Morphology, Velocity, and Intermittent Flight in Birds 1

AMER. ZOOL., 41:177 187 (2001) Morphology, Velocity, and Intermittent Flight in Birds 1 BRET W. TOBALSKE 2 Department of Biology, University of Portland, 5000 North Willamette Boulevard, Portland, Oregon 97203 SYNOPSIS. Body size, pectoralis composition, aspect ratio of the wing, and forward speed affect the use of intermittent flight in birds. During intermittent nonflapping phases, birds extend their wings and glide or flex their wings and bound. The pectoralis muscle is active during glides but not during bounds; activity in other primary flight muscles is variable. Mechanical power, altitude, and velocity vary among wingbeats in flapping phases; associated with this variation are changes in neuromuscular recruitment, wingbeat frequency, amplitude, and gait. Species of intermediate body mass (35 158 g) tend to flap-glide at slower speeds and flapbound at faster speeds, regardless of the aspect ratio of their wings. Such behavior may reduce mechanical power output relative to continuous flapping. Smaller species ( 20 g) with wings of low aspect ratio may flap-bound at all speeds, yet existing models do not predict an aerodynamic advantage for the flight style at slow speeds. The behavior of these species appears to be due to wing shape rather than pectoralis physiology. As body size increases among species, percent time spent flapping increases, and birds much larger than 300 g do not flap-bound. This pattern may be explained by adverse scaling of mass-specific power or lift per unit power output available from flight muscles. The size limit for the ability to bound intermittently may be offset somewhat by the scaling of pectoralis composition. The percentage of time spent flapping during intermittent flight also varies according to flight speed. INTRODUCTION The next time you see a bird in flight, try to pay attention to the movement of its wings. Rather than moving them up and down continuously, the bird will probably alternate flapping phases with phases in which it holds its wings motionless relative to its body. During these pauses between wingbeats, the bird will either flex its wings against its body and bound or extend its wings and glide. Flap-bounding and flap-gliding represent different forms of intermittent flight that are used by many bird species. Both flight styles are associated with fluctuations in velocity and altitude that cause a bird s movement to describe an undulating path through the air, but the fluctuations in altitude are more apparent during flap-bounding. Intermittent bounds last 1 From the Symposium Intermittent Locomotion: Integrating the Physiology, Biomechanics and Behaviour of Repeated Activity presented at the Annual Meeting of the Society for Integrative and Comparative Biology, 4 8 January 2000, at Atlanta, Georgia. 2 E-mail: tobalske@up.edu for fractions of a second; intermittent glides vary in duration from milliseconds to minutes or more. To distinguish between flapgliding and soaring with sporadic wingbeats, consider that during glides, a bird loses altitude to maintain forward air velocity, and during soaring, a bird may gain altitude and velocity either from thermals, updrafts, or gradients in wind velocity. This is an exciting time for the study of intermittent flight because, in the past decade, empirical data have emerged from laboratory and field studies that are beginning to provide new insight into the functional morphology, physiology, and biomechanics of the behavior. In the 1970s and 1980s, various authors developed mathematical models with the goal of predicting the mechanical power required for intermittent flight and furthering understanding of the evolution and ecological significance of the flight style (Lighthill, 1977; Rayner, 1977, 1985; Alexander, 1982; DeJong, 1983; Ward-Smith, 1984a, b). It is the goal of this review to summarize recent empirical work and evaluate it in relation to ex- 177

178 BRET W. TOBALSKE FIG. 1. Wing kinematics and muscle activity patterns during (A) flap-gliding and (B) flap-bounding in a Blackbilled Magpie (Pica pica) flying in a wind tunnel (Tobalske, Olson, and Dial, unpublished data). EMG electromyogram. isting hypotheses that seek to answer the question: Why do birds pause in their wing movements during sustained flight? WING KINEMATICS AND MUSCLE ACTIVITY Intermittent bounds and glides may be recognized as distinct non-flapping postures using quantitative criteria including wingspan and muscle activation (Fig. 1). However, there is a wealth of variation in the details of how birds accomplish intermittent flight, and extensive research remains to be done to frame this diversity in a comparative, phylogenetic context. Although some species appear to only use one form or the other, a variety of species flap-bound and flap-glide (Tobalske and Dial, 1994, 1996; Tobalske, 1995, 1996). At least one species, the European Starling (Sturnus vulgaris), will couple bounds and glides during the same non-flapping phase (Tobalske, 1995). Some species exhibit partial-bounds with their wings slightly extended (Tobalske and Dial, 1994; Tobalske, 1995) while other species vary wingspan during intermittent glides as a function