Center for Neuroscience and
Department of Anatomy and Neurobiology
University of Tennessee
855 Monroe Avenue
Memphis, TN 38163 USA
Neuroanatomical correlates of bird song: individual
differences and evolution
David C. Airey, Ph.D.
- Ph.D. (1999) Psychology, Cornell University
- B.S. (1992) with Honors and Summa cum Laude, Psychology, University of Maryland, College Park
- Development and evolution of neural circuits
- Distribution of cell number in neural circuits
- Behavioral or performance correlates of neuroanatomical variation
- Song control system
- 1999present: Postdoctoral Trainee to Professor Robert W. Williams, University of Tennessee, Department of Anatomy and Neurobiology
Professional Society Memberships
- Society for Neuroscience
- International Mammalian Genome Society
- Behavioral Genetics Association
Grants and Fellowships
2000present: National Research Service Award (5.10.00 to 5.9.03)
19921995: National Science Foundation Fellow
1992: Summer UMCP Undergraduate Research Award ($2500).
- Phi Beta Kappa 1992
Airey DC, Kroodsma DE and DeVoogd TJ. (2000) Differences in the complexity of song tutoring cause differences in the amount learned and in dendritic spine density in a songbird telencephalic song control nucleus. Neurobiology of Learning and Memory. 73(3):274-81
Airey DC, Buchanan KL, Székely T, Catchpole CK, and DeVoogd TJ. (2000) Song, sexual selection and a song control nucleus (HVc) in the brains of European sedge warblers. J Neurobiol. 44(1):1-6
Airey DC and DeVoogd TJ. (2000) Greater song complexity is associated with augmented song system anatomy in zebra finches. NeuroReport. 11(10):2339-44
Airey DC, Castillo-Juarez H, Casella G, Pollak EJ and DeVoogd TJ. (in press) Variation in the volume of zebra finch song control nuclei is heritable: developmental and evolutionary implications. Proc Roy Soc Lond Series B.
Airey DC, Lu L and Williams RW. (in preparation) Gene loci for quantitative differences in cerebellum size of mice. J Neurosci.
Lu L, Airey DC and Williams RW. (in review) Genetic architecture of the mouse hippocampus: indentification of quantitative trait loci with specific effects on hippocampal size. J Neurosci.
Willliams RW, Airey DC, Kulkarni A, Zhou G, and Lu L (in press) Genetic dissection of olfactory bulbs in mice: QTLs on 4, 6, 11, and 17 modulate bulb size. Behavioral Genetics.
Epelboim J, Booth J, Airey DC and Steinman RM. (1992) Eye movements while reading unspaced text. ARVO Abstracts.
Airey DC, DeVoogd TJ and Kroodsma DE. (1994) Morphometry of song control neurons in differentially tutored Eastern marsh wrens (Cistothous palustris). Society for Neuroscience Abstracts.
Airey DC, Niederer JK, Nelson AL, DeVoogd TJ and Finlay BL. (1996) High vocal center and encephalization: Specialization and developmental constraints. Society for Neuroscience Abstracts.
Collins CE, Airey DC, Guzman J, Lisi V, Nam N, Tremper K and DeVoogd TJ. (1996) Developmental effects of song deprivation in zebra finches. Society for Neuroscience Abstracts.
Airey DC, Strom R and Williams RW. (1998). Genetic architecture of normal variation in cerebellar size. Society for Neuroscience Abstracts.
Williams RW, Strom R, Zhou G, Airey DC. (1998) New QTLs that modulate eye and brain development: comparison of RIs, F2s, and an advanced intercross. 12th International Mouse Genome Conference.
Airey DC, Castillo HJ, Pollak EJ, Casella G and DeVoogd TJ. (1999) Phenotypic and quantitative genetic description of Nissl-defined volumes of song control nuclei in zebra finches. Society for Neuroscience Abstracts.
