From frond to fan: Archaeopteryx and the evolution of short-tailed birds.
It has been assumed that the transition from a frond-shaped to a fan-shaped tail improved flight ability (Heilmann 1926; Cracraft 1986; Sanz and Bonaparte 1992; Zhou et al. 1992; Chiappe 1995). Workers have suggested that the evolution of radially arranged rectrices enhanced aerial maneuverability and braking (Heilmann 1926; Lowe 1944; Maynard Smith 1952; de Beer 1956; Sereno and Rao 1992). Others have inferred greater tail strength in a fan, relative to a frondlike configuration (Bock 1986; Sereno and Rao 1992). Several researchers simply deemed a frond-shaped tail inferior (e.g., Marshall 1872; Lowe 1944). Lowe, for example, stated that "an arrangement of this sort, it is needless to say, is not only utterly different from that seen in a bird's tail, but would be quite useless in performing the guiding movements so typical of the bird's tail" (1944, p. 535).
Despite these general sentiments, however, the specific nature of presumed improvements in aerodynamic performance, neuromuscular control, or structural integrity has never been explained. Herein, we present data that elucidate the function of tail structures in living birds, particularly as they relate to fanning. These are used to help infer mechanisms of caudal control in Archaeopteryx and to discuss the evolution of bird tails.
Avian caudal morphology is best known in the domestic pigeon (Columba livia). Only a brief overview is presented here (see Baumel 1988 for a complete description). The tail skeleton consists of five or six mobile vertebrae followed by a terminal pygostyle [ILLUSTRATION FOR FIGURE 1B, 2 OMITTED]. The pygostyle is a blade-like bone that forms by the fusion of several embryonic vertebral segments. This short caudal axis acts to support and move the main tail feathers, or rectrices. Rectrices vary in number (12 in the pigeon and 18 in the turkey, Meleagris gallopavo, for example), but their relationship to the caudal vertebrae and musculature is quite consistent across species. Contrary to many descriptions, only the two most medial rectrices are directly attached to the pygostyle (Steiner 1938; Raikow 1985; Baumel 1988). The remainder are secured by two soft tissue assemblages called the rectricial bulbs ([ILLUSTRATION FOR FIGURE 2A, B OMITTED]; Baumel 1988). The bulbs are highly organized structures primarily made up of connective tissue and fat. A bulb lies on each side of the pygostyle and encases the roots (cal-ami) of the rectrices. Spiraling around the surface of each bulb is a striated muscle, the bulbi rectricium. Other muscles connect the caudal vertebrae and the rectricial bulbs to the pelvis, synsacrum, hind limb, and vent [ILLUSTRATION FOR FIGURE 2A, B OMITTED]).
The rectrices form a flight surface that is tightly coupled with the wings during aerial locomotion (Gatesy and Dial 1993, 1996). However, the functional interaction of caudal muscles during flight and other behaviors is only beginning to be understood. Hypotheses of muscle function based solely on morphology (Burt 1930; Fisher 1946; Owre 1967; Raikow 1970, Baumel 1971, 1988) have only recently been extended and refined by studies of muscle activity (Baumel et al. 1990; Gatesy and Dial 1992, 1993, 1996; Kilpatrick and Gatesy 1994). One way of approaching this complex system is to consider that the tail can modify its aerodynamic characteristics in two basic ways. First, vertebral movements can change the position of the rectrices with respect to the body. During flight, caudal musculature must produce forces to maintain and adjust tail position. Second, interrectricial movements can change the size and shape of the surface itself. The rectrices in most birds can be spread from their parallel resting position to a radial arrangement similar to the ribs of a fan. Recent theoretical analysis (Thomas 1993) and wind-tunnel experiments on wing-tail models (Hummel 1992) have confirmed the significance of fan size in modulating the tail's aerodynamic properties. Several muscular mechanisms have been suggested for the control of rectricial spreading.
Traditionally, avian myologists have considered the later-alis caudae muscle [ILLUSTRATION FOR FIGURE 2C OMITTED] to be primarily responsible for tail fanning (Burt 1930; Fisher 1946; Owre 1967; Raikow 1970). It has also been suggested that other muscles, such as the pubocaudalis externus, levator caudae, and depressor caudae, may play a role in assisting the lateralis caudae in rectricial abduction. These studies did not recognize the bulbi rectricium and, therefore, proposed no action for this muscle. Allen (1962) first described what is now known as the bulbi rectricium and proposed that it fans the rectrices "by pulling tips of calami closer together" (1962, p. 216). The lateralis caudae and depressor caudae were also thought to cause spreading, but the relative contribution of these three muscles was not discussed.
Baumel (1988) identified the rectricial bulbs and presented the most detailed model of tail fanning in his extensive treatise on pigeon tail morphology [ILLUSTRATION FOR FIGURE 2D OMITTED]. This mechanism involves both the bulbi rectricium and the lateralis caudae. The bulbi rectricium was hypothesized to produce "preliminary actions" to "provide m. lateralis caudae with an improved line of action so that its contraction can contribute to spreading of the rectrices by abduction of the outermost rectrix." Once fanned, the ventral part of the bulb "may assist the m. lateralis caudae in maintaining the outspread condition of the rectrices" (Baumel 1988, p. 82). Based on bulb morphology, it was hypothesized that contraction of the bulbi rectricium muscles distorts each bulb and squeezes the tips of the calami together. In conjunction with the shortening lateralis caudae, each rectrix would rotate about an axis distal to its tip, thus causing tail fanning. Fanning is opposed by elastic interrectricial ligaments, passive bulb recoil, and small adductor rectricium muscles. These are all thought to contribute to returning the tail to its resting, furled position (Baumel 1988).