of flight velocity (Tobalske and Dial, 1996). Upstrokes are more variable than downstrokes in birds. Non-flapping postures may represent modified, prolonged upstrokes, although this is more apparent for bounds than for glides. Birds tend to begin a bound or partial bound as the wings are being elevated as during mid-upstroke. They resume flapping by either simultaneously extending and elevating their wings (Tobalske, 1996; Tobalske et al., 1999; Fig. 1) or, as in the European Starling, by first extending the wings and then elevating them in a pull-out (DeJong, 1983; Tobalske, 1995). Intermittent glides begin immediately after a partial or complete downstroke. Flapping resumes after a glide with a partial downstroke in the European Starling and Black-billed Magpie (Pica pica; Tobalske, 1995; Fig. 1) but an upstroke in the Budgerigar (Melopsittacus undulatus; Tobalske and Dial, 1994). The primary downstroke muscle, the pectoralis, is active during intermittent glides and inactive during bounds (Meyers, 1993; Tobalske and Dial, 1994; Tobalske, 1995; Fig. 1). During flapping flight, this muscle decelerates the humerus at the end

MORPHOLOGY, VELOCITY, AND INTERMITTENT FLIGHT 179 of upstroke and depresses and pronates the humerus during downstroke (Dial, 1992). During glides, using an isometric contraction, it opposes lift that tends to supinate and elevate the wing. Compared with peak values during wingbeats, electromyographic (EMG) activity during intermittent glides is lower in amplitude and muscle force is reduced (Meyers, 1993; Tobalske and Dial, 1994; Tobalske 1995; Dial et al., 1997; Fig. 1). These data suggest that fewer motor units are recruited in the pectoralis during glides compared to during wingbeats. Consistent with this idea, in the gliding American Kestrel (Falco sparverius) activity in the pectoralis is restricted to a distinct, deep cranial region (Meyers, 1993). Activity in other forelimb muscles during non-flapping phases is variable among or within species. Some of this variation may reflect differences in how birds stabilize their wing, which may, in turn, affect hypotheses on the evolution of bird flight. For example, the primary upstroke muscle, the supracoracoideus, supinates and elevates the humerus during a wingbeat; this muscle is hypothesized to have been critical to the evolution of flapping flight in birds (Poore et al., 1997). The muscle is inactive during glides in the Budgerigar (Tobalske and Dial, 1994), but it is sporadically active during glides in the European Starling (Tobalske, 1995), American Kestrel (Meyers, 1993) and Black-billed Magpie (Pica pica, Fig. 1). During bounds, the supracoracoideus is inactive in the Budgerigar and European Starling, yet it is active in the Blackbilled Magpie (Tobalske and Dial, 1994; Tobalske, 1995; Fig. 1). The flapping phases of intermittent flight seem to be characterized by variation in mechanical power among wingbeats, although measures of in vivo mechanical power that allow a direct test of this hypothesis are only available for the Blackbilled Magpie (Dial et al., 1997). Based on observed relationships among pectoralis force, kinematic events, and EMG patterns in the Black-billed Magpie, variation in within-wingbeat power output may be tentatively inferred for other species. For example, both the Black-billed Magpie and the European Starling accelerate and gain altitude using high frequency, high amplitude wingbeats with mid-upstroke spans that suggest the use of a vortex-ring gait (Tobalske, 1995; Tobalske and Dial, 1996; Tobalske et al., 1997). Electromyographic signals from the primary flight muscles are characteristically shorter in duration and higher in amplitude and relative intensity during these wingbeats. Data from the Black-billed Magpie indicate that more power is generated by the pectoralis at such times. In contrast, deceleration and a loss of altitude occur during non-flapping intervals and during wingbeats immediately following non-flapping phases. These wingbeats feature reduced wingbeat frequency and amplitude, mid-upstroke spans suggesting the use of a continuous-vortex gait, and EMG bursts that are longer in duration and lower in amplitude and relative intensity. Data from the Black-billed Magpie indicate that pectoralis power is relatively low during these conditions. ENERGY SAVING Flapping flight requires extraordinary metabolic power output (e.g., Nudds and Bryant, 2000), and it seems obvious that intermittent flight could save energy under some circumstances. To examine this idea, it is important to distinguish between mechanical power output of the flight muscles, which should match the aerodynamic power required for flight, and the metabolic power input to the muscles, which should equal mechanical power output and heat loss (Rayner, 1999). Using aerodynamic theory, various authors have modeled the mechanical power output required for intermittent flight in comparison to continuous flapping. Although mechanical power output from the flight muscles is zero during bounds and glides, any saving offered during these nonflapping phases may be offset by a disproportionate increase in mechanical power required during flapping phases to maintain average weight support and thrust. Specific predictions from existing models vary greatly and are sensitive to assumptions about wing kinematics and the magnitudes of induced, parasite, and profile drag on the bird. Compared to continuous flapping,

