Wonderful Life: The Burgess Shale and the Nature of History

by Stephen Jay Gould

  If the spines are hard to interpret, what about the tentacles above—where prospects for modern analogues are dimmer. The pincers at their tips could have captured food, but the tentacles don’t reach the head region, and passage of food from one tentacle to another toward a frontal mouth offers little promise of efficient eating. Noting a possible connection between a hollow tube within each tentacle and a gut within the trunk (neither well enough preserved to inspire confidence), Conway Morris offered a fascinating alternative. Perhaps Hallucigenia had no frontal mouth at all. Perhaps each tentacle gathered food independently, passing the collected particles down its own personal gullet into the communal gut. You have to consider bizarre solutions when you work with such a strange animal.

  Yet Hallucigenia is so peculiar, so hard to imagine as an efficiently working beast, that we must entertain the possibility of a very different solution. Perhaps Hallucigenia is not a complete animal, but a complex appendage of a larger creature, still undiscovered. The “head” end of Hallucigenia is no more than an incoherent blob in all known fossils. Perhaps it is no head at all, but a point of fracture, where an appendage (called Hallucigenia) broke off from a larger main body (yet undiscovered). This prospect may seem disappointing, since Hallucigenia by itself forms such a wondrous beast. Hence, I am rooting for Conway Morris’s interpretation (but if forced to bet, I would have to place my money on the appendage theory). But then, the prospect of Hallucigenia as only an appendage may be even more exciting—for the whole animal, if ever discovered and reconstructed, might be even more peculiar than Hallucigenia as now interpreted. It has happened before in the Burgess. Anomalocaris (see Act 5) was once viewed as an entire arthropod, and a fairly dull crustacean at that. Then Whittington and Briggs (1985) resolved it as a feeding appendage of an animal ranking just behind Hallucigenia in Burgess oddity. We have surely not seen the last, and perhaps not the greatest, of Burgess surprises.

  DEREK BRIGGS AND BIVALVED ARTHROPODS: THE NOT-SO-FLASHY BUT JUST-AS-NECESSARY FINAL PIECE

  I must begin with an apology to Derek Briggs for an invisible slight arising from both ignorance and thoughtlessness. I made a bad mistake when I first laid out this chronological centerpiece of the book—that is, before I read the monographs in detail. I saw the Burgess transformation as a dramatic interplay between Harry Whittington, the conservative systematist who started it all, and Simon Conway Morris, the young and radical man of ideas who developed a revolutionary interpretation and dragged everyone else along. I have already indicated my error in reading this interaction according to the conventional script.

  Let me now confess another mistake, one that I should not have made. This is the classic error of those who write about science without an intuitive feel for its daily procedures; those who do the work should know better. The journalistic tradition so exalts novelty and flashy discovery, as reportable and newsworthy, that standard accounts for the public not only miss the usual activity of science but also, and more unfortunately, convey a false impression about what drives research*

  A project like the Burgess revision has potentially flashy and predictably less noticeable aspects. Both are necessary. A conventional reporter will convey only the hot ideas and the startling facts—Hallucigenia gets ink; the Burgess trilobites get ignored. But the Burgess oddballs mean little in isolation. When placed in an entire fauna, filled with conventional elements as well, they suggest a new view of life. The conventional creatures must be documented with just as much love, and just as assiduously—for they are every bit as important to the total picture.

  Derek Briggs drew the bivalved arthropods as his subject—the apparently most conventional group in the Burgess fauna. He produced an elegant series of monographs on these animals, finding some surprises, but also confirming some expectations. I had not appreciated the central role that Briggs’s work on bivalved arthropods played in the Burgess transformation. As I read Derek’s monographs, I recognized my error with some shame, and grew to understand Harry, Derek, and Simon as a trio of equals, each with a distinct and necessary role in the total drama.

  Walcott and others had described about a dozen genera of arthropods with a bivalved carapace (usually enclosing the entire head and front part of the body). Several of these genera cannot be classified with certainty, for only the carapaces have been found, not the soft parts. The other genera have always, and without any doubt or hesitation, been identified as crustaceans—as are all modern arthropods with a bivalved carapace. Derek Briggs began his project without any conscious doubts: “There were some redescriptions to be done. I assumed I would be dealing with a bunch of crustaceans.”

