Janes and Ryker • BLACK-THROATED GRAY WARBLER SONG 
341 
£ 8.0 
> 6.0 
c 
« 
t 40 
“■ 2.0 
0.0 
0.5 1.0 1.5 2.0 
Time (sec) 
RG. 2. Phrases and syllables of a sample of Bluck- 
ihroaied Gray Warbler Type I songs. Type I songs of all 
variants include two phrases (A and B). The A-phrase 
consists of repeated syllables comprised of one vibrato note 
and 1-4 tonal notes. The B-phrase consists of up to five 
syllables. The Applegate variant is illustrated which 
includes the most common form of the B-phrase consisting 
of five syllables designated B,_ s . 
hrs PDT. Wc also sampled songs from birds 
recorded beyond the study area as they relate to 
patterns observed within the study area. A varianl is 
defined as a set of songs with two or more shared 
features that were distinct from other songs. 
We defined a note when analyzing songs, as a 
continuous trace on a spectrogram. In cases where 
vibrato (buzzy) and more purely tonal notes join, 
each is considered a separate note. Syllables and 
phrases (Fig. 2) varied. Songs were analyzed from 
spectrograms generated by RAVEN Version 1.2.1 
(Cornell Laboratory of Ornithology, Ithaca, NY. 
USA). AH measures are mean ± SD unless 
otherwise indicated. 
RESULTS 
Song Variants and Dialects .—Type I songs of 
515 Black-lhroated Gray Warblers recorded in 
southwestern Oregon and northern California over 
a period of 10 years exhibited considerable 
geographic variation. Fourteen variants of Type 
I song were identified (Fig. 3). We use the term 
'ariant' to include both dialects and songs that do 
not meet the strict definition of a dialect because 
their distribution either substantially overlapped 
the geographic range of another varianl or 
transitional songs changed rapidly in structure 
across a narrow zone between two variants. We 
identify the variants by an associated geographic 
feature for convenience. 
Five variants met the definition of a dialect: 
Applegate. Days. Humbug, Agncss. and Winch- 
tidc Songs were discrete, and the geographic 
range of each exhibited little or no overlap with 
adjacent variants. Males singing these variants 
were rarely encountered outside the boundaries of 
the mapped area. Only two individuals singing 
Applegate dialect, two singing the Days dialect, 
and one singing the Agncss dialect were encoun¬ 
tered >5 km from the mapped area. 
Variants broadly overlapped in space at other 
locations. This was most obvious among several 
occupying the most restricted areas. Cow and 
Starvout variants each occupied a limited geo¬ 
graphic area, and the distribution of each broadly 
overlapped the other as well as the neighboring 
Grave variant (Fig. I ). Nowhere in their ranges 
were these two variants common. 
The Crooks variant occupied two relatively 
small areas embedded within the area occupied by 
the Caves variant. The Crooks variant occurred in 
the absence of the Caves varianl in only three 
small disjunct watersheds. In addition, a relatively 
large number of males {n = 8) were recorded 
singing the Crooks variant at distances of 14— 
41 km from the mapped area of this variant and 
within other variant areas. 
Other variants failed to meet the strict defini¬ 
tion of a dialect because the structure of the song 
was not discrete. An area of transition between the 
Caves and Grave variants was identified. Individ¬ 
uals delivered songs intermediate in form in 
small, isolated populations across a 15-km wide 
zone. Each variant beyond this transition zone was 
consistent with respect to the unique features ol 
the respective songs. 
Variation in song was observed among indi¬ 
viduals of some variants (Fig. 4). The tonal note 
in Lhe first phrase for about half the individuals 
singing the Caves song, for example, consisted of 
a single note delivered at a nearly constant 
frequency. The tonal note in others varied in 
frequency resulting in two maxima. The tonal 
notes in the songs of still other individuals became 
two distinct notes, each increasing in frequency. 
This variation occurred throughout the range of 
the Caves variant, but the double-tonal note form 
was most common in the southern part of the 
range. A similar pattern of both single and double 
tonal notes in the first phrase was also noted 
within the Agness variant. 
Song Structure .—All Type I songs in the study 
area regardless of variant were composed of two 
phrases (Fig. 2). The A-phruxe consisted of a 
series of repeated multi-note syllables, each 
including one vibrato note and 1—4 tonal notes. 
The bandwidth of the vibrato notes in the Crooks 
and Emigrant variants was narrow in most songs. 