180 BRET W. TOBALSKE however, it generally appears that flap-gliding should require less mechanical power output at slow to moderate speeds (Rayner, 1977, 1985; Ward-Smith, 1984b), and flapbounding should require less mechanical power output at fast speeds (Rayner, 1977, 1985; DeJong, 1983; Ward-Smith, 1984a, b). If flap-bounding birds support some of their weight during bounds using body lift (Csicsáky, 1977) or pull-out phases (De- Jong, 1983), they might gain an aerodynamic advantage relative to continuous flapping at moderate speeds including V mr, the maximum-range speed considered to be optimal for migration (Rayner, 1985). For example, Zebra Finch (Taenopygia guttata) generate body lift during bounds (Tobalske et al., 1999), and an analysis using Rayner s (1985) model suggests that the species might gain an advantage over continuous flapping by flap-bounding at speeds from 6 to 14 m sec 1. Caution is warranted in extrapolating predictions from models of mechanical power to predictions of energy saving for at least two reasons. First, it is not yet clear how muscle efficiency varies with flight speed and wingbeat kinematics in birds. Small changes in efficiency could have significant effects on the shape of the metabolic power curve (Thomas and Hedenström, 1998; Rayner, 1999) and could significantly alter predictions about the relative merits of intermittent flight versus continuous flapping. Secondly, although the primary flight muscles are inactive or contracting isometrically during intermittent bounds and glides, if these non-flapping phases are brief in duration, reductions in metabolic power during the bounds or glides will probably be less than reductions associated with long-duration gliding (e.g., Baudinette and Schmidt-Nielsen, 1974). Metabolic rate in mammals remains elevated during postexercise recovery (Baker and Gleeson, 1999), so the high metabolic rate observed during flapping flight in birds may not decrease for a substantial time interval after flapping has stopped. It would be worthwhile to investigate this idea in flying birds, particularly given that metabolic rate during short-duration flights interspersed with perching phases is much higher than predicted from existing aerodynamic and physiologic models (Nudds and Bryant, 2000). A direct test of the hypothesis that intermittent flight can save energy awaits further comparative study that includes direct measures of mechanical power (Dial et al., 1997; Biewener et al., 1998), metabolic power, and efficiency (Rayner, 1999). An ideal study would utilize a single species as a model that changes between continuous flapping and intermittent flight over the same range of flight speeds. BODY SIZE AND LIMITS ON FLAP-BOUNDING As size increases among bird species, flight performance declines. One explanation for this trend is that mass-specific power available from the flight muscles scales proportional with wingbeat frequency and, therefore, negatively with increasing body mass (Pennycuick, 1975). An alternative explanation is that lift per unit power output scales negatively with increasing body mass (Marden, 1994). A decline in the ability to engage in intermittent bounds is apparent with increasing body mass among woodpeckers (Picidae; Tobalske, 1995). To frame this observation in a broader comparative context for the purposes of this review, I collected new data from three additional species of woodpeckers (Williamson s Sapsucker, Sphyrapicus thyroideus; Black-backed Woodpecker, Picoides arcticus; and Three-toed Woodpecker (P. tridactylus) using the same methods as in Tobalske (1996). I also analyzed the effects of body mass on flapbounding flight in 12 species of migrating passerines (Passeriformes) using kinematic data reported in Danielson (1988). Body masses for species were obtained from Tobalske (1996) and Dunning (1993). Scaling relationships were analyzed using reducedmajor axis (RMA) regressions; to convert least-squares regression slope to RMA regression slope, divide the LS slope by the correlation coefficient (r). For clarity in this paper, I only include figures showing species (tip) data, but all tests of statistical significance were performed using phylogenetically-correct analysis of covariance (PC-ANCOVA,) and Independent Contrasts (IC; Jones et al., 1998, PDAP v. 5.0). These