One goal of our study was to test these hypotheses of rectricial abduction. We hoped that a better understanding of tail fanning in modern birds would shed light on the functional significance of the morphological transition from frond to fan and the evolution of short-tailed birds.
MATERIALS AND METHODS
Subjects, Training, Anatomy, and Nomenclature
Data were collected from adult pigeons (C. livia) of both sexes and male domestic turkeys (M. gallopavo). Five adult wild-type pigeons (mean body mass 350 g [+ or -] 21 g, SD) were taken from populations in Missoula, Montana; one domestic Show Racer (480 g) raised at the Bowman Gray School of Medicine was also used. All were housed in stainless-steel cages. Additional information was obtained from six adult male "Bourbon Red" strain domestic turkeys (mean body mass 6.5 kg [+ or -] 1.3 kg, SD) purchased from local breeders and housed in cages. All birds were provided with commercial feed and water ad libitum.
Wild-type pigeons were trained to fly within a flight corridor (minimum dimensions: 50 m long x 2.4 m wide x 2.45 m high) to a 1.15 m high perch platform after an unassisted liftoff from the ground. These birds sequentially performed three flight modes (takeoff, slow level flapping flight, and landing) between ground and perch (Gatesy and Dial 1993). The Show Racer and turkeys were trained to ascend steeply from the ground to a 1.75 m high perch platform. Turkeys also made short flights (4.5 m) between 1.75 m and 1.4 m high perches.
The caudal skeleton of modern flying birds can be divided into three distinct regions. During embryonic development, proximal caudal vertebrae become surrounded by the ilia and incorporated into the synsacrum. These segments, although homologous with caudal vertebrae in other tetrapods, are functionally part of the pelvic region. The remainder of the caudal vertebrae make up the functional tail skeleton. Henceforth, we will use the term "caudal" and "tail" to designate only these vertebrae distal to the synsacrum. The morphology of the tail and pelvis was studied in fresh, frozen, and preserved specimens (pigeons N = 10, turkeys N = 5) using binocular dissecting microscopes. The nomenclature of muscles in this study follow Baumel (1988) and Vanden Berge and Zweers (1993).
Electromyographic (EMG) data were recorded using surgically implanted bipolar electrodes as previously described (Gatesy and Dial 1993). Deep anesthesia was induced by intramuscular doses of ketamine and xylazine during all surgical procedures. Feathers were removed from the back and tail region; the rectrices and major coverts remained intact. Fine-wire electrodes were run subcutaneously from a back plug, which was sutured to thoracic interspinous ligaments. After muscles were exposed by blunt dissection, electrodes were implanted with a 23-gauge hypodermic needle and then sutured to adjacent fascia. All experiments were performed under National Institutes of Health (NIH) guidelines and monitored by the University of Montana and Wake Forest University Animal Care and Use Committees.
EMG recordings were made within 3 d after surgery. A light weight, 12-lead shielded cable was positioned around the tail base to permit free caudal movement. Signals were amplified and filtered (1000-5000x; 100-3000 Hz bandpass) before being recorded. Wild-type pigeon EMGs were converted to digital data (5000 Hz sampling rate) and stored on a microcomputer. Racer and turkey EMGs were either digitally converted or saved as analog signals on a TEAC MR-40 nine channel cassette data recorder. Selected tape sequences were then digitized as above for analysis. Data were played out on a Graphtec WR4000 thermal array recorder.
Recordings for this study were made from two caudal hypaxial muscles, the lateralis caudae (pigeons: 8 electrodes, 4 birds; turkeys: 10, 6) and the bulbi rectricium (pigeons: 9, 5; turkeys: 9, 5), and from the pectoralis (pigeons: 10, 5; turkeys: 1, 1). More specifically, electrodes were implanted in the ventral fibers of the bulbi rectricium and the sternobrachial head of the pectoralis. Electrode positions were verified by postmortem dissection following recording.
Muscles of deeply anesthetized birds were directly stimulated through implanted bipolar electrodes (100-[[micro]meter] diameter silver wire with 5-mm uninsulated tips). Although this technique is fraught with problems (see Loeb and Gans 1986), nerve stimulation is infeasible given the segmental innervation pattern. The bulbi rectricium in the pigeon is typically innervated by seven separate branches of the caudal plexus (Baumel 1988). Trains of pulses (1-4 volts, 0.1 ms duration, 200 Hz) were delivered with either a Grass SD9 or Grass S88 stimulator. The voltage was maintained at a level adequate for significant contraction of the target muscle without causing adjacent muscles to contract. In two pigeons and two turkeys, superficial muscles were sequentially cut and removed to assess the action of deeper muscles. Birds were euthanized at the end of each experiment. Several stimulation trials were recorded on videotape.
Cinematography and Video
High-speed films (150-200 frames/sec) and EMGs of wild-type pigeon were recorded simultaneously as in Gatesy and Dial (1993). An L-W Motion Analyzer projector was used to view wing and tail position. Racer and turkey flights were recorded on video. An S-VHS camera, in conjunction with a digital mixer, time code generator, and S-VHS recorder were used to collect sequences at 60 fields/sec. A voltage pulse that lit an LED positioned within the frame was used to correlate movements and muscle activity.
