Home Search Collections Journals About Contact us My IOPscience

Hovering and intermittent flight in birds

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2010 Bioinspir. Biomim. 5 045004

(http://iopscience.iop.org/1748-3190/5/4/045004)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 131.104.62.10
This content was downloaded on 15/11/2014 at 23:46

Please note that terms and conditions apply.

IOP PUBLISHING BIOINSPIRATION & BIOMIMETICS
Bioinsp. Biomim. 5 (2010) 045004 (10pp) doi:10.1088/1748-3182/5/4/045004

Hovering and intermittent flight in birds

Bret W Tobalske
Field Research Station at Fort Missoula, Division of Biological Sciences, University of Montana,
Missoula, MT 59812, USA
E-mail: bret.tobalske@mso.umt.edu

Received 8 March 2010

Accepted for publication 28 June 2010
Published 24 November 2010
Online at stacks.iop.org/BB/5/045004

Abstract
Two styles of bird locomotion, hovering and intermittent flight, have great potential to inform
future development of autonomous flying vehicles. Hummingbirds are the smallest flying
vertebrates, and they are the only birds that can sustain hovering. Their ability to hover is due
to their small size, high wingbeat frequency, relatively large margin of mass-specific power
available for flight and a suite of anatomical features that include proportionally massive major
flight muscles (pectoralis and supracoracoideus) and wing anatomy that enables them to leave
their wings extended yet turned over (supinated) during upstroke so that they can generate lift
to support their weight. Hummingbirds generate three times more lift during downstroke
compared with upstroke, with the disparity due to wing twist during upstroke. Much like
insects, hummingbirds exploit unsteady mechanisms during hovering including delayed stall
during wing translation that is manifest as a leading-edge vortex (LEV) on the wing and
rotational circulation at the end of each half stroke. Intermittent flight is common in small- and
medium-sized birds and consists of pauses during which the wings are flexed (bound) or
extended (glide). Flap-bounding appears to be an energy-saving style when flying relatively
fast, with the production of lift by the body and tail critical to this saving. Flap-gliding is
thought to be less costly than continuous flapping during flight at most speeds. Some species
are known to shift from flap-gliding at slow speeds to flap-bounding at fast speeds, but there is
an upper size limit for the ability to bound (∼0.3 kg) and small birds with rounded wings do
not use intermittent glides.
(Some figures in this article are in colour only in the electronic version)

1. Introduction (Taeniopygia guttata) can hover for ∼30 s (Norberg 1975,

Although there is a growing tradition of insects being used
as inspiration for the design of micro-air vehicles (Ellington
1999, Madangopal et al 2005, Zufferey 2008), small- and
medium-sized birds exhibit flight styles that will likely prove
useful as models for furthering the development of flying
robots. The smallest birds, hummingbirds (2–20 g in body
mass), have converged with insects upon the ability to sustain
hovering flight (Stolpe and Zimmer 1939, Greenewalt 1962,
Weis-Fogh 1972, Wells 1993, Tobalske et al 2007, Clark and
Dudley 2009), yet they are also capable of cruising flight at
speeds up to 12 m s−1 and migrating long distances (Tobalske
et al 2007, Clark and Dudley 2009, Robinson et al 1996).
Other birds may hover for very brief intervals, presumably
using anaerobic metabolism. For example, small passerines
such as pied flycatcher (Ficdula hypoleuca) and zebra finch
Tobalske et al 1999). In the first part of this paper, I explore
the unique abilities, anatomy, physiology and aerodynamics
of hummingbirds.
Unlike hummingbirds that tend to flap their wings
continuously, the vast majority of other small and medium
sized birds use forms of intermittent flight during which they
regularly interrupt flapping phases to hold their wings either in
a flexed-wing ‘bound’ posture during which the wings are held
tightly against the body or in an extended-wing ‘glide’ (Rayner
1985, Tobalske 2001). Mathematical models of mechanical
power output that are based on aerodynamic theory suggest that
these flight styles offer savings in mechanical power relative to
the cost of flapping continuously (Rayner 1977, 1985, Rayner
et al 2001, DeJong 1983, Ward-Smith 1984a, 1984b). In the
second part of this paper, I describe how intermittent flight
varies with body size, wing design and flight speed.

1748-3182/10/045004+10$30.00 1 © 2010 IOP Publishing Ltd Printed in the UK

Bioinsp. Biomim. 5 (2010) 045004 B W Tobalske

(A)

(C)

(B)

(D) (E )

Figure 1. Wing kinematics during hovering in the rufous hummingbird (Selasphorus rufus). (A) Dorsal view of posture at mid-downstroke
and traces of the paths of wingtips (filled circles) and wrists (open circles) as obtained from 500 frames s−1 video. (B) Lateral view of
posture at upstroke–downstroke transition and traces of the paths of wingtips (filled circles) and wrists (open circles) that are synchronous
with paths in (A). (C) Chord angle relative to the frontal plane of the body during one wingbeat cycle. The shaded region indicates
downstroke. The dashed line represents wrist elevation relative to frontal plane. Values are means ± SD for five birds. (D) Mid-upstroke
wing posture showing wing supination and twist along the long axis. The mid-wing is illuminated by laser light used for particle-image
velocimetry. (E) Mid-downstroke wing posture revealing different wing camber and chord angle compared with upstroke. Laser light
illuminates the middle of the wing as in (D). From Warrick et al (2005) and Tobalske et al (2007).

