Avellis • TAIL-PUMPING BEHAVIOR OF THE BLACK PHOEBE 
767 
of tree canopy. ‘Paradise Falls’ is a riparian 
habitat consisting of dense foliage and a high 
percent of tree cover. Observations were made on 
3 days between 0700 and 1500 hrs PST during 
February and April 2010. All observations were 
performed on individuals at one study site, each 
site was the focus of 1 day. and each site was 
observed only once. Three individuals were 
observed within both 'Lizard Rock’ and ‘Paradise 
Falls' and four individuals were observed at ‘Oak 
Grove'. Observations were the method used 
because birds were not banded or otherwise 
individually marked; sample size was kept small 
to avoid pseudoreplication. 
Tail pumping rates were calculated by dividing 
the number of tail movements by the duration of 
the observation period. A tail movement consisted 
of a downward flick followed by an upward flick. 
Recordings were made of six House Finches 
i Curpodaeus mexicanus) and six Black Phoebcs 
within the study location with an Olympus LS-10 
PCM recorder and Olympus ME-30 omni-direc¬ 
tional microphone in the .mp3 format at 320 kb/ 
sec and 44.1 kHz. Playback experiments consisted 
of recordings from individuals separated at least 
300 m from a focal individual to eliminate bias 
that may result from neighbor familiarity. A 
prerecorded Cooper’s Hawk (Accipiter cooperii) 
vocalization (Keller 2003) was used for anti¬ 
predator response experiments because no poten¬ 
tial predator vocalized in the field. This may 
represent a source of pseudoreplication, but 
responses to this stimulus were observed from 
several different birds from three different sites. 
All recordings were edited with Raven Lite 
(Charif et al. 2006) to ensure a uniform (3 min) 
Playback time and to eliminate background noise 
and silence. Playback experiments were per¬ 
formed using an iPod and iHome speakers hidden 
front sight and controlled with a remote. Playlists 
°f House Finches and Black Phoebes were played 
with the random order function of the iPod 
elected. Focal individuals were males that had 
i*ren observed singing (i.e.. territorial). 
Balance Hypothesis .—Tail pumping rates were 
compiled from trial data that included observation 
Periods used for the foraging enhancement 
hypothesis plus observations made during pre- 
Piayback trials for the two experimental tests. 
Seventy observations among 10 individuals (mean 
- SE, 7.1 ± 0.89 observations/individual) were 
compared between types ol substrate. A stable 
substrate consisted of those judged to support the 
weight of an individual Black Phoebe (e.g., 
ground, rocks, fences, etc.). Three 5-g fishing 
line weights were tied to the substrate when 
stability could not be judged and, if any 
displacement resulted, the substrate was catego¬ 
rized as ‘unstable’. This method was used to 
mimic displacement of perches by the mass of an 
individual bird. If a questionable substrate was out 
of the reach of the observer, it was eliminated 
from analysis. 
Fora ging Enhancement Hypothesis .—Tail pump¬ 
ing rates were calculated by counting tail move¬ 
ments of 10 individuals in 50 observations (mean ± 
SE, 5.0 ± 0.86 observations/individual) for up to 
10 min each. The trial was discontinued if an 
individual left the view of the observer. Whether or 
not an individual foraged during this trial was noted 
and foraging outcome of those that did forage was 
also recorded. 
Signal to Territorial Intruders Hypothesis .—Tail 
pumping rates were calculated by observing tail 
movements of 10 individuals in each of four trials 
consisting of 3 min each and separated by 30 sec. 
The ‘pre-playback’ trial consisted of no manipula¬ 
tion. the ‘control’ trial consisted of recorded House 
Finch vocalizations being played to a focal 
individual Black Phoebe, the ‘intruder’ trial 
consisted of a recorded Black Phoebe being played 
to a focal individual, and the ‘post-pi ay back’ trial 
consisted of the ‘intruder vocalization being 
turned off. Whether or not a focal individual called 
or approached the sound source for each trial was 
also noted. An approach was defined as an 
individual moving within 2 m of the speakers and 
approach proportions were calculated by dividing 
the number of approaches by the number of 
observations for each trial. 
Signal to Predators Hypothesis .—Tail pumping 
rates were calculated by observing tail movements 
of 10 individuals in each of four trials consisting 
of 3 min each and separated by l min. The pre- 
playback’ trial consisted of no manipulation, the 
■control* consisted of recorded House Finch 
vocalizations being played to a local individual 
phoebe. the ‘predator’ trial consisted of a recorded 
Cooper’s Hawk being played to a focal individual, 
and the ‘post-playback’ trial consisted of the 
predator sound being turned off. Whether or not a 
focal individual called and/or approached the 
sound source was noted for each trial, and 
proportions were calculated by dividing the 
number of approaches and calls by the number 
of observations for each trial. An approach was 
