THE “DISTRESS” CRY 0F INFANT DEERMICE, PEROMYSCUS:
PHYSICAL CHARACTERISTICS, SPECIFIC DIFFERENCES,
SOCIAL FUNCTION.

DISSERTATION FOR THE DEGREE OF PH. 1).. .
MECHIGAN STATE UNIVERSITY
JANE NETTING HUFF

1973

lJ ”BRADY if.

Mimics; , 3.3

g.
T ' 2 ..
anvcza‘gj

   
  

 

 

 

mull

 

 

 

 

 

 

 

llllj‘llllllllllllillll'llll '

3 3 01070 8158

.N'v ”'3' I'

. . This-is to.ceiflfy tfiat‘th'é ‘
‘5‘ l: l ‘ l
’ " thesis entitled
The ”Distress" Cry of Infant Deermice,
Peromyscus: Physical Characteristics,
Species Differences, Social Function

presented by

Jane Netting Huff'

has been accepted towards fulfillment
of the requirements for

Ph.D. Zoology

degree in

Qt (New;

‘ . “I, Major professor
..= " ‘5
Date léalflnfik 6' (393 ‘.

0—7639

. Mag“

 

 

 

 

 

 

ABSTRACT

THE "DISTRESS" CRY OF INFANT DEERMICE, PEROMYSCUS:

 

PHYSICAL CHARACTERISTICS, SPECIFIC DIFFERENCES,

SOCIAL FUNCTION.

by Jane Netting Huff

Isolated Peromyscus pups vocalize with rapid sets

 

of ultrasonic squeaks. These sounds were recorded from
pups of 8 different taxa in order to quantify species-
specific characteristics and parameters showing develop-
mental changes. The behavioral responses of adult female

Peromyscus to recorded squeaks were tested to clarify the

 

function of the calls.

 

Pups of Peromyscus maniculatus bairdi, 3.m.gracilis,

prolionotus, and £.leucopus, were recorded from day l

 

after birth until the average age of eye-opening. Pups of

E.californicus, §.melanophrys, 33m3nebracensis, and g.

 

 

 

difficulis were recorded when available.

 

Statistical analyses showed significant inter-
specific differences in the number of squeaks in a group
of squeaks (set), the duration of individual squeaks, and

in basic frequencies. The E.maniculatus subspecies showed

 

similar mean basic frequencies (bairdi - 22.6 kHz,

 

J. N. Huff

gracilis — 22.9 kHz, nebracensis — 23.7 kHz) as did

[*0

 

 

melanophrys (22.8 kHz). Mean basic frequencies for g.

 

leucopus and E.polionotus were higher (26.7 kHz and 31.4

 

kHz) while those for g.californicus and §.difficilis were

 

 

lower (18.6 kHz and 19.9 kHz). These frequencies tended
to rise with age.
Squeak duration differed significantly between the

taxa from B.polionotus (0.13 sec.) to E,m.gracilis and E.

 

californicus (0.18 sec.). Set duration did not differ sig—

 

nificantly across the groups. The number of squeaks per

set for E,m.bairdi and g.polionotus (3.70) was signifi-

 

cantly different from the mean for g.m.gracilis (4.05) and

§.leucopus (3.15). These temporal parameters all de-

 

creased significantly in mean value over age.

Comparison of oscilloscope films and sonagrams
demonstrated species~specific patterns of frequency and
intensity modulation. Each of the taxa studied differed
from the others in at least one quantifiable parameter and
in its modulation patterns. These calls also showed
characteristics which make sounds easy to locate, such as
high frequencies, frequency and intensity modulation, re-
dundancy, and sharp rise and fall.

The hypothesis that these calls might affect the
behavior of a mother mouse, (1) by increasing her activity
and stimulating her to search for a "pup—in—distress";

(2) by enabling her to efficiently locate a calling pup;

and (3) by providing sufficient long—distance information

J. N. Huff

for preferential responses to species, was tested with
groups of mice in two modified Open—field boxes. £.m.
bairdi mothers were tested in a two-way box, for dif-
ferential activation and preference for recorded squeaks
of §.m,bairdi pups versus mechanical clicks and white
noise. In terms of numbers of responses, correct choices,
and distances traveled towards the mouse squeaks, the

test and control groups preferred the squeaks to clicks or

noise. Similar groups of §.m.bairdi and g.leucopus

 

mothers preferred leuCOpus squeaks to white noise. But

 

those test groups given a choice between squeaks of the
two species showed minimal discrimination.

E,m.bairdi mothers showed accurate location of the
source of squeaks from recordings of live pups in a fan-
shaped test box. Responses to recordings were faster and
more accurate than to live if first choices alone were con—
sidered but a mean of three choices showed no significant
differences between responses to live pups and recordings.

These data are interpreted to mean that the squeak—
ing of a pup dropped by a mother mouse moving a nest or
fleeing a disturbance is both a necessary and sufficient
factcm in ensuring retrieval and survival. The sounds
aleITLthe mother mouse to the situation, allow her to lo-
Catma the calling pup efficiently, thus reducing risks of
Searching in the Open, and possibly even permit her to re-
Spond preferentially to her own Species. The ontogenetic

Changes in the characteristics of the squeaks correlate

J. N. Huff

with the decreasing amounts of maternal behaviors,possi—

bly because of reduced stimulus effectiveness. Differences

in the squeaks occur in temporal and physical parameters

across ages and between species in Peromyscus.

 

THE "DISTRESS" CRY OF INFANT DEERMICE, PEROMYSCUS:

 

PHYSICAL CHARACTERISTICS, SPECIFIC DIFFERENCES,

SOCIAL FUNCTION.

BY
Jane Netting Huff

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Zoology
1973

 

(A ACKNOWLEDGMENTS

Appreciation is expressed to Drs. Martin Balaban,
‘William Cooper, John A. King, and Oscar Tosi for their
Ihelp and advice. Thanks are also due to Dr. Hodge Markgraf
xvho obtained funds for me to finish the thesis; to Mrs.
ZEileen Sprague and the other secretaries who patiently and
carefully typed it; to Mr. Don Riggs who helped me use the
frequency analyzers; and to my sisters and friends who en-
couraged and supported me. Greatest appreciation and thanks
rmast go to the scores of mice who did "their thing," and
IMDSt of all to my family, especially Curt, Lucinda and

Anthony, who believed in me.

ii

". . . so at last I know

that its a perfectly possible thing
for a light-footed, flight-footed,
left-footed, right-footed,

White-Footed Mouse to sing."

(Fisher, 1969)

iii

TABLE OF CONTENTS

LIST OF TABLES. . . . . . . . . . .
LIST OF FIGURES C O C O 0 O O O O 0

CHAPTER 1. Introduction . . . . . .

CHAPTER 2. Literature Review: The Physical

Correlates of UltraSound in Young Rodents . . . .

Sonic and Ultrasonic Communication in

Dispersal Sounds; Contact Calls .
Voiced Sounds . . . . . . . . . .
Ultrasounds in Infant Rodents . .

Signal Characteristics. . . . .

Motivation for Signaling: Temperature

0

Motivation for Signaling: Handling.

Phylogenetic Differences. . . .
Ontogenetical Changes . . . . .
Relevant Receiver - Hearing . .

Rodents

Responses — Maternal Retrieving Behavior. . . .

General Parameters. . . . . . .
Priming. . . . . . . . . . . .
Sensory Parameters. . . . . . .
Effects of Experience . . . . .

Hormonal Effects. . . . . . .
Stimuli Emanating from Infants.

Ecological Considerations. . . . .
Localization. . . . . . . . . .
Discrimination. . . . . . . . .

Predators. . . . . . . . . . .

summary 0 O O O O O O I O . O O O 0

CHAPTER 3. Species-Specific and Ontogenetic
Characteristics of the Peromyscus "Distress" Cry.

iv

Page

vii

10

12
13
15

16
19
21
23
25
28

29

30
31
32
33
37
4O

41
41
42
43

44

Page

. . . . . . . . 45
. . . . . . . . 48
. . . . . . . . 51
. . . . . . . . 54

Introduction. . . . . . . . . . .
Procedure. . . . . . . . . . . .
Recording Equipment . . . . . . .
Sound Analysis Methods. . . . . .
Results:

Species Differences - General Observations. . . . 57
Species Differences - Frequencies. . . . . . . . 71
Species Differences - Temporal Parameters. . . . 79
Age Differences - General Observations. . . . . . 83
Age Differences - Frequencies. . . . . . . . . . 83
Age Differences - Temporal Parameters . . . . . . 98

Discussion. . . . . . . . . . . . . . . . . . . .. 103

CHAPTER 4. Effectiveness of Auditory Stimuli in
Eliciting Maternal Searching: Activation Test. . . .. 113

Subjects. . . . . . . . . . . . . . . . . . . . .. 113
Experimental Apparatus. . . . . . . . . . . . . . . 114
Test Sounds. . . . . . . . . . . . . . . . . . . . 117
Procedure. . . . . . . . . . . . . . . . . . . . . 118
Variables. . . . . . . . . . . . . . . . . . . . . 120
Results. . . . . . . . . . . . . . . . . . . . . . 121
Discussion. . . . . . . . . . . . . . . . . . . . . 130

CHAPTER 5. Species-specific Discrimination Based on
Auditory Cues: Discrimination Test. . . . . . . . . . 132

Subjects. . . . . . . . . . . . . . . . . . . . . . 133
Equipment . . . . . . . . . . . . . . . . . . . . . 135
Procedure . . . . . . . . . . . . . . . . . . . . . 135
Variables . . . . . . . . . . . . . . . . . . . . . 137
Results. . . . . . . . . . . . . . . . . . . . . . 137

Discussion. . . . . . . . . . . . . . . . . . . . . 144
Comparison of Results of Activation and
Discrimination Tests. . . . . . . . . . . . . . . . 148

CHAPTER 6. Complete and Attenuated Distress Cries
as Sound Targets: Localization Test. . . . . . . . . 158

Subjects. . . . . . . . . . . . . . . . . . . . . . 159

Equipment . . . . . . . . . . . . . . . . . . . 160
Procedure . . . . . . . . . . . . . . . . . . . . . 163
Results. . . . . . . . . . . . . . . . . . . . . . 165
Discussion. . . . . . . . . . . . . . . . . . . . . 170

(SHAPTER 7. Ultrasonic Distress Cries and Maternal
lResponse: A Simple Communication Paradigm . . . . . . 174

Why does the pup signal?. . . . . . . . . . . . . . 177
What is the signal?. . . . . . . . . . . . . . . . 179

Page
What is the message?. . . . . . . . . . . . . . . 181

Situation 0 O O O O O C O O O O O O O 181

SpeCieS. O O O O O O O O O O O O O I O O O O O 182
Age. 0 O O O O O O O O O O O O O O O O O O O O 182
Localization. O O O O O O O O O O O O O O O O I 184

Noise and Distortion. . . . . . . . . . . . . . . 185
What does the receiver do?. . . . . . . . . . . . 186
The meaning of the message?. . . . . . . . . . . 188
Literature Cited. 0 O O O O O O O O 0 O O O O O O O 194

Appendix A. O 0 O O O O O 0 O O O O O O O O O O O O 204

vi

Table

3-1

3-2

4-2

LIST OF TABLES
Ages of eye opening and ear opening in Species
of Peromyscus ................................
Distress squeak frequencies from 8 taxa of
Peromyscus ...................................
Temporal characteristics of the distress call

Of infant Peromxscus OOOOOOOIOOOOOOOO....00...

 

Distress squeak frequencies from Peromyscus

maniculatus bairdi pups of selected ages .....

 

Comparisons of estimated ear-ear head width

measurements of 3 taxa of Peromyscus with mean

 

basic frequencies of distress squeaks of pups
of the same taxa .............................
Measures of homogeneity of reproductive state
and responsiveness of Peromyscus maniculatus
mothers in 3 experimental groups .............
Mean responses and mean response latencies of
Peromyscus maniculatus mothers to P, manicula-
tus distress squeaks and control sounds in
Activation test ..............................
Mean distance scores and mean choice reversals
of Peromyscus maniculatus mothers responding

to distress squeaks of g. maniculatus pups and

vii

49

78

80

88

110

123

125

 

gable Title Page
control sounds ............................... 128
5-1 Comparative protocol of stimulus presentations
to experimental groups in the Activation and
Discrimination Tests ......................... 136
5-2 Measures of Homogeneity of reproductive state

and responsiveness of Peromyscus maniculatus

 

and P. leucopus mothers in four experimental

 

groups of the Discrimination test ............ 138
5-3 Mean number of responses and mean response

latencies of Peromyscus maniculatus and E.

 

leucopus mothers to distress squeaks of pups

 

of both species and control sounds in the
Discrimination teSt 0....OOOOOOOOOOOOOOOOOOOO. 140
5-4 Mean distance scores and mean choice reversals

of Peromyscus maniculatus and P. leucopus

 

 

mothers responding to distress squeaks of pups
of both species and control sounds ........... 143

5-5 Comparison of the Responses of g, maniculatus

 

and P. leuCOpus mothers to squeaks of con-

 

specific pups in control situations and in
ChOiCe Situations O0.0.0.0000...00.00.000.000. 152

6-1 Latencies of Peromyscus maniculatus mothers

 

entering the Fan box and arriving at the
source of live or recorded distress cries of

infant Peromyscus ............................ 167

 

viii

Table

Title Page

Alley choices of Peromyscus maniculatus

 

mothers in Fan box, searching for source of
live or recorded infant distress cries ......, 169
Paired comparisons of distress squeak para-

meters in 4 taxa of Peromyscus. S = sympatric,

 

A=a110patrico coo-9000.00.00.000000000000... 183

ix

LIST OF FIGURES

Figure Title Page
3-1 Mouse restraining chamber for squeak recordings. 53
3-2 Oscillograms of distress cries from 8-day old

Peromyscus pupSOOOIOOOOOOOOOOOOOOOO0.0.0.000...O 59

 

3-3 Oscillograms compared with sonagrams............ 61
3-4 Sonagrams of single squeaks from-recordings of

8-day old Peromyscus pups.................. ..... 64

 

3-5 Comparison of oscillogram and sound spectrum

graphs.......................................... 66
3—6 Sonagrams of single distress squeaks............ 68
3-7 Sonagrams of single squeaks..................... 70
3-8 Sound spectrograms of single squeaks compared

with oscillograms of sets of squeaks............ 74
3'9 Standard sonagrams of single distress squeaks

compared with intensity contours of the same

squeaks......................................... 77
3‘10 Sonagrams of single squeaks..................... 85
3‘11 Oscillograms and sound spectrograms for P.

mgniculatus bairdi pups of different ages....... 87

 

3-12 Oscillograms and sound spectrograms for addi-
tional distress squeaks from P, maniculatus

bairdi pups showing ontogenetical changes

X

Figure

3-14

Title Page
and differences between Type I and Type II
squeaks........................................ 92
Oscillograms and sound spectrograms from g.

maniculatus gracilis pups; distress squeaks

 

for several ages and pain squeaks.............. 95
Sound spectrograms and oscillograms comparing

P. leucopus with P, maniculatus bairdi and P.

 

 

pOlionOtuS across ages.OOOOOOOOOOOOOOOOOOOOOOOO 97
Number of squeaks per set: Daily mean numbers

of squeaks in a group plotted against age of

100

pupSOOOOOO0.00....0OOOOOOOOOOOOOOOOOOOOOOOOO...

Set durations: Mean duration, in seconds, of

sets of squeaks, plotted against age of pups... 102
Squeak duration: Mean duration of distress

squeaks, in seconds, plotted against age of
pups............................................ 105
Sound Choice Box................................ 115
Distance score to sound sources in Activation

and Discrimination tests........................ 147
Mean numbers of responses to the sound sources,
averaged for each mouse over 4 trials in the
Activation and Discrimination tests.............. 151
Mean response latencies in the Activation and
Discrimination tests............................. 154
Mean choice reversals in the Activation and

Discrimination tests............................. 156

xi

Figure Title Page
6-1 Fan Box.......................................... 162
6-2 Diagrams of relative percentages of responses

in correct and incorrect directions by Peromyscus

maniculatus mothers searching for sounds of live

and recorded distress squeaks.................... 172
7-1 Model of a simple communication system show—

ing pathways of information transfer............. 176
7-2 Schematic representation of distress squeak

model............................................ 190
A-l Schematic of the electronic recording and

analysis pathways................................ 206

INTRODUCTION

The distress cries of an infant mammal usually
bring a warm, nourishing parent to carry it back to a
secure nest. This et—epimiletic/epimiletic behavior se-
quence is frequently observed but seldom considered as
more than an example of maternal behavior. Most studies
of retrieving have concentrated on the physical and
lhormonal aspects of the behavior of the mother. However,
'the behavior of the mother is related to the behavior of
'the infant by the communication of information. The help-
lxass infant transmits a message through some channel by
nuaans of a signal and that signal is received, decoded,
arui acted upon by a receiver, the mother. In the follow-
irng study I looked at the behaviors of crying infant/re—
trtieving mother as a communication system and by analyzing
the: Characteristics of the signal and the response of the
receiver I attempted to decipher the code.

As the original impetus for this study grew out of
a genneral interest in mammalian sounds it seemed appropri—
ate tx: use rodents as experimental animals,since their
retrdxaving behavior has been well analysed. Rodent sounds,
howevery have only recently become susceptible to analysis
as equipment has been developed to cope with their very

high frequencies.

2

In terms of communication theory five channels are
available for transmitting information such as the dis-
tressed state of a rodent pup to its parent; touch, taste,
olfaction, vision and sound. Although each has intrinsic
advantages and disadvantages, sound appears to be the pre-
ferred channel for distance communication among rodents.
Sounds can be directly tuned to the ears of the receiver,
they do not linger on the air, and they can be easy or
difficult to locate, depending on characteristics of the
sound. Another peculiar advantage of the sound channel is
that signals can be produced without ceasing any ongoing
activity and they can be received by a moving or even a
sleeping animal. Distress signals emitted by infant mam-
mals are usually sounds.

Sounds are conveniently the signals most amenable to
electronic analysis and the most easily separated from
Wnoisy" surroundings. Echolocation signals of numerous
nunnals have been studied in detail but the other sounds of
otflaer mammals such as rodents, have only received incidental
anrtention. In part this has been due to the generally noc-
tnxrnal and secretive habits of rodents and in part to the
fatrt that many rodent calls are above the range of human
hearing, being partly or completely ultrasonic (i.e., above
20 kHz).

Deermice (Peromyscus) are typical of rodents in

 

thal: they make chirps and squeaks with both sonic and ultra-

SODjJZ components. Their nocturnal, retiring habits make

3
sound an important communication channel for distance sig-
nalling. Because their behavior patterns are well known
(Eisenberg, 1962; Horner, 1940; King, 1968) these mice
make excellent subjects for a study of their sounds as
behavioral stimuli in communication systems.

Although deermice are known to produce sounds when
fighting, courting, when disturbed, excited or frustrated,
and in other situations of unknown significance (Eisenberg,
1962; King, 1968; personal observations) few of these
sounds have been recorded and none analysed. As adult mice
are often inhibited from vocalizing by.the necessary re-
strictions of recording situations the only mouse sounds
which have so far been examined in any detail are those
which can be elicited involuntarily from either adults or
young.

Berlin (1966) found that adult pain squeaks to
shock had major intensities at 50 and 90 kHz. My own pre-
liminary observations and published data from Housknecht
(1968), King (1963), and Hart and King (1966) showed wide,
arui variable ranges of maximum sound energy for the infant
(histress cries of several species and subspecies of £339:
Hmnscus. The present study concentrated on infant distress
(Hties because they can be reliably elicited and because
tluey form part of a communication system in a way that the
Efixin.squeaks do not. Preliminary studies showed that these
<$ries varied across species and possibly ages in a visible

and quantifiable manner, they are reliably reproduced by

4
pups isolated momentarily from their nests, and they elicit
a predictable response from adult mice, i.e., retrieving.
Despite the small amount known about the physical
characteristics of the sounds, several studies have pro-
vided data on the hearing acuity of mice in this genus.
Dice and Barto (1952) found ear-twitch responses to sounds

in certain species of Peromyscus up to 100 kHz. Ralls

 

(1967) found evoked auditory potentials to tones of 100

kHz in P.1eucopus and P.boylei. It is therefore apparent

 

that the hearing of deermice is quite adequate for all of
the frequencies they are now known to produce.

Rowell (1960) defines retrieving as "the response
of the mother to the situation 'young outside the nest'."
Rats, mice, hamsters and many other rodents respond to
distress squeaks by approaching a squeaking infant, grasp-
ing it firmly in their teeth, and carrying or dragging it
back to the nest. Retrieving can be elicited from virgin
females and males, as well as lactating females, so it is
not strictly dependent on the hormonal state of the adult,
lout rather on the complex of stimuli presented by the in-
fant (Beach & Jaynes, 1956a, 1956b; Bell, 23 21, 1971;
Ihorner, 1947; Beniest-Noirot, 1958; Richards, 1967; Rowell,
1J960). Anosmic rats and vibrissae-less or vibrissae-
denervated rats will still retrieve (Beach & Jaynes, 1956a) .
However, Bell, SE §_l_ (1971) found that during recording
sessions when no vocalizations were detected, mother house-

mice neither retrieved infants nor groomed them. Sounds are

5

mentioned by Noirot and others as the factor which may
first attract a mother's attention (Beniest-Noirot, 1958;
Noirot & Pye, 1969; Zippelius & Schleidt, 1956).

Rodent parents appear to recognize their own young
on the basis of the various stimuli emitted by the infant
even though these stimuli change as the infant grows (Beach
& Jaynes, 1956a; Noirot, 1960; Richards, 1967; Rowell,
1960). Hamsters attack and eat strange pups depending on
the age but retrieve their own (Rowell, 1960). One major
component cue for this recognition is odor, for Beach and
Jaynes (1956a) concluded that their experimental rats could
learn to recognize their own young by odor, after having
retrieved only one. Sound is proably sufficient for dis-
tance identification, at least of species, since while

prairie deermice (Peromyscus maniculatus) do not respond

 

to the calls of infant California deermice (P.californicus)

 

they readily search for calling pups of their own species
(personal observations).
With the above considerations in mind I designed
'this study to explore the nature of the squeaks of isolated
nuouse pups, and to examine the behavioral responses of
HKJther mice to these signals. In the first part of the
Stnidy the distress cries of pups of several species of

Peromyscus were recorded and analysed with an oscilloscope

 

arni sound spectrograph. I hypothesized that if distress
Sqnaeaks of these mice have evolved as specific signals

alxerting an adult to the plight of the squeaker and

6

attracting and directing her search, as quickly as possible,
then the squeaks should show typical characteristics of
easily localizable sounds. I expected that the squeaks
would have sharp rise and fall times, pure tones with short
wave lengths and maximal sound energy at levels best heard
by adult mice and intensity modulation. These characteris—
tics would enable the relevant hearers to locate the sounds
by means of time differences, phase differences and dif-
ferential sound levels at their two ears. For information
transfer with minimal error the sounds should incorporate
some maximum amount of redundancy. As the mouse pup opens
its ears and hears its own sounds all these characteristics
could be refined to some species typical pattern and there—
fore show a reduction in variability with age.

Preliminary analyses indicated that intensity modu-
lation could only be compared visually from oscilloscope
films but not measured as it was impossible to maintain a
stable baseline. Frequencies were precisely measured with
the sound spectrograph and frequency modulations were
(depicted by the sonagraph. Temporal characteristics of
number of squeaks in a set, and the duration of squeaks
and sets could be measured on the oscilloscope films.
(3ther characteristics of the squeaks such as onsets and off—
mats and harmonic structures were clearly visible on the
sonagrams.

In the second part of this study I observed the be-

havior of lactating female deermice in modified open field

7

test boxes when they heard recorded and live distress
calls. Studies on retrieving in mice indicate, in general,
that lactating females respond very quickly to squeaking
pups but not to quiet ones (King, 1963; Noirot, 1958;
Sewell, 1970). If the sound alone is a sufficient stimulus
to a mouse to retrieve, then I hypothesized that appropri-
ately recorded sounds ought to stimulate retrieving even
in the absense of the other stimuli from a living pup. If
the stimulus is strong enough to act in a natural environ-
ment under a situation of stress, as for instance where a
mother is moving a nest disturbed in daylight, then it
ought to be sufficiently strong to lure a mother mouse out
into an open arena.

In two tests designed to test the hypothesis that
the squeaks alone were sufficient and necessary to stimulate
mother mice to search and enable them to discriminate,
females with litters were housed in the center of a long,
sound-choice box (SCB) which had a speaker at either end.
Recorded squeaks and control sounds were played alternately
from opposite speakers and the approach responses, or lack
of approach, of the mothers were recorded. Responses to
squeaks of conspecifics were compared to responses to
squeaks of pups of other species or mechanical sounds.

In a third test designed to measure the localization
abilities of the mice lactating females were allowed to
entxar a fan—shaped test box at the center when sounds were

Played from speakers randomly located on the periphery of

8

the fan. The accuracy of their responses was scored and
responses to recorded squeaks were compared with responses
to squeaks from live pups.

By means of this two—pronged approach, first, the
examination of the characteristics of the signals, and
then measurement of the responses of the relevant receiver
I attempted to define the possible and probable information
carried in the message of the distress calls of the infant
deermouse. I suggest (anthrOpomorphically) that the squeak-
ing pup is not just shouting "help" but something more like

"Here I am, a 6 day old Peromyscus maniculatus bairdi baby,

 

out of my warm nest, come and get me, right here.

CHAPTER 2

Literature Review: The Physical and Functional

Correlates of Ultrasound in Young Rodents

While it may be intrinsically obvious to the non—
ethologist that mammalian mothers usually rush to the
"rescue" of their helpless babies, the inquiring etholo-
gist cannot accept such a truism unexamined. A host of
questions is contained in this seemingly simple activity.
As this study was designed to answer some of those ques-
tions, for one species, a review of the pertinent litera-
ture is appropriate.

In this review I will attempt to establish a back-
ground for subsequent discussions of several of the most
important questions. Considering rodents, in general, and

Peromyscus, in particular, what do we know about the im-

 

petus to the rescue? What role does the helpless infant
play in regard to the mother's behavior? Are the "distress"
cries of the infants necessary stimuli to the maternal
search?; do they help the mother locate the infant? What
other information is contained in the cries, if any? Are
there other functions possible for the calls or other cues
for the retrieving mother? What prompts the pup to call?

What are the characteristics of the calls and how do they

10
vary ontogenetically and phylogenetically? Finally, of what
ecological (and selective) advantage is such a communication
system to the animals involved?

In the past there have been two general experimental
approaches to the "baby cries — mother comes" situation.
Infant rodent calls have been analyzed in terms cf rate and
intensity and the variations of these related to age, hand-
ling, or temperature, and described in terms of frequency.
Meanwhile, the stimuli eliciting various maternal behaviors
have been examined. In the experiments and analyses de-
scribed here I attempted to consider the communication sys-
tem as a whole, looking at the behavior of both mother and
pup and at their interaction. This literature survey will

reflect this attempt at synthesis.

Sonic and Ultrasonic Communication in Rodents:

F The audible squeaks and twitterings of small rodents
have long caught the ears of listening naturalists. Un-
usually long or high squeaks from "singing" mice were com-
pared to bird calls and sparked imaginative discussions as
to their mode of production and function. (Dice, 1932).

Studies of ultrasonic echolocation signals in bats
Prompted searches for sounds with similar function in other
small mammals (Gould, 33 21., 1964; Kahmann & Ostermann,
1951). In the case of rodents no specific echolocation
sounds have been demonstrated, although incidental sounds,

snufflings, footsteps and other sounds may well aid in

ll

rodent navigation in the dark (Anderson, 1954; Kahmann &
Ostermann, 1951; Rosenzweig, Riley & Krech, 1955).

Sounds used for communication have been demonstrated
in almost all rodents (Tembrock, 1963, 1968) but precise
physical and functional analyses are rare. In part this is
due to the enormous technical difficulties inherent in re-
cording very high frequency sounds from easily disturbed
animals. Within the last five years or so, advances in
acoustic recording equipment has made it more possible, al-
though expensive, to do detailed analyses of sounds of all
frequencies produced by rodents.

Studies of rodent sounds have generally been of two
types, those done as part of an ecological or behavioral
study of the animal, or those in which the frequency ranges
of single sounds of many diverse rodents have been examined.
Most of the data have been obtained from diurnal, social
rodents.

