Female preferences for unmated versus mated males in two species of voles, prairie voles (Microtus ochrogaster) and mont...


Material Information

Female preferences for unmated versus mated males in two species of voles, prairie voles (Microtus ochrogaster) and montane voles (Microtus montanus)
Physical Description:
viii, 144 leaves : ill. ; 28 cm.
Pierce, John David, 1959-
Publication Date:


Subjects / Keywords:
Sexual attraction   ( lcsh )
Sexual behavior in animals   ( lcsh )
Voles   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1989.
Includes bibliographical references (leaves 121-143).
Statement of Responsibility:
by John David Pierce.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001500395
notis - AHB3123
oclc - 21381661
System ID:

Full Text







Copyright 1989


John David Pierce, Jr.

This dissertation is dedicated to the memory of a true,

true friend and sense, Willard F. Day.


It is difficult to acknowledge all the people who

have provided assistance in the completion of this work.

I thank my committee members, Drs. Marc N. Branch, H. Jane

Brockmann, Donald A. Dewsbury, Sergio M. Pellis, and John

G. Robinson. Each has been extremely generous in sharing

time, ideas, and other valuable resources. I especially

thank my committee chair and major professor, Dr.

Dewsbury, for his invaluable contributions to all aspects

of my education. I have learned much from him.

I thank my fellow graduate students for their unique

contributions. In particular, Stephen A. Taylor has

always been there when I needed him. His help is

gratefully acknowledged.

My family has shown remarkable altruism (based on kin

recognition) and I thank them for that.

Finally, and most importantly, I thank my wife

Lorraine for all her love, support, and encouragement.

This is as much her accomplishment as it is mine.



ACKNOWLEDGMENTS...... ...............................iv

ABSTRACT.... ..................... .......................... ii

INTRODUCTION.... ...........................................1

Mate Choice............................................. 2
Methodological Issues in Preference Demonstration......4


Preferences for Conspecifics...........................9
Preferences for Members of the Opposite Sex...........20
Preferences for Different Reproductive Conditions.....23
Preferences for Dominant Individuals..................27
Preferences Based on Genetic Differences..............29
Preferences for Familiar and Novel Conspecifics........32
Preferences for Unmated versus Mated Individuals.......36
Other Preferences....................................38
Summary of Association/Mating Preferences.............40
Future Directions....................................48


Mating Systems of Prairie and Montane Voles...........51
Mating Systems and Mate Choice........................58
Summary..... ........................................... 65

GOALS OF THE PRESENT RESEARCH.............................66


General Method......................................... 69
Experiment 1 ..........................................75
Experiment 2..........................................78
Experiment 3...........................................79
Experiment 4........................................... 83
Experiment 5...........................................90
Experiment 6...........................................92
Experiment 7........................................... 93
Experiment 8........................................... 95
Experiment 9............................................99
Experiment 10........................................105

GENERAL DISCUSSION ................. ................... 107

Summary and Overview of Results......................107
Comparison to Preferences in Other Species........... 113
Preferences in Prairie and Montane Voles.............115
Preferences and Mating Systems........................119

REFERENCES................................ ................. 121

BIOGRAPHICAL SKETCH... ..................................144

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



John David Pierce, Jr.

August, 1989

Chairman: Donald A. Dewsbury
Major Department: Psychology

Mate selection has been the focus of much thought and

experimental inquiry in the field of animal behavior.

Many criteria have been identified as influencing mate

choice in rodent species. In the present study, one

aspect of mate choice, namely female preferences for

unmated versus mated males, was examined in two species of

voles, prairie voles (Microtus ochroqaster) and montane

voles (M. montanus). These two species were compared

because existing field and laboratory evidence suggests

that they may differ in their mating systems, with prairie

voles being generally monogamous and montane voles

generally polygamous. In a series of 10 experiments,

female preferences for unmated versus recently mated males

were examined using two different procedures. In tether

tests permitting copulatory behavior, prairie vole females


preferentially visited with unmated males over males that

had been mated for three ejaculations prior to testing

(Exp. 1). Prairie vole females preferentially associated

with anesthetized unmated males over anesthetized males

that had delivered three ejaculations prior to testing

(Exp. 3). When mated males delivered one, as opposed to

three, ejaculations prior to testing, prairie vole females

showed no preferences in tether tests (Exp. 2) and greatly

reduced preferences in anesthetization tests (Exp. 4).

Montane vole females showed few preferences during either

procedure regardless of the number (one or three) of prior

ejaculations by the mated male (Exps. 5-8). In

Experiments 9 and 10, females of each species were mated

to either fresh or already mated males to assess

detrimental effects on pregnancy initiation as a

consequence of mating with a depleted male. No

significant differences were noted for either species in

either probability of pregnancy or size of resulting

litters. Species differences in female preference for

unmated males parallel differences in other mating

preferences and may be related to mating system




Mate selection has been the focus of much thought and

experimental inquiry in the field of animal behavior.

Potential mating partners can vary considerably along

several dimensions related to mate quality (Halliday,

1983). Variations in characteristics such as physical

condition, dominance status, resources, and parental

abilities, among others, affect many parameters of

reproduction, including the number, survival, and quality

of offspring. A key tenet of evolutionary theory is that

animals should be selected for their ability to assess

potential mates with respect to these characteristics.

Discrimination and selection of the best potential mate

available should enhance an individual's reproductive

success. For this reason, the study of mate choice is an

important area of interest in animal behavior. In the

present study, one aspect of mate choice, namely female

preferences for unmated versus recently mated males, was

examined in two species of voles, prairie voles (Microtus

ochrogaster) and montane voles (M. montanus).

Mate Choice

In this section, mate choice is defined and discussed

in relation to mating preferences and active versus

passive choice.

A Definition

Halliday (1983) has defined mate choice as "any

pattern of behaviour, shown by members of one sex, that

leads to their being more likely to mate with certain

members of the opposite sex than with others" (p. 4).

This definition has the advantage of characterizing mate

choice solely in terms of observable phenomena. Further,

this definition makes clear that the relationship between

variations in individual behavior and variations in mating

success is the critical relationship underlying mate


Active versus Passive Choice

As Halliday's definition implies, mate choice need

not be based on specific mating preferences by the

choosing animal. For example, differences in the ease of

localizing the source of a call may lead to variations in

mating success. Alternatively, individuals may move

toward the source of a call perceived as the closest.

Parker (1983), O'Donald (1983), and others, have attempted

to clarify this issue by distinguishing between passive

and active choice. For Parker (1983), passive choice

occurs when individuals have no prior preferences for

particular characteristics, but are passively attracted to

the most intense source of a conspecific cue. Searcy and

Andersson (1986) proposed that passive choice be defined

as choice of the animal perceived to be closest to the

choosing individual.

In active choice, by contrast, choosing individuals

exhibit preferences for particular characteristics. In

these cases, individuals choose mates based on phenotypic

preferences shaped by natural selection. Whereas in

passive choice the choosing individual accepts any mate

perceived as most proximate, in active choice choosing

animals will reject certain individuals regardless of

their proximity. Thus, "evidence that females actively

reject some males in favour of others is needed to

establish the occurrence of active choice" (Partridge &

Halliday, 1984, p. 235).

Sullivan (1989) has argued that the distinction

between passive and active choice is inappropriate and

that the more important issue is the evolutionary

significance of mate choice. First, Sullivan argues, it

is difficult to distinguish between these two types of

mate choice, especially at a proximate level. For

instance, attraction to the more intense (Parker, 1983) or

most proximate (Searcy & Andersson, 1986) cue may

represent an active choice for proximity. Second,

instances of passive choice may still be shaped by natural

selection: "If we find that females prefer mates who call

frequently, we may still ask why the female's sensory

system passively responds in such a fashion" (p. 693).

Further, passive responding in this manner may still

result in differential mating success conferred upon

certain individuals, and these evolutionary implications

need to be addressed. Sullivan's approach emphasizes, as

does Halliday's (1983) definition, that the critical

relationship underlying mate choice is between variations

in individual behavior and variations in mating success.

Halliday's (1983) definition was adopted for the

present work. Following Sullivan (1989), no attempt was

made to distinguish between passive and active choice.

The word "preference" in the present paper is used as a

convenient descriptive term for associative or mating

behaviors directed toward particular individuals.

Methodological Issues in Preference Demonstration

Many studies of mate choice have involved

experimental manipulations in a laboratory setting.

Although such manipulations provide a powerful means of

investigation, there are methodological concerns which can

affect the demonstration of mating preferences. These


concerns are addressed in this section through a

consideration of stimulus presentation and response


Stimulus Presentation

Several methods have been used for the presentation

of stimuli in an experimental study of mate choice. In

most laboratory studies of mating preference, the test

subject is placed in an arena where two or more stimuli

are presented simultaneously. Stimuli can be presented

singly in a sequential manner, but such procedures are

less sensitive and more variable than two-choice

preference tasks in which two stimuli are presented

simultaneously (Johnston, 1981). Forced-choice paradigms

in which the subject must be in an area associated with a

stimulus do not allow an assessment of whether the subject

is attracted to one stimulus or repelled by the other

stimulus (Doty, 1975). Accordingly, most preference tasks

incorporate a neutral area.

Stimuli presented in a preference task may be live

animals, anesthetized animals, or odors collected from

animals. Each type of stimuli has its own particular

advantages. The use of odors as stimuli allows the

opportunity to isolate and test a specific sensory cue

without the effects of other sensory modalities. Use of

live animals has at least two advantages. First, live

stimuli allow for presentation of cues from more sensory

modalities. Second, use of live animals permits the

recording of different measures (e.g., copulatory

measures) to provide a more powerful means of assessing

preferences. Anesthetization procedures are useful in

isolating specific sensory cues while controlling for

extraneous behavioral cues by the stimulus animals.

The mode of stimulus presentation is more than a

question of convenience. In some cases, the way in which

stimuli are presented can influence the outcome of

preference testing. For example, male golden hamsters

prefer odors from estrous females over diestrous females

(Johnston, 1980), but apparently do not show this

preference when live animals serve as stimuli (Johnston,

1980; Kwan & Johnston, 1980). Preferences are most

accurately assessed when a combination of stimuli

presentations is used.

Response Measurement

A number of different measures have been used to

assess preferences. Measures taken to evaluate

preferences may be time spent in association with a given

stimulus, copulation with each stimulus animal, or

specialized behaviors (e.g., sniffing, urine marking,

solicitation behavior) directed towards the stimulus.

Ideally, mating preferences should be assessed by a

methodology that permits copulation as a measured response

(Johnston, 1983). Such a procedure is advantageous in

determining that preferences directly result in mating

behavior being directed toward a particular partner.

However, a laboratory choice test may preclude the

expression of a preference because of the proximity of and

the ease of copulating with both potential mating


An alternative approach is to assess association

preferences when copulation is prevented (e.g, by a wire

barrier) or impossible (e.g., use of anesthetized animals

or odor cues alone). In such studies, association

preferences are assumed to reflect mating preferences.

Studies using association measures have the advantages of

controlling for animal differences extraneous to the

variable under study and of allowing for a more precise

analysis of a specific criterion for mate choice. The

validity of such an approach in assessing mate choice is

dependent upon the relationship between association

preferences and mating preferences, a relationship not

fully established. Ideally, both approaches should be

used in conjunction to assess mate choice.