Lu L, Airey DC, Gilissen E, Zhou G, Williams RW. (1999) New murine hippocampus-specific QTL maps to distal Chr. 1. 13th International Mouse Genome Conference.
Airey DC, Lu L, Strom R, Gilissen E, Zhou G, Williams RW. (1999) Cerebellum-specific QTLs in the mouse brain. 13th International Mouse Genome Conference.
Johnson D, Ferkin M, Hamre KM, Matthews D, Mittleman G, Williams RW, Airey DC, Rinchik E, and Goldowitz D. (2000) Generation of Neurological mutant mice using ENU-mutagenesis of discrete regions of the genome. Society for Neuroscience Abstracts.
Airey DC, Lu L, Williams RW. (2000) Gene loci for mouse cerebellum and internal granule layer size. Society for Neuroscience Abstracts.
Lu L, Airey DC, Williams RW. (2000) Complex trait analysis of the mouse hippocampus. Society for Neuroscience Abstracts.
Kulkarni A, Airey DC, Williams (2000) Genetic architecture of the mouse retinogeniculate system: A QTL analysis of numerical matching. Society for Neuroscience Abstracts.
- Psychology 322: Hormones and Behavior, Cornell University, Spring 1998
- Psychology 322-1: Hormones and Neural Circuits, Cornell University, Spring 1998
- Psychology 326: Evolution of Human Behavior, Cornell University, Fall 1997
- Psychology 332: Neurobiol. of Learn. and Mem., Cornell University, Spring 1997
- Psychology 123: Introduction to Biopsychology, Cornell University, Fall 1996
- Psychology 101: Introduction to Psychology, Cornell University, Summer 1996
- Psychology 276: Motivation, Cornell University, Spring 1996
- Psychology 324: Research Methods in Biopsychology, Cornell University, Fall 1994
- General Honors 100: Introduction to Research Methods at UMCP, Spring 1992
Thesis Advisor: Timothy J. DeVoogd
Department of Psychology and Field of Neurobiology and Behavior
Thesis Committee Member: Barbara L. Finlay
Professor and Department Chair
Department of Psychology
Thesis Committee Member: Bob Johnston
Department of Psychology
A central goal in behavioral neuroscience is the study and
explanation of correlations between brain and behavior. In my
doctoral thesis, "Neuroanatomical correlates of bird song:
individual differences and evolution", I sought to explore
and explain a positive correlation between natural variation in
the size of a brain region in songbirds and variation in song
behavior. I describe four original studies from my thesis below
that add to our understanding of how differences in song learning
affect the brain and how differences in brain size affect the
capacity for song learning. My work also informs our understanding
of the evolution of neural capacity for behavioral differences.
I first present background and rationale.
Three articles best illustrate the fascinating and core observation
between brain and behavior in songbirds. Canady, Kroodsma, and
Nottebohm (1984) found positive correlations between variation
in volume of the song control nucleus HVc (proper name) and the
number of song types recorded from individual Eastern or Western
marsh wrens. This paper was the first to show a correlation between
HVc and song repertoire size within and across wild songbird populations.
Within either subspecies, wrens with a larger HVc had a larger
song repertoire. Western wrens, which naturally sing a repertoire
of about 150 song types, had a larger HVc than Eastern wrens that
naturally sing about 50 song types. There was no estimated difference
in overall brain size between the two marsh wren populations.