The sounds are usually differentiated in terms of
the environmental situation in which they occur, the pre-
sumed physiological state of the animal, and the subjective
appearance of the sound to the human observer. (For example,
see Balph & Balph, ground squirrels, 1966; Fitch, 1948,
ground squirrels; King, 1955, prairie dogs; Rood, 1972,
cavies; Ruffer, 1966, grasshopper mice; Waring, 1966, mar-
mots.)

Particularly in field studies, the social function

12
of a particular sound may be unclear. Listening for "mating"
calls, or "abandonment" cries may be inappropriate for
studies of rodent communication where one sound may be
uttered in several seemingly different social situations
(Brooks & Banks, 1973). When considered in terms of infor-
mation sent out by the caller, relevant to its individual
situation rather than to the social situation as a whole,
calls can be grouped into two functional systems (Tembrock,

1968).

Dispersal Sounds; Contact Calls:

The first group may be considered dispersal sounds,
serving to increase the distance between sender and receiver
in agonistic and defensive situations. The second group of
sounds are used primarily between conspecifics as self-
advertisement and contact-soliciting calls. In the simplest
form these calls carry information about the species, sex,
age, and location of the emitter. Reception of this infor-
mation may reduce the impetus to flight or fight on the part
of the relevant hearer or elicit its approach to the caller.

Both vocal and non—vocal (that is not produced by
the vocal cords) sounds may fit in either category and func-
tion in audible communication. Unvoiced sounds, however,
are predominately"white'noise, often involuntary warning
sounds. While the tempo of paw—tapping in gerbils, for ex-
ample, can be varied with the intensity of the situation,

the frequencies cannot be modified and the repertoires are

l3

extremely limited.

Voiced Sounds:

 

Voiced sounds are therefore, more interesting,having
more complex codes and variations possible. Sound reper-
toires in rodents are not as extensive as those of primates,
but sound categories may be more discrete and non—overlap-
ping.

Vocalizations reported in marmots (Armitage, 1962;
Lloyd, 1972; Waring, 1966), ground squirrels (Balph & Balph,
1966; Fitch, 1948), prairie dogs (King, 1955), squirrels
(Zolley, 1971), laboratory rats (Anderson, 1954), and sever—
al genera of mice (Eisenberg, 1968; Layne, 1968; Ruffer,
1966; Sewell, 1968; Tembrock, 1968) function in social
situations, a majority of the sounds being used in estab-
lishing or maintaining contact.

Observations of the bank vole (Clethrionomys glareo-

 

lug) convinced Schleidt (1948) that the two types of high
squeaks they produced were used in communication. A "squeak-
ing hiss" was heard in conflicts over food and a high, weak,
squeak was described as a contact sound. In 1951, Schleidt
further described the situations in which the bank vole
produces high-frequency sounds. In the presence of females
or near nest areas the males of this species gave short
squeaks of about 15.5 kHz, possibly contact calls. The
sounds of young mice were noted by Zippelius and Schleidt

(1956) in Apodemus, Mus, and Microtus. Here the authors

 

 

14

noted a loud clicking, sucking sound with the greatest in-
tensity between 5 and 10 kHz and some frequencies to 40 kHz.
A squeak with components to 80 kHz was observed and this
"abandonment" whistle caused the female to notice a calling
infant and retrieve it. When the pups were picked up by
the scruff of the neck, by the experimenter, another sound
was noted, having components up to 100 kHz.

Houseknecht (1968) used sonagrams to analyze the

vocalizations of adult Mus, Peromyscus leuCOpus and Micro-

 

tus pennsylvanicus. Vocalizations were recorded in ag-

 

gressive interactions in the laboratory. Most of the ag—
gressive encounters with all 3 genera were accompanied by
very short "squeals" or "chits." In non-contact threat
situations, where the provenance of the sound could be de-
termined, only the subordinate mice vocalized, possibly a
defense/spacing cry. Sounds of the three genera roughly
compared.

Barfield and Geyer (1972) found a sound emitted by
the post-ejaculatory male rat with components at 22 kHz and
a duration of about 1—3 seconds. Brooks and Banks (1973)
in their exhaustive catalogue of lemming sounds found a
similar male call. In both cases these appear to be
"spacing" sounds (desist contact) similar to that heard
from supine subordinate male rats. The rat sounds show
sharp onsets and offsets characteristic of easily locali—

zable sounds.

15

Brooks and Banks (1973) recorded and analysed 6

types of calls from lemmings (Dicrostonyx groenlandicus)

 

in terms of behavior and acoustic characters. They found
wide variation in the physical properties of the sounds and
emphasized that the variation in social situations in which
particular sounds were found show that emission of sounds is
dependent on the internal physiological state of the animal
and the external situation and not on a conscious attempt
on the part of the animal to communicate information (Brooks-
& Banks, 1973).

Sales (1972) detected ultrasounds during aggressive

interactions in lab rats, Apodemus, Acomys, Thamnomys,

 

 

 

Arvicanthis, Praomys, Mus, Mesocricetus, Meriones, Gerbillus,

 

 

 

 

Clethrionomys, Ligurus, Microtus and Sorex. This study grew

 

 

out of the observation that the "distress" cries of young
rats could be elicited after the usual cessation of squeaking
at eye opening if the pups were handled. The handling often
forced the older animals into submissive postures. Submissive
adult rats produced long pulses (800-1600 msec.) having
narrow bandwidths (less than 5 kHz) and attacks on these rats
were reduced. Short pulses (3-65 msec.) with wide bandwidths
(5-10 kHz) were highly correlated with escape behavior and
aggressive acts. The other Species of rodents observed
showed similar patterns of ultrasound production during
aggressive encounters.

Ultrasounds in Infant Rodents:

Ultrasounds of young rodents are easily elicited from

16

isolated pups and have been the focus of several studies.
Sewell (1968) found ultrasounds reported for young of all
species of myomorph rodents - usually having the function
of inhibiting attacks and aiding the retrieval of the pups.

Ultrasonic calls recorded from 5—day old Apodemus

 

sylvaticus pups initiated retrieving responses in listen-

 

ing mothers (Sewell, 1968).

Clicking, sucking sounds reported in small rodent
pups by Zippelius and Schleidt (1956) were also described
by Hart and King (1966), and Eisenberg (1968) for

Peromyscus pups. Personal observations agree with Eisen-

 

berg's indication that these sounds are physical con-
comitants of sucking, producing a sound when the pup con-
tinues to nurse after being removed from the nipple.
Although these sounds are sometimes discussed as vocal-
izations they are non-vocal and of no discernible
communicatory value.

Two vocalizations have been generally found in young
rodents, a long, high sound, often called an "abandonment"
or “distress" cry, and a shorter squeak resulting from
rough handling, the "pain" cry. Eisenberg (1968) also
mentions "comfort" sounds, described as "low, rapid peeps
made by the neonate while the female grooms it." No other

authors mention this sound.

§ignal Characteristics:

The "distress" cries which are the main subject of

17

this dissertation have been described as being "barely
audible, high-pitched squeaks" (Layne, 1968). More
precise descriptions are given for various species of

Peromyscus by Eisenberg (1968) and Hart and King (1966).

 

Eisenberg found 2 types of distress cries in these mice;
one, a low (fundamental about 3000 Hz) sound rising about
500 cycles and falling with sharp onsets and offsets, the
other a high (fundamental approximately 25 kHz) sound with
a similar pattern to the first but repeated rapidly. The

first was found in Peromyscus maniculatus, g. crinitus, and

 

P. californicus and the second was found in these mice but

 

also in E. eremicus. These cries have complex harmonics
and Eisenberg states that the greatest energy may be found
above the fundamental, with the important communication
sounds being ultrasonic.

Hart and King (1966) described the distress vocal—

izations of pups of 2 subspecies of a. maniculatus as being

 

short pulses (0.1 — 0.2 sec.), in groups of 2 - 5, with
dominant frequencies up to 26.5 kHz (P,m.gracilis). The
lowest frequency band was not found to be the most intense
but the greatest sound energy was usually in an upper band.
Frequencies above 30 kHz were not measured.

At the time that I began this study (1968) the
descriptions above were the most precise available. Com-
parisons of descriptions of the sounds within the genus

Peromyscus and with other rodent genera were difficult,

 

18

and still are, because of a general lack of precision and
standardization of terms such as "fundamental." Many
authors appear to consider the fundamental to be simply
the lowest frequency visible on the sonagram and often
attribute an enormous frequency spread to it, for example,
300 Hz - 25,000 Hz (Eisenberg, 1968). Ladefoged (1962)
defines the "fundamental frequency" as "the frequency of
repetition of a complex wave" or "when a repetitive wave
form is analysed into its component frequencies the
fundamental is the highest common factor of the component
frequencies." He goes on to state that "the component
with this frequency may have a large amplitude,... a small
amplitude, or...there may not be any component with this
frequency." That frequency with the most energy may, in-
stead, be called the "basic frequency" particularly when the
wave form under consideration isla non-repetitive type
which does not really have a fundamental frequency (Lade—
foqed, 1962). Some authors listing fundamentals admitted
that these might not be of major communicatory importance,
whereas others reported different frequency measures (as
"low and high frequency limits of the dominant frequency";
Hart and King, 1966). Sonagrams provided a visible com-
parison but subtle intricacies of sound patterns do not
reproduce well by this method and quantitative comparisons
are impossible without equivalently calibrated machines.
Since 1966 the distress cries of white mice, white

rats, hamsters, and lemmings have been described in detail

19

by Allin and Banks (1970), Brooks and Banks (1973), Noirot
(1966b, 1968), Noirot and Pye (1968), Okon (l970a,b, 1971),
Sewell (1968, 1970) and Smith (1972). Only Smith has dealt

with Peromyscus.

Data reported by Smith (1972) are for P. maniculatus

 

(unidentified subspecies) and describe an "audible" squeak
of 15 - 200 msec. duration, consisting of 2 or 3, occasion—
ally 4, constant frequency components; a fundamental and
2nd and 3rd harmonics. The fundamental at 8-29 kHz was
usually the principle component. Pulses are described as
rising by 5 kHz at the onset and falling by approximately
the same amount at the end. "ultrasonic" pulses of 3-45
msec. duration, consisting of a single frequency component
with a rapid downsweep were found in the 60-140 kHz range.
Smith also found a low frequency (3 kHz) multi—component

pain squeak.

Motivation for Signaling: Temperature.

Although Hart and King (1966) and Eisenberg (1968)
had suggested cold stress as a motivational factor in
signaling, no detailed studies of such factors appear until
(Dkon (1970). When Hart and King used cold stress to elicit

‘vocalizations from their Peromyscus they did not mention

 

'the exact temperature of the surroundings of the pup during
Inacordings. They suggest that the development of homeo-
1flnermy may be related to the cessation of squeaking at eye

Opening in these mice.

20

Noirot and Pye (1969) found that ultrasound produc-
tion by albino mouse pups showed a sharp drop in intensity
and frequency on days 6 and 7 and suggested that this might
correlate with the development of homeothermy. This drop
was in direct contrast to the results of a study by Okon
(l970a) who reported that albino mouse pups produced weak
ultrasounds until day 5 or 6 and stronger responses there—
after. Okon (l970a,b; 1971) considers that the response
examined by Noirot and Pye was due to another motivational
factor, handling, as these authors picked up the pups, or
rocked them in order to provoke squeaks;

Okon (1970a, 1971) studied the development of homeo-
thermy in albino mice, rats and hamsters and the concurrent
changes in ultrasounds. He found that albino mice are
almost completely poikilothermic at low temperatures from
day 1 until day 6. From about day 7 they show homeothermy
at median temperatures (22°C) and at day 15 they are
developing homeothermy at low temperatures (2-3°C). At
about this time the pups stop producing ultrasounds when
merely exposed to low temperatures. By day 19 or 20 no
calling is elicited even with temperatures as low as 2—3°C.
The acoustic response to cold is strongest from days 6-7
to day 12 and weak before day 6. Low ambient temperatures,
in this study, led to increased rate and intensity of ultra-
sounds, although the very low temperatures had an inhibiting
effect. With severe chilling, sounds cease altogether,

along with breathing, after a time which depends on the age

21

of the pup.

Allin and Banks (1970) used cold stress (2°C and
20°C) to elicit sounds from albino rat pups. They found
that after 8 days of age the call rate declined in a
fashion significantly correlated with age and possibly with
the development of homeothermy. Heat stress (40°C) was not
found to elicit calling and neutral temperatures (35°C)
elicited only infrequent calling.

There can no longer be any doubt that some of the
ultrasonic (and sonic) calls of infant rodents are casually
linked to the degree of cold stress acting on the pup. As
the physiologic temperature regulation of the pup matures,
cold stress may not be felt as soon, or as strongly, and the
lesser stimulation is reflected in reduced intensity and rate
of calling. At the age at which the homeothermic system is
sufficiently adult to protect against all but the coldest
temperatures, the pup is rarely stimulated to squeak. Co-
incidentally this is also the point at which the eyes

usually open.

Motivation for Signaling: Handling.

Noirot (1970) suggests that 2 types of distress call
exist; one emitted during cold-stress and one during handl-
ing. She postulates that the first call serves as a
stimulus for retrieving and the second results in a cessation
Of parental activity, particularly attack.

Bell, Nitschke, Gorry and Zachman (1971) considered

22

the stress of handling to be a factor in the elicitation

of squeaks from P. maniculatus bairdi. The authors re-

 

moved pups from the nest, placed them in small compart-
ments and kept them there for 3 minutes. After they were
returned to the nest, recordings were made of the ultra-
sounds produced by these "handled" pups and non-handled
(left in the nest) controls. No mention is made of whether
the temperature in the box was maintained as a neutral
temperature but the authors attribute significant dif-
ferences in rate (number of signals per session) and

higher frequencies with longer mean squeak duration, which
they found in the ultrasounds produced by the experimentals,
to the handling. However, sonagrams presented with the
data do not make the frequency differences clear, as the
sounds pictured for the "handled" mice resemble typical
cold-stress cries.

Okon (1970b) does clearly differentiate the 2 types
of signal. Handling by the observer, loss of balance, and
actual retrieving by the mother, all discussed as "tactile
stimuli" by Okon, evoked ultrasounds from albino mice. The
loudest pulses were produced by pups from day 1 on, while
at this age only weak signals are produced during temperature
stress (Okon, 1970a). After day 6 pups are less likely to
fall over and the tactile stimulus squeak is less often
produced. Okon also describes a "pain" squeak to a pinch
or tap on the tail which was found to be louder than the

other squeaks, with frequencies ranging from 80-100 kHz

23

(l970b). Okon suggests for the youngest pups the greatest
danger is from cannibalism, so these produce the "handling"
call most loudly, inhibiting attack from adults. Older
pups (after day 6) are more liable to move, or be moved,
out of the nest and so squeak most loudly to cold-stress.
In any particular situation the 2 types of call may be
mixed (Noirot, 1972).

Smith (1972) differentiates between audible (defined

as below 30 kHz) and ultrasonic calls in Peromyscus. From

 

the mouse's "point of View, of course, audible versus
ultrasonic is a spurious distinction. Neutral terms related
to the function of the calls would have been more realistic.
Whatever the name, Smith finds significantly different
frequencies and loudness between the 2 types of call. The

"audible" pulses are emitted in situations of cold stress

while the ultrasonic pulses seem to be "handled" calls.

Phylogenetic Differences:

 

Eisenberg (1968) and Hart and King (1966) have so
far been the only ones to compare sounds of different groups

of Peromyscus. Eisenberg lists different frequencies and

 

measurements of squeak duration for 4 species of
Peromyscus (1968). The two types of "abandoned" cry dis-
cussed here may be cold-stress and handled cries although
the differences are not very clear. Long (0.40-0.50 sec.),

low (4-22 kHz) cries were typical of P. californicus

 

parcasiticus shorter (0.15-0.40 sec.), low (2.5-20 kHz)

 

24

squeaks were found for P, maniculatus gambeli; while the

 

highest squeaks (to 24 kHz) and shortest (0.08-0.17 sec.)

were found for P. crinitus stephensi and P. eremicus

 

eremicus.

Hart and King (1966) compared 3, maniculatus

 

bairdi and g, maniculatus gracilis and found significant

 

differences in the characteristics of their ultrasounds.
me. gracilis pups produced more pulses per set, at a
higher rate, with a longer interpulse interval than P.m.
bairdi, but the squeaks of the latter had a higher maximum
frequency and a longer pulse duration.' Despite the lower

maximum frequency, P.m.graci1is had a wider spread in the

 

dominant frequency (3.6 - 26.5 kHz) than P,m.bairdi
(4.9 - 23.7 kHz).

Brooks and Banks (1973) compared the sounds of 2
genera of lemmings, and Noirot (1966a,b; 1968), Okon (l970a,
b; 1971), Sales (1972), and Sewell (1968; 1970) have com—
pared sounds across several genera of mice, rats and
hamsters. Nitschke, Bell and Zachman (1971) compared 3
strains of albino laboratory mice, C57BL/6J, BALB/cJ, and
C3H/HeJ, in terms of rate of signaling, peak frequency,
and signal duration. Strains and ages were found to be
significantly different on all parameters. Rate measures
were apparently made from pooled totals of 3 - 10 second
periods separated by 1 minute intervals, a measure not

directly comparable to those in any other studies. "Peak

25

frequency" was unfortunately not defined and it was not
clear whether the signals measured occurred in sets.
Comparisons with other species are thus difficult. The

Hart and King study on the subspecies of Peromyscus (1966)

 

is therefore the only useful study of sound differences

within a natural species.

Ontogenetical Changes:

 

Noirot (1966b, 1972) found that Mug pups showed a
change in squeak rate with age. The rate increased until
the date of ear-opening (day 4) and then decreased almost
to zero on the day of eye—opening. The total number of
squeaks also peaked on day 4 and the percentages of sounds
of short, constant frequency changed relative to the per-
centages of very, high squeaks of long, steady frequency.
This now appears to be similar to the relative changes
Okon (l970a,b) found between "handled" and "cold-stress"
stimulated calls. In 1968 Noirot found similar results
with white rats. Curiously the call rate appeared to be
lowest when the rate of retrieving was highest, that is at
about day 4.

Noirot and Pye (1969) found 3 age divisions for the
characteristics of the sounds. Over age a slight initial
rise in frequency became apparent. By day 4 calls showed
less variation and the calls produced on day 12 were few
in number but showed a significant rise in frequency. The

length of the squeaks decreased with age after day 4 and the

ifi—_—__—_E,—wvr - -,.

26

bandwidth and sound pressure level also dropped. Frequen-
cies were found to be more constant on day 4 than earlier
or later.

Brooks and Banks (1973) also found 3 age divisions
in the sounds of young lemmings. Younger pups, days 3-5,
gave simple calls, very short (0.08 Sec.) and high fre—
quency (to 29.1 kHz). Older pups had more complex calls
with 2 to 3 distinct segments, longer (0.61 sec.) and
lower (26.6 kHz) than earlier. Still older pups produced
longer squeaks (0.70 sec.) with lower frequencies (19.9
kHz). Under cold stress these authors found the rate of
calling by lemming pups to be highest on days 3-4 while
rates of calling to pain (tail pinch) were highest from
4-7 days. Other types of tactile stimuli, touching, roll-
ing between the fingers, elicited audible responses before
day 2 and ultrasonic responses reaching a peak on day 3.
The patterns of change in production of these calls re-
semble Okon's results (l970a,b; 1971).

In Peromyscus, Hart and King (1966) found a signif-

 

icant age trend with a decrease in rate, a decrease in
maximum frequency (P.m.bairdi) or a rise and then a drOp
in frequency (P,m.gracilis), and changes in pulse duration
and interpulse interval.

Okon (l970a,b, 1971) found 3 developmental periods
with respect to rate of calling and intensity of ultrasounds.

In the early period, where the pups are essentially

27

poikilothermic, up to 5—6 days, only few, weak responses
are made. After day 6 the intensity of cries correlates
closely with age, during cold stress, and squeaks are
actively produced. Okon found this period to be transi-
tional for thermoregulation where some control is possible.
During the last stage, after 13-14 days the intensity of
calls depend directly on temperature. Rate of calling de—
clines steadily from an early high.

Smith (1972) describes corresponding stages in

Peromyscus, with "audible" pulses at all times approximately

 

30 db greater than the "ultrasonic" pulSes. For the
"audible" pulses no changes were observed in number of
pulses per set and "ultrasonic" pulses were not produced
in any temporal pattern. "Audible" pulses reached a
maximum rate on day 7 whereas "ultrasonic" pulses were
produced by pups older than 4 days and reached a peak
rate on day 10. In this study recordings were only made
on days 1,4,7,10,12,l4,l7 so gradual trends were not
apparent.

The consensus of the data discussed above must be
that rodent pups do produce ultrasonic (20 kHz and up)
and audible (below 20 kHz) squeaks under the stimulus of
cold stress and handling. The signal may differ between
taxa as small as subspecies and certainly changes in
characteristics occur with age. Rate and intensity changes
appear to correlate closely with thermoregulatory abilities

while other changes may have other causes.

28

Relevant Receiver - Hearing:

On the other end of the communication system the
receiver must be able to hear the sound signal and then
must act on the transmitted message for the information
transfer to be complete.

Reactions to Preyer's ear muscle reflex and dorsal
skin twitches in response to sounds of Galton whistles and
sound generators, indicate sound perception at frequencies
up to 100 kHz in various rodents (Tembrock, 1968; Davis,
1927). With these reflexes Schleidt (1948) found responses

to between 25 and 30 kHz in the bank vole (Cleithrionomys).

 

Preyer's ear muscle reflex is a measure of perception which
may, at worst, overestimate the threshold (Denes, 1961).

Audiograms published as a result of continued exper-
imentation (Schleidt, 1951, 1952) show sound perception in
hamsters and mice acute up to 20-30 kHz and extending to
about 75 kHz. Dice and Barto (1952) found good sound

perception in Peromyscus up to 95 kHz. These authors also

 

found differences in responsiveness between species, with
B. nasutus responding to frequencies as high as 100 kHz.
These differences were believed to be correlated with the
different ear sizes of the mice.

Electronic recording from implanted electrodes pro-
vided Ralls (1967) with more precise measurements of sound
perception in Peromyscus leucopus, P. boylei, and Mus

musculus. The Peromyscus mice were found to have greatest

 

Sensitivity to high frequencies up to about 40 kHz and some

29

sensitivity up to 70 kHz. Younger animals showed better
responses between 30-40 kHz than 6-month old animals. The
albino mice showed a low threshold between 5-30 kHz, with
best response at about 15 kHz. These data agree with the
results of Berlin (1963), who found albino mice most
sensitive at 15 kHz, with a range of sensitivity from 1—70
kHz, using the galvanic skin response.

Thus it seems clear that deermice, and other rodents,
can hear the ultrasonic sounds produced by their stressed
young, and that in the lab at least, these sounds could

serve as stimuli in maternal retrieving.

Responses - Maternal Retrievinngehavior:

 

That the particular activities, such as nest-build-
ing, retrieving, nursing, and cleaning, commonly grouped
as "maternal" behavior must be considered separately in
any discussion of cause and function has been amply demon-
strated by Beniest-Noirot (1958), Lazar (1967), Richards
(1967) and Slotnick (1967). Although each of these ac-
tivities is epimiletic, care~giving (Scott, 1958), in re~
gard to the young of the species, they occur independently
and each appears to be related to different combinations

of casual factors (Richards, 1967). In most cases these

causal factors can be grouped into three classes; the
stimuli from the pups (extrinsic), the previous experience
of the mother (both extrinsic and intrinsic), and the

hormonal condition of the mother (Richards, 1967).

 

30

In a review of maternal behavior in rodents and
lagomorphs, published in 1967, Richards discusses the
role of each of these classes on the several, separate
maternal behaviors. In the present discussion I will
review the data with respect to retrieving behavior and

summarize the additional papers published since 1967.

General Parameters:

 

King (1963) cites field observations of P. leucoEus

and g. maniculatus mothers returning to young scattered

 

during nest disturbances and carrying them to new refuges.
In a laboratory study, P,m. gracilis and P,m. bairdi did
not show significant differences in retrieval latencies,
although there was considerable variation in the behavior
of the subjects while retrieving (King, 1963). King also
noted that the nature of the interactions between the
mother and the pups changed as the pups grew, until no
.pups were retrieved at 30 days of age.

With mice, rats, and occasionally with hamsters,
the mere presence of live pups near a nest is sufficient
to evoke retrieval responses from males as well as virgin,
pregnant, or lactating females (Beniest-Noirot, 1958;
Richards, 1966b, 1967). Richards (1966a) and Rowell
(1960) found that virgin female hamsters killed and ate
pups up to the age of 5 days, but often retrieved older
pups. Noirot (1964) found that with naive female lab mice,

Younger pups, 1 and 2 days old were most effective at

31

eliciting retrieving and other maternal behaviors. As the
pups grew their effectiveness decreased. Both Noirot and
Richards suggest that the change noticed in the retrieving
behavior of the females is related to the change in the
stimulus value of the pups - possibly the distress calls
(Noirot, 1964b; Richards, 1967). The relationship of
retrieving to the presence of ultrasonic distress calls is
also indicated by the fact that the incidence of retrieving
drops off at the age at which the young cease emitting the

calls (Noirot, l966a,b; Zippelius & Schleidt, 1956).

Priming:

Early contact with the strong stimulus of a live one-
day old pup can cause a higher percentage of both male and
female mice to respond to a "weak" stimulus (a day-old
drowned pup) even if it is presented as much as 8 days
later (Noirot, 1964a). Similar results were later found
with hamsters (Noirot & Richards, 1966). This increase in
response probably depends only on exposure to "one or more
essential cues provided by the [pup]" (Noirot, 1964a).

When naive female mice were exposed to a live, but
hidden pup, their maternal responses, including retrieving
were greatly increased when subsequently tested with a dead
pup (Noirot, 1969a). Since there was no common cues, sound
and odor lacking with the dead pup, this increase in re-
sponsiveness is more a "sensitization" or "priming." The

priming of maternal behaviors by initial contact with a

32

strong stimulus is selective. Olfactory cues appear to
increase subsequent pup cleaning, while auditory cues favor
nest-building (Noirot, 1970). Retrieving may also be primed
by auditory cues, but Noirot (1970) hypothesized that this
experimental situation elicited both distress and contact
(handled) calls from the stimulus pups and may have com-
plicated the situation by opposing effects of retrieving

behavior.

Sensory Parameters:

 

Beach and Jaynes (1956c) deprived retrieving mother
rats of vision, olfaction, and sensation around the snout
and found that lack of reception from any one of these
senses reduced their efficiency at retrieving. Earlier,
Weisner and Sheard (1933) found similar results in an ex-
periment where they also blinded the rats. Loss of any two
of these senses produced a greater effect than the lack of
only one. However, Beach and Jaynes (l956o;c) did not con-
sider audition and they regarded retrieving as a whole be-
havior sequence from the first excursion from the next
until the return. Each different sensory modality might
be primarily effective at a particular part of the retriev-
ing pattern. In another paper, the same authors do conclude
that olfaction is an important sense in the discrimination
of known and unknown young (1956a).

The papers discussed above lend support to the theory

that some part of the stimulus configuration supplied by the

33

helpless pup affects the retrieving mother's response. Evi—
dence for the existence and importance of ultrasonic sounds

as a primary cue has already been discussed.

Effects of Experience:

 

The hormonal and experiential state of the mother,
intrinsic factors, have considerable relevance to her readi~
ness to retrieve. Richards (1967) outlined several categories
of previous experience which might affect the retrieving
behavior of a rodent, including in this list: 1) the effects
of manipulations in infancy; 2) the effects of manipulations
during pregnancy; 3) the effects of parity and; 4) the effects
of practice. Other experimental factors mentioned relate
more to other epimiletic behaviors than to retrieving per se,
and so are not included in this list.

The manipulations of the mother as a pup, or her mother
affect the behavior of the mother towards her pups (Levine,
1967; Meier & Schutzman, 1968; Richards, 1967). Oppositely,
Young (1965) demonstrated that the treatments of the young
(rotation or cooling) influenced the mother's behavior towards
them. This may be particularly important if indeed the calls
of the young relate to the aggressive behavior of the mother
to them. In the Young study, non-treated pups were retrieved
more quickly than treated littermates. Meier and Schutzman
(1968) in a review of the data on the effects of experiemtnal
treatments on mother-infant interactions, stress the com—

plexity of the interaction. Any experimental intervention

 

 

34

involving the young also introduces changes in the epimiletic
behavior of the mother (Meier & Schutzman, 1968). It is the
mother's changed behavior which then influences the develop—
ment of the young.

