Wonderful Life: The Burgess Shale and the Nature of History

by Stephen Jay Gould

  Such delays and long lead times strongly suggest contingency and a vast realm of unrealized possibilities. If prokaryotes had to advance toward eukaryotic complexity, they certainly took their time about it. Moreover, when we consider the favored hypothesis for the origin of the eukaryotic cell, we enter the realm of quirky and incidental side consequences as unpredictable sources of change. Our best theory identifies at least some major organelle—the mitochondria and chloroplasts almost surely, and others with less confidence—as descendants of entire prokaryotic cells that evolved to live symbiotically within other cells (Margulis, 1981). In this view, each eukaryotic cell is, by descent, a colony that later achieved tighter integration. Surely, the mitochondrion that first entered another cell was not thinking about the future benefits of cooperation and integration; it was merely trying to make its own living in a tough Darwinian world. Accordingly, this fundamental step in the evolution of multicellular life arose for an immediate reason quite unrelated to its eventual effect upon organic complexity. This scenario seems to portray fortunate contingency rather than predictable cause and effect. And if you wish nevertheless to view the origin of organelles and the transition from symbiosis to integration as predictable in some orderly fashion, then tell me why more than half the history of life passed before the process got started.

  One final point that I find chilling with respect to the possibility of something like human evolution in an alternative world: Even though this first event took more than half the known history of life, I might be prepared to accept the probability of an eventual origin for higher intelligence if the earth were slated to endure for hundreds of billion of year—so that this initial step took but a tiny fraction of potential time. But cosmologists tell us that the sun is just about at the halfway point of existence in its current state; and that some five billion years from now, it will explode, expanding in diameter beyond the orbit of Jupiter and engulfing the earth. Life will end unless it can move elsewhere; and life on earth will terminate in any case.

  Since human intelligence arose just a geological second ago, we face the stunning fact that the evolution of self-consciousness required about half of the earth’s potential time. Given the errors and uncertainties, the variations of rates and pathways in other runs of the tape, what possible confidence can we have in the eventual origin of our distinctive mental abilities? Run the tape again, and even if the same general pathways emerge, it might take twenty billion years to reach self-consciousness this time—except that the earth would be incinerated billions of years before. Run the tape again, and the first step from prokaryotic to eukaryotic cell might take twelve billion instead of two billion year—and stromatolites, never awarded the time needed to move on, might be the highest mute witnesses to Armageddon.

  THE FIRST FAUNA OF MULTICELLULAR ANIMALS

  You might accept this last sobering scenario, but then claim, fine, I’ll grant the unpredictability of getting beyond prokaryotic cells, but once you finally do get multicellular animals, then the basic pathways are surely set and further advance to consciousness must occur. But let’s take a closer look.

  The first multicellular animals, as discussed in chapter II, are members of a world-wide fauna named for the most famous outcrop at Ediacara, in Australia. Martin Glaessner, the paleontologist most responsible for describing the Ediacara animals, has always interpreted them, under traditional concepts of the cone, as primitive representatives of modern group—mostly members of the coelenterate phylum (soft corals and medusoids), but including annelid worms and arthropods (Glaessner, 1984). Glaessner’s traditional reading evoked very little opposition (but see Pflug, 1972 and 1974), and the Ediacara fauna settled comfortably into textbooks as fitting ancestors for modern group—for their combination of maximal age with minimal complexity neatly matches expectations.

  The Ediacara fauna has special importance as the only evidence for multicellular life before the great divide separating the Precambrian and Cambrian, a boundary marked by the celebrated Cambrian explosion of modern groups with hard parts. True, the Ediacara creatures are only barely Precambrian; they occur in strata just predating Cambrian and probably do not extend more than 100 million years into the uppermost Precambrian. In keeping with their position immediately below the boundary, the Ediacara animals are entirely soft-bodied. If taxonomic identity could be maintained right through this greatest of geological transitions, and without major disruption in design to accompany the evolution of hard parts, then the smooth continuity of the cone would be confirmed. This version of Ediacara begins to sound suspiciously like Walcott’s shoehorn.