MORPHOLOGY, VELOCITY, AND INTERMITTENT FLIGHT 181 FIG. 2. Hypothesized phylogeny of woodpecker (Piciformes) and passerine (Passeriformes) species for which wing kinematics during intermittent flight have been studied in the field (derived from Short, 1982; Sibley and Ahlquist, 1990; Tennant, 1991; Cicero and Johnson, 1995; and Moore and DeFilippis, 1997). Branch lengths are either directly from, or set proportional to, the UPGMA tree in Sibley and Ahlquist (1990). Relationships are derived from molecular evidence except for branches marked with an asterisk; for these branches, anatomical data were used to infer synapomorphies and branch lengths were assumed to be identical to those between immediate sister taxa in the sample. methods account for the non-independence of species due to phylogeny (Garland et al., 1992, 1993). The hypothesized phylogeny for these species was based on molecular and morphological data and included one unresolved polytomy (Fig. 2). Among woodpecker and passerine species using flap-bounding, wingbeat frequency decreases proportional with the 0.46 power of body mass (M 0.46 ; PC-ANCOVA; F 5.8, df 1, 19, P 0.029, Fig. 3A) based on species data and proportional with M 0.62 based on independent contrasts (F 20.6, df 1, 18, P 0.001). The percentage of time spent flapping increases proportional with M 0.37 (PC-ANCOVA, F 10.2, df 1, 19, P 0.005, Fig. 3B) or, using IC, with M 0.49 (F 12.2, df 1, 18, P 0.01). Although the scaling of the percent time spent flapping was different between groups (species data; woodpeckers, RMA slope M 0.27, r 0.59; passerines RMA slope M 0.4, r 0.81), the observed difference was not statistically significant (PC-ANCOVA, F 2.04, df 1, 17, P 0.17). Likewise, the difference in RMA regression slopes for wingbeat frequency between woodpeckers and passerines was not statistically significant (PC-ANCOVA, F 0.18, df 1, 17, P 0.7). These patterns are consistent with the idea that a decline in the mass-specific power available for flight is related with the rate at which work is performed (Pennycuick, 1975). As direct measures of lift and mechanical power are lacking for these species, it is also possible that the observed pattern is due to a decline in lift per unit power output (Marden, 1994). Using observed scaling of acceleration ability in passerines (DeJong, 1983) and predictions of mass-specific power available from flight muscle (Rayner, 1977, 1985), it was previously felt that the size

182 BRET W. TOBALSKE wingbeat frequency (Fig. 3A) or lift per unit power output. The inevitable decline in flight performance must prevail, however, and this likely explains why larger birds much greater in size than 300 g engage in flap-gliding rather than flap-bounding. FIG. 3. Scaling of (A) wingbeat frequency and (B) the percent time spent flapping during flap-bounding flight in woodpeckers (Piciformes, n 9) during the breeding season and passerines (Passeriformes, n 12) during migration (data from Danielson, 1988; Dunning, 1993; Tobalske, 1996; Tobalske, unpublished data). Reduced major axis (RMA) regression lines and formulas were obtained using log-transformed data. Open circles passerines, filled circles woodpeckers. limit for the ability to flap-bound should be approximately 100 g. However, the Pileated Woodpecker (Dryocopus pileatus; 262.2 49.5 g [SD]) regularly engages in flapbounding flight (Tobalske, 1996), and my (unpublished) observations of the Black Woodpecker (Dryocopus martius; 321 30.3 g) suggest that it also engages in intermittent bounds. One factor that may help account for the ability of these larger woodpecker species to use intermittent bounds is the scaling of pectoralis composition. The diameter and percentage of intermediate (fast-oxidative glycolytic type I) fibers increases as body mass increases among woodpeckers (Tobalske, 1996; Fig. 4). It may be that these intermediate muscle fibers afford higher muscle stress and, thus, power output, at a given strain rate compared to smaller-diameter type R (red) fibers. If this is the case, the presence of type I fibers could somewhat offset the negative scaling of mass-specific power imposed by AN UNFIXED PECTORALIS GEAR A dominant hypothesis on the functional significance of intermittent bounds has been that the pectoralis muscle in small species may only be capable of a narrow range of efficient contractile velocities. Intermittent bounds could therefore represent a method for permitting within-wingbeat contractile velocity and power output to remain constant while reducing the mechanical power output below that required for hovering, takeoff or vertical ascent (Rayner, 1985). This argument developed from several lines of evidence: 1) a given muscle fiber type has a characteristic force-velocity relationship with a relatively narrow range of peak work and power for a given contractle velocity (Hill, 1950); 2) many small birds have only a single fast-oxidative-glycolytic fiber type in their pectoralis (Rosser and George, 1986; Rosser et al., 1996; Fig. 4); 3) early EMG recordings of the pectoralis muscle of the Budgerigar suggested that the entire muscle contracted simultaneously as a single motor unit or task group (Loeb and Gans, 1986) during downstroke, with no potential for force variation via variation in motor-unit recruitment (Aulie, 1970); and 4) birds such as the Zebra Finch use flap-bounding flight during hovering and slow flight even though mathematical theory predicts continuous flapping should be a better strategy for minimizing mechanical power output at such speeds (Rayner, 1985; Tobalske et al., 1999). In spite of the intuitive appeal of this hypothesis, current evidence suggests that the pectoralis muscle in small flap-bounding birds does not function strictly as a fixedgear and that this hypothesis should be revised to focus upon constraints associated with wing shape and wingbeat gait (Tobalske et al., 1999). Bipolar EMG electrodes placed intramuscularly in the Budgerigar (35 g, only type R fibers in pectoralis) show multiple spikes representing spatial and