2. Hovering Hummingbirds are unique among birds in their ability

2.1. Kinematics
Wingbeat frequency scales negatively with increasing body
mass in hummingbirds, ranging from 8 Hz in the 20 g giant
hummingbird (Patagonia gigas) to 80 Hz in the 2 g Amethyst
woodstar (Calliphlox amethystina, Greenewalt 1962). During
◦
hovering near sea level, wingbeat amplitude is ∼110 (Stolpe
and Zimmer 1939, Tobalske et al 2007, figures 1(A), (B)),
and hummingbirds increase their amplitude to accommodate
greater power demands during load lifting or with decreased
air density at higher altitude (Altshuler and Dudley 2002,
2003, Altshuler et al 2004, 2010). Although wingbeat
frequency does not vary significantly across flight speeds
(Tobalske et al 2007, Clark and Dudley 2010), it does increase
during transitory load-lifting when the flight muscles are used
anaerobically (Altshuler and Dudley 2003).
to almost fully supinate their wings during upstroke so that
the underside of the wing faces upward (Stolpe and Zimmer
1939, Tobalske et al 2007, figures 1(C) and (D)). This reverses
the pronated posture typical of all birds during downstroke
(figure 1(E)), but the postures are not mirror images in part
because upstroke features greater long-axis twist of the wing.
◦
Long axis rotation of the wing through the wingbeat is ∼140
and occurs symmetrically with wing turnaround at the end of
each half stroke. The wings are held fully extended during
the entire wingbeat cycle, and the wingtips trace a figure-
eight pattern as projected in the lateral view. Downstroke
features higher angular velocity and higher chord angle than
upstroke (Tobalske et al 2007). In contrast with the wingbeat
pattern in hummingbirds, when flying slowly, diverse species
with pointed wings exhibit hand-wing supination (tip-reversal)
and a partially flexed wing during upstroke and birds with
rounded wings fully flex them against the body during upstroke

2
Bioinsp. Biomim. 5 (2010) 045004 B W Tobalske

(Tobalske 2000). Swifts are sister taxa to the hummingbirds in

(A) 4000
the Apodiformes (Karhu 1992, Mayr 2003). Although swifts
also fly with extended wings, they do not accomplish wing
supination, and they are severely limited in their ability to fly
slowly (Warrick 1998).
Vorticity (s-1)
2.2. Aerodynamics
Hummingbirds appear to have converged upon the hovering
style of insects (Warrick et al 2005, 2009). Hummingbirds 1 cm
are the only birds that clearly demonstrate an aerodynamically -4000
active upstroke during hovering in still air (Warrick et al 2005, 2000
2009, Altshuler et al 2009). There is debate about whether the (B )

tip-reversal upstroke of other bird species is aerodynamically

active, but the evidence to date is equivocal (Tobalske 2000).
In steady hovering in a hummingbird, downstroke supports
∼75% of the body weight and upstroke supports ∼25%
(Warrick et al 2005, 2009).
In addition to having an aerodynamically active upstroke,
hummingbirds exploit two unsteady aerodynamic mechanisms
during hovering that are considered to be the key elements of Vorticity (s-1)
insect flight (Dickinson et al 1999, Lehmann 2004). These
are leading-edge vortices (LEVs) and rotational circulation.
During wing sweep in the middle of both downstroke and
upstroke, hummingbirds exhibit LEVs on the uppermost
surface of their wings (figure 2(A)). LEVs are interpreted
to represent delayed stall at high angles of attack, and they
enhance coefficients of lift and drag during wing sweep relative
to coefficients during linear translation at the same angles
-2000
of attack (Lehmann 2004). The LEV contributes ∼16% to
the total bound circulation of the hummingbird wing, and Figure 2. Velocity (vectors) and vorticity fields (background)
measurable LEVs are not always present. In contrast, LEVs illustrating near-field flow and bound circulation on the wing of a
rufous hummingbird (Selasphorus rufus) during hovering. (A) LEV
dominate fruit-fly (Drosophila melanogaster) aerodynamics
is apparent on the upper surface of the wing at mid-downstroke. The
(Birch et al 2004) and contribute up to 40% of the circulation dashed line indicates the outline of the bird’s body. The dark area
on the wing in slow-flying, nectar-feeding bats that are among behind the bird is a shadow due to the wing interrupting the laser
the smallest extant species of bats (Muijres et al 2008). light used to illuminate micron-sized particles of oil in the air.
Long-axis rotation of the wing prolongs circulation on the (B) During long-axis rotation (pronation) of the wing at the
wing from the previous half stroke through wing turnaround in transition between end of upstroke and beginning of downstroke,
circulation is maintained on the wing. Vorticity in the lower left is a
the same manner that has been shown for insects using robotic drag-induced wake (Von Kármán vortex street) created during
models (Dickinson et al 1999, Lehmann 2004, Warrick et al upstroke (Warrick et al 2009).
2009, figure 2(B)). Thus, at the end of each half stroke, the
wing is functioning as a Flettner rotor (Vogel 1994). At the
from the flight muscles. Metabolic rate in 3.5 g hummingbirds
end of long-axis rotation, when sweep begins for a new half
during hovering is ∼110 W kg−1 and efficiency in converting
stroke, circulation reverses. Circulation then increases during
metabolic energy into mechanical work is estimated to be
the rest of the half stroke (Warrick et al 2009).
∼10% (Wells 1993). To meet this high, sustained power
Vortex shedding from the wings appears to give rise to
demand, the primary muscles for downstroke (pectoralis) and
a wake pattern that consists of tilted rings with one ring
associated with each half stroke (Warrick et al 2009). It is upstroke (supracoracoideus) are proportionally larger than
further hypothesized that a separate ring structure is shed per in other species. The paired pectoralis muscles make up
wing, and that the tail functions to deflect the flow and aid in ∼25% of the body mass in hummingbirds and ∼15% of
pitch stability (Altshuler et al 2009). the body mass in other species (Greenewalt 1962, Altshuler
and Dudley 2002). The supracoracoideus is ∼50% of the
pectoralis mass in hummingbirds and ∼20% in other species
2.3. Morphology and physiology
(Greenewalt 1962, Wells 1993). Although these muscles
Hummingbird wings have a musculoskeletal architecture and have the same fast-oxidative glycolytic fiber types as in other
feather morphology that appears to be adapted for hovering. small birds (Welch and Altshuler 2009), several features help
Hovering in still air is a particularly demanding flight style for accommodate the high oxygen flux (40 ml O2 g−1 h−1 ) (Berger
an animal because it is inducing a large downward velocity into 1985, Bartholomew and Lighton 1986, Clark and Dudley
the air to support its weight, and this requires high power output 2009). These include a small fiber size, high mitochondrial

