Table 1.
Examples of types of sound sources of feather-generated sounds.
Figure 1.
Effects of removing tail-feathers on dive sound spectrograms of Allen's Hummingbird.
A: intact. B-E: after experimental removal of R2 – R5. Labeled on the left are frequencies f1, and 2nd (2f1), 3rd (3f1), 4th (4f1), and 5th (5f1) harmonics of f1 (harmonics present in B and C are not labeled). On the right is labeled f2, heterodyne interaction frequencies (f2±f1). Recorded at 48 kHz, presented with a 1024-sample FFT window. See text and Table 3 for more information.
Table 2.
Character states for Selasphorus hummingbirds.
Table 3.
Effects of removal of tail-feathers of Allen's hummingbird on fraction of dives containing dive-sound components. See also figure 1.
Figure 2.
Fundamental frequency of Allen's and Rufus tail feathers as a function of airspeed.
Gray bar indicates speed range of 17 – 26 m s−1, which corresponds to estimated dive speed for Allen's through during the dive [23]. These data show that both Allen's Hummingbird R5 and R4 are sufficient to produce sound f2, Allen's R3 is sufficient to produce sound f1, and Allen's R2 is sufficient to produce the ‘ghost frequency’ (Figure 1). Each feather, except S. sasin R5, was held at a constant orientation. Mode of flutter and sound frequency of some feathers varied as a function of orientation [14], which is why the two S. sasin R4 (gray diamonds) produced different frequencies, and is why the S. sasin R5 (gray squares) are not collinear (because feather orientation varied). Data reproduced from [11].
Figure 3.
Evolution of tail morphology and dive sounds in Selasphorus.
A. Phylogenetic reconstruction of emargination of R2 and which feather(s) are the source, s, of the dive sound. Neighboring feathers, f, are hypothesized to be filters. Branch colors indicate which tail-feathers are sound sources; dashed line indicates taxa that do not dive or produce any sound. R2 has an emarginated in shape (arrow) in the taxa in which it is a source. Fundamental frequency (in kHz) indicated for terminal taxa. Feather drawings are to-scale; photos are not. B: Spectrograms of the dive sound of male Selasphorus Hummingbirds (Hann, 1024-sample FFT window). Fundamental frequency of flutter indicated by ff and arrow. Vertical arrows indicate pulses of sound produced by individual tail-spreads. Vocalizations (v) or wing trills (wt) are also produced during the dive. Photos courtesy Anand Varma; Phylogeny from [29] and McGuire pers comm.
Figure 4.
Changes in sound production mechanism from hypothesized Selasphorus ancestor to Calliope Hummingbird and Allen's Hummingbird.
Right half of the tail is shown. Sound sources labeled s, filters f, and rectrices 2 and 3 are R2 and R3 respectively. The changes are represented in two steps, whereas the underlying trait evolution was likely continuous. C: proposed switch from a ∼1 kHz dominant frequency produced by R2 in the ancestor, to a ∼2 kHz dominant frequency produced by R3 in Allen's Hummingbird. Initially R2 is the source (top), then R3 acts as a spectral filter, amplifying the even harmonics (middle), then R3 becomes the source and the odd harmonics (black arrows) are lost (bottom). The result is the dominant frequency (d) ‘hops’ from 1 kHz to 2 kHz. See text and Figure 3 for more information.