MORPHOLOGY, VELOCITY, AND INTERMITTENT FLIGHT 183 FIG. 4. Transverse sections of woodpecker pectoralis muscle stained for succinic acid dehydrogenase (SDH) activity (adapted from Tobalske (1996). Among woodpecker species, the heterogeneity of fiber types in the pectoralis muscle increases with an increase in body size. R red, fast oxidative glycolytic; I intermediate, fast oxidative glycolytic. Bars represent 20 m. temporal variation in motor unit recruitment (Tobalske and Dial, 1994). The relative intensity and duration of EMG signals and wingbeat frequency all vary extensively during flapping phases of intermittent flight in the Budgerigar (Fig. 5) much as in the larger European Starling and Black-billed Magpie. Moreover, even though Budgerigars have only one type of fiber in their pectoralis, they exhibit continuous flapping or intermittent glides during hovering and slow flight. Although the Zebra Finch uses intermittent bounds during hovering and slow flight, the angular velocity of its wing during downstroke differs significantly among flight speeds (Tobalske et al., 1999; Fig. 5). Preliminary comparative study (Tobalske et al., 1999) suggests the observed variation is similar to that exhibited by the Budgerigar and the Ruby-throated Hummingbird (Archilocus colubris, 3 g) during flight over wide ranges of speed, and these species do not appear to regularly use intermittent bounds during slow flight. In contrast to the Zebra Finch, however, these two species have wings of higher aspect ratio, and they either change wingbeat gait or dramatically vary other wingbeat kinematics such as stroke-plane angle according to flight speed. Thus, it appears that wing design, rather than pectoralis composition, may account for the use of intermittent bounds during slow flight in some species of small birds (Tobalske et al., 1999). Intermittent bounds may represent a crude way of varying altitude for a species that seldom hovers or flies slowly. WING DESIGN Birds that vary in body mass from 19 to 158 g are known to use both flap-bounding and flap-gliding (Tobalske and Dial, 1994, 1996; Tobalske, 1995, 1996; Warrick, 1998; Warrick, personal communication; Fig. 6). Some smaller species such as the Zebra Finch only use flap-bounding flight, and larger species the size of the Rock Dove (pigeon, Columba livia; 333 g) or larger only flap-glide or soar. The behavior of the

184 BRET W. TOBALSKE FIG. 5. Evidence against a fixed muscle gear in small birds that possess only one type of fiber in their pectoralis muscle. A) Relative intensity of electromyographic (EMG) activity in the Budgerigar (Melopsittacus undulatus) varies among wingbeats during flapping phases of intermittent flight (adapted from Tobalske and Dial, 1994). B) Average angular velocity of the wing (degrees msec 1 ) of the Zebra Finch (Taenopygia guttata) varies among flight speeds (from Tobalske et al., 1999). Zebra Finch may be related to aspect ratio of the wings but not wing loading. The behavior of the pigeon may be explained by the decline in flight performance associated with increasing body mass among species (see Body size and flap-bounding). Wings that are higher in aspect ratio (AR; wing span divided by average wing chord) have higher lift:drag ratios than wings that are lower in aspect ratio (Withers, 1981). It is probably for this reason that birds such as the Zebra Finch (AR 4.4; Tobalske et al., 1999) that have wings of low aspect ratio do not regularly use intermittent glides. Curiously, however, the Black-billed Magpie, a larger bird (158 g) with similar aspect ratio (4.5) uses intermittent glides FIG. 6. Interactive effects of wing design and body mass upon the use of intermittent flight styles. A) Aspect ratio. B) Wing loading (N m 2 ). Wing kinematics for birds in this sample have been observed either in a variable-speed wind tunnel over a range of speeds or outdoors at different speeds selected by the species (data from Tobalske and Dial, 1994, 1996; Tobalske, 1995, 1996; Warrick, 1998; Tobalske et al., 1999; Warrick, personal communication). Average body mass (g) included after common name. (Fig. 6). One potential explanation for this is based upon Reynolds number. Assuming approximately the same wing shape, the zebra finch may exhibit relatively poor glide performance because it is operating at relatively lower Reynolds numbers at which viscous forces may be more significant than viscous forces for the moderately larger black-billed magpie. This idea merits broader comparative study. Glide performance is also related to wing loading (N m 2 ). Any characteristic speed for a gliding animal is proportional to the square root of wing loading (Pennycuick, 1975), so an animal with relatively low wing loading may glide at relatively slower speeds before stalling. The Zebra Finch has a wing loading (20.1 N m 2 ) that is in the middle of the range for wing loading of species that use intermittent glides (Fig. 6). Thus, the relative surface area of the wings does not appear to account for why the species only uses intermittent bounds (Fig. 6).

