.2. , 3.1.”..4.. . . ,
. 2 . .. _ .
‘ , Km .£

. .. (5.11:! .
4..«4‘)..I1..4 w
...§.::.h.§|. L
« Qty“.
2

41. . .
3:}... :2.
1‘). . 11.1.5.3: ‘
Elﬁn;
5+3... .t . .
y at}...
. . . . .
o .
. . #v A
. .. v , V . 5.: 5 ﬁ.
7‘. .. .. . 1 ..... : Saga 2..
V83:53.. tt :0... . 3. in. t
‘ . .1.
V ‘95.: at
‘ . .
. th‘tﬂilﬁi
:3. IA!
a? x
a: Li.
:v»)

as”, .
$9.3. \,

..
. .
3 Hum... Ruins?
i4... . .L. ..r.....x...
~ . ' ‘3.“
.I 1.4.0.1 I: )‘0
unuﬁqzﬂvr u
, . “.115 9”,.
. . x A
. Wu
inunnomr ‘mmlvnm
' A .svaitux‘tt .a.

 

 

 

 

7 t
s ’. ‘
“it

 

w.” , LIBRARY
5 Michigan State
i :31 University

This is to certify that the
dissertation entitled

NEUROBIOLOGY OF SONG LEARNING AND PERCEPTION IN
THE ZEBRA FINCH (TAENIOPYGIA GUTTA TA), WITH A FOCUS
ON THE ROLE OF THE HIPPOCAMPUS

presented by

David J. Bailey

has been accepted towards fulﬁllment
of the requirements for the

PhD degree in PSYCHOLOGY

 

 

\ «i; \U We
Maloi' Professor's Signature

3’3?-Oé

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution

.v-.-.-.-.-.-.-.-.-.----.--—-.-.-.-.-.-.—.-.-.-.-.-.-.---.--A--.-..--.-.-.-.-.-.-.qo-o->-A-s-.-.-l-.-.-.-.-.-.-.. -.-.-

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

NBﬂEﬁDZlI’157 DATE DUE DATE DUE

 

JUL W 23:33 09

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:lCIRC/DateDue.indd-p.1

 

NEUROBIOLOGY OF SONG LEARNING AND PERCEPTION IN THE ZEBRA
FINCH (TAENIOPYGIA OUTTA TA), WITH A FOCUS ON THE ROLE OF
THE HIPPOCAMPUS
By

David J. Bailey

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

2006

ABSTRACT
NEUROBIOLOGY OF SONG LEARNING AND PERCEPTION IN THE ZEBRA
FINCH(TAEN10PYGIA GUTTA TA), WITH A FOCUS ON THE ROLE OF
THE HIPPOCAMPUS
By

David J. Bailey

Songbirds have become one of the most widely used model systems for the
examination of neural processes underlying learning and memory. Zebra ﬁnch
( T aem’opygia guttata) song is highly sexually dimorphic in that only males produce the
behavior, yet perception of these vocalizations is critical to the reproductive success of
both sexes. Males need to hear the vocalizations of a male tutor and, later, their own
vocalizations in order to learn and develop song, but it is debatable whether experience
plays a role in the development of song perception or preference in females. Vocal
behavior in males is regulated by a system of interconnected nuclei that undergo dramatic
morphological changes during development. However, other forebrain areas that
ultimately receive input from the auditory thalamus Show differential neuronal activity
following conspeciﬁc or heterospeciﬁc song presentations in both sexes, suggesting that
they are more important for the perception or perhaps neural storage of song. The
hippocampus is also activated following playback of conspeciﬁc songs, suggesting that
this area, traditionally associated with spatial memory, might play a role in song-related
behavior. The present experiments began to test the role of early auditory experience in
the development of female song perceptions, the development of neural responses to song
in auditory regions and the hippocampus, the afferent connectivity of the hippocampal

formation, and whether destruction of hippocampal tissue affects the consolidation or

retrieval of song templates or normal preferences for particular songs. These experiments
show that (1) female zebra finches isolated from song but not biparental care during
development respond like birds raised normally during conspecifrc song presentations but
are initially limited in their ability to associate a novel environment with relevant song; (2)
early in development, the sexes display different patterns of immediate early genes that
may encode information about auditory stimuli but, over time, these responses become
sexually monomorphic and likely remain that way in adulthood; (3) together with prior
data indicating the efferent connectivity of the hippocampus, projections to the region
indicate connections (some reciprocal) with nuclei involved in song behavior and suggest
a role for the region in responses to vocal communications signals or auditory-related
memories in the zebra finch as well as homology with the mammalian hippocampus; and
(4) relatively small lesions of the hippocampus in males and females before the period
song is learned, when it begins to be practiced (at least by males), or in adulthood have
no effect on song preferences in males and females or song production in males, but do
inﬂuence memory for a particular location. Taken together, these results indicate that the
hippocampus, although not involved in song learning directly, may be integral in the
consolidation of memories relating to song or the modulation of responses of other
regions based on the salience of particular vocal signals or environments in which they

are heard or produced.

Copyright by
DAVID JOHN BAILEY
2006

To my wife, Amy, and my parents, Patricia and the late Lloyd Bailey

ACKNOWLEDGEMENTS

The research and writing of this dissertation would not have been possible without
the support and encouragement of a number of individuals and departments within and
outside of Michigan State University.

First, my gratitude to my advisor and mentor Juli Wade, for always being
available to provide direction and assistance, and for graciously allowing me to pursue
this particular area of research. To members of my dissertation committee: Laura Smale,
who was a part of each of my committees since my ﬁrst year at MSU, and Marc
Breedlove, Tony Nunez, Lyn Clemens, and Steve Maren. I also thank Jane Witten and
Fred Helmstetter of the University of Wisconsin-Milwaukee, who took me into their labs
as an undergraduate and were instrumental in stimulating my interest in research in
general and the study of biological psychology in particular. The collective efforts of all
these individuals made me a more critical thinker and a better writer and researcher.

My thanks to countless graduate students, post-docs, technicians and
undergraduates in the Wade lab who over the years provided help, friendship, and
intellectual input and procedural support for my work, namely Erin O’Bryant, Camilla
Peabody, Julia Rosebush, Stephanie Fuehring, Claudia Ruiz, Melissa Holmes, Katie
Grausum, Matt Lovem, Sean Veney, Nancy Oberg, Casey Bartrem, Jenny Stynoski,
Malik Williams, Lace Svec, Laurel Beck, Jen Neal, YuPing Tang, and Matt Burke.

Several individuals, like Shari Stockmeyer, made my day-to-day work more
manageable, and were vital to particular experiments. David McFarlane was extremely
helpful with computer, software and other technical support, as was Eric Weston with

construction of the testing chambers.

vi

I also express my extreme appreciation for funding from the National Institutes Of
Health, the Department of Psychology, the Graduate School, and the Council of Graduate
Students.

I will always look fondly on my time in East Lansing thanks in large part to
friends made there, like Russ Romeo, Dave Swender, Chris Wilson, Kalynn Schulz, Mike
Schwartz, Mary Martin, Erich Ottem, and John Morris.

Thank you to my parents, Patricia and the late Lloyd Bailey, and my siblings,
especially Thomas Bailey, for their steadfast support and encouragement, their continued
interest in my work, and for keeping me focused on my career goals.

Finally, there is one person above all others to whom I owe thanks: my wife, Amy.
She was and continues to be the foremost advocate of my career and a stabilizing force in
my life. Her unwavering love, patience, and support, especially in the face of late nights
and my occasional foul mood, was indispensable to surviving graduate school. Without

her, in just about every way, this dissertation would not have been possible.

vii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... xii
LIST OF FIGURES ................................................................................. xiii
KEY TO ABBREVIATIONS ...................................................................... xx
CHAPTER ONE
The zebra ﬁnch as a model system ....................................................................... 1
Song learning in zebra ﬁnches .......................................................... 2
Male song learning .................................................................. 2
The question of female song learning .......................................................... 3
Song control regions and circuits in zebra ﬁnches ............................................... 4
Song learning ........................................................................ 4
Song production ...................................................................... 5
Auditory perception ................................................................. 5
Song-speciﬁc responses in the adult female zebra ﬁnch hippocampus. . ....9
Avian and mammalian hippocampal function .......................................... 9
Hippocampal function in food-storing and navigating birds ................. 13
Hippocampal lesions in zebra ﬁnches and a potential role for the region
in song behavior ............................................................................ 13

Brief history of theories on hippocampal function in mammals . . . . . 14
Comparison of hippocampal connectivity in mammals and birds . ...15

Prospectus .................................................................................. 19
CHAPTER TWO

Song exposure during development modiﬁes adult behavior following song perception in
the female zebra ﬁnch ............................................................................... 22
INTRODUCTION ......................................................................... 22
MATERIALS AND METHODS ........................................................ 26
Animals and housing .............................................................. 26

Experimental manipulation ....................................................... 26

Behavioral testing ................................................................. 29

Data analysis ....................................................................... 31

RESULTS .................................................................................... 32
DISCUSSION ............................................................................... 38
Summary ........................................................................... 38

Experience vs. cue competition .................................................. 39

Song learning outside of a critical period ....................................... 39

Innate song recognition ........................................................... 40

Potential neural mechanisms of song-related memories ..................... 41

Lack oftreatment with estradiol . . . . ...42

Differences with similar studies .................................................. 43

viii

CHAPTER THREE
Expression of the immediate early genes F08 and ZENK following auditory stimulation

in 30 day post-hatch male and female zebra ﬁnches ........................................... 45
INTRODUCTION ......................................................................... 46
MATERIALS AND METHODS ......................................................... 47

Animals and housing .............................................................. 47
Stimulus exposure and tissue collection ........................................ 48
F OS and ZENK immunohistochemistry ........................................ 49
Data analysis ........................................................................ 50
RESULTS ................................................................................... 51
DISCUSSION ............................................................................... 58
Summary ............................................................................ 58
FOS expression .................................................................... 58
ZENK expression .................................................................. 59
General conclusions ............................................................... 61

CHAPTER FOUR

F08 and ZENK responses in 45 day-old zebra ﬁnches vary with auditory stimulus and

brain region, but not sex ............................................................................. 64
INTRODUCTION .......................................................................... 65

MATERIALS AND METHODS ......................................................... 68
Animals and housing .............................................................. 68
Stimulus exposure and tissue collection ........................................ 69
F OS and ZENK immunohistochemistry ....................................... 70
Data analysis ....................................................................... 71

RESULTS ................................................................................... 73

DISCUSSION ............................................................................... 79
Summary ............................................................................ 79
Differences in patterns of expression between d30 and d45 ................. 79
Medial versus lateral immediate early gene expression in NCM ........... 80
Song-speciﬁc activation in the hippocampus .................................. 81
Neuronal responses in CMM ..................................................... 82

CHAPTER FIVE

Afferent connectivity of the male and female zebra ﬁnch hippocampus determined by

iontophoretic injections of the retrograde tracer fast blue ..................................... 84
INTRODUCTION .......................................................................... 84
MATERIALS AND METHODS ........................................................ 85

Animals and housing .............................................................. 85
Tracer injection and tissue preparation ......................................... 85
Data analysis ....................................................................... 88

RESULTS ................................................................................... 88

DISCUSSION .............................................................................. 99
Summary ........................................................................... 99
Comparison with efferent connectivity ......................................... 99

ix

Interconnectivity of hippocampal subdivisions in zebra ﬁnches, pigeons

and mammals ................................................................... 100
Sexually dimorphic projections .............................................. 101
CHAPTER SD(
Effects of lesions of the hippocampus in adult and juvenile male and female zebra ﬁnches
on song-related behavior .......................................................................... 104
INTRODUCTION ........................................................................ 104
MATERIALS AND METHODS ........................................................ 107
Animals, Surgery and Housing ................................................ 107
Behavior Testing ................................................................. 109
Spatial Memory Test(s) .......................................................... 109
Song Preference Tests ........................................................... 112
Male Song Learning and Production .......................................... 114
Tissue Collection ................................................................. 118
RESULTS AND DISCUSSION OF VALIDATION OF METHODS ............ 121
Spatial Memory .................................................................. 121
Song Preference .................................................................. 128
Song Production .................................................................. 130
RESULTS — EXPERIMENTAL BIRDS .............................................. 131
Histology .......................................................................... 134
Spatial Memory ................................................................... 134
Song Preference .................................................................. 147
Song Production .................................................................. 147
Regression Analyses .............................................................. 151
DISCUSSION ............................................................................ 153
Summary .......................................................................... 153
Effects of lesions of nuclei in the song control circuit ....................... 154
Does immediate early gene activity infer a role for a region in a speciﬁc
behavior? .................................................................................................. 155
Comparison of spatial memory results with other birds and mammals... 1 55
Sex differences in spatial memory ............................................. 157
CHAPTER SEVEN
Learned song behavior may not be restricted to a sensitive period in female zebra
ﬁnches .............................................................................................. 159

Immediate early gene responses of neurons in auditory regions of male and female zebra
ﬁnches change over development as song learning progresses .............................. 161

Potential functional signiﬁcance of immediate early gene activity following song
presentations ........................................................................................ 162

The hippocampus and NCM in zebra ﬁnches may be involved in modulating responses to
the environmental or social contexts of song ................................................... 167

The structure and function of the hippocampus in zebra ﬁnches are similar to those in a
non-songbird and show homology with the mammalian hippocampus ........................... 170

Zebra ﬁnches are an ideal model system for the further examination of physiological

processes underlying speciﬁc forms of learning and memory ............................... 172
General Summary .................................................................................. 173
LITERATURE CITED ............................................................................ 175

xi

LIST OF TABLES
CHAPTER FIVE

Table 5.1. Brain regions in male and female zebra ﬁnches exhibiting fast blue labeled
neurons following iontophoresis into the dorsolateral subdivision of the hippocampus
(DLHP). Below the number of birds of each sex in which retrogradely labeled neurons
were observed (or not) in a particular brain region, the density of labeling is indicated as
follows: +4—t (high number of labeled neurons); ++ (moderate number of labeled
neurons); + (few labeled neurons); - (no labeled neurons). DMHP: dorsomedial
subdivision of the hippocampus; VHP: ventral subdivision of the hippocampus; CMM:
caudomedial mesopallium; SL: lateral septum; PVN: periventn'cular nucleus; DMP:
posterior portion of the dorsomedial nucleus of the thalamus ........................................... 90

CHAPTER SD(

Table 6.1. Similarity measurements obtained by comparing four different phrases (1-4)
of a male zebra ﬁnch’s song and those (1-4) of one of his male offspring. The “similarity
score” is an aggregate of the three other values, which indicate parallels in the types of
notes used (percent similarity), the frequencies of those notes (mean accuracy), and the
lengths of silent intervals between the notes (sequential match). Means and standard
errors for the measures are indicated at the ends of each column ................................... 131

Table 6.2. Similarity measurements Obtained by comparing three different phrases of a
male zebra ﬁnch’s song before and after a sham lesion of the hippocampus. The
“similarity score” is a product of the three other values, which indicate parallels in the
types of notes used (percent similarity), the frequencies of those notes (mean accuracy),
and the lengths of silent intervals between the notes (sequential match). Means and
standard errors (SEM) for the measures are indicated at the ends of each column ......... 133

Table 6.3. Similarity measurements of song phrases from adult male zebra ﬁnches
lesioned or sham-lesioned at d20, d45 or in adulthood. For d20 and d45 birds, similarity
measurements were obtained by comparing one phrase from the songs of each of these
males with one from their father. In birds lesioned or sham-lesioned as adults, phrases
ﬁom individual males’ songs both before and after lesion (or sham lesion) were
compared. None of the similarity measurements from the lesioned birds differed
signiﬁcantly from those of the sham controls ................................................ 150

xii

LIST OF FIGURES
CHAPTER ONE

Figure 1.1 Wiring diagram of structures in the zebra ﬁnch song circuit pulled together
from tract tracing studies, detailing the high degree of known interconnectivities between
the structures (Bottjer, Brady, & Cribbs, 2000; Foster & Bottjer, 1998; Mello, Vates,
Okuhata, & Nottebohm, 1998; Striedter & Vu, 1998; Székely, 1999; Székely & Krebs,
1996; Vates, Broome, Mello, & Nottebohm, 1996; Vates & Nottebohm, 1995; Vates,
Vicario, & Nottebohm, 1997). Solid lines specify auditory pathways, thick dotted lines
primary motor pathways, and dashed lines song learning pathways. Thin dotted black
lines designate portions of thalamo-cortical feedback loops, and the dashed + dotted lines
indicate the only known connections of the zebra ﬁnch hippocampus with the regions
indicated. DLM = medial part of the dorsolateral thalamic nucleus, DM = dorsomedial
nucleus of the intercollicular complex, DMP = posterior portion of the dorsomedial
nucleus of the thalamus, HP = hippocampus, lCMM = lateral portion of the caudomedial
mesopallium, Ll/L2/L3 = subdivisions of the Field L complex, LLV = ventral nucleus of
the lateral lemniscus, lMAN = lateral portion of the magnocellular nucleus of the anterior
nidopallium, mCMM = medial portion of the caudomedial mesopallium, mMAN =
medial portion of the magnocellular nucleus of the anterior nidopallium, RA = robust
nucleus of the arcopallium, NIf = nucleus interface of the nidopallium, nXIIts =
tracheosyringeal portion of the hypoglossal nucleus, Ov = nucleus ovoidalis, UVa =
nucleus uvaeforrnis, X = Area X of the medial striatum ........................................ 7

Figure 1.2. Density of F OS—immunoreactive cells (mean + SEM) in the NCM of adult
females exposed to conspeciﬁc (CON) and heterospeciﬁc (HET) song, tones (TON)

or no song (NS) (top panel). Different lowercase letters indicate signiﬁcant differences
between groups. The bottom panels contain photomicrographs of coronal sections
through the zebra ﬁnch brain at the level of the NCM, showing F OS immunoreactivity in
each of the stimulus conditions. Scale bar = 100 um .......................................... 10

Figure 1.3. Density of FOS-immunoreactive cells (mean + SEM) in the hippocampus
(HP) of female birds in the conspeciﬁc (CON), heterospeciﬁc (HET), tone (TON) and no
song (NS) conditions (top panels). Different lowercase letters indicate signiﬁcant
differences between groups. Photomicrographs indicate FOS-immunoreactive cell nuclei
for each group in the HP. Scale bar = 100 um ................................................... 11

Figure 1.4. Density of F OS-immunoreactive cells (mean + SEM) in the CMM of adult
females in the conspeciﬁc (CON), heterospeciﬁc (HET), tone (TON) and no song (NS)
conditions. Different lowercase letters indicate signiﬁcant differences between combined
groups (“song” versus “no song”). Photomicrographs of sections at the level of the
CMM displaying FOS immunoreactivity in response to each of the stimulus conditions
are shown in the bottom panel. Scale bar = 100 ple

xiii

CHAPTER TWO

Figure 2.1. Sonagram of vocal behavior in the presence of a novel female before (A) and
after (B) severing of the tracheosyringeal nerves of a father of birds in the song-isolated
group. While rhythmic tones associated with respiration were consistently heard, song
was never detected following axotomy ........................................................... 28

Figure 2.2. Chamber used for song presentations. The apparatus was constructed of a
wood frame, steel mesh sides, back, and top, and a Plexiglas front through which
observations and video recordings were made. Two sets of three wooden perches were
located at each end of the chamber, and a single perch was in the middle. The middle
perches in the sets of three were located 20.2 cm above the ﬂoor of the chamber, and
were spaced 6.3 cm from the adjacent perches. All other perches were 30 cm above the
ﬂoor Of the chamber. Nest boxes were located at each end of the chamber. Nest boxes
were constructed of polyvinyl chloride (PVC), and were 12.7 cm high and 12.7 cm in
diameter with a 3.8 cm diameter hole 2.5 cm from the top. A 5 cm long perch extended
outward at 90° below this hole, and the top of each nest box was covered with a
removable Plexiglas disc. Hidden in each nest box were speakers that delivered song,
and sound attenuating foam covered the walls surrounding the entire apparatus ........... 30

Figure 2.3. Mean (+ SEM) calls plus movement before song exposure on the ﬁrst testing
day. These baseline values were not signiﬁcantly different, although the song isolates
appeared somewhat more active ................................................................... 34

Figure 2.4. Mean (+ SEM) calls plus movement during ﬁve conspeciﬁc and
heterospeciﬁc song presentations (A), immediately after those songs (B), and 24 hours
after (C) by females raised with exposure to zebra ﬁnch song (CONTROLS) or isolated
from song early in life through adulthood (SONG ISOLATES). Responses 24 hours
following the ﬁrst conspeciﬁc and heterospeciﬁc songs are highlighted by dashed lines.
The “*” indicates a signiﬁcant difference between the groups. Note differences in the Y-
axis scales ............................................................................................. 36

CHAPTER THREE

Figure 3.1. Density of FOS-immunoreactive nuclei (mean + SEM) in the caudomedial
nidopallium (NCM), caudomedial mesopallium (CMM) and hippocampus (HP) in d30
female and male zebra ﬁnches exposed to conspeciﬁc (CON) or heterospeciﬁc song,
tones (TON) or no song (NS). Number of animals analyzed per immediate early gene,
brain region and sex is indicated under each bar. Different lowercase letters indicate
signiﬁcant differences between groups within brain region ................................... 52

Figure 3.2. Density of ZENK-immunoreactive nuclei (mean + SEM) in the caudomedial
nidopallium (NCM), caudomedial mesopallium (CMM) and hippocampus (HP) in d30
female and male zebra ﬁnches exposed to conspeciﬁc (CON) or heterospeciﬁc song,
tones (TON) or no song (NS). Number of animals analyzed per immediate early gene,
brain region and sex is indicated under each bar53

xiv

Figure 3.3. F 08- and ZENK-immunoreactivity in the hippocampus (HP) of female zebra
ﬁnches following conspeciﬁc or heterospeciﬁc song presentations. The arrow in the top
left panel indicates the location of the lateral ventricle, just below where immunoreactive
cells are consistently detected. The right edge of each photograph is at the midline.

Scale bar = 200 um .................................................................................. 55

Figure 3.4. Photomicrographs of FOS immunoreactivity in the caudomedial nidopallium
(NCM) of female and male zebra ﬁnches following presentations of conspeciﬁc (CON)
or no song (NS). Scale bar = 100 um ............................................................ 56

Figure 3.5. Photomicrographs of ZENK immunoreactivity in the caudomedial
nidopallium (N CM) of female and male zebra ﬁnches following presentations of
conspeciﬁc (CON) or no song (NS). Scale bar = 100 pm .................................... 57

CHAPTER FOUR

Figure 4.1. Density of ZENK- (top panels) and F 08- (bottom panels) immunoreactive
neurons in the caudomedial nidopallium (N CM), caudomedial mesopallium (CMM) and
hippocampus (HP) of 45 day post-hatch male and female zebra ﬁnches following
presentations of conspeciﬁc songs (CON), heterospeciﬁc songs (HET) or no song (NS; n
= 6 males and n = 6 females for each). Different lowercase letters indicate signiﬁcant
differences between groups within brain region and IEG. Note the y-axis scale for ZENK
immunoreactivity is greater than that for F OS .................................................. 74

Figure 4.2. Photomicrographs of ZENK- (left panels) and F 08- (right panels)
immunoreactivity in the caudomedial nidopallium (N CM) of male or female zebra
ﬁnches following presentations of conspeciﬁc songs (CON), heterospeciﬁc songs (HET)
or no song (NS). Scale bar = 100 um ............................................................ 75

Figure 4.3. ZENK- (top graphs) and 1F 08- (bottom graphs) immunoreactive nuclei in the
medial and lateral caudomedial nidopallium (NCM) following conspeciﬁc (CON),
heterospeciﬁc (HET) or no song (NS) presentations. For both ZENK and F OS,
expression within lateral NCM was signiﬁcantly greater than that in medial NCM (see
Results). Different lowercase letters in each graph indicate signiﬁcant differences
between auditory stimulus conditions within these regions. Note the y-axis scale for
ZENK immunoreactivity is greater than that for F OS .......................................... 78

CHAPTER FIVE

Figure 5.1. Coronal section through the dorsomedial surface of the zebra ﬁnch
telencephalon (top; red box indicates photographed area in bottom panel) at the level of
the hippocampus showing iontophoresed fast blue in the dorsolateral (DLHP) subdivision.
Note the retrogradely labeled cells at each arrow in the dorsomedial (DMHP) and ventral
(VHP) subdivisions of the structure and in the caudomedial mesopallium (CMM). Scale
bar = 200 um .......................................................................................... 92

XV

Figure 5.2. Coronal sections through the zebra ﬁnch brain at the level of the lateral
septum (SL; dashed red box) and periventricular nucleus of the hypothalamus (PVN;
dotted red box). Iontophoretic injections of fast blue into the dorsolateral subdivision of
the hippocampus resulted in ipsilaterally labeled cells within the SL (arrows, panel B;
photo from a male zebra ﬁnch) in both male and female zebra ﬁnches and the PVN in
females only (arrows, panel D). Thionin stained tissue through the SL (panel A) and
PVN (panel C) are provided for comparison. Scale bar for panels A and B = 200 um, and
for panels B and C = 100 um ...................................................................... 94

Figure 5.3. Coronal sections of a zebra ﬁnch brain at the level of the posterior portion of
the dorsomedial thalamic nucleus (DMP; dashed red box). Fast blue injections into the
dorsolateral subdivision of the hippocampus resulted in a few ﬁlled cells within the
ipsilateral DMP (arrow) in males but not females. HP = hippocampus, CMM =
caudomedial mesopallium; N = nidopallium. Scale bar = 100 pm .................................. 95

Figure 5.4. Coronal sections through a zebra ﬁnch brain at the level of HVC. Red,
dashed box indicates the region depicted in the thionin stained (panel A) and fast blue
labeled (panel B) sections. NCM = caudomedial nidopallium, Cb = cerebellum. Scale
bar = 200 um ...................................................................................................................... 96

Figure 5.5. Iontophoretic injections of fast blue into the hippocampus (panel A) and
ﬂuorogold into the caudomedial mesopallium (CMM; panel B at arrow) of a male zebra
ﬁnch resulted in fast blue ﬁlled cells in the CMM (panel C) and ﬂuorogold labeled
neurons in each of the hippocampal subdivisions (panels C and D). DLHP = dorsolateral
subdivision, DMHP = dorsomedial, VHP = ventral, V = ventricle separating the
hippocampus from CMM. Scale bar for panels A and B = 50 um and for C and D = 200
um ...................................................................................................................................... 98

CHAPTER SIX

Figure 6.1. (A) Chamber used for spatial memory tests following hippocampal lesions,
frontal view. Doors at the bottom of the chamber were used as entrance points. (B)
Chamber diagrammed from above. Perches (indicated in grey) at the ends of the

chamber that run parallel with the front of the maze sat 16 cm above the ﬂoor of the arms,

and those that run perpendicular in each arm were 20 cm above. In the middle of the
chamber, 3 cm above the ﬂoors of the arms, were perches in the shape of a plus, and the
part of the plus that runs parallel with the front of the chamber extended 5 cm into each
arm. The plastic cups (2 cm x 2 cm x 3 cm) were also at varying heights: cups 1a and 2a
were 19 cm above the ﬂoor of the arms, cups 1b and 2b 15 cm, and cups 1c and 2c 2 cm.
Perches were not placed directly below/adjacent to a cup so that birds would not land by
one randomly. The perches were placed nearby and just above so that birds would only
approach a cup to eat its contents ................................................................ 110

Figure 6.2. Sonagram (A; frequency over time) and spectral derivative (B; change in
power over time) of the same song portion from a male zebra ﬁnch. Introductory (i) and

xvi

‘0-

individual notes (identiﬁed as 1-5) are indicated. Zebra ﬁnch song consists of
introductory notes that are followed by notes of particular frequencies and changes in
amplitude that are repeated in a ﬁxed order (a “phrase;” three phrases indicated). Note
the “static” appearance of the background in the sonagram (A) and the elimination of that
extraneous noise in the spectral derivative (B) ................................................ 116

Figure 6.3. Example of an asymmetric similarity matrix obtained by comparing a phrase
from the song of a male zebra ﬁnch (tutor) and another from the song of one of his male
offspring (pupil). Yellow lines indicate the boundaries of individual notes or phrases
(and are used to obtain the “percent sequential match” score). High similarity values are
restricted to the diagonal, indicating that the individual notes in the tutor song were
learned and matched by the pupil in sequential order. Warmer colors in the diagonal line
indicate higher similarity (“percent similarity”). “Mean accuracy” measures the power of
the individual notes on a scale from white (high match) to gray to black (no sound or no
match) ................................................................................................. 119

Figure 6.4. Example of a symmetric similarity matrix obtained by comparing a phrase
from the song of a male zebra ﬁnch before and after a lesion of the hippocampus. High
similarity values are restricted to the diagonal, indicating that the individual notes in the
song pre—and post-lesion are in sequential order, and the warmer colors along the
diagonal line indicate higher similarity (“percent similarity”). Overall, “mean accuracy”
between the individual notes (indicated by gray and white colors) and the temporal order
of the notes (“sequential match;” black bars) are high ........................................ 120

Figure 6.5. Performance during the acquisition phase of a spatial memory test by ﬁve
hippocampal-lesioned and four intact male zebra ﬁnches. Initial latency (sec) to eat from
the baited cup by the food deprived birds upon ﬁrst exposure to the chamber, the average
time needed by each group during the last three acquisition trials, and the average number
of trials needed to reach criterion are indicated. Latencies of both groups of birds
decreased signiﬁcantly over the last three acquisition trials. No other signiﬁcant main
effects or interactions were found ............................................................... 122

Figure 6.6. Average latencies to eat from the baited cup and percent of total time in the
chamber spent in the goal arm by hippocampal- and sham-lesioned male zebra ﬁnches
over four probe trials. Sham-lesioned birds spent signiﬁcantly more time in the goal arm
over the four probe trials than lesioned birds ................................................... 124

Figure 6.7. Mean (+ SEM) number of incorrect cups examined or ﬂaps of those cups
lifted by hippocampal- and sham-lesioned male zebra ﬁnches over the four probe trials.
Hippocampal-lesioned birds examined signiﬁcantly more incorrect cups or lifted the ﬂaps
of those cups than control birds .................................................................. 125

Figure 6.8. Coronal section through an adult male zebra ﬁnch brain at the level of the

hippocampus detailing areas of cell death (traced by dashed lines) produced by injection
of ibotenic acid at d20. Scale bar = 200 um ................................................... 135

xvii

Figure 6.9. Mean (+ SEM) percent hippocampal damage measured in adulthood in male
and female zebra ﬁnches lesioned at d20, d45 or as adults. Amount of damage to the
hippocampus was determined relative to total hippocampus size in each bird. The effect
of age of lesion on amount of hippocampal damage measured in adulthood approached
signiﬁcance .......................................................................................... 136

Figure 6.10. Mean (+/- SEM) latencies to eat from the baited cup across the four probe
trials for all lesioned and sham-lesioned birds. As a group, lesioned birds differed
signiﬁcantly from sham-lesioned birds, and both groups showed a signiﬁcant decrease in
their latencies across the trials .................................................................... 138

Figure 6.11. Mean (+/- SEM) latencies to eat from the baited cup across the four probe
trials for all male and female birds. The latencies of male birds decreased at a
signiﬁcantly faster rate than those of females .................................................. 139

Figure 6.12. Mean percent time in arm with baited cup by all hippocampal-lesioned and
control birds across the four probe trials. Over the course of the probe trials, sham-
lesioned birds spent signiﬁcantly more time in the goal arms than lesioned birds ........ 140

Figure 6.13. Mean (+ SEM) number of incorrect cups examined or ﬂaps of those cups
lifted in the ﬁrst probe trial by adult male and female zebra ﬁnches lesioned or sham-
lesioned at d20, d45 or in adulthood. Overall, lesioned birds made signiﬁcantly more
mistakes than sham-lesioned ones; there were no signiﬁcant effects of age or sex or an
interaction .......................................................................................... 142

Figure 6.14. Mean (+ SEM) number of incorrect cups examined or ﬂaps of those cups
lifted in the second probe trial by adult male and female zebra ﬁnches lesioned or sham-
lesioned at d20, d45 or in adulthood. Lesioned birds made signiﬁcantly more mistakes
than sham-lesioned ones, and this effect depended on age at time of lesion. Effects of sex
approached significance ............................................................................ 143

Figure 6.15. Mean (+ SEM) number of incorrect cups examined or ﬂaps of those cups
lifted in the third probe trial by adult male and female zebra ﬁnches lesioned or sham-
lesioned at d20, d45 or in adulthood. Lesioned birds made signiﬁcantly more mistakes
than sham—lesioned ones. Effects of age and sex approached signiﬁcance ............... 145

Figure 6.16. Mean (+ SEM) number of incorrect cups examined or flaps of those cups
lifted in the fourth probe trial by adult male and female zebra ﬁnches lesioned or sham-
lesioned at d20, d45 or in adulthood. Lesioned birds made signiﬁcantly more mistakes
than sham-lesioned ones; there were no signiﬁcant effects of age or sex or an

interaction .......................................................................................... 146

Figure 6.17. Mean (+ SEM) difference scores (amount of time (seconds) spent in side of
chamber playing father’s song minus amount of time spent in side of novel male
conspeciﬁc song) of male and female zebra ﬁnches hippocampal-lesioned or sham-
lesioned at 20 days post-hatch (d20), d45 or adulthood. A positive score indicates a

xviii

preference for the father’s song, and a negative score a preference for the song of a novel
male conspeciﬁc. No signiﬁcant main effects or interactions were uncovered ........... 148

Figure 6.18. Mean (+ SEM) difference scores (amount of time (seconds) spent in side of
chamber playing conspeciﬁc song minus amount of time spent in side of heterospeciﬁc
song) of male and female zebra ﬁnches hippocampal-lesioned or sham-lesioned at 20
days post-hatch (d20), d45 or adulthood. No Signiﬁcant main effects or interactions were
uncovered ........................................................................................... 149

Figure 6.19. Spectral derivatives of samples of the songs of a male zebra ﬁnch before (A)
and after (B) a bilateral lesion of the hippocampus. Note that the types of notes used and
the order of those notes are identical before versus after the lesion ........................ 152

CHAPTER SEVEN

Figure 7.1. Drawings of coronal sections through the zebra ﬁnch brain detailing the
structures visible via Nissl stain (panels A and B). In males (A), Area X is evident as a
large nucleus distinct from the surrounding medial striatum (MSt). Area X is not
detectable in the female MSt (B). The remaining panels are photomicrographs of coronal
sections through the MSt of juvenile (d30) zebra ﬁnches detailing ZENK and F OS
immunoreactivity following conspeciﬁc song presentations. Note the almost complete
absence of ZENK immunoreactivity in Area X in males (panel C, dotted box on left)
compared to the relatively homogeneous immunoreactivity throughout the MSt in
females (dotted boxes in panel D). Little to no FOS immunoreactive neurons were
detected in both the male (E) or female (F) MSt. The midline is at the right in each
section. HA = hyperpallium apicale, HD = hyperpallium densocellulare, lMAN = lateral
portion of the magnocellular nucleus of the anterior nidopallium, X = Area X of the
medial striatum (MSt), mMST = medial portion of the medial striatum, lMSt = lateral
portion of the medial striatum. Scale bar = 1.0 mm .......................................... 165

Figure 7.2. Expression of the protein product of the immediate early gene ARC in NCM
(A) and the dorsomedial hippocampus (B). Labeled protein was observed in nuclei
(black arrows) and dendrites (white arrows). Scale bar = 15 pm (A) and 50 um (B)... 169

Figure 7.3. Intrahippocampal and intertelencephalic connections of subdivisions of the
zebra ﬁnch hippocampus compiled from Székely and Krebs (1996; dashed lines) and data
I have collected (Chapter 5; solid lines), detailing potential involvement with the song
system. A known connection between CMM and NCM is indicated by a dotted line; see
Figure 1.1 (Chapter 1) for additional connections among song control regions. A =
arcopallium, CMM = caudomedial mesopallium, DLHP = dorsolateral subdivision of the
hippocampus, DMHP = dorsomedial subdivision of the hippocampus, DMP = posterior
portion of the dorsomedial thalamic nucleus, HA = hyperpallium apicale, HD =
hyperpallium densocellulare, MSt = medial striatum, NCM = caudomedial nidopallium,
SL = lateral septum, SPf = Substance P reactive nucleus, POA = preoptic area, PVN =
periventricular nucleus, Tn = nucleus taeniae, V = lateral ventricle, VHP = ventral
subdivision of the hippocampus. Parentheses enclose structures in which retrogradely
labeled cells were observed in only one bird ................................................... 171

xix

KEY TO ABBREVIATIONS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANOVA analysis of variance

CA(1-3) Comu Amonnis (subﬁelds 1-3)

CMM caudomedial mesopallium

CON conspeciﬁc song

DA tractus dorso-arcopallialis

DLHP dorsolateral subdivision of the hippocampus

DM dorsomedial nucleus of the intercollicular complex
DLM medial part of the dorsolateral thalamic nucleus
DMHP dorsomedial subdivision of the hippocampus

DMP posterior portion of the dorsomedial thalamus
DMSO dimethylsulfoxide

dx number of days post-hatching, with day of hatch as d1
HA hyperpallium apicale

HD hyperpallium densocellulare

HET heterospeciﬁc song

HP hippocampus

HVC formal name for nucleus once referred to as “high vocal center”
IEG(S) immediate early genc(s)

LAD dorsal arcopallial lamina

LaM mesopallial lamina

LFS superior frontal lamina

LLV ventral nucleus of the lateral lemniscus

lMAN lateral portion of the magnocellular nucleus of the anterior nidopallium
mMAN medial portion of the magnocellular nucleus of the anterior nidopallium
MSt medial striatum

NCM caudomedial nidqpallium

NIf nucleus interface of the nidopallium

NMDA N-methyl—D-aspartate

NS no song

nXIIts tracheosyringeal portion of the hypoglossal nucleus
Ov nucleus ovoidalis of the thalamus

PLSD protected least signiﬁcant difference

PVC polyvinyl chloride

RA robust nucleus of the arcopallium

SFL superior frontal lamina

SPf Substance P reactive nucleus

TON randomly generated tones

UVa nucleus uvaeformis

VHP ventral subdivision of the hippocampus

X Area X of the medial striatum

ZENK Zif268, Egr-l, NGFl-A, Krox-24

 

XX

 

CHAPTER ONE

In work ﬁrst published in 1885, Herman Ebbinghaus detailed results from what is
considered the ﬁrst experimental investigation of learning and memory (Ebbinghaus,
1964), shifting away from studies of these psychological processes which were, at the
time, introspective in nature. Using himself as the only subject, he demonstrated that
associative functions could be examined experimentally and subjected to measurement.
This work opened new avenues of learning and memory research that today continue to
intrigue scientists, from a wide array of disciplines, who attempt to understand the
modiﬁcations that occur within the nervous system that result in the maintenance and
expression of behavioral change.