3
Bioinsp. Biomim. 5 (2010) 045004
Calliope hummingbird
Rock dove
Figure 3. Wing skeletons in the calliope hummingbird (Stellula
calliope) and rock dove (Columba livia) scaled so that the
handwings are of the same length (adapted from Dial (1992), used
with permission from The Auk).
density, and high ratio of the capillary surface area to the fiber
surface area (Suarez et al 1991, Mathieu-Costello et al 1992).
Hummingbirds rapidly shift their metabolism from using fatty
acids during fasting to exclusive use of ingested carbohydrates
after the fasting bout ends, and they require less oxygen when
using carbohydrates as a fuel source (Welch and Suarez 2007,
Welch et al 2007).
Several skeletal features are hypothesized to be associated
with a general commitment to aerial locomotion in swifts
(Apodidae) and hummingbirds (Trochilidae), and comparison
between the taxa highlights features that are hypothesized to
be adaptations for hovering. One shared feature of swifts
and hummingbirds is an emphasis upon the forelimb that has
been associated with a reduction in the hindlimb size. Force
from the hindlimbs contributes only 50% of the total take-
off velocity in hummingbirds (Tobalske et al 2004), whereas
the hindlimb contribution is greater than 80% in other birds
(Earls 2000). Swifts and hummingbirds have a proportionally
short humerus with a humeral head that is oriented more
orthogonally to the long-axis of the humerus compared with
near relatives such as the crested swifts (Hemiprocnidae,
Karhu (1992), Mayr (2003) or more-distantly related species
such as the rock dove (Columba livia, figure 3). These features
permit an adducted wing posture and a proximal shift in the
concentration of muscle mass for the wing muscles, both of
which should reduce the moment of inertia and inertial work
requirements of the wing (Stolpe and Zimmer 1939, Karhu
1992, Mayr 2003). The moment of inertia of a single wing of
a 4 g hummingbird is ∼2 × 10−8 kg m2 (Wells 1993).
In contrast, hummingbirds have several characteristics
that are different from swifts and, therefore, are thought to
permit their enhanced wing supination for hovering upstroke.
They have a medio-anterior condyle on their humerus (Stolpe
and Zimmer 1939, Karhu 1992, Mayr 2003), and the articular
surfaces at the elbow; wrist and distal wing elements are either
openly rounded or flat, which may permit greater rotation
about the joints (Stolpe and Zimmer 1939, Mayr 2003) than
in other species. Anatomical work suggests that grooves and
ridges in the forearm and wrist bones of other birds restrict
their ability to supinate their distal (hand) wing unless their
wrist is flexed (Vasquez 1992).
Among hummingbirds, wing length and pectoralis mass
scale allometrically (Greenewalt 1962, Altshuler et al 2010).
4
B W Tobalske
Wing length and tail length scale proportional to mass to the
0.5 power (∝ M0.5 ) instead of scaling ∝ M0.33 , which would be
consistent with geometric similarity (Greenewalt 1962, Clark
2010). Pectoralis mass scales ∝ M1.2 where M1.0 would be
expected for isometry (Altshuler et al 2010). These scaling
patterns may be morphological responses to partially offset the
adverse scaling of induced power requirements with increasing
body mass that would otherwise occur in geometrically similar
organisms (Ellington 1991). Populations living at higher
altitudes may also compensate for low air density by having
relatively longer wings (Altshuler et al 2004). Although
these trends may partially compensate for induced power
requirements, new evidence suggests that the marginal mass-
specific power available from the flight muscles inevitably
scales negatively in a broad sample of hummingbird species
(Altshuler et al 2010). This pattern is driven, in part,
by wingbeat frequency declining as wing length increases
(Greenewalt 1962, Altshuler et al 2010). Similar declines in
mass-specific power available for flight are thought to explain
variation in flight performance among other groups of birds
(Pennycuick 1975, Tobalske 1996, 2000).
It is expected from aerodynamic models that the
mechanical power required for flight varies according to a
U-shaped curve as a function of the flight speed, with the cost
of producing lift greatest during hovering and declining with
speed whereas the cost of overcoming drag on the wings and
body increases exponentially with increasing speed (Rayner
1979a, 1979b, Pennycuick 1975, Ellington 1991, Bundle et al
2007). If the mechanochemical efficiency of hummingbird
muscle, estimated to be ∼10% (Wells 1993), is constant,
the metabolic power curve should exhibit the same upwardly
concave shape (Ellington 1991). However, direct measures
of oxygen consumption during flight in hummingbirds have
produced a variety of curves with mass-specific metabolic
rates during flight varying from 30–70 ml O2 g−1 h−1 (Berger
1985, Clark and Dudley 2009, 2010). Part of the challenge
that explains why a single power curve has not emerged to
describe hummingbird flight costs is that there is significant
variation in the details of the curve shape among birds within
a species as well as among trials on different days within the
same bird (Clark and Dudley 2010).
One study in which hummingbirds were fitted with a
gas mask indicated that there was no significant increase in
the metabolic power between hovering and forward flight up
to speeds of 7 m s−1 (Berger 1985, Ellington 1991). This
indicates that hummingbirds are uniquely efficient at hovering
such that costs vary according to a ‘J-shaped’ curve with flight
speed. In contrast, recent measurements using a stationary
feeder reveal a U-shaped curve with the greatest metabolic
rates exhibited during hovering and fast forward flight (Clark
and Dudley 2009, 2010). Different methods may contribute