MORPHOLOGY, VELOCITY, AND INTERMITTENT FLIGHT 185 FIG. 7. Effect of flight speed on wing postures selected during non-flapping phases in the Budgerigar (Melopsittacus undulatus, from Tobalske and Dial, 1994). Similar patterns are exhibited by the European Starling (Surnus vulgaris; Tobalske, 1995), and Blackbilled Magpie (Pica pica; Tobalske and Dial, 1996). FLIGHT SPEED The profile drag on a bird s wings rises as a function of increasing flight speed. It is for this reason that periodically flexing the wings during bounds is predicted to offer a reduction in mechanical power output for flap-bounding flight relative to continuous flapping at fast flight speeds (Lighthill, 1977; Rayner, 1977, 1985; Alexander, 1982; DeJong, 1983; Ward-Smith, 1984a, b). As flight speed increases, species including the Budgerigar, European Starling, and Black-billed Magpie decrease the percentage of glides and increase the percentage of bounds among all non-flapping phases (Tobalske and Dial, 1994, 1996; Tobalske, 1995; Fig. 7). Among woodpeckers, the Lewis s Woodpecker is unusual because it regularly glides for long intervals while flycatching (Tobalske, 1996). During level, flap-bounding flight, the species flaps for a significantly greater percentage of flight time and flies significantly slower than other species of woodpeckers (P 0.05, Fig. 8). Limited observations of gliding speeds suggest that the Lewis s Woodpecker glides at even slower speeds than speeds used for flapbounding (6.0 compared to 7.3 m sec 1 ;Tobalske, 1996). The behavior of this species appears to be explained by flight speed rather than any component of wing shape because a variety of variables, including aspect ratio or surface area of the wing, are not ususual given the species body mass FIG. 8. Reduced major axis regressions describing the relationships between body mass in woodpeckers and (A) percentage of time spent flapping during flapbounding flight (n 9), (B) flight speed during flapbounding flight (n 9), and (C) wing area (n 7). Data from Tobalske (1996) except data for Williamson s Sapsucker (Sphyrapicus nuchalis), Three-toed Woodpecker (Picoideus tridactylus) and Black-backed Woodpecker (P. arcticus) from Tobalske (unpublished data). Filled circles represent Lewis s Woodpecker (Melanerpes lewis); regression lines were computed without including this species in the sample. and phylogeny (Tobalske, 1996; Figs. 2 and 8). Independent contrasts representing the difference between the Lewis s Woodpecker and sister taxa in the Melanerpini, the Sapsuckers (Sphyrapicus spp.) revealed the same trends indicated in Figure 8: compared to its nearest relatives, the Lewis s Woodpecker has an unusual flight speed but not wing design. Pectoralis composition is somewhat unique in the Lewis s Woodpecker, perhaps revealing correlates with its flap-gliding behavior. Fibers in the pectoralis appear rounded and relatively loosely-packed in cross-section; this morphology is potentially characteristic of birds that glide or soar

186 BRET W. TOBALSKE extensively (George and Berger, 1966; Rosser and George, 1986). Rosser and George (1986) hypothesized that it indicates that lower stress is required in the muscle for gliding compared with flapping. Unlike in other woodpecker species, yet similar to the European Starling, another species that often engages in flap-gliding, type I fibers in the pectoralis of the Lewis s Woodpcker stain intensely for alpha glycerophosphate dehydrogenase. This indicates high glycolytic capacity. These type I fibers may be selectively recruited during