Yawning may serve both social and nonsocial
functions. When budgerigars (Melopsittacus
undulatus) are briefly held, simulating capture
by a predator, the temporal pattern of yawning
changes. When this species is observed in a
naturalistic setting (undisturbed flock),
yawning and also stretching, a related behavior,
are mildly contagious. On the basis of these
findings, we hypothesized that a stressful event
would be followed by the clustering of these
behaviors in a group of birds, which may be
facilitated both by a standard pattern of
responding to a startling stressor and also
contagion. In this study, we measured yawning
and stretching in 4-bird groups following a
nonspecific stressor (loud white noise) for a
period of 1 hr, determining whether auditory
disturbances alter the timing and frequency of
these behaviors. Our results show that
stretching, and to a lesser degree yawning, were
nonrandomly clumped in time following the
auditory disturbances, indicating that the
temporal clustering is sensitive to, and
enhanced by, environmental stressors while in
small groups. No decrease in yawning such as
found after handling stress was observed
immediately after the loud noise but a similar
increase in yawning 20 min after was observed.
Future research is required to tease apart the
roles of behavioral contagion and a time-setting
effect following a startle in this species. This
research is of interest because of the potential
role that temporal clumping of yawning and
stretching could play in both the collective
detection of, and response to, local
disturbances or predation threats.
Yawning is a ubiquitous vertebrate behavior
that has been hard to characterize functionally,
primarily because there are numerous eliciting
stimuli, including general stress. The
relationship between social stress and yawning
is documented in humans and nonhuman primates
(Greco, Baenninger, & Govern, 1993; Troisi,
Aurelli, Schino, Rinaldi, & De Angelis,
1990) and may occur in other mammals. The effect
of physiological stress on yawning in rodents
was also studied, indicating that when exposed
to footshock, yawning is initially low but then
gradually increases (Moyaho & Valencia,
2002). Related to hypotheses generated by these
studies, much comparative research supports the
view that yawning is involved in the maintenance
of arousal (for review, see Baenninger, 1997)
and is also related to changes in state or
activity (Provine, 1996, 2005). Although studies
have failed to identify yawn-associated
increases in cortical arousal as measured by EEG
(reviewed by Guggisberg, Mathis, Schnider, &
Hess, 2010), recent research shows that distinct
brain temperature reductions occur following
yawning in rats (Shoup-Knox, Gallup, Gallup,
& Mc- Nay, 2010), and this cooling effect
has been hypothesized to promote arousal. In
addition, research on humans, chimpanzees, and
rats shows that yawning is associated with
behavioral arousal, as measured by modified
activity or increased locomotion (Baenninger,
Binkley, & Baenninger, 1996; Gallup, Miller,
& Clark, 2011; Giganti, Hayes, Akilesh,
& Salzarulo, 2002; Vick & Paukner,
2009). Therefore, it seems likely that the
association between stress and yawning is
connected to a general state of arousal and
activity.
To date, only in budgerigars (Melopsittacus
undulatus) is there experimental evidence that
yawning is related to stress in a nonmammalian
vertebrate (Miller, Gallup, Vogel, & Clark,
2010). In that study, handling of a bird
simulated capture by a predator, and, after
release, behavioral responses were measured over
three 20- min time blocks (Miller et al., 2010).
In comparison to control periods, yawning was
delayed and infrequent in the first 20 min
following release but then significantly
increased in frequency during the next 20
min&emdash;a temporal pattern similar to that
described in rodents (see Moyaho & Valencia,
2002). A follow-up experiment showed that a
bird's underwing temperature after handling was
negatively related to its latency to yawn
(accounting for over 38% of variance in this
response). These results suggest that
stress-induced yawning may be associated with
hyperthermia. At the same time, the delay in
yawning may be an adaptive suppression when a
predator is still near. Thus, observed patterns
of yawning may represent compromises between
conflicting adaptive responses, namely a need to
thermoregulate versus a need to freeze in face
of a potential threat.
Recent observational research provides
evidence that both yawning and stretching are
socially contagious in these birds (Miller,
Gallup, Vogel, Vicario, & Clark, in
preparation), showing tight temporal clustering
and even matching of these behaviors in
unmanipulated flocks ( 20 individuals). It is
hypothesized that contagion may function to
coordinate arousal and collective movement
within groups, and therefore we expect it to be
enhanced by the presence of an environmental
threat or disturbance.
