Wonderful Life: The Burgess Shale and the Nature of History

by Stephen Jay Gould

  Bruton dashed the final hope for traditionalism by showing that Sidneyia could not be a close relative or ancestor of merostomes. The “merostomoid” body did not define a coherent evolutionary group, but a series of disparate creatures united only by what our jargon calls a symplesiomorphic (or “shared primitive”) trait. Shared primitive traits are ancestral for large groups, and therefore cannot define subgroups within the entire assemblage. For example, rats, people, and ancestral horses do not form a genealogical group within the mammals just because all have five toes. Five toes is an ancestral trait for Mammalia as a whole. Some creatures retain this initial condition; many others evolve modifications. The “merostomoid” body form is a shared primitive trait of many arthropods. True genealogical groups, by contrast, are based on shared derived characters—the unique specializations of their common ancestors.

  True chelicerates have six pairs of appendages, and no antennae, on their head shield. Sidneyia could not be more different in this crucial respect. Its head (figure 3.46) bears one pair of antennae, and no other appendages! Bruton came to regard Sidneyia as a curious mosaic of characters. The first four of nine body segments carry uniramous walking legs like those of merostomes. But the five posterior segments bear ordinary biramous appendages, with gill branches and walking legs. The “tail” piece, formed of three cylindrical segments and a caudal fan, looks more crustacean than merostomoid. Bruton found ostracodes, hyolithids, and small trilobites in Sidneyia’s gut, and interpreted the animal as a bottom-dwelling carnivore. But with no feeding appendages on the head, and a strong, tooth-lined food groove between the legs, Sidneyia presumably fed like most arthropods, by passing food toward the mouth from the rear, not by searching and grasping from the front.

  3.46. Two views of Sidneyia: top, as seen from below, showing the form of the limbs and the attachment of eyes and antennae; and bottom, as seen from above. Drawn by Marianne Collins.

  The year 1981 was pivotal for Burgess arthropods, and for the final dispersal of the last remaining “merostomoid” hope. For, in the same year of Odaraia and Sidneyia, Whittington published his “mop-up” monograph, “Rare Arthropods from the Burgess Shale, Middle Cambrian, British Columbia.” Most or all of these animals had fallen (or would have fitted, had they been known at the time) into the “merostomoids.” But Whittington could reconstruct not one as a chelicerate. All became orphans, unique arthropods unto themselves.

  Molaria has a deep head shield, shaped like a quarter sphere, followed by eight trunk segments diminishing in size toward the rear, and capped by a cylindrical telson with a very long, jointed posterior spine, extending back more than the length of the body (figure 3.47). This basic form is faultlessly “merostomoid,” but the head bears a pair of short antennae, followed by three pairs of biramous appendages.

  3.47. Molaria, a unique arthropod of “merostomoid” form (Whittington, 1981).

  3.48. The tuberculate arthropod Habelia. Drawn by Marianne Collins.

  Habelia has the same basic shape as Molaria, but Whittington also described an impressive set of differences, some of high taxonomic significance. The carapace is covered with tubercles—a superficial though visually striking difference (figure 3.48). The trunk has twelve segments, with no cylindrical telson. The extended tail spike, ornamented with barbs and ridges, is unsegmented, but has a single joint about two-thirds of the way back. The head has a pair of antennae and only two pairs of subsequent ventral appendages. The first six trunk segments bear biramous appendages, but the last six probably bore gill branches only (in Molaria all eight body segments bear biramous appendages.)

  Whittington also discovered a new arthropod genus—a complex, tiny creature less than a half inch in length (figure 3.49). This unique and peculiar animal, named Sarotrocercus, has a head shield followed by nine body segments and a tail spike with a tuft of spines at its tip. A large pair of eyes, borne on stalks, protrudes from the bottom front end of the head shield (Molaria and Habelia are blind). In addition, the head carries one pair of thick, strong appendages terminating in a two-pronged segment. Whittington also found ten very different pairs of appendages (one pair on the head and one on each of the nine body segments)—long comblike structures, presumably gill branches, but without any evident trace of a leg branch. Whittington reconstructed Sarotrocercus as a pelagic animal, swimming on its back with Amiskwia and Odontogriphus among the rare Burgess organisms that probably lived in the water column above the stagnant basin that received the mudslide.

  3.49. The tiny arthropod Sarotrocercus, swimming on its back. Note the large eyes, the strong pair of feeding appendages, and the gill branches, presumably used for swimming, on the body segments behind. Drawn by Marianne Collins.

  Actaeus, based on a single specimen two inches long, has a head shield with a marginal eye lobe, followed by eleven body segments and an elongate, triangular terminal plate (figure 3.50). The head bears a pair of remarkable appendages, each with a stout initial portion, bent and extending downward, ending in a group of four spines. Two very long whiplike extensions attach to the inner border of the last segment, and run down and back. Behind this structure, the head probably carried three pairs of ordinary biramous appendages.

  Alalcomenaeus has a basically similar look and arrangement of appendages (see figure 3.50), and may be related to Actaeus. A head shield, bearing a marginal eye lobe, is followed by twelve body segments and an ovate terminal plate. The head bears a pair of large appendages, each with a broad initial section followed by a long thin extension—not nearly so complex as in Actaeus, but similar in style and position. The head also carries three pairs of biramous appendages. One specimen reveals an impressive set of spines on the inner surfaces of the walking legs—in proper position for passing food forward to the mouth. “These remarkable appendages,” Whittington wrote, “suggest a benthic scavenger, able to hold on to, and tear up, a carcass” (1981a, p. 331).

  Aside from a very tentative relationship between Actaeus and Alalcomenaeus, each of the five genera presented a highly specialized design based on unique features and arrangements of parts. Whittington concluded, echoing the now-familiar Burgess story:

  Many new and unexpected features have been revealed, and the morphological gaps between species greatly enlarged. Each, with rare exceptions, shows a most distinctive combination of characters. The selection [of genera] dealt with here adds further to the range of morphological characters in the nontrilobite arthropods, and to the variety of distinctive combinations of characters (1981a, p. 331).

  In 1983, Bruton and Whittington combined to deliver the coup de grâce by describing the last two major Burgess arthropods—the large Emeraldella and Leanchoilia, last two members of Størmer’s discredited Merostomoidea.

  3.50. Two arthropods that may be closely related (Whittington, 1981). (A) Actaeus. (B) Alalcomenaeus.

  3.51. Emeraldella, seen from above (A), and from the side (B), resting on the bottom. The very small gill branches of the biramous appendages indicate that this animal walked on the sea floor.

  Emeraldella possesses the basic “merostomoid” form, but accompanied by yet another set of unique structures and arrangements. The typical head shield bears a pair of very long antennae, curving up and back, followed by five pairs of appendages, the first short and uniramous, the last four biramous (figure 3.51). The first eleven trunk segments are broad, though progressively narrowing toward the rear, and each bears a pair of biramous appendages. The last two segments are cylindrical, and a long unjointed tail spine extends at the rear.

  Leanchoilia also shares the superficiality of general “merostomoid” shape, with a triangular head shield (terminating in a curious, upturned “snout”), followed by eleven body segments, narrowing and curving backward beyond the fifth. A short triangular tail spine with lateral spikes caps the nether end (figure 3.52). Leanchoilia bears thirteen pairs of biramous appendages, two at the rear of the head shield, one on each of the eleven body segments.

  But Leanchoili
a also possesses the most curious and interesting appendage of any Burgess arthropod—an exaggerated version of the frontal structure of Actaeus, a possible relative. Borrowing a term from Yohoia, and in the absence of any appropriate technical name, Bruton and Whittington simply called this structure the “great appendage.” Its basal part contains four stout segments facing down at first, but bending through ninety degrees to run forward. The second and third segments end in very long, whiplike extensions, annulated over the last half of their length. The fourth segment has a tapering shaft ending dorsally in a group of three claws, and extending ventrally as a third whiplike structure with annulations. The different orientations of various specimens indicate that this great appendage was hinged at its base (figure 3.53) and could extend forward, to help Leanchoilia repose on the substrate (figure 3.54), or bend back, perhaps to reduce resistance in swimming. Further evidence for swimming as a primary mode of life comes from the biramous appendages. Unlike Emeraldella, with its long walking legs and small gill branches, Leanchoilia bears such large gill branches that they form a veritable curtain of overlapping, lamellate lobes, completely covering and extending beyond the shorter leg branches underneath.

  The completed redescription of all “merostomoid” genera prompted Bruton and Whittington to reflect upon the incredible disparity uncovered beneath a superficial similarity of outward form. Consider only the arrangement of appendages on the head—an indication of original patterns in segmentation, and a guide to the deep anatomical structure of arthropods. Sidneyia has a pair of antennae and no other appendages. Emeraldella also bears pre-oral antennae, but has five additional pairs of appendages behind the mouth, one uniramous and four biramous. Leanchoilia does not possess antennae, but bears its remarkable “great appendages,” followed by two biramous pairs behind the mouth.

  3.52. Top view of Leanchoilia. Note the three whiplike extensions of the great appendage in front and the triangular tail spine behind.

  The Burgess had been an amazing time of experimentation, an era of such evolutionary flexibility, such potential for juggling and recruitment of characters from the arthropod grabbag, that almost any potential arrangement might be essayed (and assayed). We now recognize clear groups, separated by great morphological gulfs, only because the majority of these experiments are no longer with us. “It was only later that certain of these solutions were fixed in combinations that allow the present arthropod groups to be recognized” (Bruton and Whittington, 1983, p. 577).

  3.53. Camera lucida drawings of two specimens of Leanchoilia. The great appendages are labeled Lga and Rga, and their major segments are numbered.(A) The great appendages are folded back, presumably in the swimming position; the right appendage is flat against the body, with the left just below. A trace of the gut, or alimentary canal (al) and the tail spine (tsp) are visible. (B) The appendages extend forward, in the feeding position.

  3.54. Two views of Leanchoilia: top, in swimming position, with the great appendages folded back and the whiplike tentacles extending beyond the length of the body; and bottom, with the great appendages extending forward to aid the animal in resting on the bottom. Drawn by Marianne Collins.

  A Present from Santa Claws

  Bureaucratic entanglement provides one possible benefit amidst its own distinctive and inimitable brand of frustration. You sometimes get so angry that you do something useful as an end run around intransigence. As the old motto goes, Don’t get mad, get even. When Des Collins, after sublime patience and deep entanglement, was denied permission to excavate at Walcott’s quarry and allowed only to gather specimens from the talus slope (under further restrictions and nearly endless delays), he realized that he would have to shift his Burgess interests elsewhere.*

  Collins therefore began to search for Burgess equivalents in surrounding areas, where collection and excavation might be permitted. He succeeded abundantly, finding soft-bodied fossils at more than a dozen additional nearby localities. Most of these assemblages contain the same species as Walcott’s quarry, but Collins made a few outstanding discoveries of his own. At a locality five miles south of Walcott’s quarry (Collins, 1985), and one hundred feet below in stratigraphic sequence, Collins made the find of the decade—a large arthropod with so many spiny appendages on its head that Collins, following an old tradition of field work, gave it a nickname. As Walcott had called Marrella the “lace crab,” Collins dubbed his discovery “Santa Claws.” Working with Derek Briggs, Collins has now formalized and honored this name in his technical description (Briggs and Collins, 1988). “Santa Claws” is now, officially, Sanctacaris, which means almost the same thing.

  Sanctacaris has a bulbous head shield, wider than long and extending laterally as a flat, triangular projection on each side (figure 3.55). The body bears eleven broad segments, the first ten with a pair of biramous appendages. A wide, flat telson caps the rear end. The combination of large lamellate gill branches on the body appendages and a broad telson well designed for stabilization and steering indicates that Sanctacaris probably favored swimming over walking.