Much of what is known concerning the neural substrates of learning and memory
has resulted from work with mammals, speciﬁcally rodents. Over the past few decades,
how songbirds learn their songs, and the way in which they are able to discriminate
between songs based on memories for them, has led to insights into these processes not
possible with other model systems. Songbird research has provided, among other things,
valuable insight into adult neurogenesis (N ottebohm, 2002b), a striking correlate to the
development of human speech (Kuhl & Doupe, 1999), and a unique comparative model
for the study of brain sexual differentiation (Wade, 1999). Of the songbirds, one of the
most studied is the Australian zebra ﬁnch, T aeniopygia guttata. Zebra ﬁnches belong to
the order Passeriforrnes, which are part of a limited group of vocal learners that also
includes humans, cetaceans, parrots, hummingbirds (Jarvis, Ribeiro, da Silva, Ventura,
Vielliard, & Mello, 2000) and perhaps elephants (Poole, Tyack, Stoeger-Horwath, &

Waitwood, 2005). In their natural habitat, zebra ﬁnches are found living in social groups

in dry, wooded areas (Zann, 1996). In the wild, they breed year-round (although this
depends on rainfall), and parents divide care of the young. They are also hearty breeders
in captivity, making them an excellent model with which to study their behavior and its

development.

Song Learning in Zebra Finches

A male zebra ﬁnch will sing to court a female, and females, who produce only
structurally simple vocalizations, will choose a male to mate with based largely on his
song quality (Zann, 1996). Males learn their songs during a deﬁned stage in development
when, from approximately post-hatch day 25 (d25) to d65, sons listen to their fathers’
songs, memorize characteristics of them, and begin to rehearse their own (Eales, 1985;
Immelmann, 1969; Nordeen & Nordeen, 1997). Beginning around d40 (Nordeen &
Nordeen, 1997), males will rehearse the song template they have memorized, and through
auditory feedback ﬁne-tune it to closely match the song of their tutor (Brainard & Doupe,
2000; Nordeen & Nordeen, 1997). In general, many characteristics of adult male zebra
ﬁnch songs are reﬂected in the songs of their male offspring (BOhner, 1983; Clayton,
1987). Isolation of zebra ﬁnch males prior to the sensitive period and through adulthood
results in the production of an abnormal song (Eales, 1985; Immelmann, 1969). In fact,
the number of song elements a juvenile copies from his father is directly proportional to
the amount of time spent in contact with one another (Eales, 1985). When males are
raised in the presence of a female alone, they begin to produce abnormal song that
contains patterns characteristic of female calls (Eales, 1985). These males, however, can
modify their songs by copying elements from males they encounter after d65 (Eales,

1985, 1987b), even copying over songs already learned (Eales, 1987b). The amount of

change to elements of song depends on the degree of social deprivation during
development (Jones, ten Cate, & Slater, 1996). Also, male zebra ﬁnches, prevented from
comparing their vocalizations to the memory of those from their tutors, can develop
normal song when auditory feedback is reinstated close to maturity (Funabiki & Konishi,
2003). These results suggest that experience outside of the sensitive phase for song
learning can produce behavioral change, and that the sensitive phase in males can remain
“open-ended” if they have not learned song from a suitable tutor.

Female zebra ﬁnches clearly distinguish among songs that are relatively similar,
but it is debatable whether they need to learn conspeciﬁc song during development to
properly respond to it in adulthood. For example, females spend more time near a
speaker broadcasting the song of their mate than the simultaneously presented song of
another male (Miller, 1979a). They can also discriminate between the male songs of two
subspecies, T aeniopygia guttata guttata and T aem’opygia guttata castanotis (Clayton &
PrOve, 1989) and between their fathers’- song and that of another male conspeciﬁc (Miller,
1979b; Riebel, Smallegange, Terpstra, & Bolhuis, 2001). Studies involving isolation
from song during development suggest a potential sensitive period for learning in females
as in males, and this may be dependent upon interaction or experience with their father or
tutor. Females separated from their parents at d35 and presented with a simultaneous
choice between their father’s song and the song of another conspeciﬁc male show a
preference for their father’s song (Miller, 1979b). This preference is not detected in
females acoustically isolated at d25 (Clayton, 1988). In addition, female zebra ﬁnches
housed with only their mothers and tutored with taped male song show stronger and

repeatable preferences for them than untutored females (Riebel, 2000). Cross-fostered

females of T aem'opygia guttata guttata and T aeniopygia guttata castanotis housed apart
from their male siblings after d35 favor songs from males of the same subspecies as their
foster-father, whereas females raised normally prefer the songs of males of their own
subspecies (Clayton, 1990). Furthermore, male and female zebra ﬁnches isolated from
song from d7 and raised by Bengalese ﬁnch foster mothers until d35 show a preference
for hearing conspeciﬁc compared to heterospeciﬁc song when tested during their critical
learning period (between d28 and d44; Braaten & Reynolds, 1999). Thus, while
experience clearly guides the response of females to relatively subtle differences in song,
it is not clear whether exposure to normal vocalizations during development is required
for female zebra ﬁnches to recognize conspeciﬁc song as it is for males to produce it and

when, if at all, a sensitive period for female song learning terminates (Riebel, 2003).

Song Control Regions and Circuits in Zebra Finches

Song behavior in males is regulated by a network of nuclei consisting of a
pathway necessary for song learning interconnected with another important for song
production (Figure 1.1). Recently, the nomenclature for avian telencephalic nuclei was
revised (Reiner, Perkel, Bruce, Butler, Csillag, Kuenzel, Medina, Paxinos, Shimizu,
Striedter, Wild, Ball, Durand, Gﬁntﬁrktin, Lee, Mello, Powers, White, Hough, Kubikova,
Smulders, Wada, Dugas-Ford, Husband, Yamamoto, Yu, Siang, & Jarvis, 2004), and the
new nomenclature is used throughout these chapters. Regions necessary for song
acquisition, the lateral portion of the magnocellular nucleus of the anterior nidopallium
(lMAN) and Area X of the medial striatum (MSt), are located in the anterior forebrain

and show electrophysiological speciﬁcity for song that changes over development

(Doupe, 1997). Lesions of these nuclei during the sensitive period prevent normal song
learning, while similar lesions in adult birds do not affect song (Bottjer, Miesner, &
Arnold, 1984; Scharff & Nottebohm, 1991). Song production is regulated by HVC and
the robust nucleus of the arcopallium (RA), and lesions to these nuclei at any point
disrupt song output (Scharff & Nottebohm, 1991; Simpson & Vicario, 1990). In a
dramatic example of structure mirroring function, these brain regions (with the possible
exception of lMAN) are considerably larger in males who sing than females who do not
(Nottebohm & Arnold, 1976).

In the past decade, the function of regions involved in the perception of song in
zebra ﬁnches has been intensely investigated, made possible primarily by the study of
gene products whose expression can be selectively induced in response to speciﬁc
auditory signals (Clayton, 1997; Mello, 2002). Adult males that perceive their own song
or other males’ songs show upregulation of both the protein and message of two of these
immediate early genes, c-FOS and/or ZENK (an acronym for Zif268, Egr-l, NGFl-A,
Krox-24), in areas outside of the traditional song control nuclei described above. These
regions are the caudomedial nidopallium (N CM) and caudomedial mesopallium (CMM)
(Figure 1.1; Bolhuis, Hetebrij, den Boer-Visser, De Groot, & Zijlstra, 2001; Eda-
Fujiwara, Satoh, Bolhuis, & Kimura, 2003; Jarvis & Nottebohm, 1997; Kruse, Stripling,
& Clayton, 2000; Mello & Clayton, 1994; Mello, Nottebohm, & Clayton, 1995; Mello &
Ribeiro, 1998; Mello, Vicario, & Clayton, 1992; Stripling, Kruse, & Clayton, 2001;
Terpstra, Bolhuis, & den Boer-Visser, 2004). Conversely, when adult male zebra ﬁnches
or canaries sing, neuronal activation is detected in the motor nuclei HVC and RA (Jarvis

and Nottebohm, 1997; Kimpo and Doupe, 1997; J in and Clayton, 1997; Mello and

Figure 1.1 Wiring diagram of structures in the zebra ﬁnch song circuit pulled together
from tract tracing studies, detailing the high degree of known interconnectivities
between the structures (Bottjer et al., 2000; Foster & Bottjer, 1998; Mello et al., 1998;
Striedter & Vu, 1998; Székely, 1999; Székely & Krebs, 1996; Vates et al., 1996; Vates
& Nottebohm, 1995; Vates etal., 1997). Solid lines specify auditory pathways, thick
dotted lines primary motor pathways, and dashed lines song learning pathways. Thin
dotted black lines designate portions of thalamo-cortical feedback loops, and the
dashed + dotted lines indicate the only known connections of the zebra ﬁnch
hippocampus with the regions indicated. DLM = medial part of the dorsolateral
thalamic nucleus, DM = dorsomedial nucleus of the intercollicular complex, DMP =
posterior portion of the dorsomedial nucleus of the thalamus, HP = hippocampus,
lCMM = lateral portion of the caudomedial mesopallium, L1/L2/L3 = subdivisions of
the Field L complex, LLV = ventral nucleus of the lateral lemniscus, lMAN = lateral
portion of the magnocellular nucleus of the anterior nidopallium, mCMM = medial
portion of the caudomedial mesopallium, mMAN = medial portion of the
magnocellular nucleus of the anterior nidopallium, RA = robust nucleus of the
arcopallium, NIf = nucleus interface of the nidopallium, nXIIts = tracheosyringeal
portion of the hypoglossal nucleus, Ov = nucleus ovoidalis, UVa = nucleus uvaeformis,
X = Area X of the medial striatum.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 A
............................... E---------------------------_--------_--_---- -----
F. cochlea
tracheosyringeal nerve

to trachea & syrinx

Ribeiro, 1998). In NCM, heterospeciﬁc song presentations do not provoke such strong
ZENK responses as conspeciﬁc song (Mello et al., 1992). In addition, signiﬁcant
positive correlations exist between the number of FOS- and ZENK-immunoreactive cells
in the NCM of male zebra ﬁnches and the number of song elements copied from a tutor
(Bolhuis et al., 2000; Bolhuis et al., 2001). The NCM of males therefore appears to be
activated by speciﬁc acoustic signals that are particularly biologically-relevant.

Although females perceive subtle differences among vocalizations of male
songbirds (see above), the areas of the brain responsible for this perception have not been
intensely investigated. Areas traditionally assigned to male song behavior seem to play
some role in song perception. For example, lesions of HVC in female canaries render
them unable to distinguish conspeciﬁc from heterospeciﬁc song (Brenowitz, 1991) and to
show a strong preference for conspeciﬁc song (Del Negro, Gahr, Leboucher, & Kreutzer,
1998). In contrast, in the zebra ﬁnch, lesions of HVC do not modify song preferences in
females, but destruction of the neighboring CMM results in the display of Similar
responses to both heterospeciﬁc and conspeciﬁc song (MacDougall-Shackleton, Hulse, &
Ball, 1998). Structures in the female songbird brain that show immediate early gene
activation following song presentations have been investigated in a few studies. ZENK-
immunoreactive cells in the NCM of female European starlings are signiﬁcantly
increased following exposure to song containing longer bouts; activation in the medial
portion of CMM is uniformly high regardless of the length and number of song bouts
presented (Gentner, Hulse, Duffy, & Ball, 2001a). Like males of this species, conspeciﬁc
song presented to female European starlings results in an increase in the number of

ZENK-immunoreactive cells in the NCM and CMM, expression that does not vary with

photoperiodic condition (Duffy, Bentley, & Ball, 1999). In female zebra ﬁnches, ZENK
induction in the NCM following song presentation is qualitatively similar to that seen in
males (Mello et al., 1992).

Work I have done (Bailey, Rosebush, & Wade, 2002) has provided a more
detailed analysis of the neural response to song in adult female zebra ﬁnches. A higher
density of F OS-immunoreactive nuclei is seen in the NCM of females following
stimulation with conspeciﬁc compared to heterospeciﬁc song (Figure 1.2). This result
parallels increased ZENK mRNA in the NCM of male zebra ﬁnches and canaries
following conspeciﬁc versus heterospeciﬁc song presentations (see above). Interestingly,
this pattern is also identiﬁed in the hippocampus (Figure 1.3). In CMM, FOS activation
is equivalent following zebra ﬁnch and heterospeciﬁc song, yet both are increased
compared to presentations of tone and silence (Figure 1.4). The CMM thus acts perhaps
as an initial ﬁlter and the NCM likely provides a more detailed analysis of potential
biological relevance. The function of the hippocampus in song behavior is unknown and

warrants further attention.

A_v_iaa_ and Mammalian Hippocampal Function

The song-speciﬁc responses of a group of neurons within the hippocampus
indicate a potential role for the structure in the processing of auditory information or song
memories in zebra ﬁnches. The avian hippocampus appears to be homologous to the

mammalian hippocampus based on numerous structural and functional factors and its role

 

200 - NCM

FOS
immunoreactive cells/mm2

 

 

 

 

Figure 1.2. Density of FOS-immunoreactive cells (mean + SEM) in the NCM for adult
females exposed to conspeciﬁc (CON) and heterospeciﬁc (HET) song, tones (TON)

or no song (NS) (top panel). Different lowercase letters indicate signiﬁcant differences
between groups. The bottom panels contain photomicrographs of coronal sections
through the zebra ﬁnch brain at the level of the NCM, showing FOS immunoreactivity
in each of the stimulus conditions. Scale bar = 100 pm.

 

NE 225
\ 200 - HP
.3 175 -
8 150 -
CD 4; 125 -
E ‘53 100- a
8 75 ‘
o 50 -
G
E 25 - b b b
0 .

 

 

 

CON HET TON NS

 

Figure 1.3. Density of FOS-immunoreactive cells (mean + SEM) in the
hippocampus (HP) of female birds in the conspeciﬁc (CON), heterospeciﬁc (HET),
tone (TON) and no song (NS) conditions (top panels). Different lowercase letters
indicate signiﬁcant differences between groups. Photomicrographs indicate FOS-
immunoreactive cell nuclei for each group in the HP. Scale bar = 100 um.

 

FOS
immunoreactive cells/mm2

 

 

 

CON HET TON NS

 

Figure 1.4. Density of FOS-immunoreactive cells (mean + SEM) in the CMM for adult
females in the conspeciﬁc (CON), heterospeciﬁc (HET), tone (TON) and no song (NS)
conditions. Different lowercase letters indicated signiﬁcant differences between
combined groups (“song” versus “no song”). Photomicrographs of sections at the level of
the CMM showing FOS immunoreactivity in response to each of the stimulus conditions
are shown in the bottom panel. Scale bar = 100 um.

12

in spatial memory consolidation and retrieval (reviewed in Colombo & Broadbent, 2000;
Macphail, 2002). The recovery of stored food is common in many avian species,
speciﬁcally the parids and corvids (Kamil & Balda, 1990; Shettleworth, 1990).
Remembering locations of food caches seems to be dependent on the hippocampus. For
example, a bilateral lesion to the hippocampus in a black—capped Chickadee reduces the
accuracy of cache recovery but not the amount of caching nor attempts to ﬁnd the stored
food (Sherry & Vaccarino, 1989). Hippocampal volume in homing pigeons, which are
able to navigate great distances back to their home lofts, is greater than that of a non-
homing breed of pigeon (Rehkamper, Hasse, & F rahm, 1988). There are several
exceptions, however, to the positive correlation between the size of the hippocampus and
spatial memory ability (Bolhuis & Macphail, 2001). Lesions to the hippocampus produce
deﬁcits in recognition of the home loft and of surrounding familiar landmarks by homing
pigeons; orientation back home is normal but slower (Bingman, Ioalé, Casini, & Bagnoli,
1988; Gagliardo, Ioalé, & Bingrnan, 1999; Strasser & Bingrnan, 1996).

In zebra ﬁnches, lesions of the hippocampus result in signiﬁcant spatial memory
impairment (Patel, Clayton, & Krebs, 1997a; Watanabe & Bischof, 2004), and
transplantation of embryonic hippocampal tissue reverses this deﬁcit (Patel et al., 1997a).
Considering its importance in memory, the results I have Obtained (Bailey et al., 2002)
suggest that the hippocampus may somehow be involved in the formation or utilization of
song memories or perhaps auditory perception. One possibility is that neuronal
activation in the hippocampus may reﬂect consolidation of a “contextual song memory,”
the encoding of the environmental context in which biologically-relevant song is heard.

Substantial hippocampal activity is not observed in animals that hear heterospeciﬁc songs,

13

tones or silence (Bailey et al., 2002; Bailey & Wade, 2003), which suggests that the
hippocampus is not encoding information about a place speciﬁcally, but perhaps a
relevant event that is occurring in that place. In the case of female zebra ﬁnches, this
activity may result in an increase in preparedness to respond to male song in this context
in the future. Although documented but not discussed in detail (Kimpo and Doupe, 1997;
Bolhuis et al., 2000), activation in the hippocampus of males in response to song may
reﬂect a comparable process that in this case results in future readiness to defend a nest
site, for example. Although the hippocampus is a key component of the stress response
(McEwen, 1999), it is unlikely that the immediate early gene activation observed was
stress-related, such as that associated with removal from the aviary, handling by the
experimenter, or exposure to a novel environment, since it would have been observed in
the hippocampus in all birds. In addition, in rodents, c-FOS induction in the
hippocampus is not inﬂuenced by the degree of a particular stressor but by the animal’s
ability to explore a novel environment, which ampliﬁes the magnitude of spatial
processing by the hippocampus and in turn the immediate early gene response (Pace,
Gaylord, Topczewski, Girotti, Rubin, & Spencer, 2005).

This possible function of the zebra ﬁnch hippocampus - the integration of related
events or stimuli in the environment - is consistent with a well-studied role of the
hippocampus in mammals. The most dramatic and one of the earliest examples came
from a report on, among others, a man named HM, whose temporal lobes were surgically
removed to end his frequent epileptic seizures (Scoville & Milner, 1957). Following the
bilateral lobectomy, HM was unable to recollect new people, places or things he

encountered, and still today is unaware that he has become one of the most famous

14

subjects in the history of psychology. Shortly after the publication of HM’s case, it was
believed that the hippocampus was some sort of all-encompassing learning structure. It
was later learned that the tissue surrounding the hippocampus in addition to the region
itself that were removed from HM’s brain likely caused the profound memory deﬁcit
(Squire, 1992), but there remains no doubt about the inﬂuence the hippocampus has on
speciﬁc types of memory.

In the 19703, the theory that the hippocampus may contain a “cognitive map
(Tolman, 1948)” was put forward based on results that showed the activity of cells in the
structure of freely moving rats was closely related to the animal’s location in an open
ﬁeld (O'Keefe & Dostrovsky, 1971). Later, it was proposed that different sets of these
“place cells” within the hippocampus represented a speciﬁc region of space (O'Keefe &
Nadel, 1978). It is clear that the hippocampus contains cells responsive to an animal’s
place in its environment, and since the initial discovery a large number of studies have
corroborated the original ﬁnding (Best & White, 1998; Eichenbaum, Dudchenko, Wood,
Shapiro, & Tanila, 1999). However, contemporary theories of hippocampal function
suggest that the structure may have a more general function in forming complex
conﬁgural associations between various environmental stimuli, and these associations
need not necessarily be spatial in nature. For example, the hippocampus is integral in the
consolidation of the relationship between an unconditional stimulus (such as foot shock)
and a novel environment (conditional stimulus; Kim & Fanselow, 1992; Phillips &
LeDoux, 1995), the detection of novel stimuli (Knight, 1996), and the conveyance of an
animal’s location and what is currently present in that environment (Leutgeb, Leutgeb,

Barnes, Moser, McNaughton, & Moser, 2005). It is now widely believed that the

15

hippocampus consolidates memories not only for places and events but also the temporal
sequencing of those events (reviewed in Eichenbaum et al., 1999), processing which may
(Clark & Squire, 1998) or may not (Chun & Phelps, 1999) require conscious awareness

of cues and the context in which they are presented.

Comparison of Hippocarmial Connectivity in Mammals and Birds

The mammalian hippocampus exhibits a set of three connected pathways known
as the "trisynaptic" circuit (Amaral & Witter, 1989; Eichenbaum, Schoenbaum, Young,
& Bunsey, 1996). In the ﬁrst portion of this circuit, cells within the entorhinal cortex,
which collect information from limbic association cortex and other association areas
(Amaral & Witter, 1989; Eichenbaum et al., 1996), project to granule cells of the dentate
gyrus via the perforant path. Second, cells of the dentate gyrus project to the large
pyramidal cells of Comu Amonnis or Ammon's horn, subﬁeld 3 (CA3), via mossy ﬁbers.
Third, pyramidal cells in CA3 project to pyramidal cells of the CA1 subﬁeld, via the
Schaffer collateral system (Amaral & Witter, 1989). The output neurons of the
hippocampus, primary pyramidal neurons of CA1, form a ﬁber bundle called the fomix
that projects primarily to the mammillary bodies and septa] nuclei (Amaral & Witter,
1989; Eichenbaum et al., 1996).

Intra- and interhippocampal connectivity in avian species may not be so different
from that in mammals. In birds, the hippocampus (often cautiously referred to as the
avian hippocampal complex or formation; Colombo & Broadbent, 2000) sits on the
dorsal surface of the brain near the cerebellum. In the pigeon, sensory input from

primarily the piriform cortex innervates the dorsolateral and dorsomedial subdivisions of

16

the hippocampus (Kahn, Hough 11, Ten Eyck, & Bingrnan, 2003), the latter of which is
subdivided into dorsal and ventral portions. Cells in the dorsal dorsomedial portion then
project to ventrolateral, -central and -medial subdivisions, which project to the same
subdivisions contralaterally and to cells of the ipsilateral ventral dorsomedial subdivision,
whose axons exit the hippocampus. This network of connections in the pigeon
hippocampus is comparable to the trisynaptic pathway in the mammalian hippocampus
described above (Hough II, Pang, & Bingman, 2002).

The intratelencephalic connectivity of the homing pigeon hippocampus has
recently been determined using anterograde and retrograde tracers (Atoji, Wild,
Yamamoto, & Suzuki, 2002). Ipsilaterally, the hippocampus has reciprocal connections
with the parahippocampal area, lateral septum, and nucleus taeniae (homologous to a
portion of the mammalian amygdala; see below). Contralaterally, a unidirectional
projection is observed from the hippocampus to the parahippocampal area, lateral septum,
and MSt, and a reciprocal connection spans from the ventral portion of the hippocampus
on one side and through the pallial commissure to the other.

Until recently, no clear anatomical lateral border between the avian hippocampus
and surrounding tissue had been recognized (Kahn et al., 2003; Székely, 1999). A
nucleus lateral to the hippocampus in pigeons has been identiﬁed (Erichsen, Bingrnan, &
Krebs, 1991). Named SPf due to its high density of Substance P-immunoreactive
neurons, its position relative to the hippocampus medially and the hyperpallium apicale
(HA) laterally makes it a potential homologue to the mammalian entorhinal cortex (Kahn
et al., 2003). Retrogradely labeled neurons within the zebra ﬁnch SPf were observed

following injections of the anterograde tracer Phaselous vulgaris leucoagglutinin into the

17

dorsolateral hippocampus (Székely, 1999). In addition to these neurochemical markers,
the lateral and anterior boundaries of the hippocampus can easily be visualized using a
histochemical marker such as a Nissl stain, which reveal dramatic increases in cell
density at the boundaries of the structure.

The efferent connections of the zebra ﬁnch hippocampal complex have been
determined through injections of P. vulgaris (Székely & Krebs, 1996). The results of this
study mirror the efferent projections and intraconnectivity of the hippocampal complex in
pigeons (Atoji et al., 2002; Casini, Bingman, & Bagnoli, 1986; Krayniak & Siegel, 1978),
the main difference being only three subregions were identiﬁed based on characteristic
projections patterns: dorsolateral, dorsomedial, and ventral (Székely & Krebs, 1996).
Tracer infusion into the dorsomedial subdivision indicated almost exclusively internal
afferent projections. Terminating axons resulting from injections into the ventral region
were found in the medial septum in both hemispheres and also in the contralateral
hippocampus (Székely & Krebs, 1996). Dorsolateral injections of the tracer resulted in
mainly ipsilateral projections, which terminated in the lateral hypothalamus, areas of the
septum, nucleus taeniae, posterior portion of the dorsomedial thalamic nucleus, caudal
arcopallium and the MSt. Innervation of the caudal arcopallium may cover areas such as
the “cup” of RA that receives substantial descending auditory information from HVC
(Mello et al., 1998). It is also interesting to note that the RA and hippocampus both send
projections to the posterior portion of the dorsomedial thalamic nucleus (DMP; Vates et
al., 1997; Székely & Krebs, 1996), an area that, together with traditional song motor
control nuclei (HVC and RA), may coordinate the temporal organization of song (V ates

et al., 1997). Nucleus taeniae has been shown, at least in Japanese quail, to be involved

18

in the control of sexual behavior (Absil, Braquenier, Balthazart, & Ball, 2002). In
addition, the septum has been implicated in spatial memory and motivated behaviors in
avian species (Shiﬂett, Gould, Smulders, & DeVoogd, 2002). Thus, connections exist
that could allow the hippocampus to be involved in the processing of memories related to
responses to song. However, afferents to the hippocampus must be determined to fully
understand how the structure may interact with other regions known to play a role in song

perception or possibly production.

Prospectus

The zebra ﬁnch provides an ideal model system for examining the development of
neural mechanisms fundamental to multiple memory systems that control discrete
behaviors. Relatively few studies have addressed the role of learning in the development
of female song perceptions. To further examine this, the experiment in Chapter Two
examines whether female zebra ﬁnches, which grew to adulthood without hearing their
father’s song throughout the majority of development yet remained in social contact with
both parents, show a preference for conspeciﬁc versus heterospeciﬁc song. The
remainder of the experiments tests the hypothesis that the hippocampus in both male and
female zebra ﬁnches is important for the consolidation, retrieval or maintenance of song
templates. Although much is known about the neural substrates of auditory perception in
adult songbirds, less is known about its development. Speciﬁcally, if the hippocampus is
important for the storage and/or retrieval of song memories, then the patterns or
distribution of neuronal activity indicated by immediate early gene expression in the

hippocampus and areas important for perception should increase or change during

19

development as song templates are acquired. The experiments in Chapters Three and
Four were designed to identify the brain regions in juvenile zebra ﬁnches (at two separate
and central developmental stages) that show immediate early gene activity in response to
auditory stimulation, whether there are sex differences in that activation, and whether the
immediate early genes examined are differentially expressed.

Further, if the hippocampus is important for the storage and/or retrieval of song
memories, it should have connections with areas involved in song perception in both
sexes and perhaps song production areas in males, and destruction of hippocampal tissue
should impact song consolidation memories or perception as measured in adult males and
females. Chapter Five details the afferent connectivity of the hippocampus in male and
female zebra ﬁnches, including the areas with which it has reciprocal connections,
speciﬁcally including the regions implicated in song learning, production or perception.
Finally, Chapter Six evaluates the effects of hippocampal lesions in male and female
zebra ﬁnches at the beginning of the song acquisition and sensorimotor integration
periods (both as deﬁned in males) and in adulthood, to determine the contribution of the
structure to song learning (males and females), initial production and maintenance of
learned song patterns (males), and spatial memory as a control (males and females). If
the hippocampus is important in song learning, then males lesioned at the beginning of
the template acquisition period should produce abnormal song as adults. Decrements in
adult song production should be observed if the hippocampus is integral in the auditory
feedback necessary for the ﬁne-tuning of song that begins at sensorimotor integration. If
the structure is important for maintenance of the song template, then lesions of the

hippocampus in adulthood should result in abnormal song. These possibilities are not

20

mutually exclusive. Given the song-speciﬁc induction of immediate early genes I
observed in the hippocampus (Bailey et al., 2002), lesions of the structure may result in

an inability to properly respond to conspeciﬁc songs in adulthood. If the hippocampus is
important in song learning, then lesions during development should impact consolidation
of a song template and thus inﬂuence song preference in adulthood. As mentioned above,
if the region is important in maintenance of a song template, then lesions in adulthood
should result in abnormal song preferences. This last experiment only begins to test the
importance of the hippocampus in auditory learning in zebra ﬁnches; it does not test
whether activity within the structure is due to consolidation of a “contextual song

memory,” which links a relevant auditory stimulus with a novel or familiar environment.

21

CHAPTER TWO

Song exposure during development modiﬁes adult behavior following song perception in
the female zebra ﬁnch

INTRODUCTION

The development of responses by female birds to the songs Of males has been
described as a “neglected” but important area of study in bird song research and vocal
communication in general (Slater, 2003). It is clear that female zebra ﬁnches
(T aem'opygia guttata) can distinguish among relatively similar songs. These birds, who
produce only structurally simple vocalizations, spend more time near the songs of their
mates than the simultaneously presented song of other conspeciﬁc males (Miller, 1979a).
They can discriminate between the male songs of two subspecies, T aeniopygia guttata
guttata and T aeniopygia guttata castanotis (Clayton & PrOve, 1989), and between their
fathers’ songs and those of other zebra ﬁnch males (Miller, 1979b; Riebel, Smallegange,
Terpstra, & Bolhuis, 2002). But, females may not detect speciﬁc features of zebra ﬁnch
song as males do. They require more trials than males to distinguish between songs of
males from their own aviary (Cynx & Nottebohm, 1992), as well as songs played in
reverse (Cynx, 1993), and their preference for their fathers’ songs does not generalize to
the songs of male siblings, which share characteristics of the fathers’ songs (Riebel &
Smallegange, 2003).

Juvenile exposure to song is critical for normal masculine song development.
Isolation from song as juveniles dramatically alters the structure of adult vocalizations
produced by many species of songbirds (Marler, 1997). In zebra ﬁnches, song learning
occurs during a deﬁned stage in development when, from approximately post-hatch day

25 (d25) to d65, sons listen to their fathers’ songs, memorize characteristics of them, and

22

begin to rehearse their own (Eales, 1985; Immelmann, 1969; Nordeen & Nordeen, 1997).
In general, many characteristics of adult male zebra ﬁnch songs are reﬂected in those of
their male offspring (BOhner, 1983; Clayton, 1987). Isolation of zebra ﬁnch males prior
to the sensitive period and through adulthood results in the production of an abnormal
song, which is “slow” and contains patterns uncharacteristic of normal adult zebra ﬁnch
song (Eales, 1985; Immelmann, 1969). Indeed, the number of song elements a juvenile
copies from his father is proportional to the amount of time spent in contact with one
another (Eales, 1985). When males are raised in the presence of a female alone, they
begin to produce abnormal song that contains patterns characteristic of female calls
(Eales, 1985). These males, however, can modify their songs by copying elements from
males they encounter after d65 (Eales, 1985, 1987b), and the amount of change to
elements of song depends on the degree of social deprivation during development (Jones
et al., 1996). These studies suggest that experience outside of the sensitive phase for
song learning can produce behavioral change, and that this period can remain “open-
ended” to some extent if song was not learned from a suitable tutor. Recent work also
shows that male zebra ﬁnches, prevented from comparing their vocalizations to the
memory of those from their tutor(s), can develop normal song when auditory feedback is
reinstated close to maturity (Funabiki & Konishi, 2003).

The early stages of song learning (template formation beginning around d25) may
be similar in the two sexes (for review, see Riebel, 2003). Females separated ﬁ'om their
parents at d35 and as adults presented with a simultaneous choice between their father’s
song and the song of another conspeciﬁc male show a preference for their father’s song

(Miller, 1979b). This preference is not detected in females acoustically isolated at d25

23

(Clayton, 1988), suggesting that aspects of song learning may occur from d25 to d35.
Similar results are obtained when preference for “super-normal length” zebra ﬁnch song
is measured. Female zebra ﬁnches separated from their parents at d25 show an equal
amount of time near normal and super-length song, whereas separation at d35 (or d70)
results in the normal adult preference for the longer songs (N eubauer, 1999).
Additionally, females housed with only their mothers and tutored with taped male song
show stronger and more repeatable preferences for them than untutored females (Riebel,
2000)

While experience shapes the responses of females to relatively subtle differences
in song, it is not clear whether exposure to normal vocalizations during development is
required for adult female zebra ﬁnches to recognize conspeciﬁc song as it is for males to
produce it. A recent study documented that female zebra ﬁnches spend more time in the
side of a chamber near conspeciﬁc compared to canary song, regardless of whether they
were raised with males (Lauay, Gerlach, Adkins-Regan, & DeVoogd, 2004). This result
is consistent with the existence of an inherent preference for at least some characteristics
of zebra ﬁnch song, although other possibilities exist. The lack of preference in females
raised without adult males does not appear to be inﬂuenced by the songs of male siblings,
suggesting either that templates may be formed by females during a sensitive period
terminating before their brothers begin to vocalize (around d40) or that untutored male
song is insufﬁcient to stimulate development of this female characteristic. In a similar
study, both male and female zebra ﬁnches isolated from song from d7 and raised by
foster mothers until d35 showed a preference for hearing conspeciﬁc compared to starling

song as juveniles (Braaten & Reynolds, 1999). In both of these studies, however, the

24

responses were compared between zebra ﬁnch song and that of only one other species.
Thus, it is not clear whether the preference would broadly apply. In addition, the effects
of adult experience were not taken into account. Braaten and Reynolds (1999) did not
test mature animals, so it is unclear whether the preference they discovered would be
maintained at a time when females choose mates. In Lauay et al. (2004), the test for
conspeciﬁc versus heterospeciﬁc song was conducted after the females (even those who
developed without song exposure) heard some song as adults. Their behavior may
therefore have been inﬂuenced by recent exposure.

While this general question (whether female zebra ﬁnches need to learn the song
of their species to respond selectively to it in adulthood) has been investigated twice, the
present study was designed with methodological differences as follows. To address the
points above, adult responses to several songs of each type (zebra ﬁnch and
heterospeciﬁc) were examined alternately over the testing days in order to determine
whether responses would generalize (Kroodsma, Byers, Goodale, Johnson, & Liu, 2001;
Kroodsma, 1989). In addition, since zebra ﬁnches are extremely social (Zann, 1996), and
prohibiting contact with parents and siblings during development impacts auditory
discrimination capability in adulthood (Sturdy, Phillmore, Sartor, & Weisman, 2001),
females were kept in social contact with their fathers through what is typically considered
the sensitive period for template formation (see above). Rather than removing fathers,
the tracheosyringeal nerves were severed so they could not sing. Finally, immediate
early gene activation is increased in the hippocampus as well as auditory areas in adult
and juvenile female zebra ﬁnches following conspeciﬁc song exposure (Bailey et al.,

2002; Bailey & Wade, 2003). This result is consistent with the idea that females encode

25

information about where a favorable song was heard. Therefore, in addition to testing
responses during song, the “contextual” memory of these birds was examined by
observing their responses to the environment in which conspeciﬁc and heterospeciﬁc

song was heard the day before.