Further evidences of the dynamic interaction between
mother and infant are given by cross-fostering studies done
on lab mice and described by Meier and Schutzman (1968;
Ressler, 1962). In this case survival of fostered young was
found to be better with BALB/c mothers than with C57BL/10
mothers, whether the strain of the pups was BALB/c or C57BL/10.
The authors conclude that the behavior of the mother is "shaped"

by the responses of the pups and vice versa.

 

Any manipulations which alter the emotionality of the
mother as a pup, including her own mother's behavior, may
well affect her behavior towards her own pups. Unfortunately
few studies of these effects on rats and mice measure
differential effects on different "maternal" behaviors such
as retrieving.

A less well explored area is the effect of manipula-
tions during pregnancy. Studies of the effects of total
sound restriction during pregnancy (Levine & King, 1965) and
exposure to anxiety provoking conditions during pregnancy
(Richards, 1967) demonstrate reduced offspring survival and
reduced open field activity. Levine and King (1965) suggest
that the observed effects may be due to a general disorgan-

ization of maternal behavior.

35

Restriction of self-licking during pregnancy, may cause
total failure of maternal behaviors and this may either be
due to the failure of the development of a continuity between
the mother's own body and those of the young or to the under-
development of the mammary glands due to lack of stimulation
(Richards, 1967). At the time of parturition normal rats
switch extensive licking from their own bodies to their young.
Other maternal behaviors increase in rate of emission at this
time also, but the patterns of the switch are not clear.

Experiences with pups have more direct effects of the
development of maternal behaviors. While even primaparous
females make good mothers the consensus of published work is
that subtle behavioral tests can detect improvements in style
during the raising of any one litter and from a first to
subsequent litters (Beach & Jaynes, 1956b; Carlier & Noirot,
1965; Richards, 1967). Disagreement as to the carryover of
improvements from one litter to the next indicates that the
interactions are subtle.

With respect to retrieving in rats, Carlier and Noirot
(1965) demonstrate a reduction in time to retrieve in biparous
rats over their earlier litters. This improvement may be
due to increased skill in transporting the pups, the biparous
mothers carry the pups more efficiently by the central venter
rather than elsewhere (Beach & Jaynes, 1956b; Carlier &
Noirot, 1965; Richards, 1967). Additionally, decreases in

time to retrieve may be due to priming, as evidenced by a

36

decrease in the time spent in the nest by retrieving mice &
ground squirrels (King, 1958; Michener, 1971).

Recognition of the stimuli emanating from the pups
may also play a part in changes due to experience. Given
enough time with pups of their own species, male rodents of
several species and virgins will retrieve and attempt to care
for the pups (Beniest-Noirot, 1958; Gandelman, Zarrow &
Dennenberg, 1970; Richards, 1967; Rosenblatt, 1967, 1970).
Even in natural situations male rodents may care for pups
(Horner, 1947; Horner & Taylo-, 1969). Whether housed in
groups where males and virgins constantly were in contact with
pups (Gandelman, Paschke, Zarrow & Denenberg, 1970) or indi—
vidually offered pups of various ages for several daily
tests (Beniest-Noirot, 1958; Richards, 1967; Rosenblatt, 1967)
generally only slight differences in latency differentiated
these groups from parous females. In a single-trial test

situation, however, male 3. polionotus seemed more critical

 

of the species of test pups than parous females. The males

usually attacked test pups (g. maniculatus) but the parous

 

females never did. In this test both males and females had
lived with litters of pups (experienced). Inexperienced males
and females tended to bite both con-specific and non-specific
pups (S. Huff, unpublished observations).

Although it is apparent that a level of response to
young exists in most adult rodents, even with previously
paroud females the effects of priming are noticeable. With

Richardson's ground squirrel Michener (1973) found that

 

 

37

more non-lactating females retrieved than virgins. This
suggested that with this animal the prior experience of rais-
ing a litter has a lasting effect. In particular the results
of such tests depend on the quality of the stimuli. Very
young pups seem to be stronger stimuli than older ones.

Grota (1973) found that with new-born pups as stimuli the
maternal behavior of rats is related only to the age and the
size of the litter while the previous experience of the mother
is more important with older pups. Noirot (1964a,b,&c, 1965;
Richards, 1967) suggested that contact with a very young pup,
in mice and hamsters can increase responsiveness to older
pups. All of these studies and the discussion in Richards
(1967) lead to the conclusion that the interaction between
mother and pup in a normal situation involves both practice

in performance of the maternal behavior by the mother and
recognition of the stimuli produced by the pup. Both are
constantly being revised by the mother as a result 0f feed-
back from previous contacts with the pups. The stimuli
emanating from the pups may also change in response to the
mother's actions; i.e., a pup may squeak more if not retrieved

or handled roughly.

Hormonal Effects:

 

The hormonal state of the mother relates directly to
her readiness to perform any of these behaviors. Where re-
trieving was made difficult (leaving home cage) Gandelman,

g; a1 (1970) found lactating female mice more willing to

 

 

 

 

38

retrieve than virgins.

Much discussiOn in the literature has centered on the
role hormones play in the induction of maternal behavior.
The role of hormones in the development and maintenance of
the pregnancy which provides the pups is undisputed. However,
Rosenblatt (1970) states that "there is no clear evidence
that the mother requires either ovarian or pituitary hormones
for the maintenance of maternal behavior after parturition."
Differences between the gradual development of maternal be-
havior in virgins and newly parturient females may be as
much due to the physical events of parturition as to the
presence of specific hormones. While the induction of mater-
nal behaviors at parturition may be related to hormones its
maintenance appears to be almost completely due to stimuli
from the pups (Rosenblatt, 1970).

Confusion between the effects of hormones on maternal
behavior and pup stimulation has often arisen from failure
of authors to differentiate various maternal behaviors. For
example, Leon, Numan and Moltz (1973) conclude that maternal
behavior is normally under inhibitory gonadal control, as
demonstrated by increased levels of maternal behavior as long
as 8 weeks after gonadectomy, in both males and females.
The undoubtedly subtle effect of decreases in estrogen is
obscured by the failure to either differentiate the hormones
or behaviors. Similarly Thoman and Levine (1970) demonstrate
that a functioning adrenal is not necessary for maternal

behavior in rats. Of course, hormones, prolactin particularly,

 

 

39

contribute to the maintenance of lactation and thus have an
indirect role in the continuance of maternal behavior (Richards,
1967). ‘A study by Moltz (1970) found that ovariectomy dis-
turbs maternal behavior in primaparous rats but not multi-
parous, presumably experiential factors are superimposed on

the hormonal state here.

Much of the disagreement may be due to the different
roles hormonal preparation plays in the onset of different
maternal behaviors (Richards, 1967). While next-building is
due to the reversal of the progesterone/estrogen (Lisk,
Pretlow & Friedman, 1969; Lisk, 1971; Richards, 1967) re-
trieving may be due entirely to stimuli from the pups.

Epimiletic behavio- itself does not change the hormonal
constitution of the animal displaying it, at least in rodents.
When Terkel and Rosenblatt (1971) cross-transfused blood
from retrieving virgins to other virgins no effects of the
blood alone was observed on the retrieving behavior of the
recipients. The latencies to retrieve were found to be
directly proportional to the size of the cage and the close-
ness of contact with the pups! If an injection of maternal
blood plasma or cross—transfusion is done with mothers within
30 minutes after parturition an effect is found on lowering
the latency to retrieve (Rosenblatt, 1970; Terkel & Rosenblatt,
1968).

In summary it now appears clear that while some
"maternal" behaviors are hormonally primed at parturition,

particularly by the changing ratios of estrogen and progester-

40
one, only nursing and possibly, next-building require the
continued presence of hormones (List, 1971; List gt_al.,
1969; Rosenblatt, 1970; Terkel and Rosenblatt, 1971). Even
these behaviors and retrieving and cleaning require continuous

stimuli from the pups.

Stimuli Emanating from Infants:

 

The entire Gestalt of the neonatal rodent pup may play
a stimulus role in the elicitation and maintenance of the
care-giving behavior of adults. Specific elements of this
Gestalt are difficult to extract and study separately. Never-
theless, Noirot (1970) demonstrated that olfactory cues tend
to prime cleaning behavior while auditory cues prime nest-
building and retrieving. Many earlier studies suggest that
auditory cues are important guides in retrieving behavior and
may also serve to inhibit adult aggression (Hart & King, 1966;
Munn, 1950, Noirot, 1966; Zippelius & Schleidt, 1956). Sound
seems to stimulate retrieving; silent pups are often not re-
trieved (Gandelman, g£_al., 1970; Noirot, l964a; Sewell,

1970; Wiesner & Sheard, 1933). -

Since the stimulus value of the pup changes with age
(Noirot, l964a,b,c, 1966) and the mother's readiness to re-
trieve also changes (King, 1958), the separate elements of
the pup-stimulus must change. The odor of the pup may be
dependent on its food and the odor of its family and thus not
likely to change until weaning and dispersal. The pups are

constantly growing and growing hair, changing the visual

 

   

41
picture, but these changes have not been correlated with any

alteration of the behavior of the mother.

'However changes in the sounds and ultrasounds (above
the range of human hearing, about 20kHz) are readily measur-
able and appear to correlate with maternal readiness to re-
trieve. As silent pups are not retrieved, retrieving drops
off about the time of the development of adult temperature
regulation, which corresponds to the time of eye opening
(Beniest—Noirot, 1958; King, 1958; Noirot, 1972; Richards, 1967;

Zippelius and Schleidt, 1956).

Ecological Considerations:

Nicholson (1948) relates observations on movements of
young litters by wild female Peromyscus. Flooding of nests
or disturbance by predators may be included among factors
which would prompt a female to move her litter. It is also
probable that some nest shifting occurs spontaneously during
the nurturance of any one litter (Eisenberg, 1968; King, 1963).
When disturbed, feamle Peromyscus, Apodemus, and other mice
flee the nest with their pups attached to their nipples
(King, 1963; Sewell, 1968; Huff, pers. obs.). The pups cling
so strongly that the mother can be lifted by picking up a pup.
Tflna mother herself may have difficulty removing pups, although
'usually she disengages them before leaving the nest on routine
errands (King, 1963, 1968). The strong attachment of the pups
provides an efficient means of transporting several pups

Simultaneously in a hurry, but weaker pups do fall off and are

   

 

 

42
left behind, while the mother continues to a new refuge
(1968, Layne, King, 1963). Once in the new nest the fe-
male disengages the remaining pups and returns to retrieve
squeaking pups left behind. These observations are con-
sistent with the theory that the ultrasonic squeaks function

to keep the pups with their parent.

Localization:

 

Further confirmation of the importance of the sounds
for this behavior is found in Allin and Banks, (1972). In
this study male white rats, virgin females, and lactating
females were placed in the center of a hexagonal platform
and tested with recorded distress cries from a 7-day old
pup. Fifty percent of the lactating females made at least
one sortie out of the central nest box towards the peri-
pheral sound source. Males and virgin females did not
sortie but did orient to the sounds. These groups also re—
acted to the control noise. No appreciable differences
were found between the recorded sounds the rats heard and
the sounds originally recorded, so the authors concluded
that the sounds contained sufficient clues for accurate

orientation to the source without odor or sight of the pup.

Discrimination:

 

Adult Peromyscus show preferences for mates of their

 

own species on the basis of odor (Blair, 1953, 1954; Blair

& Howard, 1944; Bradshaw, 1965; McCarley, 1964; Moore, 1965).

 

w ‘=-—.f Y ~‘_.—r_ =9 ,; 77.1.- .

43
It should be possible for adults to discriminate pups by
odor. Adults will reject foreign pups even of their own
species and often of other species and this is commonly
assumed to be due to a "strange" odor. Whether sound plays
a part in discrimination has not been demonstrated except

as will be discussed in this paper.

Predators:

 

If the sound or the excursion to retrieve a lost pup
created a ready target for predators it would be hard to
see how the behaviors could be maintained. That other mam—
mals, including predators, can hear such sounds is known.
Peterson, Herton, and Wruble (1969) measured cochlear micro-
phonics in fissiped carnivores at the second tympanic mem-
brane. At this level those small carnivores which might
be mouse predators, such as coyotes, dogs, foxes, raccoons,
and cats, demonstrated sensitive hearing in the regions of
the frequencies of mouse squeaks, with upper limits at 60-
100 kHz. Conversely, Payne (1962) found that barn owls
appear to be quite deaf to sounds above 15-20 kHz. In-
stead, they are capable of accurately locating mice by the
sound of their movements in leaves. Predators could thus
either hear the squeaks of lost pups, or the sounds of
mothers moving to retrieve them and the situation must con-
tain substantial risk. The efficiency of the mother's
localization of the sound must be great enough for it to

be ecologically more efficient to retrieve, than to lose

 

 

44

the genetic and energetic investment on the pup.

Summary:

Data available in 1968, when this study was under-
taken, and data published since, agree that infant rodents
produce ultrasounds to cold-stress and handling. These
squeaks, when heard by mothers, trigger maternal retrieving
behavior. Information on the age of the pup and possible
species may be included in the sound. These sounds were
seen to have complex harmonics and frequency changes. The
utility of the sounds to a wild mouse mother is suggested.
The study to be described in the next chapters analyzes
distress calls of several species in one genus, compares
them in detail across age and species and shows their rele-

vance to the listening mouse.

 

CHAPTER 3

Species-Specific and Ontogenetic Characteristics

of the Peromyscus "Distress" Cry.

 

Introduction:

Noirot (1966b) found qualitative and quantitative
changes in ultrasounds emitted by albino mouse pups as
they matured. Hart and King (1966) found changes in
squeak frequency with age in Peromyscus maniculatus pups
and a decrease in numbers of sets of squeaks per 30 sec-
onds with age in the same mice. Frequency changes may be
associated with maturation of the vocal tract (Hart & King,
1966) or the opening of the ears to permit auditory feed-
back. Large changes in frequency and emission rate are
positively correlated with the opening of the ears (Noirot,
1966b). Decreases in squeak rate or frequency are cor-
related with the onset of thermoregulation and the develop-
ment of fur (Hart & King, 1966; Okon, 1970a).

The various behavior patterns defined as maternal by
Beniest—Noirot (1958), change in intensity and threshold of
action in mother hamsters, deermice, and house mice as

their litters age (King, 1963; Noirot, 1966a; Richards,

45

46

1966a,b; Rowell, 1960). This may be partly due to changes
in hormone levels (Noirot, 1969a; Terkel and Rosenblatt,
1968) or partly or wholly due to changes in the stimulus
characteristics of the pups (Noirot, 1964b; 1966a; Richards, :
1966a,b; Rosenblatt, 1967). Visual and auditory character— I
istics change more than olfactory characters during the
neonatal period but although gradual changes due to growth
and development of the body and pelage are visible to human
observers they do not correlate with the sharp decline in 1%
maternal responsiveness which occurs at about the time the ,:S
pups open their eyes. Mouse pups stop emitting ultrasonic
distress calls at this time, however, and the cessation of
squeaking correlates with the decline in maternal behavior
(Noirot, l964b; 1966a). Moreover, Noirot (1970) has demon—
strated that ultrasounds selectively prime and maintain re-
trieving and nest-building while other stimuli (olfactory)
prime other aspects of maternal behavior such as cleaning.

In the analysis reported here I concentrated on
identifying the quantifiable parameters of Peromyscus dis-
tress squeaks and describing the ontogenetical changes
through which they go. Certain aspects of these sounds
may carry particular sorts of information and may change
in different directions or at different rates. Changes in
frequency might be expected to correlate best with the
development of sound production and monotoring organs

while rate of emissions might correlate with thermo-

 

 

47

regulatory abilities. Characteristics, such as redun-
dancy, frequency and intensity modulation and sound en—
velope shape, which aid adults in locating squeaking pups,
are predicted to change only in the direction of increased ,
effectiveness until the pups no longer need to be retrieved.
The sounds produced by many animals are used for
species identification by the animals themselves and by
human observers. Species—constant sound characteristics
were therefore expected to occur in the Peromyscus squeaks fig“
both as incidental results of differential body shapes 3-
and as characters selectively divergent under pressures
of sympatry. The subspecies of Elmaniculatus should have
the most closely related frequencies if vocal tract muscles
change more slowly than behavior but these same mice should
have more divergent squeak rates, durations, and modulations
where they occur in sympatry. If the sounds could be
analyzed precisely enough it should even be possible to
suggest taxonomic relationships on the basis of differences
and similarities between the calls.
A refinement towards some species—typical pattern in
the sound would be expected if these sounds are used by the
mice for identification. Additionally, increased ability
to thermoregulate as the pups grow would reduce the motiva-
tion to squeak from cold stress and increasing mobility
would make it less important that the pup have its mother
come to retrieve it. Both these changes should be ex—

pressed in shorter squeaks, lower squeak rates and fewer

 

 

Fill F'T

 

 

48

squeaks emitted.
Procedure:

Subjects:

Systematic recordings were made of Peromyscus pups
from day one after birth to eye—opening. The day of eye-
opening was considered to be the average date for each
species as determined from colony records and preliminary
studies (Table 3—1). Pups with both eyes open did not
squeak and those with only one open rarely did, so that
most of the pups recorded did not have open eyes. Daily
recordings were made of six separate individuals from

four taxa, P.manicu1atus bairdi, P.m.gracilis, P.polio—

 

 

notus, and P.1eucopus. Recordings were also made of pups

from P.californicus, E.difficilis, P.melanophrys, and

 

P.m.nebracensis when they were available. Due, however,
to small numbers of these taxa in the colony and the in-
frequency of litters it was not possible to record as many
individuals at as many ages as the first group. Pups of
R.californicus were recorded on days 1, 2, 4, 6, 8, 10
(12 individuals), P.diffici1is on days 8, 10, 12, 14 (6
individuals), Efmelanophrys on days 2, and 4, and P.m.
nebracensis on day 4 (2 individuals). The females of these
Species often cannibalized an entire litter after one pup
had been recorded, thus reducing the supply of pups.

No more than four mice were used from the same lit-

ter and no mouse was used twice. If there were more than

 

 

 

49

 

loemH .maHm

v ao.o H H.mH as

 

maosmomubwc.m.m

 

 

 

 

 

 

.m.c m.o m.HH o.HH m mwmmmmmmmmm.m
.m.G H.H Nmo H m.mH N.o H v.5H m mHHHOHMMHU.I
.m.c h.H h.o H o.ma v.0 H N.vH h mSOHsHOwflHmM.M
Ho. H.H 4.0 H m.NH m.o H e.MH Hm mmmmmmmmmm.m
Hoo. w.o H.o H N.HH H.o M ¢.NH Hm msmoosma.m
mo. e.H m.o H m.mH m.o H e.HH mH mmmmmmmm.m.m
Hoo. N.H N.o H b.HH H.o H m.ma mm HUHHMQ.E.M
wasp x mm mwmp x mm wasp x
m mocwuwmmwc msfiswmo umolmmm mcflsmmo owwlomm s mmflommm

 

.HmHmm> .m cam HHsm .b an

aboum tmsmflabsmcs

tam mpuoown msoHoo Eoum span .hcoaoo mpmum cmmflsoflz map CH
msomNEoumm mo moflommm CH mcflcmmo Hmm paw mcflsmmo who mo moms

{Ill

Him OHQMB

 

 

 

50

two pups in a litter, a toe was clipped on each recorded
individual so that it could be recognized and avoided on
subsequent days. No more than two pups from any litter
were recorded on the same day, in order to avoid possible

family influences.
Care and Handling:

All of the mice were maintained in the mouse colony
room in clear, plastic cages (5 x 11 x 6 in deep) with re-
movable wire lids holding food and water bottles. Light-
ing was kept at a summer cycle, 12-14 hours light; 12-10
hours dark. Unless there was a history of litter cannibal—
ism males were left with the females and their litters at
all times. Pregnant females were checked twice daily for
litters, although for the purposes of this study ages were
counted by 24 hour periods.

At the time of recording, pups were removed from
the maternal nest in pairs, if possible, and carried into
the recording chamber in the experimenter's hands to re-
duce chilling. The pup to be recorded first was placed in
the apparatus while the other was held in a box under a

desk lamp a sufficient distance from the microphone.

Recording Procedure:

Before squeak recording the mouse pups were examined
for open eyelids and then placed on a wire square above a

cotton plug in the bottom of a large metal funnel (Figure

 

- “,

, ”A; 5.

 

 

 

51
3-1). This apparatus restricted the movements of the more
active pups and kept them within an average range of the
microphone. The close contact of the metal walls may
have also stimulated them to squeak more (Noirot, 1970).
The microphone was then lowered to within three inches of
the mouse, the experimenter left the room, and the sound
monitoring system was turned on.

If the mouse did not squeak in five minutes it was
returned to the nest and another pup was used. When the
pup appeared to be squeaking consistently the tape recorder
was turned on and sounds were recorded for at least one
minute. If the mouse stopped squeaking for more than 15
seconds the tape was stopped until the mouse resumed
squeaking. The recorded sound was constantly monitored
with an oscilloscope and headphones.

After recording the squeaks the mouse was removed
from the funnel and checked for raised pinnae and a startle

response to a loud click, then returned to its nest.
Recording Equipment:

Sounds were picked up from the squeaking pups with
a 1/2 inch Bruél and Kjaer condenser microphone, amplified
by a Bruel and Kjaer microphone power supply and Tektronix
amplifier (Appendix A). These sounds were recorded with an
Ampex four-channel instrumentation recorder on Ampex in-
strumentation tape (1 mil mylar) running at 15 ips. A

playback head on the recorder allowed the sounds to be

_ . .= u-4--E-d—W__—_i..

 

52

Figure 3—1

Legend

Mouse restraining chamber for squeak recordings.
The mouse pup (m) was placed at the bottom of the funnel

(F) and the microphone (M) was lowered to within 3 inches

of the pup. A ring stand (R) and clamps and a large can

(C) supported the microphone and funnel. (MC) is the

microphone cable.

 

53

FIGURE 3-1

 

 

 

 

 

Mouse Recording Chamber

 

 

54

simultaneously displayed on a Tektronix oscilloscope. The
tape recorder was the limiting factor in this system, in
terms of frequency response, as it was rated as giving a
flat response to sounds only from 4 to 40 kHz at 15 ips.
The microphone gave a flat response to 50 kHz and accept-
able sound recording was obtained to 60 kHz. Sounds above
those frequencies were observed in the recordings, but rela-
tive sound energy levels could not be compared. (For more
detailed equipment specifications see Appendix Al).

Although the recording chamber was sound-damped and
all the available 60-cyc1e filters were used, on some days
loose connections or high background noise made the record—
ings too noisy to be analysed. When a recording of a par-
ticular age was critical, more than six mice were recorded

at that age to ensure a sample of clean records.
Sound Analysis Methods:

All of the one—minute samples of squeaks for each
age of each species were filmed at normal tape speed (15
ips) using a Tektronix oscilloscope and a Grass Oscillo-
scope camera. All of them were filmed with a time base of
1 cm per second and some of the clearest were also filmed
at 2.5, 5, and 10 cm.per second. The resulting oscillo-
grams showed the temporal relationships of the squeaks and
permitted measurements to be made of the duration (in
seconds) of individual squeaks, groups of squeaks (sets),

and inter—squeak intervals, and inter-set intervals, as

 

 

 

55
well as permitting the numbers of squeaks per set to be
counted.

Most of these clear recordings were also analysed
with a Bruel and Kjaer Sound Frequency Analyser which
graphed a spectrum of sound energy, from a single squeak,
in terms of sound pressure level (db) present in particular
frequency bands. In order to register, the tapes had to be
slowed to 1-7/8 ips. These spectrograms showed relative
sound energy and harmonics and frequency modulations more
precisely than the other methods of analysis. Because the
mice were free to move somewhat in relation to the micro—
phone and because they changed their sound volume at will
it was impossible to calculate absolute sound levels from
the spectrograms. However, relative sound pressure levels
could be compared between parts of any one squeak. The
individual excursions of the pen drawing the spectrograms
could also be compared to the sound heard from the squeak
being analysed and thus frequency modulations within the
squeak could be recognized.

Those squeaks spectrographed were also examined
with a Voiceprint machine which made sonagrams of the
sounds. Again the tapes were slowed to 1-7/8 ips. Sona-
grams are a standard means of visually presenting bird
songs and human speech but, although they are used in-
creasingly for non-human mammal sounds, the quantifiable
data they provide from these simpler calls is meager.

Pictorially a sonagram furnished an attractive comparison

56

between sounds and for this reason a number were done in
this study. Sonagrams show frequency over time, and in-
tensity can be estimated from relative darkness of the burn.
Since the sonagrams depict a mouse squeak spread out over
time they were compared to the sound spectrograms which
showed precise frequencies and the oscillograms which showed
time and intensity. I

Parametric or non-parametric analyses of variance were
done on the temporal data, in detail for those taxa with samples
from more than 12 individuals (Group A, P.m. bairdi, P.m,

gracilis, P. polionotus, P. leucopus)and more superficially

 

with the smaller samples from other taxa (Group B, P.

californicus, P.m, nebracensis, P. difficilis, P. melanophrys).

 

 

 

 

Sound frequencies were calculated from the sound
spectrograms and compared with the sonagrams. Mean basic
frequencies*, fundamental frequencies**, and means for major
harmonics were calculated where possible for ages and taxa.
When fundamental frequencies were present the other frequencies
were compared to the fundamental to determine where they fit
into the overtone series. If the fundamental was absent an
attempt was made to estimate it.

Differences in intensity modulation, frequency modula-

tion, and changes in pitch were visible in the sound sprectro-

*The most important (largest amplitude) frequency com-
ponent.

**The frequency of repetition of the complex wave (Ladefoged,
1962).

 

   

 

— _—V* 7 1r‘fim —,

57
grams and like the temporal changes visible in the oscillo-
grams could be objectively compared from the pictures.
These differences are displayed in the plates included in
the results section. The sections of films shown in these
plates were chosen for typical patterns as well as clarity

and low signal—to-noise ratios (Figures 3-2 to 3-14).
Results:

Species Differences - General Observations:

The null hypothesis was that squeaks of different
species or subspecies would not be significantly different.
Data analyses disproved this and showed differences on all
measured parameters. Squeaks of 8-day old pups were ob-
served to be clear and relatively invariant within a taxon
when they were compared across all the taxa, characteristic
patterns became apparent (Figures 3—2 to 3-8). Oscillograms
and sonagrams showed that P,m.bairdi, P.1eucopus and P.3g1i;

fornicus start each squeak with a rapid frequency and in-

 

tensity rise, modulate both and then suddenly drop off

(Figures 3-2,3). Other taxa, P,m.gracilis, P,m.nebracensis,

§.pglionotus, P.diffici1is, show an initially sharp frequen-

 

 

cy rise followed by a more gradual one to a peak and then a

decrease (Figure 3-2). P,m.nebracensis, §.m.gracilis, and

 

B,Qifficilis particularly appear to put the greatest empha-
sis on the middle squeak in a set (Figures 3-2,3).
While intensity modulations are most visible on the

oscillograms, frequency variations are clearer on the

 

 

 

58

Figure 3—2

Legend

Oscillograms of distress cries from 8—day old

 

Peromyscus pups. (a) P.maniculatus bairdi; (b) g.m.grg:
gilis; (c) P.m.nebracensis; (d) P.1eucopus; (e) P.polio-
ngtus; (f) P.californicus; (g) P.difficilis. Time marks
are 1 cm equal to 1 second. Similar intensity modulation)
ggpus and P.californicus with intensity decreasing in the
center of the squeak. P.m.nebracensis, P.m.gracilis, P.‘
polionotus, and P.difficilis exhibit greatest intensity in
the center of the squeak and often also in the central

squeak of a set.

 

59
a. Rm.boirdi L—J—‘J—L—J

lcm,= lsec.
“*va 4H. ; ai'r

 

 

b. Rm. gracilis

1% 11414 41

c. Rmnobraconsls

 

 

 

 

HH” W4f—w——Mo——+fl+

L_L__L-__L_J L_l_L_L__L
d. E leucopus e. P. po Iionotus

 

 

"1 Mg: ' I. '

f. P. collfornlcus

 

g. P. dlfflcills

h. P, melanophrys FIGURE 3-2

 

 

 

 

 

Figure 3—3

Legend

Oscillograms compared with sonagrams. P.1eucopus:
(a) an oscilloscope tracing of squeaks from 8—day old pup;
time on the abcissa, 5 cm = 1 second, intensity on the
ordinate; (b) sonagram of a single squeak from an 8—day
old pup; time on the abcissa, mark = 0.1 second, frequency
on ordinate, wide-band, log plot. B.manicu1atus gracilis:
(a) oscillogram of squeaks from an 8—day old pup, 5 cm =
1 second; (b) sonagram of a single squeak from an 8-day old

pup, time mark = 0.1 second. The typical strong overtone

is visible on the P.m.gracilis sonagram.