  In the early 1980s, my friend Dolf Seilacher, professor of paleontology at Tubingen, Germany, and in my opinion the finest paleontological observer now active, proposed a radically different interpretation of the Ediacara fauna (Seilacher, 1984). His twofold defense rests upon a negative and a positive argument. For his negative claim, he argues on functional grounds that the Ediacara creatures could not have operated as their supposed modern counterparts, and therefore may not be allied with any living group, despite some superficial similarity of outward form. For example, most Ediacara animals have been allied with the soft corals, a group including the modern sea fans. Coral skeletons represent colonies housing thousands of tiny individuals. In soft corals, the individual polyps line the branches of a tree or network structure, and the branches must be separated, so that water can bring food particles to the polyps and sweep away waste products. But the apparent branches of the Ediacara forms are joined together, forming a flattened quiltlike mat with no spaces between the sections.

  For his positive claim, Seilacher argues that most Ediacara animals may be taxonomically united as variations on a single anatomical plan—a flattened form divided into sections that are matted or quilted together, perhaps constituting a hydraulic skeleton much like an air mattress (figure 5.5). Since this design matches no modern anatomical plan, Seilacher concludes that the Ediacara creatures represent an entirely separate experiment in multicellular life—one that ultimately failed in a previously unrecognized latest Precambrian extinction, for no Ediacara elements survived into the Cambrian.

  For the Burgess fauna, the case against Walcott’s shoehorn has been proven, I think, with as much confidence as science can muster. For the Ediacara fauna, Seilacher’s hypothesis is a plausible and exciting, but as yet unproven, alternative to the traditional reading, which will one day be called either Glaessner’s shoehorn or Glaessner’s insight, as the case may be.

  But consider the implications for unpredictability if Seilacher’s view prevails, even partly. Under Glaessner’s ranking in modern groups, the first animals share the anatomical designs of later organisms, but in simpler form—and evolution must be channeled up and outward in the traditional cone of increasing diversity. Replay the tape, starting with simple coelenterates, worms, and arthropods, a hundred times, and I suppose that you will usually end up with more and better of the same.

  But if Seilacher is right, other possibilities and other directions were once available. Seilacher does not believe that all late Precambrian animals fall within the taxonomic boundaries of this alternative and independent experiment in multicellular life. By studying the varied and abundant trace fossils (tracks, trails, and burrows) of the same strata, he is convinced that metazoan animals of modern design—probably genuine worms in one form or another—shared the earth with the Ediacara fauna. Thus, as with the Burgess, several different anatomical possibilities were present right at the beginning. Life might have taken either the Ediacara or the modern pathway, but Ediacara lost entirely, and we don’t know why.

  5.5. Seilacher’s classification of the Ediacara organisms according to their variations on a single flattened, quiltlike anatomical plan. These organisms are conventionally placed in several different modern phyla.

  Suppose that we could replay life’s tape from late Precambrian times, and that the flat quilts of Ediacara won on their second attempt, while metazoan
s were eliminated. Could life have ever moved to consciousness along this alternate pathway of Ediacara anatomy? Probably not. Ediacara design looks like an alternative solution to the problem of gaining enough surface area as size increases. Since surfaces (length2) increase so much more slowly than volumes (length3), and since animals perform most functions through surfaces, some way must be found to elaborate surface area in large creatures. Modern life followed the path of evolving internal organs (lungs, villi of the small intestine) to provide the requisite surfaces. In a second solution—proposed by Seilacher as the key to understanding Ediacara design—organisms may not be able to evolve internal complexity and must rely instead on changes in overall form, taking the shape of threads, ribbons, sheets, or pancakes so that no internal space lies very far from the outer surface. (The complex quilting of Ediacara animals could then be viewed as a device for strengthening such a precarious form. A sheet one foot long and a fraction of an inch thick needs some extra support in a world of woe, tides, and storms.)