  Briggs described two outstanding discoveries in his first monographs on the bivalved arthropods of the Burgess Shale. Put these together with Simon’s oddballs and Harry’s orphaned arthropods, and you have, by 1978, both a fully articulated and completely new account of how multicellular animal life evolved.

  1. Branchiocaris, the first discovery. The Crustacea are an enormous and diverse group—from the nearly microscopic ostracodes with bivalved carapaces covering the entire body like a clamshell, to giant crabs with leg spreads of several feet. Yet all are built upon a stereotyped ground plan, with a definite signature in the structure of the head The crustacean head is an amalgam of five original segments plus eyes. Five pairs of appendages are therefore present—and in a definite arrangement: two pre-oral (usually antennae) and three post-oral (usually mouth parts).* Since all modern bivalved arthropods are crustaceans, Briggs assumed that he would find this frontal signature in his Burgess subjects. But the Burgess soon provided yet another surprise.

  Back in 1929, Charles E. Resser, Walcott’s right-hand man at the Smithsonian, had described a single Burgess specimen as the crustacean Protocaris pretiosa. The genus Protocaris had been established in 1884, by none other than Charles Doolittle Walcott in his pre-Burgess days, for a Cambrian arthropod from the Parker Slate of Vermont. Resser considered the Burgess animal as sufficiently close for inclusion in the same genus. Briggs disagreed and established the new genus Branchiocaris.

  Briggs managed to amass a total of five specimens—Resser’s original, three more from the Walcott collection, and a fifth whose part was found by Raymond in 1930, but whose counterpart remained on the Burgess talus until collected by the Royal Ontario Museum expedition in 1975, as recounted in the heart-warming tale earlier in this chapter. The bivalved carapace of Branchiocaris covers the head and anterior two-thirds of the body (figure 3.36). The body itself contains some forty-six short segments, with a two-pronged telson behind. The appendages are not clearly distinguishable in the limited number of available fossils, but may have been biramous, with a short segmented branch (presumably homologous to the walking leg of most biramous arthropods, but too reduced for such a function in Branchiocaris), and a larger bladelike process, probably used for swimming near the sea floor.

  But the head of Branchiocaris provided the big surprise. Two pairs of short antenna-like appendages, pointing forward, could clearly be seen—the first more conventional in form, uniramous with many segments; the second more peculiar, stout and composed of few segments, perhaps with a claw or pincer at the end. Briggs called this second pair the “principal appendage”—just as Whittington, stumped by an analogous structure in Yohoia, had spoken of a “great appendage.”

  These appendages attached to the upper and lateral surfaces of the head. On the ventral side, three pairs of additional appendages should have followed the mouth. Briggs found nothing. The mouth stood all alone on an unadorned ventral surface. Branchiocaris, with two and only two pairs of appendages on the head, was not a crustacean. “It apparently defies classification within any group of Recent arthropods,” Briggs concluded (1976, p. 13).

  3.36. Reconstruction of Branchiocaris by Briggs (1976). (A) Side view. (B) Bottom view, showing the ventral surface of the animal surrounded by the two valves of its carapace. Note in particular the pairs of uniramous appendages, especially the uniqu
e principal appendage (labeled lpa and rpa). And note also the absence of any appendages on the head behind the mouth; this arrangement is unknown in any modern arthropod group.

  Thus, the bivalved arthropods—the group that seemed most promising as a coherent set of evolutionary cousins—also formed an artificial category hiding an unanticipated anatomical disparity. What order could possibly be found among the Burgess arthropods? Each one seemed to be built from a grabbag of characters—as though the Burgess architect owned a sack of all possible arthropod structures, and reached in at random to pick one variation upon each necessary part whenever he wanted to build a new creature. Could a biramous limb of trilobite type adorn any kind of arthropod body? Could a bivalved carapace cover any anatomy? Where was order, where decorum?