Association and mating preferences have been

established for many species by these methods and have

been the subject of much recent literature (cf., Bateson,

1983). There is now strong evidence that females play an

active role in the initiation and maintenance of

copulation (Doty, 1974) and do exert active choice in mate

selection (Partridge & Halliday, 1984). Evidence is also

accumulating that males of many species exhibit mating

preferences as well (Dewsbury, 1982a). The mating

preferences of muroid rodents are the subject of the next



Many criteria that form the basis for association and

mating preferences have been identified for muroid

rodents, a diverse taxonomic group numbering over 200

genera in three major and two minor families (Dewsbury,

1984). Olfactory cues, in particular, appear to be

critical, as preferences have been exhibited for several

classes of mammalian social odors (Brown, 1979). However,

preferences have been established for a number of non-

olfactory cues as well. Table 1 contains a listing of the

muroid rodent species which have been studied for mating

preferences and which will be discussed in this section.

Preferences for Conspecifics

A fundamental problem for animals selecting mates is

identifying conspecifics. Species-typical odors, odors

"characteristic of all members of a species or subspecies"

(Brown, 1979, p. 104), have often been shown to influence

mate choice and association preferences. Species

exhibiting a preference for conspecifics versus

heterospecifics are listed in Table 2. With few

Table 1
Muroid Rodent Species in which Association or Mating Preferences
have been Established

Common Name

Scientific Name

Montane voles

Prairie voles

Gray-tailed voles

Meadow voles

Common voles

Bank voles

Brown lemmings

Collared lemmings

Golden hamsters

Romanian hamsters

Turkish hamsters

Indian desert gerbils

Mongolian gerbils

House mice

Laboratory rats

Pygmy mice

S. grasshopper mice

Deer Mice

White-footed mice

Mexican mice

Microtus montanus

M. ochrogaster

M. canicaudus

M. pennsylvanicus

M. arvalis

Clethrionomys glareolus

Lemmus trimucronatus

Dicrostonyx groenlandicus

Mesocricetus auratus

M. neutoni

M. brandti

Meriones hurrianae

M. unguiculatus

Mus musculus

Rattus norvegicus

Baiomvs taylori

Onychomys torridus

Peromyscus maniculatus bairdi

P. leucopus

P. polionotus

Table 1-- continued

Common Name

Scientific Name

California mice P. californicus

Cactus mice P. eremicus

Wood rats Neotoma micropus

Desert wood rats Neotoma lepida

Oc 00 co co co co

a% ON 0) ON H

o 0 0o 0 > > > H H cc
SH H 4 H 0 0H 0 0 : ,.O O.
00 ( 0 0 (>0 > > CO Z

0C C C C C4 C CO C4 J C C H
F-4 r-4 rH r)l 0r l a) H

a (o i- O O W O O 4O O p
-- 0 0 0 0 4 4

OO r ,I I 0 0 0l 0 0 U) )l >
+ 0 C C C U U 0 L1 H H H 3

S- 1, .4O o o4 14 P 4 0 o o
4 o ) E D to a a )rl i a E ( ul na
p 0 0 0 0 0 0 0 0 0 H 0 0 0

) i ) U ) i ) O U 0 0 u U U u)

H (11 I 0 0
H 4 4 w o ( )
0) t 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0) u M u a u U u a u R 0 u U U U

)0 ) -4 I 1 1 0
o U) U)

) (0 0
-r U 4) V r- r H
S(0 v H H 0> 0 U) 0 0 U M as
Srl > 0 0 > > > U U U)

44 U) rq *4 -- > f rHq rHl -44I *H
U 0 0 40 04 0 4 ( O (l< 00 0

0 0C H I (C (0 0 C0 (- -i ( { 0 0 5-

S- 0 0r Pk

0 0 CO C C C C C CO C
c0 0 0


4 r-4 r-1 W- )
0 0 r-1 r-4
S> r-4 > 0 0 0
-4 > > r

P 0 0
U P >4 e
U)) 0 to
0 1 0 0 0 U U) 0

m o 0 0
CO (1 <>O O
s (0 to r
0 0 0 0 0 0

0 o C0 0 .
c cW co c -, -
m ON ON C) 1

H,- ( H H H-. 9- 4- UN 4J
U4 U U 0 N 04 0 4
,, a ~

0 0 0 0 H H H H C ) ,4

o o oo 0 0
00 0 0c4 C) P c
C) C C C

in ia i iE n En W w U) En Ei n M) to En M C
14 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ct g a En

C U U U U U U U U U U U 0 U 0

U 9 r. 3 U ) U Ul ) U )

.H 4 ) C C -ri WC C )0 4 4- 4 C

4)4 > r I M d O 0 ( (-4
0 4J 4) ( U U U U 0 p M w

4 H 44 4 It 4 C O 0 4 4.)
M 0 0 ) 0 ) to U U w 0 0 0 .
0 4 ()) ) 4 a 4 94
4 ) I *H 0 0 0 0

0 C C H H p 0 0 :) C) 0 C) ( >1 C C
0 0 P H 4. 4. 4) r4 "i4 -4 U4 4
u) 4.1 r-4 (i) C)q C) U) U) r U 00 -4 r-4
oo ) C) C ) to H A s to to 0 4. 4) 4 0 t

4 0 4 0 0 0 0
) 0H 0 0 -4 -0 H q I I >H >H
0 0 O O O i O 330 O 0 0 0

H 0 o0 0 0 0 0 0 0 0
1-. (G <0 03 0 0 0 0 0 0fl i0
x 0 0 0 0 -4 0 4 H -H *D.4 ...4 *- *'4 *z.4 *zPz

U) U

4) C) ( .

4.) < in 0 U)
H 0 0 Cq

Ci M ) rl 0
C) q mC C)
V r0 .l U)
O H 0 4 M
0 0 0 0 0 0


N 0 N
o 0 0 0 0\

m I'D > m N 0 m > e w %
ori 4i C c

0 0
H H i*H
4 r. ) 4 ) .- 4 )

S 0 0 4 0 03
M M 0 0i E O

a a a 04 04
> t-i o C i C C o C o r
-4 -4 O 4.li 4.l >

C 0 0 0 0 0 0 0 0 l 0 0 0 0 0

z z +z U u u u z z u z z z
H H (D 10 L0 (0 0 0 0 ) L0

>q H U4

0) i H 4 4-) 0 0
0 0 0 0 0 0 0 4 U0 0
000 0 0 0 0 0

-4 P*4 *-I

Qo (a o a o

0 0 0 0 0 0 -HO.qO0
4 0 4 O 0 0 0 0 0 O 0 0 O
S*H 4 44 r *
O ap O e

0- H U 0 0 U O.
0Q0)0 4 X -4 4 *-4 X 4 a) 4 0 0

0 4 1
S 0 0 0 0 0 0 0 0 0 4 0 0 0


0) U

UH 0 0

00 US X
C. 4 Q S S

0 0 H H

c c o cc 1-1 r- -
O O Ch O0 U C
SH 44 r- O H

a4a A a 0 0

C C > >
tC C0 k M Q
4J 4J 4 4 4 4- '4 0 0 U 0 4
0 0 0 0 0 0 0
E MQ o oa a a W m

S 0 0 z 0 0 c
0 U U U U U K O O U U U

0 4 0 U U U 0 O4 0 0

4-4 > "4 -H4 rH ".q

0 ( U u U (0 0 (0 (D 4) 4- ) 4
*c-: "q q r -H ui 4- u u 0 0 V ui

to p. R a.4 "A 0 0 0

4-) 4 4 4.) 4. -.4 --4 -H U) UM 4 4O 0
S0 0 0 0 0 0 0 0 40 0 0

r-I r-& *4 -4 5. 4 5-1

M 4 -4 -4 ".4 4 O- ".4 0 0 00
4 0 0 0 0

MI ) o

D W>*

0 4
S4- 0
0 IV 9:g 4

l 4 3 0 0 0 I I

) r-4 5-i rq 0 3
+* 4". 4J II

U) H U *i O
00 0 0

exceptions, animals have demonstrated clear preferences

for members of the same species.

Two main factors have been identified as modifying

conspecific association preferences. First, results from

cross-fostering work have demonstrated that these

preferences are influenced by early experience. Table 3

contains a listing of cross-fostering studies in which

species preferences have been evaluated. Not all species

studied have established preferences for foster species.

In general, however, most cross-fostered animals appear to

show a preference for the foster species or a significant

reduction in preferences for conspecifics. D'Udine and

Alleva (1983) have reviewed cross-fostering studies from

several rodent families. Their conclusions are similar to

the present ones: "Early experience resulted mainly in a

reduced preference for conspecifics and in an enhanced

preference for the cross-fostering species" (p. 312).

Foster preferences, like conspecific preferences

(Porter & Doane, 1979), are established early.

Preferences for foster species are evident as early as two

days of age in spiny mice reared by house mice (Porter,

Deni, & Doane, 1977). LeRoy, Roy, and Briley (1982)

showed an effect of cross-fostering in 15-day-old house

mice and an effect of the general rearing environment

(colony odors) in 36-day-old house mice.

Second, conspecific preferences in several species

are influenced by the estrous state of the female. In

co cc 0c 0 c
r- r, r- r- 0

oco oo oll 0 ON

H H H H 0) 0* 0
f 0 i4 4 4 (0 (*3
S- O > C

0 r 3 o O 0 0 0 0 H > > 4 -1
4H Jl rH a O L a-I m
M I Ul M ^ X rIl 4 rl (L) 4 0) r

9C C C9 C9 H H H r .4
O O O O 0 0 (0 a c

0 Cy G
o o 0 0 > > O H 0 O

(1) ) W J ) 4 ) a)
0 0 0 P4 O O 0 0 41 (0 a
0n Z m ^ ,! cc cc 01

+ e a C C C C O (T> +J

4-4 O I (1) -L O
0 > H H 0 0 0 0 H P 4H 4 4-)
0 0 VI Q VQ V V V U O0

V*O H H H H 0 0 O C

o 1 0 0 0 0 '4 >., >4 >4
.cc -4 -4 a rc u a) 01 0 0 0
o (d co a) 0) Q) 0) o1) 4) o-o r4 rl 4-) 41 4J 1'

0 0 4 0 0 0 0 0 0 0 0 C 0 0 0 0

r-4 U 4 (U U U U C H O 0 I 0

O ( 0 ( O 0 o 0 (a ( o 0
t) 0 1 0101 0l 01: 0 0 0


4-) 4 4 4 r z
01 C C CD C C t

% *0 Q 0 ) U M O O O + +)

H ( > > 4J w u u E > P
) ) 0% 0) L l C Ca 4- 4

4J 0O(1) r 4 0 to 4)
-U > 4 0 0 4
0 # m C > & m
0 1 0 01 0 0 0 4
SV V H H H H 04 r
0U 0 0 0 01 r l 0rl 1 H H. 0 l
01 4. 4.) .) o o o o 0 0 g

l rl rl > > > > j0 k k l
oI 01 m 0 5i 4 0 0 C C -' 0 5- 5 5 >i
00 0 >i >, 4. 4' -'.-H O H H 3 3 -4 0 0 0 6
( 4 0 0 O O 0 H O 0 0

