percentages of cell types and corresponding
analyses from the vaginal smears are summarized in Appendix
E-l for prairie voles, meadow voles, and montane voles.
Data for pine voles are not included because only 1 of the
15 females became perforate within the 10 week study and was
excluded from further analysis. Nonparametric tests
(Kruskal-Wallis ANOVA's and Mann-Whitney U tests) were
conducted for between-species comparisons.
The mean percentages of cell types during the first
test session (week 4) varied little between species, and did
not differ significantly. All species had smears dominated
by leukocytes (range 51% to 69%), slightly less cornified
cells (20% to 37%), and relatively few nucleated cells
(range 10% to 13%). However, by week 7, species differences
emerged. Meadow voles and montane voles had significantly
higher percentages of cornified cells, and fewer leukocytes
and nucleated cells, when compared to prairie voles during
week 7; no significant differences were detected between
meadow voles and montane voles. By week 10, meadow voles
had significantly more cornified cells than montane voles
and prairie voles, meadow voles also had significantly fewer
231

nucleated cells than the other species. During week 10,
prairie voles displayed significantly fewer cornified cells
than the other two species.
232
Repeated-measures analyses were conducted independently
for each species (Wilcoxon matched-pairs signed-rank tests).
The percentages of all cell types from prairie voles did not
differ significantly across the weeks of study. Thus,
significant species differences were due largely to changes
in the proportions of cells in meadow voles and montane
voles. Generally, the smears of these two species were
characterized by increasing proportions of cornified cells
with reductions in leukocytes and nucleated cells across the
study (see Appendix for additional post-hoc comparisons).

233
Appendix E-l. Percentages of Cells in Vaginal Smears of
Microtus (Experiment 2).
Prairie
Soecies
Meadow
Montane
Kruskal-Wallis
ANOVA
Week 4
(N = 8)
(N = 10)
(N = 8)
H(2, N = 26)
Cornified
20.24
37.33x
23.15x
1.95NS
Nucleated
10.25
11.58x
13.13
0.18NS
Leukocytic
69.51
51.09x
63.72x
2.64NS
Week 7
(N = 13)
(N = 15)
(N = 17)
H(2, N = 45)
Cornified
24.42a
55.87by
52.52by
9.66**
Nucleated
14.99a
5.56bxy
9.37b
8.34*
Leukocytic
60.59a
38.56bxy
38.llby
7.81*
Week 10
(N = 14)
(N = 15)
(N = 18)
H (2,N = 47)
Cornified
27.40a
67.91bz
55.62cy
18.35***
Nucleated
9.65a
3.40by
8.64a
12.27**
Leukocytic
62.95a
28.69bxz
35.90by
15.98***
Columns represent mean percentage values for prairie voles
ÍM. ochroaaster), meadow voles fM. pennsvlvanicus) and
montane voles (M. montanus).
Superscript letters (a,b,c) indicate results of post-hoc
Mann-Whitney U comparisons among species and are read across
a given row. Means with different letters differ
significantly (p < 0.05). The letters (x,y,z) indicate
post-hoc comparisons via Wilcoxon matched-pairs signed-ranks
tests and are read down a column for each cell type.
* _ _ _ * * * _
= p < 0.05; = p < 0.01; = p < 0.001.

BIOGRAPHICAL SKETCH
Allen Lee Salo was born in Minot, North Dakota, on
January 27, 1963. He and his family moved soon to
Manistique, Michigan, in the Upper Peninsula where he grew
to enjoy the outdoors along the shores of Lake Michigan. He
moved to Gwinn, Michigan, in 1978 where he graduated from
high school in 1981. He attended Northern Michigan
University and graduated with the Bachelor of Arts degree in
1985. His graduate study began in Omaha, Nebraska, at the
University of Nebraska at Omaha where he received the Master
of Arts degree in 1987. His graduate studies for the
doctoral degree continued at the University of Florida in
the Fall of 1987. He accepted a postdoctoral position at
the Medical University of South Carolina in Charleston.
234

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully-adequate, ÁrR scope and quality, as
a dissertation for the degree.oX .Doctor qf/Philosophy.
Donald A. Dewsbury,
Professor of Psycholij
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor x>f Philosophy.
'/?7ihoc <72
Marc N. Branch^
Professor of Psychology
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
H. Jane Brockmann„
Professor of Zoology
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
H. Kaufmann/
Eessor of Zoology
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
/ r
Carol Van Hartesveldt
Professor of Psychology

This dissertation was submitted to the Graduate Faculty
of the Department of Psychology in the College of Liberal
Arts and Sciences and to the Graduate School and was
accepted as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
December, 1992
Dean, Graduate School