61

 

 

 

Moro

intent“;

L 1 1 1 l J IOOC.

PIGUCOPUS a. oscilloscope film pf squeaks

40-

 

 

 

 

i
E '6.
kc.8__
; I .luc.
b. sonogram of asinqlo squeak
=3 ————‘—*
a. __
gr 1 1 n I IOOC.
P. maniculatus gracflis - a. oscilloscope film b.0onoqram
48-
.0- ’M
32— \ \‘\\ ‘
24-
§ I6-
° 8.
b KC I-“ l l

J

.l 80¢.

FIGURE 33

32-
24_ M

62

sonagrams and spectrograms. P,1eucopus and P.polionotus

 

rarely show harmonics (Figures 3-4, 5, 6); most of the
sound energy in their squeaks being concentrated in a

single frequency. The P.maniculatus subspecies (Figures

 

3-3, 4) and P.californicus (Figure 3-6), on the other hand,

 

usually show at least one harmonic band. In the case of

P,m,graci1is and P.ca1ifornicus this harmonic is often as

 

strong or stronger than the lower frequency (Figures 3-5,7).
Sound spectrograms (Figure 3-5) reinforce the inter—
pretation of single clear tones from P.1eucopus and §.polio-

notus and multiple harmonics from the Pgmaniculatus mice

 

and P.ca1ifornicus. In addition, careful study of these

 

records enabled me to differentiate between those mice

(P.maniculatus) which show a slight rise in frequency at

 

the beginning of the squeak and those (g.polionotus and P.

 

leucopus) with sharper onsets and offsets. Although I was
unable to observe different methods of vocalizing it may
be that whether the mouse starts squeaking with its mouth
open or closed determines the pattern. A sharp, pure tone
would be expected to emanate from an open mouth, while a
rising frequency might be produced by opening the mouth
after the squeak had already been started.

Some of the squeaks show breaks in the solid pattern
of sound (Figure 3-6d). Behavioral observations were un-
able to clarify the relationship of these breaks to the
Sflqueaks but as they also appear in the squeaks of mice re-

corded elsewhere on other equipment, it would seem that

 

 

 

63

Figure 3-4

Legend

Sonagrams of single squeaks from recordings of
8-day old pups. Time mark = 0.1 second. (a) P.1eucopus;
(b) P.polionotus; (c) P.maniculatus bairdi; (d) P.m.
gracilis. A strong harmonic is visible in the P.m.bairdi
sonagram as well as in the P.m.gragili§ which also shows
a secondary frequency. Compare with oscillograms in

Figure 3-5.

 

G4

 

 

 

 

 

 

ll
40 _
’32-
24.. F I -~
l6-
KHZB—
.1: .Isec. P leucopus 8 day
ll 48-
40..
32 -
24-
KHZ. '6';
L I .lsec. P polionotus Bdoy
c 48- W . ~- 4M 4 -ms
40- i 1.. .
'l.
32‘
24- M
16- ~
kHz.
0 I 8 .
L 4 Isec. F.’ mboirdi 8 day
48-
40—
52—:

t; n 1 1N4“. IIIHI—
M.
16-

8..
L___ l l i 7 ~ —4 | 34‘4“ quracahs 6 day

FIGURE 3'4

 

 

 

 

Figure 3-5

Legend

Comparison of oscillograms and sound spectrum graphs.
Time marked on the oscilloscope tracings. Sound spectro-
grams show a plot of sound pressure level (db) at various
frequency bands across the abcissa. Each oscilloscope
trace displays a single set of squeaks; each spectrum shows
a single squeak from the same or a different individual:
(a) and (b) P.1eucopus, two, 8-day old females; (c) and (d)
P.polionotus, one, 8—day old female; (e) and (f) g.m.bairdi,
two, 8-day old males; (g) and (h) P.m.gracilis, one, 8-day
old female. In all cases the oscillograms and the spectre-
grams (d) and (f) are from the same individuals as in Figure
3—4. Visual comparisons of the sonagrams and spectrograms
reinforce the interpretation of single clear frequencies
for P.1eucopus and P.polionotus and multiple harmonics pres-
ent for §.m.bairdi and P.m.graci1is. Apparent double images
in the spectrograms are caused by increasing or decreasing

frequencies over time.

 

P. leucopus

P. polionotus

66

Intensity

 

 

O
_—_ M
.__.v. w
—

P. maniculatus balrdl

— .2l0t

Emanlwlctw grogllll

 

Wu“

IIIIIIIIII I In III”

I6 40

FIGURE 3- 5

 

 

 

 

 

67

Figure 3—6

Legend

Sonagrams of single distress squeaks. Time mark =
0.1 second. (a) 8-day old P.californicus, unusually short
and frequency modulated; (b) l-day old P.polionotus (c) 8-
day old P.1eucopus, second squeak in a set, similar to
Figure 3—4a; (d) 7—day old P.1eucopus, showing a buzz-like

break.

 

 

68
ll 40_
A
32- ' \‘
I
24— ‘
.6_ . I‘ i
In
8- I
kHz. .
n. I I III
_ 1 sec,
P. caliormcus Bduy ‘— 1
32-
2, ~ 4
2:
I6—
8-
kHz. ‘
1 J .l sec.
P. polionotus I day
0
40'-
32-

 

 

l6—
8—
kHz. ‘
l J ,I sec.
P. leucopus 8 day
40-
32..
“- fl n
l6-
3.
W
L "a” Amalie... h: I ‘l ul-
I¥ I .I sec.
P. leucopus 7 day

FIGURE 3‘6

 

69

Figure 3-7

Legend

 

Sonagrams of single squeaks. Time mark = 0.1

second: (a) 8—day old P.ca1ifornicus; (b) 4-day old 3.

 

melanophrys; (c) 8-day cld P,m.nebracensis; (d) 8-day old

 

 

B.difficili§. Compare with Figure 3-8.

 

 

 

.oow _.

 

I

 

 

 

 

 

do... . I

352:1: l _ l ,.....s 2 55-: z p... . :_ : 2. “11-..:

um

I - re.

. . -vu

nun

III a :09
.3.
no
no.
IVN
nwn
nos

.3 .v .9.

dourI _ 6;.

In

5‘..l’.‘.ll"‘l

sew

. .... . ...II..II-....I>

/ -o¢

FIGURE 3- 7

0'

 

71
they are not experimental artifacts. Such breaks are most
common in the squeaks of species showing the greatest fre—
quency and intensity modulations. They also appear to
contain more harmonics than the rest of the squeak and to
occur at points of great intensity. It may be that these

breaks are tongue clicks.

Species Differences — Frequencies:

The hypothesis that the sounds of these taxa would

 

differ in terms of frequency was confirmed by the observa— fly
tions of different modulation patterns (envelope shape)

described above. Preliminary analyses also suggested that

more precise frequency measures would show significant

differences. Mean basic frequencies, i.g. those frequencies

with the greatest sound pressure levels in the squeak, were

computed for each taxon and each age for Group A (P.manicu-

latus bairdi, P.m.gracilis, P.1eucopus, P.polionotus).

 

Within this group, mean basic frequencies differed signifi-
cantly at P < 0.001 (Kruskal-Wallis one—way analysis of
variance). Overall mean basic frequencies for Group B (B.

californicus, P.difficilis, P.me1anophrys, P.m.nebracencis)

 

also differed significantly, 0.05 > P > 0.0.1. (Kruskal-
Wallis one—way, analysis of variance). Nevertheless the
mean basic frequency, averaged over all ages, did not appear
to be representative of the characteristics of the sounds.
On closer examination two subtleties in frequency charac-

teristics become apparent. A tendency for some species to

 

72
produce a squeak with several strong frequencies will be
discussed below and abrupt changes in squeak patterns at

critical ages will be discussed under ontogeny.

Frequency Patterns:

 

Some species (E.m.gracilis, P.diffici1is, B.califor—

 

nicus in particular) produce maximal energy at two frequen—
cies; in effect two "basic" frequencies. These frequencies
were always harmonically related to each other but appear

to be in a separate scale from other visible frequencies.

 

The higher of the two "basic" frequencies, usually 34-38
kHz, was always double the lower, usually 18-19 kHz. I
labeled those squeaks showing one strong frequency, Type I
squeaks, and those showing two strong frequencies, Type II.
When the two types of squeak were separated, the daily mean
basic frequencies for E'E'EEEEEE (22.6 kHz), P.1eucopus
(27.4 kHz), and E.polionotus (33.1 kHZ) were significantly
different from each other (Kruskal-Wallis one—way analysis
of variance: P < 0.001). When the two types were considered
separately sample sizes for the other taxa were too small
for statistical analysis but means for P.m.gracilis (Type

I X = 21.9 kHz; Type II x = 16.4 kHz and x = 33.7 kHz)
appeared to be slightly different from all the others. P.m.
nebracensis (x — 23.2 kHz and P.me1anophrys (x - 23.0 kHz)
were not significantly different from P.m.bairdi with their
Type I squeaks, and P.difficilis and B.galifornicus were

too few to determine differences.

 

73

Figure 3-8

Legend

Sound spectrograms of single squeaks compared with
oscillograms of sets of squeaks. Time, as marked, on
oscillograms. Spectrograms plot frequencies on the ab-
cissa, db on the ordinate. (a), (b), and (c) 8-day old

3.0a1ifornicus, compare with Figure 3-7 a; (d) 4-day old

 

P,melangphrys, the same individual as in Figure 3-7 b;

 

(e) and (f) 8-day old P.difficilis, the same individual as

 

in Figure 3-7 c; (g) and (h) 8—day old P.m. nebracensis,

 

(h) is from the same individual as Figure 3-7 d.

 

74

 

 

 

 

a
db
A II
lnquoncy I0 I6 2° 32 kHz
I
I

c .
‘ “M A
= “”7 VI JIM] "Vt‘r'w—V VL-‘va'n‘r': H—
E . ‘
2
E _ .‘I so:

. -—-.4

"me P. californicus

d
Panelunophr'ys
I6 20 4O 6. k"!
‘ f
.. - »,
————CI.—.———-—-4 '
A" II“ I

—— .2 10¢

 

P. dlfflcllls

23.2
a I' I 4
VIII. 1.
f—Jn.Hnr—'.:fio—«gp— "III I I"
II I I I5 2‘ 60A 3. k"!

 

P maniculatus nebracensis
FIGURE 3-8

 

 

 

 

75

Type II squeaks Show much frequency and intensity
modulation (Figure 3-9) although data were insufficient to
identify particular modulation patterns with particular
taxa or to determine the relationship between the modula-
tions and the several strong frequencies. Figures 3-3, 4,
6, 7, 9 show the 2 strong frequency bands typical of Type
II squeaks in comparison to the single bands of the Type I
squeaks.

Sound spectra also clearly demonstrate the differ-
ences between the two types of squeak (Figures 3—5, 8).
Sound pressure levels of more than a minimal 10 db above
the baseline occur in discrete frequency bands. These fre—
quencies were averaged within age periods for all the taxa
and are presented in Table 3—2. Bands of similar frequen-
cies resembling formants, are utilized by all the taxa, but
different species show concentrations of sound pressure in
different bands. Thus P.m.bairdi show a frequency band
around 34 kHz with the strongest frequency typically 23 kHz,
whereas E.polionotus sometimes shows a band at 25 kHz but
emphasizes frequencies at 34 kHz. Mean frequencies for
measurable bands are shown in Table 3—2, with the mean basic
frequency (the strongest) underlined. Where no sound rose
above the baseline a dash is used. The multiplicity of
the frequency bands of P.m.bai£di and E.m.gracili§ is clear-

ly demonstrated as is the contrasting single strong frequen-

cy of E.polionotus and P.1eucopus.

 

 

 

 

76

Figure 3—9

Legend

Standard sonagrams of single distress squeaks com-
pared with intensity contours of the same squeaks. (a)
5—day old P.m.bairdi; (b) contour of a; (c) 6-day old P.m.
bairdi; (d) contour of c. The contours show a quavering

quality in the intensity modulation and fluctuations of

intensity greatest at the beginning and end of the squeak.

 

24
I6
8.
‘l
40— II .
.h 1,
32. 34,} 5,7“
24- II‘II)
5. I
“Mi-p Lu
43.
C.
40-
32—

2.))

77

| . ,7
) .2 1} . ~ 'I I
' ,. u .- - , 0. I'
(I; \ ,I ,1
m I/
'3)‘, \ ) } )1 I» ,I/ )D ‘ ‘ ‘
\;r 3),?” ,\
I1) ‘ 'l I {

FIGURE 3‘9

 

 

 

 

h—I.l.m~h u.u__-.o:.w‘~.i;.fiwzt. ‘ E

78

Table 3-2

 

Distress squeak frequencies from 8 taxa of Peromyscus. Fre-
quencies averaged for early (days 1 & 2), middle (days 3—10),
and late (days 10 on) squeaks. Frequencies of greatest amp—
litude are shown; basic frequencies underlined. n equals
numbers of squeaks analysed. _

 

 

Band Number

 

 

 

 

 

 

 

 

 

 

 

Day n 1 2 3 4 5 6 7 8

23polionotus

Early 4 9.0 18.6(1) 25.7 36.8(1) 52.0 -

Middle 3 - - 17.5 — 34.4 — — -

Late 2 — - — - 33.8 — - 66.0

Bfleucopus

Early 2 — - 19.5 - 38.0 - 53.2 -

Middle 9 — 13.6 - 28.1 - - 51.1 -

Late 2 ~ - - - 30.0 - - 60.0

P.m.bairdi

Early 6 6.0 — 17.3 21.3 32.8 43.7 53.0 65.2

Middle 15 6.4 11.0 - 2§;6_ 33.1 45.9 53.3 63.2

Late 4 - 11.0 16.8 23;Z_ 34.2 45.1 52.0 66.6

3.111.. gracilis

Early 2 5.7 1};§_ 17.4 24;§_ 34.2 - 57.6 68.8

Middle 2 8.0 12.0 - 22.8 36.0 46.0 - 68.0
3 7.9 12.8 lfiLII 21.7 38.8 44.0 - 61.0

Late 1 — 10.0 - 20.0 — 40.0 ~ 60.0

E.I_n.nebrace_1}:;i_s_

Middle 2 - 14.0 - 23.2 - 49.2 — -

E.me1anophrys

Middle 2 — II 0 - 23_g — 40 8 - 66 8

3.9Liificil is

Middle 1 - 10.2 — 20.8 — 41.6 - 62.4
2 — 10.0 19 2 - 38.6 — 55 4 —

chalifornicus

Middle 1 9.2 - 18.2 28.0 34.0 48.0 56.0 —
2 9 4 — 18.5 26.8 36.0 ~ 56.0 —

_..-..— “—

 

-. _. _.._...

 

 

 

79

The remaining 4 taxa provided too few data for strong
comparisons but the measurements reinforced the observations
made on Group A. P.californicus and P.difficilis show Type
II squeaks, with frequency and intensity modulation (Figures
3-7,8). Squeaks of P.californicus are very long and often
low in frequency (9.2 and 18.2 kHz; 18.5 and 36.0 kHz). 3.
difficilis shows frequencies similar to E.californicus but
P.m.nebracensis and P.melanophrys are more similar to the

P.m.bairdi and P.m.gracilis.

 

Species Differences - Temporal Parameters:

 

Numbers of squeaks per set, duration of individual
squeaks, and duration of sets were called temporal para—
meters since they could be enumerated or measured in units
of time. To answer the null hypothesis of no difference,
in these parameters between taxa, two—way analyses of vari-
ance were done on daily means for each taxon in Group A,

from day 1 until day 12.

Number of Squeaks per Set:
Daily means for Group A (P.m.bairdi, P.m.gracilis, E.

leucopus, P.polionotus) are significantly different at P <

 

0.005 (parametric two-way analysis of variance). The dif-
ference between the means for P.m.gracilis and P.1eucopus
is significant at P < 0.01, and P.m.bairdi differ in mean
number of squeaks from P.polionotus and P.1eucopus at P <
0.05 (New Multiple Range test). B.m.gracilis do not differ

significantly from P.m.bairdi or P.polionotus (Table 3-3).

80

.mpmc ommsu so once Hos mOGMHHm> no mHmmHmcsa

 

 

 

 

 

 

 

40 H va.o Hm.o H Hm.o mo.o H mm.w m mflmsoomubmcamam
Ho.o H ma.o mo.o H em.a Hm.o H om.v v wwnschmaosnm
Ho.o H mH.o mo.o H Hm.o mH.o H me.m e HHHHUHHHHe.m
Ho.o H mH.o 5H.o H mH.H m~.o H ma.e NH msoHcHoHHHmo.m

mummmmm

Ho.o H HH.o mo.o H mm.o Hm.o H os.m NH msHOHOHHom.m
Ho.o H 5H.o mo.o H Hm.o mH.o H mH.m me mmmmmmmm.m
Ho.o H mH.o NH.o H mo.H mH.o H mo.e me mmammmmm~m.m
Ho.o H mH.o HH.o H NH.H eH.o H oe.m we mmmmmmamam
mnmmmmm

mm x mm H mm x

GOHHMHSUIxmoSqm cowumnspluom Hem Hem mmmmsqm c mmHommw

 

.mumumamumm m Ham so Hobuo some Eonm usmHmMMHp
.m>m© NH How mamma wafimp so comma mum Ad msonov mxwu
.msomeonmm ucmmcfl mo Hamo mmonumfic on» mo moHumHHobomucso HmuomEmB

MHHGmOHMHGmHm mum mxmw e emcee
e umuflm now made:

 

 

mlm magma

81
The mean number of squeaks in a set were not found to

differ between the taxa in Group B (P.californicus,IP,dif-

 

ficilis, E.melanophrys, P,m,nebracensis) with a non-para-

 

 

metric analysis (Kruskal-Wallis), although larger sample

sizes might have demonstrated a difference.

Set Duration:

 

The taxa in Group A differ in mean set duration at
0.05 > P > 0.025 (two-way analysis of variance). E,m,bairdi

and P,m.gracilis differ significantly from P.1eucopus and

P,polionotus, but not from each other. P.1eucopus and g.

 

poliontus do not differ significantly. .Group B taxa differ

 

at 0.01 > P > 0.011 (Kruskal-Wallis). P.californicus, and

 

P.m.nebracensis have longer sets of squeaks than P.diffi-

 

cilis and P.melanophrys.

 

For the taxa in Group A the numbers of squeaks in a
set is significantly correlated with the duration of the set
(Spearman—Rank Correlation Coefficient; P.m.bairdi, P.m.

gracilis, P.polionotus,P < 0.01; 3.1eucopus, P < 0.05). A1-

 

though the inter-squeak intervals were not measured, system-
atic sample measurements indicated that they are fairly in-
variable and average about 0.1 sec. for all taxa. When
species means were considered over all 8 taxa a significant
correlation was found between the number per set and the
duration of the set, also (Spearman—Rank Correlation Co-
efficient; 0.05 > P > 0.01) but there is no correlation be-
tween the number of squeaks per set and the duration of the

squeaks.

 

 

 

82
Squeak Duration:

The 4 taxa in Group A differ significantly in mean
squeak duration at P < 0.005 (two-way analysis of variance).
By the New Multiple Range test P.m.bairdi and E.polionotus
were shown to differ at P < 0.01; P.1eucopus and E.polio-
32223 differ at P < 0.05. P.m.bairdi and P.m.gracilis do
not differ from each other nor from P.1eucopus, and P.m.
gracilis does not differ from P.polionotus.

The species in Group B do not differ significantly

 

in terms of mean squeak duration (P.m.nebracensis had too
few means to include in the analysis).

Although these 8 taxa do not differ significantly
on every parameter they all differ from the others on at
least one measure. For example, in Group A, P.m.bairdi
and P.polionotus have similar numbers of squeaks per set
but significantly different squeak durations and frequen-
cies. The squeaks of P.1eucopus are of about the same
duration as those of P.m.bairdi but there are fewer P.
leucopus squeaks in a set and P.1eucopus sets are shorter
in duration than those of P.m.bairdi. P.m.gracilis has
the greatest mean number of squeaks per set of the taxa in
Group A, but they are of short duration. Comparing these
with the 4 taxa in Group B, P.melanophrys has the largest
mean number of squeaks per set and the longest set duration
but P.californicus and P.m.bairdi have the longest squeaks
(Table 3-3). P.polionotus has the shortest squeaks, P.

leucopus has the fewest squeaks per set, and P.1eucopus and

 

 

—._~:-:_a..;_-__._‘p --_.-a ._. J.“

83

 

 

P.difficilis have the shortest sets, while §.polionotus has

the highest mean basic frequency Tables 3-2, 3).

Age Differences - General Observations:

It was hypothesized that as a mouse pup develops it
gains increased control over its vocal apparatus. As it
begins to receive some feedback from the sounds it produces
it becomes better able to modify the sound. Ontogenetical
changes were thus expected to be apparent in the squeaks.
Observations and statistical analyses of the data confirm
this hypothesis. General patterns of change were found to
be similar across all the taxa analyses, with major changes
in frequency and temporal parameters between day 2 and days
3 or 4 and again at about day 10 (Table 3-2, 4; Figures
3-10 to 3-17). The most stable period for all parameters,
where they change most slowly and show least variation was

found to be the middle period between 3-4 and 10 days of

age.

Age Differences - Frequencies:

Squeaks recorded on day 1 and most of those from day
2 showed wide variation in frequency; all of these frequen-
cies were usually lower than the mean frequencies character-
istic of the stable middle period (Table 3-4; Figures 3-12,
13, 14). Some squeaks recorded on day 2 and all the squeaks
recorded on day 3 or later show little variation around
this "species-characteristic" frequency. Records were too

few in number and too far apart (24 hours) to determine

 

84

Figure 3-10

Legend

 

Sonagrams of single squeaks. (a) 1-day old P.m.
bairdi, time mark = 0.05 seconds; (b) to (e) time marks
= 0.1 second, (b) l3-day old P.m.bairdi, distress cry;
(c) l3—day old P.m.bairdi, pain squeak displaying mul-
tiple harmonics and great frequency modulation in a much
shorter squeak than the distress cry; (d) 8-day old P.m.
bairdi, distress cry; (e) l3—day old P.m.gracilis, dis-

tress cry.

 

 

.OSsec- 24‘
I 4————4 —’ ..

 

 

 

 

lday Pmbairdi IS-
a_ I
-I.u u .u n I . . I I nut-‘1‘ II
. 1.Iuc. 6*
”I I3 day Emboirdi
48-
40—
32—
24_ // ‘.. / ‘
I
I6- I A
6-
c I3doy Pmboirdi
. M- Inc.
48—
40—
32-
24- I
IS-
8'
‘ 4 Inc. 8601 P.m.bdrdI
0
.0- w‘u
I
32— "d I a.”
24-—
“_I
16—
M: o: I" - N
8..
AH:
L. A,_ #4 I uc. 8m.qrocIII| l3 day

FIGURE 340

 

 

 

 

 

 

86

Figure 3—11

Legend

Oscillograms and sound spectrograms for P.manicu-
1atus bairdi pups of different ages. Compare with sona-
grams of same species in Figure 3-10: (a), (b), and (c)
1-day old P.m.bairdi, the same individual as in Figure
3-10 a. The squeak (b) is the first of the set shown in
(a), while (c) is the second squeak; (d) and (e) squeaks
from a 13-day old pup, compare with sonagrams from the
same individual in Figure 3-10 b; (f), (g), and (h) pain
squeaks from a 13—day old pup, the same individual as in
Figure 3-10 c. Although the bursts in (f) appear to be of
almost normal length they are actually 2 or 3 squeaks run
together, as shown in Figure 3-10 0. Squeaks displayed in
(g), (h) show energy present at a wide band of frequencies
in these squeaks. (i) and (j) squeaks from an 8-day old

pup, (i) is from the same individual as Figure 3-10 d.

 

I

37 I
' I

I

a
Innmlty \ I I
P.munkulawn IMIVJI
— .l IIG. "m I
I .
20.1
I

     

   
  

.
....—~--u...-..-.. .
~. ...., . .I ..

III.

IIIIIIMIIIIII‘ ' ‘ ‘ ..

Is 20 32 50.4 “I

— .1 l0:

— .IIOC

 

— .I u:

 

FIGURE 3411

 

88

Table 3-4

 

Distress squeak frequencies (daily means in kHz) from Pero-
myscus maniculatus bairdi pups of selected ages.
quencies of the greatest amplitudes, equivalent to possible

The fre-

 

 

 

 

formants, are shown; basic frequencies underlined. p equals
numbers of squeaks.

Band Numbers ,
Day n 1 2 3 4 5 6 7 8
Early
1 2 6.0 - - 20.8 — 40.8 - -
2 4 - - 18.2 - 35.8 48.0 53.0 65.2
Middle
3 3 - - - 26.7 - - 54.2 -
6 3 - 11.2 16.8 26.0 - 46.0 50.4 62.8
7 4 7.4 11.3 - 23.1 - 40.0 5044 63.0
8 2 10.4 - 22.0 31.2 47.2 52.0 63.2
10 3 11.1 — 22.2 34.0 44.6 56.0 64.0
Late
12 2 - 10.7 16.8 24.0 34.4 45.2 52.8 -
l3 1 - 11.2 — 22.0 - 44.0 - 66.6
14 1 - 11.2 - 24.8 36.0 46.0 50.4 -

 

 

89
whether the change in frequency is gradual or as sudden as
it appears to be. Typically the shift in frequency was
recorded on the day corresponding to the mean date of pinnae
erection (Layne, 1968). One record of an unusual high fre-

quency for a 2-day P.polionotus pup, came from an animal

 

not observed to have its pinnae up, although the ears
might have been unsealed from the head or the nerves newly
capable of receiving sound through the skull.

From day 3 on variations observed were slight
(Tables 3-2, 4). Frequencies appeared to remain at a level
characteristic of each species (Table 3~4, Figures 3-5, 7,
8). Towards day 10 most squeaks lack visible fundamentals
and become increasingly pure tones (Tables 3-2, 4; Figures
3-10 to 3-14).

In order to compare data from P.m.bairdi and the
other taxa, daily means for major frequencies were grouped
into early - days 1 and 2, before the average date of pin-
nae erection; middle - days 3-10, after pinnae were up but
before eyes began to open; and lapg - after day 10, when
eyes and external ears were increasingly ready to open,
homothermy developing, and squeaks less frequent (Table
3-2, 4). Such groupings made frequency comparisons much
more realistic than single average frequencies for each
taxon, although total sample sizes were reduced, often be-
low the levels required for accurate statistical comparisons.

Daily means for frequencies for P.m.bairdi are shown

in Table 3—4. The mean basic frequency is lowest for day 2,

90

three of the four squeaks sampled on that day having low
frequencies not seen on other days. The mean basic fre-
quency is highest on days 3 and 4 (Figures 3-11, 12).

Only on days 1, 7, 8, and 10 are possible fundamental fre-
quencies present. Squeaks recorded on days 2, 6, and 12
show an intermediate frequency, similar to those visible

in many records in P.m.gracilis, §.difficilis and B.cali—

 

fornicus (Figures 3-5, 8, 10, 12).

Results of the multiple sonagrams and spectrograms
made of P.m.bairdi squeaks for different ages demonstrate
changes in temporal parameters and frequencies and provide
a basis for comparison with the data from the other taxa,
particularly those in group A. Visible frequencies of a
1—day old P.m.bairdi show high, pure tones (about 20 kHz)
with sharp onsets and offsets (Figure 3-10a). By the mid-
dle period, squeaks of an 8—day old pup display more ener-
gy at a second harmonic (Figure 3—10b). Spectrograms of
more squeaks from the same individuals emphasize the trend
to increasing complexity and increasingly invariant patterns
with less frequency modulation (3—12a,b,i—n). Maximum com-
plexity is seen in the "pain" squeak with its many strong
frequencies (Figures 3-1OC, 11f, g,h).

P.m.bairdi Type II squeaks do not show as much ener—
gy in the second frequency as those of P.m.gracilis or P.

californicus but rather a broader band and greater frequency

 

modulation for both strong frequencies (Figure 12f, h),

clearly different from Type I squeaks (Figure 3—12e, g, i-l).