to these different results, as parasite drag on the body may
be reduced in the wake of a stationary feeder and augmented
when a bird wears a respiratory mask. Additionally, Clark and
Dudley (2010) suggest that low wing loading will contribute
to a flatter curve shape at lower speeds because of a reduced
power cost for inducing velocity into the air.
Intermittent flight (see below) is probably a method for
reducing the power requirements of forward flight (Rayner
Bioinsp. Biomim. 5 (2010) 045004
1985), but the hummingbird species whose wing kinematics
have been measured over a wide range of speeds seldom or
never exhibit intermittent pauses in flapping while engaged in
average steady flight (Tobalske et al 2007, Clark and Dudley
2009, 2010). The reasons for this are not known but could
be due to the anatomy of the hummingbird wing (figure 3),
the relative benefits of elastic energy storage in connective
tissue (Wells 1993, Tobalske and Biewener 2008) or offsets
in the timing of neural activation of muscle relative to the
onset of shortening to permit force deactivation within a
rapid wingbeat (Tobalske et al 2010). Intermittent pauses
in flapping are exhibited during courtship display dives in
Anna’s hummingbirds (Calypte anna, Clark 2009), and much
remains to be studied regarding locomotion among the diverse
hummingbirds in Central and South America.
3. Intermittent flight
3.1. Kinematics and aerodynamics
Intermittent flight consists of brief pauses in flapping during
which a bird may produce lift to contribute to weight support
but not thrust, so that it inevitably trades potential energy for
kinetic energy until it begins flapping and producing thrust
again (figure 4(A)). This process gives rise to an average level
flight path that undulates in altitude. Variations of intermittent
flight may occur during landing. For example, woodpeckers
generally swoop toward a tree trunk where they are preparing
to perch using glide-bounding (Tobalske 1996). The wing
posture during intermittent pauses may vary between being a
fully flexed bound, identical to the perching posture, to a glide
with wings fully extended. Various species use both of these
postures as well as intermediate postures (partial bounds or
flexed-wing glides, Tobalske and Dial 1994, Tobalske 1995,
1996, Bruderer et al 2001, Lentink et al 2007, figure 4(B)).
Birds enter into intermittent non-flapping phases during
mid-upstroke and return to flapping by completing their
upstroke (Tobalske and Dial 1994, Tobalske 1995, 1996,
Tobalske et al 1999, figure 5). Generally the wings are
held motionless relative to the body, but exceptions occur.
In cockatiels (Nymphicus hollandicus), in cruising flight from
9 to 15 m s−1 , the wings may not pause completely; rather
the vertical motion of the wings markedly slows during mid-
upstroke (Hedrick et al 2002). Starlings regularly transition
between bounds and glides within the same non-flapping phase
by extending the wings horizontally (abducting them) at the
end of the bound (Tobalske 1995).
Wingbeat frequency and amplitude co-vary during
the flapping phases of intermittent flight. Instances of
peak horizontal and vertical acceleration coincide with
high-frequency, high-amplitude wingbeats (Tobalske 1995).
These kinematics suggest that lift production is enhanced
during these instances, consistent with an expectation that
aerodynamic forces are proportional to the square of the
angular velocity of the wing (Vogel 1994, Usherwood and
Ellington 2002). The flight style of the black-billed magpie
(Pica hudsonia) appears unusual in terms of the magnitude of
variation in frequency and amplitude during flapping phases
5
B W Tobalske
(A )
Flap-bounding
Flap-gliding
(B )
100
Non-flapping intervals (%)
80
60
40
20
0
7 9 11 13 16
Flight speed (m s-1)
Figure 4. Intermittent flight in birds consists of flapping phases
interspersed with non-flapping phases, and both forms give rise to
an undulating flight path. (A) During intermittent bounds, the wings
are held flexed against the body and during intermittent glides, the
wings are outstretched. (B) Wing posture during non-flapping
phases as a function of flight speed in a budgerigar (Melopsittacus
undulatus). The bird outlines indicate bounds (white), partial
bounds (gray) and glides (black, Tobalske and Dial 1994).
(Tobalske et al 1997). Rayner (1985) suggests this may
represent a form of chattering flight, similar to what can be
performed in a propeller-driven aircraft, wherein modulating
revolutions per minute (rpm) of the propeller and power output
can be used to save fuel.
The duration of intermittent pauses is variable. In the case
of bounds, pauses are generally <200 ms in duration (Tobalske
1996, Tobalske et al 1999). Intermittent glides may be as brief
as a bound but may also be well over 10 s in duration in larger
gliding birds (e.g. turkey vulture, Cathartes aura, unpublished
observation).
The wings produce lift during intermittent glides and the
magnitude of this lift varies directly with the wingspan and
area (Lentink et al 2007). Even with the wings completely
flexed during a bound, the body and tail can support up to 20%
of the body weight (Csicsáky 1977, Tobalske et al 1999, 2009).
Measurements of body acceleration and wake dynamics in live
birds as well as force measurements on prepared specimens
all support this conclusion. A flap-bounding bird sheds a pair
of trailing vortices into its wake, revealed using particle image
velocimetry (figure 6). The downward-induced velocity in the
middle of this wake indicates lift production, and this pattern
is present even when the tail is removed (Tobalske et al 2009).
Lift:Drag (L:D) ratio during bounds varies with tail length,
Bioinsp. Biomim. 5 (2010) 045004 B W Tobalske