Wild-type pigeon caudal EMGs were correlated to the upstroke and downstroke phases of the wing-beat cycle using the high speed films. Activity periods of the lateralis caudae and bulbi rectricium have been measured for takeoff, slow level flight and landing (Gatesy and Dial 1993) and were not requantified here. Videos were analyzed to correlate muscle activity with the wing-beat cycle in turkeys and the Show Racer. To assess the contributions of each muscle to fanning, EMG intensity was compared with tail span as birds ascended from the ground to the perch (turkeys: seven flights, two birds; Show Racer: four flights). Single fields of video were captured for measurement of tail span on a Macintosh computer using NIH Image (Wayne Rasband, NIH; available from the Internet by anonymous ftp from zippy.nimh.nih.gov). Signals were rectified and integrated before being divided into 40 ms bins for evaluation of intensity fluctuation. Tail-span data was then synchronized ([+ or -] 16.67 ms) with the EMG data using LED correlation pulses.
Bird tails have been modeled as thin flat delta-shaped wings and their aerodynamic performance predicted using slender lifting surface theory (Balmford et al. 1993; Thomas 1993). These workers calculated lift using the following equation:
L = [Pi]/4 [Rho][Alpha][U.sup.2][b.sup.2]max
Lift (L) is proportional to factors such as the air density ([Rho]), the angle of attack ([Alpha], the tail's angular deviation, in radians, from the incoming airstream), and the square of airstream velocity (U). More importantly, a spread tail produces a lifting force proportional to the square of its widest continuous span ([b.sub.max]; Thomas 1993). Tail fanning adjusts tail span, and thus directly influences lift production. Therefore, at any reasonable combination of airspeed and angle of attack, lift will depend solely on the degree of fanning (Thomas 1993).
Lift was calculated for a pigeon tail with 14-cm long rectrices operating at a range of speeds (5-10 m/sec), angles of attack (0 [degrees]-25 [degrees]) and fan angles (0 [degrees]-180 [degrees]) using the lift equation in Thomas (1993). Similar calculations were made for the tail of Archaeopteryx. We made no effort to set limits on Archaeopteryx performance; lift calculations were used only to compare lift production between pigeon and Archaeopteryx tails at the same flight conditions. This required several assumptions. Most critically, it must be postulated that the serially arranged rectrices in Archaeopteryx form a flight surface that can be modeled as a flat plate. Direct usage of the lift equation may not be valid. Of particular concern is the possibility that the rectrices were not held as planar as in modern birds. A second consideration is tail shape. Calculations were done using the tail of the London specimen of Archaeopteryx as fossilized (de Beer 1954). Although the tail appears well preserved, changes in rectricial position during death and burial cannot be ruled out (see Discussion). Finally, pigeon flight speeds and angles of attack were used for Archaeopteryx, some of which may not have been possible in this primitive flier. Despite these caveats, the relationship between tail span and lift is robust enough to substantiate the significance of tail-fanning ability in modern birds and to justify its cautious application to Archaeopteryx. A complete analysis of Archaeopteryx tail aerodynamics is beyond the scope of this study.
Results from stimulation experiments on anesthetized birds help clarify the roles of the bulbi rectricium and lateralis caudae in tail fanning. In both pigeons and turkeys, contraction of the lateralis caudae causes elevation and slight abduction of the outer rectrices. Stimulations of the lateralis caudae never produced substantial fanning. Unilateral abduction is typically on the order of 5 [degrees]. In contrast, stimulation of the pigeon bulbi rectricium rapidly spreads the rectrices up to approximately 90 [degrees] (ca. 180 [degrees] bilaterally). A continuous pulse train maintains the tail in a fanned position. Lower stimulus voltages, fatigue, and suboptimal electrode implants produce smaller fan angles. In the much larger turkey bulb, stimulation was more difficult and often resulted to localized contraction. In all cases, rectricial abduction occurred, but full fanning was more elusive.
The autonomy of the bulbi rectricium in tail fanning is obvious when potential synergists are eliminated. Sequential removal of muscles inserting on the bulb does not reduce rectricial abduction by the stimulated bulbi rectricium in pigeons or turkeys. Even after all other tail muscles are excised, the bulbi rectricium can still initiate and maintain a tail fan. Thus, the bulbi rectricium is capable of controlling all phases of tail fanning entirely independently [ILLUSTRATION FOR FIGURE 2D OMITTED]. Additional muscles are not necessary to initiate, produce, or sustain a tail fan.
Once spread by the stimulated bulbi rectricium, the fan forms a remarkably rigid flight surface. Manual elevation, depression, rotation, and lateral deviation of the caudal skeleton have no significant effect on fan angle or shape. If the pulse train is terminated, however, the rectrices are quickly adducted. Folding was somewhat slower than fanning and did not always return the tail to its natural resting position. This may imply that contraction of the small adductor rectricium muscles normally assist complete furling.
Fanning in Free Flight
Activity patterns of the lateralis caudae and bulbi rectricium differ dramatically. In freely flying pigeons the lateralis caudae exhibits phasic activity coupled to the wing-beat cycle ([ILLUSTRATION FOR FIGURE 3A OMITTED]; Gatesy and Dial 1993). The lateralis caudae fires twice per wing beat during takeoff, once at the upstroke-downstroke transition and again in late downstroke. After gaining altitude and forward velocity, the muscle fires once per wing beat during the upstroke and upstroke-downstroke transition. This level flapping flight EMG pattern is retained during landing. In contrast, EMGs reveal that the bulbi rectricium is active throughout the flight [ILLUSTRATION FOR FIGURE 3B OMITTED]. Fluctuations in EMG amplitude occur, but there are no gaps in activity as seen in the lateralis caudae. A similar pattern is seen in turkeys making short flights between perches; lateralis caudae EMG is phasic, whereas the bulbi rectricium shows constant activation. During both pigeon and turkey flight sequences, the tail fans rapidly during the first wing beat (liftoff) and remains widely fanned until it furls after alighting on the perch [ILLUSTRATION FOR FIGURE 4 OMITTED]. Although the constant EMG of the bulbi rectricium appears to explain this kinematic pattern better, it is possible that punctuated activation of the lateralis caudae could produce a force more constant than the EMG might suggest.