MATERIALS AND METHODS
Animals and Housing

All zebra ﬁnches were obtained from the breeding colony at Michigan State
University. Prior to group assignment, birds were housed in communal aviaries
consisting of approximately seven breeding pairs and their young. Nest boxes in the
aviaries and used in the testing chamber were constructed of polyvinyl chloride (PVC),
and were 12.7 cm high and 12.7 cm in diameter with a 3.8 cm diameter hole 2.5 cm from
the top. A 5 cm long perch extended outward below this hole, and the top of each nest
box was covered with a removable black Plexiglas disc. Seed and water were provided
ad libitum, and supplements of Spinach or oranges and a mixture of hard-boiled chicken

eggs and bread were provided weekly. The lightzdark cycle was 12: 12 (lights on at 0700).

Experimental Manipulation

As eggs batched in nest boxes in the main bird colony, the mother and father were
determined by daily observation until the identity of the male and female that consistently
entered and spent time in each nest box could be conﬁrmed. Two to four days before a
nest box was to be moved (see below), fathers were temporarily removed and their song

recorded in the presence of a novel female. An axotomy or sham surgery was then

26

performed depending on whether the nest box was to be moved to the song isolation or
control condition, respectively. Each father was anesthetized with Metofane, and an
incision was made in the skin above the trachea. For males entering the song isolation
condition, the tracheosyringeal portions of the hypoglossal nerves (XIIts) were gently
separated from the trachea and 4-6 mm was removed bilaterally. Fathers entering the
control condition underwent a sham surgery in which the nerves were gently separated
from the trachea but not severed. The axotomy produced complete cessation of normal
song production, and resulted in sounds associated only with respiration (Figure 2.1).
Following recovery from surgery, each animal was returned to its aviary. All fathers of
experimental females in this study showed no apparent difﬁculty in breathing, or any
signs of poor health, and all quickly resumed care of their young.

After all of the viable eggs hatched and before the ﬁrst hatchling reached d8, the
nest box and parents were moved to the control condition (to a separate aviary in the
main bird room with similar groups; n = 11 females from 3 different clutches) or to the
song isolation condition (to a separate room that housed only females, axotomized fathers
and their offspring; n = 10 females from 4 clutches). Around d40, when the plumage of
zebra ﬁnches was developed enough to allow distinction between males and females,
males raised in the song isolation condition were removed before they began to sing
(Immelmann, 1969; Nordeen & Nordeen, 1997). At this time both parents were also
removed, as were male Siblings and parents of control animals. Each father in the song
isolation condition was given an overdose of Equithesin following the recording of

vocalizations in the presence of a female. None produced vocalizations resembling song

27

 

6500

 

E 5200 'A‘ ,: »‘» - x-I’FH l ‘s‘ «3 st:
a, ; .—..,,:u_u_ iﬁlﬁh' g;
..c: , . . . e " .
E, 3900 a t gljgg’f‘fiil} ,3
s: ' - z l_ .. ‘i‘
S 2600 E ' . '
LL 1300 -._ f _ ”1‘2." 1 _ ‘ ~- ~ " ‘_ ~.
0 113.4; «Jame ”a
6500
A 5200- B ' ‘ ' '
E
; 3900-
2 n: In t J 1’ Jr: ”A J:
P 2600- ‘ -
35 {I «,4: 3; “A :41 vi! a
a 1300- A A A _/‘_ A if! ‘A ,A
.A J‘ .A -.J‘ J‘ .-J‘ ..J‘ .I‘

 

 

 

0 " - ' - - v ' ' . - - . ‘
0 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50

Time (seconds)

Figure 2.1. Sonagram of vocal behavior in the presence of a novel female
before (A) and after (B) severing of the tracheosyringeal nerves of a father
of birds in the song-isolated group. While rhythmic tones associated with
respiration were consistently heard, song was never detected following
axotomy.

28

(Figure 2.1). At this time, the tracheosyringeal nerves of axotomized fathers were
checked. An average of at least 3-4 mm remained between the severed ends of each

nerve; in no case did they regenerate enough to make contact with one another.

Behavioral Testing

The testing apparatus (Figure 2.2), located in a quiet room away from the main
bird colony, was constructed of a wood frame, steel mesh sides, back and top, and a
Plexiglas front through which observations and video tapes were made. Sound
attenuating foam covered the walls surrounding the entire apparatus. Two sets of three
perches were located at each end of the chamber, and a single perch was in the middle.
Nest boxes were located at each end of the chamber. Hidden in each nest box were
speakers that delivered song.

Female-directed zebra ﬁnch songs were recorded from males from our colony.
Heterospeciﬁc songs were obtained from The National Geographic Society and Cornell
Laboratory of Omithology’s Guide to Western Bird Songs. Songs from Six different
species (Bell’s vireo, Cassin’s ﬁnch, marsh wren, American robin, summer tanager,
white-breasted nuthatch) were recorded from this guide and were chosen based on their
overlapping frequency and similarity in bout length to zebra ﬁnch song. Speakers,
attached to a PC, delivered song stimuli at 72 dB at a distance of 12 cm from each. Real
Jukebox (version 1.0.0.488) controlled song delivery. Auditory responses to song
presentations were recorded using a portable cassette recorder (Marantz, Aurora, IL)

connected to an Optimus boundary microphone (Radio Shack), and a video camera

29

  

Figure 2.2. Chamber used for song presentations. The apparatus was
constructed of a wood frame, steel mesh sides, back, and top, and a
Plexiglas front through which observations and video recordings were
made. Two sets of three wooden perches were located at each end of the
chamber, and a single perch was in the middle. The middle perches in the
sets of three were located 20.2 cm above the ﬂoor of the chamber, and
were spaced 6.3 cm from the adjacent perches. All other perches were 30
cm above the ﬂoor of the chamber. Nest boxes were located at each end
of the chamber. Nest boxes were constructed of polyvinyl chloride
(PVC), and were 12.7 cm high and 12.7 cm in diameter with a 3.8 cm
diameter hole 2.5 cm from the top. A 5 cm long perch extended outward
at 900 below this hole, and the top of each nest box was covered with a
removable Plexiglas disc. Hidden in each nest box were speakers that
delivered song, and sound attenuating foam covered the walls surrounding
the entire apparatus.

30

recorded movements. Fluorescent room lights were turned off during testing, and 60 watt
incandescent light bulbs in metal reﬂectors located 30.5 cm above each end of the
chamber were illuminated. All testing was done during the light portion of the lightzdark
cycle.

To ensure similarity in recent history between the two groups, animals in the
control condition were moved in with the song isolation group when the youngest female
in each clutch was d80. As adults (mean = d121.2 +/- 2.3), females in each group
underwent behavioral testing. At the beginning of a session, a female from the song
isolation or control condition was placed through a door in the center of the chamber
described above, facing the back wall. Following a 3 minute acclimation period,
conspeciﬁc or heterospeciﬁc song was randomly delivered as follows. Each session (one
a day for twelve days) consisted of one conspeciﬁc or heterospeciﬁc song played from
the right or left speaker. All song ﬁles were 30 seconds in length (to maximize
consistency across song exposures) and were repeated six times, for a total presentation
time of 3 minutes. Over the course of the testing sessions, each female was exposed to
all six conspeciﬁc and six heterospeciﬁc songs (Kroodsma et al., 2001; Kroodsma, 1989),
with the order randomly chosen before each session. Whether conspeciﬁc or
heterospeciﬁc song was played on the ﬁrst testing day was counterbalanced, as was initial
side of delivery (right or left speaker). Birds heard each of the twelve songs once, and
conspeciﬁc and heterospeciﬁc song were alternated on testing days. Three minutes after
song presentations, birds were removed from the testing apparatus and were immediately

returned to the group aviary.

31

Data Analysis

Birds primarily responded with calls and movements; few other behaviors were
observed. Long and short calls were tallied (see Zann, 1996 for a description of call
types). If a bird turned 180° to face the opposite direction, this was scored as a “turn.”
“Perch jumps” were deﬁned as movement from one perch to another, a ﬂight from a
perch and a return to that same perch, or a ﬂight from a perch to the top of a nest box or
vice versa. As indicated above, the data were analyzed as responses during, immediately
after, and 24 hours after zebra ﬁnch and heterospeciﬁc song presentations. Since
conspeciﬁc and heterospeciﬁc song delivery was counterbalanced, some birds were
presented with zebra ﬁnch song on testing day 1 and others ﬁrst heard it on testing day 2.
Thus, while all birds were presented with six songs of each type, data from only one-half
of the animals in each group were available for the day after the sixth zebra ﬁnch or
heterospeciﬁc song presentation, so those data were discarded.

Data were scored live, as well as from video and audiotapes by an observer not
aware of treatment conditions. All data reported are from the tapes, except for that ﬁ'om
four females for one testing session in which technical difﬁculties precluded electronic
data collection. Tests for skewness and kurtosis revealed the data were not normally
distributed, so square root transformations were performed. All statistical analyses were
carried out on the transformed data, although the means and standard errors depicted are
derived from the raw data. Effects of developmental condition (control versus song
isolation; between individual birds), stimulus type (conspeciﬁc versus heterospeciﬁc song;
within individuals), time (across ﬁve tests; within individuals) and period (during,

immediately after, or 24 hours after song; within individual birds) were assessed using

32

analyses of variance (ANOVA). When signiﬁcant effects were detected, pairwise
planned comparisons were used (Maxwell, 1980). Statistical analyses were performed

using Statview and SAS (both SAS Institute).

RESULTS

The statistical signiﬁcance of effects reported for the overall ANOVA were
similar regardless of whether the number of calls or the amount of movement (turns and
perch jumps), or the sum of all of these behaviors was used, except in three instances.
These differences are indicated below, but otherwise for simplicity, data are reported only
for the total of calls plus movements.

During the 3 minute acclimation period before their ﬁrst song exposure, the
number of calls and movements by each group was equivalent and relatively low,
although a trend existed for song-isolated birds to be more active (t(l9) = 1.92, p = 0.070;
Figure 2.3). The quantity of responses differed dramatically across the periods in which
they were assessed (while song was playing, immediately after song, or 24 hours after;
5(2, 38) = 89.15, p < 0.0001; Figure 2.4). Calls and movements during song (Figure
2.4A) were two times greater than those immediately after (Figure 2.4B; 1(20) = 4.33, p =
0.0003). Interestingly, responses 24 hours after song presentation were approximately
four times more than those during (t(20) = 7.68, p < 0.0001) and 8 times those
immediately after (t(20) = 12.74, p_ < 0.0001) song (Figure 2.4C). Across the 5 analyzed
tests, responses also differed (E (4, 76) = 5.39, p_ = 0.0007), and this variable of time

interacted with period when the sum of the behaviors were considered (total responses:

33

Mean (+ SEM) calls and movement

25

 

20-

 

 

 

 

 

 

 

- CONTROLS (n = 11)
1:] SONG-ISOLATES (n = 10)

Figure 2.3. Mean (+ SEM) calls plus movement before song exposure on
the ﬁrst testing day. These baseline values were not signiﬁcantly different,
although the song isolates appeared somewhat more active.

34

Figure 2.4. Mean (+ SEM) calls plus movement during ﬁve conspeciﬁc and
heterospeciﬁc song presentations (A), immediately after those songs (B), and 24 hours
after (C) by females raised with exposure to zebra ﬁnch song (CONTROLS) or isolated
from song early in life through adulthood (SONG ISOLATES). Responses 24 hours
following the ﬁrst conspeciﬁc and heterospeciﬁc songs are highlighted by dashed lines.
The “*” indicates a signiﬁcant difference between the groups. Note differences in the
Y-axis scales.

35

Mean (+SEM) calls and movement

Mean (+ SEM) calls and movement

Mean (+ SEM) calls and movement

During conspeciﬁc song

During heterospeciﬁc song

 

A

 

 

Immediately after conspeciﬁc song

 

 

Immediately after heterospeciﬁc song

 

B

 

 

 

 

 

 

0 1%;

 

24 hours after conspeciﬁc song

 

24 hours after heterospeciﬁc song

 

60

50'

40' :1:

20‘

 

30‘ E

0 . -
SONG: n

 

 

   
 

 

 

- CONTROLS (n = 11) [:j SONG ISOLATES (n = 10)

36

5(8, 152) = 2.16, p = 0.033; calls only: E(8, 152) = 1.46, p = 0.178; movement only: £(8,
152) = 1.19, p = 0.311). Main effects of group (developmental condition) and song type
(conspeciﬁc versus heterospeciﬁc song playbacks) were not detected, and none of the
two, three, or four-way interactions involving those variables were statistically signiﬁcant
(all p_ 2 0.171).

The substantially enhanced responses 24 hours after song suggested a memory for
the testing experience the day before and warranted further attention. For control birds,
this value after the ﬁrst zebra ﬁnch song was more than four times greater than that
during the acclimation period in the ﬁrst test (before they were exposed to any songs in
adulthood). As it was in the silent periods following song exposure in each test, behavior
of the song isolates was somewhat higher and considerably more variable, but it did not
differ signiﬁcantly from the control birds during that ﬁrst exposure to the chamber (t(l9)
= 1.82, p = 0.0842). The analysis of values 24 hours after song exposure revealed a
signiﬁcant effect of time (E(4, 76) = 6.30, p = 0.0002) and song type by group interaction
(E(1, 19) = 4.87, p = 0.0399; calls only: £(1, 19) = 3.15, p = 0.0920). This signiﬁcant
effect of time appeared largely due to differences in responses of the controls and song
isolates during the early testing sessions. Thus, responses of the birds to the chamber in
which they heard conspeciﬁc or heterospeciﬁc song for the ﬁrst time 24 hours earlier
were analyzed, and a signiﬁcant interaction between song type and group was found (E (l,
19) = 6.47, p = 0.0198; data depicted before the dashed lines in Figure 2.4C). Following
zebra ﬁnch song, animals in the control group showed more total responses than the song
isolation group (t (19) = 2.74, p = 0.0129; Figure 2.4C, left). No signiﬁcant difference

was found between the two groups in the number of total responses during the same

37

period 24 hours following heterospeciﬁc song (1 (19) = 0.73, p = 0.470; Figure 2.4C,

right).

DISCUSSION

Adult female zebra ﬁnches responded considerably more during than immediately
after song. Thus, auditory stimuli modiﬁed behavior in both the song-isolated and
control groups. The behavior was further increased when females were re-introduced
into the testing chamber 24 hours following conspeciﬁc or heterospeciﬁc song
presentations. This enhanced delayed response suggests a memory for the experience the
day before. Intriguingly, females isolated from song throughout most of their
development initially showed fewer responses in the chamber in which zebra ﬁnch song
was heard 24 hours earlier than females that were raised in the presence of singing males.
This result is consistent with the idea that females differentially form representations of
environments in which song was heard depending on their prior exposure to song. Over
time, however, differences between the groups were eliminated. The average number of
calls and movements by the group raised without song was equivalent to the control birds
by their third exposure to zebra ﬁnch song, which suggests that adult experience can '
fairly rapidly overcome deﬁcits in responses due to developmental isolation from song.
In both groups, these responses remained consistently high over the testing sessions (did
not habituate across tests) likely due to the fact that birds never heard the same
conspeciﬁc or heterospeciﬁc song more than once.

At least three explanations exist for why females raised in isolation from zebra

ﬁnch song do not initially respond to the environment in which it was heard to the same

38

degree as controls. First, it is possible that responses by song-isolated females were
limited due to increased attention in anticipation of song stimuli presentations that they
experienced or learned for the ﬁrst time 24 hours earlier. In experiments using song
playback, a bird orienting to the source of a song while remaining stationary is typical,
and may be as sensitive a measure as copulation solicitation displays (Searcy, 1992).
Second, without prior experience with conspeciﬁc song, the stimulus may not have been
salient enough to create an association with the environment. Third, this deﬁcit could be
the result of cue competition as seen in studies of classical conditioning (Wasserrnan &
Miller, 1997). Females familiar with conspeciﬁc song appear capable of associating it
with where it was heard, resulting in a “conditional response” to the chamber 24 hours
later (calls and movement in anticipation of song presentation). In song isolates, zebra
ﬁnch song was novel and therefore may have competed with the environment for being
encoded in memory, resulting in an attenuation of conditional responding one day later.
Given that responses by the two groups 24 hours after song became equal over time, the
associative strength of the stimuli likely changed as a result of experience with song by
those birds previously isolated from it.

This ability of adult females raised without hearing song to “catch up” in their
response to the context in which song was heard after only a few exposures is consistent
with some other studies showing that song-related learning can occur later than the d25 to
d35 “critical” period (see Introduction; Clayton, 1988; Miller, 1979b; Neubauer, 1999;
Lauay et al., 2004). Female and male zebra ﬁnches can learn at least some features of the
songs of conspeciﬁcs as late as four to six months post-hatching (Clayton, 1988).

Additionally, like the song-isolated females in the present experiment, song features can

39

be learned or modiﬁed by males at a later age than normal (even at sexual maturity) if
they are deprived of an appropriate tutor (Clayton, 1987; Eales, 1985, 1987a; Jones et al.,
1996) or auditory feedback (Funabiki & Konishi, 2003) during the typical ages for song
acquisition. Also similar to females in the present experiment, adult males can form
memories of song following relatively little (3 hours) exposure to it (Stripling, Milewski,
Kruse, & Clayton, 2003). Thus, males and females seem capable of learning some
discriminations rather quickly, and in females this can occur even after isolation from
song during development.

It is possible that some innate knowledge of the general features of song exists in
zebra ﬁnches (Braaten & Reynolds, 1999), which facilitated the observed changes in
adult behavior based on relatively little exposure. However, it is also possible that the
developmental experience of even the song-isolated birds had some impact. That is,
birds in both groups were exposed to females who may have produced calls. These calls
have some characteristics of male song, such as harmonic stacks (V icario, Navqi, &
Raksin, 2001; Zann, 1996), which may have been sufﬁcient to induce some later zebra
ﬁnch song recognition. Similarly, three of the heterospeciﬁc songs (Bell’s vireo, marsh
wren, and white-breasted nuthatch) contained harmonic stacks and along with the
exposures to zebra ﬁnch songs may have also contributed to development of the song-
context association during the early trials. These results are similar to those in female
canaries, who recognize conspeciﬁc song when reared without exposure to it, but
acoustic tutoring inﬂuences the development of song preferences (N agle & Kreutzer,

1997).

40

The present data also suggest that female zebra ﬁnches appear able to develop the
association between conspeciﬁc song and its environment using the features of song from
any normal, adult male, since the songs on each of the testing days were randomly
selected from a bank of six. This idea is similar to what can occur in males. While they
typically learn from their fathers or males whose songs resemble their fathers’ (Bdhner,
1983; Clayton, 1987; Immelmann, 1969), juvenile males can readily learn from other
available tutors, and do not require physical contact, social attachment, or even a live bird
to do so (Bolhuis, Van Mil, & Houx, 1999; Eales, 1985; Houx & ten Cate, 1999; Mann &
Slater, 1994; Williams, 1990).

The neural mechanisms involved in the formation of a memory for conspeciﬁc
song, including the environment in which it is heard, are currently unclear. However, in
mammals and birds, the hippocampus mediates the consolidation of the association of
events in the environment (Colombo & Broadbent, 2000; Eichenbaum, 2000; Macphail,
2002). Birds possess a wide range of spatial/contextual memory abilities, that at least in
part depend on the hippocampus, which is similar structurally and functionally to its
mammalian homologue (Colombo & Broadbent, 2000; Macphail, 2002; Székely, 1999;
Székely & Krebs, 1996). Lesions of a portion of the hippocampus in male zebra ﬁnches
result in signiﬁcant spatial memory impairment (Patel et al., 1997a; Watanabe & Bischof,
2004), a deﬁcit reduced by transplantation of embryonic hippocampal tissue (Patel et al.,
1997a). Additionally, immediate early gene activation is observed in the hippocampus of
adult and juvenile female zebra ﬁnches following exposure to zebra ﬁnch song compared
to other auditory stimuli (Bailey et al., 2002; Bailey & Wade, 2003), suggesting a

potential relationship between spatial or contextual memory and conspeciﬁc song

41

exposure. Isolation from song throughout development may initially impact the ability of
the hippocampus and/or other regions in the female zebra ﬁnch brain to accurately detect
and/or respond to conspeciﬁc song and associated stimuli. Future work will determine
the potential involvement of the hippocampus in the consolidation of memories related to
song.

As indicated in the Introduction, this potential involvement of the hippocampus in
song-related memory was an important consideration in the analysis of this experiment.
It also inﬂuenced the decision about whether to treat the females with estradiol. The
hormone tends to increase responsiveness in experiments examining female preference
for male song (see Searcy, 1992). However, the hippocampus of adult zebra ﬁnches
synthesizes estradiol and contains neurons that express estrogen receptors (Gahr,
Giittinger, & Kroodsma, 1993; Saldanha, Clayton, & Schlinger, 1999). Estradiol
upregulates NMDA receptors and thus likely inﬂuences neuroplasticity in the zebra ﬁnch
hippocampus (Saldanha, Clayton, & Schlinger, 1999; Saldanha, Schlinger, Micevych, &
Horvath, 2004). These issues, combined with the noted effects of estrogen on memory in
other vertebrates (McEwen & Alves, 1999), and the possible contribution of the
hippocampus to song-related memories in the zebra ﬁnch (see above), suggested that
estradiol could have confounded the results of the present study, so the levels were not
artiﬁcially increased. Similarly, the moving of control females in with song-isolated
birds before testing was an attempt to equalize recent auditory history, thus focusing on
differences between the groups due to their auditory experiences only during
development. However, exposure to conspeciﬁc (and, surprisingly, heterospeciﬁc) song

enhances follicular development over no song playback in female canaries (Serinus

42

canaria; Bentley, Wingﬁeld, Morton, & Ball, 2000). If zebra ﬁnches are Similar in this
regard, then the lack of recent exposure to males may have further lowered their
circulating estradiol concentrations. This deﬁcit in estradiol may explain why the typical
overall preference for zebra ﬁnch song was not detected, and why behaviors such as
copulation solicitation displays were not seen. This idea could readily be tested in future
studies.

Unlike other studies (Clayton, 1988; Lauay et al., 2004; Lauay, Gerlach, Adkins-
Regan, & DeVoogd, in press; Miller, 1979b), birds in the present study did not spend
more time near individual ends of the chamber that broadcast the song stimuli, likely
because songs played from the two speakers could be heard from anywhere in the
apparatus. It is also possible that other methodological differences between the present
and previous studies prevented our females ﬁom spending more time near zebra ﬁnch
than heterospeciﬁc songs. For example, perhaps females needed visual stimuli to
encourage responses, such as models of male zebra ﬁnches (Lauay et al., 2004). Without
this measure of preference, it may be difﬁcult to know precisely what the increases in
behavioral activity by females mean (for example, whether they have a positive or
negative connotation). However, calls (which in the present study contributed
substantially to the effects detected) may indicate song preference; data from canaries
suggest that this response to playbacks is similar to that of copulation solicitation displays
(N agle, Kreutzer, & Vallet, 2002).

In any case, several points can be gleaned from this experiment on female zebra
ﬁnches. First, they clearly form memories for experiences involving auditory stimulation

that are maintained for at least 24 hours. Second, developmental exposure to conspeciﬁc

43

song initially enhances the response in the day following this experience, indicating that
the memory (song-context association) is stronger or that its potential consequences are
more salient in females who had prior experience with zebra ﬁnch song compared to
those that did not. Finally, adult exposure to song readily eliminates the effect of song
isolation during development. The interesting questions now involve the social,
physiological and neurobiological mechanisms regulating the adult plasticity

demonstrated by female zebra ﬁnches.

44

CHAPTER THREE

Bailey, D. J ., & Wade, J. (2003). Differential expression of the immediate early genes
F05 and ZENK following auditory stimulation in the juvenile male and female zebra
ﬁnch. Molecular Brain Research, 116, 147-154.

45

INTRODUCTION

Regions of the brains of adult zebra ﬁnches are tuned to different types of
auditory stimuli, although conspeciﬁc songs preferentially produce responsiveness. In
adult males, neuronal activation indicated by the expression of two immediate early
genes, F OS and ZENK, is Observed in the caudomedial nidopallium (N CM) and
caudomedial mesopallium (CMM; Bolhuis et al., 2001; Jarvis & Nottebohm, 1997; Kruse
et al., 2000; Mello et al., 1992; Mello & Clayton, 1994; Mello et al., 1995; Mello &
Ribeiro, 1998; Stripling et al., 2001) but not the traditional song control nuclei (Jarvis &
Nottebohm, 1997; Mello & Clayton, 1994; Mello & Ribeiro, 1998; Mello et al., 1992)
following presentation of their own song or the songs of other conspeciﬁcs. Conspeciﬁc
songs produce more intense effects in the NCM than heterospeciﬁc song (Mello et al.,
1992; Mello etal., 1995; Kruse etal., 2000), activation that is qualitatively similar in
males and females (Mello et al., 1992). I have shown that, in adult female zebra ﬁnches,
an increased FOS response is seen in not only the NCM, but also the hippocampus
following conspeciﬁc song but not heterospeciﬁc song, tones, or silence (Bailey et al.,
2002). FOS expression in the CMM is equivalent in animals presented with conspeciﬁc
and heterospeciﬁc songs, but is signiﬁcantly increased compared to activation with tone
or silence presentations (Bailey et al., 2002).

While much is known about the regions that respond to song in adulthood, less is
known about the development of those responses. ZENK expression in juvenile zebra
ﬁnches has been quantiﬁed, but in NCM only (J in & Clayton, 1997; Stripling et al.,
2001). At d20, the ZENK response is constitutively high, but at d30 baseline ZENK is

lower and expression is induced signiﬁcantly by conspeciﬁc and heterospeciﬁc songs

46

relative to silence controls (Stripling et al., 2001), although data from males and females
were not considered separately. Electrophysiologically, speciﬁcity for conspeciﬁc song
is exhibited by neurons in the NCM of males and females at d30 (Stripling et al., 2001) as
in adult males (Chew, Vicario, & Nottebohm, 1996a, 1996b). Therefore, the following
experiment examined the expression of the protein products of the immediate early genes
FOS and ZENK at 30 days posthatch (d30) to determine whether (1) sex differences in
juvenile immediate early gene expression exist; (2) FOS and ZENK are differentially
expressed; and (3) the pattern of immediate early gene immunoreactivity in developing
male and female zebra ﬁnches is similar to that observed in adults. Neuronal activation
was assessed at d30 because, in addition to allowing comparisons to my prior work
(Bailey et al., 2002; see Chapter 1), the age is critical for song learning in both sexes.
That is, it is during the period in which males form templates Of their fathers’ songs, but
before they produce their own (N ordeen & Nordeen, 1997). The experience of female
zebra ﬁnches at this developmental stage is critical for later responses to song, in that
they do not Show the normal choice for their father’s song over that of other males if
isolated from song at d25 (Clayton, 1988), whereas isolation at d35 does not impact their

preference (Clayton, 1988; Miller, 1979b).

MATERIALS AND METHODS
Animals and housing

Juvenile male and female zebra ﬁnches (d30; day of batch equals d1; see below
for sample sizes) were obtained from the breeding colony at Michigan State University.

Animals were housed in communal aviaries consisting of approximately seven breeding

47

pairs and their young. Free access to seed and water was provided along with once-
weekly supplements of spinach or oranges and a mixture of hard-boiled chicken eggs and
bread. All stimuli exposures were done during the light portion of the lightzdark cycle

(12:12; lights on at 7 AM).

Stimulus Exposure and Tissue Collection

Each bird was placed in an individual cage in a darkened, sound isolated room to
minimize baseline immediate early gene activity potentially produced by visual
stimulation and to replicate conditions of my prior study (Bailey et al., 2002). Following
60 min of acclimation, female-directed male zebra ﬁnch song, heterospeciﬁc song,
randomly generated tones or an empty sound ﬁle was played using Cool Edit®
(Syntrillium Software Corp.) through a speaker linked to a PC. The sound ﬁles and
stimulus delivery were the same as in my experiment in which F OS induction was
assessed in adult females (Bailey et al., 2002). Zebra ﬁnch song was obtained from ten
individual males recorded from the breeding colony; heterospeciﬁc songs were selected
from the National Geographic Society/Comell Laboratory of Ornithology’s Guide to Bird
Sounds® and were digitized to play in Cool Edit®. Heterospeciﬁc songs were selected
based on approximately comparable bout lengths and frequencies overlapping the range
of zebra ﬁnch song. Species used were American robin, Baird’s sparrow, Bell’s vireo,
Cassin’s ﬁnch, Connecticut warbler, marsh wren, Scott’s oriole, summer tanager, western
meadowlark and white breasted nuthatch. Ten tone ﬁles approximating the features of
zebra ﬁnch song were created in Cool Edit®. For each animal, three sound ﬁles were

randomly chosen from a bank of ten of a particular stimulus condition (except for birds in

48

the “no song” group, in which one empty sound ﬁle was run), based on suggestions for
song playback experiments (Kroodsma et al., 2001; Kroodsma, 1989). Each of the three
ﬁles, 30 sec in length, was repeated in a ﬁxed order, separated by a silence interval of 30
sec for a total stimulus presentation time of 30 min.

One hour following stimulus delivery, while still in the dark, birds were given an
overdose of Equithesin. They were perfused with 0.1 M phosphate buffered saline (PBS)
followed by 4% paraformaldehyde in PBS. Sex was determined post-mortem by
examination of the gonads. Brains were removed, post-ﬁxed for 1 hr in 4%
paraformaldehyde in PBS, and sunk overnight in 30% sucrose in PBS at 4°C. Three
series of frozen, coronal, 30 um sections were collected from each brain into

cryoprotectant, and stored at -20°C until processed for immunohistochemistry.

F OS and ZENK Immunohistochemistry

Tissue was rinsed for one hr in PBS. Then, immunohistochemistry for c-FOS was
performed in one set of tissue (as in Bailey et al., 2002) using a chicken primary antibody
(1220,000; D'Hondt, Vermeiren, Peeters, Balthazart, Tlemcani, Ball, Duffy, Vandesande,
& Berghman, 1999), a donkey anti-rabbit secondary antibody (1:500; Jackson
ImmunoResearch Laboratories) and diaminobenzidine (Sigma) for visualization. ZENK
expression was analyzed in a second set of tissue sections from the same animals using
the same procedures and reagents, except for the primary antibody (Santa Cruz Biotech;
catalog #sc-189; 0.1 jig/ml; Mello & Ribeiro, 1998). Brain sections were mounted on

gelatin-coated slides, dehydrated, cleared in xylene and coverslipped with Perrnount.

49

Data Analysis

For sections labeled with each antibody, immunoreactive nuclei were counted in
the NCM, HP and CMM (as in Bailey et al., 2002). When possible, six sampling areas
per brain region for each animal were quantiﬁed. However, due to histological artifact,
that was not possible for every animal. For analysis of F OS immunoreactivity, an
average of 5.3 samples was available for CMM, 5.2 for NCM, and 5.5 for the HP. For
ZENK immunoreactivity, the averages were: 4.8 (CMM), 5.3 (NCM), and 4.9 (HP). In a
few birds, no sections were available for one of the brain regions, so they were not
included in the analysis. Final sample sizes (5-8 per group) are indicated in Figures 3.1
and 3.2. Densities of FOS- and ZENK- immunoreactive nuclei were determined using an
ocular ind (as in Bailey etal., 2002) on an Olympus BX60 microscope by a rater blind to
treatment condition. Speciﬁcally, in the NCM a 0.49 mm high by 0.49 mm wide box was
placed at the level of the tractus dorso-arcopallialis (DA), ventrolateral to the HP and
dorsal to the dorsal arcopallial lamina (LAD). In the CMM, two sets of counts were
taken. Boxes measuring 0.15 mm high by 0.49 mm wide were adjacently placed below
the lateral ventricle at the level of the HP, dorsal to the mesopallial lamina (LaM) and
medial to the superior ﬁontal lamina (LF S). As in the adult female (Bailey et al., 2002),
labeling within these two sampling regions in CMM was nearly identical; the counts were
therefore summed and analyzed collectively. Cells were counted in the HP using a 0.196
mm high by 0.49 mm wide box placed neighboring the ventricle and straddling the
dorsolateral and dorsomedial subdivisions (Székely, 1999; Székely & Krebs, 1996) of the

structure. Densities of FOS- and ZENK-immunoreactive nuclei were calculated for each

50

bird by taking the average cell counts for each brain region and dividing by its area
(square mm).

Effects of sex, auditory stimulus condition and brain region were assessed using
analyses of variance (ANOVA) separately for F OS and ZENK (sex and stimulus
condition between animals; brain region within). When signiﬁcant interactions were
detected, additional ANOVAs were used to examine effects of stimulus condition within
males and females and the effects of auditory stimulus on densities of immunoreactive
cells in the individual brain regions. Where appropriate, Fisher’s PLSD was used post-

hoc for pairwise comparisons. All statistical analyses were performed using StatviewTM

(SAS Institute).

RESULTS

The most striking results were interactions between sex and stimulus type that
were in opposite directions for the two immediate early genes: the induction of FOS and
ZENK differed in males and females (Figures 3.1 and 3.2). For FOS (interaction: E (3,
49) = 4.47, p_= 0.008), females (E (3, 24) = 3.37, p = 0.035), but not males (E (3, 25) =
2.01, p = 0.138), showed a signiﬁcant effect of stimulus exposure. For ZENK (E (3, 45)
= 2.85, p = 0.049), the effect of stimulus exposure was marginally signiﬁcant in males (E
(3, 22) = 3.05, p = 0.050) but not in females (E (3, 23) = 1.27, p = 0.313; Figure 3.2).
Across the NCM, CMM and HP, FOS responses to zebra ﬁnch song in females were 200-
300% of those detected in the same regions of birds exposed to silence. In contrast,
ZENK levels in males in response to zebra ﬁnch song were 400-900% of those in the

silence-exposed birds. FOS responses in males and ZENK responses in females

51

FEMALES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

120
1001 NCM CMM HP
801
N 60«
EMA b b b
g 20-
8E 04 8 8 7 6 7
In 8
E120 MALES
€100. NCM CMM HP
"" 80‘
60- ‘ '
. 1 N
:3. a I N i
,. it ? is g
7 8 8 6 7 8 8 6 7 8 8 6

CON HET TON NS CON HET TON NS

CON HET TON NS

Figure 3.1. Density of FOS-immunoreactive nuclei (mean + SEM) in the caudomedial
nidopallium (NCM), caudomedial mesopallium (CMM) and hippocampus (HP) in d30
female and male zebra ﬁnches exposed to conspeciﬁc (CON) or heterospeciﬁc song,
tones (TON) or no song (NS). Number of animals analyzed per immediate early gene,
brain region and sex is indicated under each bar. Different lowercase letters indicate

signiﬁcant differences between groups within brain region.

52

 

 

ZENK

. . 2
1mmunoreact1ve cells/mm
00
M
\1

FEMALES
225

200 - NCM CMM HP
175 -

150 -
125 -
100 -

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

7 8 5 7 5 8 5 6 5

225
200 - NCM CMM HP
175 -
150 -
125 -
100 -
75 -
50 -
25 -

 

 

 

 

A 1—7—1
8 6 6 6 8 6 6 6 8 6 6 6
CON HET TON NS CON HET TON NS CON HET TON NS

 

 

 

 

 

 

     

mm

 

 

 

 

 

Figure 3.2. Density of ZENK-immunoreactive nuclei (mean +/- SEM) in the
caudomedial nidopallium (N CM), caudomedial mesopallium (CMM) and hippocampus
(HP) in d30 female and male zebra ﬁnches exposed to conspeciﬁc (CON) or
heterospeciﬁc song, tones (TON) or no song (NS). Number of animals analyzed per
immediate early gene, brain region and sex is indicated under each bar.

53

were not increased compared to the matching silence controls (Figures 3.1 and 3.2).

Due to the opposite responses in males and females, main effects of sex and
stimulus condition were not detected for either FOS or ZENK (all E < 1.60, all p > 0.206).
However, the brain regions appeared to differ in their responses of both immediate early
genes. For FOS (effect of brain region: E (2, 98) = 19.99, p < 0.001), the density of
FOS-immunoreactive nuclei was highest in all areas in females following zebra ﬁnch
song exposure, but the difference reached statistical signiﬁcance only the in HP (E (3, 24)
= 4.07, p = 0.018; Fisher’s PLSD, p < 0.014 for zebra ﬁnch song compared to all other
stimuli; Figures 3.1 and 3.3). For ZENK, a trend (E (2, 84) = 3.03, p = 0.053) for the
values to be slightly higher in CMM than the other regions was detected in males.

Consistent with the regional differences and the two-way interactions between sex
and stimulus type, a three-way interaction of brain region, stimulus condition and sex (E
(6, 84) = 2.34, p = 0.039) was also detected for ZENK immunoreactivity. However, in
comparing the two sexes within each of the brain regions, the only signiﬁcant effect
detected was an interaction between sex and stimulus condition in CMM (E (3, 42) = 4.73,
p = 0.006). In males, the highest density of ZENK immunoreactive nuclei was observed
following conspeciﬁc song, whereas in females, ZENK expression was highest in this
region following heterospeciﬁc song (Figure 3.2).