 

91

 

Figure 3-12

Legend

Oscillograms and sound spectrograms for additional
distress squeaks from P.m.bairdi pups, showing ontogeneti-
cal changes and differences between Type I and Type II
squeaks. (a) and (b) 2-day old pup, the first frequency
peak in the spectrogram at (a) actually corresponds to the
lower frequency at the end of the squeak. (c) and (d) 6-
day old pup; (c) shows a Type I squeak with a higher basic
frequency than average but a purer tone than in (a). (e),
(f), (g), and (h) 4 squeaks, in order, from one set emitted
by a 7-day old pup. (e) and (g) are examples of Type I
squeaks, while (f) and (h) are examples of Type II squeaks
with more overtones present and lower basic frequencies.
Notice that all four have a virtually identical overtone
present at 43.2-44 kHz. (i) and (j), 2 squeaks from dif-
ferent lO—day old pups, aligned to show similarities. (k)
and (1) 2 squeaks from different 12—day old pups, also
aligned to show similarities. A strong increase in basic
frequency is apparent. (m) squeaks from l3-day old pup
showing visible complexities of intensity and frequency
modulation. (n) a very clear tone from a l4-day old pup,

here a fundamental is present for part of the squeak.

92

. I
"mull-IIIMIIIHII _ —. J I“

 

  
  

lroquonzy IS 256 50“ kHz "In.

2.2.4 25.7 44 20 43.2
I I

   

flru squeal:

P. manlculnuu balrdl

 

hum:

DOA 20522.4 42.. 22125.0
I I I I I .

    

.- . ‘ " ‘
IL: 32 5M kHz

FIGURE 3-12

93
After day 10 only a few major frequencies are emphasized
and the squeaks give the appearance of little frequency
modulation.

P.m.gracilis shows very similar patterns to P.m.
bairdi with major differences in the greater prevalence of
Type II squeaks. Frequency modulation is much stronger in
the _.m.graci1is squeaks than in the squeaks of the other
taxa recorded. Even squeaks from the l-day old pup show
strong modulation (Figure 3-l3a). A few recordings show
pure tones of Type I squeaks (Figure 3-13,c) but the major-
ity of the samples showed the multiple harmonics and the
frequency modulations of the Type II squeaks (Figures 3-3a,
b; 3-4d; 5g,h; lOe). Mean basic frequencies for §,m.gra;
.Eiiii Type II squeaks increase with age from 11.8 kHz and
24.8 kHz in the early period to 19.8 and 38.8 kHz and de-
crease for Type I squeaks, from the middle to late periods;
22.8 kHz - middle to 20.0 kHz - late.

Pain squeaks of P.m.graoilis are virtually indisting-
uishable from those of E,m.bairdi, being short squeaks, in
continuous groups with multiple strong frequencies (Figures
3-10c, 11f, g, h). Often very short squeaks are fused into
a short chirp (i - 0.1 sec.).

Data for g.polionotus and P.1eucopus demonstrate

 

similar ontogenetical trends. These two species differ from

the E,maniculatus mice in showing no Type II squeaks and

 

little frequency modulation. In both, frequencies rise and

become more pure with age (P,leucopus Figures 3-l4a, e, f,

 

 

 

 

 

94

Figure 3-13

Legend

Oscillograms and sound spectrograms from P.manicu-

latus gracilis pups; distress squeaks for several ages and

 

pain squeaks. (a) 1-day old pup; (b) 8-day old pup, show-
ing a clearer frequency than (a); (c) and (d) lO-day old
pup, 2 types of squeaks. (c) is a Type I squeak with a
single high frequency and one harmonic and (d) is a Type
II with almost equal sound energy at two frequencies; (e)
and (f) normal distress squeaks from 13-day old pup, the
same individual as in Figure 3-lld, the squeak in (f) is

a Type I squeak; (g) oscilloscope trace of pain squeaks
from a 14-day-old pup; (h) and (i) spectrograms of single
pain squeaks from the same individual. These squeaks show
broad frequency bands similar to the P.m.bairdi pain squeaks

in Figure 3-11g and h.

 

____#—__—__——————_

95

P. maniculatus gracilis

III
IIIWIIII‘

III:

III.
IIIIIIIIIIIIIIIIIIIII I

   

 

  
   

I 500 “Mb
yam. . .
tin I . .
“ I II

.IIIIIIIIIIIIIIIIIIIIII‘I IIII I Imw.

"Nu..." Inn to at 50.4 m:
Ill-8 4|.
c d

 

,III I :
IIIIr , p .
II“ II II I I
.....III IEIIII‘II I I II II rII LI IIIEIIIIII IIIIII III III. Ewe-III“ I ‘ II . I 4: ‘ quIIII..
l6 :0 24 50.4 I“! ‘ IN:
40
o I I II
3; III. III I [Jill
%— .1 3" {W I-I' 4. : II IV ‘ II'IM'
" umlm IIII I;.|.1I ‘II. h" I 1;. |‘ WI
Hm. I16 24 32 ‘0

Wit—IIHHHWW

 

.. I .
5 ".2 I6 20 24 32 50.4 I": I "2 IS 20 2G 32 IIHI

FIGURE 3 -13

 

 

 

96

Figure 3-14

Legend

Sound spectrograms and oscillograms comparing g.
leucoEus with 3,3.bairdi and 3.201ionotus across ages.
Time as marked. (a) g.leucopus, l-day old; (b) 3.20110-
EQEEEI l—day old; (c) E.§,bairdi, 3-day old; (d) 2320110-
Egtug, 3-day old; (e) and (f) 3.1eucoEus, 3-day old, note
basic frequency identical to E,g.bairdi in (c); (g) and
(h) §.1eucogus, 7-day old, same individual but time base
l/2 that in (f); (i) 9-day old g.leucoEus; (j) lZ-day old
g.leucoEus. As the pup ages there is a general tendency

towards sharper frequencies with fewer Visible harmonics.

 

 

 

a?“ 2i:

  

III I

 

I'IIIII'
III I II
III IIIII'IIVI WI

no 2
"HWY 75 “n

P. leucopus
P. pollonol‘ul

P. In. bull-d!

 

P. leucopus

[III I
I, ,IIIEII III IIIIIIWIW IIIIILIIHIIIIIII HIIIIIIIIIIIMWIIIII Imam .-

 

 

 

_

WM~:-CKMC’Q(D’*

- .1 no: 2'3 .6

I
III.
ngEII .. “5‘

III: III... .

3° 5° ‘ k"!

FIGURE 3-14

 

 

 

 

 

.4.—

98

6c, 3b, 4a; g.polionotus Figures 3-6b, 4b, 5d, 14b, d).

 

§.polionotus squeaks are typically sharper, shorter and

 

higher than those of P.1eucopus.

Age Differences - Temporal Parameters:

 

 

Numbers of Squeaks per set:

Significant differences were found over age among
Group A taxa (0.01 > P > 0.005; two -way analysis of vari-
ance). Numbers of squeaks per set rise to a maximum for

mefbairdi, prolionotus and P.1eucopus on day 2, and main-

 

tain a level until day 10, At this point P.1eucopus and
P.m.gracilis increase the number while Psmfbairdi and Bf

polionotus precipitously decrease. Efmfgracilis starts

 

on day 1 with a high number of squeaks, increases to day 2,
slowly decreases to day 14 and then increases again. (Figure

3-15).

Set Duration:

 

Mean set duration differed significantly over age
for the 4 taxa in Group A (P < 0.005; two-way analysis of
variance). Patterns of change resembled those of numbers
of squeaks per set, except that §,m.gracilis does not show

a rise in duration after day 14 (Figure 3-16).

§queak Duration:

 

The taxa in Group A differed significantly over age
for squeak duration (P < 0.001; two-way analysis of variance).
Squeaks tended to start short, lengthen to a plateau on day

3 or 4 and to remain fairly constant until day 10, at which

 

 

 

 

99

Figure 3-15

Legend

Number of squeaks per set: Daily mean numbers of
squeaks in a group plotted against age of pups. Solid

line = _,m.gracilis; dashed line — P.m,bairdi; diamonds=

§.polionotus; elipses = P.1eucopus. Solid arrows indicMw

 

predicted age of ear opening, as demonstrated by a starfle
response. Open arrows indicate predicted age of eye opmr

ing. Circles on lines indicate estimated age at erectflm

of pinnae.

 

 

can we n>01

 

 

o. n. e. n. d. : o. o a h o m c n a .
Ls .
38:0:on 82¢ angina o.«
:23 Song.
:30 oao u... n.
\ //.\\/ fl
m:_oo.6.Em / Jed
/
/ \\ m n
< \ .
x, \ I
c I m.‘
m M“ \ ....
l 00/.II- 40“
w 9m
a...

... IL. .

a .m

w 3m .2. 9.835 S .3532 m

I. n

 

 

 

 

.g.

101

Figure 3-16

Legend

Set duration: Mean duration, in seconds, of sets

 

of squeaks, plotted against age of pups.- Solid line =
_.m.gracilis; chain = P.m.bairdi; diamonds = P.201i0notus;
elipses = P.1eucopus. Arrows and circles, as in Figure
3—15 indicate the estimated ages at eye and ear opening

and pinnae erection.

 

 

 

¢ m N

one Go 225

 

 

m. n. v. n. N. :
9:58 Sou
oFano axon»
e .cooWQMcgoqm
__:_B.em8.:: +

m___oEo.Em

1.02

FIGURE 3'16

5:950 8m

Soto cocci no

 

0

N6

*0

0.0

m6

0..

N._

seconds

¢._

w._

m._

O.N

N.N

TN

 

 

 

 

103
point they decrease (Figure 3-17). P.m.gracilis shows an
erratic lengthening and shortening of squeaks after day
12. All the parameters show the abrupt changes about day
3 and after day 10, approximately the days when the pinnae

become unsealed and the eyes and external ears open.

Discussion:

 

The observations and analyses reported in this chap-
ter demonstrate the physical differences which exist be-
tween the distress squeaks of pups of the various taxa

examined. Typically P,maniculatus pups show major energy

 

peaks between 20 and 24 kHz. Some of the subspecies, par-
ticularly P.m.gracilis, produce Type II squeaks with sever—
al strong peaks of energy while other subspecies, 23m.
bairdi, more usually produce Type I squeaks. Fundamentals
are often visible, particularly with the youngest pups and
at all times the frequency should make the pitches of
squeaks of the other Species perceptibly different to these
mice.

On temporal parameters the three subspecies of P,

maniculatus do not show clear relationships; P.m.nebracensis

 

 

being different on all three parameters (set duration, num-
ber of squeaks per set, squeak duration) from the other two
and P.m.gracilis and P,m.bairdi similar only in the duration
of sets. P.m.bairdi has longer squeaks than P.m.gracilis
but fewer of them, while P.m.gracilis has many shorter

squeaks. Related vocal musculature may explain similar

 

 

 

 

104

Figure 3-17

Legend

Squeak duration: Mean duration of distress squeaks,
in seconds, plotted against age of pups. Solid line = P.m.
gracilis; chain = P.m.bairdi; diamonds : §.polionotus;
elipses = P.1eucopus. As in preceding figures (3-15, 3—16)
arrows indicate estimated ages of ear and eye opening and

circles indicate age at pinnae erection.

 

. {I

 

 

can .0 {on
w. n. v. n. N. = o. w n o n e n N _ 0
3.3235. 05:23 .3 3 $20 02:.in o_.o
. 95:23 :03 a
33:0:on . a.» : o
- * f I
a ...... ... «.6
. O“ Illll‘ii
u-zoozooooo M \ 1n...0
_E_ua.sm .. 5.. u \
.. . .
.._. ..... m .. ... \ so
a
5 2.0
m. 4 / ...
+
9.0
\x/ to
a //.. \
n: cos. . .
E .. .IL 2.0 ,
. 2.0
.. and
‘86 «2.30m

FIGURE 3'17

 

3:23 .335

 

 

 

 

 

106
frequencies while differential emphasis on frequency bands
(formants) components may provide sufficiently divergent
pitches for specific identification by the mice. Temporal
components may not be as important as pitch in recognition
by the mice.

P.polionotus is a member of the P.maniculatus spe-
cies group, P.1eucopus is grouped in the same subgenus
(Hooper, 1958) and both may be sympatric with P.maniculatus
in areas of the Eastern United States. The P.1eucopus and E?
P.maniculatus mice used in this study were from populations
which were sympatric in some areas. At the same time the
mean basic frequencies of P.1eucopus and 3.polionotus differ
the most from frequencies of the other species. Both 3.
leucopus and P.polionotus have high squeaks (overall mean
basic frequency 33.1 kHz, for P.polionotus; 32.1 kHz, for
P.1eucopus) and the two species have similar numbers of
squeaks per set and set durations. P.1eucopus has longer
squeaks than g.polionotus. On 3 of the 4 major squeak
characteristics measured, P.1eucopus diverges from P.m.
notus is similarly different from P.maniculatus mice.

The other 3 species of mice tested are completely
allopatric and more distantly related. No particular pat-
terns of characteristics emerge except that all show fre-
quency patterns similar to the P.maniculatus mice. On
temporal characteristics §.difficilis most resembles E.

leucopus. B.californicus and P.melanophrys are alike on

 

 

—‘—P

107

all measures except frequency, where they differ slightly.

The original supposition that frequency patterns
would be more conservative to change than the temporal
patterns appears to be supported by the similarities in
frequency found in the P.maniculatus mice. Phylogenies
might be based on sounds with more data on additional spe-
cies and local populations.

A reduction in variation and a general refinement ‘

of a typical pattern Show up in the comparisons of the

I A’;’;-.:
wan—r"

squeaks from pups of increasing age. As expected, squeaks
do become shorter and less frequent as the pups grow older.
In addition, frequencies tend towards purer tones, a change
which must carry with it some'change in pitch.

Temporal and frequency characteristics do not change
at a constant rate but show the greatest variation in the
several days before the external pinnae are fully extended
and around the time of ear and eye opening. Although pups
do not respond to loud sounds before the 12th to 13th day
when their ears open (Huff pers. obs., 1968; Layne, 1968),
major changes in temporal characteristics and frequency
occur on about the third day after birth. It may be that
the ear starts to function at this point and the pup begins
to get some auditory feedback through bone conduction at
least.

More gradual changes in set duration and the number
of squeaks in a set do occur and may be related to decreasing

motivation to squeak, culminating in reduced or absent

 

108

squeaks at the time of eye opening. The auditory cues to
the mother thus become weaker stimuli correlating with her
changing hormones and resulting in diminished motivation
for search and retrieval.

Although each of the 8 taxa recorded differed sig-
nificantly from each of the others on at least one of the
4 major parameters no correlation was found to exist be—
tween rankings on means of any of the parameters and over-
all body size. §.polionotus, the smallest mouse recorded, 3%
has the highest mean basic frequency but 3.1eucopus, a 3:
middle—sized mouse, can produce even higher frequencies

at the early stages. g.californicus, 3.difficilis and §.m.

 

gracilis, all large mice, have equally high frequencies

in their Type II squeaks. E.polionotus has the shortest
the shortest sets, and 3.1eucopus has the fewest squeaks
per set. Frequency is the parameter most likely to vary
with size but the mice recorded may not be divergent enough
in size to have functionally different vocal cords. Size
of pups or of actual larynxes may correlate better with
mean basic frequency.

Changing frequencies and multiple strong frequency
bands probably make the pitch of squeaks of pups of various
ages and species perceptibly different to the mice. When
recording tapes were slowed by a standard amount (15 ips
to 1-7/8 ips) it was possible for human ears to hear pitch

and pattern differences. The relative importance of pitch

 

 

 

—4—

109
and the temporal parameters in information transfer for
the mice cannot be precisely assessed from the data presented
here. It might be suggested that both would be important
as elements of redundancy in the information transfer, with
frequency carrying more subtle cues on age, species, and
perhaps individual identity.

Efficient localization of the source of a distress
squeak by a mother mouse depends on her detection of dif-
ferences of intensity, phase and time of arrival of the i
sound at her two ears. Localization by means of intensity
is most effective when the head is wider than the wave—
length of the sound and can block the sound from directly
reaching one ear or the other (Marler, 1961, 1967). Locali-
zation by means of phase differentials is best when the
sound reaches both ears directly but out of phase, when the
wavelength is longer than the width of the head between the
ears. Abrupt discontinuities and transient frequencies
permit localization by comparison of the arrival times at
the two ears.

This last characteristic is apparent in all the dis-
tress squeaks recorded, although lacking in the pain squeaks.
Sharp onsets and offsets and grouped short squeaks provide
ample discontinuities for time comparisons and thus locali—
zations of distress squeaks by means of time differences.

Measurements of the widths of the skulls of the
mothers of the pups in this study were not made, but rough

estimations were made on the basis of skull measurements

 

 

 

 

110
given for g.m. gracilis, g.m. bairdi and g. leucopus in
Burt (1957). Width measurements for skulls used in that
source were made on distance between zygomatic arches.
Ear to ear measurements made on living mice with calipers
agreed, in general, with these data. The following table
collects frequency data and estimates head width for the

3 taxa.

 

Table 3-5

Comparison of estimated ear—ear head width measurements
of adult mice of 3 taxa of Peromyscus with mean basic
frequencies of distress squeaks of pups of the same taxa.

< . 5,;5. :
x...“

 

 

 

Mean width Mean width Wavelength of mean
Species zygomatic arch(mm) ear-ear(mm) basic frequency (mm)
23m, bairdi 12.0 11.7 early 15.9
middle 13.8
late 13.8
33m, gracilis 13.1 - early 27.6 G 13.3
middle 17.4 G 8.5
late 16.6
E; leucopus 12.6 12.7 early 8.7
late 11.8
late 11.1

 

As the measurements of head width on the living mice

might differ slightly from skull measurements due to the
presence of fur, skin, and pinnae, the effective width
would be expected to be somewhere lower than the estimated

ear-ear measurement and the zygomatic area measurements.

 

 

 

111

Since the wavelengths of the mean basic frequencies are
longer than the estimated head widths of the relevant
hearers for E'E' bairdi these mice would be able to local-
ize the sources of the sounds by means of phase differen-
tials. Wavelengths for frequencies for B.m. gracilis are
also longer than the estimated head width with the ex-
ception of the short mean for the Type II squeaks for the
early and middle age groups. If these high frequencies are

heard as separate waves E'E' gracilis might be able to

.....<-:.:

utilize both localization methods, phase differences and
differences in intensity. All of the wavelengths calcu-
lated for mean basic frequencies for g. leucopus are shorter
than the maximum head width. It is thus possible that

these mice could also use both phase and intensity differ-
ences but more precise measurements of living mice are needed
for a definitive answer. All of the species considered have
at least two methods of accurately locating the sounds pro-
duced by their pups. Redundancy provided by the repetition
of the squeaks would improve efficiency.

Pain squeaks do not provide easily localizable charac-
teristics and are analagous to the warning calls of birds
which alert others to danger but are difficult for a pred-
ator to locate. The pain squeaks recorded for the g.
maniculatus mice were very similar between the two sub.
Species, with gradual onsets and offsets, vague intensity
modulation and multiple frequencies. Intensity differences,

phases and times of arrival would be almost impossible to

 

 

 

112
distinguish. In addition multiple repetitions of these very
short squeaks would insure that the information was trans-
mitted.

The observations based on the analyses of these squeaks
suggest that the obvious characteristics of the sounds have
important functions in the communication systems of these
animals. Even greater subtleties of information transfer
probably remain to be explicated by further analyses and

l".
experimentation. ’fl
. .

 

CHAPTER 4

Effectiveness of Auditory Stimuli in Eliciting Maternal

Searching: Activation Test.

Schleidt (1956), Noirot (1960; 1964c), Sewell (1970)
and others have mentioned that mouse pups which do not squeak q
are not retrieved. In pilot studies, lactating Peromyscus ft
maniculatus rarely left their pups and nests to explore
the experimental chamber unless they heard a squeaking pup
or a recording of one. The following experiment was de-
signed to test the hypothesis that the squeaks of mouse pups
out of their nests stimulate a mother mouse to search for
the sound source. The pilot studies suggested that this is
a sufficiently strong stimulus to cause a lactating female
to leave her nest and nursing pups and enter a relatively
strange, open arena. Without this stimulus a mouse might
spend most or all of the daylight hours in her nest with

her young (King, 1963).

Subjects:

Forty-two lactating Peromyscus maniculatus bairdi

 

first generation daughters of wild—caught parents were
used as the experimental animals in 3 groups of 14 mice

each. All the females were between 6 and 18 months of

113

 

 

114
age. Each mother mouse was assigned to one of the three
experimental groups, in sequential order, as her litter
was born. Those females which cannibalized their young
before the test were not used in the experiment. All the 1
litters were randomly culled to four young one day after
birth. The females were maintained with their mates and
litters in standard colony cages in the mouse colony room
until 1—3 days prior to testing. At this time the females .=W
and their litters were moved into a room adjacent to the Em

test chamber. i?
Experimental Apparatus:

During the pretest and test days each mother and
her litter was housed in a modified colony cage which
had a 1-1/2 in opening in the lower center of each end.
These openings were glued to plexiglas tunnels (2" x 2")
and until test time they were both blocked with guillotine
doors. The plexiglas tunnels connected the cage to two
square (2' X 2' x 18" high) plywood arenas (Figure 4—1).
These bottomless arenas were placed on the vinyl-tile
floor and were covered with 1/4 in screen. The entire
Sound Choice Box (SCB), with the exception of the mouse
cage and tunnels, was painted flat black. At each end of
the SCB, opposite the tunnel entrance, a screened opening
(8" x 6" high), covered with black cloth, concealed a
high—frequency speaker (University Sphericon Tweeter).

As diagramed in Figure 4—1, photocells placed beside the

   

 

115

Figure 4—1

Legend

Sound Choice Box. A mouse placed in the central
cage could be exposed to sounds presented from either
speaker (S) or both simultaneously. Subsequent activity
could be monitored automatically by means of photocells
(P) triggered by lights (L) placed on the opposite side
of the apparatus. Each arena is an open field 24 inches
square. Photocells are placed at critical position

assessment points rather than equidistantly.

 

 

 

116

FIGURE 4-I

:3 3.23 5:3»

5 a

 

Eu: :3

 

m

 

 

 

“:9

I I

 

 

 

 

 

Eu: 2:: .

 

 

   

   

 

117

SCB arenas and the tunnels and connected to an Esterline-
Angus Event Recorder monitored the movements of the mouse
in the apparatus.
The Sound Choice Box and speakers were placed in a
sound—damped room and connected by cables to the tape re-
corder, amplifiers, and event recorder in an adjacent room.
From 0600 to 1800 hours the test room was illuminated by
a 60—watt light bulb suspended from the ceiling and the
rest of the time only by the six 10-watt red bulbs used ;

with the photocells.
Test Sounds:

The three sounds used in this test were recorded
with the same equipment and procedures as the squeaks for
analysis (Chapter 3). Sounds were recorded for five
minutes on one tape segment, each on a separate channel.

The squeaks of a five-day—old E.maniculatus bairdi were

 

' noise from

used as a test sound and compared to "white'
radio static and mechanically generated clicks. The
"white" noise was presented as a control for reactions to
possibly intermittent speaker hiss and the mechanical
clicks were presented as a control for responses to sounds
having some similar characteristics to the squeaks. Since
all of the sounds were on one segment of tape it was possi-
ble to present them simultaneously,in any pairing desired,

by plugging the appropriate channel into the speaker.

Each of the three experimental groups was presented

 

 

 

 

118

with a different combination of the sounds. The "squeak—
control" group heard the recorded mouse squeaks and the
control "white" noise, the "click—control" group heard the
clicks and the "white" noise, and the "choice" group heard
the squeaks versus the clicks. In this way the mice were
tested for their responses to the mouse squeaks alone, for
responses to a temporally similar sound alone, and for

their choices between the mouse squeak and the partial copy,
differing primarily in frequency. That sound to which each
group was expected to respond was defined as the "correct"
stimulus. Thus for the first group, squeak-control, the
"correct" stimulus was the mouse squeaks, for click-control,
it was the clicks, and for the choice group, the mouse squeaks

were defined as "correct" and the clicks as "incorrect."

Procedure:

Each mouse was placed with her 4—6 day old litter in
the central cage in the Sound Choice Box on the afternoon
of the day preceding the test day. She was permitted free
access to both arenas for the 22—24 hours prior to the test
but her activity was monitored by means of the photocells
and event recorder. The activity records were examined
before testing in order to discover preferences for one
arena over the other. Where preferences appeared, the
initial test sound was presented from the non-preferred
side, otherwise initial presentations were random. I

Immediately before testing, the tunnel doors in the

 

   

 

 

119
central cage were closed and the mouse pups were removed
to another room. Then the doors were raised and the
mother was permitted to search for them. The length of
time until she returned to the nest and remained there
for one minute (as indicated by the activity record) was
scored as the "latency—to—calm."

Once the mother mouse had remained in the nest for
one minute, the test sound was presented at one end of the
SCB. The taped sound was played continuously until the
mouse entered one of the tunnels or until five minutes had
passed. If the mouse responded the sound was continued
for only one more minute.

Most females returned to their nests within one
minute after the sound had been turned off. All were allowed
a one minute intermission between the time they returned to
their nests and the next presentation of sound. In the
second trial the test sound was presented from the opposite
speaker. These reversals could be accomplished without
entering the experiment room by alternating the speaker
connections to the recorder. The protocol for the second,
third and fourth trials was identical and each time the
test sound emanated from the alternate speaker. The test-

ing protocol can be outlined as follows:

 

 

 

 

 

120
cage closed + litter out + cage opened
+

"time—to-calm" Z 3 minutes

Arena A l Arena B
1 minute intermission

sound 3 5 minutes test control
1 minute intermission

control : 5 minutes test sound
1 minute intermission

sound 3 5 minutes test control
1 minute intermission

control i 5 minutes test > sound
litter returned

Variables:

The independent variable was the sound presented.
All other factors which might have influenced the response
of the mouse, such as age, sex, experience, reproductive
state, age of own litter, and some attributes of the stimu—
lus were held as constant as possible.

The Sound Choice Box was designed to permit measure—
ment of the dependent variables of response latency, direc-
tion of choice, number of choice changes,and the distance
traveled to or from the stimulus. Latencies were measured
in seconds with a stopwatch, from the time the sound was
turned on until a recording pen clicked as the mouse moved

into the beam of a photocell. These measurements were

. tau 1'4: .1. .;....

 

   

 

121

compared for accuracy against the same time measured by the
event recorder. Choices of direction were indicated on the
chart of the event recorder. Distance traveled by the
mouse was recorded as the farthest photocell triggered in
any excursion from the nest.

The experiment was planned so that parametric or non—
parametric analyses of variance could be done on the test

averages for the subjects in each group. Comparisons could ‘

 

be made between groups for all variables and within groups

‘v-' r“-
OFT—h

between numbers of "positive" and "negative" responses, 3:

and distances to the different sounds.
Results:

The dependent variables in this test can be con-
sidered to fall into two categories: variables related to
general activation and those related to preference for one
sound over another. Response latencies and total number
of responses were measures of activation, while numbers of
"positive" responses, reversals, and distances traveled in
particular directions were measures of preference. Two
variables related to the homogeneity of the experimental
groups will be reported first as they support the contention
that the experimental groups were similar in composition.
Litter Age

In order to ensure a relatively homogeneous repro-
ductive state among the experimental animals and a similar

level of responsiveness to pups the mice were tested when

 

 

 

 

122

their litters were about six days old. Since any one mouse
occupied the test apparatus for 24 hours prior to the actual
test, logistics occasionally demanded testing a mouse with
a litter that was one or two days older or younger than
six days. Nevertheless the mean litter age did not differ
significantly across groups (one—way parametric analysis of
variance) and the variances were homogeneous (Table 4—1).
Latency—to—Calm

Once the mice had been established in the SCB and
their litters removedythe length of time required for them
to cease searching, to return to their nests, and remain
there for at least one minute was measured. It was expected
that this latency indicated the level of arousal of the
mouse after the disturbance of the removal of the pups.
Although some mice might be more excited and more active
following this treatment than others, the experimental
groups should not differ significantly in latency since
this procedure occurred prior to the experimental treatment.
All of the mice took at least three minutes (counting the
one minute in the nest) to reach the desired level of calm
and there were no significant differences in mean latencies
between the groups (Table 4—1).