continuous flapping and then turned off during prolonged glides (Tobalske, 1996). During wind-tunnel flight, the percentage of time spent flapping varies significantly with flight speed (Fig. 9). In the Budgerigar and European Starling, this variation describes a -shaped curve as speed increases (Tobalske and Dial, 1994; Tobalske, 1995). In the Zebra Finch, the curve declines with each increase in flight speed (Tobalske et al., 1999; Fig. 9). It is generally expected that the mechanical power curve for flight varies as a -shape with flight speed (Pennycuick, 1975; Rayner, 1985, 1999). As mechanical power output during intermittent glides and bounds is zero, the percentage of time spent flapping provides a rough indication of the shape of the mechanical power curve for flight in these species. It is noteworthy, then, that the curve for the percentage of time spent flapping decreases with speed in the Zebra Finch even though minimum power speed for this species is estimated to be near 4 m sec 1 (Rayner, 1985). One possible explanation for this discrepancy is that body lift during intermittent bounds helps forestall an increase in mechanical power output as this species increases flight speed up to 14 m sec 1 (Tobalske et al., 1999). It would, therefore, be worthwhile to test the body lift:drag ratios of other species during intermittent bounds. FIG. 9. Effect of flight speed on the percent time spent flapping during intermittent flight in the Zebra Finch (Taenopygia guttata), Budgerigar (Melopsittacus undulatus) and European Starling (Sturnus vulgaris). ACKNOWLEDGMENTS I wish to thank all of the individuals who have contributed in substantial ways to my study of intermittent flight in birds: Andrew Biewener, Ken Dial, Nate Olson, Wendy Peacock, Jeremy Rayner, Jerred Seveyka, Claudine Tobalske, and Doug Warrick. I also thank Randi Weinstein for organizing the symposium on intermittent locomotion and inviting me to contribute to the occasion. This research was supported, in part, by Murdock grant 99153. REFERENCES Alexander, R. McN. 1982. Optima for animals. Arnold, London. Aulie, A. 1970. Electrical activity from the pectoral muscle of a flying bird, the Budgerigar. Comp. Biochem. Physiol. 36:297 300. Baker, E. J. and T. T. Gleeson. 1999. The effects of intensity of the energetics of brief locomotor activity. J. Exp. Biol. 202:3081 3087. Baudinette, R. V. and K. Schmidt-Nielsen. 1974. Energy cost of gliding flight in herring gulls. Nature 248:83 84. Biewener, A. A., W. R. Corning, and B. W. Tobalske. 1998. In vivo pectoralis muscle force-length behavior during level flight in pigeons (Columba livia). J. Exp. Biol. 201:3293 3307. Cicero, C. and N. K. Johnson. 1995. Speciation in sapsuckers (Sphyrapicus): III. Mitochondrial-DNA sequence divergence at the cytochrome-b locus. Auk 112:547 563. Csicsáky, M. J. 1977. Body-gliding in the Zebra Finch. Fortschr. Zool. 24:275 286. Danielson, R. 1988. Parametre for fritflyvende småfugles flugt. Dan. Ornithol. Foren. Tidsskr. 82:59 60. DeJong, M. J. 1983. Bounding flight in birds. Ph.D. Diss., 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 109:874 885. Dial, K. P., A. A. Biewener, B. W. Tobalske, and D. R. Warrick. 1997. Mechanical power output of bird flight. Nature 390:67 70. Dunning, J. B., Jr. (ed.) 1993. CRC handbook of avian body masses. CRC Press, Boca Raton. Garland, T., Jr., P. H. Harvey, and A. R. Ives. 1992. Procedures for the analysis of comparative data using phylogenetically independent contrasts. Syst. Biol. 41:18 32. Garland, T., Jr., A. W. Dickerman, C. M. Janis, and J.