The relationship between EMG activity and fan size was analyzed to ascertain which muscle or muscles are responsible for tail fanning. Caudal muscle activity patterns and tail span were quantified for turkeys ascending vertically to a perch. As in longer horizontal flights, the bulbi rectricium EMG is continuous, whereas the lateralis caudae EMG is phasic [ILLUSTRATION FOR FIGURE 4B OMITTED]. Tail span begins at an initial furled level before rising to its fully fanned size in ca. 130 ms. Once open, the fan fluctuates in width with each wing beat before closing. Birds would often elevate and abduct the rectrices upon landing to help balance on the perch.
The intensity of EMG activity in the lateralis caudae and bulbi rectricium was quantified and divided into 40 ms bins [ILLUSTRATION FOR FIGURES 4C, D OMITTED]. Following an initial high level of activity, the bulbi rectricium EMG intensity decreases and undergoes fluctuations with each wing beat. When compared with tail span, bulbi rectricium EMG intensity is found to correlate well with the timing and degree of tail fanning. Peaks in EMG intensity precede peaks in span by 30-50 ms. When tail span is shifted forward in time by 30-50 ms (presumably a combination of electromechanical and inertial delays) and rescaled, a remarkably good fit can be found with bulbi rectricium EMG intensity [ILLUSTRATION FOR FIGURE 4D OMITTED]. Given the complexities of activation history and changes in contraction velocity, a perfect relationship is not expected. Yet, despite these caveats a direct relationship between bulbi rectricium activity and fan size can be inferred. The cessation of bulbi rectricium EMG does not cause immediate furling of the fan. Presumably the bulbi rectricium continues to produce force subsequent to EMG activity. These results have been duplicated in an ascending pigeon. Unlike the bulbi rectricium, lateralis caudae EMG intensity correlates poorly with fan size [ILLUSTRATION FOR FIGURE 4C OMITTED]. Major bursts of activity do not always translate into increased rectricial abduction. Similarly, major changes in fan size are not always accompanied by similar fluctuations in lateralis caudae intensity. This is particularly true in the last wing beat of an ascent, when the lateralis caudae is often silent while the tail continues to fan.
Tail Fanning and Lift
The relationship between lift and rectricial spreading is best shown by fanning the tail to different degrees under identical flight conditions. For example, consider a pigeon tail in an airstream of 5m/sec held at a 15 [degrees] angle of attack. If fully furled, no lift would be produced (Thomas 1993). By slightly fanning to an 8-cm span, the tail now creates approximately 0.04 N of lift. Maximum abduction of the rectrices to a 26-cm span would yield 0.42 N of lift. This tripling of the tail span increases lift by an order of magnitude. Because lift is proportional to span squared, this relationship holds for any reasonable combination of flight speed and angle of attack.
The London Archaeopteryx has a tail span of approximately 9 cm. If held at a 15 [degrees] angle of attack in an airstream of 5m/sec, it would produce 0.05 N of lift. This is only 12% of the lift from the fully fanned tail of a pigeon of comparable body size. Note that changing flight speed or angle of attack alters the absolute lift values for each bird but that a fully fanned pigeon tail will always provide more than eight times the lift of an Archaeopteryx tail under the same conditions.
Prior to Baumel's (1988) comprehensive treatise on the pigeon tail, the morphology and functional significance of the rectricial bulb were almost completely unknown. Experimental studies of muscle activity during locomotion (Gatesy and Dial 1993), respiration (Baumel et al. 1990), and display (Kilpatrick and Gatesy 1994) have provided additional data to help understand the functional interaction of tail structures, but only in two species. Interspecific variation in bulb morphology may correlate with tail use and flying ability (Baumel 1988) but is also related to the number and size of the rectrices (Gatesy, unpubl. obs.). However, preliminary surveys of representatives of various orders have revealed that the basic organization of the tail apparatus is remarkably conservative in all but the flightless ratites (Baumel 1988; Gatesy, unpubl. obs.). Further investigation into avian bulb diversity is warranted, but it is likely that functional mechanisms shared by the pigeon and turkey are similar in most modern flying birds. In the following sections, we will use our new data to analyze caudal function in these living forms, speculate on tail mechanisms in Archaeopteryx and discuss avian tail evolution.
Interpretation of Muscle Function
Our results reveal that pigeons and turkeys control tail-fan size by graded activation of just one muscle, the bulbi rectricium [ILLUSTRATION FOR FIGURE 2D OMITTED]. Data from studies of EMG activity in non-turning (Gatesy and Dial 1993) and turning pigeon flight (Gatesy and Dial 1992, unpubl. manuscript), EMG activity during turkey flight and sexual display (Kilpatrick and Gatesy 1994), and muscle action through electrical stimulation all support a nonfanning role for the lateralis caudae. Asymmetrical activation during turning and display is consistent with our interpretation of the lateralis caudae as a muscle for holding or moving the fan rather than producing it. When active unilaterally or asymmetrically, the lateralis caudae force should shift the fan laterally and tilt the fan by lifting its leading edge. Symmetrical bilateral activity would primarily elevate the entire fan. We hypothesize that the pulsatile activity of the lateralis caudae during flight is related to adjusting pitch, particularly during takeoff, to help stabilize the body as it experiences dramatic forces from the wings.