Total Number of Responses

 

The total number of responses were defined as the
numbers of trials in which the mouse responded at least
once; the highest total a mouse could have was four. If

the mouse changed its direction and went to the other side

 

 

 

 

 

 

123

 

 

$.31 H Nmémm 3.0 H ma 3 .885
Hm.mm H H>.vmm Hm.o H v.m ea HoupsOOIMoflHU
Nw.mm H mh.owm Hm.c H m.m «a Honpsoolxmmsvm
.m.m m .m.m m a wmsouw
Easeloulmocmpmq 0mm poppflq
.Ammusaflfi m Hmpwm mpcoommv SAMUIOplmosmvmq
paw Amhmpv mmm kuuflq "mQSOHm Hapsmfiflnmmxm cough QH mudguoa msumHSOquE

wsommfioumm mo mmocm>flmaommmn paw wumum m>fluosooummn mo wuflmqwmofios mo monommmz

HIv magma

 

 

 

 

 

124

of the box, the switch counted as a reversal. A complete
response was scored if the mouse only triggered the photo-
cell in the tunnel between the cage and an arena.
Heterogeneity of the variances required a non-
parametric statistical test (the STP) to evaluate the dif—
ferences between the mean numbers of responses. The groups
did not vary significantly in the mean number of trials in
which there was a response (Table 4-2). However, the
variation in the means was much lower in the choice group, .fi
indicating that more mice in that group responded in at
least three trials than in the other groups.
Number of First Responses Correct
The mean numbers of first responses which were to
the correct side, i.e., to the squeaks for the squeak—
control and choice groups and to the clicks for the click-
control group, differed significantly among the groups
(parametric one-way analysis of variance 0.05 > P > 0.01).
By the New Multiple Range test the greatest part of the
difference was found to be between the click—control and the
choice groups (Table 4—2), at the 1% level.
More mice in the squeak-control and choice groups
directed greater than 50% of their first responses to the
mouse squeaks than would have been expected by a hypothesis
of random response. The preferences in these two groups
were significant by the chi—square test (P < 0.01 and P <

0.01 respectively).

 

 

125

 

Ampqoommv

mo.o om.mHH. mm.naa mH.ooa mmsommon nomnnoo ow moneyed
Ampcoommv

Ho.o mo.an ma.mHH em.em mmqommmu umnflm on socmuma

mo.o «H.m no.m mv.m pomnuoo

mmmcoamwn #mnflm Mogadz

.m.a mm.m em.m mm.m mchQmmu a nun; mamaue
ea ea ea c
m oofiogo Honucoo I MUHHU Honucoo I xmmswm

 

#mmH COHHMNVHHOM I
d3.” mmVGDOm HOHHGOU mod—.6 mxmmgvm mmwhHmHU wDHMHSUHQME.nH OH MHQQHOE
mDHMHSOHGGE WSOWWEOHQQ MO mwflOCwHMH @mflOQmQH GMQE UGM mmeOQmmh QMGE

 

va GHQMB

 

126

Latency to First Response

 

The time from onset of the test sound until the
mouse triggered either of the tunnel photocells was de—
fined as the latency to first response. Variances in these
latencies were quite large and the data skewed to the left,
so a non-parametric Kruskal-Wallis one-way analysis of
variance was used. This test showed that the groups did
differ in latency (0.01 > P > 0.001). The squeak—control ”
group had the shortest mean latency to respond and the jim
click—control group took the longest time to make any re- I
sponses (Table 4—2).

Latency to Correct Response

 

This was the measure of the time from the onset of
the test sound until the mouse triggered the tunnel photo-
cell on the side of the SCB where the "correct" sound was
being played (defined as the mouse squeaks for the squeak—
control group and the choice group, and the clicks for the
Click—control group). The variances in latency were again
quite large and the data negatively skewed so the Kruskal-
Wallis test was used. Although the means of these three
groups were closer when correct responses were considered
than when any responses were considered they were never—
theless significantly different at the 5% level
(0.05 > P > 0.02). Again the squeak—control group had the
shortest latency and the click-control the longest.
Distance

The distance a mouse moved in either arena was

 

 

 

 

—__

127
measured by the photocells triggered as the mouse passed.
It was impossible to tell whether the path of the mouse
was straight from one photocell to the next or composed'of
numerous advances, retreats, and excursions from one side
of the box to the other. Observation showed that both
types and all sorts of intermediate variations occurred.
In order to avoid "losing" the mouse, the photocells were
positioned at critical points in the SCB; near the speak— '
ers, across the entrance to the arenas, and across the .%
tunnels between the arenas and the cage, rather than equi-
distantly. Since they thus measured not absolute distance,
but just the number of times the mouse passed by, the re-
sults are tabulated in the form of scores. Each photocell
was numbered by its position relative to the central cage;
the number ones crossed the tunnels, number twos caught
the mouse just entering the arenas, and the number threes
were nearest the speakers. A mouse that ran all the way
to the speaker in each trial would accumulate a mean score
of 3 while a mouse which never did more than stick its
head into the tunnel would have a mean score of 1 (Table
4-3).

Distance Score to Correct Response

 

These data showed homogeneity of variance so the
null hypothesis that the groups did not differ was tested
by a parametric one—way analysis of variance. The groups
were found to differ significantly overall (P < 0.05).

By the New Multiple Range test the click-control group

 

 

 

 

 

128

‘(w,. . int-30:.

 

mo.o

Ho.o

mo.o

mm.H

mn.o

mm.N
«a

moflozo

mm.m. mv.m mammnm>mn moaogo

va.m mn.a mmaommmu uomuuoosfi mocmumflo

mm.H mm.m mmsommmn pomuuoo wocmwmflo

vH ea c
Houps00|x0HHO Houpsoolxmwsvm

 

mUGSOm Honpcoo paw wand

msumasoflcme.m mo mxmwsvm mmmuumflp ou msflpcommmu mHmSpOE

 

mspmasoflc

ME WDOWNEOHQAH m0 MHMWH®>®H QUHOSU CMOE USN mmHOUm QOQMHWHU EGGS

Mlv GHQMB

 

129
differed from the squeak-control group at P < 0.05 and from
the choice group at P < 0.01 (Table 4—3).
Distance Score to Incorrect Response

Here also, variance homogeneity permitted a para-
metric one—way analysis of variance. This time the dif-
ferences in mean score were significant at the 1% level
and the STP test showed that these differences were
concentrated between the squeak-control and choice groups
and the click-control and choice groups. There was no _1
significant difference between the two control groups
(Table 4—3).

A T-test on the differences in mean scores to the
correct side and to the incorrect side for each group
showed that all three groups differed significantly in
their distance scores (squeak—control = P < 0.05, click-
control = P < 0.025, choice = P < 0.001) although the
choice group differed the most. The click—control group
went farthest to the incorrect side while the squeak-control
and the choice groups went farthest to the correct side,
i.e., to the recorded squeaks.

Number of Choice Reversals

 

The number of choice reversals was defined as being
the number of times a mouse triggered a photocell on the
Opposite side of the SCB from the previous mark in any one
trial. A mean number of reversals was calculated for
each mouse. The variances in this measure were non-homo-

geneous so the Kruskal—Wallis one-way analysis of variance

 

 

 

 

130

was used to test for a difference in means. The results
of this test showed that the mean number of reversals did
differ between the groups (0.02 > P > 0.01). The click-
control group ran back and forth the most and the choice

group the least (Table 4—3).
Discussion:

The results of this test of distress squeaks as a
factor in activating mother mice to search were interpreted ,i
as supporting the hypothesis that the squeaks do cause the ffi
mother mice to come out of their nests to search for the
source of the sound. Since the only cues present were the
recorded squeaks there was no possibility that other
channels of communication were available to carry this in—
formation to the mouse. Had live pups been present with
their unique combination of odors, sounds, and tactile
sensations the retrieving responses of the mothers might
have been faster and even better directed, but with sound
alone the mice went as close to the sound source as they
could get and struggled to get closer.

There can be no question that the mother mice readily
distinguished the recorded squeaks from the'White'noise.
They showed by their differential approach behavior that
they preferred the squeaks. There did not appear to be
any aversive factor to the White"noise as the mice readily
entered the arenas with the White"noise when they continued

their searches for the invisible squeaking pups. The mice

 

 

 

 

j

131
would have meant going to the clicking side, and went less
far to the clicks. The mice appeared to have no trouble

differentiating the three sounds.

 

 

 

 

CHAPTER 5

Species-specific Discrimination Based on Auditory Cues:

Discrimination Test.

Peromyscus maniculatus females remained in their

 

nests and did not respond to recorded distress squeaks of {
P.californicus pups in a pilot study (personal observations, {fl
1968), although in the same test they did respond to the
calls of pups of their own species. This suggested that
some species-specific discrimination of distress calls was
possible and that comparisons of the responses of mice of
different species to recorded distress calls of each species
might elucidate the ability of mice to distinguish related
sounds.
If the differences between the distress calls of dif-
ferent species, as shown in Chapter three, are the result of
character divergence,then sympatric species should not only
show physical differences in their calls but should also
be able to differentiate them. In our trapping areas in

central lower Michigan we had trapped P.maniculatus bairdi

 

and P.1eucopus in the same trap line. Whether this shows
territorial overlap as well as geographical sympatry is not
clear but the mice were surely within hearing range of each
others' squeaks. For the maintenance of separate niches

and efficient propagation of the species it would be

132

 

 

 

 

—

133
advantageous for these mice to have some means of recogniz—
ing conspecifics at a distance.

I therefore hypothesized that if there was no con—
specific recognition of infant distress calls, mother mice
would be unable to demonstrate a choice between calls re-
corded from con—specific and non-specific pups and played
in the SCB. As it was evident from observations in the
colony that P.1eucopus mice tended to be more "nervous" and
generally active than the P.maniculatus, I decided to test in
similar choice and sound-control groups of both species.
Differential discriminatory abilities were not unexpected
although discrimination due to increased activity could not

be separated from discrimination due to superior hearing.
Subjects:

Twenty—four lactating P.maniculatus bairdi, and 16

 

lactating P.1eucopus were used in this test. Each species
provided two experimental groups to which mothers were as—
signed, in sequential order, as their litters were born.
Maintenance and pretest procedures were identical to those
in the Activation test. The 24 P.maniculatus were all Fl
daughters of wild-caught parents. Four of the P.1eucopus
had been wild—caught themselves as adults and the rest were
first to third generation offspring of wild—caught mice.
Earlier studies on mice from the same stocks had shown that
even 17 generations of semi-domestication in the laboratory

did not significantly alter activity in a novel environment

 

 

 

jun—-

134

in the squeak-control group searched for the source of
mouse squeaks and went correctly to the source a signifi-
cant number of times.

Mice in the click-control groups did not hear the
mouse squeaks but rather sounds only superficially repre-
senting them. In this case their correct responses did
not significantly differ from random, i.e., that which
would have been expected if they had just been alternating
sides. The sounds did activate them to leave their nests. :j
But they neither came out so much, nor so fast as the 4
groups hearing mouse squeaks. When distances were calcu—
lated this group was seen to have gone farther to the
negative side, that is, the "white" noise, suggesting an
avoidance of the clicks.

When they were given a choice between squeaks and
clicks all of the mice preferred the squeaks. Although
intermediate response latencies suggested that the mice
in the choice group actually had to chose sides, when they
came out they significantly chose the squeaks. Once in the
arena with the squeaks they tended to go farther, and more
of them went all the way to the speaker where they tended
to stay,leaping at the walls, continually triggering the
photocell, as if attempting to get closer to the source.

In part,the long time spent at the squeaking speaker ex—
plains the greatly reduced choice reversals in this group.
The results from the click-control group suggest that the

mice in the choice group had fewer choice reversals, which

 

 

 

135

(Price, 1967). Although a correlation between semi-domesti-
cation and latencies to enter an open field was found in
that study (Price, 1967) only three generations of labora-
tory conditions were not expected to have any discernable
difference. In addition some of the P.maniculatus had been
used in the Activation or Localization tests,or both,but as
the object was to demonstrate species differentiation a
certain practice in retrieving or additional familiarity

with the test box was not expected to affect the test results.

.m. —.

Equipment:

All of the apparatus used in this test was identical
to that used in the Activation test. The sounds played to
the mice differed,in that the recorded squeak of a 7—day

old P.1eucopus was substituted for the clicks.
Procedure:

In order to allow comparisons between the groups in
both the Activation test and this Discrimination test, test—
ing protocols, with one exception, and all pretest procedures
were the same in both tests (see Procedure — Chapter 4).

For each species the squeak-control group heard the P.1euco-
pus squeaks versus white noise and the choice groups heard
the recorded squeaks of the two species simultaneously

(Table 5-l).

 

 

 

 

136

 

wWHOC :®#H£3=

 

mmaoc .mufla3=

msmoosma mapmasoflcmfi m
msmoosma m
msmoosma msumasoflcme NH
msmoost NH

BmMB ZOHBflZHZHmomHQ

 

meHHo msumasoflcme ea
wxoaao mmaoq =mpfls2= ea
wmfloc =mpH33= msumasoflcmfi ea

flasfiflum ccsom c

EmmB ZOHB<>HBU¢

wofloco

Houucoouxmmswm

mOHoco

Honpcoolxmmsvm

mofloso
HOHuGOOIMoHHU

Houucoolxmmsvm

mdouo

msmooswa.m

mSuMHSOHcmE.M

msumasoflcmfi.m

mmflowmm

 

.xom mofloco ccsom on» ca
mumxmmmm mcfimommo Eoum mamsomcmuaseflm mUGDOm OBu chum: moHE mo
msoum comm .mummu scaumcflfifiuomfla can cofium>fip0< map ca wmsoum
Hopcmfiflnmmxm 0p mGOHuMpcwmem wsasfiflpm mo HOUOpOHm m>aumummfioo

Him OHQMB

 

 

 

 

 

137

Variables:

Since this test was an outgrowth of the Activation
test the independent variable was again the sound presented.
Other factors which might have influenced the response of
the mouse, such as age, sex, experience, reproductive state,
age of own litter, and some attributes of the stimulus were
also held constant. The dependent variables of latency—to—
enter, choice,and distance were recorded by the event-re-
corder and analysed. For the squeak-control groups a posi- {fl
tive response was considered to be a response to the P.1euco-
pus squeaks and for the choice groups a positive response

was one to the con—specific squeaks.
Results:

As in the Activation test the dependent variables were
considered to fall into the two categories of those related
to general activation and those related to preference. Re-
sponse latencies and total numbers of responses were con-
sidered to be measures of activation and numbers of positive
responses, reversals, and distances traveled were measures
of preference.

Litter Age

Over all the groups in this test the mean litter age
was 7.6 days, slightly older than the litters of the mothers
in the Activation test (Table 5—2). Within the four groups
of the Discrimination test the mean litter age did not differ

significantly (Kruskal-Wallis one-way analysis of variance)

 

 

 

 

138

Table 5—2

Measures of homogeneity of reproductive state and
responsiveness of Peromyscus maniculatus and P.
leucopus mothers in four experimental groups of
the Discrimination test: Litter age (days) and
Latency—to-calm (seconds).

 

 

«4...:

'c-_

Litter age Latency—to—calm
Groups n x S.E. i S.E.

P.maniculatus

squeak-control 12 8.2 i 0.32 194.5 : 43.58*

choice 12 7.6 i 0.38 135.5 : 47.77
P.1eucopus

squeak—control 8 7.9 i 0.50 410.5 : 95.67

choice 8 6.8 + 0.78 222.9 + 73.76

*n's reduced in all groups in calculation of mean and standard
errors due to missing data. '

 

139

and I thus considered them to be at essentially the same
point in their hormone cycles.
Latency-to-calm

Over all four groups the mean time for the mothers
to calm down after the removal of their pups, until they had
remained in their nests for 30 seconds did not differ sig-
nificantly (Kruskal-Wallis) one way analysis of variance).
In the Activation test the latency—to-calm was measured at
the end of one minute but in this test the 3.1eucopus mice
were found to be so active that they were rarely still for
one minute after their pups were removed. In order to have
a standard of comparison between the two species the waiting
period was therefore shortened to 30 seconds (Table 5-2).-

Total Number of Responses

 

All of the P.maniculatus mice responded in every
trial, thus reducing the over-all variance sufficiently to
permit a parametric analysis of variance. The one-way ana-
lysis of variance demonstrated that the groups differed sig-
nificantly in mean number of trials in which there was at
least one response (P < 0.005). This significance was at—
tributed to the greater number of trials in which the P.
maniculatus groups and the P.1eucopus squeak-control groups
responded than the P.1eucopus choice group (New Multiple
Range Test; P.maniculatus groups differed from P.1eucopus
choice P < 0.01; P.1eucopus control differed from P.1eucopus

choice P < 0.05: Table 5—3). The greater response of the

 

 

 

140

.m.C

.m.G

moo.o

m>.HNH ne.om hm.o> mm.om Amwcoommv
mmcommwn poonuoo op hocwumq

VN.NvH mo.mm ov.mm VN.mm Amficoommv
mmcommmu pmuflm ou mocwumq

mN.N oo.m oo.m hH.M HOOHHOO
momsommwn pmuflm HmQEsz
om.m mm.m oo.e oo.e mmcommmu m and; mamflua

m m NH, «H c
oofloso aonucoolxmmsvm moflosu HOMpGOOIxmwsqm
msmoost.M mSDMHDOHcmE.m

 

.pmmu GOHpmcflEHHomHQ may EH mccsow

Honucoo cam mofloomm nuOQ mo mmsm mo mxmmswm mmwuumflp ou mnwnpOE msmooswa.m can

 

WSHMHDUHGME mSUW%EOHmm MO mmflUCQHMH waOQWwH GMGE USN mmeOQme MO HwQESQ Qmwz

Mlm mHQMB

 

 

 

141
P.maniculatus mice was possibly related to the prior ex-
perience of some of them in the SCB, as well as to a genuine
species difference in excitability.

Number of First Responses Correct

 

The mean numbers of first responses to the "correct"
side of the SCB, i.e., to the squeaks for the squeak—control
groups and to the conspecific squeaks for the choice groups,
did not differ significantly when analysed with a Kruskal-
Wallis non—parametric analysis of variance (Table 5-3). I
Considering total initial responses for all the mice
in each group, all the P.maniculatus mice responded in all
trials. The two squeak-control groups (Eleucopus & P.manicu-
latus) responded non-randomly, that is, chose squeaks more
than noise, 79.2 & 78.2 percent of the time, respectively,
significant at 0.001. In contrast, the P.maniculatus choice
group did not respond in a manner distinguishable from ran-
dom, being correct only 50% of the time. The P.1eucopus
choice group differs from the random expected percentage
correct (50%) at the 0.05 level.

Latency to First Response

 

As in the Activation test this latency was measured
as being the time between the onset of the test sound and
the first recorded response of the mouse. These latencies
showed a heterogeneous variance which required the use of
the non-parametric Kruskal-Wallis one-way analysis of vari-
ance. The results showed no differences in mean latency to

respond (Table 5-3).

   

 

 

142

Latency to Correct Response

 

The time for a mouse to respond to the "correct"
stimulus did not differ significantly between the four
groups. Variances were large and again heterogeneous so
the Kruskal-Wallis test was used (Table 5-4).
Distance
Defined, as in the Activation test, as the distance
a mouse moved in either arena towards a stimulus sound,
this variable was measured in terms of scores of photocells .i“
triggered. A mouse that ran all the way to the speaker in L
each trial would accumulate a mean score of 3 while move—
ments into the tunnels only would accumulate a mean score
of 1. For additional comparisons between the two species
the differences between correct and incorrect scores for
each group were calculated (Table 5-4).

Distance Score to Correct Response

 

These scores were found to show heterogeneous vari-
ances requiring the Kruskal—Wallis one-way analysis of
variance. The means were not found to differ between the
groups (Table 5—4).

Distance Score to Incorrect Response

Since these data showed homogeneous variances a
parametric one—way analysis of variance was possible. This
test showed significant differences between the groups
(P < 0.05). Individual groups were then paired and tested
with the New Multiple Range test showing that the P.manicu-

latus squeak-control and choice groups differed at P < 0.01,

143

mo. m *
Ho. m *x
moo. m xxx

aanm.o me.H *aHh.o ¥««mm.H mHoom pomuuoo
IGH paw uowuuoo
comaumn mocmHoMMHQ

.m.c mm.N mm.N Nv.m oo.m memHm>mH GOHOQU

mo.o mo.H MH.H . Hb.H mm.o mmcommmu
pownnoocH mocmumHQ

.m.c oo.N vv.N Nw.N Hm.N mmcommmu
uoouuoo monopmHQ

m m NH NH c
m moHoco HOMHGOOIMmmswm moHoco HouuGOOIxmmSWm
msmoosmH.M .mSHMHsoHcmE.M

 

.HHSEHum uomuuoocH paw uomunoo Op pmHm>muu mmocmuch cmwzuwn

mwocmeMMHp mo cocooHMHcmHm OHMHSUHMO ow poms umoplu m.uchSHm .mocmHHm> mo mommHmcm
oHHumEmumm can AmHHHMBIHMMmSHMV OHHumEMMmmlsoc SHHK .mHmme>mH moHoao How can mmmcomm
Imu uowMHoocH 6cm pomuuoo m0 wmuoom wocmuch How cwusmfioo mQSOHm cmm3umn mmocmummeo
.mccSOm Honucoo can meommm buon mo wmsm m0 mxmmDWm mmmhuch ou mcHUGommmH mnmcuofi
msmooSwH.M cam mDHmHSUHGmE msomhfioumm mo mHmmHm>mH moHoco smog can mwuoom mocmume com:

 

Vim mHQMB

 

 

 

f—

144

with the mice in the choice groups going farther to the
incorrect sound (leucopus squeaks), than the squeak—control
group. The P.1eucopus groups did not differ on thismeasure.
But the P.maniculatus choice group differed from the P.
leucopus choice group at P < 0.05.

A Student's t-test on the differences of the paired
scores to the correct side and to the incorrect side for

each group showed that all mice went significantly further W
to the speaker producing the squeaks of their own species I

”vs—fur.

‘p—

(or any squeaks for P.maniculatus squeak-control group)
than to the opposite side.

Number of Choice Reversals

 

The variances in the mean number of choice reversals
were homogeneous so a parametric one-way analysis of vari-
ance was run. The differences in these numbers were not

found to be significant over all four groups (Table 5-5).
Discussion:

The results of this test confirm the findings in
the Activation test that deermouse mothers search for the
source of infant distress squeaks. Here, a second species,
P.1eucopus, showed increased activity and left the nest
when distress calls were played. These mice approached
mouse squeaks more frequently, and closer, than the "white"
noise. In the choice test, with recorded squeaks from a
6-day old P.maniculatus and a 7-day old P.1eucopus, the

P.maniculatus mothers responded randomly on all but one

 

 

145
variable. One measure, the comparison between the distance
score to the correct and incorrect stimuli, showed a sig-
nificant difference from random. This difference in scores
was smaller for the choice group than for the squeak-control
group but it was significant. The P.maniculatus mothers
went closer to the P.maniculatus squeaks than to the P.
leucopus squeaks when they had a choice (Figure 5-1).

The P.1eucopus mothers also showed a significant
difference in distances traveled to the conspecific versus 'l“
the non-specific sounds. In addition the P.1eucopus choice v
group showed a higher percentage of correct responses to
the conspecific sounds.

These data are interpreted to mean that while it
appears to be possible for deermice (at least of these two
species) to differentiate distress calls of pups of their
own species from others, the recognition is difficult and
the sounds are perceptibly similar. It may not be necessary
to differentiate the calls in the natural situation if the
mice live in different niches and rarely meet.

Bradshaw (1965) and Moore (1965) have shown that
adult mice of these species can differentiate conspecifics
from mice of other species on the basis of odor alone, so
it is probable that the adults could recognize pups of
their own species by odor. If the pups can be recognized
as appropriate to retrieve when the mother arrives, it may
not be necessary for her to be able to differentiate be-

fore she leaves her nest. It may be that even with

 

146

Figure 5—1

Legend

Distance score to sound sources in Activaticni and
Discrimination tests. Photocells recording mouse anztivity
were numbered 1, 2, or 3 depending on their distaxnze from
the nest/start box in the SCB. Mean scores reflrurted the
distance traveled by the mouse towards a particulax: sound
source as averaged for each mouse over four trials. Scores
towards the stimulus defined as "positive" above the abcissa,
those defined as "negative" below the abcissa. I"values fin
differences between "positive" and "negative" scuares printw

on top of each bar.

 

1117

 

      

«zoos—3 mane—5.... E ate—30m O>=DOOZ

2 23m 3555
3.0033 man—32.55.. -

m

030: 0.33 D

       
 

39o
268:0. m

a.»
unto—33:05 m

FIGURE 5-1

mucsom 3:33. 3 23m 0.2.239

   

 

148

sympatric species the situation rarely arises where the
distress cries of a pup of one species are overheard by a
mother of the other species. Differences in the squeak
recognizable by the mice may be due to physical differences
in the mouth and vocal cords and not the result of charac-
ter'divergence. The test described here merely demonstrates
a perceptible difference but cannot clarify the cause of the

difference.

Comparison of Results of Activation and Discrimination

Tests:

The results of these two tests can be compared in

terms of the responses of P.maniculatus mice to squeaks of

 

pups of their own and the other species and in terms of
differences in the behavioral responses of the adults of
both species to similar stimuli in similar situations.
Although the experimental groups were unequal in size, the
experimental apparatus and protocols were identical and so
I felt that non-parametric statistical comparisons of the
results were justified.

When the retrieving responses of the mice were con-
sidered within species but across experimental groups the
two species demonstrated similar trends. Both Pgmanicula-
tus and P.1eucopus mice chose squeaks of pups of their own
species over "white" noise. They made more correct choices
when they were offered squeaks versus the control sound than

when they had to choose between two squeaks. Distance scores

 

 

149

showed that both species went farthest towards the con-
specific squeaks when presented with the choice situation
(Figure 5-1).

§,maniculatus and P.1eucopus also demonstrated a
preference for conspecific squeaks in terms of total re-
sponses (Figure 5-2, Table 5—5). Again both species showed
strong preferences for squeaks of either type over the con-
trol sounds. Response latencies, however, were longer for

P.1eucopus in the choice trials than for P.maniculatus

 

(Figure 5-3). A certain portion of this difference can be

ascribed to the greater familiarity of the P.maniculatus

 

mice with the apparatus since their earlier latencies (in
the Activation test) were comparable to the P.1eucopus
latencies.

Choice reversals reflect the difficulty of the choice
between the squeaks of the two species, even for the more

experienced 33maniculatus. Reversals are significantly

 

greater for the squeak choice groups (Figure 5-4).
These comparisons support the conclusions already
reached from the Discrimination test results - i.e. that

while it may be possible for Peromyscus to differentiate

 

pups of closely related species on the basis of sound
alone, they do not always do it. The cues which make the
sounds acceptable to the females of the different species
may differ simply from anatomical differences due to physi-
cal divergences rather than from specific character di—

vergence of the sound characteristics. More widely

 

 

 

 

150

Figure 5—2

Legend

Mean numbers of responses to the sound sources,
averaged for each mouse over four trials in the Activa—
tion and Discrimination tests. Responses to "positive"
sounds above the abcissa, those to "negative" sounds be—

low abcissa. Random response levels marked with arrows

at the level of 50% total response for each side.

 

151

FIGURE 5'2

9.02.3 Snooze. m :—
mxuoaum m=.o.=o.coe.n. I

9.0.? m

030... 02:? D

m>....<wmz

m>_..._m0n.

$2033. 3......

do .3532

maneuaoa

n30.39233.5—

      
       

O.n

O.N

0..

0..

02¢.

O.n

 

 

152

Table 5-5

Comparison of the responses of P.maniculatus and
P.1eucopus mothers to squeaks of con—specific pups

in control situations (squeaks versus "white“ noise)
and in choice situations (con—specific squeaks versus
squeaks of non—specifics). Numbers of mice respond-
ing more than 50% of the time. »

gun—-9
..._F.r- -

 

con—specific con—specific ‘
squeaks versus squeaks versus
"white" noise non-specific squeaks
obs. exp. obs. exp.
P.maniculatus
n—14 12 7 n=12 2 - 6
P.1eucopus
n=8 6 4 =8 4 4

Chi—square 0.01 > P > 0.005

 

 

 

153

Figure 5-3

Legend

 

Mean response latencies in the Activation and
Discrimination tests. White bars are latencies to any'
(first) response. Dark bars are latencies for responses
to the "positive" (correct) sound stimulus. Mean values

for latencies in seconds marked at the top of each bar.

 

 

02.0030. m

32.0000: 32.00005
.2200 .8280
00.2.0 0.02.0» 00.2.0 .02...»