-H 3 0 0 M 0 0 0 E 4 i 0

M *-4 < s< r' 0 0 0 0 < '

-'.4 011

-4 a) 0' 0c 4 4
5-s 01 CO 4. CO

I 0 4CO
0W C01 0

(04 0 0 5 s 0 0 3 0
-iu w s u s o U u ^ ac

m 0 (ON 0 0 -
S-4 r- r- r- -,
SH H H cO O O CO p O O
S1 5 00 H 3 3

>4f I I
-4C O O TI C C rO 4J 4J 44
m m fu (0 (0 to 0 0 0 0

0 44 44 4 0 0 % 0 % O

0 M M 0 9 z O > > O u u u u
4 O 10 r. M
0 r 1 >1 0 (0 p) p p p $

3 -r4 -4O 3 O t O MC U O U
(Y x .-i oY ci u u u o z z

U .4 $ 54 $4
a ) 2 4 U) 4) u4 0)
U 4-) (U U) 4-) En 4-) WU U 4Ji 4-) U) U) 4-) U)
^ M U) z En z r. En (n U r r. Ut r
) 0 0 0 0 0 0 0 0 0 0 0 0 0 0
4-4 f4 z z rL u P z 4 D 4 z z O< z

4) U) (1) 4 4
U) US) -4 *4

S 0) D 0 (0 4 U 0 0 O
U p4 41 0 0 4 4 4U
0at H q 0 0 0 0 w 0 0

R 4 4 4) 4 4 *WI W IM
U tt 0 0 1 1 (0 AS 9z
U) p) U) 4U) u) 4 P 0
M) >p pU Uo) U U 'l t0 4.) 4.) 0 0 U

) 0 0 ( UQ U 0 0 CI I (i 14

) 0 O 4 4 4 4 4 4 -4 4 P Ou t U
o l U Q) H 3 U4 U H H-i -r-i *
1 ) o 0 (0 tt M 0 M

H 0 0 0 0 0 0 0 p p
3 0 --I -4 *4 *H- 4-4 *- 0 0 0 0
0 c z c o 1V
-, 0 0 <4 4 4 4< 0 4 0 0 0 0


)H *t o

0 U 0
0 U --4 -4 0

a) z (0 Uq 0
0 0) > 00 0
o) UP U I +i
'I, I- t r
U U U) a aa e

deer mice, estrous females, but not diestrous females,

preferred conspecific males (Doty, 1972). Estrous state

in females used as stimulus animals can influence male

preferences. When given a choice between conspecific and

heterospecific females, prairie and meadow vole males

showed no preferences when females were diestrus but

significant preferences when females were estrus (Pierce,

Ferguson, & Dewsbury, 1989). In contrast, it appears that

estrous state does not influence conspecific preferences

by males in two species (brown and collared) of lemmings

(Huck & Banks, 1980a; see also Huck & Banks, 1980b),

cactus mice (Carter & Brand, 1986), and California mice

(Carter & Brand, 1986). It is likely that estrous

condition affects the display of a preference rather than

the ability to discriminate between conspecific and

heterospecific stimulus animals, although no evidence yet

exists to support this hypothesis.

There is evidence to suggest that conspecific

preferences can be influenced by adult experiences. Work

with strain preferences in house mice suggests that adult

social experience may be necessary for the expression of a

conspecific preference (Albonetti & D'Udine, 1986).

The work discussed here has demonstrated conspecific

preferences in laboratory settings. Field investigations

of this phenomenon are scarce. Stoddart (1986) reported

that wood mice and short-tailed voles preferentially

entered conspecifically scented traps over either odorless

or heterospecifically scented traps. These trapping data

supplement the existing laboratory evidence and suggest

that preferences for conspecifics are a real field


Preferences for Members of the Opposite Sex

Like conspecific preferences, preferences for members

of the opposite sex are a fundamental component of mate

choice. Preferences for heterosexuals have been assessed

by allowing a test animal a choice between 1) a male and a

female conspecific, 2) the odors alone from male and

female conspecifics, or 3) the odor of an opposite-sexed

animal versus a clean, no-odor choice. Both males and

females of almost all species studied have displayed

heterosexual association and mating preferences.

In tests using opposite-sexed conspecifics as

stimuli, preferences have been shown by female laboratory

rats (de Jonge, Burger, van Haaren, Overdijk, & van de

Poll, 1987; de Jonge, Eerland, & van de Poll, 1986; Dudley

& Moss, 1985; Gilman & Westbrook, 1978) and male and

female golden hamsters (Beach, Stern, Carmichael, &

Ranson, 1976; Carmichael, 1980; Landauer, Banks, & Carter,

1977, 1978). Preferences based on odors from opposite-

sexed animals have been demonstrated in male prairie voles

(Taylor & Dewsbury, 1988), female montane voles (Sawrey &

Dewsbury, 1987), male and female meadow voles (Ferkin &

Seamon, 1987; Ferkin, 1988a), male deer mice (Dewsbury,

Ferguson, Hodges, & Taylor, 1986), male and female wood

rats (August, 1978), male brown lemmings (Huck & Banks,

1980a), male collared lemmings (Huck & Banks, 1980a), and

male golden hamsters (Landauer, Banks, & Carter, 1978).

Comparisons between a heterosexual odor and a control

or no-odor condition have shown generally consistent

preferences for the heterosexual odors in several species

including deer mice (Doty, 1973 for females but not for

males; Dewsbury et al., 1986 for males and females),

golden hamsters (Gregory, Engel, & Pfaff, 1975; Johnston,

1974; Kwan & Johnston, 1980; Murphy, 1973 for males;

Johnston, 1979 for females), and male laboratory rats

(Lydell & Doty, 1972), but not in either male or female

white-footed mice (Doty, 1973).

Drickamer (1984) examined capture rates of deer mice

and white-footed mice in traps baited with either male or

female odors and noted a "strong heterosexual odor

preference" (p. 701) by males and females of both species.

Similar findings from trap entry data have been reported

for short-tailed voles by Stoddart (1986).

There is evidence to suggest that social and sexual

experience influences heterosexual preferences. Sexually

experienced male laboratory rats preferred female urine

odor over distilled water odor, whereas sexually

inexperienced males did not (Lydell & Doty, 1972). An

influence of sexual experience has also been noted for

male laboratory rats (Pfaff & Pfaffmann, 1969), female

laboratory rats (de Jonge et al., 1987), and male montane

voles (Sawrey & Dewsbury, 1987). Further, socially

isolated male laboratory rats exhibited decreased

preferences for investigating female odors compared to

group-housed male laboratory rats (Brown, 1985). In

contrast, there appears to be no effect of sexual

experience in golden hamsters (Gregory et al., 1975;

Johnston, 1974; Landauer et al., 1978), collared lemmings

(Huck & Banks, 1984), or brown lemmings (Huck & Banks,


Similarly, reproductive condition may play a role in

heterosexual preferences in laboratory rats.

Gonadectomized laboratory rats of both sexes showed no

preferences (Brown, 1977, 1978). In ovariectomized female

laboratory rats, injections of testosterone propionate or

estradiol benzoate significantly increased partner

preference for males over estrous females (de Jonge et

al., 1986). Meadow voles of both sexes exhibited

preferences for heterospecific odor during the breeding

season. During the nonbreeding season, females preferred

the odors of other females, and males showed no preference

(Ferkin & Seamon, 1987). Juveniles of this species

preferred opposite-sexed adult conspecifics, regardless of

the season in which they were tested (Ferkin, 1988b).

Preferences for Different Reproductive Conditions

Male Preferences for Estrous Females

The data concerning male preferences for estrous over

diestrous females have recently been reviewed by Taylor

and Dewsbury (1988, in press). These authors note that

relatively few species have been studied and species

differences exist: "Males of some species do not prefer

estrous females, and males of other species display

preferences only under particular conditions" (Taylor &

Dewsbury, in press, p. 3).

Species in which males have shown a preference

include laboratory rats (Carr, Loeb, & Dissinger, 1965;

Carr, Wylie, & Loeb, 1970; Lydell & Doty, 1972; Merkx,

1984; Merkx, Slob, & van der Werff Ten Bosch, 1988; Pfaff

& Pfaffmann, 1969; Stern, 1970; but see Brown, 1977),

desert woodrats (Fleming, Chee, & Vaccarino, 1981), Indian

desert gerbils (Kumari & Prakash, 1984), Mongolian gerbils

(Block, Volpe, & Hayes, 1981), house mice (Hayashi &

Kimura, 1974; Rose & Drickamer, 1975), brown lemmings

(Huck & Banks, 1984), collared lemmings (Huck & Banks,

1984), and prairie voles (Taylor and Dewsbury, 1988).

Males of two species, deer mice (Dewsbury et al., 1986)

and montane voles (Taylor and Dewsbury, in press),

apparently do not prefer estrous females over diestrous


Male golden hamsters have not shown preferences when

the choice is between an estrous and a diestrous female

(Johnston, 1980; Kwan & Johnston, 1980; Landauer et al.,

1978), but do show a preference when odors from soiled

bedding are used as the stimuli (Johnston, 1980; Huck,

Lisk, Kim, & Evans, 1989), or if the estrous female

receives physical contact from another male prior to

testing (Carmichael, 1980). Male golden hamsters

discriminate and prefer an intact female over a

vaginectomized one both when odors cues alone are present

and when the stimulus females are present but physical

contact is prevented (Kwan & Johnston, 1980). Even in

this study, male golden hamsters did not exhibit a

preference for estrus versus diestrus in either intact or

vaginectomized females.

There is evidence to suggest that prior sexual

experience by the male may be necessary for the

demonstration of a preference in most, but not all,

species (see Taylor & Dewsbury, 1988, in press).

Reproductive condition provides a similar influence.

Castrated male laboratory rats did not show any

preference, but implants of dihydrotestosterone (DHT)

restored estrous preferences in such laboratory rats

(Merkx, 1984).

Estrous preferences appear to be at least partially

mediated by vaginal bacterial flora. Estrous female

laboratory rats treated with a vaginal antibiotic solution

were less attractive to males than were untreated estrous

laboratory rats (Merkx et al., 1988).

Female Preference for Intact Versus Castrated Males

Female preference for males of different reproductive

condition has been assessed in five species by giving

females a choice between intact and castrated males.

Female house mice preferred intact over castrated males

when odors alone provided the stimulus (Scott & Pfaff,

1970) and when intact and castrated males were used as

stimulus animals (Hayashi & Kimari, 1978). Female montane

voles, when presented with tethered intact and castrated

males, preferentially associated and copulated with the

intact males (Webster, Williams, & Dewsbury, 1982).

Estrous Mongolian gerbils likewise preferred intact males

over castrated males (Agren & Meyerson, 1977, 1978).

Female golden hamsters preferred the odors of intact males

over the odors of castrated males and of females

(Johnston, 1979). This preference was expressed most

strongly by estrous females, but was also shown by

diestrous and lactating females. Pregnant females showed

"a high degree of generalized interest in odors of

conspecifics but are not differentially attracted to

potential sexual partners" (Johnston, 1979, p. 33).