Flapping Bound Flapping Bound Flapping

8

Altitude (cm)
4
0
-4

-8

elevation (cm)
8
Wingtip
4
0
-4
-8
Muscle length

1.2
(L / Lrest)

1.1

1.0

0.9

1.5
EMG (mV)
Pectoralis

-1.5
0.0 0.25 0.50 0.75
Time (s)
Figure 5. Kinematics and muscle activity during zebra finch (Taeniopygia guttata) flight at 12 m s−1 in a wind tunnel. Two bounds and
13 wingbeats are included in this interval of flight. Altitude is for the estimated center of mass relative to average altitude during the flight
interval. Muscle activity is measured using electromyography and muscle length is measured using sonomicrometry (Tobalske et al 2005).

reaching greater values when the tail is artificially extended.
Presumably other design criteria such as the need to minimize
drag or avoid damage to the tail function to limit tail length in
flap-bounding birds.
3.2. Muscle activity
In all species studied to date, the pectoralis is inactive during
bounds and exhibits an isometric contraction during glides
(Meyers 1993, Tobalske 1995, 2001, Tobalske et al 1999,
figure 5). Sonomicrometry reveals that the pectoralis does
not change length during intermittent pauses (Tobalske et al
2005, Askew and Ellerby 2007). Neuromuscular activity is
limited to the cranial portion of the pectoralis during glides
in American kestrels (Falco sparverius, Meyers 1993). Some
larger birds that glide and soar extensively have a deep belly
of their pectoralis that consists of slow muscle fibers that
are hypothesized to be particularly economical for sustaining
isometric contractions during gliding (Rosser and George
1986, Rosser et al 1994). In contrast, the main belly of the
pectoralis consists of fast fibers in all flying birds (Rosser
and George 1986), and, therefore, fast fibers appear adequate
for intermittent gliding (Tobalske and Dial 1994, Meyers and
Mathias 1997).
The use of the supracoracoideus during intermittent
pauses is more variable. In budgerigars and European
starlings, the supracoracoideus is inactive during bounds,
whereas it is active during bounds in black-billed magpies,
and it appears to be engaged in isometric contraction during
glides in various species (Meyers 1993, Tobalske and Dial
1994, Tobalske 1995, 2001). The isometric contraction of this
upstroke muscle during glides indicates its role in providing
stability at the shoulder joint (Tobalske and Biewener 2008).
Regardless of the muscle being considered, the amplitude
of electromyographic signals is always significantly less
during non-flapping phases compared with during flapping
(Meyers 1993, Tobalske and Dial 1994, Tobalske 1995, 2001).
This indicates a reduction in motor unit recruitment, and is
consistent with a reduction in metabolic cost during fixed-
wing phases compared with flapping (Baudinette and Schmidt-
Nielsen 1974). Because metabolic rates require time to
decrease following intense activity (Baker and Gleeson 1999,
Borsheim and Bahr 2003), it is likely that any metabolic
savings during brief pauses is less than the reduction in cost
associated with prolonged glides.
Formerly, it was hypothesized that flap-bounding flight
was a mechanism for varying muscle work and power
output without having to vary force or contractile velocity
in the major downstroke muscle, the pectoralis (Rayner 1977,
1985). This hypothesis might be particularly appealing for
the design of autonomous vehicles, but it does not appear
to be correct for birds. Direct measures of muscle contractile
activity during flight may be obtained using electromyography
and sonomicrometry. Such experiments have revealed that
pectoralis motor-unit recruitment and contractile velocity vary
significantly among flight speeds and flight modes in flap-
bounding birds (Tobalske et al 2005, Ellerby and Askew 2007).
Moreover, the variation in contractile velocity exhibited by
the flap-bounding zebra finch exceeds that exhibited by larger
species that do not use flap-bounding flight (Tobalske et al
2005).
3.3. Predictions of energy saving from mathematical models
Flapping flight has the highest power requirement of any
form of vertebrate locomotion (Schmidt-Nielsen 1972), and

