Eight Little Piggies
The straightening rod tries to push a jutting thumb of oddness back into a linear array by designating the small and peculiar group as intermediary between two large and conventional categories. The Onychophora owe whatever small recognition they possess to this strategy—for they have most commonly been interpreted as living relicts of the evolutionary transition between two great phyla: the Annelida (segmented worms, including leeches and the common garden earthworm) and the Arthropoda (about 80 percent of animal species, including insects, spiders, and crustaceans). In this argument, Peripatus is a superworm for its legs and a diddly fly for building these legs without true segments.
A third possibility obviously exists and clearly bears interesting implications. This third way has been supported, often by well-respected taxonomists, but our general preference for shoehorns and straightening rods has given it short shrift. The Onychophora, under this view, might represent a separate group, endowed with sufficient anatomical uniqueness to constitute its own major division of the animal kingdom, despite the low diversity of living representatives. After all, the criterion for separate status should be degree of genealogical distinctness, not current success as measured by number of species. A lineage may need a certain minimal membership just to provide enough raw material so that evolution can craft sufficient difference for high taxonomic rank. But current diversity is no measure of available raw material through geological history. Evolution is ebb and flow, waxing and waning; once-great groups can be reduced to a fraction of their past glory. A great man once told us that the last shall be first, but just by the geometry of evolution, and not by moral law, the first can also become last. Perhaps the Onychophora were once a much more diverse group, standing wide and tall in their distinctness, while Peripatus and its allies now form a pitifully reduced remnant.
(By speaking of potential distinctness, I am not making an untenable claim for total separation without any relationship to other phyla. Very few taxonomists doubt that onychophorans, along with other potentially distinct groups known as tardigrades and pentastomes, have their evolutionary linkages close to annelids and arthropods. But this third view places onychophorans as a separate branch of life’s tree—splitting off near the limbs of annelids and arthropods and eventually joining them to form a major trunk—whereas the shoehorn would stuff onychophorans into the Arthropoda, and the straightening rod would change life’s geometry from a tree to a line and place onychophorans between primitive worms and more advanced insects.)
We can only test this third possibility by searching for onychophorans in the fossil record—a daunting task because they have no preservable hard parts and therefore do not usually fossilize. I write this essay because several striking new discoveries and interpretations, all made in the past year or two, now point to a markedly greater diversity for onychophorans right at the beginning of modern multicellular life, following the Cambrian explosion some 550 million years ago. These discoveries arise from two fortunate circumstances: First, onychophorans have been found in the rare soft-bodied faunas occasionally preserved by happy geological accidents in the fossil record; second, some ancient onychophorans possessed hard parts and can therefore appear in ordinary fossil deposits.
I fully realize that this expansion in onychophoran diversity at the beginning of multicellular animal life can scarcely rank as the hottest news item of the year. Most readers of these essays, after all, have probably never heard the word onychophoran and, lamentably, have no acquaintance with poor, lovely Peripatus. So why get excited about old onychophorans if you never knew that modern ones existed in the first place? Do hear me out if you harbor these doubts. Much more than Peripatus lies at stake, for validation of the third position—that onychophorans represent a separate branch of life’s tree—has broad and interesting implications for our entire concept of evolution and organic order. I also think that you will marvel at the details of these early onychophorans for their own sake—and their weirdness.