Our conclusions are based on data from electrical stimulation of anesthetized birds and from EMG recordings of freely flying birds. Additional insight might come from flying birds after removing, bisecting, denervating, or anesthetizing specific muscles such as the lateralis caudae. We intentionally avoided such manipulations of the caudal system in hopes of reaching conclusions from intact birds. The inherent problems of compensation, learning, and localization of anesthetic detract from the apparent simplicity of studying muscular complexes by impairing components of the system. We feel our results are persuasive, but others may wish to test our interpretations by these means.
Potential Benefits of the Bulb Fanning Mechanism
Changes in bulbi rectricium activity produce rapid alterations in fan size. Because lift is proportional to the square of tail span, this mechanism allows birds to exploit a wide range of potential lift forces. Most modern birds appear to take advantage of this flexibility. Specific combinations of fan size and fan position are used to adjust the magnitude and direction of the tail lift vector during different modes of flight. Birds in gliding (Pennycuick 1968; Tucker 1992) and slow flapping flight (Spedding et al. 1984; Gatesy and Dial 1993) modulate tail spread and angle of attack to control pitch (Thomas 1993). In high speed, nonturning flight in which caudal lift is not needed, the tail is typically furled to minimize drag (Pennycuick 1975; Baumel 1988). During landing, tail movements maximize lift to prevent stalling (Thomas 1993) and may increase wing performance (Pennycuick 1975). Rotations, lateral deflections, and other deviations of the tail are observed in soaring birds, presumably to fine-tune roll and yaw. Tail forces are also thought to counteract sideslip and adverse yaw when banking (Hummel 1992; Thomas 1993). Such forces likely are produced by the asymmetrical activation of caudal muscles, as recorded in turning pigeons (Gatesy and Dial 1992, in prep.). Of course, minute adjustments of the tail are common in all of these flight modes.
A second, and perhaps equally important, feature of the bulbi rectricium fanning mechanism is its independence. The bulbi rectricium controls only the position of the rectrices with respect to one another. The remaining caudal muscles function to elevate, depress, rotate, laterally deviate, or stabilize the variable-sized tail fan with respect to the body. This division of labor may have significant benefits for control of the caudal flight surface. Because the rectrices and bulb attach only to the pygostyle, the tail fan is isolated on this terminal bone. The pygostyle, bulb, and rectrices thus form an anatomical and functional unit (Baumel 1988). Spreading of the rectrices is accomplished without directly affecting other caudal segments. Conversely, movement of intervertebral joints need not disrupt tail fanning. Deflections of the caudal vertebrae can displace the tail fan without deforming it. Independence of fanning and vertebral movement has been proposed by Raikow (1985). However, he invoked the lateralis caudae in his rectricial-abduction mechanism. Because this muscle crosses each intervertebral joint, its hypothesized role in fanning would preclude mechanical independence. Our data indicate that the lateralis caudae is neither capable of nor necessary for fanning, making the rectricial bulb fanning mechanism significantly more independent. In modern flying birds, therefore, tail fanning is relatively decoupled from fan movement. The decoupling of primitively linked components may play a significant role in the evolution of complex systems (Young 1938; Lauder 1981, 1989; Lauder and Liem 1989; Gatesy and Dial 1996).
Functional Implications for Archaeopteryx
Our data from living birds can be used to make inferences about tail structure and control in Archaeopteryx. First, it is highly implausible that Archaeopteryx could fan its tail in the same manner as modern flying birds. Each pair of rectrices attached, presumably by a ligament, to a caudal vertebra ([ILLUSTRATION FOR FIGURE 1C OMITTED]: Owen 1862; Steiner 1938; de Beer 1954). Abduction, if any, must have occurred by rotation of each rectrix about an axis at or near its tip. Because the tips of adjacent rectrices could not be approximated, a bulbi rectricium fanning mechanism was not operational. In all likelihood, there was no rectricial bulb.
Second, the tail feathers must have been firmly connected to the caudal skeleton and one another to prevent unfavorable deflection, separation, or collapse during flight. The tail, when present, is always preserved as a cohesive unit, even in the London specimen in which other parts of the body are disarticulated [ILLUSTRATION FOR FIGURE 1A OMITTED]. The rectrices themselves appear relatively undisturbed in the London, Berlin and Solenhofer Aktien-Vereins specimens (de Beer 1954; Wellnhofer 1993; pers. obs.). Archaeopteryx tail fronds are all preserved dorsoventrally; the right and left feathers are never folded over like the pages of a book. When the pelvis lies on its side, the distal caudal skeleton is sometimes rotated into a dorsoventral plane (Berlin and Solnhofen specimens; Wellnhofer 1988). Alternatively, the vertebrae can be preserved laterally (Eichstatt and Solenhofer Aktien-Vereins specimens; Wellnhofer 1974, 1993), but the rectrices are dorsoventrally flattened, not folded. Extensive connective tissue reinforcement must have maintained tail integrity and prevented the rectrices from forming a dramatic dihedral. This rigidity, however, probably limited rectricial movement.