Laboratory rats have been the most studied species

with respect to this preference. Experienced females of

this species, regardless of their estrous state, can

discriminate (Carr & Caul, 1962) and prefer intact males

(Drewett, 1973; Dudley & Moss, 1985) and their odors

(Brown, 1977; Carr et al., 1965; Taylor, Haller, & Regan,

1982; Taylor, Regan, & Haller, 1983) over castrated males.

Inexperienced females prefer intact male odors only when

they are in estrus. This relationship has been

demonstrated in adult (Carr et al., 1965) and immature

(36-54 days old) female laboratory rats (Carr, Wylie, &

Loeb, 1970). Ovariectomized female laboratory rats fail

to demonstrate any preferences for intact males (Brown,


Female preference for intact males is almost surely

mediated by androgen levels of the stimulus animals.

Drewett and Spiteri (1979) injected castrated male

laboratory rats with either dihydrotestosterone (DHT-

castrated males smell like intact males), DHT and

estradiol benzoate (EB- castrated males smell and behave

like intact males), or nothing. Preferences assessed by

speed of approach in a straight runway apparatus indicated

that DHT-EB males were preferred over DHT males, and DHT

males were preferred over castrated controls. In another

study, exogenous injections of testosterone propionate

(TP) increased the attractiveness of castrates and

eliminated a female preference for intact males (Taylor et

al., 1982). The amount of TP injected affected

preferences; female laboratory rats consistently preferred

to urine mark in the area vacated by the male with higher

testosterone levels (Taylor et al., 1982). These studies

support the conclusion that females are choosing based on

"changes in androgen-dependent urinary byproducts" (Taylor

et al., 1983, p. 43).

A related female choice that has been examined is a

preference for sexually experienced males over sexually

naive males. To date, such a preference has only been

reported for female laboratory rats in studies using

stimulus animals (Gilman & Westbrook, 1978; Hill & Thomas,

1973) and odors alone (Taylor et al., 1983). Female

golden hamsters showed no preference for sexually

experienced males over naive ones, but did choose males

primed by contact with an estrous female immediately prior

to testing over unprimed males (Carmichael, 1980).

Sexually inexperienced prairie voles showed no preference

when given a choice between a tethered sexually

experienced and a tethered naive male (Pizzuto, 1988).

Examination of female preferences for experienced males is

a rich topic for future study.

Preferences for Dominant Individuals

Females of several species prefer dominant males or

their odors over subordinate males. Typically, males are

classified as dominant/subordinate based upon the outcome

of staged aggressive interactions, then females are

allowed a choice of a dominant and subordinate male (or

their odors) for association and/or copulation. In

studies using tether choice testing, preferential

association and copulation with the dominant male have

been demonstrated in laboratory rats (Carr, Kimmel,

Anthony, & Schlocker, 1982), brown lemmings (Huck & Banks,

1982a), golden hamsters (Brown, Humm, & Fischer, 1988;

Huck, Lisk, Allison, & Van Dongen, 1986), and prairie

voles (Shapiro & Dewsbury, 1986), but not montane voles

(Shapiro & Dewsbury, 1986). Female house mice associated

preferentially with subordinates but copulated

preferentially with dominant males (Parmigiani, Brunoni, &

Pasquali, 1982). In experiments where copulation was

prevented, preferential visiting of dominant males has

been shown by female prairie voles (Shapiro & Dewsbury,

1986) and female bank voles (Hoffmeyer, 1982), but not

female montane voles (Shapiro & Dewsbury, 1986).

Preferences for dominant males based on odor cues alone

have been shown by brown lemmings (Huck & Banks, 1982b;

Huck, Banks, & Wang, 1981), bank voles (Hoffmeyer, 1982),

house mice (Jones & Nowell, 1974), and golden hamsters

(White, Fischer, & Meunier, 1984a, 1986; Huck, Lisk, &

Gore, 1985).

In two species, brown lemmings (Huck et al., 1981)

and golden hamsters (White, Fischer, & Meunier, 1984b),

estrous females preferred the odors of certain males

before dominance testing was begun. Preferred males

subsequently emerged as the dominant males following

testing. This effect is most probably mediated by

androgen by-products in the urine as several researchers

have established that urine odors, in some way, indicate

social status (e.g., Jones & Nowell, 1973, 1974; Lombardi

& Vandenbergh, 1977; Maruniak, Desjardins, & Bronson,

1977). Dominance may also be indicated by the increase in

scent marking by dominant males (Huck et al., 1985) and

the reduction in scent marking by subordinates

(Desjardins, Maruniak, & Bronson, 1973).

As is the case for many preferences, the estrous

condition of the female influences the preference for

dominant males. Whereas estrous brown lemmings preferred

dominant males, diestrous females preferred the odors of

subordinate males (Huck et al., 1981). Similarly,

diestrous golden hamsters preferred the odors of

submissive males (White et al., 1984a), although a

replication of this work did not confirm the initial

findings, as diestrous females preferred control odors

over either dominant or subordinate odors (White et al.,


Preferences Based on Genetic Differences

The optimal balance between inbreeding and

outbreeding has been discussed by Bateson (1983). Bateson

noted that while the evidence for optimal outbreeding is

"still relatively meagre" (p. 273), an analysis of the

costs associated with inbreeding and outbreeding suggests

a great importance to finding a mate that is optimally


Dewsbury (1988a) has recently reviewed the literature

on kinship-based mating preferences. Five paradigms for

the study of this aspect of mate choice were presented,

but it was noted that the most valid paradigm, choice

tests permitting actual copulation, has received very

limited use.

Male house mice (D'Udine & Partridge, 1981; Gilder &

Slater, 1978; Hayashi & Kimura, 1983; Winn & Vestal, 1986;

Yanai & McClearn, 1972) have exhibited preferences for

unrelated individuals (see Dewsbury, 1988a). Barnard and

Fitzsimons (1988) showed that male house mice preferred

individuals less closely related than half siblings.

Whether female house mice show a similar preference

is difficult to assess. In studies where female house

mice did show a preference, the estrous state of the

female was artificially induced via the injection of

exogenous hormones (Winn & Vestal, 1986; Yanai & McClearn,

1972) or not assessed (D'Udine & Partridge, 1981; Gilder &

Slater, 1978). Hayashi and Kimura (1983) and Barnard and

Fitzsimons (1988) noted no female preference in this

species but again estrous state was not controlled in

either study.

Mongolian gerbils (Agren, 1984a) and common voles

(Bolhuis, Strijkstra, Moor, & van der Lende, 1988) have

been reported to show a preference for unrelated

individuals. In contrast, male laboratory rats preferred

female siblings in the work of Deni, Vocino, and Epstein

(1980). Much more work is needed in this area before

definitive conclusions can be reached (see Dewsbury,


Incest avoidance and reproductive suppression provide

supporting evidence that animals prefer to mate with

unrelated individuals. In laboratory studies, sibling

pairs housed together generally show a reduction in

breeding compared to non-sibling pairs housed together.

Species in which reproductive suppression and inbreeding

avoidance have been demonstrated include prairie voles

(Batzli, Getz, & Hurley, 1977; Carter & Getz, 1985;

Gavish, Hofmann, & Getz, 1984), gray-tailed voles (Boyd &

Blaustein, 1985), pine voles (Schadler, 1983), California

voles (Batzli et al., 1977), deer mice (Dewsbury, 1982b;

Hill, 1974), cactus mice (Dewsbury, 1982b), white-footed

mice (Wolff & Lundy, 1985; Wolff, Lundy, & Baccus, 1988),

Mongolian gerbils (Agren, 1981, 1984b), laboratory rats

(Hepper, 1983 [quoted in Hepper, 1986]), and house mice

(Yanai & McClearn, 1972). Meadow voles (Batzli et al.,

1977) and Djungarian hamsters (Ferguson & Dewsbury, 1987)

apparently do not show inbreeding avoidance.

In addition to mate choice based on kinship and

degree of relatedness, muroid rodents may base a choice of

mates on other genetic factors. Work with inbred house

mice suggests that the major histocompatibility complex

(MHC) of genes influences mate choice in this species (see

Beauchamp, Yamazaki, Bard, & Boyse, 1988; Yamazaki et al.,

1976, 1988). Males of most strains prefer estrous females

of different MHC types; a preference that may maintain MHC

heterozygosity (Beauchamp et al., 1988) and is affected by

rearing experience (Yamazaki et al., 1988).

Mating preferences of house mice may also be

influenced by the T locus. In tests of social and

association preferences, male and female house mice showed

consistent preferences for individuals not carrying the

lethal t-allele (Lenington, 1983), regardless of their own

genotype (Egid & Lenington, 1985; Lenington & Egid, 1985).

Preferences based on t-alleles and MHC genes are the best

examples of choice based solely on genetic quality of

potential mates.

Preferences for Familiar and Novel Conspecifics

The complexity and variety of mammalian chemical

signals are sufficient to provide individuals with unique

odors that provide cues for individual recognition

(Johnston, 1983). Individual recognition of conspecifics

plays an important mediating role in the social

interactions of many species. Dominant-subordinate

relations, territorial interactions, parent-sibling

relations, and mating encounters are affected by prior

experience with the same individuals and the outcomes of

these interactions may be contingent upon individual

recognition. The role of odors in individual recognition

has been summarized by Brown (1979) and Halpin (1980) who

note the many mammalian species in which there are

recognizable individual differences in body odors of


The typical laboratory paradigm for assessing

individual recognition involves testing subjects for a

preference for familiar versus novel heterosexual

conspecifics. Table 4 provides a listing of species in

which individual recognition has been studied with this

method. The data are inconsistent in many instances. For

instance, depending on the study, female laboratory rats

either prefer the familiar male (Carr et al., 1979),

prefer the novel male (Krames, 1970), or show no

preference at all (Carr, Krames, & Costanzo, 1970; Krames

& Mastromatteo, 1973). Data from male collared lemmings

also are inconsistent (see Table 4), although this

discrepancy may be attributable to different testing

conditions. Further, there are interspecific and

intersexual differences in the display of preferences.

There may be differences in male and female mating


* i




















a) a) a) 0)
rr4 r-l r r-

) a) a) 0 ) 0
r4 (4 S Z h S







i41 ,--4 r1 ,-I 4-*

-4 *- -4- -4-
4 I 4


4 0 4<
C a C



r l
to )

4) r)

4 (0 r-4
x u P4








a) Q) a) Q)
0 w0 0 0
0 0 0 0




0 0 0 0
r~z zi

0 0
o C
0 z

2 2

0 0
0 0

0) 0 0) 0

U) a) Q ) a)
4 fj4 r4 r4
,-I ,-. ,-

1.4 .4
0 (0 ( 00
M* -4 *M-
l-I r- r -4 l-i
* ll 0) -4 -r a) a)
(0 0 0 A0 0 0

.-f -r H 0 0 "C
C C C 3 0 4
4 04 0
r a, 0~ r u o

strategies in laboratory rats, brown lemmings, and gerbils

(see Table 4).

The pattern displayed generally correlates well with

mating system. Monogamous species, such as prairie voles

(Carter, Getz, & Cohen-Parsons, 1986) and Mongolian

gerbils (Agren, 1984c), have generally shown preferences

for familiar individuals, whereas polygamous species have

preferred novel partners or shown no preference. Carr,

Krames, and Costanzo (1970) have suggested that prior

sexual experience influences familiarity preferences.

