Reprinted from
II March 1977, Volume 195, pp. 1002-1004
Species Identification in the North American Cowbird:
Appropriate Responses to Abnormal Song
Andrew P. King and Meredith J. West
Copyright© 1977 by the A merican Association fOT the Advancement oj Science
•
Species Identification in the North American Cowbird:
Appropriate Responses to Abnormal Song
Abstract. Female cowbirds raised in auditory isolation from males responded to
the songs of male cowbirds with copulatory postures. The songs of males reared in
isolation were more effective in eliciting the posture than the songs of normally
reared males. The females did not respond to the songs of other species. These re­
sults indicate one mechanism of species identification for this parasitic species.
M0-
courtship and is accompanied by a bow­
lothrus ater, is reported to parasitize
more than
200 species (1). Thus, unlike
ing display. We collected data from a
small group of males that suggested that
most vertebrates, young cowbirds are
male peer experience was sufficient to
not
con­
produce normal adult courtship song.
specific parents during their earliest inter­
However, males reared in isolation from
actions with adults or peers. Although
both male peers and adults, developed
much remains to be learned about the de­
songs that differed from the normal ver­
velopmental impact of such stimulation,
sion, particularly in the extent to which
there is growing evidence to demonstrate
some of the notes are mod�lated (Fig.
The
brown-headed
exposed
to
cowbird,
stimulation
by
its facilitative role in avian species identi­
2).
The following experiments were de­
(2). How, then, does the cow­
signed to elaborate upon these observa­
bird, naturally deprived of such experi­
tions by testing other naive females. The
ence, come to identify other cowbirds?
songs of both normal and isolate males
This question has puzzled students of de­
were included to help clarify the role of
velopment and evolution for many years,
peer experience in song development.
fication
and, until now, only speCUlative solu­
The subjects were six female cowbirds
tions existed (3). We report a series of ex­
raised in auditory and visual isolation
periments detailing a mechanism of spe­
from adult cowbirds. Their prenatal audi­
cies identification that allows us to under­
tory experience was controlled in that
stand how cowbirds have evolved to
eggs were obtained from a captive colo­
overcome the fact that they were not
ny of cowbirds and incubated in isola­
reared by conspecifics.
tion. Their postnatal social contact was
Two findings led to the discovery of
limited to a
1- to 5-day period during
this mechanism. First, we had raised a fe­
which the individual eggs hatched into
male cowbird in the laboratory in com­
barn swallow (Hirundo rustica) nests,
plete auditory and visual isolation from
and the young birds were fed by their
other cowbirds. At 8 months of age, she
hosts. After being returned to the labora­
was exposed to a recording of male cow­
tory, they were hand-reared as a group
bird courtship song. She adopted a pos­
without contact with adult cowbirds.
ture we term the "copulatory response":
They were placed in soundproof cham­
her
wings were lowered and
spread
bers between the ages of 35 and
60 days
apart, her neck and body were arched,
(4). The birds were housed in pairs to al­
and the feathers around the cloacal re­
gion were separated (Fig.
1). A series of
low the opportunities for social inter­
action and auditory experience thought
playbacks confirmed the specificity of
to be essential for the birds to come into
the response: she showed no response to
breeding condition. Chambers
the songs of a red-winged blackbird
each housed two female cowbirds; cham­
1 and 2
(Agelaius phoeniceus) or to the cowbird
ber 3, a female cowbird and a female red­
flight call.
winged blackbird; chamber 4, a female
The second finding related to the de­
cowbird and a male cowbird; chamber 5,
velopment of the male song. Male cow­
a male cowbird and a male cardinal
bird song, which can be transcribed as
(Richmendena
"burble
phase of their photoperiods was gradu-
burble
.
j
tsee," occurs during
cardinalis).
The
light
�'---'.
--
"
cowbird, was transferred to a chamber
housing a female canary 3 weeks before
the
1976 testing. Also, two samples of
16
-
NORMAL
12
8
parts of the abnormal song were used: (i)
the "burble-burble" phrase alone, (ii)
the "tsee" alone, and (iii) a new sample
4
16
of abnormal song from another isolate
12
male.
The abnormal song retained its superi­
or effectiveness, and abnormal, normal,
and
partial-abnormal
songs
were
all
more effective than the controls (Table
Fig. 1. Female cowbird displaying the copula­
tory response.
ABNORMAL
1). The differences among the conditions
were reliable (Friedman two-way analy­
sis of variance, Xr2
=
23.3, P <
.001).