154

3.839. .2: D

FIGURE 5-3

02.2.09. .0230 m

mag—aoEoE m
$2030.30:

.2200 .2200

00.2.0 .0000» 30:0

 

seconds

 

   

155

Figure 5—4

Legend

Mean choice reversals in the Activation and Dis—

 

crimination tests. Bars indicate mean number of times
mice ran from one arena through the nest/start box and
entered the other arena. As the entire length of the SCB
could be traversed by a running mouse in less than a
second the distance of each entry into an arena is not
computed here. P values for significance of the differemms
between groups were P < 0.05 for the maniculatus squeak-
control,click—control and maniculatus—squeak/click choice

groups and P > 0.05 for the four leucoupus squeak groups.

 

156

0.00000 0300000
020.00.00.00 00.0: 2...: 0000000. m

 

 

 

 

 

 

 

 

 

 

0x000 0
9.00000 00.3000...
0.00800. m
0.00000
020.0020... .n.
4
—
5

0000000.m

FIGURE

 

00.0:0

  

0000000
00.0.0250...

00.0... 0......»

0x00000
0000000.m

 

0.00.3.0

0.0.0... 02...; 00.00 0:...»

II

  

 

 

Nu. .. I"
0000000
00.0.0255... I n
1 9.0....

..Q

In

to

030.00.006m

Number of reversals

 

 

157

divergent species would be expected to reject sounds of

the other species more conclusively.

 

 

CHAPTER 6

Complete and Attenuated Distress Cries

as Sound Targets: Localization Test.

Since both a squeaking pup and a searching mother
mouse expend energy and risk alerting predators, it is
important that the mother find the pup as fast as possi—
ble. Short, sharp, high—frequency sounds, such as these
mouse squeaks, are theoretically easy to localize with
binaural sound reception devices, i.e. a pair of ears
(Marler, 1967). Observations on wild mice indicate that
pups dropped by a mother fleeing a disturbed nest are
quickly located and retrieved. Obstacles and darkness
do not appear to affect the accuracy of the search.

I hypothesized that neither visual nor olfactory
cues are necessary for localization of a squeaking pup
but that sounds provide sufficient and essential informa-
tion for a directed search. Sound sources providing only
audible cues should be located as precisely as sources with
the additional visual and olfactory information from living
pups. In nature while odors may be limited in effective—
ness as signals at any distance,particularly downwind,
sounds should carry well in any direction from a point

source. Using any or all of the possible methods of

158

   

 

me--.’ —. __

159

localizing sounds (discussed in Chapter 3) a mouse should
be better able to follow a sound to its source than an
odor,even in dark or obstructed habitats where sight is
impossible.

Accordingly I designed a box which permitted alter—
native presentation of sounds devoid of accompanying visual
or olfactory cues (recorded squeaks) and sounds with all
the additional cues of a living pup except visible ones
(live pups behind a light-opaque, sound-transparent bar-
rier). The box was semicircular with alleys fanning out
from the center of the circle so that sOunds could be pre-
sented at eight positions equidistant from the subject.
The alleys helped to channel the sound and forcethe mouse
to chose a direction while still at a distance from the

sound source.
Subjects:

The test animals comprised one group of 42 lactating

P.maniculatus bairdi. Four of these animals were wild—

 

caught as adults, and 38 were Fl daughters of wild-caught
mice. The scores of the two groups did not differ sig-
nificantly on any measure so the results were lumped into
one group. All of the mice were nursing litters of 5—10
day old pups (P age of litter = 7.4 days). The F1 mice
had almost all been used in either the Activation or the
Discrimination tests or in both, but as their scores did

not differ significantly from the naive mice,experience

 

   

 

160

at retrieving in another apparatus did not seem to produce
inhibitions. As the point of the test was to test a maxi-
mum accuracy in locating a pup and not ability to retrieve,

experience was not judged to be a negative factor.
Equipment:

Test squeaks were recorded in the same manner as for
all the other tests, using the same equipment and they were
played back to the mice in the same way. The experimenter
stayed in the room with the mouse in order to visually ob—
serve her behavior. Tape recorder and amplifier remained
in an outer room and were operated by remote controL leav—
ing the Speaker and remote control switch as the only
equipment in the room. Illumination was supplied by one
lO-watt,red—light bulb hanging over the center of the cir-
cle formed by the Fan box. This provided just barely
enough light for the observer to see the movements of the
mice and there were no shadows formed by the sides of the
alleys. Despite the light and the presence of the ob-
server the mice left the start cage and readily searched
in the Fan box when there were squeaks.

The Fan box (figure 6—1) was designed to make each
sound source equidistant from the exit of the start cage.
The outer edge of the box circumscribed a portion of a
circle having an arc of 120° and radia which, if continued
into the center, formed angles of 15°. The Fan box was

divided into eight wedge—shaped compartments with the small

 

 

   

 

 

161

Figure 6—1

Legend

Fan Box: A mouse placed in the cage at the center
of the fan is exposed to sounds emanating from a speaker
(S) placed at the end of any alley (A). Perforated bar-
riers (B) prevent the mouse from seeing either the speaker
or a live pup placed at the end of the alley but do not
restrict odors and sounds. Walls of the alleys form
angles of 15° at the center. Complete outer walls are

36 in. long, and inner alley walls are 18 in. long.

162

.ENqu. I

z<.

 

uc<u

 

 

 

   

 

163

end of the wedge (4-2/3") open to the center and the large
end (9-1/3") at the edge of the circle. A rectangular
opening (6" X 5" high) cut into the plywood outer ends of
these wedges and covered with screen and black fabric,pro-
vided a sound-transparent shield for the speaker outside
the box. The plywood sides of the wedges extended into
the box radially for 18", half-way to the center. Each
wedge thus formed an alley down which the mouse could ap—
proach the source of the sound. Three inches from the
wide outer end of each wedge a perforated barrier (8-1/2"
X 2" high) prevented the mouse from seeing the pups used
as live sound sources. The outer walls of alleys number 1
and 8 continued to the center of the box and joined the
entrance to the start cage. This cage was a standard
colony cage modified so as to have a door (2" X 3" high)
which could be raised from above to let the subject into
the Fan box. Except for the start cage the entire Fan box

was painted flat black.
Procedure:

Each mouse was put into one of the modified cages
with her litter, food, water,and clean bedding, at least
48 hours prior to the test. On the day preceding the test
the mouse was allowed to explore the Fan box for 30 min—
utes. During this time her litter was randomly placed in
the box to encourage her to enter and retrieve them.

At the time of the test the mouse's cage was

 

 

 

164

positioned at the opening to the Fan box, her litter was re-
moved to another room, and she was allowed to calm down for
five minutes. Then the first stimulus was presented, either
a live pup of the mouse's own or the recorded squeaks of a
6-day old pup. The stimuli were initially presented from
a randomly chosen alley; additional presentations were made
in similar fashion except that no more than one presenta-
tion of each type was made from any alley. Alleys number 1 and
8 (at the far right and left) were never used, in order to avoid
a strong bias due to the tendency of the mice to follow
along the walls of the box.

Sounds were presented for 30 seconds before the
door of the start cage was raised letting the mouse into
the box. The sound continued for five minutes or until
the mouse had made three choices, whichever came first.
A choice was defined as an excurSion past the halfway
mark in any alley. After three choices, correct or in-
correct, the sound was turned off and the mouse returned
to the cage. The next sound was presented after a one min-
ute waiting period. If a mouse did not respond in five
minutes it was scored as zero response and 5 minute latency.
Mice with three consecutive zero scores were replaced as
subjects.

As stimuli, recorded squeaks alternated with live
pups. Each subject was given six trials, three with each
stimulus. As each thus served as her own control, her

reSponses to live pups and recorded squeaks could be

 

165

compared. Almost all of the mice retrieved the live pups
and these pups were allowed to remain in the nests during
subsequent trials with no apparent effect on the mother's
behavior. Unused pups were returned to their mothers

at the end of the tests.«

In this experiment the independent variable was
the nature of the stimulus, whether live pup or recorded
sound. The dependent variables were the correctness of
the choices of the mice, their response latency, and the
numbers of responses. Since each mouse served as her own
control, there were, in effect, two paired experimental
groups which could be compared in terms of responses to

live pups or recordings.
Results:

All but 8 of the mice responded in all trials.
Four of the mice which refused did so in 2 consecutive
trials; four mice refused to respond in only one trial.
These null trials (no reSponse) were equally divided be-
tween live stimuli and recordings. Whether the mouse's
first trial stimulus was live or recorded made no sig-
nificant difference in the number of reSponses. Mice
which refused to enter the box often spent the trial time
flipping in their cages, a behavior which was interpreted
as an indication of a conflict between fear of the appara—
tus and the impulse to search for the squeaking pup. When

the door was lowered and the sound turned off at the end

 

 

166

of the trial, the mouse always stopped flipping.

Latencies

 

Latency to enter the Fan box after the door was
opened and the latency to arrive (touch the barrier) at
the end of the alley were separately measured. Mean la-
tency to enter for recorded sound was 35.78 seconds and
38.51 seconds for live pups (Table 6-1). Since each mouse
served as her own control the Wilcoxon non-parametric test
was run on paired data but the differences in entrance la—
tencies for the two stimuli were not found to be signifi-
cant.

On the average it took a mouse 15.15 seconds for a
mouse to reach the end of an alley after it entered the
box on hearing the cry of the live pup,and 25.38 seconds
when it heard the recorded squeaks. Arrival latencies,
calculated as the time elapsed from the opening of the
door until the mouse arrived at the distal end of the
alley, differed between the stimuli (Wilcoxon, P = 0.05,
Table 6-1). Mean arrival latencies were 53.66 seconds to
live and 61.16 seconds to recorded sounds.

Choices

Predictions of the behavior of the test animals in
the box were based on the number of alleys and the number
of positions where the mouse could end up if it behaved in
an undirected manner. Although sounds were only presented
in 6 alleys the mice were free to enter all 8 alleys. Any

mouse responding to the stimulus might be correct, or wrong

 

 

 

167

endow How m>HHHm 0H HmOGOH mpcoomm HH.0H mocmummmHU x

mo.o

Amocoomm.
00.m H mH.Hm mN.m H mm.mm N0 m>HHH4 OH hocmumq
m>HH now moved 0H HmmGOH mpcoowm m0.N mostmHHHp m
.mpcoomm.
Nm.m H m0.mm 0m.m H Hm.mm N0 Hmpcm 0H wocmHMH
mmm mmm
UGDOm pmcnoomn mesa m>HH c

 

mousom mcp.pm

 

.msommEoumm HcmmcH Ho moHHo mmmHHmHU ompuoomu Ho m>HH Ho
mcH>HHHm cam xon cam 03H mcHHchm mHmSHOE mSHMHSOHGME wsomhfiouwm Ho mmHocmHMH

 

Him OHQMB

   

 

 

- . ... . . __ . _ l... _ r . 7* '7 "" ‘ '- "_ .'___""—‘—*"""“‘

168

("off") by one alley (15°), 2 alleys (30°), 3 alleys (45°),
4 alleys (60°), 5 alleys (75°) or 6 alleys (90°). Predic-
tions were complicated by the fact that since the box was
only semi-circular, not all choices were open for all of
the 6 possible positions of the speaker. If the stimulus
came from alley number 4, for example, then a mouse could
only be off by 4 alleys on the right and 3 on the left.

Probabilities of a mouse being correct or wrong by
a particular number of alleys (degrees) were calculated
(Table 6-2). Expected numbers of responses and percent—
ages were compared with observed responses with the Chi~
square test. Predictions of random behavior, unrelated
to the source of the sound, produced the ratios in the
expected columns. The observed reSponses were found to
be highly significantly different and overwhelmingly
oriented to the source of the sound (Table 6-2).

If only the first response of the three permitted
in any one trial was considered there was a significant
trend to more correct orientations when the stimulus was
the live pups than when it was recorded squeaks. How—
ever if the best choice of the three was considered there
was no significant difference between responses to recorded
and live squeaks.

First choices are schematically represented in
Figure 6—2, where the arrows correspond to the percent-
ages of the total responses to correct and other alleys

when the stimulus was in various positions. Solid arrows

   

 

 

169

Hoo.o v v m Hoo.o v v m mumswmlHfio Umpoomxm .m> pm>umwno
m.0 o.N m.0 o.N o o mOHOQO 0c
H.N m.o o o o.mN om.oH wHOE HO 0
0.H 0.o m.N o.H 0.mH Ho.0 HMO m
m.0 o.N 0.N o.H m.0N 00.m HHO N
m.mH 0.0 m.m 0.m o.mN om.0H HMO H
m.mm 0.mN o.Hm 0.0m m.NH mN.m Homunoo
0 m 0 m 0 m muHoro

Unsom Umpuoomu pm>Hmeo pcsom 0>HH pm>nmmno UmHommxo

onoao 0000

Hoo.o A m A Ho.o Hoo.o v v m onwdelHnu Umuomaxm .m> ©0>Hmmno
m.0 o.N m.0 o.N o o mOHOSO 0c
m.HH 0.0 m.0 m.m o.mN m.OH QHOE Ho 0
m.0 m.m H.0 o.m 0.mH Ho.0 HHO m
o.mH 0.0 m.0 m.m m.om 00.m HMO N
m.mN m.HH 0.0m 0.m o.mN om.OH HHO H
m.mm o.0H 0.0m m.HN m.NH mN.m Homnuoo
w m w m w x mOHOQU

pcsom Uwpuoomu ©m>ummno OGSOm w>HH pm>Hmeo pmuommxm

oOHZ HO mHmQESZ mmmuo>< N0 n z mOHOQO HmHHm

 

.mmeHm HcmOMHOM OH Hmmmu ..OH0 .HHO 030 .HHO woo

.monsOm OQSOm 03H mchHMHcOO mmHHm 03H mo oOHOQO m on muomma GOHOcO uomnuoo .mHMHHH mwucp
Hm>o pmmcmnnm HmHHH m cH moncH mo wOHO£O 0009 cam HmHHH 0 OH mOHono HmHHm .HmMH mumsww
IHco AHHz ©0>Hmwno OH Umnmmfioo 00Hoco pmpowmxm Eoccmm .mmHuo mmmHHch HQMHQH pmpuoomn HO
0>HH Ho mOHsom HOH mGHQOHmmm .xon cmm cH meSHOE mSHMHSOHcmE msommEOHmm Ho mmOHOSO mmHHd

 

mlm OHQMB

 

 

 

170
represent responses to live squeaks and broken arrows re-
sponses to recorded squeaks. The diagrams dramatize the
accuracy with which the mice directed themselves to the
sound source. Had best choice of three been diagramed
the picture would have been even more clear cut.

Although the design of the experiment made it rare
for a stimulus to be presented twice for the same alley,
in one test, on those occasions when it did happen there
was no observational evidence that the mice returned to
the site of a successful retrieve rather than searching
for the source of the sound. There was, however, a tend-
ency for the mice to follow along the walls of the box and
then cross the opening to the alleys, listening at each
one. At the rapid rate at which they moved this behavior
often caused them to mistakenly enter the alley preceding
the correct one. Those mice which were correct most con-
sistently, tended instead to linger at the entrance to
the box, turning their ears until they appeared to get
some sort of fix on the sound and then went directly to

the end of the apprOpriate alley.
Discussion:

The results of the analyses of the choices of the
mice responding to both live and recorded squeaks indicate
that the sound alone is sufficient to provide cues for ac-
curate location at the distance of 3 feet within 15°, in

this apparatus. While first choices were better with

 

 

 

 

 

171

Figure 6-2

Legend

Diagrams of relative percentages of responses in

correct and incorrect directions by Peromyscus maniculahm

 

mothers searching for sounds of live and recorded distnws
squeaks. Live squeaks as stimuli — sold arrows; recordai
squeaks — broken arrows. Arrows represent percentagesof
responses, not relative distance since all responses re-
corded went the minimum distance. S marks placement of
the sound source. Multiple diagrams are presented as um
positions of the sound source are logically independentam
are paired on the page so as to show right and left sides
of the Fan box. Notice increased edge effect, especiafly
with recorded sound when the source is in alleys 2 and 7
(diagrams e & f). Long arrows indicate high percentages,
short arrows lower percentages, and alleys with few orsing1e

responses are not indicated. Only the highest, not alltm

percentages are printed.

 

 

 

 

172

IIGIIIIE 6-2

 

   

 

173

live pups as the stimulus I propose that this was not due
to non-sonic cues emanating from them, but rather from the
fact that the sound was not attenuated by the recording
process and the volume was more natural, although by no
means constant. In addition the longer latency to enter
for mice responding to live squeaks correlates with the
observation that the mice lingered more often at the door-
way when the stimulus was live than when it was recorded.
Those that lingered were more accurate at finding the cor-
rect alley. When the stimulus was recorded squeaks,mice
tended to dash out into the Fan box and run along the walls
in a less directed, less efficient search. This behavior
is reflected in the longer arrival latencies when they
responded to the recorded sounds.

Certainly,when best choices are considered and even
when only first choices are counted,these deermice were
all highly successful in finding the "pups." When pups
were actually present they could get them back to the
nest, and this meant carrying them over the 2—1/2" bar-
rier, in less than a minute from this distance. Even in
a dim and obstructed natural habitat, especially one
familiar to a mouse, squeaking pups could easily be lo-

cated and retrieved in a minimal time.

 

 

 

CHAPTER 7

Ultrasonic Distress Cries and Maternal Response:

A Simple Communication Paradigm

The assumptions on which this study rested were

1) that infant deermice (Peromyscus) emit sounds when re-

 

moved from their nests and deprived of maternal warmth and
tactile stimulation; 2) that these sounds are predominately
composed of frequencies above the limits of human hearing
(i.e. above 20 kHz, ultrasonic) but are susceptible to
electronic analysis; and 3) that these cries play some
stimulatory role in the retrieving behavior of lactating
female deermice. I suspected, in addition, that the cries
might vary according to phylogenetic relationships and
that they would show.ontogenetic changes. Reviews of the
pertinent literature supported the primary assumptions and
confirmed the possible existence of phylogenetic variety

and ontogenetical change. This study showed that Peromys-

 

cps pups do emit ultrasonic distress cries when subjected
to the mild cold stress Of being out of their nests and
that these sounds are sufficient to alert and direct mother
mice to retrieve them. Phylogenetic and ontogenetic dif—
ferences in the sounds were demonstrated.

Following the model presented in Figure 7-1, the

174

 

 

 

 

 

Figure 7—1

Legend

Model of a simple communication system, showing pafir
ways of information transfer. In the paradigm discussed
here the information source becomes the mouse pup, its
vocal cords and mouth are the transmitter, the signal is a
sound with a message on the state of the pup coded in the
physical characteristics Of the signal. The sound heard
(received signal) by the mother mouse (receiver) may have
been distorted by extraneous noises. The arrival of the
message at its destination is demonstrated by the behavior

of the receiver.

 

176

FIGURE 7 - I

 

 

zo_h(z_hmwo

 

 

 

lorry-Ohm:—
02( was:

 

 

 

 

um<mmm8

au>.u0m¢

 

003.0000

 

J<zgm

om>_u0u¢

 

szz<=o

 

 

 

cwhki
amz<mb

 

 

mezzo»

 

 

4(zo.m

ou...tlmz<¢k

20:32:00.2.

 

 

”@(mmul

omhh.lmz<¢h

 

   

 

-.7.fi_—___————.—__-‘ .~w—w—————---—-.— _.. v. H .-

177

information transfer between pup and mother would have the
pup as the information source, the emitter, and the mother
as the receiver. The pup transmits a message coded in an
age-specific, species—specific signal, using its vocal cords
and mouth as the transmitter. The signal is transmitted in
the sound channel, where other sounds of laboratory or
woodland may intrude as noise and distortion. If the pup
repeats the signal enough (redundancy), the information
will be received by the relevant receiver, the mother.

When she acts on the information received then we know the
message has been perceived.

The results of the experiments described here, now
provide us with sufficient information to consider the
functional aspects of the source, the signal, and the re—
ceiver, and place them in the communication system. We
can fit the pieces together in this auditory system as has
been done before for visual communication systems such as

the classic stickleback mating sequence.

 

Why does the pup signal?

An infant rodent pup often emits ultrasonic cries
when Stressed by cold temperature or by being touched,
bumped, or handled. The recent discussions of the physiolo-
gical trigger for the onset of squeaking in rodent pups
leave no doubt of the importance of low temperatures.

Hart and King (1966) suggested a connection between dis-

tress cries in Peromyscus and temperature, and Smith (1972)

 

 

 

 

178

presents confirming data on the close relationship between
these calls and temperature. Okon (l970a, b, 1971) clear-
1y relates distress calls in white mice, white rats, and
hamsters with developing homeothermy. As pups of these
species develop temperature regulation they cease squeak-
ing. With the exception of the Hart and King study, these
data were published after the completion of the experiments
discussed here, but the results are in complete agreement.

The temperature in my recordingapparatuswas main-
tained at about 18° C (65° F), similar to the 20° C used
by Okon, and thus constituted a condition of cold stress.
Smith (1972) considers 35° to be nest temperature in Berg:
myscus, so my recordingapparatusshould have been suffi-
ciently cold to induce the spontaneous production of dis-
tress cries to cold. Once removed from the nest the pups
lay quietly in the recording funnel and were neithertouched
nor could they fall over during the recording session.
Pain squeaks deliberately elicited by pinching tails were
analysed separately. Thus handling can be discounted as
a stimulus for the squeaks I recorded, even though I did
not suspect the existence of separate "handling" and "cold
stress" distress cries, at the time. As I was primarily
concerned with the physical characteristics of the sounds
I did not vary the temperature.

Preliminary observations convinced me that the rate

of calling was highly variable and also dependent on the

 

 

 

 

 

179

:motivation of the pup, so the number of calls in a pre-
scribed time were not measured. I predicted that the
number of squeaks in a set were probably a characteristic
related to phylogenetic differences, as data in Hart and
King suggested (1966). Subsequent studies (Okon, 1970a;
Smith, 1972) confirmed that squeak rate was closely related
to temperature and the homeothermic ability of the pup.
Similarly, intensity was not measured because the
movements of the mouse's head were impossible to control
without restraint, which wOuld have introduced a handling
factor, and I also felt that the intensity of squeaking was
a motivational factor. Okon (l970b) and Smith (1972) also
found intensity (SPL) correlated with temperature in 3
general age groups. Up to 5-6 days, squeaks of albino mice
are generally weak, stronger, then, until 12-13 days and

thereafter weak again.

What is the signal?

 

The squeaks produced by the mice I recorded are
generally lOng, high sounds with multiple harmonics and a
small amount of frequency and intensity modulation. The
greatest energy is usually found in the second or fourth
harmonic and the fundamental may be unheard. There is a
sharp onset and a sharp Offset, usually with a steady tone
in the middle. Observations suggest, although definitive
tests have not been made, that the sharpness of the onset

and offset are achieved by the pup Opening its mouth after

   

 

180

the start of the sound production and closing it again
before the end of the squeak.

Two types of distress squeak became apparent in
the phylogenetic and ontogenetic comparisons. Examples of
both types are sufficiently common and clear to show that
they are not experimental artifacts. Type I squeaks show
one strong band of energy, usually the second harmonic,
Often with few, weak, or no other harmonics visible. Type
II squeaks are complications of the same pattern, having
equally strong energy components at a frequency and its
harmonic, usually the second and fourth harmonics. It
should be mentioned that these 2 types Of squeak have no
relationship to the 2 types of squeaks and chits described
for adult Peromyscus by Houseknecht (1968). Rather, I in-
terpret them as variations on a theme by different taxa,
perhaps occasioned by differentially thick vocal cords.
There may be a correlation between the size of a mouse and
the tendency to produce Type II squeaks, as I recorded these
squeaks mostly from the larger mice, P.californicus, P.di£:
ficilis and P.m.gracilis. The fact that these Type II
squeaks have 2 strong energy bands must give them a percep-
tibly different sound quality from the Type I squeaks with
a single band.

Published frequency averages for these species which
do not take the entire spectrum into consideration give an

unrealistic measure of the sounds. Species means for

 

 

 

181

frequency averaged over all ages unrealistically simplify
the sounds also. All of the groups for which I analyzed

the sounds showed 3 age divisions for frequency modulation
patterns, similar to the 3 age divisions later found for
rate and intensity (Okon, 1970a, b; Smith, 1972). It is

not clear, however, whether the frequency periods correspond
to the development of homeothermy, as the rate and intensity
changes apparently do (Okon, 1970a, b). Pinnae in several

Of the taxa, P.polionotus, P.m.gracilis, and _.m,bairdi,

 

 

become unsealed from the head and erect at about the end

of the "early" period, day 3. Ears and eyes generally Open
by the end of the "middle" period, about day 12. The rais-
ing Of the pinnae and particularly the Opening of the ears
may provide the squeaking pup with some feedback on the
sounds it is making and provide information on-the basis of
which the pup can modify the sounds. Whether the pup does
use what it hears to modify its squeaks cannot be decided

DOW .

What is the-message?

 

Situation:

If Smith (1972) is correct in distinguishing the
"isolation" cry from the "handling" cry on the basis of
frequency, rate and intensity as well as situation, then the
calls discussed here, from pups stressed by cold, could
carry very specific situational information. Increasingly

this call appears to be emitted solely by pups in the cold,

   

 

182
either out of the nest or alone in it.
Species:

My data show clear differences in physical charac-
teristics between species and even subspecies. All of the
taxa I recorded differed from each of the others on at least
one parameter (Table 7-1). The highest mean basic frequen-

cies were found in P.1eucopus and P.polionotus, and these

 

 

mice always produced Type I squeaks. In contrast. 2. cali—

fornicus and P.m.gracilis almost always produced Type II

 

 

squeaks with strong frequency components around 17—19 kHz.
Frequency patterns, as predicted, were most similar in the
most closely related mice, in general. Temporal parameters
do not show the same clear relationships. Temporal patterns
may shift more easily with physical changes in activity and
body type associated with divergence in habitat usage in
allopatric groups. The physical differences existing be-
tween the squeaks of mice of different species and even sub-
species could provide sufficient information for species
identification.

Age:

Squeaks become shorter and less frequent as pups
mature, harmonics fade out and pattern variations are reduced.
It is important to note that temporal and frequency charac-
teristics do not change continuously but rather in bursts
at about day 3 and days 10—12. Some gradual trends are

visible in set duration and the number of squeaks in a set

   

 

183

.05000000 0005 0:0 000 HH 0000 0:0 :msonun 0000000 HH 0&00 050 H 0000 0000 00000000 HwnHmbnmnm .H
00:0HommH0 ucmuHchme o: .0.:
00.0 0 .0000000000 00000000000 0.000; ..

00.0 0 .0050H0mmH0 ucmonchHm 0

 

 

 

 

 

0500 00 00 .0.5 0 0:50Mu000 m .mmmmmmmwnm\0suocoHHomnm
H0HHH0 00 0 0 0 00000-00H00d0 < 0SHOcoHHomam\MHHHmmmwumum
000000 . .0.c . . .. 00:00-000 <\0 mmmmmmMHnmxmeHmmmwam m

nu H0MHH0 00 00 4 .00: 00000-00H0000 < 0OpocoHHomnM\Hmmmmmnmnm
H0030 .. .0 . c .. .. 00000000 <\ 0 a .mHmmHlnaM
.H0500 .0.5. H 0 00 00H0000 m .mwwmmmmwsaxmxmmMHMMdmnm
H00 H00
0000 005030000 00cou 50H00000 coHumhsm 0000000 0H0050H00H0m .00500 0x00
#00500 0H000 uH0>o 000000 000 000502 pmpmoHu

 

OHHpmmoHHw u < .0H0000500 u m .000005OH00 00 0x00 0 5H 0000050500 M00000 000000H0 mo 0500H900500 00HH00

Hu0 0Hnmk

 

 

184

(Figures 3-15, 3-16). No particular parameter varies con-
sistenly across all ages or all species. It seems likely,
therefore, that the listening mouse may recognize a familiar
sound Gestalt, instead of a particular frequency or number
of squeaks. Combinations of harmonics and energy invest-
ments in different frequency bands would make perceptibly
different pitches. Repetition of sets may simply add re—
dundancy, insurance against information loss due to noise or
distortion.