Because song is learned by juveniles from singing adults, a key
question was raised. Does learning cause HVc size or does HVc
size limit learning? Kroodsma and Canady (1985) brought nestling
Eastern and Western marsh wrens into the laboratory where they
were artificially tutored with a supernormal repertoire of combined
Eastern and Western song types. Despite this common rearing environment,
individual Eastern wrens developed repertoire sizes typical for
Eastern wrens and Western wrens developed repertoire sizes typical
of Western wrens. In addition, the volume estimate of HVc for
Western wrens was larger than for Eastern wrens. The differences
between Eastern and Western wrens in repertoire size and the volume
of HVc was thus presumed to have a genetic basis. Brenowitz, Lent,
and Kroodsma (1995) addressed the direction of causation for the
observed HVc and repertoire size correlation by asking if learning
a small or large repertoire of song types caused differences in
HVc size. Nestling Eastern marsh wrens were brought into the laboratory
and artificially tutored with either a small repertoire or a large
repertoire of songs. While the tutoring successfully created a
large behavioral difference, there were no differences in the
volumes of HVc between groups. Within the group that heard a large
repertoire, the correlation between the volume of HVc and repertoire
size approached significance.
The results of these three studies suggest that the size of the
song control nucleus HVc is unaffected by early acoustic learning
experiences related to hearing or producing different repertoire
sizes, and that the size of HVc may set an upper limit to the
size of the song repertoire that can be acquired. This working
hypothesis claims that birds with large HVc can sing either large
or small repertoires, but that birds with small HVc may not sing
large repertoires. The hypothesis does not, however, explain how
or why the correlation occurs.
Good explanations of biological causation address questions related
to both how and why a behavior occurs. Tinbergen (1963) usefully
partitioned biological causation into four complementary descriptions.
To ask how a bird sings a large song repertoire is to inquire
after mechanism or development. To ask why a bird sings a large
song repertoire is to inquire after fitness consequences or ancestral
qualities. Mechanism, ontogeny, phylogeny, and adaptation are
each necessary pieces of a complete biological causal statement
of repertoire size (e.g., Kroodsma and Byers 1998; DeVoogd and
Szekely 1998). I organized my study of the correlation between
HVc and repertoire size in songbirds with three questions: (1)
What is the developmental basis for differences in HVc size? (2)
Do differences in HVc size translate to performance differences
in song behavior? (3) Do performance differences in song behavior
(and differences in HVc size) translate to differences in reproductive
Chapter 2, "Sexual selection and brain size", provides
a broad background of material related to why repertoire size
is important to songbirds. I define sexual selection and provide
a brief survey of reviews describing sexual selection of song
in animals, particularly songbirds. I conclude that song has two
primary functions, territory defense and mate attraction, and
that repertoire size is a cue used by females to choose mates.
I follow this review with another broad background of material
related to the developmental sources of size variance in brains
and brain parts, particularly nucleus HVc. The intersection of
sexual selection and brains has been addressed by Jacobs (1996).
She asserted, "if females...can assess a male's ability to
learn more information, or learn more quickly than rival males,
then the demonstration of enhanced learning abilities could be
a sign of a superior mate. The rationale for this is based on
the idea of limited 'brain space'; learning requires an investment
in brain tissue, which is extremely metabolically expensive. Hence,
if enhanced learning demands increased allocation to 'brain space',
then conspicuous indicators of learning ability (and their underlying
brain structures) could be subject to sexual selection."
Chapter 3 presents a study titled, "Differences in the complexity
of song tutoring cause differences in the amount learned and in
dendritic spine density in a songbird telencephalic song control
nucleus" (Airey et al., 2000a). This study tested whether
differential experience that leads to differences in adult song
repertoire affects dendritic spine density in HVc and its efferent
target RA (robustus archistriatalis). Juvenile eastern marsh wrens
were tape tutored with either 5 or 45 song types. As adults, the
small repertoire group had learned mostly 5 or 6 song types, the
large repertoire group 36 to 47. Wrens that learned the large
song repertoires had a greater dendritic spine density for the
most spiny neurons present in HVc (mean difference, 36%), but
not in RA. Recent physiological evidence describes HVc as a premotor
area coding syllables, motifs, and higher-order song patterns,
and my data now clearly reveal that differences in the size of
the song repertoire that are experienced lead to differences both
in song learning and in the density of dendritic spines in HVc.