Overall, the density of ZENK-immunoreactivity was greater than that observed
for F OS (Figures 3.4 and 3.5). As in prior reports in which birds heard song but did not
produce it (Bolhuis, Zijlstra, den Boer-Visser, & Van der Zee, 2000; Jarvis & Nottebohm,

1997; Mello & Clayton, 1994; Mello & Ribeiro, 1998; Mello etal., 1992), immediate

54

CONSPECIFIC HETEROSPECIFIC
SONG ,- SONG

FOS

 

Figure 3.3. FOS- and ZENK-immunoreactivity in the hippocampus (HP) of
female zebra ﬁnches following conspeciﬁc or heterospeciﬁc song
presentations. The arrow in the top left panel indicates the location of the
lateral ventricle, just below where immunoreactive cells are consistently
detected. The ﬁght edge of each photograph is at the midline. Scale bar = 200
pm.

55

FOS

 

Figure 3.4. Photomicrographs of FOS immunoreactivity in the caudomedial
nidopallium (N CM) of female and male zebra ﬁnches following
presentations of conspeciﬁc (CON) or no song (NS). Scale bar = 100 um.

56

ZENK

 

Figure 3.5. Photomicrographs of ZENK immunoreactivity in the
caudomedial nidopallium (N CM) of female and male zebra ﬁnches following
presentations of conspeciﬁc (CON) or no song (NS). Scale bar = 100 um.

57

early gene activity was not induced in the high vocal center (HVC), robust nucleus of the
arcopallium (RA), lateral portion of the magnocellular nucleus of the anterior

nidopallium (lMAN) or in Area X (region present in males only).

DISCUSSION
Summary

The brains of juvenile (d30) male and female zebra ﬁnches respond differently to
auditory stimuli. Juvenile females Show speciﬁcity for zebra ﬁnch song across the NCM,
CMM and HP via F OS expression but not ZENK, whereas the same areas in males show
speciﬁcity for zebra ﬁnch song as measured by ZENK but not FOS expression. Although
direct comparisons to adult data cannot be made because tissue was not processed
concurrently, the patterns observed in the present and previous studies suggest that some
aspects of adult perceptual function are in place in males and females at this age.
However, it is unclear whether males and females continue to use different patterns of
immediate early genes to regulate responses to speciﬁc auditory stimuli, which would
imply sexually dimorphic neural mechanisms for comparable tasks. In contrast, the
responses of the two immediate early genes may equalize in the two sexes as the birds

mature.

F OS Expression
The pattern of F OS-immunoreactivity in juvenile females resembles that of adult
females. In the present study, presentations of zebra ﬁnch song produce a higher density

of F OS-immunoreactive neurons across the regions in juvenile females, reaching

58

statistical signiﬁcance in the HP. In adult females (Bailey etal., 2002), both the NCM
and HP Show selectivity for zebra ﬁnch song, suggesting that the HP response may
develop first. The contribution of the HP to song perception is not yet understood,
although FOS activation in response to conspeciﬁc song is seen in this region in adults of
both sexes (Bailey et al., 2002; Bolhuis et al., 2001; Bolhuis et al., 2000; Kimpo & Doupe,
1997). The response in CMM appears to become less selective as females mature, since
in the adult responses to conspeciﬁc and heterospeciﬁc songs are equivalent (Bailey et al.,
2002). In juvenile males, no differences were observed in the densities of F OS-
immunoreactive neurons across the auditory stimulus conditions in the areas analyzed. A
detailed analysis of F OS expression in the NCM, CMM and the HP under song exposure
conditions similar to mine has not been investigated in adult males. So, it is presently
unclear whether males lag behind females in acquiring the selective FOS responses or

whether this functional difference remains into adulthood.

ZENK Expression

Unlike FOS expression, the density of ZENK-positive cells was increased
following exposure to zebra ﬁnch song in juvenile males. Qualitatively, this result
parallels data from adult male zebra ﬁnches: ZENK expression is higher in NCM and
CMM following conspeciﬁc song presentations than other stimuli (Jarvis & Nottebohm,
1997; Kruse et al., 2000; Mello & Clayton, 1994; Mello etal., 1995; Mello & Ribeiro,
1998; Mello et al., 1992). In contrast, ZENK-immunoreactivity in juvenile females was
statistically indistinguishable across the auditory stimulus types in the regions analyzed.

These results differ from those of adult zebra ﬁnches and other songbirds. In the adult

59

female zebra ﬁnch, ZENK mRN A induction in NCM is qualitatively similar to that seen
in males following stimulation with conspeciﬁc song; it appears to be higher than that
following heterospeciﬁc song, tones or silence (Mello et al., 1992). Similarly,
presentations of conspeciﬁc songs to adult female European starlings produce more
ZENK-immunoreactive neurons in NCM relative to silence controls (Duffy et al., 1999).

It is unknown when the response of ZENK-immunoreactivity becomes speciﬁc
for zebra ﬁnch song in the female NCM or whether it does in CMM and the HP.
However, like those on FOS expression, the present data on ZENK suggest a sex
difference in neural function at d30. One possibility for the conspeciﬁc song response in
these regions in juvenile males is that ZENK may be important for song template
formation during this period. Although song perception is undoubtedly necessary for
female zebra ﬁnches, and exposure to their fathers’ songs at this time seems critical for
their preferences in adulthood (Clayton, 1988; Miller, 1979b), they may not form a
template in a manner comparable to that of males. The memories necessary for future
song production by males may be more specialized than those required solely for
recognition, and thus may utilize different neural mechanisms.

ZENK protein in NCM in the present study closely resembles the expression
observed by Stripling et al. (Stripling et al., 2001), who report no differences in levels of
ZENK mRNA in d30 zebra ﬁnches following stimulation with conspeciﬁc compared to
heterospeciﬁc song. The immediate early gene was apparently analyzed for the two
sexes combined in their study. When the densities of ZENK-immunoreactive nuclei in
the NCM of males and females were considered together, I also detected no difference

between the conspeciﬁc and heterospeciﬁc song groups, and a substantial induction over

60

the silence control. Thus, although different methods (immunohistochemistry vs. in situ
hybridization) were used in this study and others (J in & Clayton, 1997; Stripling et al.,
2001), the results for NCM were replicated. Stripling et a1. (Stripling etal., 2001) also
report using the HP overlying NCM as a control for immediate early gene expression,
because it shows a low level of ZENK mRN A. This result, too, appears to be consistent
with our data. That is, the ZENK immunoreactivity is localized to a relatively small

pocket within the HP which is not directly dorsal to the NCM (Figure 3.3).

General Conclusions

The dorsal telencephalic areas essential to the perception of song, both during
development and adulthood, have been determined largely through examination of
immediate early gene activity and electrophysiological responses of neurons. The
challenge in understanding the function of these brain regions is in pulling these results
together. Cells in the NCM of d30 male and female zebra ﬁnches Show
electrophysiological speciﬁcity for zebra ﬁnch song, and the responses between the sexes
is equivalent (Stripling et al., 2001). The same paper, however, describes equivalent
ZENK responses to heterospeciﬁc and conspeciﬁc song, although both are increased
compared to silence controls. My immediate early gene data suggest that this conclusion
may relate to the pooling of male and female data. Importantly, the present results also
suggest that at this juvenile age, the sexes use different patterns of immediate early gene
expression to encode information about auditory stimuli. While collectively the
electrophysiological and immediate early gene data are consistent with the idea that on

average the speciﬁc induction of both F OS and ZENK in response to zebra ﬁnch song

61

develops after the speciﬁcity of electrophysiological responses, and may be associated
with the formation of song memories to some degree in both sexes, it is also possible that
. the presumably random recordings of cells in NCM (Stripling et al., 2001) were biased
toward FOS-expressing cells in females and ZENK-expressing cells in males.
Alternatively, the results might reﬂect the fact that FOS and/or ZENK induction may not
always occur in cells that are electrophysiologically active, or that electrophysiological
recording is a more sensitive measure than the density of cells expressing immediate
early genes. The latter appears to be true in HVC. Signiﬁcant increases in immediate
early gene responses are not detected when a male listens to his own song (Jarvis &
Nottebohm, 1997; Mello & Clayton, 1994; Mello & Ribeiro, 1998; Mello et al., 1992),
yet discrete peaks of neural activity can be recorded in this area in awake birds during
exposure to portions of the song (Nealen & Schmidt, 2002). To help reconcile these
ideas, though beyond the scope of this dissertation, it would be worth analyzing
immediate early genes in zebra ﬁnches exposed to their own song while anesthetized;
under these conditions, the electrophysiological responses are robust (Lewicki & Konishi,
1995; Margoliash & Fortune, 1992; Nealen & Schmidt, 2002).

The present data also indicate the value of investigating the responses of both
FOS and ZENK at more developmental stages. This song-induced response of ZENK at
least is clearly affected by experience. Not only does it develop between d20 and d30,
but the onset is delayed in males who are raised without the other members of their clutch
(J in & Clayton, 1997; Stripling et al., 2001). Interestingly, exposure to adult male song is
not required (J in & Clayton, 1997). Questions remain as to whether the FOS response

develops in a similar manner, when the sex differences appear, and how long they persist.

62

While additional experiments are needed to elucidate the details of how neural
responses to song develop, this study documents a sex difference in juveniles in the
function of cells in brain regions responsive to song. While male and female zebra
ﬁnches both acquire appropriate responses to song, they appear to do so via divergent
mechanisms. This sex difference is intriguing because it exists in areas that are not
obviously morphologically distinct. Sexual dimorphisms are very clear in the motor
pathway for song production (HVC, RA, nXIIts) and in Area X. For nearly 30 years,
birdsong researchers have known that these regions are larger (or only present) in males,
who sing, compared to females who do not (Arnold, 1997; Nottebohm & Arnold, 1976;
Wade, 2001; Wade & Buhlman, 2000). The differences between the sexes can now be

expanded to include areas involved in song perception, at least on a mechanistic level.

63

CHAPTER FOUR

Bailey, D. J ., & Wade, J. (2005). FOS and ZENK responses in 45 day-old zebra ﬁnches
vary with auditory stimulus and brain region, but not sex. Behavioural Brain Research,
162,108-115.

64

INTRODUCTION

Communication in zebra ﬁnches (T aeniopygia guttata) involves the production
and perception of a variety of vocal signals by both sexes. The most widely studied of
these vocalizations, male song, is learned during a sensitive period in development when,
beginning at approximately post-hatch day 25 (d25), sons memorize characteristics of
their fathers’ songs. Although vocalizations are made by male and female zebra ﬁnches
shortly after hatching, particularly in the form of begging calls (Zann, 1996), males do
not begin to sing their own song until around d45, at which point they actively rehearse it
(Johnson, Soderstrom, & Whitney, 2002) to closely match that of their tutor
(“sensorimotor integration;” Brainard & Doupe, 2000; Eales, 1985; Immelmann, 1969;
Nordeen & Nordeen, 1997). Sensorimotor integration can be delayed or modiﬁed by
preventing males from making or hearing their own vocalizations (Pytte & Suthers, 2000;
Solis, Brainard, Hessler, & Doupe, 2000; Solis & Doupe, 1999). Furthermore, isolation
of zebra ﬁnch males prior to the sensitive period for song learning until adulthood results
in the production of an abnormal song (Eales, 1985; Immelmann, 1969). Females may
undergo a similar period for learning as well: acoustic isolation at d25 (Clayton, 1988)
eliminates their preference for a father’s song over that of another conspeciﬁc male,
whereas this preference is maintained if isolation begins at d35 (Miller, 1979b).

These results point to the importance, in both sexes, of hearing song during
development. Electrophysiological recordings and the measurement of immediate early
gene (IEG) responses, like those of c-FOS and ZENK (an acronym for 211268, Egr-l ,
NGFl-A, Krox-24), have revealed regions of the brain important for the perception of

song. A variety of auditory stimuli elicit activity in brain regions of adult male and

65

female songbirds outside of the regions that control song learning and production. One
of these regions, the caudomedial nidopallium (N CM), responds primarily to conspeciﬁc
songs (Bailey et al., 2002; Bolhuis et al., 2001; Cheng & Clayton, 2004; Chew et al.,
1996a, 1996b; Duffy et al., 1999; Gentner, Hulse, Duffy, & Ball, 2001b; Hernandez &
MacDougall-Shackleton, 2004; Jarvis & Mello, 2000; Jarvis & Nottebohm, 1997; Jin &
Clayton, 1997; Kruse et al., 2000; Mello & Clayton, 1994; Mello et al., 1995; Mello &
Ribeiro, 1998; Mello et al., 1992; Nastiuk, Mello, George, & Clayton, 1994; Park &
Clayton, 2002; Phillmore, Bloomﬁeld, & Weisman, 2003; Ribeiro, Cecchi, Magnasco, &
Mello, 1998; Stripling, Volman, & Clayton, 1997; Terpstra et al., 2004; Vignal, Attia,
Mathevon, & Beauchaud, 2004; Whitney, Soderstrom, & Johnson, 2003). Responses
within another auditory perceptual region, the caudomedial mesopallium (CMM), are
more variable, but activation similar to that observed in NCM has been shown (Bailey et
al., 2002; Bailey & Wade, 2003; Bolhuis etal., 2001; Bolhuis et al., 2000; Cheng &
Clayton, 2004; Duffy et al., 1999; Eda-Fujiwara et al., 2003; Gentner et al., 2001b;
Hernandez & MacDougall-Shackleton, 2004; Jarvis & Nottebohm, 1997; Kimpo &
Doupe, 1997; Kruse et al., 2000; Kruse, Stripling, & Clayton, 2004; Maney,
MacDougall-Shackleton, MacDougall-Shackleton, Ball, & Hahn, 2003; Mello & Clayton,
1994; Mello & Ribeiro, 1998; Mello et al., 1992; Park & Clayton, 2002; Phillmore et al.,
2003; Sockman, Gentner, & Ball, 2002; Terpstra et al., 2004). FOS and ZENK
expression in the zebra ﬁnch hippocampus (HP) following conspeciﬁc song presentations
has also been detected in adult and juvenile males and females (Bailey et al., 2002;
Bailey & Wade, 2003; Bolhuis et al., 2001; Bolhuis et al., 2000; Cheng & Clayton, 2004;

Eda-Fujiwara et al., 2003; Kimpo & Doupe, 1997; Kruse et al., 2004). However, other

66

studies examining ZENK have reported little, if any, IEG activity in response to song
stimulation in the HP (J in & Clayton, 1997; Mello & Clayton, 1994), and one published
report indicates that electrophysiological responses to auditory stimuli have not been
observed in the structure (Chew, Mello, Nottebohm, Jarvis, & Vicario, 1995).

Although a large volume of data from many oscine species details neural
responses to auditory stimuli in adulthood, few studies have examined responses to
auditory stimuli in the developing zebra ﬁnch. In d20 males and females, no ZENK
response in NCM is detected following conspeciﬁc song presentations relative to a
silence control, whereas both conspeciﬁc and heterospeciﬁc songs (but not tones) induce
ZENK expression at d30 (J in & Clayton, 1997; Stripling et al., 2001). Interestingly,
electrophysiological speciﬁcity for conspeciﬁc song does exist in NCM at d20 and d30
and does not differ between males and females. We have shown that the brains of d30
male and female zebra ﬁnches respond differently to auditory stimuli. Overall, across the
NCM, CMM and HP, juvenile females show speciﬁcity for zebra ﬁnch song as measured
by levels of FOS- but not ZENK-immunoreactive cells, whereas the same areas in males
show speciﬁcity for zebra ﬁnch song with ZENK but not FOS expression (Bailey &
Wade, 2003).

Whether ZENK or FOS continue to be expressed in males and females,
respectively, in response to conspeciﬁc songs, or whether both IEG responses become
equivalent in the two sexes as the birds become adults was unknown. The one study to
simultaneously examine IEG activity in adult male and female zebra ﬁnches documented
only that no qualitative difference existed in ZENK expression in the NCM following

conspeciﬁc song presentations (Mello et al., 1992); the lack of a difference in the IEG

67

response mirrors the equivalence in electrophysiological recordings from the NCM of
adult males and females in response to novel conspeciﬁc songs (Chew et al., 1996b). The
speciﬁcity of NCM neurons to conspeciﬁc songs in adult female zebra ﬁnches, measured
by FOS-immunoreactivity (Bailey et al., 2002), parallels the distribution of FOS
expression in adult males following tutor song exposure (Bolhuis et al., 2001), but novel
conspeciﬁc songs were not tested in these males. Although direct comparisons of both
IEGs in the two sexes are needed, these results suggest that neurons within auditory
perceptual regions in adult male and female zebra ﬁnches respond to speciﬁc auditory
stimuli with similar patterns. To further gauge whether song-selective genomic responses
in these regions become ﬁne—tuned with developmental experience, the present
experiment examined IEG responses at d45 in male and female zebra ﬁnches. This point
in development was chosen because it is a time when a clear functional difference

emerges, in that males but not females are beginning to produce song.

MATERIALS AND METHODS
Animals and housing

Male and female zebra ﬁnches at 45 days post-hatching (n = 6 per stimulus
condition; see below) were obtained from breeding aviaries at Michigan State University.
Birds were provided free access to seed, water, cuttleﬁsh bones and ﬁne gravel along
with once—weekly supplements of spinach or oranges and a mixture of hard-boiled

chicken eggs and bread. Experiments were initiated during the light portion of the

light:dark cycle (12: 12; lights on at 0700).

68

Stimulus Exposure and Tissue Collection

Auditory stimuli and their delivery were the same (except for one minor change
indicated below) as used in previous studies from our lab (Bailey et al., 2002; Bailey &
Wade, 2003). Each juvenile was placed in an individual cage in a dark, sound isolated
room, and following an acclimation period of 60 minutes was presented with either
female-directed male conspeciﬁc song, heterospeciﬁc song or no song. Songs (and in the
case of the “no song” group, an empty sound ﬁle) were burned onto compact disc and
played via a Sony CD Walkman (#D-E220) connected to a speaker located directly in
front of the cage (as opposed to PC-delivered song in our prior studies; Bailey et al., 2002;
Bailey & Wade, 2003). Conspeciﬁc songs were recorded ﬁom ten individual males from
our breeding colony and were unfamiliar to the birds used in this study. Heterospeciﬁc
songs (American robin, Baird’s sparrow, Bell’s vireo, Cassin’s ﬁnch, Connecticut
warbler, marsh wren, Scott’s oriole, summer tanager, western meadowlark and white
breasted nuthatch) were selected from the National Geographic Society/Cornell
Laboratory of Omithology’s Guide to Bird Sounds. Each bird was presented three
conspeciﬁc or heterospeciﬁc song ﬁles randomly chosen from a bank often of each
stimulus type, except for those in the “no song” group to which one 30 minute empty
sound ﬁle was played. The conspeciﬁc and heterospeciﬁc song groups received 30
minutes of auditory stimulation with eachof the three respective song ﬁles, all 30
seconds in length, repeated in a ﬁxed order and separated by silence intervals of 30
seconds. Birds were euthanized one hour following the end of stimulus delivery (see

below).

69

In order to determine whether any movements or vocalizations made by the birds
could have confounded the interpretation of our results, a one hour tape recording was
made from each bird that captured the 30 minutes of stimulus exposure and the following
30 minutes. No male sang during this period, although one male and two females
presented with conspeciﬁc songs called at the initiation of the stimuli. The male called
four times and the two females one and six times, respectively. No calls were made by
any other bird. In addition, no movements that were audible to the experimenter were
made by any of the birds.

Birds were injected with an overdose of Equithesin and perfused with 0.1 M
phosphate buffered saline (PBS) followed by 4% paraformaldehyde in PBS. Although at
this age the plumage of male and female zebra ﬁnches has begun to differentiate, the
gonads were examined post-mortem to conﬁrm the sex of each bird. Brains were
removed, post-ﬁxed for 1 hr in 4% paraformaldehyde in PBS, sunk overnight in 30%
sucrose in PBS at 4°C, cut frozen into three series of 30 um coronal sections, and stored

at -20°C in cryoprotectant until immunohistochemistry.

F OS and ZENK Immunohistochemistry

Immunohistochemical procedures were the same as those used previously (Bailey
etal., 2002; Bailey & Wade, 2003). Brieﬂy, following the rinsing of cryoprotected tissue
with PBS, the protein products of c-FOS and ZENK were visualized using separate,
alternate sets of tissue. For FOS, a chicken primary antibody (1 :20,000; D'Hondt et al.,
1999), a donkey anti-rabbit secondary antibody (1:500; Jackson ImmunoResearch

Laboratories) and diaminobenzidine (Sigma) were used. For ZENK, the same procedures

70

and reagents were used, except for the primary antibody (Santa Cruz Biotech; catalog
#sc-189; 0.1 ug/ml). Following protein visualization, brain sections were mounted on

gelatin-coated slides, dehydrated, cleared in xylene and coverslipped with Permount.

Data Analysis

F OS and ZENK immunoreactivity were determined in the NCM, CMM and HP
of each bird from six sampling areas per brain region, when available. For FOS, the
averages were 5.7 sections for NCM and 5.2 each for the CMM and HP; for ZENK, the
averages were: 5.7 (NCM), 4.9 (CMM) and 5.1 (HP). FOS- and ZENK-immunoreactive
nuclei were counted within an ocular grid on an Olympus BX60 microscope by an
observer not aware of auditory stimulus conditions or sex of the animals. Cells within
NCM were counted two ways. First, to replicate analyses of immunoreactivity observed
in prior studies (Bailey et al., 2002; Bailey & Wade, 2003), a 0.49 mm high by 0.49 mm
wide box was placed approximately 0.2 mm from the midline at the level of the dorsal
arcopallial tract (DA), ventrolateral to the HP and dorsal to the dorsal arcopallial lamina
(LAD). Recently, however, the analysis of ZENK immunoreactivity in medial and lateral
divisions of NCM has been separately evaluated in response to a bird’s own song, tutor
song, or that from a novel conspeciﬁc (Terpstra et al., 2004). Therefore, cells within
these regions were then counted in two individual boxes 0.49 mm high by 0.49 mm wide
placed at the same dorsal-ventral level of the brain as described above. For cell counts
within medial NCM, one edge of the grid was placed at the midline of the telencephalon;
for those in the lateral NCM, the medial edge was placed next to the ﬁrst, 0.5 mm from

the midline. As in our previous studies (Bailey et al., 2002; Bailey & Wade, 2003), cells

71

within the CMM were counted in two boxes measuring 0.15 mm high by 0.49 mm wide
adjacently placed below the lateral ventricle at the level of the HP, dorsal to the
mesopallial lamina (LaM) and medial to the superior frontal lamina (SFL). Similar to
data obtained from adult females (Bailey et al., 2002) and d30 juvenile males and females
(Bailey & Wade, 2003), labeling within these two sampling regions in CMM was nearly
identical, so the counts were summed and analyzed collectively. In the HP, cells were
counted using a 0.196 mm high by 0.49 mm wide box placed neighboring the ventricle
across regions deﬁned as “dorsolateral” and “dorsomedial” (Székely, 1999; Székely &
Krebs, 1996).

Average cell counts for each brain region were divided by ocular grid area (in
square mm; NCM = 0.240, CMM = 0.0735 + 0.0735 = 0.147 and HP = 0.096) to
determine densities of FOS- and ZENK-immunoreactive nuclei. Analyses of variance
(ANOVAS) were used to determine the effects of sex (between individuals), auditory
stimulus condition (between individuals), and brain region (within individuals)
independently for F OS and ZENK, using measurements from NCM (the ﬁrst analysis
described above), CMM and the HP. A separate ANOVA was used to determine
differences between males and females and auditory stimulus conditions in the medial
and lateral NCM (within individuals). For each analysis, Fisher’s PLSD was used post-
hoc as appropriate for pairwise comparisons when signiﬁcant main effects were detected.

Statistical analyses were performed using Statview (SAS Institute).

72

RESULTS

Signiﬁcant main effects of auditory stimulus condition (ZENK: E (2, 30) = 3.64, p
= 0.038; FOS: E (2, 30) = 4.87, p = 0.015) and brain region (ZENK: E (2, 60) = 19.79, p
< 0.001; FOS: E (2, 60) = 6.36, p_ = 0.003) were found for both IEGs analyzed. No main
effect of sex for either IEG was detected (ZENK: E ( 1, 30) = 0.39, p = 0.535; FOS: E (1,
30) = 1.90, p_ = 0.179). No interactions between sex and auditory stimulus condition were
found for either ZENK (E (2, 30) = 0.28, p = 0.755) or FOS (E (2, 30) = 0.81, p = 0.453),
and there were no signiﬁcant two- or three-way interactions among the variables
measured (all E < 1.51, p_ > 0.209). Across the brain regions, ZENK and FOS
immunoreactivity were increased following conspeciﬁc song presentations (both sexes
combined; ZENK: E (2, 33) = 3.88, p = 0.031; FOS: E (2, 33) = 4.79, p = 0.015; Figures
4.1 and 4.2). However, the details were slightly different for the two IEGs. For ZENK,
immunoreactivity following zebra ﬁnch song presentations differed only from silence
(Fisher’s PLSD p = 0.009), whereas for FOS, conspeciﬁc song increased IEG expression
compared to both other stimuli (12 < 0.016). Similarly, while the HP consistently showed
the lowest response (AN OVA, main effect of brain region with combined sexes; ZENK:
E (2, 66) = 20.07, p < 0.001; FOS: E (2, 66) = 6.36, p = 0.003), it differed signiﬁcantly
from both the NCM and CMM in the ZENK analysis (1; < 0.001) but only from the CMM
in the FOS analysis (2 = 0.001). Interestingly, F OS expression in the CMM only was
higher than in the NCM (p = 0.008), an effect likely due to increased basal expression
rather than an enhanced speciﬁc response (Figure 4.1). The levels of ZENK and FOS

expression in these regions in the two females and one male that called during

73

300
250

200*
150-

“I

 

NCM

MM HP
a1)

C
gl—

150
100-

a
50 - ,

g b
0. Rte—1 _, AN-

CON HET NS CON HET NS CON HET NS

ZENK

o . o e 2
immunoreactrve cells/mm2 Immunoreactrve cells/mm

lb

 

100‘
50 '

 

 

 

 

 

 

   

 

 

 

 

 

 

 

FOS

    

 

 

 

 

 

 

 

 

 

 

Figure 4.1. Density of ZENK- (top panels) and FOS- (bottom panels)
immunoreactive neurons in the caudomedial nidopallium (N CM),
caudomedial mesopallium (CMM) and hippocampus (HP) of 45 day post-
hatch male and female zebra ﬁnches following presentations of conspeciﬁc
songs (CON), heterospeciﬁc songs (HET) or no song (NS; 11 = 6 males and
n = 6 females for each). Different lowercase letters indicate signiﬁcant
differences between groups within brain region and IEG. Note the y-axis
scale for ZENK immunoreactivity is greater than that for F OS.

74

HET

NS

 

Figure 4.2. Photomicrographs of ZENK- (left panels) and FOS- (right panels)
immunoreactivity in the caudomedial nidopallium (N CM) of male or female zebra
ﬁnches following presentations of conspeciﬁc songs (CON), heterospeciﬁc songs
(HET) or no song (NS). Scale bar = 100 um.

75

conspeciﬁc song presentations did not noticeably differ from the means of their
respective groups.

The response patterns were generally similar among the brain regions, but the
magnitude of effects varied somewhat. For example, auditory stimulus exposure affected
ZENK levels within the NCM (E (2, 33) = 4.05, p = 0.027). Animals exposed to zebra
ﬁnch songs had a higher density of immunoreactive neurons in this region than animals
not exposed to song (9 = 0.008), although ZENK-positive cells did not differ signiﬁcantly
between birds presented with conspeciﬁc and heterospeciﬁc songs (2 = 0.173) or
heterospeciﬁc versus no songs (p = 0.156). In the CMM and HP, auditory stimulus type
did not produce signiﬁcant differences in ZENK immunoreactivity (E (2, 33) = 2.33, p =
0.113 and E (2, 33) = 2.89, p = 0.070, respectively), although birds presented with zebra
ﬁnch songs had approximately 50—100% more labeled neurons in these two regions than
those in the other two groups.

Similar to results for ZENK expression, the density of F OS-immunoreactive
nuclei in the NCM was signiﬁcantly affected by auditory stimulus type (If (2, 33) = 6.26,
p_ = 0.005). In this case, immunoreactivity was increased in birds presented with
conspeciﬁc songs compared to both heterospeciﬁc songs (2 = 0.012) and silence (p =
0.002). Responses following heterospeciﬁc songs did not differ from birds presented
with no song stimuli (p = 0.501). Also similar to the ZENK data, although FOS
immunoreactivity was on average increased more than two-fold following conspeciﬁc
songs, auditory stimulus condition did not result in signiﬁcant differences in expression

in the CMM (E (2, 33) = 2.02, p = 0.149) or HP (1: (2, 33) = 2.31, p = 0.115).

76

In both the medial and lateral NCM, conspeciﬁc song presentations produced the
highest IEG responses (Figure 4.3); a signiﬁcant effect of auditory stimulus condition
was found for FOS (E (2, 30) = 4.55, p = 0.019) and approached signiﬁcance for ZENK
(E (2, 30) = 3.10, p = 0.060). Overall, both ZENK (E (1, 30) = 21.60, p_ < 0.001) and
FOS (E (1, 30) = 13.52, p = 0.001) immunoreactivity differed signiﬁcantly between
lateral and medial NCM, with levels higher in the lateral than medial portion for both
IEGs. Densities of ZENK- (E (1, 30) = 0.08, p = 0.785) or FOS- (E (1, 30) = 0.97, p =
0.332) immunoreactive nuclei did not signiﬁcantly differ between males and females, nor
were interactions between sex and auditory stimulus condition uncovered (ZENK: E (2,
30) = 0.59, p = 0.559; FOS: E (2, 30) = 0.74, p = 0.486). None of the two- or three-way
interactions were signiﬁcant (all E < 2.67, p > 0.085).

Based on the lack of a sex difference and interactions, data from males and
females were pooled to further examine differences between the lateral and medial
portions of NCM, and the results were statistically identical to those above. For both
IEGS, expression was more dense in the lateral than medial NCM (Figure 4.3; ZENK: E
(1, 33) = 22.08, p < 0.001; FOS: E (1, 33) = 13.73, p = 0.001). The effect of stimulus
condition approached signiﬁcance for ZENK (E (2, 33) = 3.27, p_ = 0.051) and differed
for FOS expression (E (2, 33) = 4.63, p_ = 0.017). No signiﬁcant interaction between
stimulus type and NCM region was found for either ZENK (E (2, 33) = 2.73, p_ = 0.080)
or FOS (E (2, 33) = 1.36, p = 0.270). Conspeciﬁc song presentations resulted in
signiﬁcantly more ZENK-positive cells in the combined lateral and medial NCM than
birds not exposed to song (Fisher’s PLSD p = 0.015), but not signiﬁcantly more than

those resulting from heterospeciﬁc songs Q; = 0.251). ZENK-immunoreactivity

77

 

     

 

 

NE 250 .
\ a .medralNCM
% 200 ‘ lateral NCM
U
% a; 150. a,b
LI-l '23,
100
Q)
N a
g 50«
E 0-
80

 

 

 

 

FOS
immunoreactive cells/mm2

CON HET NS

Figure 4.3. ZENK- (top graphs) and FOS- (bottom graphs) immunoreactive
nuclei in the medial and lateral caudomedial nidopallium (NCM) following
conspeciﬁc (CON), heterospeciﬁc (HET) or no song (NS) presentations.
For both ZENK and F OS, expression within lateral NCM was Signiﬁcantly
greater than that in medial NCM (see Results). Different lowercase letters
in each graph indicate signiﬁcant differences between auditory stimulus
conditions within these regions. Note the y-axis scale for ZENK
immunoreactivity is greater than that for FOS.

78

following heterospeciﬁc songs did not differ from that observed in silence controls (p_ =
0.175). For F OS, conspeciﬁc songs produced more immunoreactive nuclei than
heterospeciﬁc songs (9 = 0.049) and no song (p = 0.006); immunoreactivity between
groups exposed to heterospeciﬁc songs and no song did not signiﬁcantly differ (p =

0.360).

DISCUSSION

The expression of ZENK- and FOS-immunoreactive neurons to auditory stimuli
was sexually monomorphic in d45 zebra ﬁnches, but varied across the brain regions
analyzed and the type of auditory stimulus presented. Conspeciﬁc song produced the
highest densities of ZENK and FOS immunoreactivity across the NCM, CMM and the
HP. Although the pattern of expression was the same for all regions analyzed, overall
ZENK expression in the HP was lower than that in the NCM and CMM. Signiﬁcant
differences in IEG expression produced by auditory stimuli were found only in the NCM,
and within the structure the expression of ZENK and F OS was higher in the lateral
compared to the medial portion.

The similar response in the two sexes is probably more like adults (see
introduction to this chapter) and differs qualitatively from the pattern we detected at d30.
At that age, across the NCM, CMM and HP, males express increased ZENK but not F OS,
and females show the reverse, in response to conspeciﬁc songs. Although direct
comparisons to this prior work cannot be made because tissue was not processed
concurrently, at d45 the FOS and ZENK responses in both sexes are higher following

conspeciﬁc songs than the other auditory stimulus conditions; the sex difference observed

79

in animals 15 days younger has disappeared. Regarding the selectivity of responses,
others have found no effect of stimulus type in the ZENK responses of NCM cells at d20,
but conspeciﬁc and heterospeciﬁc song increased ZENK expression relative to basal
levels at d30 (Stripling et al., 2001). In our study using d30 birds (Bailey & Wade, 2003,
Chapter 3) conspeciﬁc songs produced the highest densities of IEGS in all regions
analyzed, but the effect was only statistically signiﬁcant in the HP using the FOS analysis,
and marginally signiﬁcant in the CMM for ZENK. In the present study, signiﬁcant
effects of auditory stimulus condition are found only within the NCM for both IEGS,
which is the one effect consistently detected over all studies of this type on adults (see
Introduction to this chapter). Also, in contrast to what was seen in d30 birds (Bailey &
Wade, 2003), the density of ZENK- expressing cells in NCM is higher than that in the HP,
and FOS-immunoreactivity in the two regions is equivalent, similar to what is seen in
adult females (Bailey et al., 2002). Therefore, it appears that birdsong begins to
selectively induce IEGS between d20-d30 but, by d45, FOS and ZENK responses begin
to be sharpened. These results parallel those from another experiment at d45, in which
the NCM in both males and females exposed to normal conspeciﬁc song showed two
times more ZENK-immunoreactive neurons than birds presented with abnormal song
(Tomaszycki, Sluzas, Newman, Adkins-Regan, & DeVoogd, 2004). Thus, as in adults,
neurons within NCM at this age respond based on song characteristics and quality.

FOS- and ZENK-immunoreactivity in response to novel song playback (and other
auditory stimulus conditions) in the present study was greatest in the lateral but not
medial portion of NCM, contrary to data recently reported in Terpstra et al. (Terpstra et

al., 2004) for adult songbirds. Interestingly, despite the increased responses in the medial

80

NCM, a signiﬁcant positive correlation between ZENK immunoreactivity and the
number of song elements copied from a tutor was found in the lateral but not medial
NCM. This effect was detected when tutor song was presented; no signiﬁcant correlation
between number of elements copied and immunoreactivity following a bird’s own song
or a novel song was uncovered (Terpstra et al., 2004). Furthermore, a marginally
signiﬁcant negative correlation between ZENK immunoreactivity in medial NCM and
number of copied elements was also found (Terpstra et al., 2004). In conjunction with
the present data, these ﬁndings may suggest that the lateral region of NCM plays a role in
acquiring or utilizing the song template. That is, at d45 the lateral NCM may Show an
increased IEG response to conspeciﬁc song because of enhanced activity associated with
comparing zebra ﬁnch vocalizations to the template during sensorimotor integration. In
contrast, this function is diminished in adults.

The functional signiﬁcance of song-induced IEG activation in the HP and CMM
also requires further investigation. A number of studies from several labs have detected
variable responses in the HP in terms of quantity of IEG labeling, whether FOS or ZENK
is measured, whether the response is speciﬁc to particular auditory stimuli, and perhaps
the number of song stimuli used. In adult female zebra ﬁnches, conspeciﬁc songs
produce a higher density of FOS-immunoreactive neurons in the HP than heterospeciﬁc
songs, tones or no songs (Bailey et al., 2002). In d30 females, FOS-positive cells are
highest in the HP following zebra ﬁnch song exposures compared to heterospeciﬁc songs,
tones or silence (Bailey & Wade, 2003). Other studies have reported IEG activation in
the HP in response to auditory stimulation with novel conspeciﬁc songs and tutor song

(Bolhuis et al., 2001; Bolhuis et al., 2000; Eda-Fujiwara et al., 2003; Kimpo & Doupe,

81

1997) and by the pairing of conspeciﬁc song with multi-colored lights (Kruse etal.,
2004). Compared to silence controls, male zebra ﬁnches presented with conspeciﬁc song
stimuli also showed a 40% increase in phosphorylated extracellular-signal regulated
kinase (pERK) in the HP, which is integral to ZENK gene activation (Cheng & Clayton,
2004). Other studies have stated explicitly that no IEGS are observed in the HP following
song presentation (Mello & Clayton, 1994; Stripling et al., 2001), and another normalized
staining in the NCM to the HP (J in & Clayton, 1997), suggesting minimal IEG activity
within the structure. Although it has been reported that electrophysiological responses to
auditory stimuli are not found in HP neurons (Chew et al., 1995), auditory responses have
been observed in some HP cells using conspeciﬁc song and white noise as stimuli to
examine spike-rate discrimination in the zebra ﬁnch (Dr. Mark Hauber, personal
communication). It is clear that the HP in some cases responds to conspeciﬁc song but
studies must be designed to evaluate the speciﬁc nature of that response, including the
extent of its anatomical distribution (see Bailey et al., 2002; Bailey & Wade, 2003).