6
Bioinsp. Biomim. 5 (2010) 045004
(A)
Z
Y
X
(B)
= 5 m s-1
Z (∝ M−0.46 ).
1 cm
Y Fiber type in the pectoralis of woodpeckers changes as a
Figure 6. Body and tail lift during intermittent bounds in a zebra
finch (Taeniopygia guttata). (A) With the wings fully flexed, the
body and tail generate a pair of trailing vortices that counter-rotate.
The induced downwash between the vortices represents the reaction
of the air to lift production. (B) Caudal view of the wake of a
zebra-finch specimen mounted on a force transducer with velocity
calculated using particle image velocimetry (Tobalske et al 2009).
intermittent flight styles may reduce this power requirement.
Mathematical models developed from aerodynamic theory
indicate that flap-bounding can be an attractive strategy when
flying relatively fast (Rayner 1985, Ward-Smith 1984a) while
flap-gliding may offer greater advantages at slower speeds
(Ward-Smith 1984b, Rayner 1985). The production of lift by
the body and tail may help extend the range of aerodynamically
attractive speeds for flap-bounding to include the maximum
range speed—the speed predicted to be optimal for sustained
cruising flight (Rayner 1985, Tobalske et al 1999, 2009,
figure 6). The contribution of ‘turn-out’ phases during which
the wings are extended after a bound may allow flap-bounding
to offer an advantage over a broad range of speeds (DeJong
1983). Likewise, variation in flight speed and thrust can result
in predicted energetic advantages for both flap-bounding and
flap-gliding over a wide range of speeds (Rayner et al 2001).
3.4. Effects of morphology
There are prominent effects of body size and wing shape upon
the performance of intermittent flight. Within an intermediate
7
B W Tobalske
range of body mass, from 19 to 160 g, birds are capable
of using flap-bounding and flap-gliding flight, regardless of
their aspect ratio (wing length/average width) or wing loading
(body weight/wing area, Tobalske 1996, 2001). In contrast,
small birds with rounded, low-aspect ratio wings such as the
13 g zebra finch (aspect ratio, AR = 4.2) only use intermittent
bounds (Tobalske et al 1999, 2005). This pattern may be
explained by coupling the effect of a rounded, low-aspect ratio
wing shape increasing drag (Withers 1981) with the relatively
low inertia of small birds. Species of about the same body
mass but with more pointed, high-aspect ratio wings such as
the barn swallow (Hirundo rustica, 20 g, AR = 6.2) and house
martin (Delichon urbica, 17 g, AR = 6.5) use both bounds (or
partial bounds) and glides (Bruderer et al 2001).
There is an upper limit of body size upon the ability to
engage in intermittent bounds such that birds larger in size
than the pileated woodpecker (Dryocopus pileatus, 270 g,
Tobalske 1996, 2001) appear to use intermittent glides only.
This pattern appears to be explained by an adverse scaling
of the mass-specific power available for flight (Pennycuick
1975, Altshuler et al 2010). The percentage of time spent
flapping during cruising flight increases with increasing body
mass among passerines and woodpeckers (Danielson 1988,
Tobalske 1996, 2001), scaling ∝ M0.37 as would be expected
if wingbeat frequency, which scales ∝ M−0.46 , functions to
limit mass-specific power output. Likewise, the duration of
bound phases decreases as mass increases in woodpeckers
function of body size, and this may partially offset the adverse
scaling of the marginal mass-specific power available that
permits intermittent bounds (Tobalske 1996). The pectoralis
of woodpeckers is composed of red, fast-oxidative fibers, and
intermediate, fast-oxidative glycolytic fibers. The diameter
of red fibers does not vary with mass, but the percentage and
diameter of intermediate fibers both increase with increasing
size, scaling ∝ M0.79 and ∝ M0.15 , respectively. If the
intermediate fibers offer proportionally more power, muscle
anatomy could explain why the percentage of time spent
flapping in woodpeckers, ∝ M0.15 , does not increase at the
same rate as it does when exploring a larger comparative
sample that includes passerines (Tobalske 1996, 2001).
3.5. Effects of flight speed
The aerodynamics of flight at different speeds has significant
effects upon intermittent flight behavior. Induced power,
required for weight support, is greatest during hovering and
slow-speed flight, and profile and parasite power, required
to overcome pressure and skin-friction drag on the wings and
body, respectively, both increase exponentially with increasing
speed. As a consequence, mechanical power required for
flight varies as a U-shaped curve as a function of flight speed
(Pennycuick 1975, Rayner 1979a, 1979b, 1985, Tobalske et al
2003, Askew and Ellerby 2007). The shape of the power
curve varies with morphology and flight style, and intermittent
bounds and glides likely contribute to this variation (Tobalske
et al 2003).
Bioinsp. Biomim. 5 (2010) 045004
Among the species that use both bounds and glides, there
is a tendency to flap-glide at slow speeds and flap-bound during
faster flight (Tobalske and Dial 1994, Tobalske 1995, 1996,
Bruderer et al 2001, figure 4(B)). This behavior is consistent
with seeking to minimize power output at any given speed
(Rayner 1985). For these species, time spent flapping and
effective wingbeat frequency (total wingbeats per unit time)
varies according to an upwardly concave, U-shaped curve
(Bruderer et al 2001, Tobalske 2001). This behavior varies
with the goals of the animal. For example, Lewis’ woodpecker
(Melanerpes lewis) slowly glides when foraging on the wing
for insects, but it shifts toward bounds when flying faster and
more directly toward a destination (Tobalske 1996). In zebra
finch, a species that only flap-bounds, there is a decrease in
time spent flapping from 89% during brief hovering episodes
to 55% during fast forward flight (14 m s−1 ).
4. Conclusions
The smallest birds exhibit distinctly different strategies
to fly over a range of speeds. Hummingbirds have
converged upon the kinematics and aerodynamics of insects
to accomplish sustained hovering flight (Warrick et al
2005, 2009). They are thought to be unique among birds
in having a lift-producing upstroke during hovering, and the
morphological features associated with their ability to supinate
their wings (Stolpe and Zimmer 1939, Mayr 2003) appear to
be associated with a wing that stays fully outstretched through
the entire wingbeat cycle, even during fast forward flight.
The hummingbird species studied to date do not appear to
regularly use intermittent pauses during hovering or cruising
flight (Tobalske et al 2007, Clark and Dudley 2009, 2010).
In contrast, other relatively small birds use intermittent flight,
and they flex their wings to some extent during the upstroke
of their flapping phases.