We have actually known a bit about ancient onychophorans for most of this century, thanks once again to that greatest of treasure troves for soft-bodied fossils, the Burgess Shale. In 1911, two years after discovering the Burgess Shale, C. D. Walcott gave the unpronounceable name Aysheaia (we generally call it “a-shy-a” in the trade) to an animal that he described as an annelid worm. Many taxonomists, just viewing Walcott’s illustrations, immediately saw that the creature looked much more like an onychophoran. In 1931, G. Evelyn Hutchinson, who became the world’s greatest ecologist and was, perhaps, the finest person I have ever had the privilege of knowing, published a definitive account on the onychophoran affinities of Aysheaia. Hutchinson had studied Peripatus in South Africa and he knew onychophoran anatomy intimately. As an ecologist, he was powerfully intrigued by the issue of how an ordinary marine invertebrate like Aysheaia could evolve into a terrestrial creature like Peripatus with such minimal change in outward anatomy. (Aysheaia had fewer pairs of legs and fewer claws per leg than do modern onychophorans. It also bore a terminal mouth at the body’s end, while living onychophorans have a ventral mouth on the underside. In addition, Aysheaia had no slime papillae and could not use such a device to shoot sticky stuff through ocean waters in any case. But, all told, the differences are astonishingly slight for more than 500 million years of time and a maximal ecological shift from ocean to land.)
One other ancient onychophoran was recognized before last year—a European form named Xenusion, found during the 1920s. But Aysheaia and Xenusion did not shake the shoehorn or the straightening rod. Only two fossils, both so similar to modern forms, do not make an impressive show of diversity. Onychophorans remained a tiny and uniform group, ripe for stuffing in or between larger phyla and not meriting a status of its own.
In the last essay, I described the beginning of the onychophoran coming of age (I was going to say “renaissance,” but a renaissance is a rebirth, and onychophorans never had an earlier period of flowering in our consciousness). I described the discovery in China of the animal that bore the small, circular, meshwork plates known for many years from lowermost Cambrian rocks as Microdictyon. This fossil comes from the remarkable Chengjiang fauna of south-central China, a Burgess Shale equivalent (although slightly older), with beautiful soft-bodied preservation of many animals already known from the more famous Canadian site (and several novelties as well, including the Microdictyon animal). The plates called Microdictyon are attached in pairs to the side of the animal just above the junction of paired lobopods with the trunk of the body. The animal itself looks like an onychophoran. If this interpretation holds, then some ancient onychophorans had hard parts. The Chengjiang fauna also contains a second probable onychophoran with plates, named Luolishania.
Thus, the early fossil record of onychophorans had begun to expand in numbers and anatomical variety, including fully soft-bodies forms like Aysheaia and creatures with small pairs of plates like Microdictyon and Luolishania. But the big boost, the event that might finally push onychophorans over the border of distinct respectability, finally occurred on May 16, 1991, when the Swedish paleontologist L. Ramsköld and his Chinese colleague Hou Xianguang published an article in the British journal Nature (science, at its best, is truly international—see bibliography).
Ramsköld and Hou dropped a bombshell that makes elegant sense of a major paleontological puzzle of recent years. In 1977, Simon Conway Morris described the weirdest of all Burgess Shale organisms with the oddest of all monikers: Hallucigenia, named, as Simon wrote, for “the bizarre and dream-like appearance of the animal.” He described Hallucigenia (see figure) as a tubular body supported by seven pairs of long, pointed spines—not jointed arthropod appendages or fleshy lobopods, but rigid spikes. In Conway Morris’s reconstruction, a single row of seven fleshy tubes, each ending in a pair of little pincers, runs along the back, with a tuft of six smaller tubes, perhaps in three pairs, behind the larger seven. The head, not well preserved on any specimen, was depicted as a bulbous extension and the tail a
s a straight, upward-curving tube.
Hallucigenia was bizarre enough in appearance, but even more puzzlement attended the issue of how such a creature could function. In particular, how could a tiny animal, no more than an inch in length, be stable on seven pairs of rigid spikes for “legs”? And if stable, how could it possibly move? Some of our best functional morphologists, including Mike Labarbera of the University of Chicago, struggled with this issue and found no resolution.
The unlikely morphology, and the even more troublesome question of function, led many paleontologists to dispute Conway Morris’s reconstruction (and Simon himself also began to doubt his original conclusions). In my book Wonderful Life, I presented Conway Morris’s original version and then opted for a different interpretation proposed by several colleagues before me—that Hallucigenia is a part broken off from a larger (and still unknown) animal. I wrote:
Left: Conway Morris’s original reconstruction of Hallucigenia. Simon Conway Morris (1977): Reprinted by permission of Palaeontology. Right top: Ramsköld and Hou’s inversion of Hallucigenia as an onychophoran. Reprinted by permission from Nature; Copyright © 1991 Macmillan Magazines Limited. Right bottom: Ramsköld and Hou’s reconstruction of the new Chengjiang onychophoran with side plates and spines. Reprinted by permission from Nature; Copyright © 1991, Macmillan Magazines Limited.
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.
I received several dozen much appreciated letters from readers of my book, suggesting different reconstructions for some of the oddball creatures of the Burgess Shale. Hallucigenia received the lion’s share of attention—and one suggestion cropped up again and again, in at least twenty separate letters. These correspondents, nearly all amateurs in natural history, pointed out that Hallucigenia would make much more sense turned upside down—for the spines, which never made any sense as organs of locomotion, could then function far more reasonably for protection!
I responded to these letters with, for me, the decisive rejoinder that a single row of tentacles (Simon’s version of the upper surface) would work even more poorly than paired spines as devices of locomotion. How could an animal balance, not to mention hop around, on a single row of tentacles? Yet I couldn’t deny that everything else made more sense upside down.
It doesn’t happen often, but if Ramsköld and Hou are correct—and I think they are—then the gut feeling of amateurs has triumphed over the weight of professional opinion. For Ramsköld and Hou have, unbeknownst to them of course, followed the advice of my correspondents. They have turned Hallucigenia upside down, but with an added twist (intellectual, not geometric) as well—they have inverted it into an onychophoran!
Ramsköld and Hou present two major arguments for their inversion of Hallucigenia. First, they must tackle the issue that hung me up: How can a single row of tentacles function as legs? They acknowledge the problem, of course, but suggest that Conway Morris was wrong and that two rows of paired tentacles are actually present along the surface that he called dorsal, or topmost. If Ramsköld and Hou are correct, then the major objection to reversing Hallucigenia disappears—for two rows of flexible tentacles look like the ordinary legs of a bilaterally symmetrical creature. Moreover, when you turn Hallucigenia upside down on the assumption that two rows of tentacles adorn the topside of Conway Morris’s version, then the inverted beast immediately says “onychophoran” to any expert, for the little paired pincers at the end of each tentacle become dead ringers for onychophoran claws. Ramsköld and Hou have not yet developed enough evidence to prove the double row of tentacles conclusively, but our museum at Harvard contains the sample best suited for resolving this issue—a slab of rock with more than a dozen Hallucigenia specimens. I have lent this slab to Ramsköld and Hou, and I suspect that an answer will soon be forthcoming.
Second, they must explain how an onychophoran could possess the several pairs of long, pointed, upward-protruding spines that an inversion of Hallucigenia places along the top edge of the animal—for some fossil onychophorans bear plates. (Microdictyon and Luolishania as previously discussed), but none yet described carry spines. Here, Ramsköld and Hou present compelling evidence in a form much favored by natural historians—a sequence or continuum linking a strange and unexpected form to something familiar through a series of intermediates.
Start the series with Microdictyon. This animal, probably an onychophoran, carried pairs of flat plates along the side of its body just above the insertion of lobopods. Go next to a new and as yet unnamed “armored lobopod,” again from the prolific Chengjiang fauna. This clear onychophoran also bore paired plates in the same position as in Microdictyon. But each plate now carries a small spine (see figure)—nothing like the elongation in Hallucigenia, but evidence that onychophoran plates can support spines. For a third step, go to isolated plates collected in lower Cambrian rocks of North Greenland by J. S. Peel and illustrated by Swedish paleontologist Stefan Bengtson in a commentary in Nature written to accompany Ramsköld and Hou’s paper. These Greenland plates have the same meshwork structure as those of Microdictyon—and onychophoran affinity seems a reasonable conjecture (although in this case, we have only found preserved plates, not the entire body). But the Greenland plates carry spines verging on the length of Hallucigenia spikes. We now require only a small step to a fourth term in the series—to an onychophoran bearing plates with highly elongated spines: in other words, to Hallucigenia turned upside down!
We are witnessing a veritable explosion of Cambrian onychophorans—Aysheaia and Xenusion with their soft bodies, Microdictyon and Luolishania with plates, the unnamed Chengjiang creature with plates and short spines, the Greenland form with longer spines, and finally, inverted Hallucigenia with greatly extended spines.
The reversal of Hallucigenia has capped and sealed the tale. The larger conclusion seems inescapable: In the great period of maximal anatomical variety and experimentation that followed right after the Cambrian explosion first populated the earth with multicellular animals of modern design, the Onychophora represented a substantial and independent group of diverse and successful marine organisms. The modern terrestrial species are a tiny and peripheral remnant, a bare clinging to life for a lineage that once ranked among the major players. The shoehorn and straightening rod have served us poorly as strategies of interpretation. Groups with few species may be highly distinct in genealogy.
Onychophorans, moreover, are not the only small cluster of straggling survivors within groups that were once major branches of life’s tree. The distinct phylum of priapulid worms, for example, contains fewer than 20 species worldwide, compared with some 8,000 for marine polychaete worms, members of the dominant phylum Annelida. Yet, in the Cambrian period at the beginning of multicellular history, priapulids and polychaetes were equally common and similarly endowed (or so it seemed) with prospects for long-term success. Moreover, just as onychophorans have held on by surviving in the peripheral habitat of terrestrial life (for a formerly marine group), modern priapulids all live in harsh and marginal environments—mostly in cold or deep waters and often with low levels of oxygen.
In recognizing the Onychophora as a distinct group with an ancient legacy of much greater breadth, we may regret the loss of tidiness provided by the shoehorn and straightening rod, but we should rejoice in the interesting conceptual gains. For by our latest reckoning of life’s early history, “uncomfortable” groups like the Onychophora should exist today. We once thought that the history of life moved upward and outward from simple beginnings in a few primitive, ancestral lines to ever more and ever better—the conventional notion that I have called the
cone of increasing diversity. On this model, an ancient and distinctive genealogical status for several small groups (like the Onychophora) makes no sense—for life’s early history, at the point of the cone, shouldn’t have featured many distinct anatomies at all. The large living groups of mollusks, arthropods, annelids, vertebrates, etc.—all of which have fossil records extending back to this beginning—provide quite enough material for legacies from these early times of limited simplicity. But the reinterpretation of the Burgess Shale, and our burgeoning interest in the early history of multicellular life in general, have indicated that the cone model is not only wrong but also backward. Life may have reached a maximal spread of anatomical experimentation in these early days—and later history may be epitomized as a diminution of initial possibilities by decimation, rather than a continual expansion.
In this reversed model of a grass field, with most blades clipped off and just a few proliferating wildly thereafter, we should expect to find some blades that survived the mower but never flowered extensively again—whereas, in the cone model, the forest of blades never existed, and the early history of life provides insufficient raw material for many distinct modern groups like the Onychophora.
However much I may regret the loss of a wonderful weirdo in the reversal of Hallucigenia, and in its consequent change in status from oddball to onychophoran, I am more than compensated by fascinating insight into the history of ancestry for my favorite name bearer, Peripatus. I revel in the knowledge that these marginal and neglected animals belong to a once-mighty group that included armored members with plates and long spines. And I rejoice in the further knowledge thus provided about the strange and potent times of life’s early multicellular history. (My regret, in any case, could not possibly be more irrelevant to nature’s constitution, either now or 500 million years ago. Hallucigenia was what it was. My hopes, and those of any scientist, are only worth considering as potential biases that can block our understanding of nature’s factuality.)