SHORT COMMUNICATIONS 
373 
TABLE 1. Trails of male and female Common Yellowthroats breeding at Saratoga Springs, New York, USA in 2011 in 
comparison to ihe singing female. The measure for each trait is shown for the singing female and the number of standard 
deviations from the mean of other breeding females and males. 
Rdtiant hypmheses 
Trail 
Breeding males 
Breeding females 
Singing female 
n 
X = SD 
n 
X ± SD 
Obs. 
SD from F 
SD from M 
Mass (g) 
37 
10.1 ± 0.35 
23 
10.8 ± 0.78 
9.3 
-1.9 
-2.3 
Male-like 
Tarsus (mm) 
37 
19.9 ± 0.48 
24 
19.5 ± 0.61 
19.2 
-0.5 
-1.5 
Male-like 
Wing (mm) 
37 
54.9 ± 1.26 
24 
51.8 ± 1.40 
51.0 
-0.6 
-3.1 
Male-like 
Bib (mm 1 ) 
37 
671 ± 161 
24 
397 ± 142 
283 
-0.8 
-2.4 
Male-like 
UV B. a 
37 
10.0 ± 2.24 
24 
3.5 ± 2.96 
3.8 
0.1 
-2.8 
Male-like 
Yellow B.“ 
37 
31.2 ± 3.42 
24 
24.9 ± 4.30 
24.7 
-0.1 
-1.9 
Male-likc 
Bih C c „ c 
37 
0.91 ± 0.04 
24 
0.84 ±0.10 
0.89 
0.5 
-0.5 
Hematocrit 11 
34 
0.53 ± 0.04 
22 
0.45 ± 0.04 
0.47 
0.7 
-1.5 
TAC 
34 
1.6 ± 0.55 
22 
1.9 ± 0.69 
1.0 
-1.3 
-1.1 
Androgens 
T 
23 
3.3 ± 1.45 
9 
1.7 ± 0.82 
2.6 
1.2 
-0.5 
Functional 
Density* 
23 
4.0 ± 1.90 
23 
4.0 ± 1.90 
2 
-1.1 
—1.1 
" l : V brightness of bib feathers: average refleclanee from 320 to 400 nm. 
h Yellow brightness of bib feathers: average reflectance from 550 to 625 nm. 
r Carotenoid chroma of bib feathers: (Reflectance,,,! Reflectance^i/ReflectanceToo- 
“ Ratio of blood cells to plasma: (blood cell volume + plasmaVwhole sample volume. 
Total antioxidant capacity: measured in antioxidants iniM). 
Testodcrone: measured in pg/ml. 
8 Number of neighboring pairs holding a territory with centroid to centroid distance <300 in. 
and total antioxidant capacity) were normal for 
females in our population (Table I). 
DISCUSSION 
Rare but regular observations of singing females 
occur in many species where only males normally 
sing, The persistence of these observations has led 
to speculation about the causes of atypical female 
singing. However, the scarcity of observations does 
not allow hypotheses to be evaluated because the 
relevant correlates have not been measured. We 
had observations of only a single singing female 
hut measured covariates that were relevant to most 
of the proposed explanations for atypical female 
singing. We found no compelling evidence that any 
proposed hypothesis was adequate to explain 
singing by the female yellowthroat wc observed, 
and the basis for atypical female singing remains 
unclear. 
The most commonly accepted explanations tor 
atypical female singing arc an over-expression ol 
androgens or the acquisition of male-like traits by 
older females (Baldwin et al. 1940. Nolan 1078, 
Byers and King 2000). Our singing female’s 
testosterone levels were no higher than those in 
other females at our study site that did not sing. 
She may, however, have had high levels ot 
androgens during critical developmental stages 
before our observations. Alternatively, she may 
have had abnormally high sensitivity to the effects 
of androgens; in this case, even normal levels of 
testosterone may have been sufficient to produce 
song. Finally, although our singing female’s 
measured testosterone was not abnormal, it was 
relatively high, and this may have predisposed her 
to sing based on factors that were not measured 01 - 
in combination with other traits. Thus, we cannot 
rule out the possibility that testosterone had a role 
in her singing. Both the unconfirmed singer in 
2010 and the singing female in 2011 were first- 
lime breeders at our site, and may have been 
young. We observed eight breeding females 4 or 
more years of age (extremely old lot females in 
this species) in previous years and we did not 
observe singing in these older individuals. None 
of the morphological traits we measured lor this 
female (i.e.. wing and tarsus length and parame¬ 
ters associated with bib size and color) was 'male- 
like*. further suggesting that age- and androgen- 
related hypotheses are not adequate explanations 
for singing in this case. 
Common Yellowthroats have been well stud¬ 
ied and female song is unlikely to have been 
routinely missed. However, if females use song 
only during short windows each breeding season, 
female song may not be as rare as previously 
suspected. The female that we observed in 2011 
sang repeatedly for only 5-6 days and the 
suspected singing female in 2010 sang much 
less frequently and for only 2-3 days. Females in 
both cases sang a highly abnormal song that 
would not be easily recognized as a Common 
