SHORT COMMUNICATIONS 
371 
in concert. First, female singers may have 
anomalously high androgen levels, resulting in 
the development of male-like characteristics (e.g., 
Baldwin et al. 1940, Byers and King 2000). 
Similarly, song may develop in older females that 
have begun to acquire more androgen-mediated, 
male-like characteristics (Nolan 1978). Alterna¬ 
tively. atypical female singing may be functional, 
but only used by a limited number of individ¬ 
uals in specific situations, for example, at high 
breeding densities (Hobson and Sealy 1990) or 
during pair-bond formation (Gilbert and Carroll 
1999). 
The most frequent vocalizations of the Com¬ 
mon Yellowthroat, (Geothlypis trichas), have 
been well studied; males produce two main non¬ 
singing calls, chips and rattles, along with two 
songs, a commonly repeated perch song and a less 
frequently used flight song (Wunderlc 1978, 
1979; Kowalski 1983; Ritchison 1991, 1995; 
Guzy and Ritchison 1999). Females also chip 
and produce infrequent rattles (CCT. pers. obs.). 
Female Common Yellowthroats also have been 
observed singing on rare occasions (pers. comm, 
from B. E. Byers to D. A. Spector; Spector 1992) 
but. to the best of our knowledge, no previous 
publications have described their song or provided 
recorded evidence of their singing. We observed 
one potentially singing female in 2010 (KAL, 
pets, obs.), and a confirmed singer in 2011 in 
observations of 141 individual females over 
7 years. We compare measurements of morpho¬ 
logical. physiological, and plumage characteris¬ 
tics of the 2011 singing female with those of 
typical males and females at our study site, and 
discuss the implications these measurements may 
have on understanding reasons for atypical female 
singing. 
METHODS 
We monitored a breeding population of Com¬ 
mon Yellowthroats in Saratoga Springs, New 
York. USA (43° 10' 24.6" N. 73 53' 19.7" W) 
from early May to late July 2011. Yellowthroats at 
our study site had been closely observed for seven 
breeding seasons (2005-2011). All adults were 
captured and banded each year with a U S. 
Geological Survey aluminum band and a unique 
combination of colored plastic leg bands. We 
collected information at time of capture on bib size 
and coloration, and morphology. We also collected 
blood samples to measure hematocrit and paternity 
following Freeman-Gallant et al. (2010. 2011) and 
Taff et al. (2011). We used plasma from our blood 
samples to measure total antioxidant capacity and 
testosterone levels with commercially available 
microplate kits (total antioxidant capacity: Cayman 
Chemical, Ann Arbor. MI. USA; testosterone: 
Enzo Life Sciences. Furmingdale. NY. USA). We 
mapped the boundaries of each breeding territory 
and measured the distance from the centroid of 
each territory to the centroid of nearby territories. 
We calculated breeding density as the number of 
territories with centroids <300 m from the centroid 
of each focal territory. 
We recorded the singing female with a 
Sennheiser ME 66 shotgun microphone and a 
Sony MZ-M200 Hi-MD recorder. The archived 
recording is available through the Macaulay 
Library at Cornell University (www.macaulay 
library.org. catalog number: ML Audio 166385). 
We monitored only one singing female and were 
unable to perform statistical tests to evaluate 
possible associations between female singing and 
correlated parameters. We compared the traits ot 
the singing female with the average trait values 
for all males and females breeding at our site in 
2011. Specifically, we compared measures for 
trails relevant to three hypotheses for female 
singing: (1) acquisition oi male-like traits (wing 
and tarsus length, bib size and coloration), (2) 
high androgen levels (plasma testosterone), and 
(3) functional singing at high breeding density 
(neighborhood density). We also compared 
several other traits as potential indicators the 
singing female was anomalous in some way that 
was not accounted for by the three hypotheses 
considered imass, total antioxidant capacity, and 
hematocrit). 
RESULTS 
Wc monitored 141 individual females in the 
7 years that we studied our field population. Many 
of these females returned to the study site over 
multiple years and wc observed 193 female 
breeding attempts. One female may have sung in 
2010 (KAL, pers. obs.), but this observation was 
not confirmed and we were unable to obtain a 
recording. Only the female observed in 2011 sang 
conclusively. Thus, female song appears to be rare 
in this species. 
We first observed this female singing on 27 
May 2011. The female continued to sing each day 
until 2 June 2011, at which time she had 
completed nest building. We did not observe the 
female singing again after that date for the 