Both samples of abnormal song were
ally extended during the spring months.
equally effective, and one of the partial
The playback period began in May
songs was more effective than the full
(6). Female CB, despite her
1975, when the birds were 9 months old,
normal song
and was conducted in two phases. In the
sexual experience and her lack of re­
first phase, the females were exposed to
sponding in the earlier experiment, re­
recordings of normal cowbird song ob­
sponded as strongly as the sexually naive
tained from a wild-caught male and the
females
8
4
r
•
O�
____
�
.
r \��,�
,.. III �
'�.
__"L�
'
�
'� ��
" I
'
_
II
R
________
_
'"
,
1----112 II< -----I
Fig. 2. These recordings were made on a sono­
graph (Kay) with a wide-band filter and set for
flat response and linear display. The units on
the ordinate are kilohertz. The most striking
difference is the rapid modulation between 2
and 8 khz present in the abnormal but not the
normal song. This type of modulation was not
present in any of a sample of 522 song spectro­
graphs obtained from a captive colony of nor­
mally reared male cowbirds (10).
(7).
control songs of male red-winged black­
Another piece of evidence attests to
tory and sexual experience with males
birds, meadowlarks (Sturnella magna),
the potency of the abnormal song as a
does not diminish the effectiveness of the
and Baltimore orioles (Icterus galbula).
sexual releaser. Two wild female cow­
abnormal song.
There
birds were captured and placed in the lab­
were
four
playbacks
spread
1976. While in
In summary, these data indicate that
throughout each day; the order of pre­
oratory in the winter of
sentation varied each day but was the
these
but
experience with adult cowbirds in order
same for each bird. In the second phase,
could not see, caged male cowbirds and
for species identification to occur. Naive
the procedure was identical except that
male and female free-living cowbirds. Al­
females
... the playback series included the abnor­
'tal song of an isolate male. The re­
though a full series of playbacks was not
song; naive males produce an acoustical­
carried out owing to the birds' nervous­
ly atypical, but highly effective, song.
presence or ab­
ness in captivity, the first time the abnor­
Isolate song, then, appears to have quali­
mal song was played to them, one of the
ties of a "supernormal" releaser. Why
During the first phase, four of the six
females immediately adopted the copUla­
normal song appears less effective is as
female cowbirds responded more often
tory posture. In subsequent exposures,
yet unspecified, but it may reflect the
to normal song than to any of the control
fact that males sing in contexts other
songs (Table
1). One of these, female
she continued to respond to the abnor­
mal song but never to the normal. Thus,
than courting. The songs of males, there­
DB, showed a dual preference for cow­
response to abnormal song was not re­
fore, may contain more complex mes­
bird and meadowlark song. The fifth cow­
stricted to females raised in isolation.
sages indicating not only the species,
bird, female W, did not come into breed­
Furthermore, the response of this female
sex, or breeding condition of the singer,
ing condition and hence did not respond.
and that of female CB suggest that audi-
but agonistic or territorial information as
•
\......sponse measure was the
sence of the copulatory posture
(5).
surroundings,
they
heard,
neither male nor female cowbirds require
respond
adaptively
to
male
The sixth, female CB, was housed with
the male cowbird and also showed no
response. She was observed, however,
to copUlate frequently. The variability
across the females was probably due to
variations in breeding condition, which
was indicated by the numbers of eggs
laid.
During the second phase, four females
demonstrated a clear preference for the
abnormal song (Table
1). For each of the
four birds, the abnormal song produced
the highest rates of response, the normal
song the second highest and the control
songs of the other species the lowest
rates.
/.
In order to pursue this finding in great­
�
'"
detail, the birds were maintained in
\,--tSolation for another year and retested in
\
the spring of
1976. Several procedural
changes were also made. Female CB,
which had been reared with the male
•
.
well. A closer assessment of the function­
another. Little is known about cowbird
al properties of the songs of normally
social behavior, yet data from our labora­
reared males may help to clarify these
tory indicate that cowbirds have highly
speculations. In addition, these data sug­
structured dominance hierarchies and
gest the need to examine in more detail
complex intraspecific behaviors that may
the developmental role of self-stimula­
also facilitate identification or maintain
tion. It may be that the self-generated
social integration among acquainted cow­
sounds of the isolate males channeled
birds
their song development toward an em­
scribed might represent an independent
phasis on the acoustic properties that
system designed to ensure identification
contain the sexual message in cowbird
during the most important context, the
song.
Males reared with other males,
breeding season, and to enhance repro­
however, whether age mates or adults,
ductive opportunities by inducing sex­
may receive acoustically more varied
ually appropriate behavior in the female.
stimulation, may learn to modify their
Thus,
song components as a result of auditory
multiple means to identify one another,
and behavioral feedback from their com­
the presence of a mechanism geared di­
panions (thereby leading to a dilution of
rectly for reproduction may represent a
the purely sexual message), or both. In
critical adaptation for a parasitic species.
any case, these data implicate self-stimu­
ANDREW P. KING
(9).