Discussion Our findings show that stretches,
and yawns to a lesser degree, were nonrandomly
clumped in time following the auditory
disturbances, indicating that the degree of
temporal clustering is sensitive to
environmental stimuli. Two potential
explanations emerge from the increased temporal
clustering following a loud noise. First, the
behaviors themselves may have become more
contagious among flockmates following the
auditory stressor. Alternatively, clumping may
have been a result of individuals having a
similar behavioral response to this stimulus,
thus promoting synchronization that was not
socially influenced. Given the interaction
between time and group, the latter explanation
is unlikely, because if nonsocial
synchronization was responsible for all results,
one would expect the time courses of all groups
to be similar. Furthermore, the lack of any
difference in latency measures between
conditions and the extension of temporal
clustering into the later two time blocks
provides support that this pattern resulted
largely from behavioral contagion, and not
result of an initial, nonsocial
coordination.
We did not find any temporal clumping under
control conditions like that documented in our
observational study of larger flocks ( 20
individuals; Miller et al., in review), but this
could be due to a number of factors including
differences in group size and separation,
disruptions in rest-activity cycle associated
with experimental testing, and the limited time
frame in which recordings occurred. Although
observational studies of contagious yawning have
also been presented in other species, such as
gelada baboons (Theropithecus gelada; Palagi,
Leone, Mancini, & Ferrari, 2009),
unequivocal experimental evidence of this effect
is limited to humans (Platek, Critton, Myers,
& Gallup, 2003; Provine, 1989) and
chimpanzees (Pan troglodytes; Anderson, Myowa-
Yamakoshi, & Matsuzawa, 2004; Campbell,
Carter, Proctor, Eisenberg, & de Waal,
2009). Further research is needed to determine
whether yawning and stretching can be
contagiously triggered among budgerigars in a
controlled experiment, as well as identify the
various factors associated with the temporal
clustering of these behaviors in this and other
species.
The proximate function of stretching is not
well understood, but, like yawning, is probably
a homeostatic behavior, possibly involved in
behavioral arousal. In fact, when mildly
disturbed (e.g., person entering a room without
close approach) or motivated by external events
including food, budgerigars commonly stretch and
then initiate flight (A. B. Clark, A. C. Gallup,
M. L. Miller, & A. R. Vogel, personal
observation), suggesting that stretching is
involved in preparation for movement. This makes
it a possible signal of activity, and similar to
yawning, the temporal clustering of stretching
could coordinate changes in a group's activity
level or state.
When comparing the results from this study
to those of Miller et al. (2010), one obvious
difference is that the latency to yawn in this
study did not differ between white noise and
control conditions. This suggests that, unlike
handling restraint, the auditory disturbance did
not suppress yawning. Because handling stress
simulates an encounter with a predator, release
is essentially "escape" and the subsequent
inhibition in yawning may be an adaptive
response to decrease detection by remaining very
still. The white noise disturbance was
presumably a nonspecific stressor that was less
threatening, producing mild vigilance without
prolonged stillness. In fact, there was no
difference in the latency to move after the
white noise compared to the control trials:
noise, 185 89 s; control, 106 86 s, F(1, 12)
3.653, p .080. In contrast, we did observe
an increase in yawning frequency 20 min
following the stressor, which is comparable to
that shown with handling (Miller et al.,
2010).
Taken together with the previous report on
handling stress (Miller et al., 2010), these
findings suggest that yawning is adaptively
suppressed or elicited, both socially or
nonsocially, depending on factors such as the
immediacy of the threat and the degree of
physiological stress experienced, although more
precise measures of stress response are needed
for confirmation. Furthermore, the tight social
clumping of stretching after an auditory
disturbance may be a collective response
preparing the group for flight in the event that
a more salient threat was presented. Because
flight synchronization is believed to function
in antipredator avoidance (Monus & Barta,
2011), we propose that future studies of group
vigilance should consider yawning and stretching
as potential variables of interest. Consistent
with the view that contagious yawning evolved to
coordinate arousal, which in turn may improve
vigilance within the group (Gallup & Gallup,
2007), the temporal clumping of yawning and
stretching witnessed after the auditory
disturbance may serve to enhance both the
collective detection of, and response to, local
disturbances or predation threats.