The greatest flexibility in terms of forward flight styles
appears to be for birds of mass between 20 and 160 g that
have relatively pointed wings. These species approach a
hummingbird-style upstroke during take-off and slow flight by
using a supinated ‘tip-reversal’ upstroke, and they shift toward
an aerodynamically active upstroke in faster flight in which the
wings are progressively flexed as speed increases (Tobalske
2000). Moreover, they shift from flap-gliding during slow
forward flight toward flap-bounding in fast flight (Tobalske
and Dial 1994, figure 4(B)).
Less flexibility is apparent in small birds with rounded,
low-aspect ratio wings. These species tend to fully flex
their wings during upstroke, and the smallest species, such
as the zebra finch, use flap-bounding only. Flap-bounding
is not a useful strategy for saving energy during slow flight
(Rayner 1985). Thus the smallest flap-bounding species may
be specialized for flying relatively fast.
An inevitable adverse scaling of mass-specific power
required for flight versus mass-specific power available from
the flight muscles limits the ability to use flap-bounding as size
increases among species. This adverse scaling can be partially
offset by morphological changes such as increased wing
length or pectoralis mass in hummingbirds (Greenewalt 1962,
8
B W Tobalske
Altshuler et al 2010), increased glycolytic fiber content in
the pectoralis of woodpeckers (Tobalske 1996), or departures
from dynamic similarity such as an increase in muscle strain
as observed during take-off flight in pheasants and related
species (Phasianidae, Tobalske and Dial 2000). Nonetheless,
the largest birds that can take advantage of an ability to turn
off their flight muscles and momentarily act like a projectile
while still maintaining average altitude in the air are  0.3 kg.
Acknowledgments
I sincerely thank my collaborators who have contributed
greatly to my studies of hovering and intermittent flight. My
research summarized in this paper was supported by grants
from the Murdock Charitable Trust (99153 and 2001208)
and the National Science Foundation (IBN-0327380 and IOB-
0615648).
References
Altshuler D L and Dudley R 2002 The ecological and evolutionary
interface of hummingbird flight physiology J. Exp. Biol. 205
2325–36
Altshuler D L and Dudley R 2003 Kinematics of hovering
hummingbird flight along simulated and natural elevational
gradients J. Exp. Biol. 206 3139–47
Altshuler D L, Dudley R, Heredia S M and McGuire J A 2010
Allometry of hummingbird lifting performance J. Exp. Biol.
213 725–34
Altshuler D L, Dudley R and McGuire J A 2004 Resolution of a
paradox: hummingbird flight at high elevation does not come
without a cost Proc. Natl Acad. Sci. USA 101 17731–6
Altshuler D L, Princeva M, Pan H and Lozano J 2009 Wake patterns
of the wings and tail of hovering hummingbirds Exp. Fluids
46 835–46
Askew G N and Ellerby D J 2007 The mechanical power
requirements of avian flight Biol. Lett. 3 445–8
Baker E J and Gleeson T T 1999 The effects of intensity of the
energetics of brief locomotor activity J. Exp. Biol. 202 3081–7
Bartholomew G A and Lighton J R 1986 Oxygen consumption
during hover-feeding in free-ranging Anna hummingbirds
J. Exp. Biol. 123 191–9
Baudinette R V and Schmidt-Nielsen K 1974 Energy cost of gliding
flight in herring gulls Nature 248 83–4
Berger M 1985 Sauerstoffverbrauch von Kolibris Colibri coruscans
und C thalassinus beim Horizontalflug BIONA Report 3
ed W Nachtigall (Stuttgart: G Fischer) pp 307–14
Birch J M, Dickson W B and Dickinson M H 2004 Force production
and flow structure of the leading edge vortex on flapping wings
and high and low Reynolds numbers J. Exp. Biol. 207 1063–72
Borsheim E and Bahr R 2003 Effect of exercise intensity duration
and mode on post-exercise oxygen consumption Sports
Medicine 33 1037–60
Bruderer L, Liechti F and Bilo D 2001 Flexibility in flight behaviour
of barn swallows Hirundo rustica and house martins Delichon
urbica tested in a wind tunnel J. Exp. Biol. 204 1473–84
Bundle M W, Hansen K S and Dial K P 2007 Does the metabolic
rate-flight speed relationship vary among geometrically similar
birds of different mass? J. Exp. Biol. 210 1075–83
Clark C J 2009 Courtship dives of Anna’s hummingbird offers
insights into flight performance limits Proc. R. Soc. B
276 3047–52
Clark C J 2010 The evolution of tail shape in hummingbirds Auk
127 44–56
Clark C J and Dudley R 2009 Flight costs of long sexually selected
tails in hummingbirds Proc. R. Soc. B 276 2109–15
Bioinsp. Biomim. 5 (2010) 045004
Clark C J and Dudley R 2010 Hovering and forward flight energetic
in Anna’s and Allen’s hummingbirds Physiol. Biochem. Zool.
83 654–62
Csicsáky M 1977 Body-gliding in the zebra finch Fortschr. Zool. 24
275–86
Danielson R 1988 Parametre for fritflyvende småfugles flugt Dan.
Ornithol Foren Tidsskr. 82 59–60
DeJong M J 1983 Bounding flight in birds PhD Thesis University of
Wisconsin, Madison
Dial K P 1992 Avian forelimb muscles and nonsteady flight: can
birds fly without using the muscles in their wings? Auk 109
874–85
Dickinson M H, Lehmann F-O and Sane S P 1999 Wing rotation
and the aerodynamic basis of insect flight Science
284 1954–60
Earls K D 2000 Kinematics and mechanics of ground take-off in the
starling Sturnus vulgaris and the quail Coturnix coturnix
J. Exp. Biol. 203 725–739
Ellerby D J and Askew G N 2007 Modulation of pectoralis muscle
function in budgerigars Melopsitaccus undulatus and zebra
finches Taenopygia guttata in response to changing flight speed
J. Exp. Biol. 210 3789–97
Ellington C P 1991 Limitations on animal flight performance J. Exp.
Biol. 160 71–91
Ellington C P 1999 The novel aerodynamics of insect flight:
applications to micro-air vehicles J. Exp. Biol. 202 3439–48
Greenewalt C H 1962 Dimensional relationships for flying animals
Smithson. Misc. Collect. 144 1–46
Hedrick T L, Tobalske B W and Biewener A A 2002 Estimates of
circulation and gait change based on a three-dimensional
kinematic analysis of flight in cockatiels Nymphicus
hollandicus and ringed turtle-doves Streptopelia risoria J. Exp.
Biol. 205 1389–409
Karhu A 1992 Morphological divergence within the order
Apodiformes as revealed by the structure of the humerus
Papers in Avian Paleontology Honoring Pierce Brodkorb ed K
E Campbell (Science Series, Natural History Museum of Los
Angeles County vol 36) pp 379–84
Lehmann F-O 2004 The mechanisms of lift enhancement in insect
flight Naturwissenschaften 91 101–22
Lentink D, Müller U K, Stamhuis E J, de Kat R, van Gestel W,
Veldhuis L L M, Henningsson P, Hedenström A, Videler J J
and van Leeuwen J L 2007 How swifts control their glide
performance with morphing wings Nature 446 1082–5
Madangopal R, Khan Z A and Agrawal S K 2005 Biologically
inspired design of small flapping wing air vehicles using