Male laboratory rats given extensive polygamous experience

preferred the odor of novel laboratory rats, whereas

"monogamous" males showed no preference.

Preferences for Unmated versus Mated Individuals

Mating with recently mated individuals represents a

potential risk for both males and females (see below).

Accordingly, both sexes should have a similar mating

strategy for this criterion; namely, an avoidance of

recently mated individuals and a corresponding preference

for unmated individuals.

Preferences for unmated individuals have been tested

in several species of muroid rodents. Male and female

laboratory rats respond preferentially to the odors from

fresh individuals over those that had mated for one

ejaculation (Krames & Mastromatteo, 1973). These authors

concluded that in laboratory rats it "would appear that

during copulation, 2 crucial odor variables to which the

male responds are freshness and novelty" (p. 533). In

contrast, Zucker and Wade (1968) noted that male

laboratory rats indiscriminately mated when fresh and

mated females were presented. Sexually satiated male

golden hamsters likewise showed no preference when given a

choice between fresh and mated anesthetized females in a

lordosis position (Johnston & Rasmussen, 1984). Female

montane voles showed no preferences for unmated males in

tests permitting copulation and in tests where copulation

was prevented (Sawrey, 1983). However, female golden

hamsters discriminated and preferred rested anesthetized

males over recently mated anesthetized ones (Huck, Lisk,

Parente, & Principato, 1986). Male brown lemmings and

male collared lemmings preferred the odors of unmated over

mated females (Huck, Banks, & Coopersmith, 1984). This

preference was shown by sexually experienced males and by

sexually satiated males, but not by sexually naive males.

Male preference for unmated over mated females has

been demonstrated in prairie voles. Ferguson, Fuentes,

Sawrey, and Dewsbury (1986) presented male prairie voles

with a choice between two females, one which had recently

been mated to a nonexperimental male for one ejaculation

and one that had not. Males preferred unmated females in

both tether tests and anesthetization tests. Procedures

were then repeated using montane voles. No preferences

were exhibited in the tether tests; only a weak preference

was shown in anesthetization trials.

Other Preferences

Blonde coat color is a simple recessive trait in

prairie voles. The blonde color morph is extremely rare

in wild prairie vole populations, attributable to the

recessive nature of this trait and increased

susceptibility to predation. Getz and Pizzuto (1987)

conducted a tether choice study with blonde and normal-

colored voles as stimulus animals. Results revealed a

preference by each color morph for individuals of the

opposite coat color. These authors speculate that such

preferences "contribute to suppression of expression of

the blonde allele in prairie vole populations" (p. 230).

Edwards and Barnard (1987) infected female house mice

with larvae of the parasitic Trichinella spiralis, then

staged encounters with uninfected males. Sexually

experienced males investigated infected females more and

were less likely to associate and mate with them compared

to males paired with uninfected females. Sexually

inexperienced males responded similarly to infected and

uninfected females. In a follow-up study, infected and

uninfected male house mice were grouped together (Edwards,

1988). Social behaviors by infected house mice were

reduced, whereas investigation of infected house mice by

uninfected house mice was increased. Although these

studies have not shown directly a preference for healthy

individuals, the results suggest such a relationship (see

also Hamilton & Zuk, 1982).

Fischer, Olsen, & Meunier (1985) presented outbred

female golden hamsters with odors from inbred and outbred

males. Significant preferences for outbred males were

noted for estrous females; diestrous females exhibited no

preferences. If these results are generalizable, then

female mate selection may provide a further advantage to


Rodents of many species produce ultrasonic

vocalizations (USVs) prior to, during, and after

copulation (see Pierce, Sawrey, & Dewsbury, 1989 for a

review). Golden hamsters of both sexes approached natural

and synthetic ultrasounds produced in one arm of a Y-maze

(Floody & Pfaff, 1977). Association preferences for

vocalizing males over devocalized males have been reported

in estrous and diestrous house mice (Pomerantz, Nunez, &

Bean, 1983). In a similar design, female laboratory rats

did not differ in their association times, but did direct

more solicitation behavior (darting) to the intact males

(Thomas, Howard, & Barfield, 1982). Production of USVs

may be a possible preference-based criterion.

In two-choice tests, females have expressed

preferences for one male in the absence of any

experimental manipulation of the stimulus animals (French,

Fitzpatrick, & Law, 1972; Taylor & Weiss, 1987; Webster et

al., 1982) and the bases for these preferences remain

unclear. Preferred males achieved higher levels of

copulatory activity than non-preferred males (Taylor &

Weiss, 1987; Webster et al., 1982), although no evidence

yet exists to suggest there is a reproductive advantage in

terms of pregnancy initiation and successful production of

a litter. Still, the conclusion is clear. The

identification of the criteria by which choices are made,

and the relationship of these criteria to other variables

such as mating system and reproductive success, is still


Summary of Association/Mating Preferences

Table 5 provides a summary listing of preferences

exhibited in a variety of species. Muroid rodents choose

mates based on a surprisingly large number of criteria.

It appears that variations in many diverse characteristics

provide the bases for mate selection by these species.

The criteria for mating preferences can be classified

into two types. First, "identification" criteria identify

individuals as potential and appropriate mating partners

but provide no assessment of individuals that fit these

criteria. Examples of identification criteria include

preferences for conspecifics and members of the opposite

sex. Discrimination of these criteria are necessary for


54 0

Da U

O) VO 4. 4.
a) a) *
to o o*

U) > 0 *
o0 t:o Z P 4 -4

0 -4

3 O C3

-,0 2 3 .J

0 r0 OO
a) a) r4l

u *
a )
a 0) 0

0) U)
C 41 1
0 a u 0
o u 0 p

0 a
to x o< a4

C E 0 0 0 0 0)

0 0 0 0 0 0 0 0 0 0 0

(4.4 w)
0 to En


-4g CD )
4. 0) 0) CD
>D >0 0

LO > t 0) 0) 00
4) t r4 (0 4 0
4U 4 *H V a
) 0 c to z
0 p O4) (0
E-4U ) z 1 0 z)

o o

P UlU) U3)
0 > > .,1
3 0

S 4.) 4 4.) 4

z :D
0 0 0 0

0 o
0 00 00 00

p) r. 40 40 4) 41

U 0



S0 0 0 0
o Vi 2 4 a [ a a

0 000 0 0 0
S U U U u

~4 44
a C

C3 0

0 *H 0 C
oD 0
U) to

00 0 R 0
U0 0 rV

0u 0 :

4 0 0 p 0 0
Q)U UU ) U

r-4 r-4 r-1
4) a) *.-4
> > e
0 0 it
z z Z4

.1 4-),
I1 (n C
mw H

0 0
u u

a rZ z F4


M .M

0 4

* M
O 3j
O E-4

4 E4 z




0 4)




U0 0 0
C 0-4 0
cp a o c 00
o r- U r- *H U
O B I CM I 3

0 00 I0
04 0 0 0 0

U \ r
04 0 *r 0 0
0 m4 OO

r6 0 0


0 0q
* \

0 N 0 N
0, O) U
o 0 $4 0 $U 0

0U N H N H

0 0 0 0 0 0



0 0)<

$ 4 0 >1
) 0 U )
&i 0ao>

a a N
0 0 0
0 0 Z
s &U

0 0 0
O a &

0 0 0 0

O O Z 0


a) a)

o0 0
a C) 0

Sa 0
U) a)-e

0) X '-



N 0)
V 4-)

4 R






4 1
o I



0 0 0
O 0

0 N r4 0 .
oo A 4
4) w rq r:i
-,o -4
c 0 to
44-4 ) 0
4) W 04

C II r I
)*4 *4 C

0 k 0 4
0 0 1 II0

0 UU O

En0 0 0
o *P P4 0 4.1 Q
4 r) a e

S0 II *H 0)

00 i ao

d4 a W I
0 m 0 0 0 U
0 C 0P 1 04 C
a O 0000 4.

4* 0
X 0C O
0 M 4 0
a) .*. (0
0 4) 0 0

to C ai o

0 O 11 4

0 v-I 0 O

successful mating, and as would be expected, most (but not

all) species have shown clear discrimination and

preferences based on identification criteria.

The second type of criteria, "evaluative" criteria,

provide criteria by which individuals can be assessed for

quality. Here, individuals may be appropriate mating

partners but variations in important characteristics make

some individuals better potential mates than others.

Characteristics such as dominance rank, familiarity to the

choosing animal, and unmated status are examples of

evaluative criteria. Whereas most species display clear

preferences based on identification criteria, there is

greater interspecies variation in the display of

preferences based on evaluative criteria.

The classification scheme proposed here is

arbitrary, but may have utility in aiding understanding of

interspecific variation in mating preferences. Variation

in social and mating systems should affect preferences for

evaluative, but not identification, criteria. Preferences

for evaluative criteria also should provide an indication

of the "choosiness" of each sex in various species.

There is no reason to expect that all preferences are

mutually exclusive and dependent on uniquely associated

cues. For instance, androgen by-products in the urine

provide information concerning an animal's sex,

reproductive condition, and dominance status. A simple

strategy of attraction to stronger levels of androgen

odors would account for female preferences for several

criteria. Similarly, prior association mediates not only

a preference for novel versus familiar, but also

apparently for kin based preferences.

It is apparent from this review that preferences are

not "hard-wired" components of mate selection. Many

preferences are directly influenced by early experience

(see also Bateson, 1978; D'Udine & Alleva, 1983) and are

labile even in adulthood. In many instances, expression

of mating preferences is a function of the individual's

reproductive condition. Differences as a function of

season, prior sexual experience, and female's estrous

condition have been noted for one or more criteria.

Future Directions

The work of many researchers has yielded valuable

insights into mate choice by muroid rodents. It is

apparent that more work is needed to elucidate further the

importance of these criteria for all rodent species.

Additionally, other aspects of mate choice need to be

studied. A few aspects can be mentioned here.

First, the relationship of association preferences to

mating preferences needs to be assessed more carefully.

Studies which have used both of these measures to assess

mate choice have generally found an agreement between time

spent in association and amount of copulatory behavior.

However, association preferences and mating preferences

may differ for some species in some instances (e.g.,

Parmigiani et al., 1982). Interpretations of association

preferences as indicative of mating preferences should be

made with caution, until this relationship is assessed

more fully.

Little work on physical characteristics such as size

and age has been conducted. Age in particular is a

criterion for choice in such diverse groups as primates

(Robinson, 1982), feral pigeons (Burley & Moran, 1979),

crickets (Zuk, 1988), and butterflies (Wiklund & Forsberg,

1985). Age may be a criterion for mate choice in rodents.

There is existing research for laboratory rats which

indicates deleterious effects of age on pregnancy

initiation and cycling in females (Davis, Gray, &

Dewsbury, 1977; Hendricks, Lehman, & Oswalt, 1979) and on

the potency of the male's ejaculates (Toner & Adler,

1985). Swanson, Desjardins, and Turek (1982) failed to

find significant decrements in copulatory behavior in aged

male golden hamsters, but Huck, Lisk, and Guyton (1988)

have reported that aged females of this species required

greater amounts of vaginal stimulation for the initiation

of pregnancy. The effects of age on mating preferences

merit further attention.

Future research with pregnant or lactating females

also holds great promise, as it can be argued that females

in the field are most frequently in this condition.