Localization:

The cues necessary for accurate localization of
sounds are all present in these distress squeaks. Marler
(1961, 1967) considers that sounds can be localized by means
of differences in intensity, phase, and time of arrival of
the sound at paired sound receptors, ears. Localization
by means of intensity differences is most effective when
the sound shadow of the head prevents sound from reaching
both ears simultaneously, which occurs when the wavelength
of the sound is shorter than the width of the head. Data in
Chapter 3 (Table 3—5) suggest that the mice in this study
would only be able to use this method if fundamentals were
heard. However, localization by means of phase differentials
is best when the sound frequency is longer in wavelength
than the width of the head. Most of the basic frequencies
measured in this study are longer in wavelength than the
distance between the ears of their producers so that phase—

differential localization is probable. Sharp onsets and

   

 

 

185
offsets and transients are effective in localization by
means of time differences. These squeaks are composed of
many discontinuities and the short, sharp squeaks are re-
peated rapidly in sets, making localization by means of
time also probable.

In contrast, the pain squeaks analyzed had gradual
onsets and offsets, many blurred frequencies and extremely
fast repetition. These characteristics suggest to me that
there is no need for the pain squeaks to be easily localized
and may even be some advantage in disguising their source.

The analyses indicate that the distress squeak of
infant rodents is far more complex, physically, and capable
of carrying more information, than has previously been sup-
posed. The squeaks are characteristically easy to locate
and significantly different from species to species and for
major groups of ages (or developmental stages). It is not
even unreasonable to suspect a certain amount of individual
variation, given the high variance noted within species,
and it might be possible to demonstrate individual recog-
nition on the basis of sound. These squeaks are undoubtedly

the "audible" cries described in Smith (1972).

Noise and Distortion:

 

High frequency sounds are an excellent medium for
information transfer in a dark, obscured habitat. They re-
quire no other energy for transmission than that invested

in them at emission, they can go around obstacles, and do

 

 

186

not linger to give directions to predators. Short, sharp,
high sounds probably encounter few other natural noises
which would interfere with them and within a very short
time period can transmit a large amount of information.
Other receivers besides the relevant one, the mother
mouse, can of course, pick up the communication, but that
is no worse a problem with sound than with any other chan-
nel. In this case, some predators cannot receive the com-
munication; owls do not appear to hear sounds this high,

and snakes cannot hear at all.

What does the receiver do?

 

Listening mother Peromyscus are stimulated to emerge

 

from secure nests, search for, and retrieve calling pups.
The results of the Activity and Discrimination tests con-
firm studies with other rodents and establish the importance
and efficacy of these sounds in initiating retrieving be-
havior in at least 2 species of deermice. Mother g.manicu-

latus bairdi and g.leucopus overwhelmingly responded to re-

 

 

corded sounds of calling pups of both species. Very few
responses were given to control sounds and no responses were
observed when there were no squeaks.

These mice also responded in a manner suggesting
discriminatory abilities. P.m3bairdi showed a preference

on 1 measure while P.1eucopus discriminated on 2 measures.

 

Conspecific identification would have been expected to have

been even more clear had Species with more strongly divergent

 

 

187

sounds been used. 3.1eucopus and §,maniculatus were chosen

 

in the first place for this test because they are essential-
ly allopatric in Michigan. If there is any adaptive value
in a mouse being able to differentiate pups of its own
species by sound alone, it was expected to show moSt strong-
ly where chance encounters between species are possible,
with allopatric species. The results suggest that the
physical differences in the sounds may be discernable to
relevant receivers, mother mice and that they may act pre-
ferentially.

Location of the source of distress squeaks by re-
trieving mothers was immediate and highly accurate (Chapter
6). Squeaks were located, as well as live pups which pre-
sented the additional cues of odor and visible form. Re-
sults are conclusive enough in this test to state that
sound alone is a sufficient cue for localization. Odor may
be utilized on contact, for family identification, but this
suggestion has not been explored.

The equipment used in presenting the recorded squeaks
as stimuli attenuated the very high frequencies and undoubt-
edly introduced additional noise. It also presented the
sounds at an unnatural sound pressure level. Nevertheless
the mice emerged to search and found the sound sources with
little difference in performance from the trials where live
pups were used. This indicates that the frequencies which

carry the location information, at least, are those middle

   

 

188
ones (20-40 kHz) which have the greatest energy in the
squeak.

While a specific frequency may carry the location
information to the listening mother, the combination of all
the other characteristics of the sound, its Gestalt, prob-
ably informs her of the age and species of the caller. It
is either recognizable as a familiar sound (conspecific) or
tone variations which may be heard constantly from her own
litter. As the pups in the litter develop, their sounds
change, but as the mother must hear them constantly her
sound search image, or recognition image, must change also.
As the mother's physiological maternal motivations diminish,
the pups, developing their own temperature regulation and
abilities to move, squeak less and less loudly. Consequent-

ly retrieving responses diminish.

The meaning of the message?

 

Meaning, that is, the information in the context in
which it is emitted and received, can now be better under-
stood (Smith, 1965). Figure 7-2 is a diagram of the infor-
mation system of Figure 7—1 with the behavior of the mice
superimposed on it. The information of situation (isolated
pup), species (conspecific), age (same as the listener's
pups), and location, is contained in the squeak that the
female hears. When she emerges and retrieves the pup, the
information transfer is complete, the message has been re-

ceived, the meaning comprehended.

—--.-r_v .

 

189

Figure 7—2

Legend

Representation of distress squeak communication systenn
After nest disturbance female mouse may "lose" pup exnd
then must return, guided by its squeaks to retreive :Lt

carrying it to the nest, where the squeaks stop.

 

190

00:0.

0.

0:300 2 020.

g

00.00300 080:

FIGURE 7-2

000: 30c 00 003:0

to m
1/

000: 00 t 00:08

> 000.000.0000 \ a: 0:0 00.20%

0mm @

      
  

\...0x00:00 ./

a /
\\ xv.

to 000.6 030

>\/

\

0000 05000.6

000: .w
000.20% 000:

SUMMARY

Distress vocalizations were recorded from deermouse
pups of four taxa (Peromyscus maniculatus bairdi, E,m.

gracilis,.§.leucopus, and g.polionotus) for each day from

 

 

 

birth until eye opening. Recordings were also made of four

other taxa (E.californicus, §.m.nebracensis, 23melanophrys,

 

 

and §.difficilis). The vocalizations were analyzed, by

 

means of magnetic tapes, oscillograms, sound spectrograms,
and sonagrams, for frequency characteristics, numbers of
squeaks in a set, set duration, and squeak duration.

It was hypothesized that l) the frequency and tem-
poral patterns of the squeaks carry information on the
species, age and location of the pup; 2) the characteristics
of the squeaks change with age; 3) the squeaks serve as a
stimulus for maternal retrieving behavior; 4) they also
carry sufficient information to enable retrieving mice to
identify species; and 5) the source of these squeaks can
accurately be located by the mother mouse.

The results of sound analyses and behavioral tests
confirmed these hypotheses, in particular as follows:

1. Temporal parameters; Peromyscus distress squeaks

 

are generally short, sharp sounds uttered in groups. Squeak

duration, number of squeaks in a set, and set duration were

191

 

 

 

 

192
characteristically constant within species but differed
significantly between species. All of the taxa differed from
the others on at least one of these parameters. Numbers of

squeaks per set ranged from 3.15 (3.1eucgpus) to 4.60 (g,

 

 

melanophrys) squeak duration ranged from 0.13 (g.poliono-

 

tus) to 0.18 sec. (g.m,bairdi and g.ca1ifornicus); set dura-

tion ranged from 0.81 sec. (§,m.nebracensis) to 1.48 sec.

 

(g.californicus).

 

2. Frequency parameters; Two frequency patterns were
found, pure tone squeaks with no visible harmonics (Type I)
and squeaks with 2 main bands of sound.energy and much in-
tensity modulation (Type II). The first type were found in

3.1eucopus and £.polionotus and the second mostly in g.

 

 

difficilis, and §,melanophrys had similar mean basic fre-

 

 

quencies (21-24 kHz). while E3polionotus and §.1eucopus had

 

 

higher mean basic frequencies (32-33 kHz) and E.californicus

 

had a low frequency of 18 kHz.

3. Phylogenetic differences; Frequency characteristics
were most similar within species. Temporal parameters varied
even between subspecies. All species differed on some fre-
quency pattern.

4. Ontogenetic changes; Frequencies tended to rise
with age and older pups modulate frequency and intensity
less than younger ones. Harmonics gradually disappear with
age. Numbers of squeaks per set and set duration decreased
and squeak duration increased, peaked and then dropped with

age.

 

193

5. Recorded squeaks activated lactating §,manicu1a-

 

tus mothers to leave nests and search for the source of the
sound. The mice preferred the squeaks to control sounds.

6. Lactating Efmaniculatus and 3.1eucopus mice

 

slightly preferred recorded squeaks of conspecific pups to
those of other species.

7. Recorded squeaks alone provided sufficient infor-
mation for accurate location of sound sources by lactating
§,manicu1atus mice. Recorded squeaks were located as ac-
curately as squeaks of live pups.

The results of these tests confirm the popular in-
tuition that the distress cry on an infant mouse carries
information to properly primed adults which draws their at-
tention, stimulates them to search for the source of the
sound, and enables them to locate it. In addition, the
squeaks carry information on species identity, age of the

pup and cues for accurate location.

LITERATURE CITED

   

 

LITERATURE CITED

Allin, J. T. and Banks, E. M. 1970. Effects of temperature
on ultrasound production by infant albino rats.
Devel. Psychobiol., 4 (2):149-56.

Allin, J. T. and Banks, E. M. 1972. Functional aspects of
ultrasound production by infant albino rats; Rattus
norvegicus. Anim. Behav., 20(1):l75-85.

 

Anderson, J. W. 1954. The production of ultrasonic sounds
by laboratory rats and other mammals. Science, 119:
808.

Armitage, K. B. 1962. Social behavior of a colony of the
yellow-bellied marmot (Marmota flaviventris). Anim.
Behav., 10(3-4):3l9-31.

 

Balph, D. M. and Balph, D. F. 1966. Sound communication of
Uinta ground squirrels, Citellus armatus. J. Mammal.,
47(3):440—50.

 

Barfield, R. J. and Geyer, L. A. 1972. Sexual behavior:
ultrasonic postejaculatory song of the male rat.
Science, 176:1349—50.

Barnett, S. A. and Evans, C. S. 1965. Questions on the
social dynamics of rodents. in Social Organization

of Animal Communities. Symp. 2001. Soc. Lond. 14,
Academic Press.

Beach, F. and Jaynes, J. 1956a. Studies of maternal re-
trieving in rats 1: Recognition of young. J. Mammal.,
37(2):177—80.

Beach, F. and Jaynes, J. 1956b. Studies on maternal re-
trieving in rats II: Effects of practice and previous
parturitions. Amer. Nat., 90:103-109.

Beach, F. and Jaynes, J. 1956c. Studies of maternal retriev-
ing in rats III: Sensory cues involved in lactating
female's response to her young. Behaviour, 10:104—25.

Bell, R. W., Nitschke, W., Gorry, T. H. and Zachman, T. H.
1971. Infantile stimulation and ultrasonic signal-
ling: A possible mediator of early handling pheno-
mena. Devel. Psychobiol., 4(2) 181-91.

194

Elij

 

 

195

Beniest-Noirot, E. 1958. Analyse du comportement dit mater-
nel chez 1a souris. Monog. Franc. de Psych. #1.
Paris: Centre Nation. Recherche Sci.

Berlin, C. I. 1963. Hearing in mice via GSR audiometry. J.
Speech & Hear. Res., 6:359-368.

Blair, W. F. 1953. Experimental evidence of species dis-

crimination in the sympatric species Peromyscus truei
and g. nasutus. Amer..Nat., 87(3):103-105.

Blair, W. F. and Howard, W. E. 1944. Experimental evidence
of sexual isolation between three forms of mice of
the cenospecies Peromyscus maniculatus. Cont. Lab.
Vert. Biol., Univ. Mich., 26:1-19.

Bolles, R. C. and Woods, P. J. 1964. The ontogeny of be-
haviour in the albino rat. Anim Behav., 12(4):427-41.

Bradshaw, W. N. 1965. Species discrimination in the Peromys-
cus leucopus group of mice. Texas J. Sci., 27(3):

278-93.

 

Brooks, R. J. and Banks, E. M. 1973. Behavioral biology
of the collared lemming [Dicrostonyx groenlandicus
(Trai11)]: An analysis of acoustic communication.
Anim. Behav. Monog., 6.

Carlier, C. and Noirot, E. 1965. Effects of previous ex-
perience on maternal retrieving by rats. Anim. Be-
hav., 13(4): 423-6.

Denes, P. 1961. The reliability of Preyer's ear reflex in
the mouse (Mus musculus). Naturwiss., 48:82.

 

Dice, L. R. 1932. The songs of mice. J. Mammal., 13(3):
187—196.

Dice, L. R. and Barto, E. 1952. Ability of mice of the

genus Peromyscus to hear ultrasonic sounds. Science,
116:110-11.

 

Eisenberg, J. F. 1962. Studies on the behavior of Peromys-
cus maniculatus gambelii and Perom scus californicus
parasiticus. Behaviour, 19(3):l77-207.

 

 

 

Fisher, A. 1969. Sing, Little Mouse. New York: T. Y.
Crowell Co.

Fitch, H. S. 1948. Ecology of the California ground squirrel
on grazing lands. Amer. Midl. Nat., 39(3) 513-596.

196

Gandelman, R., Paschke, R. E., Zarrow, M. X. and Denenberg,
V. H. 1970. Care of young under communal conditions
in the mouse (Mus musculus). Devel. Psychobiol.,
3(4); 245-50.

 

Gandelman, R., Zarrow, M. X. and Denenberg, V. H. 1970.
Maternal behavior: differences between mother and
virgin mice as a function of the testing procedure.
Devel. Psychobiol., 3(3):207-14.

Gould, E., Negus, N. G. and Novick, A. 1964. Evidence for
echolocation in shrews. J. Exp. Zool., 156(1):19-
38.

Gourevitch, G. and Hack, M. H. 1966. Audibility in the rat.
J. Comp. Phys. Psych., 62(2):289—9l.

Grota, L. J. 1973. Effects of litter size, age of young,
and parity on foster mother behaviour in Rattus
norvegiCus. Anim. Behav., 21:78-82.

 

Harris, V. T. 1954. Experimental evidence of reproductive
isolation between two subspecies of Peromyscus manicu-

 

latus. Cont. Lab. Vert. Biol., Univ. Mich., 70:1-13.

Hart, F. M. and King, J. A. 1966. Distress vocalizations
of young in two subspecies of Peromyscus maniculatus.
J. Mammal., 47(2)287-93.

 

Horner, B. E. 1947. Parental care of young mice of genus
Peromyscus. J. Mammal., 28(1):l-36.

 

Horner, B. E. and Taylor, J. M. 1969. Paternal behavior in
Rattus fuscipes. J. Mammal., 50(4)803-5.

 

Houseknecht, C. R. 1968. Sonographic analysis of vocaliza-
tions of three species of mice. J. Mammal., 49(3):
555-60.

Huestis, R. R. 1933. Maternal behavior in the deer-mouse.
J. Mammal., 14(1):47-9.

Kahmann, H. and Ostermann, K. 1951. Wahrnehmen und Her-
vorbingen hoher Tone bei kleinen Saugetieren. Ex-
perientia, 7:268-9.

King, J. A. 1955. Social behavior, social organization,
and population dynamics in a black-tailed prairie
dog town in the Black Hills of South Dakota. Cont.
Lab. Vert. Biol., Univ. Mich., 67:1-123.

King, J. A. 1958. Maternal behavior and behavioral develop-
ment in a subspecies of Peromyscus maniculatus. J.
Mammal., 39:177-90.

 

 

197

King, J. A. 1963. Maternal behavior in Peromyscus. in.
Maternal Behavior in Mammals. Rheingold, H. (ed).
New York: Wiley and Sons.

King, J. A. 1968. Psychology. in King, J. A. (ed) Biology
of Peromyscus (Rodentia). Amer. Soc. Mammal.

King, J. A. 1970. Ecological Psychology: An approach to
motivation. Neb. Symp. Motiv. 18:1-33.

Ladefoged, P. 1962. Elements of Acoustic Phonetics.
Chicago: Univ. Chicago Press.

Layne, J. N. 1968. Ontogeny. in King (ed) Biology of Pero—
myscus (Rodentia). Amer. Soc. Mammal.

Leon, M., Numan, M. and Moltz, H. 1973. Maternal behavior
in the rat: facilitation through gonadectomy. Sci—
ence, 179:1018-19.

Levine, S. 1967. Maternal and environmental influences on
the adrenocortical response to stress in weaning rats.
Science, 156:259-60.

Levine, S. and King, D. L. 1965. The effect of auditory re-
strictions during pregnancy on offspring survival.
Psychonom. Sci., 3:275—6.

Lisk, R. D. 1971. Oestrogen and progesterone synergism and
elicitation of maternal nest—building in the mouse
(Mus musculus). Anim. Behav., 19:606-10.

Lisk, R. D., Pretlow, R. A. and Friedman, S. M. 1969. Hor-
monal stimulation necessary for elicitation of mater-
nal nest-building in the mouse (Mus musculus). Anim.
Behav., 17:730-7.

Lloyd, J. E. 1972. Vocalizations in Marmota monax. J. Mam-
mal., 53(1):214-16.

Marler, P. 1967. Animal communication signals. Science,
175:769-74.

McCarley, W. H. 1964. Ethological isolation in the ceno—
species Peromyscus leucopus. Evolution, 18(4):33l-
42.

 

Meier, G. W. & Schutzman, L. H. 1968. Mother-infant inter-
actions and experimental manipulations: confounding
or misidentification. Devel. Psychobiol., 1:141—5.

Michener, G. R. 1971. Maternal behavior in Richardson's
ground squirrel, Spermophilus richardsonii

 

 

198

richardsonii:' retrieval of young by lactating fe-
males. Anim. Behav., 19:653-6.

 

Michener, G. R. 1973. Maternal behavior in Richardson's
ground squirrel (Spermophilus richardsonii richard-
sonii): retrieval of young by non-lactating females.
Anim. Behav., 21(1): 157-9.

 

Moltz, H. 1970. Effects of breeding experience on the
maternal behavior of the laboratory rat. in Aronson
33 El (eds) Development and Evolution of EEhavior.
San Francisco, Calif.: Freeman.

Moore, R. E. 1965. Olfactory discrimination as an isolating
mechanism between Peromyscus maniculatus and Peromys-
cus polionotus. Amer. Midl. Nat., 73(1):85-100.

 

 

 

Munn. N. L. 1950. Handbook of Psychological Research on
the Rat. Boston: Houghton waflin Co.

Nicholson, A. J. 1941. The homes and social habits of the
Wood-mouse (Peromyscus leucopus noveboracensis) in
Southern Michigan. Amer. Midl. Nat., 25(1):196—223.

 

Nitschke, W., Bell, R. W., and Zachman, T. 1972. Distress
vocalizations of young in three inbred strains of
mice. Devel. Psychobiol., 5(4):363-70.

Noirot, E. l964a. Changes in responsiveness to young in the
adult mouse 1. The problematical effect of hormones.
Anim. Behav., 12(1):52-8.

Noirot, E. 1964b. Changes in responsiveness to young in
the adult mouse IV. The effect of an initial contact
with a strong stimulus. Anim. Behav., 12(4):442-5.

Noirot, E. 1964c. Changes in responsiveness to young in the
adult mouse: The effect of external stimuli. J.
Comp. Phys. Psych., 57(1):97-9.

Noirot, E. 1965. Changes in responsiveness to young in the
adult mouse III. The effect of immediately preceding
performances. Behaviour, 24(3-4): 318-25.

Noirot, E. 1966a. Ultrasons et comportements maternels chez
les petits rongeurs. Ann. Soc. Roy. 2001. Belg., 95:
47-56.

Noirot, E. 1966b. Ultrasounds in young rodents I. Changes
with age in albino mice. Anim. Behav., 14(4):459-62.

Noirot, E. 1968. Ultrasounds in young rodents II. Changes
with age in albino rats. Anim. Behav., 16:129-34.

 

   

 

199

Noirot, E. 1969a. Changes in responsiveness to young in the
adult mouse V. Priming. Anim. Behav., 17:542-6.

Noirot, E. 1969b. Serial order of maternal responsiveness
in mice. Anim. Behav., 17(3): 547-50.

Noirot, E. 1970. Selective priming of maternal responses by
auditory and olfactory cues from mouse pups. Devel.
Psychobiol., 2(4):273-6.

Noirot, E., 1972. Ultrasounds and maternal behavior in
small rodents. Devel. Psychobiol. 5(4):37l-87.

Noirot, E. and Pye, D. 1969. Sound analysis of ultrasonic
distress calls of mouse pups as a function of their
age. Anim. Behav., 17:340-9.

Noirot, E. and Richards, M. P. M. 1966. Maternal behaviour
in virgin female golden hamsters: changes consequent
upon initial contact with pups. Anim. Behav., 14:
7-10 0

Okon, E. E. 1970a. The effect of environmental temperature
on the production of ultrasounds by isolated non—
handled albino mouse pups. J. Zool. Lond., 162:71—
83.

Okon, E. E. 1970b. The ultrasonic responses of albino mouse
pups to tactile stimuli. J. Zool. Lond., 162:485-92.

Okon, E. E. 1971. The temperature relations of vocalizations
in infant Golden hamsters and Wistar rats. J. Zool.
Lond., 164:227-37.

Payne, R. S. 1962. How the barn owl locates prey by hear-
ing. Living Bird, 1:151-9.

Peterson, E., Heaton, W. C. and Wruble, S. D. 1969. Levels

of auditory response in fissiped carnivores. J. Mam-
mal, 50(3):566-78. -

Price, E. O. 1967. Behavioral changes in the deermouse,
Peromyscus maniculatus bairdi, after seventeen years
of domestication: Reaction to novel stimuli. Unpub-
lished Doctoral Thesis. Michigan State University.

Ralls, K. 1967. Auditory sensitivity in mice: Peromyscus
and Mus musculus. Anim. Behav., 15(1):123-8.

 

 

Ressler, R. H. 1962. Parental handling in two strains of
mice reared by foster parents. Science, 137:129-30.

 

 

200

Richards, M. P. M. 1966a. Maternal behaviour in virgin fe-
male golden hamsters (Mesocricetus auratus Weterhousa:
the role of the age of the test pup. Anim. Behav.,
14:303-9.

 

Richards, M. P. M. 1966b. Maternal behaviour in the golden
hamster: responsiveness to young in virgin, pregnant
and lactating females. Anim. Behav., 14:310-13.

Richards, M. P. M. 1967. Maternal behaviour in rodents and
lagomorphs. Adv. in Reprod. Phys., 2:53-110.

Richards, M. P. M. 1969. Effects of oestrogen and projes-
terone on nest-building in the golden hamster. Anim.
Behav., 17(2):356-6l.

Rood, J. P. 1972. Ecological and behavioural comparisons
of three genera of Argentine cavies. Anim. Behav.
Monog. 5.

Rosenblatt, J. S. 1965. The basis of synchrony in the be—
havioral interaction between the mother and her off-
spring in the laboratory rat. in Foss, B. M. (ed)
Determinants of Infant Behavior III. London.

Rosenblatt, J. S. 1967. Nonhormonal basis of maternal be-
havior in the rat. Science, 156:1512—14.

Rosenblatt, J. S. 1970. Views on the onset and maintenance
of maternal behavior in the rat. in Aronson 92 al_
(eds.) Development and Evolution of Behavior. San
Francisco, Calif.: Freeman.

Rosenzweig, M. R., Riley, D. A. and Krech, D. 1955. Evi—
dence for echolocation in the rat. Science, 121:600.

Rowell, T. E. 1960. On the retrieving of young and other
behaviour in lactating golden hamsters. Proc. Zool.
Soc. Lond., 135(2):265—82.

Ruffer, D. G. 1966. Observations on the calls of the grass-
hopper mouse (Onychomys leucogaster). Ohio J. Sci.,
66(2);219.

Sales, G. D. 1972. Ultrasouns and aggressive behavior in
rats and other small mammals. Anim. Behav., 20:88-
100.

Schleidt, W. M. 1948. Tone Hoher Frequenz bei Mausen I.
Experientia, 4:145-6.

 

 

201

Schleidt, W. M. 1951. Tone Hoher Frequenz bei Mausen II.
Experientia, 7(2):65-66.

Schleidt, W. M. 1952. Reaktonen auf Tone hoher Frequenz
bei Nagern. Naturwiss., 39:69—70.

Scott, J. P. 1958. Animal Behavior. Chicago: Univ. Chi-
cago Press.

Sebeok, T. A. 1965. Animal communication. Science, 147:
1006.

Sewell, G. D. 1968. Ultrasound in rodents. Nature, 217:
682-3.

Sewell, G. D. 1970. Ultrasonic communication in rodents.
Nature, 277:410

Smith, J. C. 1972. Sound production by infant Peromyscus
maniculatus (RodentiazMyomorpha). J. Zool. Lond.,
168:369-79. -

 

 

Smith, W. J. 1963. Vocal communication of information in
birds. Amer. Nat., 47:117-25.

Smith, W. J. 1965. Message, meaning, and context in ethol-
Ogy. mere Nat., 99:405-90

Sotnick, B. M. 1967. Intercorrelations of maternal activi-
ties in the rat. Anim. Behav., 15:267-9.

Tembrock, G. 1963. Acoustic behavior of mammals. in Busnel,
R.-G. (ed) Acoustic Behaviour of Animals. New York:
Elsevier.

Tembrock, G. 1968. Land mammals. in Sebeok, T. A., Animal
Communication. Bloomington, Ind.: Indiana Univ.
Press.

Terkel, J. and Rosenblatt, J. S. 1968. Maternal behavior
\ induced by internal blood plasma injected in virgin
rats. J. Comp. Phys. Psych., 65(3):479-82.

Terkel, J. and Rosenblatt, J. S. 1971. Aspects of nonhor-
monal maternal behavior in the rat. Hormones and Be-
havior, 2(2):161-71.

Thoman, E. B. and Levine, S. 1970. Effects of adrenalectomy
on maternal behavior in rats. Devel. Psychobiol.,
3(4):237-44.

 

   

 

202

Waring, G. H. 1966. Sounds and communications of the
Yellowbellied marmot (Marmota flaviventris). Anim.
Behav., 14:177-183.

Wiesner, B. P. and Sheard, N. M. 1933. Maternal Behaviour
in the Rat. London: Oliver & Boyd.

Young, R. D. 1965. Influence of neonatal treatment on
maternal behavior: a confounding variable. Psychonon.
Sci., 3:295—6.

Zippelius, H.-M. and Schleidt, W. M. 1956. Ultraschall-
Laute bei Jungen Mausen. Naturwiss., 21:502.

 

 

APPENDICES

203

APPENDIX A

Electronic Equipment

The following list and drawing of the recording and
playback pathways are included to clarify the electronic

setup for comparative purposes.

1. Bruél and Kjaer condensor microphone, % in., #4133/34;
with cathode follower; frequency response rated at 20
Hz - 40 kHz. '

2. Bruél and Kjaer microphone power supply, #2801.

3. Tektronix #129 plug-in power supply with #2A61 plug-in
units; used to step up the signal from the microphone
to the tape recorder; essentially unlimited frequency
response.

4. Ampex SP-300 Instrumentation tape recorder (multi-chan—
nel, 4 speed), 10% in. reels; 3 db s/n ratio; frequency
response flat 50 Hz - 40 kHz. A11 recording done at 15
ips; playback at 1-7/8 ips for analysis. Ampex Instru-
mentation tape, 1 m1 mylar, #641.

5. Tektronix Oscilloscope #3A74.

6. Grass Oscilloscope Camera

7. Bruél and Kjaer frequency analyzer, #2107, frequency re-
sponse 20 Hz - 20 kHz; with Bruél & Kjaer logarythmic
potentiometer #2305.

8. Acoustic spectrograph "Voice Print" #4177B; 85-8000 Hz
response, Kay Sonagraph paper type B.

204

 

_‘

 

-.- _.

 

 

 

 

 

205

Figure A-l

Legend

Schematic of the electronic recording and analysis

pathways. Pieces of equipment are marked and connections

are indicated by arrows.

 

 

206

 

 

50150u05

_o>o_

 

Lax—ac.»

 

 

 

FIGURE A-I

xucoaro:
h

 

 

 

m><31h<m m_m>4<2< DZ< 02.nzOUmu

 

 

 

 

 

 

501530..
onbt

 

 

 

 

..o$=nEu
-9...

 

 

 

 

 

 

 

 

ULUEUU

o

 

 

magma—:30

m

 

 

 

09.30:.—

ace—39.39:

 

 

 

 

 

 

 

 

 

 

 

MICHIGAN sum: UNIV. LIBRARIES
WW11111111111“ll“WNWHIWWHI
31293010708158