Third, motion of the caudal vertebrae was linked to rectricial movement. The frond-shaped tail of Archaeopteryx is organized much like a single, large feather; the rectrices were bound to a flexible, multisegmented skeletal shaft. Although vertebral mobility appears greatest proximally, most specimens show gentle curving of the distal caudals. Any deflection of the caudal axis not restricted to the most proximal vertebrae must have distorted the tail fan. Conversely, muscles that abducted the rectrices, if present, would have affected intervertebral joints. Each vertebra could have had its own abductor muscles connected only to the two rectrices it bore, but such a system would make coordinated movement of all rectrices highly unlikely. If the feathers could be abducted, muscles and tendons spanning many segments are much more plausible. Together with evidence for connections between adjacent rectrices, such muscles would make it impossible to abduct the feathers without influencing the position of tail vertebrae or requiring compensatory activity in other muscles. Thus, the muscular control of vertebral and rectricial movement was tightly coupled, rather than being independent as they are in modern birds.
These tail properties would have important consequences for the overall aerodynamic performance of Archaeopteryx. Of great significance is the likelihood that changes in tail span were minor. A rudimentary abduction mechanism and limited rectricial mobility probably constrained the range of lift forces produced by the tail. Even if fully spread, which is untenable, the frondlike arrangement produces a span that is small relative to the fan of an aerobatic modern bird of similar body size [ILLUSTRATION FOR FIGURE 1 OMITTED]. The longest rectrices on the London specimen are only 40% to 45% of the total tail length (Owen 1863; de Beer 1954). In contrast, pigeon and turkey rectrices make up 75% to 90%, forming a fan almost twice as wide as long when fully abducted.
The tail's contribution to pitch control during flight was limited primarily to changing its angle of attack. If tail fanning was limited, lift production and the position of the center of lift would have been relatively fixed at any speed and angle of attack. The long, low aspect ratio tail had a center of lift with a large moment to rotate the body when elevated or depressed (Caple et al. 1983). Such an arrangement is more stable than the more proximally located lift center of most modern birds (Maynard Smith 1952).
Maneuverability was doubtless more limited and inefficient than in most modern birds. This stems from many factors. First, forelimb morphology indicates that a modern wing-beat cycle was not present in Archaeopteryx (e.g., Dial 1992; Vasquez 1992; Jenkins 1993). Nonsteady flight may have been extremely difficult and aerobatic ability highly restricted. Next, as stated above, the tail could produce only a limited range of lift forces, thus reducing its ability to assist the wings. Finally, the mechanical linkage between vertebral and rectricial movement in Archaeopteryx probably hampered coordinated turning. Movements of the caudal axis probably deformed the tail surface. In living birds, natural or induced fan asymmetry is known to affect turning performance dramatically (Moller 1991). Asymmetry skews the lift force and requires compensatory wing and tail movements, leading to increased cost (Thomas 1993).
Bird Tail Evolution
A transformation from frond to fan bestows two primary advantages: (1) A variable-sized fan can modulate lift forces dramatically. Modern flying birds alter tail span with flight mode but also adjust fan size extremely rapidly during acute braking and turning. A frond-shaped tail would produce lower, more consistent forces and would have to be depressed or elevated to alter lift substantially. (2) Isolation of the fan on the pygostyle effectively decouples the fanning mechanism from other tail movements. Fan symmetry can be maintained easily without deformation by muscles that position the fan relative to the body. In contrast, when rectricial and vertebral movements are coupled, any bend in the frond's axis will distort the flight surface.
How did the transition from long- to short-tailed birds take place? The disparity between Archaeopteryx and modern birds is relatively large, yet can be attributed to three processes: shortening of caudal centra, sequestering of proximal caudals into the synsacrum, and fusing of distal caudals to form a pygostyle. These changes produced a functional tail skeleton of diminutive size, with only a minor reduction in the total number of caudal segments ([ILLUSTRATION FOR FIGURE 5 OMITTED]; Marshall 1872; van Oort 1904; Steiner 1938). The sequence of these modifications has remained elusive because Late Cretaceous bird fossils, such as Hesperornis and Ichthyornis, already possess a reduced tail (Marsh 1880). Intermediate stages were purely hypothetical prior to recent discoveries of Early Cretaceous birds with more primitive tail skeletons (Sereno and Rao 1992; Zhou et al. 1992; Sanz et al. 1988; Sanz and Bonaparte 1992). Iberomesornis (Sanz et al. 1988; Sanz and Bonaparte 1992) is in many ways intermediate between Archaeopteryx and birds with modern tail morphology [ILLUSTRATION FOR FIGURE 5 OMITTED]. There are eight free caudal vertebrae and an extremely large pygostyle, estimated to be composed of 10 to 15 fused vertebrae (Sanz and Bonaparte 1988, 1992). Sinornis (Sereno and Rao 1992) and Cathayornis (Zhou et al. 1992) also have compressed caudal centra, eight free caudals, and a pygostyle.
If our interpretation of modern bird tails is correct, the acquisition of a pygostyle would have profound functional significance. First, the coalescence of distal vertebrae may have been the initial step in the evolutionary insulation of the rectrices from axial bending. Fusion of segments would provide a rigid foundation for the rectrices and foster symmetry of the tail surface in that region. Second, the presence of a pygostyle may indicate the appearance of a rectricial bulb; Baumel (1988) suggested that these structures coevolved. Although the feather configuration is not known, calami of the distal rectrices could have been freed from their primitive segmental connection. The organization of soft tissues into an incipient rectricial bulb may have afforded Iberomesornis and other early birds a greater range of lift forces independent of other caudal movements.