Drickamer (1989) has examined odor preferences in pregnant

female house mice. During the early days of gestation,

females preferred the odors of the stud male and avoided

the odors of a strange male. As gestation continued, odor

preferences were reduced. Females avoided both male odors

toward the end of gestation. Avoidance of males and their

odors has also been noted for prairie vole females during

gestation (Prater & Rigley, 1982), although methodological

issues make this interpretation subject to caution.

Pregnant female golden hamsters approached conspecific

odors from intact and castrated males, but did not prefer

either odor (Johnston, 1979).


In this section, the mating systems characteristic of

prairie and montane voles are described. Mate choice is

then discussed in relation to their mating system.

Finally, a rationale is provided for why monogamous

females should prefer unmated males as mating partners.

Mating Systems of Prairie and Montane Voles

Mating Systems

Mating systems have been defined as the "ensemble of

behaviors and physical adaptations to mating, as well as

some of the social consequences of these behaviors"

(Vehrencamp & Bradbury, 1984, p. 251). This behavioral

ensemble reflects ecological and physiological constraints

on the species, and produces a mating system

characteristic of most members of a given population

(Dewsbury, 1988b).

One type of mating system, monogamy, is

characteristic of less than three percent of mammalian

species (Kleiman, 1977). Dewsbury (1988b) has proposed

that monogamy can be defined in terms of three dimensions.

The three dimensions are exclusivity of mating, joint

parental care, and preferential or exclusive association

between a particular male and female. Animals are

considered monogamous if they meet two of the three


Although other definitions vary, they are generally

based on one or more of these dimensions. Thus,

Wittenberger and Tilson (1980) propose that monogamy is "a

prolonged association and essentially exclusive mating

relationship between one male and one female" (p. 198).

Wickler and Seibt (1983) distinguish between monogamy as a

social system and monogamy as a mating system. They argue

that the latter is defined by "exclusivity of copulation"

(p. 37). Kleiman (1977) does not provide a specific

definition for monogamy, but proposes that a useful

distinction can be made between obligate and facultative

monogamy. Obligate monogamy is defined by family unit

cohesiveness, biparental care, and exclusivity of mating.

Facultative monogamy occurs at low densities and is

characterized by reduced pair association, minimal

paternal care, and more frequent extra-pair copulations.

Mating systems can vary across species within a

genus. For example, voles of the genus Microtus display

substantial interspecies variation in mating systems.

These species are widespread in North America and found in

most grassland and tundra communities (Rose & Birney,

1985). Typically, microtine rodents live sympatrically

with other mammalian species, but are usually the most

numerous in any given region in which they occur.

Microtine species display a range of different social and

mating systems, due in part to the diversity of the

ecological habitats which different species occupy. The

Microtus species of the present study, prairie voles and

montane voles, represent two diverse points in that range.

Mating System of Prairie Voles

Prairie voles, Microtus ochroqaster, evolved in the

stable habitats of prairie grasslands of North America,

habitats suitable for the formation of monogamous mating

systems (Getz, Hoffman, & Jike, 1986). Diet consists

mainly of succulent forbs (broadleaf herbaceous plants)

and grasses (Getz, 1985). Forbs are critical for survival

and reproduction (Cole & Batzli, 1979), and their low

availability and dispersed distribution provide the

ecological conditions for territoriality, and, in the case

of prairie voles, monogamy (Ostfeld, 1985).

Extensive field and laboratory research has

established that prairie voles exhibit a monogamous mating

system under most ecological conditions (see Carter &

Getz, 1985; Carter et al., 1986; Getz & Carter, 1980).

Trapping data have provided strong evidence for continued

male-female associations. Captures of male-female pairs

were frequent and equally likely during the breeding and

non-breeding seasons (Getz, Carter, & Gavish, 1981).

Male-female pairs were frequently recaptured together,

some as late as 20 weeks after the initial capture.

Hofmann, Getz, & Gavish (1984) tracked heterosexual pairs

via radiotelemetry. Results indicated that 11 or 12 pairs

remained together and shared a common nesting burrow

during the 10-day tracking period. Trapping data derived

from nests of two field populations indicated that 50% of

281 breeding units were monogamous (defined as a single

resident male and female pair); only 23% contained more

that one male or one female (Getz & Hofmann, 1986).

Population density strongly influenced mating system in

this study, with monogamous units significantly more

frequent at low population densities that at high

population densities, although monogamous units were the

most frequent at all densities. Visits by nonfamily

members were infrequent.

Laboratory studies corroborate the field data.

Family pairs in a semi-natural environment displayed

cohabitation of a common nest and exhibited biparental

care of the young (Gruder-Adams & Getz, 1985). Biparental

care by prairie voles has also been documented by other

researchers (Hartung & Dewsbury, 1979; McGuire & Novak,

1984; Oliveras & Novak, 1986). Thomas and Birney (1979)

established 27 vole societies of varying sex ratios in

laboratory pens. All but one of these societies mated

monogamously and displayed extensive biparental care to

offspring. In related work, Gavish, Carter, and Getz

(1981) showed that reproductive success, as measured by

number of pups born, was significantly higher in

monogamous pairs than in experimentally-created breeding

units consisting of three animals.

Preferential associations between pair members have

been assessed in a series of experimental studies. In

staged dyadic encounters, aggressive behaviors were

significantly greater between a breeder animal and an

unfamiliar, opposite-sexed animal than between members of

established pairs (Getz et al., 1981). Shapiro (1987)

examined rates of contact in dyadic encounters and

concluded that prairie voles exhibit "contact proneness".

In multiple female situations, male prairie voles

preferentially mated with a single female (Fuentes &

Dewsbury, 1984). Mating preferences for familiar mates

(Newman & Halpin, 1988; Shapiro, Austin, Ward, & Dewsbury,

1986) and unmated individuals (Ferguson et al., 1986) have

also been reported for this species and are indicative of

a monogamous mating system.

Together, these data provide strong evidence that

prairie voles mate monogamously under low to moderate

population densities. Laboratory evidence has been well

established for the three dimensions of monogamy proposed

by Dewsbury 1988b); exclusivity of mating, joint parental

care, and association. Field evidence has been

established for the latter two of the three dimensions.

Dewsbury (1981) suggested nine correlates of monogamy.

Prairie voles scored highly on this Monogamy Scale.

Evidence for monogamy for prairie voles meets the

definition established by other researchers (i.e.,

Kleiman, 1977; Wittenberger & Tilson, 1980).

Mating System of Montane Voles

The behavior of montane voles, in contrast, is more

consistent with a polygamous mating system. Montane voles

are native to mountain valley regions of the northwestern

United States (Jannett, 1980). These regions contain

sedges and grasses, but the feeding habits of this species

have not been fully established (Ostfeld, 1985). It is

known that a critical nutritional requirement for

reproductive activation is 6-MBOA, a compound found in

green vegetation (Sanders, Gardner, Berger, & Negus, 1981;

Berger, Negus, Sanders, & Gardner, 1981).

Field work has established that male and female

montane voles are territorial (Jannett, 1978, 1980), but

the size and shape of territories differ between males and

females. Male territories are larger and overlap the

ranges of several females (Jannett, 1980). Although a

male and female may be in close association during female

estrus, there is no indication of pairs nesting together

(Jannett, 1982). In addition, in instances of female

dispersal, males do not follow but rather remain behind.

There is little effect of density upon the mating system.

Polygamy occurs at all population levels except very low

densities where facultative monogamy generally occurs

(Jannett, 1980).

Laboratory evidence confirms the polygamous

tendencies of montane voles. As noted above, montane

voles show little preference for familiar (Shapiro et al.,

1986) or unmated individuals (Ferguson et al., 1986).

Male-female dyadic encounters in a seminatural enclosure

revealed that montane voles show low levels of contact

proneness and high rates of aggression (Shapiro, 1987).

Males in multiple female test situations copulated with

more females and distributed ejaculations more equally

than did prairie voles (Fuentes & Dewsbury, 1984).

Montane voles scored toward the polygamy end of the

Monogamy Scale of Dewsbury (1981).

Summary of Mating Systems of Prairie and Montane Voles

Different forms of social organization and mating

system result from environmental regulation and stable

differences in behavior. These factors have led to

prairie voles exhibiting a territorial, monogamous system

at low to moderate population densities, and montane voles

exhibiting a territorial, polygamous system at most

population densities.

Mating Systems and Mate Choice

The mating system characteristic of a given

population has implications for the mating preferences

exhibited by individuals in that population. It has been

argued that mate choice should be more pronounced in

monogamous species than polygamous ones (Kleiman, 1977;

O'Donald, 1983). Preferences related to a potential

mate's ability to provide resources and parental care

should be critical for monogamous animals, as should

characteristics that identify a potential mate as

"available"; i.e., not already a member of a long-term,

stable association. In polygamous species where the male

does not provide resources or parental care, preferences

based on these characteristics would be expected to be

less critical.

One example of a characteristic from the latter

category is the recency of mating by the potential mate.

Wittenberger and Tilson (1980) have argued that the

ability to ascertain the "true mated status of potential

mates" (p. 199) represents one of three critical

preconditions for the evolution of monogamous mating

systems. Although their argument was for female choice,

the ability to detect mated status should be critical for

both sexes. For males displaying high male parental

effort (MPE), pairing with a recently mated female leads

to the dangers of sperm competition and cuckoldry (see

Parker, 1970). For females of a species with high MPE,

copulating with a recently mated male is risky for several

reasons, including (1) failure to initiate pregnancy, (2)

failure to maintain pregnancy, and (3) failure to raise a


Failure to initiate pregnancy

The basic organizational unit of copulatory behavior

in many mammalian species is the ejaculatory series, which

begins with an intromission and terminates with an

ejaculation. During ejaculatory series, males of many

species provide not only sperm, but also the stimulation

necessary to initiate a series of neuroendocrine reflexes

critical for sperm transfer, ovulation, and the initiation

of pregnancy (Dewsbury, 1978; Diamond, 1970, 1972).

Providing adequate stimulation to initiate pregnancy

may require copulation beyond the completion of one

ejaculatory series; stimulation that a recently mated male

may not be able to provide. Copulatory stimulation has

been shown to facilitate sperm transfer in golden hamsters

(Diamond, 1972; Lanier, Estep, & Dewsbury, 1975, but see

Huck & Lisk, 1985b), and laboratory rats (Matthews &

Adler, 1977, 1978). In montane voles, more than two

complete ejaculatory series were required to induce

ovulation in all females tested (Davis, Gray, Zerylnick, &

Dewsbury, 1974). Ovulation rates dropped to 25% in

females that received only one ejaculatory series. One

ejaculatory series was sufficient to produce ovulation and

implantation in female prairie voles (Gray, Zerylnick,

Davis, & Dewsbury, 1974).

In addition to possibly failing to provide

stimulation, a recently mated male is probably a depleted

male. A characteristic pattern of sexual activity in many

mammalian species is the completion of several ejaculatory

series followed by a general decrease in copulatory

activity to the point of cessation of copulation for long

periods of time. This cessation point has been generally

referred to as sexual satiation or sexual exhaustion

(e.g., Beach, 1956). Thus, whereas sperm production is

"cheap," the capacity to produce ejaculates, and

consequently the capacity to impregnate females, may be

limited (Dewsbury, 1982a; see Nakatsuru & Kramer, 1982).