The mechanism we have de­
although
cowbirds
lation as possibly a very important form
Department of Psychology,
of species-typical experience for cow­
Cornell University,
birds as has been hypothesized
Ithaca, New York
(8).
Final­
5.
may
have
14850
MEREDITH J. WEST*
Iy, although it is tempting to label the
cowbird's response to isolate song as id­
Department of Psychology,
iosyncratic or unrepresentative of other
University of North Carolina,
songbirds, this cannot be done, as little
Chapel Hill
6.
7.
27514
late songs. Such information would be
useful in species that, like the cowbird,
have courtship songs. Songs of a primari­
ly territorial nature may be more difficult
to assess.
Although these results provide one an­
swer to the question of the mechanism
by which cowbirds identify one another,
they do not explain how young or juve­
nile cowbirds first come to recognize one
References and Notes
I. H. Friedmann, U.S. Natl. Mus. Bull. 233,
6,
8.
9.
(1963).
2. G. Gottlieb,
Development 0/ Species Identifica­
tion: An Inquiry into the Prenatal Determinants
o/Perception (Univ. of Chicago Press, Chicago,
1971).
3. D. Lehrman, in Ethology
and Psychiatry, N. F.
White, Ed. (Univ. of Toronto Press, Toronto,
1974) pp. 187-197; E. Mayr, Am. Sci. 62, 650
(1974).
4. Each chamber consisted of two concentric
boxes constructed of elywood and sheetrock.
Wood and acoustic ule baffles between the
boxes were designed to be most effective be­
tween 2 and 16 khz. Suppression was greater
tested with t-tests. Reliable differences were
obtained for the following comparisons: abnor­
mal one or two versus normal song (abnormal
11.04, P < .001; abnormal two:
one: t
t = 9.95; P < .(01); abnormal versus control
songs (abnormal one: t = 8.85, P < .001; abnor­
mal two: t = 19.43, P < .(01); normal versus
control song (t = 6.09, P < .(01); the "burble"
phrase versus the "tsee" phrase (t = 3.17,
P < .05); and either partial song versus the nor­
mal "burble" (t = 4.22, P < .(01). There was
no reliable difference between the two samples
of abnormal song or between the "tsee" phrase
and normal so
. ng.
Female CB's lack of response the previous year
probably resulted from her being housed with a
male cowbird, who repeated his sorog to her
many times each day, thus raising her threshold
for song responsiveness. Furthermore, her com­
panion's song was undoubtedly of superior
acoustic quality, a factor that might also have
diminished her interest in the recorded songs.
G. Gottlieb, Psychol. Bull. 79, 362 (1973).
A. King, paper presented at the meeting of the
Animal Behavior Society, Wilmington, N.t., 22
to 27 May 1975.
D. Eastzer, thesis, Cornell University (1975).
We thank G. Wilcox who designed the isolatior
chambers, R. Chu and D. Eastzer who helped i
the field by collecting eggs and nestlings, R. E��
Johnston for valuable suggestions during the
initial stages of this - project, and P. Cabe, W.
Dilger, D. Miller, P. Ornstein, and H. Rhein­
gold for helpful comments on the manuscript.
We are especially grateful to G. Gottlieb for
both technical and conceptual assistance.
Request for reprints should be sent to M,J. W.
=
comparative information exists regard­
ing the responses of other species to iso­
than 39 db at 1000 hertz, and it increased with
higher frequencies to greater than 50 db between
8 and 16 khz. The interior box was a I. l-m cube,
fabric-lined to reduce sound reflection, lighted
by two 40-watt Vita Lite tubes and continuously
ventilated. White noise was broadcast in the
room housing the chambers.
The songs were recorded with a dynamic mien"
/
phone (Uher 5(7) and tape recorders (Uher-4000). The recordings were played through a
driver (JBL 2420) and hom (JBL 2340). Play­
back levels were determined with a sound pres­
sure meter (General Radio 1933). The same
sound pressure levels (SPL's) \\ere used for
playbacks of the normal and the abnormal
songs. At 0.35 m from the speaker along the
axis, the SPL was 86 ± 1.5 db slow reading and
104 + 1.5 db impulse. Control songs of the other
species were adjusted to the slow reading. The
A-weighted SPL inside the chamber was 50 db
slow reading.
Differences between categories of song were
10.
11.
15 September 1976; revised 15 November 1976
"