four-bar mechanisms and quasi-steady aerodynamics J. Mech.
Des. 127 809–16
Mathieu-Costello O, Suarez R K and Hochachka P W 1992
Capillary-to-fiber geometry and mitochondrial density in
hummingbird flight muscle Respir. Physiol. 89 113–32
Mayr G 2003 Phylogeny of early tertiary swifts and hummingbirds
Aves: Apodiformes Auk 120 145–51
Meyers R A 1993 Gliding flight in the American kestrel Falco
sparverius: an electromyographic study J. Morphol.
215 213–24
Meyers R A and Mathias E 1997 Anatomy and histochemistry of
spread-wing posture in birds. 2. Gliding flight in the California
Gull Larus californicus: a paradox of fast fibers and posture
J. Morphol. 233 237–47
Muijres F T, Johansson L C, Barfield R, Wolf M, Spedding G R
and Hedenström A 2008 Leading-edge vortex improves lift in
slow-flying bats Science 319 1250–3
Norberg U M 1975 Hovering flight in the pied flycatcher Ficedula
hypoleuca Swimming and Flying in Nature vol 2 ed T Y-T Wu,
C J Brokaw and C Brennen (New York: Plenum)
pp 869–81
Pennycuick C J 1975 Mechanics of flight Avian Biology vol 5 ed
D S Farner and J R King (New York: Academic) pp 1–75
9
B W Tobalske
Rayner J M V 1977 The intermittent flight of birds Scale Effects in
Animal Locomotion ed T J Pedley (New York: Academic)
pp 37–55
Rayner J M V 1979a A vortex theory of animal flight: part I. The
vortex wake of a hovering animal J. Fluid Mech. 91 697–730
Rayner J M V 1979b A vortex theory of animal flight: part II. The
forward flight of birds J. Fluid Mech. 91 731–63
Rayner J M V 1985 Bounding and undulating flight in birds
J. Theor. Biol. 117 47–77
Rayner J M V, Viscardi P W, Ward S and Speakman J R 2001
Aerodynamics and energetics of intermittent flight in birds Am
Zool. 41 188–204
Robinson T R, Sargent R R and Sargent M B 1996 Ruby-throated
Hummingbird Archilochus colubris The Birds of North
America vol 204 ed A Poole and F Gill (Philadelphia, PA: The
Academy of Natural Sciences and Washington, DC: American
Ornithologists’ Union)
Rosser B W C and George J C 1986 The avian pectoralis:
histochemical characterization and distribution of muscle fiber
types Can. J. Zool. 64 1174–85
Rosser B W C, Waldbillig D M, Wick M and Bandman E 1994
Muscle fiber types in the pectoralis of the white pelican a
soaring bird Acta Zool. 75 329–36
Schmidt-Nielsen K 1972 Locomotion: energy cost of swimming
running and flying Science 177 222–8
Stolpe V M and Zimmer K 1939 Der Schwirrflug des Kolibri im
Zeitlupenfilm J. Ornithol. 87 136–55
Suarez R K, Lighton J R B, Brown G S and Mathieu-Costello O
1991 Mitochondrial respiration in hummingbird flight muscles
Proc. Natl Acad. Sci. USA 88 4870–3
Thomas A L R and Hedenström A 1998 The optimum flight speeds
of animals J. Avian Biol. 29 469–77
Tobalske B W 1995 Neuromuscular control and kinematics of
intermittent flight in European starlings Sturnus vulgaris
J. Exp. Biol. 198 1259–73
Tobalske B W 1996 Scaling of muscle composition wing
morphology and intermittent flight behavior in woodpeckers
Auk 113 151–77
Tobalske B W 2000 Biomechanics and physiology of gait selection
in flying birds Physiol. Biochem. Zool. 73 736–50
Tobalske B W 2001 Morphology velocity and intermittent flight in
birds Am. Zool. 41 177–87
Tobalske B W, Altshuler D L and Powers D L 2004 Take-off
mechanics in hummingbirds Trochilidae J. Exp. Biol.
207 1345–52
Tobalske B W and Biewener A A 2008 Contractile properties of the
pigeon supracoracoideus during different modes of flight
J. Exp. Biol. 211 170–9
Tobalske B W, Biewener A A, Warrick D R, Hedrick T L
and Powers D R 2010 Effects of flight speed upon muscle
activity in hummingbirds J. Exp. Biol. 213 2515–23
Tobalske B W and Dial K P 1994 Neuromuscular control and
kinematics of intermittent flight in budgerigars Melopsittacus
undulatus J. Exp. Biol. 187 1–18
Tobalske B W and Dial K P 2000 Effects of body size on take-off
fight performance in the Phasianidae Aves J. Exp. Biol. 203
3319–32
Tobalske B W, Hearn J W D and Warrick D R 2009 Aerodynamics
of intermittent bounds in flying birds Exp. Fluids
46 963–73
Tobalske B W, Hedrick T L, Dial K P and Biewener A A 2003
Comparative power curves in bird flight Nature 421 363–6
Tobalske B W, Olson N E and Dial K P 1997 Flight style of the
black-billed magpie: variation in wing kinematics,
neuromuscular control and muscle composition J. Exp. Zool.
279 313–29
Tobalske B W, Peacock W L and Dial K P 1999 Kinematics of
flap-bounding flight in the zebra finch over a wide range of
speeds J. Exp. Biol. 202 1725–39
Bioinsp. Biomim. 5 (2010) 045004
Tobalske B W, Puccinelli L A and Sheridan D C 2005 Contractile
activity of the pectoralis in the zebra finch according to mode
and velocity of flap-bounding flight J. Exp. Biol.
208 2895–2901
Tobalske B W, Warrick D R, Clark C J, Powers D R, Hedrick T L,
Hyder G A and Biewener A A 2007 Three-dimensional
kinematics of hummingbird flight J. Exp. Biol.
210 2368–82
Usherwood J R and Ellington C P 2002 The aerodynamics of
revolving wings: I. Model hawkmoth wings J. Exp. Biol. 205
1547–64
Vasquez R J 1992 Functional osteology of the avian wrist and the
evolution of flapping flight J. Morphol. 211 259–68
Vogel S V 1994 Life in Moving Fluids: The Physical Biology of
Flow 2nd edn (Princeton, NJ: Princeton University Press)
Ward-Smith A J 1984a Analysis of the aerodynamic performance of
birds during bounding flight Math. Biosci. 68 137–47
Ward-Smith A J 1984b Aerodynamic and energetic considerations
relating to undulating and bounding flight in birds J. Theor.
Biol. 111 407–17
Warrick D R 1998 The turning and linear maneuvering performance
of birds: the cost of efficiency for coursing insectivores Can. J.
Zool. 76 1063–79
10
B W Tobalske
Warrick D R, Tobalske B W and Powers D R 2005 Aerodynamics of
the hovering hummingbird Nature 435 1094–7
Warrick D R, Tobalske B W and Powers D R 2009 Lift production
in the hovering hummingbird Proc. R. Soc. B 276 3747–52
Weis-Fogh T 1972 Energetics of hovering flight in hummingbirds
and in Drosophila J. Exp. Biol. 56 79–104
Welch K C and Altshuler D L 2009 Fiber type homogeneity of the
flight musculature in small birds Comp. Biochem. Physiol. B:
Biochem. Mol. Biol. 152 324–31
Welch K C, Altshuler D L and Suarez R K 2007 Oxygen
consumption rates in hovering hummingbirds reflect
substrate-dependent differences in P/O ratios: carbohydrate as
a ‘premium fuel’ J. Exp. Biol. 210 310–6
Welch K C Jr and Suarez R K 2007 Oxidation rate and turnover of
ingested sugar in hovering Anna’s Calypte anna and rufous
Selasphorus rufus hummingbirds J. Exp. Biol. 210 2154–62
Wells D 1993 Muscle performance in hovering hummingbirds
J. Exp. Biol. 178 39–57
Withers P C 1981 An aerodynamic analysis of bird wings as fixed
aerofoils J. Exp. Biol. 90 143–62
Zufferey J-C 2008 Bio-Inspired Flying Robots: Experimental
Synthesis of Autonomous Indoor Flyers (Lausanne,
Switzerland: EPFL Press)