In modern birds that lack a pygostyle (most ratites and some tinamous), the rectrices are either highly reduced or absent. The skeletal axis, however, is still short; no modern birds have reelongated the caudal vertebrae to resemble Archaeopteryx. The recently described bird Mononykus, a flight-less Late Cretaceous form ([ILLUSTRATION FOR FIGURE 5A OMITTED]; Perle et al. 1993, 1994), is particularly intriguing. The distal tail is not known, but the 19 long caudal vertebrae are unlike any bird other than Archaeopteryx. If Mononykus is the sister group of all birds other than Archaeopteryx as proposed, it suggests that tail reduction took place after several changes in the hind limb and pelvis (Perle et al. 1993). Alternatively, if Mononykus is a degenerate member of an advanced flying clade, it represents the only known example of dramatic tail reelongation among birds. This example points out the need for more information about the early evolution of birds and avian flight.
We thank R. Trenary, B. Tobalske, N. Olsen, D. Warrick, J. Kilpatrick, C. Simpson, and J. Rodgers for their assistance in this project. Discussions with J. Baumel, J. Ostrom, G. E. Goslow Jr., F. A. Jenkins Jr., and A. Thomas were very helpful. This work was supported by National Science Foundation grant IBN-92-03275 and a Wake Forest University RE-CREAC award to S.M.G. and grants BNS-89-08243 and IBN-92-11393 to K.P.D.
ALLEN, T. T. 1962. Myology of the limpkin. Ph.D. diss., Univ. of Florida, Gainesville.
BALMFORD, A., I. L. JONES, AND A. L. R. THOMAS. 1993. On avian asymmetry: Evidence of natural selection for symmetrical tails and wings in birds. Proc. R. Soc. Lond. B Biol. Sci. 252:245-251.
BAUMEL, J. J. 1971. Morphology of the tail apparatus in the pigeon (Columba livia). Proc. XIX World Vet. Congr. 3:849.
-----. 1988. Functional morphology of the tail apparatus of the pigeon (Columba livia). Adv. Anat. Embryol. Cell Biol. 110:1-115.
BAUMEL, J. J., J. A. WILSON, AND D. R. BERGREN. 1990. The ventilatory movements of the avian pelvis and tail: Function of the muscles of the tail region of the pigeon (Columba livia). J. Exp. Biol. 151:263-277.
BEEBE, C. W. 1915. A tetrapteryx stage in the ancestry of birds. Zoologica 2:39-52.
DE BEER, G. 1954. Archaeopteryx lithographica. Brit. Mus. (Nat. Hist.), Lond. Publ. 224:1-68.
-----. 1956. The evolution of ratites. Bull. Brit. Mus. (Nat. Hist.) 4:59-70.
BOCK, W. J. 1986. The arboreal origin of avian flight. Mem. Calif. Acad. Sci. 8:57-72.
BURT, W. H. 1930. Adaptive modifications in the woodpeckers. Univ. Calif. Publ. Zool. 32:455-524.
CAPLE, G., R. P. BALDA, AND W. R. WILLIS. 1983. The physics of leaping animals and the evolution of preflight. Am. Nat. 121: 455-476.
CARROLL, R. L. 1988. Vertebrate paleontology and evolution. Freeman, New York.
CHIAPPE, L. M. 1995. The first 85 million years of arian evolution. Nature 378:349-355.
CRACRAFT, J. 1986. The origin and early diversification of birds. Paleobiology 12:383-399.
DIAL, K. P. 1992. Activity patterns of the wing muscles of the pigeon (Columba livia) during different modes of flight. J. Exp. Zool. 262:357-373.
FISHER, H. I. 1946. Adaptations and comparative anatomy of the locomotor apparatus of New World vultures. Am. Midl. Nat. 35: 545-727.
GATESY, S. M., AND K. P. DIAL. 1992. How pigeons turn: Asymmetrical caudal muscle activity during maneuvering flight. Am. Zool. 32:143A.
-----. 1993. Tail muscle activity patterns in walking and flying pigeons (Columba livia). J. Exp. Biol. 176:55-76.
-----. 1996. Locomotor modules and the evolution of avian flight. Evolution 50:331-340.
HEILMANN, G. 1926. The origin of birds. Witherby, London.
HUMMEL, D. 1992. Aerodynamic investigations on tail effects in birds. Z. Flugwissenschaften Weltraumforschung 16:159-168.
JENKINS, F. A. J., JR. 1993. The evolution of the avian shoulder joint. Am. J. Sci. 293:253-267.
KILPATRICK, J. E, AND S. M. GATESY. 1994. Flight versus display: The functional design of avian tail muscles. Am. Zool. 34:107A.
LAUDER, G. V. 1981. Form and function: Structural analysis in evolutionary morphology. Paleobiology 7:430-442.
-----. 1989. Group report - How are feeding systems integrated and how have evolutionary innovations been introduced? Pp. 97-115 in D. B. Wake and G. Roth, eds. Complex organismal functions: Integration and evolution in vertebrates. Wiley, New York.
LAUDER, G. V., AND K. F. LIEM. 1989. The role of historical factors in the evolution of complex organismal functions. Pp. 63-78 in D. B. Wake and G. Roth, eds. Complex organismal functions: Integration and evolution in vertebrates. Wiley, New York.
LOEB, G. E., AND C. GANS. 1986. Electromyography for Experimentalists. Univ. of Chicago Press, Chicago.
LOWE, P. R. 1944. An analysis of characters of Archaeopteryx and Archaeornis. Were they reptiles or birds? Ibis 86:517-543.
MARSH, O. C. 1880. Odontornithes, a monograph of the extinct toothed birds of North America. Rep. Geol. Explor. 40th Parallel 7:1-201.
MARSHALL, W. 1872. Beobachtungen uber den vogelschwanz. Niederlandishes Archiv fur Zoologie 1:194-210.
MAYNARD SMITH, J. 1952. The importance of the nervous system in the evolution of animal flight. Evolution 6:127-129.