  2. Canadaspis, the second discovery. Consider the story of Burgess arthropods as published by the end of 1976. Marrella, a supposed relative of trilobites, was an orphan. Yohoia, with its great appendage, was uniquely specialized and unaffiliated, not a precursor of anything. Burgessia, namesake of the fauna, was another orphan. Even Branchiocaris, firm candidate for a crustacean, sported a unique anatomy under its bivalved carapace. Moreover, these four orphans showed no propensity for coherence among themselves; each reveled in its own peculiarities. Would any Burgess arthropod ever accept the allegiance to a modern group that Walcott, wielding his shoehorn, had once forced upon all?

  Canadaspis is the second most common animal in the Burgess Shale. It is large by Burgess standards (up to three inches in length) and tends to be preserved with a conspicuous reddish color. It has a bivalved carapace, but as Briggs soon discovered, an underlying anatomy very different from Branchiocaris.

  In a short paper of 1977, Briggs placed two bivalved species in the new genus, Perspicaris. His reconstructions suggested something exciting, but the rarity of specimens and their poor preservation precluded any firm conclusion. He couldn’t prove the affiliation, but nothing about these two species precluded membership in the Crustacea. Had a representative of a modern group finally been found?

  In 1978, Briggs resolved this issue with elegance and finality. His long monograph on the well-preserved, superabundant Canadaspis perfecta finally placed a Burgess creature in a successful modern group. Canadaspis was not only a crustacean, but its home within the Crustacea could be established. Canadaspis is an early malacostracan—a representative of the great group of crabs, shrimp, and lobsters. Briggs found all elements of the intricate malacostracan stereotype in the anatomy of Canadaspis: a head bearing five pairs of appendages, and built of five segments plus eyes; a thorax (middle section) of eight segments, and an abdomen (back section) of seven segments plus a telson. Further, the head appendages are arranged just right, with two pairs of short, uniramous antennae in front of the mouth, and three pairs of ventral appendages behind the mouth.* The abdominal segments bear no appendages, but each thoracic segment carries a pair of standard biramous appendages, with an inner leg branch and a broad outer gill branch (figures 3.37 and 3.38).

  3.37. Reconstruction of Canadaspis by Briggs (1978). This animal has the typical structure of a true crustacean of the malacostracan line: two pairs of appendages in front of the mouth (labeled an1 and an2), three pairs of appendages behind the mouth (ma, mx1, and mx2), a thorax of eight segments (beginning with the segment labeled t1), and an abdomen of seven segments (ab1-ab7). Each thoracic segment bears a pair of biramous appendages.

  The brevity of this description is no denigration of the importance of Canadaspis in the Burgess reformulation. A weird animal needs a longer write-up to explain its uniqueness; a familiar creature can simply be characterized as “like Joe whom everyone knows.” But Canadaspis is both a key and an anchor to the Burgess story, a creature every bit as important as any of Simon’s weird wonders. Suppose that every Burgess animal were a bizarre denizen of a lost world. What then would we make of the assemblage? A failed experiment, a washout, a first attempt totally bypassed by a reconstituted modern fauna, and therefore offering no clues and no connection to the origin of later life. But the presence of Canadaspis, and other creatures of modern design, suggests a different and more enlightening view. The Burgess fauna does include modern prototypes, and, in this key respect is an ordinary Cambrian fauna; but the vastly broader range of designs that disappeared may reveal the most important of all patterns in life’s early history.

  3.38. The true crustacean Canadaspis. The five head segments bear two pairs of antennae and three pairs of appendages behind the mouth, the last two of which are continuous with, and similar in form to, the biramous appendages of the body. Drawn by Marianne Collins.

  As Derek resolved Canadaspis, Simon had left behind his whirlwind of wonders to work on the main subjects of his project, the true Burgess worms. His results, published in two monographs (1977 and 1979), beautifully affirmed the lesson of Canadaspis. Some Burgess organisms, even among soft-bodied members of the fauna, fit comfortably into modern groups—thus accentuating and highlighting the importance of the oddballs as additions to normality. In 1977, Conway Morris recognized among forms that Walcott had scattered across three phyla (as polychaetes, crustaceans, and echinoderms) six or seven genera of priapulid worms. The Priapulida form a small phylum of ten genera or so in today’s oceans, but they dominated the worm fauna of the Burgess Shale. (The Burgess priapulids form a major part of my story in chapter V.)