Not only are there limitations in the number of

ejaculates produced, there are also significant decreases

in sperm counts of later ejaculates. Decrements in

potency and fertilizing capacity of successive ejaculates

have been shown in several rodent species. Mating with

depleted golden hamsters resulted in reduced fertility

(Huck et al., 1986) and smaller litter sizes, presumably

because of smaller ejaculates (Huck & Lisk, 1985a).

Successive ejaculates in laboratory rats were

characterized by decreasing sperm counts and reduced

fertility (Austin & Dewsbury, 1986; Toner & Adler, 1985).

Limited ejaculate production and decreases in sperm

counts across series have also been noted in prairie and

montane voles. Characteristic of both species is the

production of a large, initial ejaculate with subsequent

ejaculates being a diminishing fraction of the initial

ejaculate (Pierce et al., manuscript submitted for

publication). In early laboratory tests, montane voles

achieved a mean of 5.0 (Dewsbury, 1973) and prairie voles

a mean of 2.0 (Gray & Dewsbury, 1973) ejaculations before

reaching a satiety criterion of 30 min without an

intromission. In more recent work with different

colonies, a satiety criterion of 30 min without an

intromission or 60 min without an ejaculation has been

used. This criterion was adopted to correct for the

occurrence of "incomplete series", intromissions at

irregular intervals for hours after the last ejaculation,

in species such as montane voles (Dewsbury, 1973) and

cactus mice (Dewsbury, 1974). Using this more recent

criterion, a montane vole mean of 3.4 ejaculations (Pierce

et al., submitted) and prairie voles means of 2.7 (Pierce

et al., submitted) and 2.9 (Pierce et al., 1988)

ejaculations have been reported. Thus, these species are

limited in ejaculate production with montane voles

generally showing more copulatory behavior than prairie


Limited ejaculate production, the comparatively small

ejaculates following the initial ejaculate, and the

stimulation requirements to initiate pregnancy suggest

that females of these species may not receive the

necessary copulation to initiate pregnancy by mating with

depleted males. This may be more the case for prairie

voles, however. Satiated male montane voles display a

Coolidge effect, the reliable resumption of copulatory

activity following introduction of a novel female

(Dewsbury, 1973). This resumption may be sufficient to

initiate pregnancy, as has been demonstrated in golden

hamsters (Huck & Lisk, 1985a, 1985b). Prairie voles fail

to show a Coolidge effect (Gray & Dewsbury, 1973). Thus,

it can be argued that mating with a depleted male in

prairie voles may be more likely to lead to a failure to

ovulate and conceive than in montane voles.

Failure to maintain pregnancy

Even if pregnancy is successfully initiated, mating

with a recently mated male may have two consequences for

maintaining pregnancy because of the risk of desertion.

First, in several species, the presence of the male during

the early stages of pregnancy is critical for the

maintenance of that pregnancy by the female. This effect

has been reported for prairie voles by Richmond and Stehn

(1976) who discovered that 1-4 days of cohabitation were

necessary for pregnancy maintenance. Highest rates of

pregnancy occur when the male is present throughout

pregnancy (Hofmann, Getz, & Gavish, 1987). A similar

facilitation by male presence has been reported for

montane voles (Berger & Negus, 1982).

Second, a deserted female is at risk of having the

pregnancy disrupted. The Bruce effect (Bruce, 1959) is a

pregnancy block that occurs when a female is exposed to a

strange male or the odor of a strange male. This effect

is common in many rodent species (see reviews by Labov,

1981; Schwagmeyer, 1979), especially microtine rodents

(Stehn & Jannett, 1981). Montane (Stehn & Jannett, 1981)

and particularly prairie voles (Stehn & Jannett, 1981;

Stehn & Richmond, 1975) are susceptible to the Bruce

effect. In prairie voles, post-implantational pregnancy

blocks can occur as late as 16 or 17 days past copulation

(Stehn & Richmond, 1975) and in semi-natural enclosures

where the female can potentially avoid the male (Heske &

Nelson, 1984). Evidence suggests that the presence of the

original male facilitates the maintenance of pregnancy

(Richmond & Stehn, 1976) and may block the occurrence of

the Bruce effect. Thus, the male's presence during

gestation can be critical for pregnancy maintenance.

Failure to raise a litter

Even if a female is able to deliver a litter

successfully, the absence of a partner that would

otherwise provide parental care may result in increased

pup mortality or even the catastrophic loss of an entire

litter. Kleiman and Malcolm (1981) have identified

several categories of direct and indirect paternal

behavior; including resource acquisition, nest

construction and maintenance, care of mother and young,

defense, pup retrieval, etc. Without male parental care,

females of some species, such as prairie voles, may be

unable to raise a litter successfully. Prairie vole males

provide extensive parental care in laboratory (Dewsbury,

1985; Hartung & Dewsbury, 1979) and field (Getz, 1962) and

existing evidence suggests that this paternal care results

in higher pup survivorship (Thomas & Birney, 1979).

Male presence during development may also help

prevent infanticide. Many laboratory studies have

reported infanticidal behavior in several rodent species,

although field data of this phenomenon are scarce (see

reviews by Brooks, 1984; Labov, Huck, Elwood, & Brooks,

1985). Male absence during litter development may leave a

litter more susceptible to infanticide.


In monogamous species, males may provide resources

and parental care critical for the survival and

development of the offspring. Thus, females of these

species should discriminate and prefer males that are not

already a member of a long-term stable association. One

proximate cue that may indicate the "availability" of a

male is the recency of his prior mating. The ability of

females to discriminate between unmated and recently mated

males was assessed in the present research.


As noted above, male preferences for unmated over

mated females have been demonstrated in prairie voles,

whereas male montane voles show little preference

(Ferguson et al., 1986). These results suggest that

animals of monogamous species are more discriminating than

are polygamous species with respect to the proposed

criterion of recency of mating. The studies reported here

were designed 1) to consider a similar effect in female

prairie and montane voles, 2) to assess whether the amount

of recent mating by the potential partner influences the

expression of a preference, and 3) to determine if there

is a functional consequence in terms of pregnancy

initiation that underlies a preference for an unmated


The general structure of the 10 experiments was as

follows. Experiments 1-8 represented a 2 x 2 x 2 design

with two species (prairie voles and montane voles), two

levels of recent mating (one versus three ejaculations),

and two types of apparatus (tether tests and

anesthetization tests) used to study female preferences

for unmated versus mated males (see Figure 1). In









3 1

EXP. 1 EXP. 2

EXP. 3 EXP. 4



3 1

EXP. 5 EXP. 6

EXP. 7 EXP. 8

Figure 1. Experimental design of Experiments 1-8.

Experiments 9 and 10, pregnancy rates in virgin females

which mated with a recently mated, depleted male or a

sexually experienced, non-depleted male were examined.

The levels of recent mating selected for the present

experiment were based on the number of ejaculations

typically attained before satiety and the requirements for

pregnancy in the two species (see above). At one prior

ejaculation, males can not be considered as depleted. At

three ejaculations, however, most male prairie voles have

reached sexual satiety and future copulation would likely

be at infrequent and irregular intervals. Montane voles

after three ejaculations will still display sexual

behavior, but may not have the more than two series

required to initiate pregnancy (Davis et al., 1974; see


It was predicted that prairie vole females would

prefer unmated males in both tether and anesthetization

trials when the amount of recent mating was sufficient to

provide the cues necessary for the discrimination, whereas

female montane voles would show no preference. With

respect to a functional consequence of this preference, it

was predicted that prairie vole females would show reduced

probability of pregnancy when mated with a depleted male

as compared to mating with a fresh male. It was expected

that montane voles would show no difference.


General Method


Subjects were 78 female and 74 male prairie voles and

80 female and 70 male montane voles derived from the

laboratory colonies at the University of Florida. Females

were 90-180 days old at the time of testing; males were no

older than 240 days by the completion of testing. Prior

to testing, subjects were individually housed in 23 x 19 x

13 cm polycarbonate cages in colony rooms on a reversed

16L:8D photoperiod with lights on starting at 20:00 hours.

Water and Purina rabbit chow were available ad lib.

All male voles were pretested for copulatory behavior

prior to testing in order to provide them with sexual

experience and ensure that they copulated under test

conditions. Pretesting consisted of mating each male to a

nonexperimental female for two complete ejaculatory

series, then leaving the pair together overnight in the

same cage for additional sexual experience. All protests

were conducted at least 10 days prior to using the male as

a subject. Some males were used as stimulus animals more

than once in an experiment. When being used a second

time, the stimulus male was paired with a different

stimulus male and in the opposite treatment condition

(i.e., fresh versus depleted) than the first time he was

used as a stimulus male.

All female voles used as subjects were without known

sexual experience prior to testing and were behaviorally

receptive at the time of testing. Each female was a

subject in only one experiment and had no prior exposure

to either of the stimulus males prior to testing.


All tests for Experiments 1-8 were conducted in a

tether box made of plywood painted flat gray, with a

Plexiglas front and a layer of San-i-cel bedding covering

the floor (see Figure 2). The box measured 68 x 20 x 33

cm and was divided into three equal-sized compartments;

end compartments for the two males and a central neutral

area. For Experiments 1, 2, 5, and 6, males were tethered

at opposite ends of the box and a female were free to move

between them. In Experiments 3, 4, 7, and 8, untethered,

anesthetized males were placed ventrum up at opposite ends

of the box and a female was free to move between them.

Males were placed ventrum up to allow females maximal















exposure to the olfactory cues assumed to underlie the


Experiments 9 and 10 were conducted in 48 x 27 x 13

cm polycarbonate cages.


Females were brought into estrus by placing them

across a wire barrier from a male in a divided cage three

days prior to testing. Receptivity was determined on the

day of testing. For prairie voles, daily vaginal smears

were taken using tap water and a thin wire loop. Only

those females whose smears indicated a first-day estrus

(defined as a virtual lack of leukocytes and a

predominance of non-nucleated squamous cells) on the test

day and which showed behavioral receptivity to a stud male

were used. For montane voles, receptivity was assessed by

the female's behavioral responses to a stud male on the

test day, as vaginal smears are not a good predictor of

behavioral estrus in this species (Sawrey, 1989).

For Experiments 1-8, one male was mated for the

appropriate number of complete ejaculatory series with a

nonexperimental female. These females were brought into

behavioral receptivity by injection of 0.06 mg of

estradiol benzoate three days before testing. Following

the last ejaculation, both males were either tethered

(Experiments 1, 2, 5, and 6) or anesthetized (Experiments

3, 4, 7, and 8) at opposite ends of the box. The side on

which the mated male was placed was counterbalanced across

trials. The female was then introduced into the back of

the central area and data recordings were begun.

Anesthetization tests lasted 30 min following the

introduction of the female. Tether tests lasted for 60

min following the first intromission by either male.

Tests were aborted if the female failed to visit each male

at least once within the first five minutes.

Anesthetization was accomplished by subcutaneous

injections of 0.04 ml of a ketamine HCL- xylazine mixture.

Males were placed in the tether box when all activity had

ceased and the animal was breathing deeply and slowly.

The female was introduced immediately thereafter into the

back of the central area.