MOLLER, A. P. 1991. Sexual ornament size and the cost of fluctuating asymmetry. Proc. R. Soc. Lond. B Biol. Sci. 243:59-62.
VAN OORT, E. D. 1904. Beltrag zur osteologie des vogelschwanzes. Tijdschr. Ned. dierk. Vereen. 9:1-144.
OSTROM, J. H. 1976. Archaeopteryx and the origin of birds. Biol. J. Linn. Soc. 8:91-182.
OWEN, R. 1862. On the fossil remains of a long-tailed bird (Archaeopteryx macrurus, Ow.) from the lithographic slate of Solenhofen. Proc. R. Soc. Lond. B Biol. Sci. 12:272-273.
-----. 1863. On the Archaeopteryx of von Meyer, with a description of the fossil remains of a long-tailed species, from the lithographic stone of Solenhofen. Phil. Trans. R. Soc. Lond. B Biol. Sci. 153:33-47.
OWRE, O. T. 1967. Adaptations for locomotion and feeding in the Anhinga and the Double-crested Cormorant. Ornithol. Monogr. 6:1-138.
PENNYCUICK, C. J. 1968. Power requirements for horizontal flight in the pigeon Columba livia. J. Exp. Biol. 49:527-555.
-----. 1975. Mechanics of flight. Pp. 1-75 in D. S. Farner and J. R. King, eds. Avian biology. Vol. 5. Academic Press, New York.
PERLE, A., M. A. NORELL, L. M. CHIAPPE, AND J. M. CLARK. 1993. Flightless bird from Cretaceous of Mongolia. Nature 362:623-626.
PERLE, A., L. M. CHIAPPE, R. BARSBOLD, J. M. CLARK, AND M. A. NORELL. 1994. Skeletal morphology of Mononykus olecranus (Theropoda: Avialae) from the Late Cretaceous of Mongolia. Am. Mus. Novit. 3105:1-29.
PROCTOR, N. S., AND P. J. LYNCH. 1993. Manual of ornithology: Avian structure and function. Yale Univ. Press, New Haven, CT.
RAIKOW, R. 1970. Evolution of diving adaptations in the stifftail ducks. Univ. Calif. Publ. Zool. 94:1-52.
-----. 1985. Locomotor system. Pp. 57-147 in A. S. King and J. McLelland, eds. Form and function in birds. Vol 3. Academic Press, London.
SANZ, J. L., AND J. F. BONAPARTE. 1992. A new order of birds (Class Aves) from the Lower Cretaceous of Spain. Nat. Hist. Mus. Los Angeles Co. Sci. Ser. 36:39-49.
SANZ, J. L., J. F. BONAPARTE, AND A. LACASA. 1988. Unusual Early Cretaceous birds from Spain. Nature 331:433-435.
SERENO, P. C., AND C. RAO. 1992. Early evolution of avian flight and perching: New evidence from the Lower Cretaceous of China. Science 255:845-848.
SPEDDING, G. R., J. M. V. RAYNER, AND C. J. PENNYCUICK. 1984. Momentum and energy in the wake of a pigeon (Columba livia) in slow flight. J. Exp. Biol. 111:81-102.
STEINER, H. 1938. Der "Archaeopteryx" - Schwanz der vogelem-bryonen. Vierteljahrsschrift der Naturf. Gesellschaft in Zurich 83:279-300.
THOMAS, A. L. R. 1993. On the aerodynamics of bird tails. Phil. Trans. R. Soc. Lond. B Biol. Sci. 340:361-380.
TUCKER, V. A. 1992. Pitching equilibrium, wing span and tail span in a gliding Harris' hawk, Parabuteo unicinctus. J. Exp. Biol. 165:21-41.
VANDEN BERGE, J. C., AND G. A. ZWEERS. 1993. Myologia. Pp. 189-247 in J. J. Baumel, A. S. King, J. E. Breazile, H. E. Evans, and J. C. Vanden Berge, eds. Handbook of avian anatomy: Nomina anatomica avium. Publ. Nuttall Ornithol. Club 23, Cambridge.
VASQUEZ, R. J. 1992. Functional osteology of the avian wrist and the evolution of flapping flight. J. Morphol. 211:259-268.
WAGNER, A. 1862. On a new fossil reptile supposed to be furnished with feathers. Annu. Mag. Nat. Hist. 3:261-267.
WELLNHOFER, P. 1974. Das funfte skelettexemplar von Archaeopteryx. Palaeontographica A 147: 169-216.
-----. 1988. Ein neues Exemplar von Archaeopteryx. Archaeopteryx 6:1-30.
-----. 1993. Das seibte exemplar von Archaeopteryx aus den Solnhofener schichten. Archaeopteryx 11:1-47.
YOUNG, J. Z. 1938. The evolution of the nervous system and of the relationship of organism and environment. Pp. 179-204 in G. R. de Beer, ed. Evolution, essays on aspects of evolutionary biology, presented to Professor E. S. Goodrich on his seventieth birthday, Clarendon Press, Oxford.
ZHOU, Z, F. JIN, AND J. ZHANG. 1992. Preliminary report on a Mesozoic bird from Liaoning, China. Chin. Sci. Bull. 37:1365-1368.
|Printer friendly Cite/link Email Feedback|
|Author:||Gatesy, Stephen M.; Dial, Kenneth P.|
|Date:||Oct 1, 1996|
|Previous Article:||Life-history variation and comparative phylogeography of darters (Pisces: percidae) from the North American central highlands.|
|Next Article:||Body size and harem size in male red-winged blackbirds: manipulating selection with sex-specific feeders.|