Fossils Feathers and Flight

We can use comparative anatomy to investigate the morphology of ancestors even when the ancestors are not preserved in the fossil record. Paleontologists employ a technique called phylogenetic bracketing (Witmer 1995, Bryant and Russell 1992). Despite its fancy name, the idea of phylogenetic bracketing is relatively simple. At its core, it presumes that any features shared by two or more related animals were also present in their last common ancestor. This hypothesis may be rejected if evidence to the contrary is discovered during the investigation of a specific case. With a phylogenetic bracket and some fossils, we can determine when certain features arose along the line to avian flight.

Using evidence from the fossil record, we now know that birds are living dinosaurs that descended from cursorial (running), bipedal predators (Ostrom 1974, 1976, 1979, 1997; Gauthier and Padian 1985; Gauthier 1986). The fossil record of the evolution of avian flight is extensive and constantly growing; in particular, we now have a detailed record showing how the skeletal and muscular systems were modified along the route to flight. What we see in this record is that all of the skeletal, ligamentous, and muscular features just discussed arose gradually along the lineage leading to birds. We can further infer that those features arose not for flight but in conjunction with predation (Gauthier and Padian 1985; Gishlick 2001b, 2001c).

Figure 5.2. A simplified phylogeny for birds, plotting the appearance of characters relevant to flight: (1) four-chambered heart; (2) four-fingered hand; (3) three-fingered hand, showing fused distal carpals 1 and 2 with a trochlea restricting the motion in a lateral plane; (4) filamentous feathers, longer hands; (5) enlarged semilunate carpal, with M. extensor metacarpi ulnaris inserting onto back of hand (automatic flexion), pennaceous feathers on hands and tail; (6) M. extensor metacarpi radialis insertion on flange of metacarpal 1 (automatic extension); (7) flight feathers; (8) alula; (9) fully integrated and irreducibly complex flight system.

Figure 5.2. A simplified phylogeny for birds, plotting the appearance of characters relevant to flight: (1) four-chambered heart; (2) four-fingered hand; (3) three-fingered hand, showing fused distal carpals 1 and 2 with a trochlea restricting the motion in a lateral plane; (4) filamentous feathers, longer hands; (5) enlarged semilunate carpal, with M. extensor metacarpi ulnaris inserting onto back of hand (automatic flexion), pennaceous feathers on hands and tail; (6) M. extensor metacarpi radialis insertion on flange of metacarpal 1 (automatic extension); (7) flight feathers; (8) alula; (9) fully integrated and irreducibly complex flight system.

To see how these features were assembled, paleontologists explore the origin of flight-related features in the context of the phylogeny. The phylogeny, or evolutionary relationships, of the organisms under study provides a framework on which to trace the order and evolutionary timing of the appearance of flight-related features (Padian 1982, 2001; Padian and Chiappe 1998b). Figure 5.2 shows a simplified phylogeny of birds and dinosaurs with the changes in the "irreducible" components of the flight system plotted when they occur. A myriad of detailed anatomical changes in the forelimb occur along this lineage; we will deal with only a few significant features here.

The theropod ancestors of birds were bipedal predators; they did not use their arms for locomotion but for seizing prey. This conclusion is based on numerous independent phylogenetic tests based on comparative anatomy (Ostrom 1976, 1979; Gauthier and Padian 1985; Gauthier 1986). This phy-logenetic context gives us an independent way to test the adaptive route to avian flight (Padian 1982, 2001; Padian and Chiappe 1998b). Any explanation for how the avian flight system arose requires an explanation of how theropod arms were used in a way that enabled them to acquire features that could then be exapted for flight. Basal theropods, those lower on the tree (see figure 5.2), such as Dilophosaurus or Coelophysis, had four-fingered hands and largely moved their hands in palmar-antipalmar direction (the direction we move our hands at the wrist when dribbling a basketball), as opposed to birds, who have three-fingered hands and move their wrists in a radial-ulnar direction (try moving your hands in the direction of the thumb or pinky). By comparison, the principal direction of manual movement in humans is in the palmar-antipalmar direction, with a high degree of pronation and supination: you can turn your hands more than 180 degrees, either palms up or palms down.

Birds and other theropods cannot pronate or supinate much, and they do not do so by crossing the radius and ulna as in humans. They hold their hand so that the plane of the hand is perpendicular to the ground. To move your hand like a bird's, hold your arms at your sides, bent at the elbows, with your hands out in front of you. Now turn your hands so that your palms face each other with your thumb pointing up and your pinky pointing down. If you were a bird, you could bend your hand at the wrist only in the direction of your thumb or the direction of your pinky, and you could not rotate your hands so that your palms can be up or down. You don't have a lot of range of motion in that axis (radial-ulnar), but birds and the theropods closest to them can move their wrists only along that axis.

The first flight-related feature that evolved on the line to birds is the trochlea carpalis of the wrist, which the radiale slides along. The evolution of the trochlea leads to the restriction of the wrist motion into a lateral (radial-ulnar) plane. The appearance of the trochlea along with the initial wrist motion restriction occurs at the base of the group called tetanurines (figure 5.2) and is accompanied by a reduction to three fingers, as seen in avians. This is first apparent in torvosaurs (Gauthier 1986); the trochlea carpalis becomes deeper in allosaurids (Gilmore 1915, Ostrom 1969, Madsen 1976, Chure 2001). A good example of this simple trochlea can be seen in theropods such as Allosaurus, where there is an enlarged first distal carpal fused to a small second distal carpal. Together these bear a shallow trochlea along which a somewhat enlarged radiale can slide (Madsen 1976, Gauthier 1986, Chure 2001). These bones enabled a degree of radial-ulnar transverse motion between the distal carpals and radiale. The distal carpal trochlea was not very deep, however, so it still allowed the ancestral dorsal-palmar motion between the car-pals. As you move up the tree (figure 5.2) toward birds, the distal carpals continue to enlarge, increasing the range of lateral motion in the wrist, along with a steady enlargement of the radiale and deepening of the trochlea. The arms are steadily lengthened, particularly the hand.

The next important feature occurs within a group called the Coelurosauria. It is here that the first feathers evolve. When feathers first appear, they are simple and hairlike, much like the natal down of chicks (Sumida and Brochu

2000; Prum and Brush 2002, 2003). These are first visible on the small compsognathid dinosaur Sinosauropteryx (Chen et al. 1998). The feathers most likely served as insulation, much as they do in juvenile birds today (Brush 2000; Prum and Brush 2002, 2003), rather than being used for flight. Phylogenetic bracketing indicates that all dinosaurs phylogenetically above Sinosauropteryx (that is, to the right of Sinosauropteryx in figure 5.2) would have inherited at least these simple feathers from a common ancestor. Imagine a fuzzy Tyranno-saurus rex, at least a juvenile one. (This was nicely illustrated in the November 1999 issue of National Geographic.) In basal coelurosaurs, the hand is longer, approaching the long hands of maniraptors and avians (Ostrom 1974, Gauthier 1986).

Many of the important changes in the forelimb occur within the maniraptors. In maniraptors, we first see a large semi-lunate surface of the fused distal carpals, as in avians, along with a deep trochlea and large triangular radiale. Further, we first see an avianlike range of motion for the wrist and forelimb. Such a range of motion is first seen clearly in theropods such as Oviraptor, but it was probably present in the more basal theropod Ornitholestes as well.

At this point in the phylogeny, we also see evidence of an EMU insertion on the dorsal surface of the second metacarpal, indicative of the automatic flexion system seen in birds (Gishlick 2001a). Thus, the automatic flexion of the forelimb may have appeared before automatic extension. Pennaceous feathers (which have a rachis and vanes, what we think of as true feathers) appear at this level in the phylogeny, as exhibited on the hands and tails of the oviraptorosaurs Caudipteryx and Protarchaeopteryx (Ji et al. 1998; for good pictures, see the July 1998 issue of National Geographic). All theropods following this point in the phylogeny, therefore, would have had true feathers on their hands and tails. These feathers are symmetrical, indicating that they were not used for flight: flight feathers are asymmetrical for aerodynamic efficiency.

