The Great Fossil Enigma
Page 39
Discover magazine, writing about the top fifty science stories of 1992, proclaimed, “If Sansom is right, vertebrates are as old as the oldest conodonts – that is, at least 515 million years old.” Here Sansom recounts the moment of first discovery in Newcastle: “We were sitting there, and these images appeared that were quite spectacular…. It was very, very clear we were looking at enamel, bone, and cartilage.” He continued: “It came as quite a shock…. The structures were actually clearer in the conodonts than in human teeth.” The Northern Echo, a British regional paper, gave the story the headline “Meet your great, great, great, great…grandad.”12
Those pondering the origins of vertebrate life could no longer ignore the conodont. Indeed, conodont workers of all persuasions might have felt reason to celebrate the suddenly elevated status of their fossils. Chris Barnes, who was immersed in events in Earth history, wondered if a particular event in the Ordovician had marked an important moment in the evolution of vertebrate tissue. For those already advancing on the vertebrate from other directions, especially the Leicester team, there could be no better piece of news; it would form a vital component in their arguments for years to come. Briggs, who had been asked by Science to write a “Perspectives” piece to accompany the article, immediately celebrated the impact of this discovery, which had thrown ideas regarding early vertebrate evolution into disarray and in an instant expanded the number of known vertebrate genera in the Cambrian-Ordovician from 5 to 150 (nearly all of which were conodonts). “In any event the vertebrates can now be added to the list of major metazoan taxa that appeared during the Cambrian radiation,” Briggs observed. Aldridge now felt he could at last state, without fear of contradiction, that “conodonts are extinct primitive soft-bodied vertebrates” and even go so far as to suggest “the soft tissue evidence now confirms their place among our earliest ancestors.” Henry Gee, an editor of Nature and an important sounding board for the British workers, later reflected on the implications of finding this new piece in the vertebrate jigsaw. He recalled that almost since the beginning of paleontology, it had been believed that vertebrate life began with sluggish armored fishes, but now these were as nothing compared to the diversity of conodonts that joined them and in most cases were known only by their tooth-like hard parts: “Like so many geological Cheshire cats leaving their smiles to posterity.” “Could later vertebrates have had instead a common ancestor among the conodonts animals?” he wondered. Reflecting on his childhood inspirations and the bizarre armored fishes that once populated his mind, he reimagined the natural history museum of that time: “The Hall of Fossil Fishes might have presented a truer picture of the Age of Fishes had the bulk of its space been devoted to conodonts, with just one small case, in a dark corner, devoted to pteraspids, sarcopterygians and their relatives.” But precisely where conodonts fitted into the vertebrate family tree remained unresolved: “The mystery continued: it was a bizarre Borgesian plot made real, of future palaeontologists obliged to reconstruct human history based on several million sets of dentures, and nothing else.” For an object vanishingly small, its impact was never less than remarkable. Even Gee, overcome as he was with lyrical metaphors, found himself affected.13 The discovery was a gauntlet thrown at the feet of those who were shaping the picture of early vertebrate evolution. It demanded that science return to its microscopes and that biologists consider fossils.
The question of whether conodont elements really were vertebrate teeth was now being tackled on two fronts. While in Birmingham, Smith, Sansom, and their collaborators were revealing the fossils’ material affinity, in Leicester, Mark Purnell was discovering if they could have functioned like teeth. Purnell had drawn a line in the sand at 1982 and corralled everything before that year, including all those functional models produced in the 1970s, into a period of prehistory he called “pre-animal.” The animal discoveries had provided sufficient new and quite specific data on many key aspects of apparatuses that had previously been matters of speculation. In a paper with Canadian conodont worker Peter von Bitter, Purnell attempted to reinterpret the function of the pair of robust blade-like elements that sat at the back of the apparatus. Were they covered in soft tissue? Did they roll particles of food between them? Did they bite, shear, or operate like a gate? Noticing some small nodes at one end of each blade, they reasoned that these nodes were set opposite each other, offering a rasp-like functionality in a scissor-like action. By implication this meant that the animal must have grasped “large” food particles with the anterior part of the apparatus rather than used these elements as filter devices.14
It had been while still working in Canada that Purnell had concluded that a fundamental objection to all previous functional models was a reliance on a test of “plausibility.” Now he asked, could this be avoided? Were there other aspects of the assemblage that might more objectively point to function? At its most fundamental, the argument over function that had developed in the 1970s had revolved around whether elements were teeth or components in a filtering apparatus. Zoologists had long known that different parts of an organism grow at different rates according to the animal's needs, that growth patterns in some respects reflect mode of life. Comparative data had suggested to Purnell that a filter-feeding animal would need to preferentially grow the net – the comb-like elements near the opening of the mouth. What he found, however, was that the robust mashing elements grew more rapidly. This suggested to him that they had a molar-like function. The other elements grew at the same rate as, or slightly more slowly than, the animal as a whole. Purnell could also quash rumors that conodont teeth were, like those in the hagfish, replaced in life or lost or reabsorbed by the living animal. He deduced that the conodont animal must have been predatory and used the front of its apparatus to grasp and the rear elements to “chew.”15
Bob Nicoll, however, was unimpressed both by the vertebrate and these food-munching teeth; he continued to defend his filtering model.16 Hit lacked fully formed jaws, Nicoll argued, how could the animal – with a muscular arrangement something like that in amphioxus – exert sufficient force to cut?