  The striking suite of head appendages identifies this relatively large Burgess arthropod (up to four inches long) as a carnivore specialized for direct pursuit. The first five pairs make a coordinated and formidable array that inspired Collins’s field name. They are biramous, with the outer branches reduced to antenna-like projections (not gills) and the inner branches arranged as a fierce-looking set of jointed feeding appendages with sharp spines on the inner borders. These feeding branches gain in length from front to back, starting with four segments on the first pair, and increasing to eight or more on the fifth. The sixth pair, different in both form and position, lies behind the first five and well to the side. The outer branch is, again, similar to an antenna in form, but much larger than the corresponding branch of the five feeding appendages. The inner branch is short, but terminates in an impressive fringe of radiating spines.

  One might think at first assessment, Oh, just another of those Burgess “merostomoids”—with a forest of head appendages as its distinctive specialization, just as Habelia has its tubercles, Sidneyia its stout walking legs, and Leanchoilia its great appendage. Interesting, but not my advertised “find of the decade.”

  Not so. The difference between Sanctacaris and the others is taxonomic, and conceptually stunning: Sanctacaris seems to be a genuine chelicerate, the first known member of a line that eventually yielded horseshoe crabs, spiders, scorpions, and mites. Sanctacaris bears the requisite six pairs of appendages on its head. None of these appendages has been specialized to form the distinctive claw, the chelicera, that defines and names the group, but the absence of a structure early in the geological run of a group may simply mean that such a specialization has not yet evolved.

  Briggs and Collins (1988) have also identified other derived chelicerate characters (including the differentiation of head from body appendages, and the position of the anus), thus corroborating the status of Sanctacaris by more than a single feature. They state:

  Such a combination is unique to the chelicerates. The apparent lack of chelicerae, an advanced character present in all other chelicerates, is consistent with the primitive biramous appendages on both the head and trunk. It places Sanctacaris in a primitive sister group to all other chelicerates.

  3.55. Sanctacaris. Drawn by Marianne Collins.

  The limbs of modern chelicerates are uniramous, with the outer branch lost on the head appendages (yes, the walking legs of spiders are all on the prosoma, or head portion), and the inner branch lost on the trunk (yes again, spider gills are on the opisthosoma, or body portion). Sanctacaris, by preserving the full set of possibilities before selective elimination in later specialized lines, serves as an interesting structural precursor for its great group.

  But the chief excitement of Sanctacaris lies in its key role in completing the fundamental argument for Burgess arthropods. With the discovery of Sanctacaris, we now have, in the Burgess, members of all four great arthropod gro
ups—trilobites in fair abundance, crustaceans represented by Canadaspis, uniramians by Aysheaia* (accepting Robison’s interpretation, as I do), and chelicerates by Sanctacaris. They are all there—but so are at least thirteen other lineages (and perhaps as many again yet to be described) of equal morphological uniqueness. Some of these thirteen are among the most specialized (Leanchoilia) or, at least by numbers, the most successful (Marrella) of Burgess arthropods. I challenge any paleontologist to argue that he could have gone back to the Burgess seas and, without the benefit of hindsight, picked out Naraoia, Canadaspis, Aysheaia, and Sanctacaris for success, while identifying Marrella, Odaraia, Sidneyia, and Leanchoilia as ripe for the grim reaper. Wind back the tape of life, and let it play again. Would the replay ever yield anything like the history that we know?

  CONTINUING THE MARCH OF WEIRD WONDERS

  The last decade, so satisfying for arthropods, has also witnessed the resolution of two additional weird wonders—unique and independent anatomies that would merit classification as separate phyla if we felt comfortable about bestowing so high a taxonomic rank on a single species (see Briggs and Conway Morris, 1986, for a list of such Burgess creatures still unstudied). These two works may be the most elegant and persuasive in the entire Burgess canon. They stand as a fitting end to my play, for they combine the greatest intellectual and aesthetic satisfaction with an assurance that this particular drama has no foreseeable end.