IEG responses in the CMM are also highly variable. For example, cells in the
CMM are-active in response to a bird’s own song (Jarvis & Nottebohm, 1997; Kimpo &
Doupe, 1997), tutor song (Bolhuis et al., 2000), novel conspeciﬁc song (when it is the
only stimulus tested; Duffy et al., 1999; Kimpo & Doupe, 1997; Kruse et al., 2000, 2004;
Mello & Ribeiro, 1998), song of a familiar rather than foreign dialect (Maney et al., 2003)
and the length of song stimuli (Sockman et al., 2002). Some studies have shown that the
CMM may, in particular species and at certain points in development, respond more to
conspeciﬁc than heterospeciﬁc songs (Bailey & Wade, 2003; Hernandez & MacDougall-

Shackleton, 2004), and others have observed more general auditory responses (Bailey et

82

al., 2002; Bailey & Wade, 2003; Eda-Fujiwara et al., 2003; Gentner et al., 2001b; Mello
et al., 1992; Phillmore et al., 2003; Terpstra et al., 2004). Similarly, the density of IEG
expression in response to auditory stimuli has been reported as both greater (Bailey et al.,
2002; Bailey & Wade, 2003; Bolhuis et al., 2001; Bolhuis et al., 2000; Gentner et al.,
2001b; Hernandez & MacDougall-Shackleton, 2004; Maney et al., 2003) and less
(Phillmore et al., 2003) in the CMM than NCM. It is clear that, as for the HP, additional
work is needed to clarify the role of the CMM. Still, the results of the present report
document a maturation of the perceptual regions, NCM in particular, such that by d45

several adult-like characteristics are in place in both males and females.

83

CHAPTER FIVE

Afferent connectivity of the zebra ﬁnch hippocampus determined by iontophoretic
injections of the retrograde tracer fast blue

INTRODUCTION

Zebra ﬁnches learn, produce and respond to vocal signals via regions that are
linked in generally well-deﬁned neuronal circuits. The regions important for song
template acquisition and the interconnections with the motor control circuitry integral for
sensorimotor integration and song production in male zebra ﬁnch vocalization are well
delineated (see Chapter 1; Bottjer et al., 2000; Mello et al., 1998; Vates et al., 1996;
Vates & Nottebohm, 1995; Vates et al., 1997). In addition, connections among other
telencephalic structures, speciﬁcally the caudomedial nidopallium (NCM), caudomedial
mesopallium (CMM) and Field L (homologous to mammalian primary auditory area;
receives information from auditory thalamus, at least in males; Vates et al., 1996), have
been described and are consistent with the idea that the regions work together to process
song information.

Expression of the protein products of the immediate early genes c-FOS and
ZENK (to some extent) in the hippocampus following conspeciﬁc songs only (Chapters 3
and 4; Bailey et al., 2002; Bailey & Wade, 2003) suggests that the region may be
involved in auditory discrimination or memory formation related to song behavior.
However, direct connections from auditory perceptual areas to the hippocampus have not
been documented in the zebra ﬁnch brain. Connections exist that could allow the
hippocampus to be involved in the processing of auditory information, speciﬁcally to the
dorsomedial portion of the thalamic nucleus (Székely & Krebs, 1996; DMP), part of a

“thalamo-cortical” circuit potentially involved in song learning or perception (Vates et al.,

84

1997), as well as portions of the arcopallium, perhaps including neurons associated with
the motor nucleus RA or its “cup” (Székely & Krebs, 1996). However, afferents to the
hippocampus must be determined to fully understand how the region may interact with
other areas known to play a role in song perception, production, or memory formation.
Together with data already published detailing the efferent connectivity of the structure
(Székely & Krebs, 1996), the afferent connections of the hippocampus were investigated
here to enable the generation of a working model of hippocampal circuitry in male and
female zebra ﬁnches to allow us to better understand its role in song-related behavior,
and to examine potential structure and function relationships with homologous pathways
in mammals. Injections were made only into the dorsolateral subdivision of the
hippocampus (DLHP) to mirror the site of injection Of an anterograde tracer in a prior
study (Székely & Krebs, 1996) and because the DLHP is the region of the hippocampus
where immediate early gene activity is most robust following conspeciﬁc song

presentations (see Figure 3.3).

MATERIALS AND METHODS
Animals and housing, tracer injection and tissue preparation

Adult male (11 = 4) and female (11 = 4) zebra ﬁnches were anesthetized with
isoﬂurane. Glass micropipettes were prepared from borosilicate glass capillaries (World
Precision Instruments, Inc., Sarasota, FL) using a Vertical Micropipette Puller (Sutter
Instrument Co., Novato, CA). The tip of each micropipette was broken to a diameter of
10-12 um, and ﬁtted in the electrode holder of a stereotaxic frame (David Kopf

Instruments, Tujunga, CA). Birds were mounted in the stereotaxic apparatus with head

85

level in a beak clamp (two points picked along sagittal suture and their ventral
coordinates measured and equalized; bar adjusted as necessary). The skull was exposed,
and the intersection of the lambdoidal and sagittal sutures was measured. A small pencil
mark was then made at coordinates relative to that intersection point for injections into
the hippocampus (anterior/posterior (AP) = +1.0, lateral (L) = +/- 0.5) and/or CMM (for
conﬁrmation of afferent HP connectivity (see Results); AP = +1.8, L = +/- 1.5).

Using a 1 mm diameter engraving cutter in a Dremel® rotary tool (S-B Power
Tool Co., Racine, WI), a hole was made at the coordinates indicated above. The dura
mater underneath this hole was carefully teased apart, and if necessary blood was
removed with cotton swabs or GelFoam® (Pharmacia and Upjohn Co., Kalamazoo, MI).
A small volume of fast blue (0.5 pl; 30 pg/ml saline; Illing Plastics, Bergfeld, Germany)
or ﬂuorogold (0.5 ul; 3% + 1% dimethylsulfoxide (DMSO) in 0.1 M phosphate buffered
saline (PBS); F luorochrome, Inc., Englewood, CO) was transferred to each glass
micropipette by capillary action or suction. Excess tracer on the outside of the
micropipette was absorbed with a cotton swab prior to injection. A silver wire electrode
was lowered into the micropipette, and the positive output lead of a MidgardTM Precision
Current Source (Stoelting Co., Wood Dale, IL) was attached to it. A ‘ground’ output lead
from the current source was attached to a skin ﬂap overlying the skull. The current
source was turned on, and the output current adjusted to 7 11A. Micropipettes were
lowered through the hole in the skull to a ventral (V) depth of -0.3 relative to the surface
of the brain for the hippocampus and -0.5 for the CMM. Alternating (7 seconds on, 7

seconds off) current was used to unilaterally iontophorese the retrograde tracer(s) into the

86

region(s). After 10 min, the alternating output switch was turned off, and the
micropipette remained in place for 10 min before removal.

The hole in the skull was covered with bone wax (Fine Science Tools Inc., Foster
City, CA) and the ﬂaps of skin reconnected with silk suture (Ethicon, Inc., Somerville,
NJ) and sealed with collodion (Mallinckrodt Baker, Inc., Phillipsburg, NJ). Birds were
removed from the stereotaxic apparatus and then placed individually in cages in the main
bird colony with free access to food, water and supplements of spinach or oranges and
hard boiled chicken eggs and bread.

Five days following tracer injection (survival time based on pilot work), each bird
was overdosed with Equithesin and perfused transcardially with 0.1 M phosphate
buffered saline (PBS) followed by 4% paraformaldehyde in PBS. Brains were post-ﬁxed
for 1 hr in 4% paraformaldehyde and then sunk overnight in 30% sucrose in PBS while
continuously protected from exposure to light.

Brains were cut under dim light coronally on a cryostat at 30 um and mounted
onto Superfrost Plus slides (Fisher Scientiﬁc, Pittsburgh, PA). Two alternate sets were
cut. One set from each zebra ﬁnch was rapidly dehydrated, cleared in xylene, and
coverslipped with DPX (Sigma-Aldrich, St. Louis, MO). Tissue remained in the dark
while drying. The alternate set of tissue from each bird was stained with thionin,
dehydrated, cleared with xylene, and coverslipped with Permount to aid identiﬁcation of
the locations of administration Of the tracers and labeled cells, as well as any potential

damage to the tissue due to the injection.

87

Data Analysis

Under ﬂuorescent light, the injection site and cells labeled with fast blue and
ﬂuorogold were digitally photographed using PictureFrame (version 2.2) software and a
MacroF IRE digital camera (model #S99831; Optronics, Goleta, CA) connected to an
Olympus BX51 microscope. Images in this chapter of the dissertation are presented in
color. Labeling in a brain region was determined to be absent (no ﬁlled neurons), low (1-

5 labeled cells per tissue section), moderate (6-10 labeled cells) or high (> 10 cells).

RESULTS

Birds with injection sites limited to within or just above the area of the DLHP in
which FOS-immunoreactive cells were observed following conspeciﬁc song
presentations (Bailey, Rosebush & Wade, 2002; Bailey & Wade, 2003) were included in
the analysis. Retrogradely labeled cells ﬁom injections medial or lateral to this region or
that penetrated other portions of the telencephalon ventral to the lateral ventricle were not
examined in detail. Injection Sites were typically 100-200 pm in diameter (e.g., Figure
5.1) and were determined by the largest diameter of fast blue product (designated by a
brown color under the wavelength of ﬂuorescent light used) in a series of coronal
sections. Labeled cells were clearly visible clustered around the injection sites, and the
iontophoretic injections exhibited no observable leakage around the micropipette path.

A summary of the main brain regions in which retrogradely labeled cells were
observed is presented in Table 1. In both male and female zebra ﬁnches, retrogradely
labeled neurons were consistently detected in the dorsomedial (DMHP) and ventral (VHP)

subdivisions of the hippocampus, CMM (Figure 5.1) and lateral septum (SL; Figure 5.2).

88

In some cases, the projections seemed to be sexually dimorphic. In females but not males,
cells ﬁlled with fast blue were found in the periventricular nucleus of the hypothalamus
(PVN; Figure 5 .2), and in males but not females, labeled cells were detected in the
posterior portion of the dorsomedial thalamic nucleus (DMP; Figure 5.3). In all but one
of the males and in only one female, retrogradely labeled cells were found in HVC
(Figure 5.4). In addition to these major projections, a few cells were observed
(inconsistently among the animals) in nucleus taeniae (Tn), the rostromedial portion of
the medial striatum (MSt), nucleus rotundus (Rt), the pedunculopontine tegmental
nucleus (TPc), and the Substance P reactive nucleus (SPf; data not shown).

To begin to conﬁrm the afferent connectivity of CMM with the hippocampus, fast
blue was injected into the hippocampus and ﬂuorogold into the CMM of a male zebra
ﬁnch (Figures 5.5A and B, respectively). In the brain of this animal, cells ﬁlled with fast
blue were detected in the CMM (Figure 5.5C), and ﬂuorogold-labeled neurons were
observed in the hippocampus (Figure 5.5C) throughout the subdivisions (Figure 5.5D).
Results similar to these were obtained when fast blue was iontophoresed into the CMM
of a female zebra ﬁnch: retrogradely labeled cells were observed in the DLHP and

DMHP (data not shown).

89

 

Number of birds and
relative density of fast
blue labeled neurons

 

Region Males Females

 

 

 

 

 

 

 

DMHP 4/4 3/3
++ ++
VHP 4/4 2/ 3
++ ++
CMM 3/4 3/3
+++ +++
SL 4/4 2/3
+ +
PVN 0/4 3/ 3
— ++
DMP 3/4 0/3
+ -
HVC 3/4 1/3
++ ++

 

Table 5.1. Brain regions in male and female zebra ﬁnches exhibiting fast blue
labeled neurons following iontophoresis into the dorsolateral subdivision of the
hippocampus (DLHP). Below the number of birds of each sex in which
retrogradely labeled neurons were observed (or not) in a particular brain region,
the density of labeling is indicated as follows: +++ (high number of labeled
neurons); ++ (moderate number of labeled neurons); + (few labeled neurons); -
(no labeled neurons). DMHP: dorsomedial subdivision of the hippocampus;
VHP: ventral subdivision of the hippocampus; CMM: caudomedial mesopallium;
SL: lateral septum; PVN: periventricular nucleus; DMP: posterior portion of the
dorsomedial nucleus of the thalamus. ‘

90

Figure 5.1. Coronal section through the dorsomedial surface of the zebra ﬁnch
telencephalon (top; red box indicates photographed area in bottom panel) at the level of
the hippocampus showing iontophoresed fast blue in the dorsolateral (DLHP)
subdivision. Note the retrogradely labeled cells at each arrow in the dorsomedial
(DMHP) and ventral (VHP) subdivisions of the structure and in the caudomedial
mesopallium (CMM). Scale bar = 200 um.

91

 

 

92

Figure 5.2. Coronal sections through the zebra ﬁnch brain at the level of the lateral
septum (SL; dashed red box) and periventricular nucleus of the hypothalamus (PVN;
dotted red box). Iontophoretic injections of fast blue into the dorsolateral subdivision
of the hippocampus resulted in ipsilaterally labeled cells within the SL (arrows, panel
B; photo from a male zebra ﬁnch) in both male and female zebra ﬁnches and the PVN
in females only (arrows, panel D). Thionin stained tissue through the SL (panel A) and
PVN (panel C) are provided for comparison. Scale bar for panels A and B = 200 um,
and for panels B and C = 100 um.

93

\.
c

. -,.-...
k4. ;

'0
3.5"

A.
'0

 

 

94

 

Figure 5.3. Coronal sections of a zebra ﬁnch brain at the level of the
posterior portion of the dorsomedial thalamic nucleus (DMP; dashed red
box). Fast blue injections into the dorsolateral subdivision of the
hippocampus resulted in a few ﬁlled cells within the ipsilateral DMP
(arrow) in males but not females. HP = hippocampus, CMM =
caudomedial mesopallium; N = nidopallium. Scale bar = 100 pm.

 

Figure 5.4. Coronal sections through a zebra ﬁnch brain at the
level of HVC. Red, dashed box indicates the region depicted in
the thionin stained (panel A) and fast blue labeled (panel B)
sections. NCM = caudomedial nidopallium, Cb = cerebellum.
Scale bar = 200 um.

96

Figure 5.5. Iontophoretic injections of fast blue into the hippocampus (panel A) and
ﬂuorogold into the caudomedial mesopallium (CMM; panel B at arrow) of a male
zebra ﬁnch resulted in fast blue ﬁlled cells in the CMM (panel C) and ﬂuorogold
labeled neurons in each of the hippocampal subdivisions (panels C and D). DLHP =
dorsolateral subdivision, DMHP = dorsomedial, VHP = ventral, V = ventricle
separating the hippocampus from CMM. Scale bar for panels A and B = 50 um and
for C and D = 200 um.

97

 

DLl ll) DMl ll)

 

98

DISCUSSION

The projections to the hippocampus detailed in this study indicate connections
with nuclei involved in song behavior and, together with song-speciﬁc responses
observed in the hippocampus (Chapters 1, 3 and 4, Bailey et al., 2002; Bailey & Wade,
2003), suggest a role for the region in responses to vocal communication signals or
auditory-related memories in the zebra ﬁnch. Moreover, based on the present study and
the one examining the efferent projections of the hippocampus by Székely & Krebs
(1996), the zebra ﬁnch hippocampus appears to possess reciprocal connections with
several of these regions that are directly or may indirectly be involved in song perception
or memory, such as the CMM and SL in males and females and DMP in males.

Injections of a retrograde tracer into DLHP gave rise to only ipsilaterally labeled
I cells, which is similar to observations in zebra ﬁnches and pigeons by others who have
examined the projections of this region using an anterograde tracer (Casini et al., 1986;
Casini et al., 1986; Krayniak & Siegel, 1978; Székely & Krebs, 1996). For example,
following anterograde tracer injections into the DLHP, terminating axons were
consistently Observed in the rostromedial medial striatum (MSt), preoptic regions, PVN,
DMP, arc0pallium, SL and the central gray (Székely & Krebs, 1996). In parids like the
black-capped Chickadee, the volume of the SL is larger than in a species that does not
store food and increases in fall when the bird is exhibiting peak caching behavior (Shiﬂett
et al., 2002). The septum may therefore be involved in spatial or relational memory
along with the hippocampus, and in the zebra ﬁnch a relationship may exist between the
regions that results in the processing of memory for relevant auditory signals in the form

of a “contextual song memory” mentioned above. In a non-songbird, the

99

intertelencephalic connectivity of the hippocampus is quite similar. Ipsilaterally, the
hippocampus in homing pigeons has reciprocal connections with the SL and Tn, and a
contralateral projection is observed from the hippocampus to the SL and MSt (Atoji et al.,
2002)

The interconnectivity of subdivisions within the zebra ﬁnch hippocampus also
parallels that from a non-songbird. DLHP projects to both the DMHP and VHP (Székely
& Krebs, 1996), and retrogradely-labeled cells in these regions were consistently detected
in the present study in both males and females. In the pigeon, a more detailed analysis of
intrahippocampal connectivity has been performed by stimulating individual regions of
the hippocampus and recording evoked ﬁeld potentials from neighboring subdivisions.
Based on the cytoarchitecture of the region, the pigeon hippocampus also has dorsolateral,
dorsomedial and ventral subdivisions that are further subdivided and connected in a
manner similar to that observed in the zebra ﬁnch (Hough II et al., 2002). For example,
the dorsal DLHP projects to the dorsomedial subdivision, which send axons that synapse
in the medial VHP, which ultimately project back to the DLHP to its ventral portion
where axons of that region exit the hippocampus. These networks of connections suggest
a similarity to the trisynaptic pathway (Amaral & Witter, 1989; Eichenbaum et al., 1996)
in the mammalian hippocampus. However, the topology of the region present in
mammals, dentate gyrus to Ammon’s born to subicular and entorhinal cortices, appears to
be different from that present in birds. If one presumes that adjacent parts of the dorsal
cortex are homologous to the entorhinal and subicular cortices (Hough H et al., 2002;
Butler and Hodos, 1996), then the topological order in zebra ﬁnches would be Ammon’s

horn (“hippocampus proper,” or the subdivisions described above) to dentate gyrus (the

100

parahippocampal area as described in pigeons) to subicular and entorhinal cortices
(perhaps SPf; Székely & Krebs, 1996; Butler and Hodos, 1996).

The primary differences between this study and that by Székely & Krebs (1996)
are my observations of retrogradely labeled cells in the CMM and HVC. In their work,
hippocampal projections observed are only in rostral portions of the mesopallium just
dorsal to the most lateral extent of the CMM where the lateral ventricles terminate
(Székely & Krebs, 1996), whereas in the present study, retrogradely-labeled cells were
detected in CMM from its border with the lateral ventricle to the medial extent of the
region. Although this difference may be based on the uptake or kinetics of the different
tracers, the hippocampus in mammals receives substantial auditory input and the
electrophysiological properties of neurons within the hippocampus can be modulated by
lesion or stimulation of those regions. For example, in rats, sensory evoked activity in
the hippocampus is modulated by damage to the entorhinal cortex or medial septum
(Foster, Hampson, West, & Deadwyler, 1988), and in rabbits, auditory stimuli are more
potent than visual stimuli in their ability to modulate electrophysiological properties of
dorsal hippocampal neurons (Astikainen, Ruusuvirta, & Korhonen, 2005). In zebra
ﬁnches, cells in the CMM appear to be the primary means for conveying auditory
information to the hippocampus, and the two structures may be part of a circuit that is
integral to the consolidation of song-related memories.

The sexually dimorphic projections observed in the present study are intriguing
also, although more birds may be needed to conﬁrm statistically that the projections
differ between the sexes. Although the efferent connections of the hippocampus were

previously examined in both sexes (Székely & Krebs, 1996), whether sex differences

101

were uncovered was not reported. In the present study, in males but not females, the
hippocampus appears to have reciprocal connections with DMP, a thalamic nucleus that
projects to the left and right medial portions of the magnocellular nuclei of the
nidopallium (mMAN), which project ipsilaterally to the motor nucleus HVC (see Chapter
1; Vates et al., 1997). Further, the downstream motor nucleus RA projects to DMP,
creating a “thalamo-cortical” circuit that is involved in the coordination of vocal signals
(Vates et al., 1997). That the hippocampus has reciprocal connections with DMP
suggests perhaps a modulatory role for hippocampal neurons on the ﬁrnction of cells
within DMP. In addition, in three out of four males and in only one out of four females
were labeled cells observed in HVC following hippocampal injections of the retrograde
tracer, although this difference may be due to the dramatic difference in size between the
male and female HVC. In female but not male zebra ﬁnches, retrogradely-labeled
neurons were found in the PVN, suggesting a sexually dimorphic role for the region that
may or may not be related to song behavior. However, in male European starlings, axons
from another hypothalamic nucleus, the medial preoptic nucleus, terminate in a region of
the arcopallium surrounding RA, suggesting an indirect pathway through which the
region may inﬂuence song behavior and speciﬁcally motivation to sing (Riters & Alger,
2004)

The hippocampus in zebra ﬁnches has several intratelencephalic connections that
provide the region access to auditory information and suggest a modulatory ﬁrnction for
the structure in the processing of or response to song stimuli. For example, in adult
females, immediate early gene activity in response to conspeciﬁc and heterospeciﬁc song

is observed in CMM, but only conspeciﬁc songs induce immediate early gene activity

102

within the hippocampus (Bailey et al., 2002). Although further examination of the neural
loop between these regions is necessary, these data collectively point to a means by
which zebra ﬁnches may be able to form memories relating particular, relevant songs to
speciﬁc environments (see Chapter 1) or perhaps distinguish other individual
conspeciﬁcs. The questions that remain are the speciﬁc nature of these connections,
particularly the sexually dimorphic ones, when they develop, and how they are regulated.
The answers will be important to further our understanding of the nature of vocal
communication and the continuous neural modiﬁcations that may inﬂuence the sender-

receiver relationship.

103

CHAPTER SD(

Effects of lesions of the hippocampus in adult and juvenile male and female zebra ﬁnches
on song-related behavior

INTRODUCTION

Responses to conspeciﬁc song have been detected in the hippocampus using a
number of different neurochemical and behavioral paradigms. I found that in adult
female zebra ﬁnches conspeciﬁc song presentations produce a higher density of FOS-
immunoreactive neurons in the hippocampus than heterospeciﬁc songs, randomly
generated tones or no songs (Bailey et al., 2002). In juvenile females at d30, FOS-
positive cells are also highest in the hippocampus following conspeciﬁc song exposures
compared to heterospeciﬁc songs, tones or silence (Bailey & Wade, 2003). In d45 birds,
FOS and ZENK immunoreactivity are highest (but not signiﬁcantly higher) in the
hippocampus after conspeciﬁc song presentations compared to heterospeciﬁc songs and
silence (Bailey & Wade, 2005). Several other studies have reported immediate early
gene activation in the hippocampus in response to auditory stimulation with novel
conspeciﬁc songs and/or tutor song (Bolhuis et al., 2001; Bolhuis et al., 2000; Eda-
Fujiwara et al., 2003; Kimpo & Doupe, 1997) and after pairing of conspeciﬁc song with a
discrete visual stimulus (Kruse et al., 2004). Male zebra ﬁnches presented with
conspeciﬁc song stimuli showed a 40% increase in phosphorylated extracellular-signal
regulated kinase (pERK) in the hippocampus compared to a silent control (Cheng &
Clayton, 2004); pERK activation is integral to induction of the ZENK gene. Additionally,
neural activity in the hippocampus of male and female zebra ﬁnches measured using
optical imaging following the injection of a voltage sensitive dye was higher during than

before and after conspeciﬁc song presentations (Kakeue, Kaminosono, Kitamura, Ogawa,

104

& Oka, 2004). However, other studies have stated explicitly that no immediate early
genes are observed in the hippocampus following song presentations (Mello & Clayton,
1994; Stripling et al., 2001), and another normalized staining in the NCM to the
overlying hippocampus (J in & Clayton, 1997), suggesting minimal or baseline expression
within the region. Although it was reported that electrophysiological responses to male
zebra ﬁnch song are not found in neurons within the hippocampus (Chew et al., 1995),
auditory responses have been observed in some hippocampal cells using conspeciﬁc song
and white noise as stimuli (Dr. Mark Hauber, personal communication).

The telencephalic connections of the zebra ﬁnch hippocampus detailed in Chapter
5 suggest a potential role for the region in song-related behavior. For example, in males
and females, the hippocampus appears to have reciprocal connections with the auditory
area CMM and the lateral septum, a region known to be involved in motivated behavior
in birds (see Chapter 5). The hippocampus in males also projects to and receives
efferents from DMP, a thalamic nucleus part of a circuit involved in song production.
That neurons within the hippocampus receive information from regions involved in song
perception, motivation and production and that the hippocampus projects back to those
areas suggests a potential modulatory role for the hippocampus in speciﬁc song-related
behaviors.

It is clear that the hippocampus in some cases responds to conspeciﬁc song and
that intertelencephalic connections exist that provide song-related information to the
region, but the speciﬁc nature and purpose of these responses and connections have not
been examined. Conspeciﬁc song induced activation within the hippocampus suggests a

relationship between auditory perception and song memory, potentially involving (1) the

105

 

consolidation, retrieval or maintenance of a song template, and/or (2) the encoding of
information about the environment in which song was heard, consistent with the
relational memory function of the structure observed in a multitude of studies using avian
and mammalian species (Colombo and Broadbent, 2000; Eichenbaum, 2000; Squire,
1992). Given the role of the hippocampus in speciﬁc types of memory formation
(Oberlander, Schlinger, Clayton, & Saldanha, 2004; Patel et al., 1997a; Watanabe &
Bischof, 2004), it may be involved in the development or utilization of song memories in
a manner consistent with one of or both of the two hypotheses listed above.

The following experiment was designed to determine the effects of lesions to the
hippocampus during development or adulthood on song-related behavior in the zebra
ﬁnch. Following a lesion or sham control, birds were given tests of song production
(males), two tests of song perception (tutor’s song versus that of another conspeciﬁc, and
conspeciﬁc versus heterospeciﬁc song; both males and females), and spatial memory
(both sexes) as a control. If, for example, the hippocampus is important in adult song
perception or production, then destruction of the structure in mature birds should result in
impaired performance. This study also tested the possibility that the hippocampus is
involved in the song learning process. Decrements in song production or song perception
following lesions Of the structure at d20 or d45 in males and females, just as they enter
the template formation and sensorimotor integration phases (at least as deﬁned in males;
Nordeen & Nordeen, 1997), respectively, would indicate an integral role for cells within

the hippocampus in the development and adult expression of normal song behavior.

106

MATERIALS AND METHODS
Animals, Surgery and Housing

To conﬁrm the validity of the behavioral measures used in this experiment, 9
adult male (5 hippocampal lesion and 4 sham control, see below) and 8 intact adult
female zebra ﬁnches were obtained from group cages in which birds of the same sex
were housed together. They were in visual and auditory but not physical contact with the
other sex. All of the females were from Magnolia Bird Farm (Riverside, CA) and were in
their home aviaries at least two weeks prior to testing. Eighty-nine birds from the
breeding colony were used in the actual experiment. Of those, 15 died following surgery
or shortly thereafter, 12 had lesions ventral to the hippocampus, and three (one female
lesioned at d20 and two males lesioned or sham-lesioned in adulthood) did not eat at all
from the baited cup during the spatial memory test (see details on the test below) and
were removed from the experiment. The number of birds remaining for the groups was
as follows: d20, n = 10 males and 9 females; d45, n = 10 males and 10 females; and d100,
n = 10 males and 10 females.

All birds were maintained on a 12:12 light:dark cycle (lights on at 0700). While
housed in the communal aviaries, animals were provided with ad libitum access to seed,
water, gravel, and once-weekly supplements of oranges, spinach, or hard boiled chicken
eggs and bread. Birds in each group were removed from their communal aviaries at d100
and placed in individual cages (11” x 9” x 12”; see rationale below) with free access to
water, gravel, and seed (except before Spatial memory test, see below). While in
individual cages, birds continued to have visual and auditory contact with other males

and females.

107

Birds were deeply anesthetized with isoﬂurane and bilateral injections of 0.1 111 of
10 mg/ml ibotenic acid (ICN Biomedicals, Costa Mesa, CA; dissolved in 0.1 M
phosphate buffered saline (PBS), pH 7.4; lesion group) or 0.1 ul PBS (sham) were made
into the hippocampus using coordinates and procedures described in the experiment in
Chapter 4, determined through pilot work, and aimed only at the dorsolateral subdivision
of the structure where I detected conspeciﬁc song-induced immediate early gene
activation (Chapters 2 and 3; Bailey et al., 2002; Bailey & Wade, 2003, 2005). In d45
and adult birds, holes were drilled through the skull and the micropipette was lowered 0.3
mm relative to the dura. The skull of d20 zebra ﬁnches is not ﬁrlly ossiﬁed, so no holes
were drilled, and the micropipettes easily pierced through the skull’s surface. The ventral
measurement for birds of this age was 0.55 mm relative to the surface of the skull to
account for the 250-300 um thickness. Since zebra ﬁnches need parental care for feeding
until approximately d35, and because song learning in the presence of a tutor begins
around d25 and continues through approximately d65, juveniles were returned to their
home cages following surgery and recovery. Birds were returned only after I determined
that the effects of the isoﬂurane had dissipated, mainly by ensuring that locomotion was
ﬂuid and that birds (at least those d45 and adult) were able to balance on perches without
trouble. Lesions to juvenile birds were performed near the end of the light cycle to lessen
the number of encounters with other birds immediately following the surgery and to

allow for additional, unimpeded recovery overnight.

108

Behavior Testing

All behavioral tests (except those for song production in adult males before the
lesion, noted below) were done in the same room. Therefore, the spatial memory test was
run ﬁrst in all animals to avoid effects on performance in the task that may have been due
to pre-exposure to the environment. The order of the remaining tests (two tests of song
preference in males and females and “after lesion” song production tests in all males)

were counterbalanced and all tests were separated by intervals of approximately 24 h.

Spatial Memory T est(s)

I designed a modiﬁed “t-maze” to examine spatial memory in zebra ﬁnches. The
apparatus (see Figure 6.1), consisted of a wood frame and steel mesh on all sides through
which behavior could be observed and recorded. This portion of the chamber was
designed so the birds could clearly see the environment outside of it. At the base was a
wood frame/steel mesh enclosed column that led vertically to left and right arms. Cut
into the bottom of the column were four identical doors used as release points. In the
apparatus were six identically colored and patterned cups with closable ﬂaps, three in
each arm, arranged in such a way that the arms looked identical at their entrance points.
The chamber was situated in a room such that landmarks were easily identiﬁed based on
their relation to the walls of the apparatus (a bookshelf, curtain, video camera, light above
the center of the chamber, proximity to walls of the room, different shapes of the same
color on adjacent walls, patterns of acoustic tile afﬁxed to walls, etc.). The natural

response of birds tested was to ﬂy up the column, and the intent of the task is that they

109

 

 

 

A ﬁ 1'210cmll -
a "
it

 

 

 

 

 

 

 

B I I
1b 1c 2a
= :1
la H 2c 2b
I I

Figure 6.1. (A) Chamber used for spatial memory tests following hippocampal
lesions, frontal view. Doors at the bottom of the chamber were used as entrance
points. (B) Chamber diagrammed from above. Perches (indicated in grey) at the
ends of the chamber that run parallel with the front of the maze sat 16 cm above the
ﬂoor of the arms, and those that run perpendicular in each arm were 20 cm above.
In the middle of the chamber, 3 cm above the ﬂoors of the arms, were perches in
the shape of a plus, and the part of the plus that runs parallel with the front of the
chamber extended 5 cm into each arm. The plastic cups (2 cm x 2 cm x 3 cm)
were also at varying heights: cups 1a and 2a were 19 cm above the ﬂoor of the
arms, cups 1b and 2b 15 cm, and cups 1c and 2c 2 cm. Perches were not placed
directly below/adjacent to a cup so that birds would not land by one randomly.

The perches were placed nearby and just above so that birds would only approach a
cup to eat its contents.

 

 

 

 

 

 

110

entered and spent most of their time in the arm where seed was found previously, even
from random release points, based on spatial cues in the room.

Prior to testing, birds were individually housed without food for approximately 4
h to increase motivation to perform the task. Food deprivation began immediately after
the onset of the light portion of their cycle, Since zebra ﬁnches ﬁll their crops with seed
over the course of the morning hours (Zann, 1996). Following this deprivation period,
each bird underwent “acquisition” trials, which began by placing them individually into
the chamber through the front door in the bottom of the column. The ﬂaps of all cups
were open, and only one of the cups was ﬁlled with seed. Pilot work determined that
almost all birds took 20 min or less to acclimate to the chamber and eat from the baited
cup. Thus, birds were given 20 min (maximum latency) to explore the chamber and
locate the seed. Once the seed was found (or if 20 min had passed), each bird was given
30 sec to eat, after which it was returned to its individual cage in a separate room. Every
10 min, birds were placed back into the chamber through the same door until the arm that
contained food was entered within 30 sec on three consecutive trials (criterion). Upon
reaching criterion, birds were returned to their individual cage for one hr.

The cups that did not contain seed were ﬁlled and then emptied to eliminate any
olfactory cues birds may use to locate the seed, although the only avian species that
appear to rely heavily on olfactory cues are those that migrate or are carnivorous
(Malakoff, 1999). Also, zebra ﬁnches forage for and choose seeds based mainly on
visual cues like size and color to limit time spent foraging (Zann, 1996). Following the
one hr interval upon reaching criterion, birds underwent four “probe” trials, each

separated by 10 min, which tested their memory for the food location. Each bird was

111

placed through each of the four doors of the chamber at random and arm entries, time
spent in each arm, and the number of cups examined or ﬂaps lifted were measured. A
cup was noted as “examined” whenever a bird leaned from an adjacent perch toward the
cup or ﬂew from a perch to the steel mesh next to a cup. Cups closest to the ﬂoors of the
arms of the chamber could be examined by a bird that hopped along the ﬂoor to them. A
bird was given 30 sec to eat if it found the seed by lifting the ﬂap of a speciﬁc cup. Birds
were removed from the chamber if seed was not eaten within 20 min.

For the males used to pilot the spatial memory task, initial latencies and the
number of trials to reach criterion were compared by means of unpaired t-tests. For the
experimental birds, a mixed model analysis of variance (ANOVA) was used to determine
whether differences existed in these measures by treatment group, age and sex. A mixed
model analysis of variance was used to determine differences in the latencies to ﬁnd the
food between the hippocampal- and sham-lesioned animals in the subsequent testing
sessions, the change in the latencies over time, the number of incorrect cups examined or

ﬂaps lifted during those trials, and any potential interactions.

Song Preference Tests

Each zebra ﬁnch was placed in the chamber described in Figure 2.2. In order to
determine preference for one song over another, songs, burned onto compact disc and
played via a Sony CD Walkman (#D-E220), were broadcast synchronously from speakers
at each end of the apparatus (“approach zones”; Similar to procedures used by Clayton,
1988 and Miller, 1979b). The middle of the apparatus between the two ends was

considered a “neutral zone” (Clayton, 1988). Birds were placed in the chamber and

112

allowed to acclimate for 1 hour. Then, the songs of two zebra ﬁnch males (one from the
father and a dissimilar, novel conspeciﬁc song chosen randomly from a bank of six) or
those of a conspeciﬁc and heterospeciﬁc (both novel and randomly chosen from banks of
six) were played simultaneously for 20 min. For each bird’s ﬁrst preference test, the
song that should be preferred (father’s song over that of another conspeciﬁc, and
conspeciﬁc over heterospeciﬁc song) was played from the speaker on the side of the
room opposite that where the food was located in the spatial memory test (see below) to
eliminate the preference for a “zone” of the chamber that may be conferred by memory
for the food location. Following song delivery, birds were removed from the chamber
and returned to their individual cages. Behavior was recorded with a Sony Digital
Handycam (model DCR-TRV230) and uploaded to a PC as described above.

In nature, zebra ﬁnches respond to song with a suite of behaviors. Females, for
example, will ﬂy to and from a perch several times before beginning to “hop” from perch
to perch near the song of a male, and will produce “tet” and distance calls in a male’s
presence (Zann, 1996). In the laboratory, tests of song recognition and preference have
measured the amount of time a bird spends near a speaker broadcasting a particular song.
Previous studies using this paradigm have described song discrimination as spending
twice as much time in one approach zone than the other for a signiﬁcant portion of the
trial (Clayton, 1988). In pilot work described below, I also measured activity during song
exposure by recording the number of calls and movement (turns on a perch and jumps
from perch to perch) made in each zone of the chamber. Difference scores for each
measure (responses to father minus other conspeciﬁc and zebra ﬁnch song minus

heterospeciﬁc song) were Obtained and analyzed between the sham and lesion groups

113

within each treatment age and sex by ANOVA. For the 8 females used to pilot this

procedure, differences scores for each measure were compared using one sample t-tests.

Male Song Learning and Production

The father of each subject was determined by observing the male that entered a
particular nest box consistently over a number of days. Fathers were temporarily
removed and their songs recorded in the presence of a novel female. These songs were
later uSed as indicated below (1) to compare to the songs of their male offspring to
determine the inﬂuence of hippocampal lesions on song learning, and (2) as one of two
simultaneously played conspeciﬁc songs in tests Of song preference. All songs were
recorded via an Optimus boundary microphone (Radio Shack) connected to a Sony
Digital Handycam (model DCR-TRV230) and uploaded to a PC as Windows Media
Video ﬁles using Windows Movie Maker (version 2.1.4026.0). Each .wmv ﬁle was
converted to Audio Video Interleave format using RiverPast Video Cleaner Lite (version
6.5.0.50717). Audio was extracted from the .avi ﬁles using Adobe Audition (version 1.0)
and background noise was reduced in fathers’ songs only (since only they were used for
tests of song preference) using a Fast Fourier Transform size of 512 points. All song ﬁles
were saved as Windows PCM (.wav) with a sample rate of 44,100 Hz and a resolution of
16 bits.

All song was recorded when birds were adults. To determine whether lesions of
the hippocampus affect song production or stability in adulthood, songs pro-lesion (and
pre-sham) were recorded from adult males by placing each in a cage with a novel female.

Most males sang within 20 min following the introduction of a female, so to ensure equal

114

experience between the sexes, females were placed in the same room in a cage with a
novel male for 20 min before and after the lesion/sham. Following the lesion (after
approximately 9 days; range = 7 to 10 days), adult males were again tested for song
production in the presence of a female. Over development, the songs of male zebra
ﬁnches become almost identical to those of their tutor, usually their father (Immelmann,
1969). To determine whether lesions of the hippocampus affect the matching of tutor
song, the songs of males lesioned as juveniles were recorded in adulthood in the presence
of a novel female. The songs of males lesioned at d20 were recorded at an average age of
approximately 152 days post-hatch, and those lesioned at d45 at d148. Females lesioned
or sham-lesioned as juveniles were put in a cage with a novel male once for 20 min.

The comparison of songs between tutor and pupil or pre- and post-manipulation
has been a longstanding means of determining the similarity of songs and/or whether a
particular neural or behavioral manipulation impacts song learning, production or
stability (Jones, ten Cate, & Bijleveld, 2001). Zebra ﬁnch song consists of syllables or
“notes” that, when viewed as a sonagram (see example in Figure 6.2), appear as distinct
frequency traces separated by intervals of silence or with abrupt changes in amplitude
(Zann, 1996). The order of these notes, including the intervals of silence between them,
are grouped and repeated in “phrases” or “motifs” that in total last roughly 3-4 see,
resulting in a “song.” Song comparisons made by human observers involve the
examination of qualities of song such as the types, order and frequencies of notes as well

as the length of silent intervals. While a basic visual assessment of song is necessary

115

Frequency (Hz)

 

phrase I phrase 2 phrase 3

 

01 1'1234 5123451234

t—I

 

Time (seconds)

Figure 6.2. Sonagram (A; frequency over time) and spectral derivative (B; change in
power over time) of the same song portion from a male zebra ﬁnch. Introductory (i)
and individual notes (identiﬁed as 1-5) are indicated. Zebra ﬁnch song consists of
introductory notes that are followed by notes of particular frequencies and changes in
amplitude that are repeated in a ﬁxed order (a “phrase;” three phrases indicated).
Note the “static” appearance of the background in the sonagrarn (A) and the
elimination of that extraneous noise in the spectral derivative (B).

116

(and is used as part of the procedure in this experiment; see below), individual observers
may differ considerably in determining which elements of song match and to what degree
(Jones et al., 2001).

To eliminate the subjective bias or inexperience that may lead to unreliable
determinations of song similarity, an automated procedure for measuring and comparing
the acoustic features of song has been developed (Tchemichovski, Nottebohm, Ho,
Pesaran, & Mitra, 2000). The freeware, Sound Analysis Pro ‘
(http://ofer.sci.ccny.cuny.edu/html/body_sound_analysis.htrnl), creates spectral ‘
derivatives of song syllables based on the change of power over time, unlike traditional
sonagrams that represent the power of sound as a change in frequency over time.
Because of this, background noise is omitted almost entirely, and syllables and the
phrases they comprise are more visually distinct (Figure 6.28).

Similarity scores of “pupil” versus “tutor” songs (for birds lesioned as juveniles)
and songs “before” versus “after” lesion (for birds lesioned as adults) were obtained as
follows. First, in Sound Analysis Pro, adjustments were made to each song based on
guidelines detailed in the user’s manual (Tchemichovski, Swigger, & Mitra, 2004). The
amplitude thresholds of songs were calibrated by adjusting an unscaled dB sound level
within a frequency range of 860-8600 Hz. For each song, amplitude was adjusted to
eliminate extraneous points in the tracing that were clearly not associated with a syllable,
and Wiener entropy, a measure of tonal versus harsh sounds, was adjusted to eliminate
low amplitude background noises. Similarity scores were then calculated, with one
phrase from each song used for comparison (see below). The overall similarity score

calculated by Sound Analysis Pro is an aggregate of three values: (1) percent similarity,

117

the percentage of a tutor’s sounds (individual note types) present in the song of the pupil
or between a bird’s song before and after a manipulation; (2) mean accuracy, the spectral
characteristics of those individual notes; and (3) sequential match, a comparison of the
temporal order of the notes, their duration, and the length of silent intervals between them.
Pupil versus tutor songs were compared asymmetrically (Figure 6.3; used when one song
is a model for another; Tchemichovski et al., 2004), and the songs of birds before and
after hippocampal lesion were compared symmetrically (Figure 6.4; used when one song
is not categorized as a model for another).

All comparisons were done blind to treatment condition. Unpaired t-tests were
used to determine whether similarity measurements in birds lesioned as adults differed
signiﬁcantly from the adult sham-lesioned controls. In addition, the sonagrams of adult
birds before and after lesion were qualitatively examined to determine if notes were
added or subtracted, and whether the order of the notes remained consistent following the
lesion. Similarity measurements for birds lesioned at d20 and d45 were analyzed using a

factorial ANOVA for age and treatment.

Tissue Collection

Following behavior testing, animals were given an overdose of Equithesin and
perfused with 0.1 M phosphate buffered saline (PBS) followed by 4% paraformaldehyde
in PBS. Brains were post-ﬁxed for one hour in paraformaldehyde, sunk overnight in 30%
sucrose in PBS, cut frozen on a cryostat into 30 um coronal sections onto Superfrost Plus
slides (Fisher) and thionin stained as described in Chapter 5. The hippocampus of each

lesioned bird was reconstructed using Neurolucida software (version 6.02.1,

118

Tutor

4 i
‘A

ll !\

inf!"

‘. Hid}
411),,
(I...

. :Yf’

 

Figure 6.3. Example of an asymmetric similarity matrix obtained by comparing
a phrase from the song of a male zebra ﬁnch (tutor) and another from the song
of one of his male offspring (pupil). Yellow lines indicate the boundaries of
individual notes or phrases (and are used to obtain the “percent sequential
match” score). High similarity values are restricted to the diagonal, indicating
that the individual notes in the tutor song were learned and matched by the
pupil in sequential order. Warmer colors along the diagonal line indicate
higher similarity (“percent similarity”). “Mean accuracy” measures the power
of the individual notes on a scale from white (high match) to gray to black (no
sound or no match).

119

After lesion

Before lesion

 

 

i“
14,—; n____ ‘ 1 ‘ u H "a-

 

 

Figure 6.4. Example of a symmetric similarity matrix obtained by comparing a
phrase from the song of a male zebra ﬁnch before and after a lesion of the
hippocampus. High similarity values are restricted to the diagonal, indicating
that the individual notes in the song pre-and post-lesion are in sequential order,
and the warmer colors along the diagonal line indicate higher similarity
(“percent similarity”). Overall, “mean accuracy” between the individual notes
(indicated by gray and white colors) and the temporal order of the notes
(“sequential match;” black bars) are high.

120

MicroBrightField, Inc.). Contours were traced in every fourth section around the border
of the hippocampus and the area within the structure in which cell bodies were destroyed
to conﬁrm the accuracy and extent of the hippocampal lesion. Total hippocampal and
lesion volumes were computed using NeuroExplorer (version 4.01.1, MicroBrightField,
Inc.) to determine the percent of the structure ablated in each animal injected with

ibotenic acid.

RESULTS AND DISCUSSION OF VALIDATION OF METHODS
Spatial Memory

Following the hippocampus or sham lesion, males did not signiﬁcantly differ in
their initial latency to eat from the baited cup (3 (7) = 0.90, p = 0.398) or the number of
trials needed to reach the criterion level (t (7) = 0.14, p = 0.894; Figure 6.5). Analyzing
across the last three acquisition trials (during the last of which, criterion was reached),
there was no signiﬁcant difference in latency to eat from the baited cup between the
groups (E (1, 14) = 0.73, p = 0.422) nor any interaction with time (F (2, 14) = 0.65, p =
0.537). For both groups, latency to eat from the baited cup decreased over the last three
trials and approached signiﬁcance (13 (2, 14) = 3.49, p = 0.059).

Latency to eat from the covered, baited cup in the ﬁrst probe trial was 600.25 +/-
99.73 sec in sham-lesioned males, an approximately 50% increase from the latency to eat
from the opened cup during the last acquisition trial (1(3) = 2.73, p = 0.072).
Hippocampal-lesioned males had an average latency 961.6 +/-112.50 sec in the ﬁrst
probe trial, up approximately 60% from the last acquisition trial (1(4) = 2.70, p = 0.054).

No signiﬁcant difference in latencies were seen between the treatment groups over the

121

 

 

1600 1 3

I

I
A 1400 -
o I
3 I t:
Z 1200 ' as»
:3 q I " 6 :0-0
o I 5
B I .c:
,4: 1000 - . 8
.8 : g
E 800 - : L 4 ,2
o 600 - : ,H
O O
H | H
§ 400 - I -2 3
6 I E
’5 I 2
-‘ 200 - g

I

0 . . 1 . ' o

 

 

 

 

 

 

 

initial last three
latency acquisition trials

- Hippocampal lesion (n = 5)
1: 311311101 = 4)

Figure 6.5. Performance during the acquisition phase of a spatial memory test by ﬁve
hippocampal-lesioned and four intact male zebra ﬁnches. Initial latency (sec) to eat
from the baited cup by the food deprived birds upon ﬁrst exposure to the chamber,
the average time needed by each group during the last three acquisition trials, and the
average number of trials needed to reach criterion are indicated. Latencies of both
groups of birds decreased signiﬁcantly over the last three acquisition trials. No other
signiﬁcant main effects or interactions were found.

122

four probe trials (E (l, 21) = 2.97, p = 0.129), but there were differences in their patterns
across the trials. Both groups showed a decrease in latency to eat from the baited cup
from trial one to four (F (3, 21) = 12.14, p < 0.0001). The decrease in average latencies
by the lesioned birds was steeper (Figure 6.6), and a signiﬁcant interaction between
latencies and treatment was found (E (3, 21) = 3.27, p = 0.042).

Overall, sham-lesioned birds spent signiﬁcantly more time in the goal arm in the
probe trials than hippocampal-lesioned animals (Figure 6.6; E ( 1, 21) = 11.59, p_ = 0.011).
No signiﬁcant change in time spent in the goal arm occurred across the trials (_F_’ (3, 21) =
1.27, p = 0.312), and this factor did not interact signiﬁcantly with treatment group (If (3,
21) = 0.87, p = 0.470). Hippocampal-lesioned birds examined signiﬁcantly more
incorrect cups or lifted the ﬂaps of those cups than control birds over the four probe trials
(Figure 6.7; E (l, 21) = 11.84, p = 0.011). Mistakes were made by only two control birds
on the second probe trial. The mistakes by the lesioned birds decreased signiﬁcantly over
the probe trials, resulting in a signiﬁcant trial effect (E (3, 21) = 3.31, p = 0.040). No
signiﬁcant interaction between treatment group and incorrect cups examined over the
trials was found (E (3, 21) = 2.35, p = 0.101).

The hippocampus of each lesioned bird in this pilot work was reconstructed as
detailed above and, on average, just over 11% (range: 4.1-18.5%) of the volume of the
hippocampus in lesioned birds was destroyed by the excitotoxin. Simple regression was
used to determine whether the amount of hippocampal damage correlated with any of the
behavioral measures. NO signiﬁcant correlations were found when the percent of
hippocampal lesion was compared with the initial latency to eat from the baited cup

during acquisition, the number of trials to reach criterion, the difference in latency to eat

123

 

 

 

 

 

 

 

 

 

 

 

 

 

1200 I 100

I
o ' F -—
30’. 1000 - I — T 8 0
9 . g
o I
B 800 - l s
'3 l ‘60 ..‘é’
E l o
3 600 - : .5
3 : .. 40 E
i 400 i : 3
: °‘
3' 200 - : l 20

I

l

0 I I I I — — — —— 0

Probe trial: 1 2 3 4 1 2 3 4

- Hippocampal lesion (n = 5)

[:1 Sham (n = 4)
Figure 6.6. Average latencies to eat from the baited cup and percent of total time in the
chamber spent in the goal arm by hippocampal- and sham-lesioned male zebra ﬁnches

over four probe trials. Sham-lesioned birds spent signiﬁcantly more time in the goal
arm over the four probe trials than lesioned birds.

124

Mean (+ SEM) number of incorrect cups

 

 

 

examined/ﬂaps lifted

 

 

 

   

 

O .
Probe trial: 1 2 3 4

- Hippocampal lesion (n = 5)

E] Sham (n = 4)
Figure 6.7. Mean (+ SEM) number of incorrect cups examined or ﬂaps of those
cups lifted by hippocampal- and sham-lesioned male zebra ﬁnches over the four

probe trials. Hippocampal-lesioned birds examined signiﬁcantly more incorrect
cups or lifted the flaps of those cups than control birds.

125

from the baited cup between the last acquisition and ﬁrst probe trials, the number of
incorrect cups or ﬂaps lifted in the ﬁrst probe trial, the percent time in the goal arm
during the ﬁrst probe trial, or the latency to eat from the baited cup in the ﬁrst probe trial
(all E < 8.52, all p > 0.062).

To my knowledge, three published studies have examined the spatial memory
capabilities of zebra ﬁnches. In two of these studies (Oberlander et al., 2004; Patel et al.,
1997a), food-deprived adult zebra ﬁnches were trained to remember the location of a
food source in a compartment covered by a cardboard ﬂap. Other equally-sized
compartments were also covered with ﬂaps but contained no food. Male and female
zebra ﬁnches with bilateral hippocampal lesions (Patel et al., 1997a) were inhibited in
their retrieval of the food source location as indicated by the number of ﬂaps lifted over
compartments that did not contain food, and there was no sex difference in performance.
In a separate study using a similar task, the performance of males treated with estradiol
improved more rapidly than that of control-treated ones (Oberlander et al., 2004).
Another study examined the ability of hippocampal-lesioned adult males to ﬁnd
previously located seed in one of four feeders on the ﬂoor of an experimental chamber
(Watanabe & Bischof, 2004). Lesioned birds made more visits than sham-lesioned
controls to feeders that did not contain seed. However, lesioned males did not
signiﬁcantly differ from controls in their latency to ﬁnd the seed, suggesting use of an
alternative, non-spatial strategy.

Zebra ﬁnches, however, do not cache food, and therefore these “window
shopping” tasks, commonly used with food-storers such as those in the parid and corvid

families (Capaldi, Robinson, & F ahrbach, 1999; Krebs, Sherry, Healy, Perry, &

126

Vaccarino, 1989; Macphail, 2002), may not be suitable or relevant. Although zebra
ﬁnches in the wild search for seed primarily on the ground, they prefer to remain perched
when not foraging (Zann, 1996). The task I designed may more accurately reﬂect the
natural behavior of the animals and eliminate stress associated with searching for seed on
the ground (and that may also be compounded by the food deprivation).

The setup of the apparatus I designed allows for the recording of additional
dependent variables besides the number of incorrect locations examined for a previously
visited food source detailed in the studies above. Birds must choose one of two arms that
via intra-maze cues are identical, but that can be differentiated based on the diverse extra-
maze cues present. The arm entered upon placement into the base of the chamber is one
indication of memory for a spatial location. However, an arm entered ﬁrst may not be an
accurate indicator of memory for the food source because I place the birds in the chamber
and then leave the room; my movements and the sound of the testing room door closing
cause birds to initially ﬂy around within the chamber. This is why thirty seconds was
chosen as the time for arm entry to reach criterion during the acquisition trials and
whether the arm with the baited cup was entered “ﬁrst” in the probe trials. After removal
of distractions, birds make a choice to enter or remain in a particular arm, and given the
length of food deprivation and a memory for the food source, correct arm entry for birds
tested in the pilot study was within that timeframe. In addition to choosing the correct
arm, the arm birds spend the most time in is also indicative of their memory for the food
that was previously located in that arm. This type of variable cannot be measured in the
experiments brieﬂy described above (Oberlander et al., 2004; Patel et al., 1997a;

Watanabe & Bischof, 2004) given the size and setup of the chambers. The number of

127

incorrect cups examined or ﬂaps liﬁed in a speciﬁc arm of the chamber I designed may
therefore be a more precise measure of spatial memory performance.

In this pilot work, lesioned males were clearly deﬁcient in their ability to
accurately locate the seed, but like hippocampal-ablated birds in another study (W atanabe
& Bischof, 2004), they were, over time, able to ﬁnd the seed in a manner consistent with
the performance of the control birds, suggesting again some type of non—spatial strategy.
Rats can switch from place learning dependent on the hippocampus to response learning
dependent on the caudate nucleus (Packard & McGaugh, 1996). It is reasonable to
suggest that a similar switch could occur in avian species as well. Importantly, the initial
deﬁcit in performance I show in hippocampal-lesioned birds, coupled with histology
indicating cell loss, is proof that the structure is not functioning normally due to loss of
the dorsolateral subdivision and that the apparatus is a reliable measure with which to

detect the associated decrements in spatial memory behavior.

Song Preference

The amount of time spent or behaviors displayed in the side of the chamber
playing heterospeciﬁc song was subtracted from that observed in the conspeciﬁc song
side, and one sample t-tests were used to determine whether a preference existed for one
side over another based on a hypothesized difference of zero for the difference score.
The amount of time females spent in the side that played conspeciﬁc song was
approximately 20 times more than in the side of the chamber that played heterospeciﬁc
songs (t (7) = 17.81, p < 0.0001). Other measures were greater on average in the

conspeciﬁc side of the chamber, but only approached signiﬁcance. On average, about six

128

times more calls were made in the conspeciﬁc song side of the chamber (t (7) = 2.06, p <
0.078). Approximately ﬁve times more turns on a perch were made in the side of
conspeciﬁc song than the heterospeciﬁc song side (1(7) = 2.32, p = 0.053). Four and-a-
half times more jumps from perch to perch were made in the conspeciﬁc side of the
chamber (t (7) = 2.05, p = 0.080). As in Chapter 1, calls and movement (turns and perch
jumps) were summed and analyzed collectively. In the conspeciﬁc song side, about six
times more calls and movement were made than in the heterospecifrc song side (t (7) =
2.17, p = 0.067). Finally, the number of entries birds made into each side of the chamber
did not statistically differ (t (7) = 0.19, p_ = 0.852).

In the pilot work, females’ preference for conspeciﬁc song was rather robust.
Although only time spent in the side of conspeciﬁc song was signiﬁcantly different from
that in the heterospeciﬁc song side, other behavioral measures (calls, turns, perch jumps)
were greatest in the zebra ﬁnch song side and approached signiﬁcance, perhaps based on
the small sample size. Other studies have shown that male and female zebra ﬁnches
spend more time in the side of a chamber near conspeciﬁc compared to heterospeciﬁc
song (Braaten & Reynolds, 1999; Lauay et al., 2004). The apparatus used for preference
tests is similar to those used in prior studies (Clayton, 1988; Miller, 1979b) and the pilot
data demonstrate its validity in the conﬁrmation of a song preference.

Prior research using a number of different paradigms demonstrate that male
(Adret, 1993; Clayton, 1990) and female (Clayton, 1988; Miller, 1979b; Riebel et al.,
2002) zebra ﬁnches prefer their father’s (tutor’s) song to that of a novel conspeciﬁc or
their own song (Adret, 1993). Given these ﬁndings, taken together with the robust

preference by females for the conspeciﬁc over heterospeciﬁc song side of a chamber

129

detailed above, the test of preference for a father’s song versus that of a novel conspeciﬁc

was not piloted.

Song Production

As mentioned above, the song of a zebra ﬁnch typically consists of one phrase
that is repeated several times. To ensure that the selection of only one phrase would be
adequate in the determination of song similarity, multiple phrases from the songs of a
tutor and his pupil and the songs of a male before and aﬁer lesion of the hippocampus
were compared. Similarity measurements from the comparisons of four separate phrases
from a tutor and pupil song are shown in Table 6.1. The percent similarity and mean
accuracy were on average high between all the phrases examined. Percent sequential
matches were low in four phrase comparisons due, I believe, to extraneous auditory
stimuli and placement of the microphone. Since these males sang in the presence of a
female, other sounds such as calls to the male by the female and courtship behaviors by
both sexes (ﬂying away from the perch and back, beak wipes against the perch or cage,
tail quivers, dances on a perch, etc.; Zann, 1996) were picked up by the microphone.
Also, it is possible that the microphone may not have picked up portions of a song
depending on the direction a male was facing when singing. In the spectral analyses
provided by the program, it is clear that in some phrases the power of individual notes is
not as strong as the same notes in other phrases. These factors could have resulted in
tracings that affected the power and sequential matches of particular song phrases

determined and analyzed by Sound Analysis Pro.

130

 

Comparison Phrase

 

 

 

Percent Percent
Similarity Percent Mean Sequential
Tutor Pupil Score Similarity Accuracy Match
1 1 62.3 87 71.7 100
l 2 67.4 96 70.2 100
1 3 39.9 96 79.2 4.9
1 4 38.9 97 77.9 2.9
2 1 75.6 95 79.6 100
2 2 72.2 93 77.7 100
2 3 72.6 92 78.9 100
2 4 69.7 92 75.8 100
3 1 73.2 94 77.9 100
3 2 41.6 97 78 9.9
3 3 41 98 77.3 8.3
3 4 57.9 82 80.9 74.6
4 1 55.2 75 73.7 100
4 2 67.6 98 69 100
4 3 49.2 82 68.4 75.6
4 4 48.2 85 67.4 68.3
MEAN: 58.3 91.2 75.2 71.5
SEM: 13.5 6.9 4.5 40.2

 

Table 6.1. Similarity measurements obtained by comparing four different phrases (1-4)
of a male zebra ﬁnch’s song and those (1-4) of one of his male offspring. The “similarity
score” is an aggregate of the three other values, which indicate parallels in the types of
notes used (percent similarity), the frequencies of those notes (mean accuracy), and the
lengths of silent intervals between the notes (sequential match). Means and standard
errors for the measures are indicated at the ends of each column.

131

The same type of analysis was performed on three separate phrases from the
songs of a zebra ﬁnch pre- and post-sham lesion (Table 6.2). Although one would expect
the similarity between the song phrases from an unmanipulated bird to be very close to
100 percent, extraneous noise at the time of recording or proximity to the microphone
when singing would affect song similarity as described above. In two out of nine
comparisons, the percent similarity was less than the 90% or above desired for tests of
self-similarity (Tchemichovski et al., 2004), in contrast to the range of 94-100 for the
remaining seven comparisons. It is likely, again, that in these instances other auditory
stimuli were factoring into the analysis.

The comparisons of multiple phrases in both types of song analyses (tutor vs.
pupil and pre- vs. post-lesion) yielded generally similar results. For this reason, and due
to the potential confounds mentioned above, only one phrase from each song type was
used for comparison. A phrase was picked (1) if it was clear that no extraneous noises
were part of it and (2) by the “sharpness” of its spectral derivative.

Given the overall equivalence in similarity measurements of individual phrases
from the songs of a tutor and his pupil (Table 6.1) and those from a male before and
following a lesion of the hippocampus (Table 6.2), Sound Analysis Pro appears to be a
valid and reliable means of determining the relationship between song phrases. Also,
similarity scores I obtained are equivalent to others that related tutor and pupil song
phrases (Liu, Gardner, & Nottebohm, 2004) and examined the stability of adult song
following lesion of a speciﬁc brain region (Coleman & Vu, 2005). The additional
analysis comparing song phrases of adults pre- and post-lesion or sham by eye involves a

straightforward, qualitative detection of note order. This method is standard in

132

 

Comparison Phrase

 

 

 

Percent Percent
Similarity Percent Mean Sequential

Before After Score Similarity Accuracy Match
1 1 87.2 97 91.6 97.7
1 2 65.2 83 83 95.3
1 3 89.1 96 93.8 98.5
2 1 87.9 98 92.1 97.5
2 2 65.5 83 82.8 94.9
2 3 90.9 97 95.2 98.1
3 1 95.9 100 96.7 99.3
3 2 81.4 94 89.1 97.2
3 3 95.2 99 96.1 99.9
MEAN: 84.3 94.1 91.2 97.6
SEM: 11.6 6.5 5.2 1.7

 

Table 6.2. Similarity measurements obtained by comparing three different phrases of a
male zebra ﬁnch’s song before and aﬁer a sham lesion of the hippocampus. The
“similarity score” is a product of the three other values, which indicate parallels in the
types of notes used (percent similarity), the frequencies of those notes (mean accuracy),
and the lengths of silent intervals between the notes (sequential match). Means and
standard errors (SEM) for the measures are indicated at the ends of each column.

133

determining song stability following a behavioral or physiological manipulation and is
not open to subjective bias as are measurements of the speciﬁc frequency and temporal

characteristics of song (Jones, ten Cate, & Bijleveld, 2001).

RESULTS — EXPERIMENTAL BIRDS

Histology

The amount of hippocampal damage produced by the excitotoxin ranged from
approximately 10-23% for the birds used in this study (see Figure 6.8 for a sample
bilateral hippocampus lesion). The effect of age of lesion on the amount of hippocampal
damage observed in adulthood approached signiﬁcance (E (2, 24) = 3.15, p = 0.061), with
birds lesioned at d20 having the least damage, birds lesioned as adults the most, and those
at d45 an amount intermediate between those two (Figure 6.9). There was no effect of
sex (E (1, 24) = 0.03, p = 0.869) or interaction between age and sex (E (2, 24) = 0.36, p =

0.704) on the percent damage to the hippocampus.

Spatial Memory

Birds with hippocampal lesions did not signiﬁcantly differ from their sham
controls in their initial latencies to eat from the uncovered, baited cup during the ﬁrst
acquisition trial (all E < 1.61, all p_ > 0.211 for main effects of age, sex, treatment group
and interactions) or the number of trials needed to reach the criterion level during

acquisition (all E < 1.74, all p > 0.187; data not shown).

134

 

Figure 6.8. Coronal section through an adult male zebra ﬁnch brain at the
level of the hippocampus detailing areas of cell death (traced by dashed
lines) produced by injection of ibotenic acid at d20. Scale bar = 200 pm.

135

Percent hippocampal damage

 

25

20-

 

 

 

 

 

 

 

d20 d45 adult

- males
'3 females

Figure 6.9. Mean (+ SEM) percent hippocampal damage measured in
adulthood in male and female zebra ﬁnches lesioned at d20, d45 or as adults.
Amount of damage to the hippocampus was determined relative to total
hippocampus size in each bird. The effect of age of lesion on amount of
hippocampal damage measured in adulthood approached signiﬁcance.

136

However, latencies to eat from the covered, baited cup during the four probe trials
was signiﬁcantly different between the combined sham and lesion groups (Figure 6.10; E
(2, 141) = 9.83, p = 0.003), and these latencies signiﬁcantly decreased over the course of
the probe trials for both groups (E0, 141) = 37.81, p < 0.0001). Further, the mean
latencies of males in both the sham and lesion groups decreased more abruptly over the
probe trials than all females (Figure 6.11; £(3, 141) = 2.95, p = 0.035). No other main
effects or two-, three-, or four-way interactions were signiﬁcant (all E < 1.95, all p >
0.077).

The percent time birds spent in the arm that contained the baited cup over the four
probe trials was signiﬁcantly greater in birds that received the sham lesion (Figure 6.12;
E_(1, 141) = 26.02, p = <0.0001). It is expected that birds without a memory for the
spatial location of the seed would spend, on average, equal time in both arms (“chance”
level). In fact, lesioned birds spent approximately 59% in the goal arm during the ﬁrst
two probe trials, while sham-lesioned birds spent 77 and 84%, respectively (Figure 6.12).
Over the probe trials, the amount of time spent in the goal arm increased in both groups
(E_(3, 141) = 6.20, p = 0.0005). No other main effects or interactions were uncovered (all
5 < 1.41, all p > 0.244).

The most robust effects in this test of spatial memory were the number of
mistakes birds made during the probe trials, speciﬁcally lifting the ﬂap of or examining a
cup that did not contain seed. Overall, lesioned birds made signiﬁcantly more mistakes
than the sham controls (Li-(1, 141) = 64.52, p < 0.0001). The number of mistakes

depended on the age at which the animal was lesioned (£(2, 141) = 5.03, p = 0.011), and

137

Latencies (sec) to eat from baited cup

 

1200

l

1000

800 r

600 -

400 .

200

 

 

 

O I I I I
1 2 3 4

Probe trial

0 Lesioned birds (n = 30)
O Sham-lesioned birds (n = 29)

Figure 6.10. Mean (+/- SEM) latencies to eat from the baited cup across the
four probe trials for all lesioned and sham-lesioned birds. As a group,
lesioned birds differed signiﬁcantly from sham-lesioned birds, and both
groups showed a signiﬁcant decrease in their latencies across the trials.

138

7

Latencies (sec) to eat from baited cup

 

1 000

800 r

600 -

 

400 ‘

 

200 ~

 

 

 

 

O . u . .
l 2 3 4

Probe trial

0 Males (n = 30)
0 Females (n = 29)

Figure 6.11. Mean (+/- SEM) latencies to eat from the baited cup across the
four probe trials for all male and female birds. The latencies of male birds
decreased at a signiﬁcantly faster rate than those of females.

139

Percent time in goal arm

100

 

 

 

m
C
1

 

O\
O
I

A
O

N
O
1

 

 

 

 

 

 

 

 

 

 

Probe trial

- Lesioned birds (11 = 30)

1:] Sham-lesioned birds (n = 29)
Figure 6.12. Mean percent time in arm with baited cup by all hippocampal-
lesioned and control birds across the four probe trials. Over the course of

the probe trials, sham-lesioned birds spent signiﬁcantly more time in the
goal arms than lesioned birds.

140

the number of mistakes made by females tended to be greater than that of males (F_(l,
141) = 3.70, p = 0.060). Mistakes decreased over the course of the probe trials for all
animals (Ij_(3, 141) = 4.18, p = 0.007) but at different rates for the lesioned and sham-
lesioned birds (1(3, 141) = 6.94, p = 0.0002). The four-way interaction of mistakes, age,
sex and treatment also approached signiﬁcance (_F_(6, 141) = 2.03, p = 0.065),

and it was clear that birds made mistakes dependent on these factors (see below). No
other interactions were signiﬁcant (all E < 2.29, all p > 0.112).

Given the main effects and interactions indicated above, mistakes over the course
of the probe trials were broken down for further analysis. Examining the ﬁrst probe trial
only (Figure 6.13), lesioned birds made more mistakes than sham-lesioned birds (EU, 47)
= 46.87, p < 0.0001) regardless of age (£(2, 47) = 1.19, p = 0.314), sex (_F_(1, 47) = 0.18,
p_ = 0.670), or any interaction of those variables (all E < 2.08, all p > 0.137). Examining
within the treatment groups, mistakes by the sham-lesioned or lesioned birds were not
dependent on age at time of surgery or sex, nor was there an interaction (all _F_ < 2.3], all
p > 0.120).

In probe trial two, lesioned birds again made more mistakes than the controls
(Figure 6.14; E (1, 47) = 22.96, p < 0.0001) and this was dependent on age at the time of
lesion (E (2, 47) = 3.64, p = 0.039). Post-hoe tests revealed that adult birds made
signiﬁcantly more mistakes than d20 birds (Tukey/Kramer; p < 0.05); no signiﬁcant
differences were found between d20 and d45 birds or between adults and d45 birds. Also
in probe trial two, females made more mistakes than males, although this effect did not

reach signiﬁcance (E (1, 47) = 3.71, p = 0.061). None of the two- or three-

141

Number of mistakes — Probe trial 1

 

 

 

          

d20 d20 d45 d45 adult adult
males females males females males females

 

   

 

   

 

- Lesioned birds
I: Sham-lesioned birds

Figure 6.13. Mean (+ SEM) number of incorrect cups examined or ﬂaps of
those cups lifted in the ﬁrst probe trial by adult male and female zebra
ﬁnches lesioned or sham-lesioned at d20, d45 or in adulthood. Overall,
lesioned birds made signiﬁcantly more mistakes than sham-lesioned ones;
there were no signiﬁcant effects of age or sex or an interaction.

142

 

 

Number of mistakes — Probe trial 2

 

 

    

 

 

 

 

 

 

 

 

 

     

d20 d20 d45 d45 adult adult
males females males females males females

- Lesioned birds
1:] Sham-lesioned birds

Figure 6.14. Mean (+ SEM) number of incorrect cups examined or ﬂaps of
those cups lifted in the second probe trial by adult male and female zebra
ﬁnches lesioned or sham-lesioned at d20, d45 or in adulthood. Lesioned
birds made signiﬁcantly more mistakes than sham-lesioned ones, and this
effect depended on age at time of lesion. Effects of sex approached
signiﬁcance.

143

way interactions were signiﬁcant (all 1;“ < 2.52, all p > 0.091). Analysis of mistakes made
by sham birds only revealed an effect of sex (E (1, 23) = 7.98, p = 0.01): control females
made more mistakes than control males. There were no effects of age or an interaction of
age and sex (all E < 1.41, all p > 0.264). Mistakes by lesioned birds in probe trial two
were not dependent on sex (13 (1, 24) = 0.77, p = 0.389) but were affected by age (E (2, 24)
= 3.66, p = 0.041); there was no signiﬁcant interaction between the two variables

(E (2, 24) = 0.92, p_ = 0.414). Birds lesioned as adults made signiﬁcantly more mistakes
than birds lesioned at d20 (Tukey/Kramer; p < 0.05), but no signiﬁcant differences were
found between adult and d45 birds or between the d20 and d45 groups.

In probe trial three, lesioned birds again made signiﬁcantly more mistakes than
sham-lesioned animals (Figure 6.15; E (1, 47) = 47.94, p < 0.0001). Birds lesioned at
d20 made fewer mistakes than the other groups, and females made more errors than
males, although the effects of age and sex only approached signiﬁcance (E (2, 47) = 2.75,
p = 0.074 and E (1, 47) = 3.46, p = 0.069, respectively). None of the two- or three-way
interactions was signiﬁcant (all E < 2.24, all p_ > 0.117). Examining sham birds alone, no
signiﬁcant effects of sex, age or an interaction were uncovered (all F < 1.40, all p_ >
0.249). The mistakes by lesioned birds analyzed alone were not signiﬁcantly inﬂuenced
by the sex of the animal (E (1, 24) = 2.50, p = 0.127), but the effect of age approached
signiﬁcance (E (2, 24) = 2.94, p = 0.072). No interaction of sex and age was found (E (2,
24) = 0.99, p = 0.386).

More mistakes were made by lesioned birds in probe trial four (Figure 6.16; E (l,

47) = 5.77, p = 0.021). No other signiﬁcant effects or interactions were uncovered when

144

Number of mistakes — Probe trial 3

 

              

 

 

 

 

d20 d20 d45 d45 adult adult
males females males females males females

   

   

- Lesioned birds
Sham-lesioned birds

Figure 6.15. Mean (+ SEM) number of incorrect cups examined or ﬂaps of
those cups lifted in the third probe trial by adult male and female zebra
ﬁnches lesioned or sham-lesioned at d20, d45 or in adulthood. Lesioned
birds made signiﬁcantly more mistakes than sham-lesioned ones. Effects
of age and sex approached signiﬁcance.

145

Number of mistakes — Probe trial 4

 

  

 

 

 

 

 

 

 

   

d20 d20 d45 d45 adult adult
males females males females males females

- Lesioned birds
[:I Sham-lesioned birds

Figure 6.16. Mean (+ SEM) number of incorrect cups examined or ﬂaps of
those cups lifted in the fourth probe trial by adult male and female zebra
ﬁnches lesioned or sham-lesioned at d20, d45 or in adulthood. Lesioned
birds made signiﬁcantly more mistakes than sham-lesioned ones; there were
no signiﬁcant effects of age or sex or an interaction.

146

all animals were analyzed collectively or when control or lesioned birds were examined

alone (all E < 2.04, all p > 0.147).

Song Preference

Lesions of the hippocampus did not affect the preference for tutor song over the
song of a novel male zebra ﬁnch (Figure 6.17). No signiﬁcant main effects were
uncovered for age (E12, 47) = 0.74, p = 0.485), sex (£(l, 47) = 0.01, p = 0.933) or
treatment group (E_(1, 47) = 0.48, p = 0.494). No signiﬁcant interactions were found
between any of these variables (all E < 1.06, all p > 0.356). Only one group (females
lesioned in adulthood) did not prefer one song over another based on time spent in the
part of the chamber broadcasting a particular song (Figure 6.17).

Preference for novel conspeciﬁc song over the song of a heterospeciﬁc (Figure
6.18) was not affected by lesions of the hippocampus (£(l, 47) = 0.00, p = 0.961), nor
was it signiﬁcantly inﬂuenced by age (_F_(2, 47) = 0.41, p = 0.669) or sex (E1 1, 47) =
0.37, p = 0.546). None of the two- or three-way interactions between the variables were

signiﬁcant (all E < 2.45, all p > 0.098).

Song Production

Ablation of the hippocampus in adulthood did not affect song production in male
zebra ﬁnches. Percent similarity (t (8) = 0.33, p = 0.747), mean accuracy (1(8) = 0.40, p
= 0.699), percent sequential match (t (8) = 1.07, p = 0.317) and overall similarity scores (1
(8) = 0.45, p = 0.665) did not signiﬁcantly differ between the groups based on the

comparisons of phrases from songs recorded pre- and post-sham or lesion (Table 6.3).

147

 

 

1200

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

§ 1000 - "
8
w 800 . _-
Q) -P
8
23 600 l
«“3
ft:
1: 400 i
E
m 200 .
a:
C O . _—
(6
d)
2 -200 -
~400 . v u I . .

 

d20 d20 d45 d45 adult adult
males females males females males females

- Lesioned birds
[:I Sham-lesioned birds

Figure 6.17. Mean (+/- SEM) difference scores (amount of time (sec) spent in side of
chamber playing father’s song minus amount of time spent in side of novel male
conspeciﬁc song) of male and female zebra ﬁnches hippocampal-lesioned or sham-
lesioned at 20 days post-hatch (d20), d45 or adulthood. A positive score indicates a
preference for the father’s song, and a negative score a preference for the song of a
novel male conspeciﬁc. No signiﬁcant main effects or interactions were uncovered.

148

1200

 

1000 -

800 -

600 ~

 

 

400 ‘

 

200 ‘

 

 

 

 

 

 

Mean (+ SEM) difference scores

             

d20 d20 d45 d45 adult adult
males females males females males females

- Lesioned birds
[:J Sham-lesioned birds

Figure 6.18. Mean (+ SEM) difference scores (amount of time (sec) spent in side of
chamber playing conspeciﬁc song minus amount of time spent in side of
heterospeciﬁc song) of male and female zebra ﬁnches hippocampal-lesioned or sham-
lesioned at 20 days post-hatch (d20), d45 or adulthood. No signiﬁcant main effects
or interactions were uncovered.

149

 

Means (+/- SEM)

 

 

Similarity Percent Mean Sequential
Group Treatment Score Similarity Accuracy Match

d20 lesion 64.3 (5.9) 92.6 (1.5) 77.4 (0.6) 80.2 (17.7)
d20 sham 55.6 (3.5) 84.4 (3.8) 71.9 (2.5) 83.7 (6.8)
d45 lesion 65.9 (2.3) 89.8 (2.7) 74.4 (2.1) 95.0 (5.0)
d45 sham 64.6 (5.0) 84.4 (2.4) 76.5 (2.1) 90.3 (6.0)
adult lesion 86.3 (3.2) 96.8 (0.7) 92.8 (2.8) 96.9 (0.5)
adult sham 88.7 (4.25) 96.0 (2.3) 94.1 (1.5) 98.1 (1.1)

 

Table 6.3. Similarity measurements of song phrases from adult male zebra ﬁnches
lesioned or sham-lesioned at d20, d45 or in adulthood. For d20 and d45 birds, similarity
measurements were obtained by comparing one phrase from the songs of each of these
males with one from their father. In birds lesioned or sham-lesioned as adults, phrases
ﬁom individual males’ songs both before and after lesion (or sham lesion) were
compared. None of the similarity measurements from the lesioned birds differed
signiﬁcantly from those of the sham controls.

150

No additional notes or changes in note order were detected in any of the adult-lesioned
birds (Figure 6.19).

Analyzed together, the overall similarity scores of birds lesioned at d20 or d45 did
not signiﬁcantly differ (main effect of age: E (1, 16) = 1.47, p_ = 0.243; treatment: E (1,
16) = 1.30, p = 0.271; Table 4). Percent similarity with a tutor’s song phrase was not
inﬂuenced by age of lesion (E (1, 16) = 0.27, p = 0.613) but was, interestingly, higher in
the lesioned animals compared to the sham controls (E (l, 16) = 6.29, p = 0.023). The
mean accuracy and sequential matches between tutor and pupil’s song phrases was not
signiﬁcantly affected by the hippocampal lesion (E (1, 16) = 0.79, p = 0.387 and E (1, 16)
= 0.004, p = 0.952) or age at time of lesion (E (1, 16) = 0.17, p_ = 0.689 and E (1, 16) =
1.10, p_ = 0.310, respectively). No signiﬁcant interaction between age and group was
uncovered through analysis of any of the song similarity measures (all E < 3.73, all p >

0.071).

Regression Analyses

To conﬁrm whether the extent of a lesion correlated with behavior on a particular
task, simple regressions were performed within each age group for each dependent
variable in the experiment: initial latency to eat from the baited cup during the ﬁrst
acquisition trial, number of trials to reach criterion, latencies to eat from the baited cup,
percent time in the goal arm, and mistakes during the four probe trials, difference scores
in the preference tests (time in side of chamber broadcasting father’s song minus side

playing novel male conspeciﬁc song and novel conspeciﬁc song minus heterospeciﬁc

151

 

i: l ilmkﬁvr‘ﬁ

\~

 

 

$83: 1118b Its-m:
i 1 2 3 4 5 1 2 3 4
phrase I phrase 2

Figure 6.19. Spectral derivatives of samples of the songs of a male zebra ﬁnch before
(A) and aﬁer (B) a bilateral lesion of the hippocampus. Note that the types of notes
used and the order of those notes are identical before versus aﬁer the lesion.

152

song), and song similarity measurements (overall similarity score, percent similarity,
mean accuracy and sequential match) for males only.

In adult males and females, the amount of hippocampal damage was negatively
correlated with the amount of time spent in the goal arm on probe trial two (E (1, 9) =
5.64, p_ = 0.049). A positive correlation between the percent of hippocampus lesioned
and number of mistakes approached signiﬁcance on probe trial two (E (1, 9) = 5.05, p_ =
0.055) and reached signiﬁcance on probe trial three (E (1, 9) = 13.02, p = 0.007). No
other correlations, including males only and similarity scores, were uncovered (all E <
3.21, all p > 0.111). No correlations on any measure were uncovered for birds lesioned

at d20 (all E < 4.05, all p > 0.079) or d45 (all E < 8.37, all p > 0.063).

DISCUSSION

Lesions of the dorsolateral subdivision of the hippocampus, where I detected
immediate early gene activity in response to conspeciﬁc songs (Bailey et al., 2002;
Bailey and Wade, 2003; Bailey and Wade, 2005), had no effect on (1) the ability of birds
lesioned during development to learn song in order to normally respond to and produce it
in adulthood or (2) song stability and responses to song in adulthood, following lesions of
the region in developed animals. These results suggest that immediate early gene activity
in the hippocampus does not confer a role for the region in song learning per se, but they
do not exclude the possibility that the structure consolidates memories relating to song or
modulates the responses of other regions based on the salience of particular vocal signals
or environments in which they are heard or produced. Although not the primary focus of

this study, the spatial memory task developed is a valid and reliable tool with which to

153

a n u——‘—
4 .‘l
L

test the ability of normal male and female zebra ﬁnches to remember the location of a
particular stimulus in a fairly complex environment and to measure the attenuation in that
behavior that results from speciﬁc neural damage.

Ablations of different regions within the song circuit in zebra ﬁnches have
resulted in speciﬁc effects on song behavior. For example, lesions of Area X and lMAN
during the sensitive period for song learning prevent normal song development,
speciﬁcally by affecting the circuitry of the motor nucleus RA in the case of lMAN
lesions(Kittelberger & Mooney, 1999), while similar lesions in adult birds do not affect
song (Bottjer et al., 1984; Scharff & Nottebohm, 1991). Destruction in development or
adulthood of cells within HVC and RA disrupts song output (Scharff & Nottebohm, 1991;
Simpson & Vicario, 1990). The present results clearly indicate that the hippocampus
does not inﬂuence song learning or production like these “traditional” song control nuclei.
Based on similarity scores with tutor phrases, lesions of the hippocampus at d20 and d45
did not disrupt formation of a song template or the later sensorimotor integration required
for song stabilization. In adult males, song remained stable following the hippocampus
lesion, again suggesting no involvement of the region in song maintenance or expression
of the song template. In all birds, normal song preferences for tutor song over that of a
novel male conspeciﬁc and conspeciﬁc over heterospeciﬁc song were maintained, unlike,
for example, lesions of CMM in adult female zebra ﬁnches that result in the display of
similar responses to both heterospeciﬁc and conspeciﬁc song (MacDougall-Shackleton et
al., 1998).

The question remains as to why, given that neurons in the hippocampus respond

selectively to conspeciﬁc songs, do lesions of the structure not impact normal song

154

preferences? Other than the possibility of encoding information about the environment in
which relevant auditory signals are heard (discussed in further detail below), the
hippocampus may modulate the activity of other regions to apportion proper responses to
speciﬁc song stimuli. Evidence for this potential function of the hippocampus comes
from male zebra ﬁnches that underwent their ﬁrst courtship of a female: F OS
immunoreactive neurons were observed throughout the hippocampus (Sadananda &
Bischof, 2002), expression that was higher than that seen in males from an aviary with
social and sexual contact with conspeciﬁcs or males chased around a cage by a human
experimenter. Again, this activity within the structure broadens the role of the
hippocampus over and above spatial memory processing. Immediate early gene activity
in the hippocampus following acquisition in a spatial memory paradigm (see Herdegen &
Leah, 1998) may not necessarily infer the sole function of the region during the task.
Hippocampal c-FOS mutant mice exhibit normal spatial learning compared to control
mice in the Morris water maze task, suggesting no involvement of the immediate early
gene in this type of memory consolidation. Therefore, activity within the region may be
modulatory in nature, affecting other brain regions to gauge the relevance of particular
situations and stimuli, and linking them appropriately with some sort of behavioral
response.

The most robust ﬁndings of the present study are the differences in spatial
memory ability between control birds and hippocampal-lesioned animals that mirrors
effects obtained in this and other avian species and provides further evidence for a
functional homology with the mammalian hippocampus. Food-deprived zebra ﬁnches

with lesions to the hippocampus are deﬁcient in their ability to locate a previously found

155

food source (Patel eta1., 1997a; Watanabe & Bischof, 2004). Also, spatial memory in
pigeons (Bingman, Bagnoli, Ioalé, & Casini, 1984) and black-capped chickadees (Sherry
& Vaccarino, 1989) is disrupted by lesions to the hippocampus. In the present study,
lesioned birds displayed signiﬁcantly higher latencies than the controls to eat from the
baited cups during the probe trials, spent signiﬁcantly less time in the arm that contained
the seed, and made more mistakes at locating the food than the sham-lesioned animals.
Interestingly, the performance of lesioned animals increased over the course of the probe
trials. A gradual decrease in latencies to eat from the baited cup, an increase in the time
spent in the arm that contained the cup, and decreases in mistakes over the probe trials all
indicate, perhaps, that alternate strategies were used to determine the location of the food
source and that learning occurred outside of the realm of spatial processing as in a prior
study (W atanabe & Bischof, 2004), or that what remains of the hippocampus could
function better with time. The mechanisms through which this compensation may occur
have not been studied in avian species. It is possible, however, that the 10 min interval
between probe trials was short enough to permit birds in both groups to retain the
location of the baited cup in short-term memory. There is no doubt, however, that
lesioned birds were initially deﬁcient in spatial long-terrn memory based on their
performance in the ﬁrst probe trial, which followed the acquisition trials by 1 hr and is
thus outside the domain of short-term memory.

Two other ﬁndings from the present results on spatial memory ability in zebra
ﬁnches also warrant further attention. First, the performance of birds lesioned at d20 was
better than adults on a few measures (see Results). Also, although not statistically

different, the percent lesion to the DLHP was lower in birds lesioned at d20 than those

156

lesioned in adulthood. To my knowledge, neurogenesis in the zebra ﬁnch hippocampus
has not been reported, although environmental change induces neurogenesis in other
brain regions, such as HVC and Area X (Lipkind, Nottebohm, Rado, & Barnea, 2002).
Given the occasional increased performance of these birds, their smaller lesion volume,
that neurogenesis within the avian brain (Nottebohm, 2002a; Patel, Clayton, & Krebs,
1997b) is well documented, and that the hippocampus in mammals is a primary site for
functional neurogenesis (van Praag, Schnider, Christie, Toni, Palmer, & Gage, 2002), this
is a possibility. It is also unknown whether the lesion occurred before a period of
increased development of the hippocampus, such that an increase in cell birth within the
region following the lesion may have compensated slightly for the loss of neurons.

Also interesting is the occasional sex difference in ability to remember the
location of the baited cup. Females made more errors than males in some of the probe
trials, and their latencies to eat from the baited cup declined at a slower rate than males
over the four probe trials. Males of many species show consistent advantages in a variety
of spatial tasks (see J onasson, 2005). Zebra ﬁnches do not migrate, are not territorial,
and both sexes forage for food and take care of the young (Zann, 1996), which suggests
equal spatial memory demands between the sexes, yet several instances of male-biased
performance are seen. Although this difference is not robust throughout the probe trials,
due possibly to the number of subjects in each group, additional questions remain as to
Whether the morphology, neurochemistry or electrophysiological activity of the

hippocampus differs between the sexes in the zebra ﬁnch, or whether differences in non-

SPatial learning tasks exist.

157

This experiment serves as an important ﬁrst step in determining the involvement
of the hippocampus in song-related behavior. Although the region does not appear to be
important for song learning directly, the song-speciﬁc responses of the region (Chapters 3
and 4), its afferent and efferent connections (detailed in Chapter 5), and its role in spatial
memory (current Chapter) are suggestive of, at the very least, a modulatory role in vocal

communication in zebra ﬁnches.

158

CHAPTER SEVEN

Zebra ﬁnches are an ideal model system for the examination of multiple memory
systems, in particular for exploring how they develop and work in concert to control
discrete, readily observable behaviors. The results presented in the preceding chapters
begin to illustrate several points regarding particular aspects of song learning in female
zebra ﬁnches, neuronal responses to song during development, and the potential role of
the hippocampus in song—related behavior in zebra ﬁnches. These points are summarized

and expanded on in the sections below.

L_earned sorrg behavior may not be restricted to a sensitive period in female zebra ﬁnches
Male zebra ﬁnches learn song during a distinct period in development in order to
later produce it in adulthood to attract a female (Immelmann, 1969). Females do not sing,
but are selective in their responses to particular song stimuli, suggesting that they, too,
may learn song as males do (see Riebel, 2003). As detailed in Chapter 2, females
isolated from song but not biparental care during development showed normal responses
to the playback of conspeciﬁc songs in adulthood but their responding to the environment
in which conspeciﬁc (but not heterospeciﬁc) songs were heard 24 hours earlier was
attenuated. However, as the number of song exposures increased, the responses of song
isolated females increased to the levels of birds raised normally. These results suggest
that (1) female zebra ﬁnches may differentially form representations of environments in
which relevant songs are heard depending on prior exposure to those songs and (2) that
experience with song in adulthood can overcome deﬁcits in responding to song due to

developmental isolation from it, similar to the ability of male zebra ﬁnches to learn song

159

outside of the traditionally-deﬁned sensitive period (Chapter 2; Clayton, 1987; Eales,
1985, 1987a; Jones et al., 1996).

The interesting points that result from this behavioral study are the apparent
ability of the female zebra ﬁnch brain to remain plastic into adulthood and the speciﬁc
neural mechanisms that may be involved in the formation of memory for conspeciﬁc
songs or, perhaps, the ability of a bird to associate salient vocal communication with a
particular environment. The brains of males and females show distinct neuronal
responses to song that can be measured by the activity of immediate early genes (see
Chapters 1, 3 and 4). Through this type of analysis, I have shown that the hippocampus
responds to conspeciﬁc song stimuli more so than other types of auditory signals,
suggesting a role in the consolidation or modulation of behaviors necessary for the
generation of proper responses to song. If, as hypothesized, conspeciﬁc song-induced
activation in the hippocampus is important for the consolidation of a “contextual song
memory (Chapter 1),” then signiﬁcantly less immediate early gene expression would be
expected in the hippocampus of a female isolated during development (as in this study)
and exposed to playback of normal conspeciﬁc song. Although females in the
experimental group were not exposed to normal vocalizations from their fathers during
development, these males still produced sounds associated with respiration. It is possible
that song isolated females formed memories of their fathers’ non-song vocalizations
during development. If this is correct, then song isolated females should at least initially
prefer this auditory stimulus in a test of song preference in adulthood. Additional insight
into the function of cells in the hippocampus could also be obtained: immediate early

gene induction in the hippocampus of song isolated females following playback of

160

vocalizations from axotomized fathers should be as robust as the responses of neurons in

the hippocampus of birds raised normally and presented with normal zebra ﬁnch song.

Immediate eagly gene responses of neurons in auditory ggions of male and female zebra

 

ﬁnches change over development as scgdearning progresses

The bulk of the published data regarding this immediate early gene expression in
neurons in response to song have used adult birds. In adult male and female zebra
ﬁnches, song playback induces immediate early gene expression in NCM (Bailey et al.,
2002; Mello et al., 1992), CMM (Bailey etal., 2002; Bolhuis et al., 2000) and the
hippocampus (Bailey etal., 2002; Bolhuis et al., 2000; Kimpo & Doupe, 1997; Kruse et
al., 2004). The data I collected, along with that from others, have established that
neuronal responses to song at two periods in development, when song templates begin to
be acquired (d30) and when the sensorimotor integration period in males begins (d45),
are differential between males and females (at d30) and over time equalize (between d30
and d45). At d30, neuronal responses to song in female zebra ﬁnches are indicated via
FOS expressing neurons, and those in males via ZENK expressing cells. This differential
activation indicates the value of investigating the responses of both FOS and ZENK, and
future studies that examine auditory responses in songbirds would beneﬁt from
employing antibodies to the protein products of both genes, especially studies involving
both males and females. Also in d30 females, the hippocampus is the only region in
which conspeciﬁc song signiﬁcantly increases immediate early gene activity compared to
other auditory stimuli (Chapter 3). At d45, responses in the two sexes equalize, and the

NCM is the only structure in which signiﬁcant effects of auditory stimuli are uncovered

161

(Chapter 4). This suggests that the hippocampus may be important for song learning in
females during what could be their sensitive period (between d25-35; see Chapter Two),
and by d45, the NCM is the primary region of the brain in both sexes that responds to
conspeciﬁc song stimuli. Although many nuclei in the zebra ﬁnch brain are sexually
dimorphic (N ottebohm & Arnold, 1976), in adulthood the gross morphological structure
as well as the function of the auditory regions NCM and CMM appear equivalent
between the sexes. Why cells of these regions initially respond to song via divergent
neural mechanisms and how and why these cellular responses equalize over development

remain unknown.

Potential functional signiﬁcance of immediate early gene activity following song
presentations

The great interest that has surrounded the examination of the expression of
immediate early genes in the brain has been their potential involvement in synaptic
plasticity that could result in long-term memory formation. Indeed, this hypothesis has
been reproduced in countless studies across diverse species in which immediate early
gene activation correlates with acquisition of some particular behavior, suggesting
involvement in learning (see Tischmeyer & Grimm, 1999).

Immediate early gene induction triggers a cascade of post-transcriptional activity
that affects signal transduction, neuronal excitability, synaptic function or cytoskeletal
structure (Clayton, 2000; Mello, 2002). While induction of these immediate early genes
is only one of several regulatory events that take place following neuronal activation, it is

tempting to speculate in zebra ﬁnches what this activation may modify. For example,

162

additional data I have collected point to one such mechanism that may promote long-terrn
memory formation related to song. Area X of the medial striatum (MSt) is present in
males and integral to song learning (see Chapter 1) but is not identiﬁable via a Nissl stain
in females. In d30 and d45 birds, ZENK is not induced in Area X in males following
presentations of conspeciﬁc song but is activated in the surrounding MSt. ZENK
expression is homogenous throughout MSt in females, including the region that
corresponds to Area X (Figure 7.1). Neurochemical data from other studies show a u‘
similar pattern of expression. Neurons immunoreactive for an antibody to the NMDA
receptor subunit NR1 are not detected within Area X in adult male zebra ﬁnches but are
found in the surrounding striatum; interestingly, these NR1-immunoreactive cells are
found throughout the striatum in females (Saldanha et al., 2004). These patterns of
expression suggest a potential interaction between ZENK and NMDA receptors in song
development. For example, it is possible that ZENK induced by hearing conspeciﬁc song
inﬂuences the expression of NR1 in the MSt excluding Area X in males but throughout
the region in females. ZENK is integral in the modulation of neuronal excitability
(reviewed in Knapska & Kaczmarek, 2004) and expression of zif-268 (of which ZENK is
an avian homolog) in rodents appears to play a role in maintaining plastic changes
associated with long-term potentiation (LTP) (Knapska & Kaczmarek, 2004).
Glutamatergic receptors are important in LTP as well (Nordeen & Nordeen, 2004), and
activation of them during song template acquisition is vital to song learning and perhaps
even the opening of the sensitive period for it (Aamodt, Nordeen, & Nordeen, 1996;
Basham, Nordeen, & Nordeen, 1996; Heinrich, Singh, Sohrabji, Nordeen, & Nordeen,

2002; Nordeen & Nordeen, 2004; Scott, Singh, Nordeen, & Nordeen, 2004; Singh,

163

Basham, Nordeen, & Nordeen, 2000). The next step in understanding the function of
immediate early genes, speciﬁcally how they relate to song learning and perception, is an
analysis of what they are co-expressed with at various developmental stages following
presentations of conspeciﬁc song. To start, analysis of the co-expression of ZENK and
NMDA receptors under conditions of song exposure can lead to more of an
understanding of their roles in inﬂuencing song learning and behavior.

In addition to work needed to further evaluate the cascade of cellular events
triggered or modiﬁed by immediate early gene induction, future studies should correlate
immediate early gene expression with a bird’s observable preference for a stimulus. In
my work with adult females (Bailey et al., 2002) as well as d30 (Bailey and Wade, 2003)
and d45 (Bailey and Wade, 2005) males and females, auditory stimuli were presented in
the dark to minimize immediate early gene induction associated with visual stimuli and to
replicate the conditions of prior work (Bolhuis et al., 2000) in zebra ﬁnches. Because of
this, only auditory recordings were made while birds underwent the acclimation, song
exposure and post-song exposure periods. Hardly any vocalizations were made by birds,
and although movement was occasionally heard, it was of course impossible to know in
detail how the birds responded to the song stimuli. Recent work in female white-
crowned sparrows quantiﬁed both the ZENK response and behavioral responses to song
of hatch or local dialect (which is preferred) versus foreign dialect. Similar to results
obtained with zebra ﬁnches, ZENK induction in the NCM and CMM was greater in the
group exposed to the preferred song of local dialect (Maney et al., 2003). Additionally,
ZENK immunoreactivity in NCM was correlated with a non-vocal copulation solicitation

display, “wing quivers,” and induction in CMM with a vocal copulation solicitation

164

Figure 7.1. Drawings of coronal sections through the zebra ﬁnch brain detailing the
structures visible via Nissl stain (panels A and B). In males (A), Area X is evident as
a large nucleus distinct from the surrounding medial striatum (MSt). Area X is not
detectable in the female MSt (B). The remaining panels are photomicrographs of
coronal sections through the MSt of juvenile (d30) zebra ﬁnches detailing ZENK and
FOS immunoreactivity following conspeciﬁc song presentations. Note the almost
complete absence of ZENK immunoreactivity in Area X in males (panel C, dotted box
on left) compared to the relatively homogeneous immunoreactivity throughout the
MSt in females (dotted boxes in panel D). Little to no FOS immunoreactive neurons
were detected in both the male (E) or female (F) MSt. The midline is at the right in
each section. HA = hyperpallium apicale, HD = hyperpallium densocellulare, lMAN
= lateral portion of the magnocellular nucleus of the anterior nidopallium, X = Area X
of the medial striatum (MSt), mMST = medial portion of the medial striatum, lMSt =
lateral portion of the medial striatum. Scale bar = 1.0 mm.

165

FEMALE

MALE

 

 

HA

lMAN \®

 

 

HA

lMAN

 

 

166

display, “trills.” The behavioral correlate to immediate early gene induction is necessary
to further deduce the function of ZENK and FOS, for example, in responses or future
responses to song, and how their activation within the auditory areas NCM and CMM

inﬂuences the motor regions necessary for production of observable preference.

The hippocampus and NCM in zebra ﬁnches mgay be involved in modulating responses to
the environmental or social contexts of song

Immediate early gene expression is observed in several regions of the avian brain
following context-dependent song production, perception or spatial memory performance.
For example, in homing pigeons that were transported to a familiar training site and
released, ZENK immunoreactivity was four times greater in the lateral portion of the
hippocampus than that in birds transported to the familiar site and that did not home
(Shimizu, Bowers, Budzynski, Kahn, & Bingman, 2004). In addition, two times more
ZENK expressing cells were found in the medial portion of the MSt (mMSt) in birds that
navigated back to the home loft compared to birds that did not. The authors suggest that
the mMSt may participate in route like spatial learning that operates in parallel with more
maplike spatial representations that are consolidated by, stored in or mediated by the
hippocampal formation (also see Gagliardo et al., 1999). Given that immediate early
gene expression in the MSt and hippocampus are highest in birds exposed to conspeciﬁc
song and that connections exist between the two regions (Chapter 5), they may work
together in the consolidation or retrieval of spatial or relational memories, such as the

“contextual song memory” hypothesized above.

167

 

 

An additional experiment I designed further examines the contextual activation
within auditory regions by determining whether the activity-regulated cytoskeletal-
associated protein (ARC), an immediate early gene modulated by synaptic activity
(Guzowski, Lyford, Stevenson, Houston, McGaugh, Worley, & Barnes, 2000; Guzowski,
Setlow, Wagner, & McGaugh, 2001; Lyford, Yamagata, Kaufmann, Barnes, Sanders,
Copeland, Gilbert, Jenkins, Lanahan, & Worley, 1995), is differentially expressed in
NCM following exposure to particular auditory and/or non-auditory stimuli. In this
experiment, female zebra ﬁnches were moved from group housing to individual cages in
a ‘home’ environment, and after an acclimation period of one day were presented with
song or no auditory stimulus in either that room or a new one. Some of the birds were
euthanized, and others were returned to their home facility until the following day, at
which time they were either re-exposed to the same or a novel combination of visual and
auditory stimuli. In both cases, the density of somatic and dendritic ARC
immunoreactivity was determined (Figure 7.2). Increased ARC immunoreactivity in
NCM was observed only in females exposed to song following a change in environment;
this effect existed under both single and double exposure conditions (Bailey, Beck, Svec,
& Wade, 2005). Surprisingly, levels of ARC in each subdivision of the hippocampus
(dorsolateral, dorsomedial and ventral) were not affected by song exposure or change in
environment. These data suggest a role for ARC in the ability of NCM neurons to
associate environmental and conspeciﬁc song stimuli, and parallel data from others
(Kruse et al., 2004) that suggest non-auditory stimuli (such as the change of environment
in the present study) inﬂuence levels of ZENK expression within the NCM. Work is

ongoing to determine the levels of ARC expression in CMM and PVN (given the

168

 

Figure 7.2. Expression of the protein product of the immediate early
gene ARC in NCM (A) and the dorsomedial hippocampus (B).
Labeled protein was observed in nuclei (black arrows) and dendrites
(white arrows). Scale bar = 15 pm (A) and 50 um (B).

169

reciprocal connections between CMM and NCM (Chapter 1) and between the PVN and
hippocampus (Chapter 5)) to more fully understand the changes that occur in regions
important for song behavior in females.

The expression of immediate early genes can also depend on a bird’s immediate
environment, speciﬁcally the presence or absence of other birds. Adult songbirds show

little to no immediate early gene expression in Area X following song exposure (Jarvis &

 

Nottebohm, 1997; Jarvis, Scharff, Grossman, Ramos, & Nottebohm, 1998; Jin & Clayton, ‘
1997) or when a male sings to a female. However, ZENK is induced in this region when

males sing in the presence of another male or alone (Jarvis etal., 1998). Additionally,

levels of ZENK immunoreactivity are greater in the male zebra ﬁnch NCM when female

calls are presented in the presence of other males than when presented in social isolation

(Vignal, Andru, & Mathevon, 2005), suggesting that social context modulates immediate

early gene activity within this auditory region.

The structure and function of the hippocampus in zebra ﬁnches are simLar to those in a
non-songbird and show homology with the mammalian hippocampus

Together with data detailing the efferent connectivity of subdivisions of the zebra
ﬁnch hippocampus (Székely & Krebs, 1996), data from Chapter 5 indicate similarity with
the intrahippocampal circuitry of the pigeon hippocampus, and both are reminiscent of
the trisynaptic pathway in the mammalian hippocampus (Figure 7.3). This suggests a
similar means by which the multisynaptic, feedforward excitation of the mammalian
hippocampus by stimulation of the entorhinal cortex (Yeckel & Berger, 1990) can also

occur in the avian hippocampus. Based on these anatomical similarities and the

170

 

dorsal

 

 
    
     
 

lateral
A’fascicle

 

 

 

 

 

 

VHP
I)’
ventral "'l
fascicle
'-.
A SL PVN
POA DMP (MSt)
(Tn)

 

 

 

Figure 7.3. Intrahippocampal and intertelencephalic connections of subdivisions of
the zebra ﬁnch hippocampus compiled from Székely and Krebs (1996; dashed
lines) and data I have collected (Chapter 5; solid lines), detailing potential
involvement with the song system. A known connection between CMM and NCM
is indicated by a dotted line; see Figure 1.1 (Chapter 1) for additional connections
among song control regions. A = arcopallium, CMM = caudomedial mesopallium,
DLHP = dorsolateral subdivision of the hippocampus, DMHP = dorsomedial
subdivision of the hippocampus, DMP = posterior portion of the dorsomedial
thalamic nucleus, HA = hyperpallium apicale, HD = hyperpallium densocellulare,
MSt = medial striatum, NCM = caudomedial nidopallium, SL = lateral septum, SPf
= Substance P reactive nucleus, POA = preoptic area, PVN = periventricular
nucleus, Tn = nucleus taeniae, V = lateral ventricle, VHP = ventral subdivision of
the hippocampus. Parentheses enclose structures in which retrogradely labeled cells
were observed in only one bird.

171

functional similarities detailed in Chapter 6, the hippocampus in zebra ﬁnches appears to

be a functional homologue of that in mammals.

Zebra ﬁncheaam an ideal model system for the further examipation of physiological
processes underlying_speciﬁc forms of learning and memog

Given the song-speciﬁc induction of immediate early genes in the hippocampus
(Chapters 3 and 4), its connections with regions involved in song behavior, motivation
and arousal (Chapter 5), and the effects on performance in a spatial memory task
following lesions of a subdivision of the region (Chapter 6), the hippocampus in male and
female zebra ﬁnches is an excellent model with which to study the physiological
underpinnings of learned behavior. Along with NCM, the hippocampus of adult male
and female zebra ﬁnches is high in aromatase (Saldanha, Popper, Micevych, & Schlinger,
1998), the enzyme responsible for estrogen synthesis, and estrogen is important for
various forms of learning and increased cognitive function (McEwen & Alves, 1999;
Oberlander et al., 2004; Rissman, Heck, Leonard, Shupnik, & Gustafsson, 2002).
Aromatase levels remain high in the hippocampus from just before the critical period for
song learning through adulthood (Jacobs, Arnold, & Campagnoni, 1999). Estrogen is
important in song learning in juvenile male zebra ﬁnches: castration and systemic
treatment with the estrogen antagonist tamoxifen results in the production of abnormal
song (Bottjer & Hewer, 1992). It is reasonable to hypothesize that hippocampal estrogen
speciﬁcally is involved. Estrogen receptors are present in the avian hippocampus

(Metzdorf, Gahr, & Fusani, 1999), so the hormone could act locally, or a high rate of

172

 

 

hippocampal estrogen production could allow for diffusion to other telencephalic regions
important for song production or perception.

Thus, perhaps lesions of only the dorsolateral subdivision (Chapter 6) may not
have been enough to disrupt normal song behavior because aromatase availability (and
thus estrogen production), for one, in other regions was enough to compensate for the
loss of only 10—20% of the region. In addition, aromatase mRNA and protein are
increased following damage to the brain in zebra ﬁnches (Peterson, Saldanha, &
Schlinger, 2001), which could also compensate. Interestingly, aromatase and NMDA
receptors are coexpressed in neurons of the hippocampus (Saldanha et al., 2004).
Estrogen treatment increases the innervation of NMDA receptor-positive neurons and the
density of presynaptic NMDA autoreceptors. As indicated above, glutamatergic
receptors are integral to LTP (Nordeen & Nordeen, 2004) and song learning (Aamodt et
al., 1996; Basham et al., 1996; Heinrich et al., 2002; Nordeen & Nordeen, 2004; Scott et
al., 2004; Singh et al., 2000). So, if estrogen synthesis in the hippocampus is important
for the function of the song system in either sex, then perhaps its inhibition will produce
deﬁcits. A study like this would provide additional insight into the function of the
hippocampus in song behavior speciﬁcally and, more generally, a potential physiological

dissociation of the nature of memory encoding by the hippocampus.

General Summgy
Studies in the zebra ﬁnch have focused primarily on the development of and
neural mechanisms involved in male song learning and production and, in association, the

sexual differentiation of brain and behavior. The studies of this dissertation have

173

 

 

furthered our knowledge regarding the development of neural responses to song in both
males and females, the regions and mechanisms involved, and have begun to elucidate
additional neuronal circuits and structures in the sexes that may modulate song memory

formation and expression.

174

 

In

LITERATURE CITED

Aamodt SM, Nordeen EJ, Nordeen KW (1996). Blockade of NMDA receptors during
song model exposure impairs song development in juvenile zebra ﬁnches. Neurobiology

of Learning and Memogg 65:91-98.

 

Absil P, Braquenier J B, Balthazart J, Ball GF (2002). Effects of lesions of nucleus taeniae

on appetitive and consummatory aspects of male sexual behavior in Japanese quail. Brain,
Behavior and Evolution 60: 13-35.

Adret P (1993). Operant conditioning, song learning and imprinting to taped song in the
zebra ﬁnch. Animal Behaviour 46: 149-159.

Amaral DG, Witter MP (1989). The three-dimensional organization of the hippocampal
formation: A review of anatomical data. Neuroscience 31 :571-591.

Arnold A (1997). Sexual differentiation of the zebra ﬁnch song system: positive evidence,
negative evidence, null hypotheses, and a paradigm shift. Journal of Neurobiology
33:572-584.

Astikainen P, Ruusuvirta T, Korhonen T (2005). Longer storage of auditory than of
visual information in the rabbit brain: evidence from dorsal hippocampal
electrophysiology. Experimental Brain Reseaﬂh 160: 189-193.

Atoji Y, Wild J M, Yamamoto Y, Suzuki Y (2002). Intratelencephalic connections of the

hippocampus in pigeons (Columba Iivia). Journal of Comparative Neurology 447:177-
199.

Bailey DJ, Beck LA, Svec LA, Wade J (2005). Increased expression of activity-regulated
cytoskeletal-associated protein in the caudomedial nidopallium of adult female zebra
ﬁnches following song exposure coupled with a change in environmental context. Society
for Neuroscience 31.

Bailey DJ, Rosebush J C, Wade J (2002). The hippocampus and caudomedial neostriatum
show selective responsiveness to conspeciﬁc song in the female zebra ﬁnch. Journal of
Neurobiology 52:43-51.

Bailey DJ, Wade J (2003). Differential expression of the immediate early genes FOS and
ZENK following auditory stimulation in the juvenile male and female zebra ﬁnch.
Molecular Brain Resea_rc_h 116:147-154.

Bailey DJ, Wade J (2005). F OS and ZENK responses in 45 day-old zebra ﬁnches vary

with auditory stimulus and brain region, but not sex. Behavioural min Resea_rg
162:108-115.

175

Basham ME, Nordeen EJ, Nordeen KW (1996). Blockade of NMDA receptors in the
anterior forebrain impairs sensory acquisition in the zebra ﬁnch (Poephila guttata).
Neurobiology of Learning and Memory 66:295-304.

Bentley GE, Wingﬁeld J C, Morton ML, Ball GF (2000). Stimulatory effects on the
reproductive axis in female songbirds by conspeciﬁc and heterospeciﬁc male song.
Hormones and Behavior 37:179-189.

Best P], White AM (1998). Hippocampal cellular activity: A brief history of space.
Proceedings of the National Academy of Science, USA 95:2717-2719.

Bingman VP, Bagnoli P, Ioalé P, Casini G (1984). Homing behavior of pigeons aﬁer
telencephalic ablations. B_rain Behaﬁor and Evolution 24294-108.

Bingman VP, Ioalé P, Casini G, Bagnoli P (1988). Hippocampal ablated homing pigeons
show a persistent impairment in the time taken to return home. Journal of CompaLative
Physiology A 1632559-563.

Bohner J (1983). Song learning in the zebra ﬁnch (T aenopygia guttata): selectivity in the
choice of a tutor and accuracy of song copies. Animal Behaviour 31:231-237.

Bolhuis JJ, Hetebrij E, den Boer-Visser AM, De Groot J H, Zijlstra GGO (2001).
Localized immediate early gene expression related to the strength of song learning in
socially reared zebra ﬁnches. European Journal of Neuroscience 13:2165-2170.

Bolhuis JJ, Macphail EM (2001). A critique of the neuroecology of learning and memory.
Trends in Cognitive Sciences 5:426-433.

Bolhuis JJ, Van Mil DP, Houx BB (1999). Song learning with audiovisual compound
stimuli in zebra ﬁnches. Animal Behaviour 58: 1285-1292.

Bolhuis JJ, Zijlstra GGO, den Boer-Visser AM, Van der Zee EA (2000). Localized
neuronal activation in the zebra ﬁnch brain is related to the strength of song learning.
Proceedings of the National Academy of Science, USA 97:2282-2285.

Bottjer S, Hewer S (1992). Castration and antisteroid treatment impair vocal learning in
male zebra ﬁnches. Joumal of Neurobiology 23:337-353.

Bottjer SW, Brady JD, Cribbs B (2000). Connections of a motor cortical region in zebra
ﬁnches: Relation to pathways for vocal learning. Journal of Compamve Neurology
420:244-260.

Bottjer SW, Miesner EA, Arnold AP (1984). F orebrain lesions disrupt development but
not maintenance of song in passerine birds. Science 224:901-903.

176

Braaten RF, Reynolds K (1999). Auditory preference for conspeciﬁc song in isolation-
reared zebra ﬁnches. Animal Behaviour 58:105-111.

Brainard MS, Doupe AJ (2000). Auditory feedback in learning and maintenance of vocal
behaviour. Nature Reviews Neuroscience 1:31-40.

Brenowitz EA (1991). Altered perception of species-speciﬁc song by female birds after
lesions of a forebrain nucleus. Science 251:303-305.

 

Capaldi EA, Robinson GE, Fahrbach SE (1999). Neuroethology of spatial learning: The
birds and the bees. Annual Review of Psychology 50:651-682.

Casini G, Bingman VP, Bagnoli P (1986). Connections of the pigeon dorsomedial
forebrain studied with WGA-HRP and 3H-proline. Journal of Comparative Neurology
245:454-470.

Cheng H-Y, Clayton DF (2004). Activation and habituation of extracellular signal-
regulated kinase phosophorylation in zebra ﬁnch auditory forebrain during song
presentation. Journal of Neuroscience 24:7503-7513.

Chew SJ, Mello CV, Nottebohm F, Jarvis ED, Vicario DS (1995). Decrements in
auditory responses to a repeated conspeciﬁc song are long-lasting and require two periods

of protein synthesis in the songbird forebrain. Proceedings of the National Academyof
Science, USA 92:3406-3410.

Chew SJ, Vicario DS, Nottebohm F (1996a). A large-capacity memory system that
recognizes the calls and songs of individual birds. Proceedings of the National Academy
of Science. USA 93: 1950-1955.

Chew SJ, Vicario DS, Nottebohm F (1996b). Quantal duration of auditory memories.
Science 274: 1909-1914.

Chun MM, Phelps EA (1999). Memory deﬁcits for implicit contextual information in
amnesic subjects with hippocampal damage. Nature Neuroscience 2:844-847.

Clark RE, Squire LR (1998). Classical conditioning and brain systems: The role of
awareness. Science 280277-81.

 

Clayton DF (1997). Role of gene regulation in song circuit development and song
learning. J oumal of Neurobiology 33:549-571.

Clayton DF (2000). The genomic action potential. Neurobiolggy of Learning and
Memog 74:185-216.

Clayton NS (1987). Song tutor choice in zebra ﬁnches. Animal Behaviour 35:714-721.

177

Clayton NS (1988). Song discrimination learning in zebra ﬁnches. Animal Behaviour
36: 1016-1024.

Clayton NS (1990). Subspecies recognition and song learning in zebra ﬁnches. Animal
Behaviour 40: 1009- 101 7.

Clayton NS, Prove E (1989). Song discrimination in female zebra ﬁnches and bengalese
ﬁnches. Animal Behaviour 38:352-354.

Coleman MJ, Vu ET (2005). Recovery of impaired songs following unilateral but not
bilateral lesions of nucleus uvaeformis of adult zebra ﬁnches. J ourgal of Neurobiology
63:70-89.

Colombo M, Broadbent N (2000). Is the avian hippocampus a functional homologue of
the mammalian hippocampus? Neuroscience and Biobehavioral Reviews 24:465-484.

Cynx J (1993). Conspeciﬁc song perception in zebra ﬁnches (T aeniopygia guttata).
Journal of Comparative Psychology 107:395-402.

Cynx J, Nottebohm F (1992). Role of gender, season, and familiarity in discrimination of
conspeciﬁc song by zebra ﬁnches (T aeniopygia guttata). Proceedings of the National
A_cademy of Science. USA 89: 1368-1371.

Del Negro C, Gahr M, Leboucher G, Kreutzer M (1998). The selectivity of sexual
responses to song displays: effects of partial chemical lesion of the HVC in female
canaries. Behavioural Mn Research 96: 151-159.

D'Hondt E, Vermeiren J, Peeters K, Balthazart J, Tlemcani 0, Ball GF, Duffy DL,
Vandesande F, Berghman LR (1999). Validation of a new antiserum directed towards the
synthetic c-terminus of the F OS protein in avian species: immunological, physiological
and behavioral evidence. Joumal of Neuroscience Methods 91 :3 1-45.

Doupe AJ (1997). Song- and order-selective neurons in the songbird anterior forebrain
and their emergence during vocal development. Journal of Neuroscience 17:1147-1167.

Duffy DL, Bentley GE, Ball GF (1999). Does sex or photoperiodic condition inﬂuence
ZENK induction in response to song in European starlings? Brain Research 844:78-82.

Eales LA (1985). Song learning in zebra ﬁnches: some effects of song model availability
on what is learnt and when. Animal Behaviour 33: 1293-1300.

Eales LA (1987a). Do zebra ﬁnch males that have been raised by another species tend to
select a conspeciﬁc song tutor? Animal Behaviour 35: 1347-1355.

Eales LA (1987b). Song learning in female-raised zebra ﬁnches: another look at the
sensitive phase. Animal Behaviour 35: 1356-1365.

178

Ebbinghaus H. (1964). Memory: A Contribution to Experimental Psychology (H Ruger, C
Bussenius, Trans). New York: Dover Publications, Inc.

Eda-Fujiwara H, Satoh R, Bolhuis JJ, Kimura T (2003). Neuronal activation in female
budgerigars is localized and related to male song complexity. European J oumgpf
Neuroscience 17:149-154.

 

Eichenbaum H (2000). A cortical-hippocampal system for declarative memory. Nature
Reviews Neuroscience 1 :41-50.

Eichenbaum H, Dudchenko P, Wood E, Shapiro M, Tanila H (1999). The hippocampus,
memory, and place cells: Is it spatial memory or a memory space? Neuron 23:209-226.

Eichenbaum H, Schoenbaum G, Young B, Bunsey M (1996). Functional organization of
the hippocampal memory system. Proceedings of the National Academy of Science. USA
93:13500-13507.

Erichsen JT, Bingman VP, Krebs JR (1991). The distribution of neuropeptides in the
dorsomedial telencephalon of the pigeon (Columba Iivia): A basis for regional
subdivisions. Journal of Compmtive Neurology 314:478-492.

Foster EF, Bottjer SW (1998). Axonal connections of the high vocal center and
surrounding cortical regions in juvenile and adult male zebra ﬁnches. Journal of
Compar__a£ive Neurology 397: 1 18-138.

Foster TC, Hampson RE, West MO, Deadwyler SA (1988). Control of sensory activation
of granule cells in the fascia dentata by extrinsic afferents: septal and entorhinal inputs.
Journial of Neuroscience 8:3869-3878.

Funabiki Y, Konishi M (2003). Long memory in song learning by zebra ﬁnches. Journal
of Neuroscience 23:6928-6935.

Gagliardo A, Ioalé P, Bingman VP (1999). Homing in pigeons: The role of the
hippocampal formation in the representation of landmarks used for navigation. Journal of
Neuroscience 19:311-315.

Gahr M, Gtittinger HR, Kroodsma DE (1993). Estrogen receptors in the avian brain:
survey reveals general distribution and forebrain areas unique to songbirds. Journal of
Comparative Neurology 327: 1 12-122.

Gentner T, Hulse S, Duffy D, Ball G (2001a). Response biases in auditory forebrain

regions of female songbirds following exposure to sexually relevant variation in male
song. Joumi of Neurobiology 46:48-58.

179

 

Gentner TQ, Hulse SH, Duffy D, Ball GF (2001b). Response biases in auditory forebrain
regions of female songbirds following exposure to sexually relevant variation in male
song. J ournal of Neurobiology 46:48-58.

Guzowski JF, Lyford GL, Stevenson GD, Houston FP, McGaugh JL, Worley PF, Barnes
CA (2000). Inhibition of activity-dependent Arc protein expression in the rat
hippocampus impairs the maintenance of long-term potentiation and the consolidation of
long-term memory. Journal of Neuroscience 20:3993-4001.

Guzowski JF, Setlow B, Wagner EK, McGaugh JL (2001). Experience-dependent gene
expression in the rat hippocampus after spatial learning: A comparison of the immediate-
early genes Arc, c-fos, and zif268. Journal of Neuroscience 21 :5089-5098.

Heinrich J E, Singh TD, Sohrabji F, Nordeen KW, Nordeen EJ (2002). Developmental
and hormonal regulation of NR2A mRN A in forebrain regions controlling avian vocal
learning. Journal of Neurobiology 51: 149-159.

 

Herdegen T, Leah JD (1998). Inducible and constitutive transcription factors in the
mammalian nervous system: control of gene expression by Jun, F 03 and Krox, and
CREB/ATF proteins. min Research Reviews 28:370-490.

Hernandez AM, MacDougall-Shackleton SA (2004). Effects of early song experience on
song preferences and song control and auditory brain regions in female house ﬁnches
(Carpodacus mexicanus). Journal of Neurobiology 59:247-258.

Hough 11 GE, Pang KCH, Bingman VP (2002). Intrahippocampal connections in the .
pigeon (Columba livia) as revealed by stimulation evoked ﬁeld potentials. Journal of
Compﬂrﬁve Neurology 452:297-309.

1

Houx BB, ten Cate C (1999). Do stimulus-stimulus contingencies affect song learning in
zebra ﬁnches (T aeniopygia guttata)? Journal of Comparative Psychology 113:235-242.

Immelmann K. (1969). Song development in the zebra ﬁnch and other estrildid ﬁnches.
In R Hinde (Ed.), Bird Vocalizations (pp. 61-74). Cambridge: University Press.

Jacobs EC, Arnold AP, Campagnoni AT (1999). Developmental regulation of the
distribution of aromatase- and estrogen-receptor-mRNA-expressing cells in the zebra
ﬁnch brain. Developmental Neuroscience 21 :453-472.

Jarvis ED, Mello CV (2000). Molecular mapping of brain areas involved in parrot vocal
communication. J ougal of Compa_gtive Neurology 419: 1-3 1.

Jarvis ED, Nottebohm F (1997). Motor-driven gene expression. Proceedings of the
National Academy of Science, USA 94:4097-4102.

180

Jarvis ED, Ribeiro S, da Silva ML, Ventura D, Vielliard J, Mello CV (2000).
Behaviourally driven gene expression reveals song nuclei in hummingbird brain. Nature
406:628-632.

Jarvis ED, Scharff C, Grossman MR, Ramos JA, Nottebohm F (1998). For whom the bird
sings: context-dependent gene expression. Neuron 21:775-788.

 

J in H, Clayton DF (1997). Localized changes in immediate-early gene regulation during
sensory and motor learning in zebra ﬁnches. Neuron 19:1049-1059.

 

Johnson F, Soderstrom K, Whitney O (2002). Quantifying song bout production during
zebra ﬁnch sensory-motor learning suggests a sensitive period for vocal practice.
Behavioural Brain Researal; 131:57-65.

Jonasson Z (2005). Meta-analysis of sex differences in rodent models of learning and
memory: a review of behavioral and biological data. Neuroscience and Biobehavioral
Reviews 28:811-825.

Jones AE, ten Cate C, Bijleveld CJH (2001). The interobserver reliability of scoring
sonagrams by eye: a study on methods, illustrated on zebra ﬁnch songs. Animal
Behaviour 62:791-801.

Jones AE, ten Cate C, Slater P] B (1996). Early experience and plasticity of song in adult
male zebra ﬁnches (T aeniopygia guttata). Journal of Comparative Psychology 110:354-
369.

Kahn MC, Hough 11 GE, Ten Eyck GR, Bingman VP (2003). Internal connectivity of the
homing pigeon (Columba livia) hippocampal formation: An anterograde and retrograde
tracer study. Journal of Compaaative Neurology 459: 127-141.

Kakeue T, Kaminosono S, Kitamura Y, Ogawa H, Oka K. (2004). T he function of the
hippocampus in auditory perception revealed by in viva optical imaging in zebra ﬁnch
(T aeniopygia guttata). Paper presented at the Society for Neuroscience.

Kamil AC, Balda RP (1990). Spatial memory in seed-caching corvids. Psychology of
Learning and Motivation 26: 1-25.

Kim JJ, Fanselow MS (1992). Modality-speciﬁc retrograde amnesia of fear. Science
256:675-677.

Kimpo RR, Doupe AJ (1997). F OS is induced by singing in distinct neuronal populations
in a motor network. Neuron 18:315-325.

 

Kittelberger J M, Mooney R (1999). Lesions of an avian forebrain nucleus that disrupt
song development alter synaptic connectivity and transmission in the vocal premotor
pathway. Journal of Neuroscience 19:9385-9398.

181

 

 

Knapska E, Kaczmarek L (2004). A gene for neuronal plasticity in the mammalian brain:
Zif268/Egr-1/NGF1-A/Krox-24/T188/ZENK? Progress in Neurobiology 74:183-211.

Knight RT (1996). Contribution of human hippocampal region to novelty detection.
Nature 383:256-259.

Krayniak PF, Siege] A (1978). Efferent connections of the hippocampus and adjacent
regions in the pigeon. Erin Behavior and Evolution 15:372-388.

Krebs JR, Sherry DF, Healy SD, Perry VH, Vaccarino AL (1989). Hippocampal
specialization of food-storing birds. Proceedings of the Natiﬂl Academy of Science,
USA 86:1388-1392.

Kroodsma D, Byers B, Goodale E, Johnson S, Liu W-C (2001). Pseudoreplication in
playback experiments, revisited a decade later. Animal Behaviour 61 :1029-1033.

Kroodsma DE (1989). Suggested experimental designs for song playbacks. Animal
Behaviour 37:600-609.

Kruse AA, Stripling R, Clayton DF (2000). Minimal experience required for immediate-

early gene induction in zebra ﬁnch neostriatum. Neurobiology of Learning and Memog
74:179-184.

Kruse AA, Stripling R, Clayton DF (2004). Context-speciﬁc habituation of the zenk gene
response to song in adult zebra ﬁnches. Neurobiology of Learning and Memopy 82:99-
108.

Kuhl P, Doupe A (1999). Birdsong and human speech: Common themes and mechanisms.
Annual Review of Neuroscience 22:567-631.

Lauay C, Gerlach NM, Adkins-Regan E, DeVoogd TJ (2004). Female zebra ﬁnches
require early song exposure to prefer high quality song as adults. Animal Behaviour
68:1249-1255.

Lauay C, Gerlach NM, Adkins-Regan E, DeVoogd TJ (in press). Female zebra ﬁnches
require early song exposure to prefer high quality song as adults. Animal Behaviour.

Leutgeb S, Leutgeb JK, Barnes CA, Moser E1, McNaughton BL, Moser M-B (2005).
Independent codes for spatial and episodic memory in hippocampal neuronal ensembles.
Science 309:619-623.

Lewicki MS, Konishi M (1995). Mechanisms underlying the sensitivity of songbird

forebrain neurons to temporal order. Proceedings of the National Aﬁdemy of Science,
USA 92:5582-5586.

182

Lipkind D, Nottebohm F, Rado R, Barnea A (2002). Social change affects the survival of
new neurons in the forebrain of adult songbirds. Behavioural Brain Resea_r_c_h_ 133:31-43.

Liu W-C, Gardner TJ, Nottebohm F (2004). Juvenile zebra ﬁnches can use multiple
strategies to learn the same song. Proceedings of the National Academy of Science, USA
101:18177-18182.

Lyford GL, Yamagata K, Kaufrnann WE, Barnes CA, Sanders LK, Copeland NG, Gilbert
DJ, Jenkins NA, Lanahan AA, Worley PF (1995). Arc, a grth factor and activity-
regulated gene, encodes a novel cytoskeleton-associated protein that is enriched in
neuronal dendrites. Neuron 14:433-445.

 

MacDougall-Shackleton SA, Hulse SH, Ball GF (1998). Neural bases of song preferences
in female zebra ﬁnches (T aeniopygia guttata). NeuroReport 923047-3052.

Macphail EM (2002). The role of the avian hippocampus in spatial memory. Psicologica
23:93-108.

Malakoff D (1999). OLFACTIONzFollowing the Scent of Avian Olfaction. Science
286:704-705.

Maney DL, MacDougall-Shackleton EA, MacDougall-Shackleton SA, Ball GF, Hahn TP
(2003). Immediate early gene response to hearing song correlates with receptive behavior

and depends on dialect in a female songbird. Journal of Comparative Physiology A
189:667-674.'

Mann NI, Slater PJ B (1994). What causes young male zebra ﬁnches, Taeniopygia guttata,
to choose their father as song tutor? Animal Behaviour 47:671-677.

Margoliash D, Fortune ES (1992). Temporal and harmonic combination-sensitive
neurons in the zebra ﬁnch's HVc. Journal of Neuroscience 12:4309-4326.

Marler P (1997). Three models of song learning: evidence from behavior. Journal of
Neurobiology 33:501-516.

Maxwell SE (1980). Pairwise multiple comparisons in repeated measures designs.
Journal of Educational Statistics 5:269-287.

McEwen BS (1999). Stress and hippocampal plasticity. Annual Review of Neuroscience
22:105-122.

McEwen BS, Alves SE (1999). Estrogen actions in the central nervous system. Endocrine
Reviews 20:279-307.

Mello CV (2002). Mapping vocal communication pathways in birds with inducible gene
expression. Journal of Compa_rative Physiology A 188:943-959.

183

 

Mello CV, Clayton DF (1994). Song-induced ZENK gene expression in auditory
pathways of songbird brain and its relation to the song control system. Journal of
Neuroscience 14:6652-6666.

Mello CV, Nottebohm F, Clayton D (1995). Repeated exposure to one song leads to a
rapid and persistent decline in an immediate early gene's response to that song in zebra
ﬁnch telencephalon. Journal of Neuroscience 15:6919-6925.

Mello CV, Ribeiro S (1998). ZENK protein regulation by song in the brain of songbirds.
Journal of Comparative Neurology 393:426-438.

Mello CV, Vates GE, Okuhata S, Nottebohm F (1998). Descending auditory pathways in
the adult male zebra ﬁnch (T aeniopygia guttata). Journal of Comparative Neurology
395: 137-160.

 

Mello CV, Vicario DS, Clayton DF (1992). Song presentation induces gene expression in
the songbird forebrain. Proceedigs of the National Academy of Science. USA 89:6818-
6822.

Metzdorf R, Gahr M, Fusani L (1999). Distribution of aromatase, estrogen receptor, and
androgen receptor mRN A in the forebrain of songbirds and nonsongbirds. Journal of
Comﬂtive Neurology 407:1 15-129.

Miller DB (1979a). The acoustic basis of mate recognition by female zebra ﬁnches
(T aeniopygia guttata). Animal Behaviour 27:376-3 80.

Miller DB (1979b). Long-term recognition of father's song by female zebra ﬁnches.
Nature 280:389-391.

Nagle L, Kreutzer M, Vallet E (2002). Adult female canaries respond to male song by
calling. Ethology 108:463-472.

Nagle L, Kreutzer ML (1997). Song tutoring inﬂuences female song preferences in
domesticated canaries. Behaviour 134289-104.

Nastiuk KL, Mello CV, George JM, Clayton DF (1994). Irnmediate-early gene responses
in the avian song control system: cloning and expression analysis of the canary c-jun
cDNA. Molecular BraiaReseapca 27:299-309.

Nealen PM, Schmidt MF (2002). Comparative approaches to avian song system function:

insights into auditory and motor processing. Journal of Compmtive Physiology A
188:929-941.

184

Neubauer RL (1999). Super-normal length song preferences of female zebra ﬁnches

(T aeniopygia guttata) and a theory of the evolution of bird song. Evolutionm Ecology
13:365-380.

Nordeen KW, Nordeen EJ (1997). Anatomical and synaptic substrates for avian song
learning. Journal of Neurobiology 33:532-548.

Nordeen KW, Nordeen EJ (2004). Synaptic and molecular mechanisms regulating
plasticity during early learning. Annals of the New York Academy of Sciences 1016:416-
437.

Nottebohm F (20023). Neuronal replacement in adult brain. grain Research Bulletin
57:737-749.

 

Nottebohm F (2002b). Why are some neurons replaced in adult brain? Journal of
Neuroscience 22:624-628.

Nottebohm F, Arnold A (1976). Sexual dimorphism in vocal control areas of the songbird
brain. Science 194:211-213.

 

Oberlander JG, Schlinger BA, Clayton NS, Saldanha CJ (2004). Neural aromatization
accelerates the acquisition of spatial memory via an inﬂuence on the songbird
hippocampus. Hormones and Behavior 45:250-25 8.

O'Keefe J, Dostrovsky J (1971). The hippocampus as a spatial map. Preliminary
evidence from unit activity in the freely-moving rat. Brain Research 34: 171-175.

O'Keefe J, Nadel L. (1978). The hippocampus as a cognitive map. London: Oxford
University Press.

Pace TWW, Gaylord R, Topczewski T, Girotti M, Rubin B, Spencer R (2005).
Immediate—early gene induction in hippocamp hippocampus us and cortex

as a result of novel experience is not directly related to the

stressfulness of that experience. European Journal of Neuroscience 22: 1679-1690.

Packard MG, McGaugh J L (1996). Inactivation of hippocampus or caudate nucleus with
lidocaine differentially affects expression of place and response learning. Neurobiology
of Learning and Memory 65:65-72.

Park KHJ, Clayton DF (2002). Inﬂuence of restraint and acute isolation on the selectivity

of the adult zebra ﬁnch zenk gene response to acoustic stimuli. Behavioural Brain
Research 136: 185-191.

Patel SN, Clayton NS, Krebs JR (1997a). Hippocampal tissue transplants reverse lesion-

induced spatial memory deﬁcits in zebra ﬁnches (T aeniopygia guttata). Journal of
Neuroscience 17:3861-3869.

185

 

 

Patel SN, Clayton NS, Krebs JR (1997b). Spatial learning induces neurogenesis in the
avian brain. Behavioural Brain Research 89:115-128.

Peterson RS, Saldanha CJ, Schlinger BA (2001). Rapid upregulation of aromatase mRNA
and protein following neural injury in the zebra ﬁnch (T aeniopygia guttata). Journal of
Neuroendocrinology 13:3 17-323.

Phillips RG, LeDoux JE (1995). Lesions of the fomix but not the entorhinal or perirhinal
cortex interfere with contextual fear conditioning. Journal of Neuroscience 15:5308-5315.

Phillmore LS, Bloomﬁeld LL, Weisman RG (2003). Effects of song and calls on ZENK
expression in the auditory telencephalon of ﬁeld- and isolate-reared black capped
chickadees. Behavioural Brain Research 147: 125-134.

Poole J H, Tyack PL, Stoeger-Horwath AS, Waitwood S (2005). Elephants are capable of
vocal learning. Nature 434:455-456.

Pytte CL, Suthers RA (2000). Sensitive period for sensorimotor integration during vocal
motor learning. Jorﬂal of Neurobiology 42: 172-189.

Rehkamper G, Hasse E, F rahm HD (1988). Allometric comparison of brain weight and
brain structure volumes in different breeds of the domestic pigeon, Columba livia f.d.
(fantalis, homing pigeons, strassers). Brain Behavior and Evolution 31: 141-149.

Reiner A, Perkel DJ, Bruce LL, Butler AB, Csillag A, Kuenzel W, Medina L, Paxinos G,
Shimizu T, Striedter G, Wild M, Ball GF, Durand S, Gﬁntﬁrlciln 0, Lee DW, Mello CV,
Powers A, White SA, Hough G, Kubikova L, Smulders TV, Wada K, Dugas-Ford J,
Husband S, Yamamoto K, Yu J, Siang C, Jarvis ED (2004). Revised nomenclature for
avian telencephalon and some related brainstem nuclei. Journal of Cormaa‘ative
Neurology 473:377-414.

Ribeiro S, Cecchi GA, Magnasco MO, Mello CV (1998). Toward a song code: Evidence
for a syllabic representation in the canary brain. Neuron 21:359-371.

Riebel K (2000). Early exposure leads to repeatable preferences for male song in female
zebra ﬁnches. Proceedings of the Royal Society of London B 267:2553-2558.

Riebel K (2003). Developmental inﬂuences on auditory perception in female zebra
ﬁnches - is there a sensitive phase for song preference learning? Animal Biology 53:73-
87.

Riebel K, Smallegange I, Terpstra N, Bolhuis J (2001). Sexual equality in zebra ﬁnch

song preference: evidence for a dissociation between song recognition and production
learning. Proceedings of the Royal Society of London B 269:729-733.

186

Riebel K, Smallegange IM (2003). Does zebra ﬁnch (T aeniopygia guttata) preference for
the (familiar) father's song generalize to the songs of unfamiliar brothers? Journal of
Comparative Psychology 1 17:61-66.

Riebel K, Smallegange IM, Terpstra NJ, Bolhuis JJ (2002). Sexual equality in zebra ﬁnch
song preference: evidence for a dissociation between song recognition and production
learning. Proceedings of the Royal Society of London B 269:729-733.

Rissman EF, Heck AL, Leonard JE, Shupnik MA, Gustafsson J-A (2002). Disruption of

estrogen receptor [3 gene impairs spatial learning in female mice. Proceediggs of the
ﬂitional Academy of Science, USA 99:3396-4001.

Riters LV, Alger SJ (2004). Neuroanatomical evidence for indirect connections between
the medial preoptic nucleus and the song control system: possible neural substrates for
sexually motivated song. CeLand Tissue Resear_ch 316:35-44.

Sadananda M, Bischof H-J (2002). Enhanced F os expression in the zebra ﬁnch
(T aenopygia guttata) brain following ﬁrst courtship. Journal of Compa_pative Neurolagy
448: 150-164.

Saldanha CJ, Clayton NS, Schlinger BA (1999). Androgen metabolism in the juvenile
oscine forebrain: A cross-species analysis at neural sites implicated in memory function.
Journal of Neurobiology 40:397-406.

Saldanha CJ, P0pper P, Micevych PE, Schlinger BA (1998). The passerine hippocampus
is a site of high aromatase: inter- and intraspecies comparisons. Hormones and Behavior
34:85-97.

Saldanha CJ, Schlinger BA, Micevych PE, Horvath TL (2004). Presynaptic N-methyl-D-
aspartate receptor expression is increased by estrogen in an aromatase-rich area of the
songbird hippocampus. Journal of Compa'ative Neurology 469:522-534.

Scharff C, Nottebohm F (1991). A comparative study of the behavioral deﬁcits following
lesions of various parts of the zebra ﬁnch song system: Implications for vocal learning.
J oumﬂf Neuroscience 1 1:2896-2913.

Scott LL, Singh TD, Nordeen EJ, Nordeen KW (2004). Developmental patterns of
NMDAR expression within the song system do not recur during adult vocal plasticity in
zebra ﬁnches. Journal of Neurobiology 58:442-454.

Scoville WB, Milner B (1957). Loss of recent memory after bilateral hippocampal lesions.
Journal of Neurology, Neurosurgery and Psychiatg 20:11-21.

Searcy WA. (1992). Measuring responses of female birds to male song. In P McGregor

(Ed.), Playback and Studies of Animal Communication (pp. 175-190). New York: Plenum
Press.

187

Sherry DF, Vaccarino AL (1989). Hippocampus and memory for food caches in black-
capped chickadees. Behavioral Neuroscience 1032308-318.

Shettleworth SJ (1990). Spatial memory in food-storing birds. Philosophical Transactions
of the Royal Society of London B 329: 143-151.

Shiﬂett MW, Gould KL, Smulders TV, DeVoogd TJ (2002). Septum volume and food-
storing behavior are related in parids. Journal of Neurobiology 51:215-222.

Shimizu T, Bowers AN, Budzynski CA, Kahn MC, Bingman VP (2004). What does a
pigeon (Columba livia) brain look like during homing? Selective examination of ZENK
expression. Behavioral Neuroscience 1 18:845-851.

Simpson HB, Vicario DS (1990). Brain pathways for learned and unleamed vocalizations
differ in zebra ﬁnches. Journal of Neuroscience 10: 1541-1556.

Singh TD, Basham ME, Nordeen EJ, Nordeen KW (2000). Early sensory and hormonal
experience modulate age-related changes in NR2B mRN A within a forebrain region
controlling avian vocal learning. Jorml of Neurobiology 44:82-94.

Slater PJB (2003). Fifty years of bird song research: a case study in animal behavior.
Animal Behaviour 65:633-639.

Sockman KW, Gentner TQ, Ball GF (2002). Recent experience modulates forebrain
gene-expression in response to mate-choice cues in European starlings. Proceedings of
the Royal Society of London B 269:2479-2485.

Solis MM, Brainard MS, Hessler NA, Doupe AJ (2000). Song selectivity and
sensorimotor signals in vocal learning and production. Proceedings of the National
Academy of Science. USA 97:11836-11842.

Solis MM, Doupe AJ (1999). Contributions of tutor and bird's own song experience to
neural selectivity in the songbird anterior forebrain. Journal of Neuroscience 1924559-
4584.

Squire LR (1992). Memory and the hippocampus: A synthesis from ﬁndings with rats,
monkeys, and humans. Psychological Review 99: 195-231.

Strasser R, Bingman VP (1996). The relative importance of location and feature cues for
homing pigeon (Columba livia) goal recognition. Journal of Comparative Psychology
110:77-87.

Striedter GF, Vu ET (1998). Bilateral feedback projections to the forebrain in the
premotor network for singing in zebra ﬁnches. Journal of Neurobiology 34:27-40.

188

Stripling R, Kruse AA, Clayton DF (2001). Development of song responses in the zebra
ﬁnch caudomedial neostriatum: Role of genomic and electrophysiological activities.
Journal of Neurobiology 48: 163-180.

Stripling R, Milewski L, Kruse AA, Clayton DF (2003). Rapidly learned song-
discrimination without behavioral reinforcement in adult male zebra ﬁnches
(Taeniopygia guttata). Neurobiology of Learning and Memory 79:41-50.

Stripling R, Volman SF, Clayton DF (1997). Response modulation in the zebra ﬁnch
neostriatum: Relationship to nuclear gene regulation. Journal of Neuroscience 17:3 883-
3893.

Sturdy CB, Phillmore LS, Sartor JJ, Weisman RG (2001). Reduced social contact causes

auditory perceptual deﬁcits in zebra ﬁnches, T aeniopygia guttata. Animal Behﬂiour
62: 1207-1218.

 

Székely AD (1999). The avian hippocampal formation: subdivisions and connectivity.
Behavioural Brain Resezaah 98:219-225.

Székely AD, Krebs JR (1996). Efferent connectivity of the hippocampal formation of the
zebra ﬁnch (T aenopygia guttata): An anterograde pathway tracing study using Phaseolus
vulgaris leucoagglutinin. Journal of Comparative Neurology 368: 198-214.

Tchemichovski O, Nottebohm F, Ho CE, Pesaran B, Mitra PP (2000). A procedure for an
automated measurement of song similarity. Animal Behaviour 59:1167-1176.

Tchemichovski O, Swigger D, Mitra PP. (2004). Sound Analysis Pro User Manual
(Version 1.04).

Terpstra NJ, Bolhuis JJ, den Boer-Visser AM (2004). An analysis of the neural
representation of birdsong memory. Journal of Neuroscience 24:4971-4977.

Tischmeyer W, Grimm R (1999). Activation of immediate early genes and memory
formation. Cellular and Molecular Life Sciences 55:564-574.

Tolman EC (1948). Cognitive maps in rats and men. Psychological Review 55:109-208.

Tomaszycki ML, Sluzas EM, Newman SW, Adkins-Regan E, DeVoogd TJ (2004).
Differential immediate early gene responses to intact and isolate song occurs in NCM by
45 days in zebra ﬁnches. Society for Neuroscience Abstracts 34.

van Praag H, Schnider AF, Christie BR, Toni N, Palmer TD, Gage PH (2002). Functional
neurogenesis in the adult hippocampus. Nature 415: 1030-1034.

 

Vates GE, Broome BM, Mello CV, Nottebohm F (1996). Auditory pathways of caudal
telencephalon and their relation to the song system of adult male zebra ﬁnches
(T aeniopygia guttata). Journal of Compzaative Neurology 366:613-642.

189

Vates GE, Nottebohm F (1995). Feedback circuitry within a song-learning pathway.
Proceedings of the National Academy of Science; USA 92:5139-5143.

Vates GE, Vicario DS, Nottebohm F (1997). Reafferent thalamo-"cortical" loops in the
song system of oscine songbirds. Journal of Comparative Neurology 380:275-290.

Vicario DS, Navqi NH, Raksin JN (2001). Sex differences in discrimination of vocal
communication signals in a songbird. Animaﬂehaviour 61:805-817.

Vignal C, Andru J, Mathevon N (2005). Social context modulates behavioural and brain
immediate early gene responses to sound in male songbird. European J oumw
Neuroscience 22:949-955.

Vignal C, Attia J, Mathevon N, Beauchaud M (2004). Background noise does not modify
song-induced genic activation in the bird brain. Behaviouﬁ Brain Researc_h_ 153:241-248.

Wade J (1999). Sexual dimorphisms in avian and reptilian courtship: Two systems that
do not play by mammalian rules. min. Behavior and Evolution 54:15-27.

Wade J (2001). Zebra ﬁnch sexual differentiation: the aromatization hypothesis revisited.
Microscopy Research and Technique 54:354-363.

Wade J, Buhlman L (2000). Lateralization and effects of adult androgen in a sexually
dimorphic neuromuscular system controlling song in zebra ﬁnches. Journal of
Comparative Neurology 426: 154-164.

Wasserrnan EA, Miller RR (1997). What's elementary about associative learning? Annual
Review of Psychology 48:573-607.

Watanabe S, Bischof H-J (2004). Effects of hippocampal lesions on acquisition and
retention of spatial learning in zebra ﬁnches. Behavioural Brain Research 155:147-152.

Whitney O, Soderstrom K, Johnson F (2003). CB1 cannabinoid receptor activation
inhibits a neural correlate of song recognition in an auditory/perceptual region of the
zebra ﬁnch telencephalon. Journal of Neurobiology 56:266-274.

Williams H (1990). Models for song learning in the zebra ﬁnch: fathers or others?
Animal Behaviour 39:745-757.

Yeckel MF, Berger TW (1990). Feedforward excitation of the hippocampus by afferents
from the entorhinal cortex: Redeﬁnition of the role of the trisynaptic pathway.
Proceedings of the National Academy of Science, USA 87:5832—5836.

Zann RA. (1996). The Zebra Finch: A Synthesis of Field and Laboratory Studies. London:
Oxford University Press.

190

MMMMMMMM

Illllllllljlljllll