MORPHOLOGY, VELOCITY, AND INTERMITTENT FLIGHT 187 A. Jones. 1993. Phylogenetic analysis of covariance by computer simulation. Syst. Biol. 42:265 292. George, J. C. and A. J. Berger. 1966. Avian myology. Academic Press, London. Hill, A. V. 1950. The dimensions of animals and their muscular dynamics. Sci. Prog. 38:209 230. Jones, J. A., P. E. Midford, and T. Garland, Jr. 1998. Phenotypic Diversity Analysis Program v 5.0. Univ. of Wisconsin, Madison. Lighthill, M. J. 1977. Introduction to the scaling of aerial locomotion. In T. J. Pedley (ed.), Scale effects in animal locomotion, pp. 365 404. Academic Press, New York. Loeb, G. E. and C. Gans. 1986. Electromyography for experimentalists. University of Chicago Press, Chicago. Marden, J. 1994. From damselflies to pterosaurs: How burst and sustainable flight performance scale with size. Am J. Physiol. 266:R1077 R1084. Meyers, R. A. 1993. Gliding flight in the American Kestrel (Falco sparverius): An electromyographic study. J. Morph. 215:213 224. Moore, W. S. and V. R. DeFillipis. 1997. The window of taxonomic resolution for phylogenies based on mitochondrial cytochrome b. In D. P. Mindel (ed.), Avian molecular evolution and systematics, pp. 83 119. Academic Press, San Diego. Nudds, R. L. and D. M. Bryant. 2000. The energetic cost of short flights in birds. J. Exp. Biol. 203: 1561 1572. Pennycuick, C. J. 1975. Mechanics of flight. In D. S. Farner (ed.), Avian biology, Vol. 5, pp. 1 75. Academic Press, London. Poore, S. O., A. Sánchez-Haiman, and G. E. Goslow, Jr. 1997. Wing upstroke and the evolution of flapping flight. Nature 387:799 802. Rayner, J. M. V. 1977. The intermittent flight of birds. In T. J. Pedley (ed.), Scale effects in animal locomotion, pp. 37 55. Academic Press, New York. Rayner, J. M. V. 1985. Bounding and undulating flight in birds. J. Theor. Biol. 117:47 77. Rayner, J. M. V. 1999. Estimating power curves of flying animals. J. Exp. Biol. 202:3449 3461. Rosser, B. W. C. and J. C. George. 1986. The avian pectoralis: Histochemical characterization and distribution of muscle fiber types. Can. J. Zool. 64: 1174 1185. Rosser, B. W. C., M. Wick, D. M. Waldbillig, and E. Bandman. 1996. Heterogeneity of myosin heavychain expression in fast-twitch fiber types of mature avian pectoralis muscle. Biochem. Cell. Biol. 74:715 728. Short, L. L. 1982. Woodpeckers of the world. Delaware Museum of Natural History, Greeneville. Sibley, C. A. and J. E. Ahlquist. 1990. Phylogeny and classification of birds: A study in molecular evolution. New Haven, Yale University Press. Tennant, M. R. 1991. Phylogenetic systematics of the Picinae. Ph.D. Diss., Wayne State University, Detroit, Michigan. Thomas, A. L. R. and A. Hedenström. 1998. The optimum flight speeds of animals. J. Avian Biol. 29: 469 477. Tobalske, B. W. 1995. Neuromuscular control and kinematics of intermittent flight in European Starlings (Sturnus vulgaris). J. Exp. Biol. 198:1259 1273. Tobalske, B. W. 1996. Scaling of muscle composition, wing morphology, and intermitent flight behavior in woodpeckers. Auk 113:151 177. Tobalske, B. W. and K. P. Dial. 1994. Neuromuscular control and kinematics of intermittent flight in Budgerigars (Melopsittacus undulatus). J. Exp. Biol. 187:1 18. Tobalske, B. W. and K. P. Dial. 1996. Flight kinematics of Black-billed Magpies and pigeons over a wide range of speeds. J. Exp. Biol. 199:263 280. Tobalske, B. W., N. E. Olson, and K. P. Dial. 1997. Flight style of the Black-billed Magpie: Variation in wing kinematics, neuromuscular control, and muscle composition. J. Exp. Zool. 279:313 329. Tobalske, B. W., W. L. Peacock, and K. P. Dial. 1999. Kinematics of flap-bounding flight in the Zebra Finch over a wide range of speeds. J. Exp. Biol. 202:1725 1739. Ward-Smith, A. J. 1984a. Analysis of the aerodynamic performance of birds during bounding flight. Math. Biosc. 68:137 147. Ward-Smith, A. J. 1984b. Aerodynamic and energetic considerations relating to undulating and bounding flight in birds. J. Theor. Biol. 111:407 417. Warrick, D. R. 1998. The turning- and linear-maneuvering performance of birds: The cost of efficiency for coursing insectivores. Can. J. Zool. 76: 1063 1079. Withers, P. C. 1981. An aerodynamic analysis of bird wings as fixed aerofoils. J. Exp. Biol. 90:143 162.