True feathers evolved for some function other than flight, perhaps as a display feature or for brooding behaviors (Padian and Chiappe 1998a, 1998b; Prum and Brush 2002, 2003). Or perhaps the first feathers were employed as "blinders," causing the hands and arms to appear larger and thus prevent small prey from dodging to either side, or employed to herd prey (Gishlick 2001b). Feathers on the hands would have also had aerodynamic properties and may have aided in thrust during running (Burgers and Chiappe 1999; Burgers and Padian 2001) or traction while running up inclined surfaces (Dial 2003). Further, because of the way the feathers attach to the hand, they would not have greatly inhibited the grasping ability of the forelimbs (Gishlick 2001b, 2001c).

If you continue further into the Maniraptora, troodontids and dromaeosaurs

(such as Velociraptor and Deinonychus) evolve an enlarged flange on the first metacarpal for the insertion of the EMR, giving the first evidence for the automatic extension of the hand with the elbow in theropods (Gishlick 2001a). The very constrained motion of the forelimb at this point in the phylogeny could be regarded as a rudimentary form of the flight stroke previously detailed. Thus, the motions of avian flight and its muscular control evolved before flight and did not appear all at once, perhaps under selective pressure for grasping of prey (Gauthier and Padian 1985, 1989; Padian and Chiappe 1998a, 1998b; Gishlick 2001b, 2001c).

Archaeopteryx displays the first evidence of powered flight in dinosaurs, indicated by its asymmetrical feathers (Feduccia and Tordoff 1979, Rietschel 1985), large wing surface capable of supporting body mass (Padian and Chiappe 1998a, 1998b), and the capability of elevating the shoulder for the upstroke (Jenkins 1993). But even at this point, Archaeopteryx still maintained a functional grasping hand (Gishlick 2001b, 2001c). From this point on, however, selection for flight becomes stronger than selection for a grasping hand, and thus the flight system starts to become fine-tuned. By the time of Confuciusornis, we see the beginning of the fusion of the wrist elements into a carpometacarpus as well as the loss of a functional (for grasping) second digit (Chiappe et al. 1999). As the functional demands of flight increased, however, it became impossible to maintain dual use (Gishlick 2001b, 2001c). This transfer of function from grasping, through a grasping-flying stage, to pure flying is documented by the sequence of structures found in the fossils of Deinonychus, Archaeopteryx, Confuciusornis, and Eoalulavis, in that order. With the addition of an alula (Sanz et al. 1996)—one grasping finger is not particularly useful—the hand became used solely for flight, and the third digit was reduced and fused to the second thereafter. The alula modifies the flow of air over the wing to reduce turbulence. It allows birds to improve low-speed flying and gives them the maneuverability that make them the lords of the air. From here on, the avian flight system is something that can be considered irreducibly complex. Hence, we may conclude that irreducible complexity does not imply unevolvability.

By no means did the avian flight mechanism assemble all at once in its irreducible form. Rather, it was assembled piecemeal over millions of years and millions of generations. A grasping strike happened to produce thrust when the wing was large enough. The kinematics worked in the right way, so the feathered arm was exapted for flight rather than selected for flight originally. The historical record has enabled scientists to dissect such irreducible structures into reducible components and allows us to understand how supposedly irreducibly complex structures can evolve. With that in mind, we ask, what is the biological significance of irreducible complexity?

The answer, I think, is not much. Irreducible complexity would be biologically significant if the proponents of intelligent design were correct that it meant "unevolvable." But the example of the avian flight system contradicts their central claim. Irreducibly complex systems are evolvable, and that evolvability can easily be documented when a fossil record is available to put the structure into a historical context. I would bet that if a fossil record were available for the bacterial flagellum, we would see the same type of exaptation and mosaic evolution that we see in the avian-flight system. It's that simple.

When Behe (1996) proposed irreducible complexity, he treated it like a fundamental property of the biological world that evolutionary biologists had been intentionally or unintentionally not acknowledging since Darwin. That was not the case. Irreducibly complex systems have not been missed; they just were not considered insurmountable obstacles for evolution because of exaptation and mosaic evolution. The example of the avian flight system shows clearly that it is not evolutionary biologists who are missing something fundamental.

Was this article helpful?

0 0

Post a comment