  If Ediacara represents this second solution, and if Ediacara had won the replay, then I doubt that animal life would ever have gained much complexity, or attained anything close to self-consciousness. The developmental program of Ediacara creatures might have foreclosed the evolution of internal organs, and animal life would then have remained permanently in the rut of sheets and pancake—a most unpropitious shape for self-conscious complexity as we know it. If, on the other hand, Ediacara survivors had been able to evolve internal complexity later on, then the pathways from this radically different starting point would have produced a world worthy of science fiction at its best.

  THE FIRST FAUNA OF THE CAMBRIAN EXPLOSION

  Our hypothetical advocate of the cone and ladder might be willing to give ground on these first two incidents from the dim mists of time, but he might then be tempted to dig his entrenchment across the Cambrian boundary. Surely, once the great explosion occurs, and traditional fossils with hard parts enter the record, then the outlines must be set, and life must move upward and outward in predictable channels.

  Not so. As noted in chapter II, the initial shelly fauna, called Tommotian to honor a famous Russian locality, contains far more mysteries than precursors. Some modern groups make an undoubted first appearance in the Tommotian, but more of these fossils may represent anatomies beyond the current range. The story is becoming familiar—a maximum of potential pathways at the beginning, followed by decimation to set the modern pattern.

  The most characteristic and abundant of all Tommotian creatures, the archaeocyathids (figure 5.6), represent a long-standing problem in classification. The familiar litany plays again. These first reef-forming creatures of the fossil record are simple in form, usually cone-shaped, with double wall—cup within cup. In the traditional spirit of the shoehorn, they have been shunted from one modern group to another during more than a century of paleontological speculation. Corals and sponges have been their usual putative homes. But the more we learn about archaeocyathids, the stranger they appear, and most paleontologists now place them in a separate phylum destined to disappear before the Cambrian had run its course.

  5.6. An archaeocyathid, showing the basic organization of cup within cup.

  Even more impressive is the extensive disparity just now being recognized among organisms of the “small shelly fauna.” Tommotian rocks house an enormous variety of tiny fossils (usually one to five millimeters in length) that cannot be allied with any modern group (Bengtson, 1977; Bengtson and Fletcher, 1983). We can arrange these fossils by outward appearance, as tubes, spines, cones, and plates (figure 5.7 shows a representative sample), but we do not know their zoological affinities. Perhaps they are merely bits and pieces from an era of early, still imperfect skeletonization; perhaps they covered familiar organisms that later developed the more elaborate shells of their conventional fossil signatures. But perhaps—and this interpretation has recently been gaining favor among aficionados of the small shelly fauna—most of the Tommotian oddballs represent unique anatomies that arose early and disappeared quickly. For example, Rozanov, the leading Russian expert on this fauna, concludes his recent review by writing:

  Early Cambrian rocks contain numerous remains of very peculiar organisms, both animals and plants, most of which are unknown after the Cambrian. I tend to think that numerous high-level taxa developed in the early Cambrian and rapidly became extinct (1986, p. 95).

  5.7. Representative organisms of unknown affinity from the Cambrian “small shelly fauna” (Rozanov, 1986). (A) Tommotia. (B) Hyolithellus. (C) Lenargyrion.

  Once again, we have a Christmas tree rather than a cone. Once again, the unpredictability of evolutionary pathways asserts itself against our hope for the inevitability of consciousness. The Tommotian contained many modern groups, but also a large range of alternative possibilities. Rewind the tape into the early Cambrian, and perhaps this time our modern reefs are built by archaeocyathids, not corals. Perhaps no Bikini, no Waikiki; perhaps, also, no people to sip rum swizzles and snorkle amidst great undersea gardens.

  THE SUBSEQUENT CAMBRIAN ORIGIN OF THE MODERN FAUNA

  Our traditionalist is now beginning to worry, but he will grant this one last point pour mieux sauter. OK, the very first Cambrian fauna included a plethora of alternative possibilities, all equally sensible and none leading to us. But, surely, once the modern fauna arose in the next phase of the Cambrian, called Atdabanian after another Russian locality, then the boundaries and channels were finally set. The arrival of trilobites, those familiar symbols of the Cambrian, must mark the end of craziness and the inception of predictability. Let the good times roll.

  This book is quite long enough already, and you do not want a “second verse, same as the first.” I merely point out that the Burgess Shale represents the early and maximal extent of the Atdabanian radiation. The story of the Burgess Shale is the tale of life itself, not a unique and peculiar episode of possibilities gone wild.

  THE ORIGIN OF TERRESTRIAL VERTEBRATES

  Our traditionalist is now reeling. He is ready to abandon virtually all of life to contingency, but he will make his last stand with vertebrates. The game, after all, centers on human consciousness as the unpredictable product of an incidental twig, or the culmination of an ineluctable, or at the very least a probable, trend. To hell with the rest of life; they aren’t on the lineage leading to consciousness in any case. Surely, once vertebrates arose, however improbable their origin, we could then mount confidently from ponds to dry land to hind legs to big brains.

  I might grant the probability of the most crucial environmental transition—from water to land—if the characteristic anatomy of fishes implied, even for incidental reasons, an easy transformation of fins into sturdy limbs needed for support in the gravity of terrestrial environments. But the fins of most fishes are entirely unsuited for such a transition. A stout basal bar follows the line of the body axis, and numerous thin fin rays run parallel to each other and perpendicular to the bar. These thin, unconnected rays could not support the weight of the body on land. The few modern fishes that scurry across mud flats, including Periophthalmus, the “walking fish,” pull their bodies along and do not stride with their fins.

  Terrestrial vertebrates could arise because a relatively small group of fishes, only distantly related to the “standard issue,” happened, for their own immediate reasons, to evolve a radically different type of limb skeleton, with a strong central axis perpendicular to the body, and numerous lateral branches radiating from this common focus. A structure of this design could evolve into a weight-bearing terrestrial limb, with the central axis converted to the major bones of our arms and legs, and the lateral branches forming digits. Such a fin structure did not evolve for its future flexibility in permitting later mammalian life; (this limb may have provided advantages, in superior rotation, for bottom-dwelling fishes that used the substrate as an aid in propulsion). But whatever its unknown advantages, this necessary prerequisite to terrestrial life evo
lved in a restricted group of fishes off the main line—the lungfish-coelacanth-rhipidistian complex. Wind the tape of life back to the Devonian, the so-called age of fishes. Would an observer have singled out these uncommon and uncharacteristic fishes as precursors to such conspicuous success in such a different environment? Replay the tape, expunge the rhipidistians by extinction, and our lands become the unchallenged domain of insects and flowers.

  PASSING THE TORCH TO MAMMALS

  Can we not grant the traditionalist some solace? Let contingency rule right to the origin of mammals. Can we not survey the world as mammals emerged into the realm of dinosaurs, and know that the meek and hairy would soon inherit the earth? What defense could large, lumbering, stupid, cold-blooded behemoths provide against smarts, sleekness, live birth, and constant body temperature? Don’t we all know that mammals arose late in the reign of dinosaurs; and did they not then hasten the inevitable transition by eating their rivals’eggs?

  This common scenario is fiction rooted in traditional hopes for progress and predictability. Mammals evolved at the end of the Triassic, at the same time as dinosaurs, or just a tad later. Mammals spent their first hundred million year—two-thirds of their total history—as small creatures living in the nooks and crannies of a dinosaur’s world. Their sixty million years of success following the demise of dinosaurs has been something of an afterthought.

  We have no indication of any trend toward mammalian hegemony during this initial hundred million years. Quite the reverse—dinosaurs remained in unchallenged possession of all environments for large-bodied terrestrial creatures. Mammals made no substantial moves toward domination, larger brains, or even greater size.