Darwin's Doubt

by Stephen C. Meyer

  The idea that enormous, solid plates actually move, recycling themselves through the plate-tectonic processes of subduction and seafloor spreading, had not yet been proposed. Yet modern plate tectonic theory now affirms that oceanic crustal material eventually plunges back into the earth and melts in a process known as subduction. After surface rocks melt during subduction, they form a new supply of molten magma. Eventually, magma from other locations deep in the earth wells up at mid-oceanic ridges to form new igneous rocks, in a process known as seafloor spreading. It follows that any oceanic sediments deposited atop the oceanic igneous crust will have a limited “life span” on the surface of the earth. Eventually, these sedimentary rocks collide with the continental margin, plunge deep into the upper mantle, and melt to form magma.

  As a consequence of this cycle, the maximum age of any marine sediment is strictly limited. And according to modern estimates, the oldest section of oceanic crust has existed only since the Jurassic (or about 180 million years ago8)—far too young to contain fossil ancestors of the trilobites. As the evidence for plate tectonics mounted, scientists discarded Walcott’s artifact hypothesis and Lipalian interval as nonstarters. Paleontologists today do not expect to find any Precambrian ancestors of the trilobites in oceanic sediments, since they realize that there are no Precambrian sediments in the ocean basins. If Precambrian strata are to be found anywhere, continents are the place.

  Other Versions of the Artifact Hypothesis

  Although Walcott’s proposals to explain away the absence of fossilized ancestors of the Cambrian animals came to naught, other versions of the artifact hypothesis continued to circulate. These proposals take two basic forms. Some scientists claimed, though for different reasons, that the expected Precambrian fossil ancestors had simply not yet been found—that missing fossils were an artifact of the incomplete sampling of the fossil record. Others suggested that Precambrian sedimentary rocks had not preserved the missing fossils—that the incomplete preservation of the Precambrian animals meant the missing fossils were no longer there to be found.

  Walcott rejected the idea that paleontologists simply had not looked in, or sampled, enough places. He noted that geologists already had extensively investigated “the great series of Cambrian and Precambrian strata in eastern North America.” Though they had looked “from Alabama to Labrador; in western North America [and] from Nevada and California far into Alberta and British Columbia, and also in China” their investigations had turned up nothing of significant interest.9 In Walcott’s view the continents simply had not preserved the fossilized remains of the Cambrian ancestors.

  Before Walcott, some geologists had gone a step farther and suggested that all Precambrian sedimentary rocks had been destroyed via extreme heat and pressure, a process called “universal metamorphism.” Walcott rejected this hypothesis, since he himself had encountered a “great series of pre-Cambrian sedimentary rocks on the North American continent” among other places. Other geologists suggested that major bursts of evolutionary innovation occurred only during periods when sedimentary deposition had ceased, thus again resulting in a lack of fossil preservation. But, as Gould remarked of Walcott’s artifact hypothesis, this explanation also appeared to many scientists “forced and ad hoc … born of frustration, rather than the pleasure of discovery.”10

  Contemporary Versions of the Artifact Hypothesis: Too Soft or Too Small

  After the demise of the “universal metamorphism” idea, some paleontologists proposed simpler, more intuitively plausible versions of the artifact hypothesis. They claimed that the proposed intermediate forms leading to the Cambrian animals may have been either too small or too soft, or both, to have been preserved.

  Developmental biologist Eric Davidson, of California Institute of Technology, has suggested that the transitional forms leading to the Cambrian animals were “microscopic forms similar to modern marine larvae” and were thus too small to have been reliably fossilized.11 Other evolutionary scientists, such as Gregory Wray, Jeffrey Levinton, and Leo Shapiro, have suggested that the ancestors of the Cambrian animals were not preserved, because they lacked hard parts such as shells and exoskeletons.12 They argue that since soft-bodied animals are difficult to fossilize, we shouldn’t expect to find the remains of the supposedly soft-bodied ancestors of the Cambrian fauna in the Precambrian fossil record. University of California, Berkeley, paleontologist Charles R. Marshall summarizes these explanations:

  [I]t is important to remember that we see the Cambrian “explosion” through the windows permitted by the fossil and geological records. So when talking about the Cambrian “explosion,” we are typically referring to the appearance of large-body (can be seen by the naked eye) and preservable (and therefore largely skeletonized) forms… . If the stem lineages were both small and unskeletonized, then we would not expect to see them in the fossil record.13

  Though intuitively plausible, several discoveries call into question both of these versions of the artifact hypothesis. As for the idea that the ancestors of the Cambrian animals were too small to be preserved, paleontologists have known for some time that the cells of filament-shaped microorganisms (probably cyanobacteria) have been preserved in ancient Precambrian rocks. Paleobiologist J. William Schopf, of the University of California, Los Angeles, has reported an extremely ancient example of these fossils in the Warrawoona Group strata of western Australia. These fossilized cyanobacteria are preserved in 3.465-billion-year-old bedded cherts (microcrystalline sedimentary rocks).14 The same strata have also preserved stromatolite mats, an organic accretionary growth structure usually indicating the presence of bacteria, within slightly younger dolostone sediments of roughly 3.45 billion years in age (see Fig. 3.5).15

  FIGURE 3.5

  Figure 3.5a (left): Photographs of Cambrian-age fossil stromatolites. Courtesy Wikimedia Commons, user Rygel, M. C. Figure 3.5b (right): Alternating fine-and coarse-layered structure of Precambrian stromatolite fossils shown in cross section. Courtesy American Association for the Advancement of Science, Figure 2B, Hoffman, P., “Algal Stromatolites: Use in Stratigraphic Correlation and Paleocurrent Determination,” Science, 157 (September 1, 1967): 1043–45. Reprinted with permission from AAAS.

  These discoveries pose a problem for the idea that the Cambrian ancestors were too small to survive in the fossil record. The sedimentary rocks that preserve the fossilized cyanobacteria and single-celled algae are far older and, therefore, far more likely to have been destroyed by tectonic activity than those later sedimentary rocks that should have preserved the near-ancestors of the Cambrian animals. Yet these rocks, and the fossils they contain, have survived just fine. If paleontologists can find tiny fossilized cells in these far older and rarer formations, shouldn’t they also be able to find some ancestral forms of the Cambrian animals in younger and more abundant sedimentary rocks? Yet few such precursors have been found.

  There are also several reasons to question the second version of this hypothesis—the idea that the presumed Cambrian ancestors were too soft to be preserved. First, some paleontologists have questioned whether soft-bodied ancestral forms of the hard-bodied Cambrian animals would have even been anatomically viable.16 They argue that many animals representing phyla such as brachiopods and arthropods could not have evolved their soft parts first and then added shells later, since their survival depends upon their ability to protect their soft parts from hostile environmental forces. Instead, they argue that soft and hard parts had to arise together.17 As paleontologist James Valentine, of the University of California, Berkeley, has noted in the case of brachiopods, “The brachiopod Bauplan [body plan] cannot function without a durable skeleton.”18 Or as J. Y. Chen and his colleague Gui-Qing Zhou observe: “Animals such as brachiopods … cannot exist without a mineralized skeleton. Arthropods bear jointed appendages and likewise require a hard, organic or mineralized outer covering.”19

  Because these animals typically require hard parts, Chen and Zhou assume that the ancestral forms of these an
imals should have been preserved somewhere in the Precambrian fossil record if in fact they were ever present. Thus, the absence of hard-bodied ancestors of these Cambrian animals in the Precambrian strata shows that these animals first arose in the Cambrian period. As they rather emphatically insist: “The observation that such fossils are absent in Precambrian strata proves that these phyla arose in the Cambrian.”20

  It should be pointed out that this argument cannot be made for all Cambrian animal groups and, in my view, does not achieve the standing of a “proof” in any case. Many Cambrian phyla, including phyla characterized by mostly hard-shelled animals such as mollusks and echinoderms, do have soft-bodied representatives. The earliest known mollusk, Kimberella, for example, lacked a hard external shell (though it did have other hard parts).21 So, clearly, some mainly hard-shelled Cambrian groups could have had soft-bodied ancestors.

  It is also possible to postulate the existence of an arthropod or brachiopod ancestor—especially some extremely distant ancestor—lacking a hard shell. Soft-bodied onychophorans (velvet worms) were once proposed as ancestors of the arthropods, though more recent studies challenge this idea. Onychophorans themselves arise well after arthropods in the fossil record and cladistics analysis suggests that onychophorans may be a sister, rather than an ancestral, group to arthropods.22 Even so, it’s difficult to disprove a negative: in particular, to foreclose the possibility that arthropods or brachiopods might have had a soft-bodied ancestor deep in the Precambrian.

  Nevertheless, it seems unlikely on a Darwinian view of the history of life that all Cambrian arthropod or brachiopod ancestors, especially the relatively recent ancestors of these animals, would have lacked hard parts entirely. There are many types of arthropods that arise suddenly in the Cambrian—trilobites, Marrella, Fuxianhuia protensa, Waptia, Anomalocaris—and all of these animals had hard exoskeletons or body parts. Moreover, the only known extant group of arthropods without a hard exoskeleton (the pentastomids) have a parasitic relationship with arthropods that do.23 Thus, surely, it seems likely that some of the near ancestors of the many arthropod animals that arose in the Cambrian would have left at least some rudimentary remains of exoskeletons in the Precambrian fossil record—if, in fact, such ancestral arthropods existed in the Precambrian and if arthropods arose in a gradual Darwinian way.

  Moreover, the arthropod exoskeleton is part of a tightly integrated anatomical system. Specific muscles, tissues, tendons, sensory organs—and a special mediating structure between the soft tissue of the animal and the exoskeleton called the endophragmal system—are all integrated to support the process of molting and exoskeletal growth and maintenance that is integral to the arthropod mode of existence. A best-case Darwinian scenario for the origin of such a system would, therefore, envision the “co-evolution” of these separate anatomical subsystems in a coordinated fashion, since some of these anatomical subsystems confer a functional advantage to the animal largely by supporting, and promoting, the growth and maintenance of the exoskeleton (and vice versa). Others would be vulnerable to damage without it. Thus, it seems unlikely that these interdependent subsystems would evolve independently first without an exoskeleton, only to have the exoskeleton arise suddenly as a kind of accretion atop an already integrated system of soft parts at the end of a long evolutionary process.

  This, again, makes it reasonable to expect that at least some rudimentary arthropod hard parts would have been preserved in the Precambrian if arthropods were present then. That such parts are unknown for all Cambrian arthropods (and brachiopods) in a fossil record that presumably favors hard-part preservation, seems at least curious. And it appears, on its face, to support the assertions of those Cambrian paleontologists such as Chen and Zhou who take the absence of any hard parts in the Precambrian record as evidence of the absence of those groups that typically depend on hard parts for their existence.

  In any case, advocates of the artifact hypothesis must at least explain a Cambrian explosion of hard body parts, if not whole Cambrian animals. As paleontologist George Gaylord Simpson noted in 1983, even if it’s true that Precambrian ancestors were not preserved simply because they lacked hard parts, “there is still a mystery to speculate about: Why and how did many animals begin to have hard parts—skeletons of sorts—with apparent suddenness around the beginning of the Cambrian?”24

  There is an additional, more formidable, difficulty for this version of the artifact hypothesis. Although the fossil record generally does not preserve soft body parts as frequently as hard parts, it has preserved many soft-bodied animals, organs, and anatomical structures from both the Cambrian and the Precambrian periods.

  As we saw earlier, Precambrian sedimentary rocks in several places around the world have preserved fossilized colonial blue-green algae, single-celled algae, and cells with a nucleus (eukaryotes).25 These microorganisms were not only small, but they also entirely lacked hard parts. Another class of late Precambrian organisms called the Vendian or Ediacaran biota included the fossilized remains of many soft-bodied organisms, including many that may well have been lichens, algae, or protists (microorganisms with cells containing nuclei). Cambrian-era strata themselves preserve many soft-bodied creatures and structures. The Burgess Shale in particular preserved the soft parts of several types of hard-bodied Cambrian animals, such as Marrella splendens,26 Wiwaxia,27 and Anomalocaris. The Burgess Shale also documents entirely soft-bodied representatives28 of several phyla, including:

  Cnidaria (represented by an animal called Thaumaptilon, a feather-shaped colonial organism formed from smaller soft sea anemone–like animals)29

  Annelida (represented by the polychaete worms Burgessochaeta and Canadia)30

  Priapulida (represented by Ottoia, Louisella, Selkirkia—all worms with a distinctive proboscis)31

  Ctenophora (represented by Ctenorhabdotus, a gelatinous animal with a translucent body similar to a modern comb jelly)32

  Lobopodia (represented by Aysheaia and Hallucigenia, segmented soft-bodied animals with many legs)33

  The Burgess also preserves soft-bodied animals of unknown affinities, such as Amiskwia, a gelatinous air mattress–like animal;34 Eldonia, a jellyfish-like animal with a much more complex anatomy than a modern jellyfish;35 and the aforementioned, difficult to classify, Nectocaris.36 As Simon Conway Morris notes, “The existing [Burgess] collections represent approximately 70,000 specimens. Of these, about 95 percent are either soft-bodied or have thin skeletons.”37

  The Chengjiang Explosion

  Any doubts about the ability of sedimentary rocks to preserve soft and small body parts were permanently laid to rest by a series of dramatic fossil finds in southern China beginning in the 1980s.

  In June 1984, paleontologist Xian-Guang Hou journeyed to Kunming, in southern China, to prospect for fossilized samples of a bivalved arthropod called a bradoriid.38 The area around Kunming in the Yunnan Province was well known for its lower Cambrian strata and typical Cambrian-era fossils, such as bradoriids and trilobites, both of which were relatively easy to preserve because of their characteristically hard exoskeletons. In 1980, Hou had found many bradoriid samples in a geological formation called the Qiongzhusi Section near Kunming.

  In the summer of 1984 Hou traveled to the town of Chengjiang to look for bradoriids in another geological formation called the Heilinpu Formation. His efforts there yielded little success. As a result, he turned his attention to another outcrop, a sedimentary sequence now called the Maotianshan Shale. Hou’s team set farmworkers to digging out and scouring the mudstone blocks. His book, The Cambrian Fossils of Chengjiang, China, describes what happened next:

  At about three o’clock in the afternoon of Sunday July 1, a semicircular white film was discovered in a split slab, and was mistakenly thought to represent the valve of an unknown crustacean. With the realization that this … represented a previously unreported species, breaking of the rock in a search for additional fossils continued apace. With the find of another specimen, a 4–5 cm long anim
al with limbs preserved, it became apparent that here was nothing less than a soft-bodied biota.39

  Hou remembers the Cambrian specimen vividly, for it appeared “as if it was alive on the wet surface of the mudstone.”40 Redoubling their efforts, the researchers quickly uncovered the fossilized remains of one extraordinary soft-bodied animal after another. Most of the fossils were preserved as flattened two-dimensional imprints of three-dimensional organisms, although, as Hou observes, “some retain a low three-dimensional relief.”41 Most important, he notes, “The remains of hard tissues, such as the shells of brachiopods or the carapaces of trilobites, are well represented in the Chengjiang fauna, but less robust tissues, which are usually lost through decomposition, are also beautifully preserved.”42

  As the result of the very fine, small-grained sediments in which they were deposited, the Chengjiang fossils preserved anatomical details with a fidelity surpassing even that of the Burgess fauna.43 The Maotianshan Shale also preserved an even greater variety of soft-bodied animals and anatomical parts than the Burgess Shale had done. In the ensuing years, Hou and his closest colleagues, J. Y. Chen and Gui-Qing Zhou, found many excellent examples of well-preserved animals lacking even a keratinized exoskeleton, including soft-bodied members of phyla such as Cnidaria (corals and jellyfish), Ctenophora (comb jellies), Annelida (a type of “ringed” segmented worm), Onychophora (segmented worms with legs), Phoronida (a tubular, filter-feeding marine invertebrate), and Priapulida (another distinctive type of worm).44 (See Fig. 3.6.)

  They found fossils preserving the anatomical details of numerous soft tissues and organs such as eyes, intestines, stomachs, digestive glands, sensory organs, epidermes, bristles, mouths, and nerves.45 They also discovered jellyfish-like organisms called Eldonia, which exhibit delicate, soft body parts such as radiating water canals and nerve rings. Other fossils even revealed the contents of the guts of several animals.46