836 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123. No. 4. December 2011 
Numerous studies demonstrate that oscine 
songs are individually distinctive and can be used 
to identify individuals based on song alone. 
However, relatively little is known about songs 
of suboscines, particularly individual distinctive¬ 
ness (Lovell and Lein 2004). Suboscine songs are 
generally thought to be less variable than those of 
oscines as they are not learned (Kroodsma 1984). 
which leaves less chance for errors in the learning 
process (Raposo and Hofling 2003. Wiley 2005). 
Reduced variability could diminish the ability of 
songs ot suboscines to provide information on 
individual identity and quality or geographic 
population structure. 
Several recent studies have explored suboscine 
song distinctiveness, song discrimination, and 
even individual recognition (e.g., Sedgewick 
2001, Bard et al. 2002. Rios-Chelen et al. 2005, 
Wiley 2005, Fitzsimmons et al. 2008, Lein 2008 
Femfmdez-Jurieic et al. 2009). Lovell and Lein 
(2004) demonstrated the suboscine Alder Fly¬ 
catcher (Empulonax cilnorum) could discriminate 
between the songs of neighbors versus strangers 
and (hat this species was also capable of 
individual recognition via vocalizations alone 
(Lovell and Lein 2005). the first time this ability 
was shown in a suboscine. 
We explored vocal distinctiveness in the 
Eastern Wood-Pewee (Contopus virem). a sub¬ 
oscine of eastern North America (McCarty 1996) 
with several distinct male songs, including the 
onomatopoeic 'pee-ah-wee* song, for which this 
species ,s named, as well as this species’ main 
secondary vocalization, the ‘wee-ooo’ song 
l0ng f0CUSed Astern 
Wood-Pe vee songs (e.g.. Craig 1943. Smith 
1988), although studies of this species' songs 
have not explored individual distinctiveness g 
Vocalizations should have relatively little 
within individual variation to effectively convey 
information while being comparatively more 
variable between individuals (Falls 1982) The 
production of a unique set of vocalizations by an 
individual constitutes a vocal signature. Our 
pee ahwe WaS ? *** ^ prediction thal both the 
pee-ah-wee and wee-ooo songs of male Eastern 
C ° main individual| y-specific vocal 
methods 
producer bJ de mal P e ee E^t VVee w nd Wee "°°° son S s 
natural Condi,^ifw , ^ 00d ‘ Pe »“ a under 
DS W “tuhe st er County. New 
York and Fairfield County, Connecticut, USA 
during early June 2008. We used a Maranfr 
PMD660 digital recorder with a Sennheiscr MEW 
directional microphone to record songs, We 
visually tracked focal birds, and all songs of cac 1 
focal male were recorded in a single recordin' 
session to ensure all recordings were of focal 
males. We ended the recording session and all 
recordings from that session were excluded from 
analyses if we lost visual contact with a foal 
individual or another conspecific individual 
appeared, Wc recorded and analyzed 323 pee- 
ah-wee songs from 16 males and 201 wee-ooo 
songs from 17 males. Songs of individuals could 
vary over time or with motivational circumstanc¬ 
es. Our method is unlikely to capture such 
variation. 
Songs were digitized at a sampling rale of 
22.05 kHz. and a 16-bit depth, and presented as 
spectrograms generated with FFT of 1,024 and a 
Hamming window using Syrinx sound analysis 
software (John Burt, www.syrinxpc.comi. We 
selected a maximum of 20 pee-ah-wee and wee- 
ooo songs for each focal male, choosing record¬ 
ings with the highest signal-to-noise ratio. We 
measured or calculated temporal, temporal pro¬ 
portion, frequency, and frequency modulation 
variables for each song type using cursors on 
each spectrogram displayed on a computer screen 
(Tables 1 and 2; Fig. 1). Measurement variable' 
were examined to confirm assumptions of nnr- 
mality and equality of variance were met. 
Spectrograms represent a trade-off between 
precision in temporal and frequency measure' 
We elected to use a single spectrogram for both 
temporal and frequency measures to great!} 
reduce the number of individual measurement' 
required. For example, a single placement m 
cursors on the computer screen can p vldt j 
simultaneous measurements of duration, as rtcl! 
as starting, ending, minimum, and maximum 
frequencies. Any loss of precision could fedi# 
our ability to detect small differences within and 
between individual song variables. However the 
relative tonal clarity and brevity of pee-ah-wee 
and wee-ooo song traces on the specfrog^ 
somewhat minimize the loss of measurement 
precision as compared to acoustic sound' • 3 
cover larger frequency bands and are longer 
We calculated a value, the potential ,or 
information coding’ (PIC), for all song M easure . 
ments to evaluate and compare the within am 
between individual variation of both pee-ah' 