All tests were conducted under red light conditions

during the dark phase of the photoperiod.


Tests involving the tethering procedure permitted the

collection of visitation and copulatory data. Visitation

measures, which provide a measure of association

preferences, included the number and duration of visits to

each male. A visit to a male was scored when any part of

the female entered the section of the tether box

containing a male.

The copulatory pattern of voles consists of three

classes of behaviors: mounts (with no vaginal insertion),

intromissions (mounts with insertion and thrusting but no

ejaculation), and ejaculations (mounts with intromission,

thrusting, and seminal emission). Copulatory measures

recorded included the number of mounts, the number of

intromissions, the number of thrusts, and the number of

ejaculations. Hit rate, the ratio of intromissions to the

number of mounts plus intromissions, was calculated for

each male (Dewsbury, 1973; Gray & Dewsbury, 1973).

Tests involving anesthetized males permitted the

collection of visitation data only. Measures related to

visitation included the number and duration of 1) visits

to each male, 2) bouts of sniffing/licking each male, and

3) bouts of lying beside/across each male (Ferguson et

al., 1986). Sniffing/licking was scored whenever the

female's snout was within 1 cm of any part of an

anesthetized male.

Measures for tests involving pregnancy in females

(Experiments 9 and 10) were the number of pregnancies in

each condition and the number of pups born.

Experiment 1

The first experiment was an attempt to determine

whether a behaviorally receptive prairie vole female would

discriminate between and copulate preferentially with an

unmated, tethered male versus a tethered male which had

recently mated for three ejaculatory series.


Within three days prior to testing, prairie vole

males were fitted with plastic collars and given 1-hr

adaptation to the tether procedure and apparatus. Females

were given 1-hr adaptation to the tether box in the

absence of males. Just prior to testing, one male was

mated to a nonexperimental, hormone-injected female for

three complete ejaculatory series in a separate 48 x 27 x

13 cm polycarbonate cage. The mated and unmated males

were then tethered at each end of the tether box for 10

min. Following this adaptation period, the female was

placed in the back of the neutral area to start the test.

Recording of copulatory and visitation measures started

immediately and continued until 1 hr after the first

intromission by either male. A test was considered

aborted if the female failed to begin copulation within 30

min following introduction. A total of 12 trials was run.


Results of paired t-tests are presented in Table 6.

There was a substantial difference in time spent with each

male, as females preferentially associated with unmated

Table 6
Preferences of Female Prairie Voles for Unmated Versus Mated
(for Three Ejaculations) Tethered Male Voles in Experiment 1

Measure Unmated Mated t E

(M) Male Male (df=ll)

Total Visit Time (sec) 2244.4 616.8 3.26 .01

No. Visits 29.3 24.1 1.44 n.s.

No. Ejaculations 2.3 1.0 2.11 .056

No. Intromissions 12.0 4.2 2.03 .06

No. Thrusts 71.3 24.2 2.24 .05

No. Mounts 13.7 7.1 1.57 n.s.

Hit Rate 0.47 0.34 0.98 n.s.

No. Trials with >50% of 11 1

Total Visit Time

males. Mean number of visits to each male did not differ

significantly, indicating that the results were not due to

differential sampling of each male.

Differences in the amount of copulatory behavior

engaged in by each male fell just short of statistical

significance. Females received significantly more thrusts

from unmated males than from mated males.

Experiment 2

In the first experiment, a significant female

preference for associating with unmated males over males

recently mated for three ejaculations was revealed. The

second experiment was an attempt to assess whether females

would continue to show a preference for unmated males when

the amount of copulatory activity by the mated male was

reduced. In this experiment, it was determined whether a

behaviorally receptive female would discriminate between

and copulate preferentially with an unmated, tethered male

versus a tethered male which had recently mated for one

ejaculatory series.


In Experiment 2, mated males copulated for one, as

opposed to three, ejaculatory series with a

nonexperimental female before being tethered in the tether

apparatus. The procedures for Experiment 2 were otherwise

similar to those of Experiment 1. A total of 10 trials

was run.


Results of paired t-tests are presented in Table 7

and show that females did not establish significant

preferences for unmated versus mated males.

None of the visitation or copulatory measures was

significantly different, although number of visits

approached significance (t = 2.07, R <.07) and there was a

substantial difference of 535.5 sec in total visit time

between males. These results demonstrate that the amount

of copulatory behavior by a recently mated male has an

important influence on female preferences.

Experiment 3

Females in Experiment 1 may have shown a preference

for an unmated male based on both behavioral and

nonbehavioral cues. Behaviorally, informal observations

suggested that recently mated males interacted less with

the female, at least initially. Conversely, unmated males

were frequently more persistent in their approaches and

interactions. This difference in behavior may have been

the cue for female preference. Experiments 3 and 4 were

Table 7
Preferences of Female Prairie Voles for Unmated Versus Mated
(for One Ejaculation) Tethered Male Voles in Experiment 2

Measure Unmated Mated t E

(M) Male Male (df=9)

Total Visit Time (sec) 1796.9 1261.4 0.89 n.s.

No. Visits 26.4 23.8 2.07 .07

No. Ejaculations 2.2 1.6 0.90 n.s.

No. Intromissions 19.1 12.9 0.62 n.s.

No. Thrusts 97.1 52.9 1.24 n.s.

No. Mounts 8.9 12.1 0.40 n.s.

Hit Rate 0.55 0.46 0.55 n.s.

No. Trials with >50% of 6 4

Total Visit Time

attempts to determine if prairie vole females would

discriminate and associate preferentially with an unmated

male over a recently mated male when behavioral cues were

eliminated. In Experiment 3, prairie vole females were

given a choice between an anesthetized unmated male and an

anesthetized male which had recently mated with a

nonexperimental female for three ejaculatory series.


Just prior to testing, a prairie vole male was mated

to a nonexperimental, hormone-injected female for three

ejaculations. Following the last ejaculation, the mated

male and an unmated male were injected subcutaneously with

0.04 ml of the ketamine HCL-xylazine mixture and placed,

ventrum up, in opposite sides of the tether box. The

female was introduced immediately thereafter in the back

of the neutral area to start the test. Visitation and

association measures were recorded for 30 min following

introduction of the female. A total of 12 trials was run.


Results of paired t-tests are presented in Table 8.

Prairie vole females spent significantly more time

visiting and lying beside the unmated male compared to the

mated male. The amount of sniffing to each male

Table 8
Preferences of Female Prairie Voles for Unmated Versus Mated
(For Three Ejaculations) Anesthetized Male Voles in
Experiment 3

Measure Unmated Mated t E

(M) Male Male (df=ll)

Total Visit Time (sec) 1093.6 363.3 3.90 .01

No. Visits 22.1 14.0 1.61 n.s.

Time Sniffing (sec) 243.2 171.3 1.97 .07

Time Lying Beside/ 99.2 24.1 2.18 .05

Across (sec)

No. Trials with >50% of 10 2

Total Visit Time

approached significance (R <.07). Mean number of visits

did not differ significantly, indicating similar sampling

rates to each male.

The pattern of visitation indicates the establishment

of a preference during testing. A mean time spent

visiting each male was determined for each third of the 30

min observation session. These data are presented in

Figure 3. As the figure shows, females established

preferences for unmated males during the first 10 min,

then progressively increased visitation to the unmated

male as the session continued. By the final third of the

session, females were spending virtually all of the time

with the unmated male.

Experiment 4

In Experiment 3, female prairie voles preferred to

associate with anesthetized unmated males over

anesthetized males recently mated for three ejaculations.

Experiment 4 was undertaken to determine if a prairie vole

female, in the absence of behavioral cues, would

discriminate and associate preferentially with an unmated

male over a male which had recently delivered one

ejaculation to a nonexperimental female. Females failed

to show this preference when males were tethered

(Experiment 2).

w .

4 -)


-Il X
F 04Ez














The procedures for Experiment 4 were similar to those

of Experiment 3, except that the mated male copulated for

one, as opposed to three, ejaculatory series with a non-

experimental female. A total of 12 trials was run.


Results of paired t-tests are presented in Table 9.

There was no significant difference in number of visits,

time visiting, or time spent lying beside or across either

male. Females sniffed unmated males significantly longer

than mated males. Females initially preferred unmated

males, but increased time spent visiting with the mated

male as the session continued (see Figure 4). Overall,

however, there was no significant difference in time spent

with either male.

These data are consistent with the data from

Experiment 2 in that females did not preferentially

associate with unmated males when mated males deliver one,

as opposed to three, ejaculations.

Prairie vole females, like prairie vole males,

display preferences for unmated versus recently mated

potential partners. Further, in this species, the amount

of prior copulatory behavior is one factor influencing

female preference for unmated males. The next four

Table 9
Preferences of Female Prairie Voles for Unmated Versus Mated
(For One Ejaculation) Anesthetized Male Voles in Experiment

Measure Unmated Mated t R

(M) Male Male (df=ll)

Total Visit Time (sec) 769.1 840.4 0.41 n.s.

No. Visits 10.3 10.7 0.20 n.s.

Time Sniffing (sec) 282.2 149.9 2.18 .05

Time Lying Beside/ 169.4 84.0 0.71 n.s.

Across (sec)

No. Trials with >50% of 5 7

Total Visit Time

















as ~


< II








< -

experiments were an attempt to assess whether montane vole

females would show a similar preference.

Experiment 5

Experiment 5 was an attempt to determine whether a

behaviorally receptive montane vole female would

discriminate between and copulate preferentially with an

unmated tethered male versus a tethered male which had

recently mated for three ejaculatory series.


The procedures for Experiment 5 were similar to those

of Experiment 1, except that montane voles were used as

subjects and stimulus animals. A total of 12 trials was



Results of paired t-tests are presented in Table 10.

Female montane voles visited each male similarly, as no

significant differences were noted for number or duration

of visits. Unmated males achieved significantly more

mounts than did recently mated males, but no difference

was noted for number of intromissions or thrusts. The

Table 10
Preferences of Female Montane Voles for Unmated Versus Mated
(for Three Ejaculations) Tethered Male Voles in Experiment 5

Measure Unmated Mated t R

(M) Male Male (df=ll)

Total Visit Time (sec) 1604.7 1510.7 0.17 n.s.

No. Visits 58.9 49.8 1.16 n.s.

No. Ejaculations 2.7 1.6 2.11 .057

No. Intromissions 27.3 22.4 0.75 n.s.

No. Thrusts 152.0 55.5 1.02 n.s.

No. Mounts 27.0 6.6 2.80 .05

Hit Rate 0.49 0.63 1.16 n.s.

No. Trials with >50% of 7 5

Total Visit Time

difference in ejaculation frequency for each male

approached significance.

Experiment 6

In Experiment 5, female montane voles failed to

associate preferentially with unmated males over males

which had recently mated for three ejaculations.

Experiment 6 was designed to determine whether the amount

of prior copulatory behavior by the mated male influences

preference formation by females. In this experiment, it

was determined whether a behaviorally receptive female

will discriminate between and copulate preferentially with

an unmated tethered male versus a tethered male which had

recently mated for one ejaculatory series.


The procedures for Experiment 6 were similar to those

of Experiment 2, except that montane voles were used as

subjects and stimulus males. A total of 12 trials was