  In 1979, Conway Morris sorted out one of Walcott’s greatest confusions—the Burgess polychaetes. Walcott had used the Polychaeta (marine representatives of the phylum Annelida, or segmented worms) as a dumping ground for many Burgess oddities. Within Walcott’s polychaetes, Conway Morris found two genera of priapulids and four genera of weird wonders. But Walcott had also identified some true polychaetes. From this mixture, Conway Morris identified and established six genera of Burgess polychaetes. This group, so dominant in today’s seas, was overshadowed by priapulids (with the same number of genera, but many more specimens) in Burgess times. But both groups proclaimed the same general message. The Burgess fauna contained both ordinary and unique anatomies in abundance.

  ACT 4. Completion and Codification of an Argument: Naraoia and Aysheaia, 1977–1978

  After such an extended third act, we need a sparer fourth to make a largely symbolic point amidst the resolution of two important Burgess genera distinguished by more than their maximally unpronounceable, vowel-laden names.

  Harry Whittington had started this drama by orphaning some arthropods that everyone had previously placed in established groups (Act 1). He had upped the ante by showing that Opabinia was not an arthropod at all, but a creature of strange and unique anatomy (Act 2). His students and associates then converted these anomalies into a generality about the Burgess and its time by documenting the same pattern throughout the fauna (Act 3). When Harry Whittington finally accepted the new interpretation, and began to view anatomical oddity as a preferred hypothesis a priori, rather than a last resort, the story had reached its logical end; the Burgess transformation had been completed (Act 4). In conceptual terms, the rest would be mopping up, but with the best of all particular stories still to be told (Act 5).

  Naraoia added the last substantial piece to the logical structure of the new view. This old Burgess standby, described by Walcott as a branchiopod crustacean, has a carapace composed of two flat, smooth, oval valves, meeting at straightened borders one behind the other. These valves, discrete and shiny on most fossils, make Naraoia one of the most striking and attractive of Burgess organisms, but they also impose a severe problem in interpretation. They cover almost all the soft anatomy; most specimens show only the distal tips of the appendages, protruding out beyond the edge of the carapace (figure 3.39). Since the proximal (and invisible) ends of the appendages provide the primary taxonomic basis for identifying arthropod groups—both by their form and by their pattern of insertion into the body—Naraoia could never be properly interpreted.

  Whittington resolved this dilemma with his dis
covery of three-dimensional structure in the Burgess fossils. He realized that he could dissect through the firm carapace to reveal the proximal ends of the appendages, and their points of insertion. When he cut through the carapace of Naraoia (figure 3.40), he uncovered enough of the appendages to count their segments and resolve their proximal ends, including gnathobases and food grooves. Whittington also received one of the great surprises of his professional life. He was looking at a leg branch of the animal he knew best—a trilobite. But beyond a vague similarity in general outline, the carapace, with its two valves, hardly resembles the exoskeleton of a trilobite. Most trilobites have a threefold division, into head, thorax, and pygidium. (Contrary to popular belief, this division, stem to stern, is not the source of the name “trilobite,” or “three-lobed.” Trilobation refers to the threefold side-to-side division into a central axis and two side regions, called pleurae.)

  3.39. Camera lucida drawing of an excellent specimen of Naraoia (Whittington, 1977). The two valves of the carapace cover almost all the soft anatomy, and only the ends of the appendages protrude beyond them.

  3.40. Determination of the taxonomic affinity of Naraoia by dissection. (A) A complete specimen before dissection. (B) The same specimen, dissected to reveal the legs at their point of attachment to the body. (C) Camera lucida drawing of the dissected specimen. Since the legs are of typical trilobite form, Naraoia is identified as the first known bivalved trilobite.

  Whittington also found other key trilobite characters in Naraoia, notably the defining segmentation of the head, with one pair of uniramous pre-oral antennae and three pairs of ventral post-oral appendages. Naraoia, despite its curious outer covering, was surely a trilobite. Whittington therefore described this genus as a new and separate class within the Trilobita. He wrote with barely disguised joy and an uncharacteristic personal touch—and why not, for Harry is the world’s expert on trilobites. These are his babies, and he had just given birth to a stunning and different child: