Darwin's Doubt

by Stephen C. Meyer

  The discoveries near Chengjiang demonstrated beyond any reasonable doubt that sedimentary rocks can preserve soft-bodied fossils of great antiquity and in exquisite detail, thereby challenging the idea that the absence of Precambrian ancestors is a consequence of the fossil record’s inability to preserve soft-bodied animals from that period. At the same time, the sedimentary rocks near Chengjiang had other surprises in store.

  Precambrian Secrets

  Paul Chien is a Chinese-American marine biologist who, as a boy, left mainland China with his family to escape the 1949 Communist takeover under Mao Tse-tung. Eventually, after completing Ph.D. studies in the United States, he became a biology professor at the University of San Francisco. He first learned about the discoveries in southern China after reading the Time magazine cover story in 1995. Then he learned, ironically, from a story in People’s Daily, the official newspaper of the Chinese Communist party, that some Chinese paleontologists thought that these discoveries challenged a Darwinian view of the history of life.

  The great variety of the marine invertebrates present in Chengjiang fascinated Chien and convinced him to return to the country of his birth. After making his first trip in the summer of 1996, he met J. Y. Chen. As Paul Chien returned over several successive summers to do research of his own, he and J. Y. Chen continued to share their findings and compare notes.

  Upon arriving in China in 1998 for his third summer of research, Paul Chien learned that Chen had discovered a fossil of an adult sponge in the late Precambrian rocks of a sedimentary formation called the Doushantuo Phosphorite—a formation that lies beneath the Maotianshan Shale. As the two scientists examined the sediments that encased Chen’s fossil sponge, they made a discovery that would doom the most popular remaining version of the artifact hypothesis.

  FIGURE 3.6

  Cambrian explosion fossils from Chengjiang fauna, artist depictions and fossil photos. Photos in 3.6a, b, e, and f courtesy J. Y. Chen. Photos in 3.6c and d courtesy Paul Chien.

  As J. Y. Chen began to examine the sedimentary rocks that enclosed his fossilized sponge, he decided to look at them in a so-called thin section under a light microscope. Chen wondered whether smaller embryonic forms of these Precambrian animals might also have been preserved in these phosphorite rocks. Sure enough, under magnification he found little round balls that he and Paul Chien identified as sponge embryos. In 1999, at a major international conference about the Cambrian explosion held near Chengjiang, J. Y. Chen, Paul Chien, and three other colleagues presented their findings.47

  A number of Chinese paleontologists questioned them at first, suggesting that the little round balls were not sponge embryos at all, but instead the remains of brown and green algae.48 Here Paul Chien’s expertise came to the fore. Early in his career, Chien had perfected a technique for examining the embryos of living sponges under a scanning electron microscope. He now adapted his technique to examine these microscopic fossilized structures using a more powerful microscope. What he found startled him and amazed other scientists.

  Sponges are nature’s glassworks. They are made of a soft and flexible lattice of cells from which protrude silica-encrusted “spicules.” Though sponges come in a variety of shapes and sizes, they are one of the simplest known forms of animal life, with between six and ten distinctive cell types.49 In comparison, the typical arthropod has between thirty-five and ninety cell types.

  As Chien examined the little balls of the Doushantuo Phosphorite under the scanning microscope, he noticed what looked like cells undergoing cell division. At first, he had no way to determine what type of cells they might be. But as he examined cross sections of these cells more carefully, he identified a distinctive structure that he knew from his prior research.

  Only sponges have spicules, and the fossilized cells he was examining preserved microscopic spicules in the early stages of their development.50 Clearly, these were not algal balls; they were sponge embryos. Even more surprising, Chien was able to observe the internal structure of these embryonic cells, allowing him to identify the nuclei of some of these cells within the fossilized remains of the larger outer cell membrane (see Fig. 3.7).

  The discovery of these sponge embryos has proven decisive in the case against the remaining versions of the artifact hypothesis, for several reasons.

  First, though spicules in sponges are encased in a thin layer of glassy silica, sponges are generally considered to be a soft-bodied organism because of the predominantly soft tissues out of which the rest of their bodies are made. Moreover, the cells of all embryos during their earliest embryonic stages are soft. Even in animals that have internal or external skeletons, the nascent forms of these hard parts do not emerge until gastrulation, partway through the process of embryological development. Thus, discovery of an embryo in the earliest stages of cell division shows beyond a doubt that Precambrian sedimentary rocks can, under the right circumstances, preserve soft-bodied organisms.

  It also established something else. J. Y. Chen found these sponge embryos beneath the Cambrian–Precambrian boundary in late Precambrian rock. Yet these Precambrian layers did not preserve remains of any clearly ancestral or intermediate forms leading to the other main groups of Cambrian animals. This raised an obvious question. If the Precambrian sedimentary strata beneath the Maotianshan Shale preserved the soft tissues of tiny, microscopic sponge embryos, why didn’t they also preserve the near ancestors of the whole animals that arose in the Cambrian, especially since some of those animals must have had at least some hard parts as a condition of their viability? If these strata could preserve embryos, then they should have preserved fully developed animals—at least, if such animals were present at the time. That well-developed, clearly ancestral animal forms were not preserved, when tiny sponge embryos51 were, strongly indicates that such forms were simply not present in the Precambrian layers.

  FIGURE 3.7

  Figure 3.7a (above): Photographs of fossilized sponge embryos in the early stages of cell division, showing the sponge at the 8-cell stage of division, with four of its cells marked in the foreground. Figure 3.7b (right): A close-up image of a fossil of a sponge embryo cell revealing numerous yolk granules inside. Courtesy Paul Chien.

  Of course, there are conditions under which fossils are unlikely to be preserved. We know, for example, that near-shore sands do not favor preservation of detail, let alone the fine detail of very small organisms a millimeter or less in length.52 Even so, such considerations do little to bolster the artifact hypothesis. The sedimentary environments that produced the carbonates, phosphorites, and shales of the Precambrian strata beneath the Maotianshan Shale, for example, would have provided a congenial environment for fossilizing all sorts of creatures during Precambrian times.

  Advocates of the artifact hypothesis need to show not just that certain factors discourage preservation in general. No one disputes that. What they need to show is that these factors were ubiquitous in Precambrian depositional environments worldwide. If near-shore sands characterized all Precambrian sedimentary deposits, then paleontologists would not expect to find any fossils there, at least not any tiny ones. Yet clearly this is not the case. Precambrian strata include many types of sediments that can preserve—and in the case of the Doushantuo formation in China, have preserved—animal remains in fine detail, including small and vulnerable sponge embryos.

  Moreover, geologists Mark and Dianna McMenamin have noted that in many other Cambrian locales around the world, including one in Newfoundland that they have studied extensively, the pattern of sedimentation changes very little across the Cambrian–late Precambrian boundary, suggesting that many Precambrian environments would have provided equally good environments for the preservation of fossils.53

  In their 2013 book, The Cambrian Explosion, paleontologists James Valentine and Douglas Erwin go further. They note that many late Precambrian depositional environments actually provide more favorable settings for the preservation of fossils than those present in the Cambrian period. As th
ey note, “a revolutionary change in the sedimentary environment—from microbially stabilized sediments during the Ediacaran [late Precambrian] to biologically churned sediments as larger, more active animals appeared—occurred during the early Cambrian. Thus, the quality of fossil preservation in some settings may have actually declined from the Ediacaran to the Cambrian, the opposite of what has sometimes been claimed, yet we find a rich and widespread explosion of [Cambrian] fauna.”54

  Statistical Paleontology

  Recent work in a field known as statistical paleontology casts further doubt on the artifact hypothesis. Since the discovery of the Burgess Shale, Precambrian-and Cambrian-era discoveries have repeatedly uncovered fossil forms that either establish radically disparate new forms of life or, increasingly, forms that fall into existing higher taxonomic groups (such as class, subphylum, or phylum).

  As a result, the fossil record amply documents organisms corresponding to the terminal branches on the Darwinian tree of life (animal forms representing new phyla or classes, for example), but it fails to preserve those organisms representing the internal branches or nodes leading to these representatives of novel phyla and classes of Cambrian-era animals. Yet these intermediates are the very forms required to connect the terminal branches to form a coherent evolutionary tree and establish that the representatives of the Cambrian animals did arise by means of a gradual evolutionary process from simpler Precambrian ancestors.

  Recall that Louis Agassiz thought that this pattern could not be explained by appealing to an incomplete fossil record, because the fossil record was strangely selective in its incompleteness, preserving abundant evidence of the terminal branches but consistently neglecting to preserve the representatives of the internal branches or nodes.

  Contemporary paleontologists, such as Michael Foote at the University of Chicago, have come to a similar conclusion. Foote has shown, using statistical sampling analysis, that as more and more fossil discoveries fall within existing higher taxonomic groups (e.g., phyla, subphyla, and classes), and as they fail to document the rainbow of intermediate forms expected in the Darwinian view of the history of life, it grows ever more improbable that the absence of intermediate forms reflects a sampling bias—that is, an “artifact” of either incomplete sampling or preservation.

  This kind of analysis merely quantifies what, in other circumstances, we would sense intuitively. Imagine that you reach into an enormous barrel full of marbles and randomly pull out a yellow, a red, and a blue marble. At this point your brief sampling should leave you undecided as to whether you have a representative sample of the barrel’s contents. You might at first imagine that the barrel also contains marbles representing a rainbow of intermediate colors. But as you continue to sample from every place in the barrel and find that the barrel disgorges only those same three colors you begin to suspect that it may offer a much more limited selection of colors than, say, the rack of color samples at your local paint store.

  Over the past 150 years or so, paleontologists have found many representatives of the phyla that were well known in Darwin’s time (by analogy, the equivalent of the three primary colors) and a few completely new forms altogether (by analogy, some other distinct colors such as green and orange, perhaps). And, of course, within these phyla there is a great deal of variety. Nevertheless, the analogy holds at least insofar as the differences in form between any member of one phylum and any member of another phylum are vast, and paleontologists have utterly failed to find forms that would fill these yawning chasms in what biologists call “morphological space.” In other words, they have failed to find the paleontological equivalent of the numerous finely graded intermediate colors (Pendleton blue, dusty rose, gun barrel gray, magenta, etc.) that interior designers covet. Instead, extensive sampling of the fossil record has confirmed a strikingly discontinuous pattern in which representatives of the major phyla stand in stark isolation from members of other phyla, without intermediate forms filling the intervening morphological space.

  Foote’s statistical analysis of this pattern, documented by an ever increasing number of paleontological investigations, demonstrates just how improbable it is that there ever existed a myriad of as yet undiscovered intermediate forms of animal life—forms that could close the morphological distance between the Cambrian phyla one tiny evolutionary step at a time. In effect, Foote’s analysis suggests that since paleontologists have reached repeatedly into the proverbial barrel, sampled it from one end to the other, and found only representatives of various radically distinct phyla but no rainbow of intermediates, we shouldn’t hold our breath expecting such intermediates to eventually emerge. He asks “whether we have a representative sample of morphological diversity and therefore can rely on patterns documented in the fossil record.” The answer, he says, is yes.55

  By this affirmation, he doesn’t mean that there are no biological forms left to discover. He means, rather, that we have good reason to conclude that such discoveries will not alter the largely discontinuous pattern that has emerged. “Although we have much to learn about the evolution of form,” he writes, the statistical pattern created by our existing fossil data demonstrates that “in many respects our view of the history of biological diversity is mature.”56

  Chengjiang and the Cambrian Conundrum57

  The Cambrian and Precambrian fossils from southern China have rendered the mystery associated with the Cambrian explosion more acute in other ways as well. First, the fossil finds in southern China coupled with advances in radiometric dating techniques applied to other Cambrian-era strata have allowed scientists to reassess the duration of the Cambrian explosion. As the name implies, the fossils documenting the Cambrian explosion appear within a relatively narrow slice of geologic time. Until the early 1990s, most paleontologists thought the Cambrian period began 570 million and ended 510 million years ago, with the Cambrian explosion of novel animal forms occurring within a 20-to 40-million-year window during the lower Cambrian period.

  Two developments have led paleontologists and geochronologists to revise those estimates downward. First, in 1993, radiometric dating of zircon crystals from formations just above and below Cambrian strata in Siberia allowed for a precise redating of Cambrian strata. Radiometric analyses of these crystals fixed the start of the Cambrian period at 544 million years ago,58 and the beginning of the Cambrian explosion itself to about 530 million years ago (see Fig. 3.8). These studies also suggested that the explosion of the novel Cambrian animal forms occurred within a window of geologic time much shorter than previously believed, lasting no more than 10 million years, and the main “period of exponential increase of diversification” lasting only 5 to 6 million years.59

  FIGURE 3.8

  The Cambrian explosion occurred within a narrow window of geological time.

  Geologically speaking, 5 million years represents a mere 1/10 of 1 percent (0.11 percent, to be precise) of earth’s history. J. Y. Chen explains that “compared with the 3-plus-billion-year history of life on earth, the period [of the explosion] can be likened to one minute in 24 hours of one day.”60

  Some geologists or evolutionary biologists dispute these numbers, but they do so by redefining the Cambrian explosion as a series of separate events rather than using the term to refer to the main radiation of new body plans in the lower Cambrian. In 2009, I participated in a debate in which one of my opponents, paleontologist Donald Prothero, from Occidental College, used this common rhetorical strategy to minimize the severity of the Cambrian mystery. In his opening statement, he claimed that the Cambrian explosion actually took place over an 80-million-year period of time and that consequently those who cited the Cambrian as a challenge to the adequacy of the neo-Darwinian theory were mistaken. As I was listening to his opening statement, I consulted his textbook to see how he had derived his 80-million-year figure. Sure enough, he had included in the Cambrian explosion three separate pulses of new innovation or diversification, including the origin of a group of late Precambrian organisms
called the Ediacaran or Vendian fauna. He also included not only the origin of the animal body plans in the lower Cambrian, but also the subsequent minor diversification (variations on those basic architectural themes) that occurred in the upper Cambrian. He included, for example, not just the appearance of the first trilobites, which occurred suddenly in the lower Cambrian, but also the origin of a variety of different trilobite species later from the upper Cambrian.

  In my response to Prothero, I noted that he was, of course, free to redefine the term “Cambrian explosion” any way he liked, but that by using the term to describe several separate explosions (of different kinds), he had done nothing to diminish the difficulty of explaining the origin of the first explosive appearance of the Cambrian animals with their unique body plans and complex anatomical features. Beyond this, as we’ll see in the next chapter, the Vendian organisms may not have been animals at all, and they bear little resemblance to any of the animals that arise in the Cambrian. We’ll also see that most, if not all,61 of these organisms actually went extinct well before the origin of the animals that first appear in the lower Cambrian and so they do little to minimize the problem of the explosive origin of animals.

  In any case, expanding the definition of the Cambrian explosion only obscures the real challenge posed by the event, a challenge underscored by the discoveries at Chengjiang. An analysis by MIT geochronologist Samuel Bowring has shown that the main pulse of Cambrian morphological innovation occurred in a sedimentary sequence spanning no more than 6 million years.62 Yet during this time representatives of at least sixteen completely novel phyla and about thirty classes first appeared in the rock record. In a more recent paper using a slightly different dating scheme, Douglas Erwin and colleagues similarly show that thirteen new phyla appear in a roughly 6-million-year window.63 As we’ve seen, among these animal forms were the first trilobites, with their lens-focusing compound eyes among other complex anatomical features. The problem of explaining how so many new forms and structures arose so rapidly in the first explosive period of the Cambrian remains, whether or not one decides to include within the designation “Cambrian explosion” other distinct events (see Part Two).