In the forebrain song nuclei of these songbirds, as in some other
vertebrate systems, differences in learning and performance are
associated with differences in synaptic anatomy specifically in
the region that organizes the learned pattern. This study was
conducted in collaboration with Donald E. Kroodsma on the alternate
hemispheres of the brains of the same marsh wrens used
in the study by Brenowitz et al. (1995) described above. Although
those authors showed that learning either a small or a large song
repertoire does not cause gross changes in the volume of HVc,
my results suggest that the working hypothesis above should be
modified to respect the distinction between fine and gross anatomical
correlates of learning a small or a large song repertoire.
Chapter 4 presents a study titled, "Song, sexual selection
and a song control nucleus (HVc) in the brains of European sedge
warblers" (Airey et al., 2000b). Prior studies showed that
repertoire size is a cue female sedge warblers use to choose mates
(Buchanan and Catchpole, 1997). This study tests two predictions
concerning HVc size in male sedge warblers: first, that males
with more complex songs will have a larger HVc, and second, that
males who pair successfully will have a larger HVc than unpaired
males. Data on song composition and pairing status were collected
from wild sedge warblers breeding in Hungary. I found that male
HVc volume standardized over telencephalon volume predicts three
measures of song complexity, including repertoire size. I did
not detect a group difference in HVc volume between six paired
males and six unpaired males. I could not claim female choice
as a source of directional selection for increased HVc size in
sedge warblers. However, power analysis also suggests caution
in claiming evidence that the groups are not different. Further
study of this population is required to address the effects of
age and female choice on HVc size in male sedge warblers. The
results contribute to a growing body of evidence supporting a
functional relationship between individual differences in song
performance and HVc size.
Chapter 5 presents a study titled, "Greater song complexity
is associated with augmented song system anatomy in zebra finches"
(Airey and DeVoogd, 2000c). In this study I revisited the relation
between brain anatomy and song behavior in zebra finches, believing
that two prior studies were not conclusive for all zebra finches
in all environments. I found that differences in song behavior
(repertoire size and phrase duration) are related to differences
in HVc size in zebra finches that have learned song in a complex
social setting. This is similar to previous investigations of
song complexity and HVc size in canaries, marsh wrens, starlings,
and sedge warblers. I advanced the study of brain and behavior
correlations in songbirds by showing that repertoire size in zebra
finches can be well predicted by the combined variance in several
brain nuclei (multiple R = 0.871), providing the first demonstration
that volumetric differences across multiple components of a neural
network are predictive of song behavior. The result apparently
depends on the ratio of the neostriatal nuclei (HVc and lMAN)
to the paleostriatal nucleus Area X. Thus, larger song repertoires
and longer songs may require greater amounts of the nucleus HVc
to represent more song units or the temporal sequences of linked
acoustic units, but in zebra finches a greater amount of HVc per
se may not be as important as the amount relative to the volumes
of other circuit nuclei.
Chapter 6 presents a large-scale neuroanatomical study titled
"Variation in the volume of zebra finch song control nuclei
is heritable: developmental and evolutionary implications"
(Airey et al., 2000d), in which I estimated the additive genetic
determinants of size variance in HVc and other song control nuclei
by measuring the volume of all major song nuclei and of the telencephalon
in more than 100 adult male zebra finches from 38 families. In
many songbird species, females prefer males who sing a larger
repertoire of syllables. Males with more elaborate songs have
a larger nucleus HVc, the highest structure in the song production
pathway. HVc size is thus a potential target of sexual selection.
I found evidence that the size of HVc and other song production
nuclei are heritable across individual males within a species.
In contrast, I found that heritabilities of other nuclei in a
song learning pathway are lower and are not different from zero,
suggesting that variation in the sizes of these structures is
more closely tied to developmental and environmental differences
between individuals. I found that evolvability, a statistical
measure that predicts response to selection, is higher for HVc
and its target for song production, RA, than for all other brain
volumes measured. This suggests that selection based on the functions
of these two structures would result in rapid major shifts in
their anatomy. I also found that the size of each song control
nucleus is significantly phenotypically correlated with the song
related nuclei to which it is monosynaptically connected, and
that the correlation between HVc and RA contains a significant
genetic component. This suggests that selection on HVc through
female choice on male song complexity may produce correlated shifts
in the anatomy of RA. Finally, I found that the volume of the
telencephalon is 8% larger in males than in females, a result
which suggests that selection on HVc size may be more costly than
previously suspected, as HVc correlates significantly with telencephalon
These findings begin to join theoretical analyses of the role
of female choice in the evolution of bird song to neurobiological
mechanisms by which the evolutionary changes in behavior are expressed.
Because aspects of song behavior in songbirds are learned, song
is in part a culturally transmitted behavior, and this may be
the strongest immediate factor determining song complexity for
individual songbirds. But evidence suggests that the size of the
song nucleus HVc may set an upper limit to the size of the song
repertoire that can be learned. A role for the genetic transmission
of repertoire size must also be considered. I raised three questions
at the outset, (1) What is the developmental basis for differences
in HVc size? (2) Do differences in HVc size translate to performance
differences in song behavior? (3) Do performance differences in
song behavior (and differences in HVc size) translate to differences
in reproductive success? I have shown that the developmental basis
of HVc size and other song control regions includes additive genetic
contributions of the parental genome. I have also shown that performance
differences in male song behavior correlate robustly with male
HVc size. Continued research may also show that females that choose
males with more complex song or larger HVc are ensuring the reproductive
benefits of these characters in their sons.
Airey DC, Kroodsma DE and DeVoogd TJ. (2000a) Differences
in the complexity of song tutoring cause differences in the amount
learned and in dendritic spine density in a songbird telencephalic
song control nucleus. Neurobiology of Learning and Memory:
Airey DC, Buchanan KL, Székely T, Catchpole CK, and DeVoogd
TJ. (2000b) Song, sexual selection and a song control nucleus
(HVc) in the brains of European sedge warblers. J Neurobiol.
Airey DC and DeVoogd TJ (2000c) Greater song complexity is associated
with augmented song system anatomy in zebra finches. NeuroReport.
Airey DC, Castillo-Juarez H, Casella G, Pollak EJ andDeVoogd TJ.
(2000d) Variation in the volume of zebra finch song control nuclei
is heritable: developmental and evolutionary implications. Proc
Roy Soc Lond Series B. In review.
Buchanan KL and Catchpole CK. (1997) Female choice in the sedge
warbler, Acrocephalus schoenobaenus: multiple cues from song and
territory quality. Proc Roy Soc Lond Series B 364, 521-526.
Brenowitz EA, Lent K and Kroodsma DE. (1995) Brain space for learned
song in birds develops independently of song learning. J Neurosci
Canady RA, Kroodsma DE and Nottebohm F. (1984) Population differences
in complexity of a learned skill are correlated with the brain
space involved. Proc Natl Acad Sci USA 81, 6232-6234.
DeVoogd TJ and Székely T. (1998) Causes of avian song:
Using neurobiology to integrate proximate and ultimate levels
of analysis. In: Animal Cognition in Nature (pp. 337-380): Academic
Jacobs LF. (1996) Sexual selection in the brain. Trends in
Ecology and Evolution 11(2), 82-86.
Kroodsma DE and Canady RR. (1985) Differences in repertoire size,
singing behavior, and associated neuroanatomy among marsh wren
populations have a genetic basis. Auk 102(3), 439-446.
Kroodsma DE and Byers BE. (1998) Songbird song repertoires: an
ethological approach to studying cognition. In: Animal Cognition
in Nature (pp. 305-336): Academic Press.
Tinbergen N. (1963) On aims and methods of ethology. Zeitschrift
fur Tierpsychologie 20, 410-433.
CV formatted for web by Alex Williams (firstname.lastname@example.org).