Meanwhile, in Bristol, Briggs was crafting yet another interpretive lens. He had set his postdoctoral researcher, Amanda Kear, to watch over the dead; to gather information on the decay of amphioxus (the lancelet or Branchiostoma). Their aim was to better understand the preservation of the conodont animals and other so-called primitive vertebrates.17 Lancelets were considered possible “living descendents of the ancestors of vertebrates.” Briggs and Kear's experiments revealed that the noto-chord was resilient to decay, recording its former existence in two lines like those seen in the Scottish fossils. This work permitted interpretation of these specimens to move beyond a priori assumptions about the nature of the animal and even explain some of the inadequacies in the fossils themselves. It altered expectations and clarified interpretations; it, too, moved arguments away from the test of plausibility.
These were heady days for a science that would soon be reaching the peak of its activity. In 1993, and now in possession of ten Granton specimens, the three original authors, plus Paul Smith and Neil Clark, published the final, definitive, anatomical account of the Scottish animals in the Philosophical Transactions of the Royal Society.18 Aided by discoveries in South Africa, the two lobe-like head structures were now interpreted as hollow sclerotic cartilages that once supported the eyes. One specimen also seemed to preserve the animal's hearing apparatus. These were capsules that Aldridge and Briggs had once thought eyes. They had had their minds changed by the discovery of similar structures in the Mazon Creek lamprey, Mayomyzon, and hagfish Myxinikela. They also believed they could see faint traces of gill structures like those described in the Silurian relative to the lamprey, Jamoytius. The twin lines running down the body of the animal were now compelling evidence of a notochord, and in one heavily phosphatized specimen the area between these lines displayed a fibrous structure that just might indicate the notochord sheath. Again something similar had been seen in th
e Mazon Creek chordate Gilpichthys. Some aspects, however, remained frustratingly indefinite – a whisker away from being the proof they needed. The dorsal nerve cord was among these hazy features. But with each new specimen the beautifully preserved V-shaped segmentation could not be doubted. It formed “one of the most compelling pieces of evidence for chordate affinity.” The separation of the chevrons could now be attributed to postmortem shrinkage as could be seen in Briggs and Kear's decay experiments. In one specimen, Aldridge's team even wondered if they could see evidence of original muscle fibers. Their conclusions were emphatic: “The evidence of the soft-part anatomy, together with features of element histology, show that the conodonts are vertebrates.” The Birmingham team's corroborative work had given Aldridge and his colleagues a sense of certainty: “Other hypotheses that have been forwarded in recent years can now be refuted.”
Sweet's tests had, in part at least, been answered. The notochord had gained solidity, and the nerve cord and gill structures tentatively interpreted. Sweet's suggestion that other invertebrate groups might preserve the trunk structures seen in the conodont animal had already been firmly denied. Again, Conway Morris and Janvier provided important supporting voices for believing the animal a chordate. Indeed, for many, including Aldridge and his colleagues, the question had since 1987 become “which chordate?” Nowlan and Carlisle, for example, had been amongst those who had placed the animal with amphioxus. However, new certainty regarding the eyes suggested that interpretation was incorrect.
The worm – the worm that had grown out of necessity in response to the discomfort caused by Pander's vertebrate – was now very dead and the reinstatement of Pander's imagined fish in full swing. However, for many involved in the debate, elevation into the hallowed realms of human sisterhood depended not simply on the fossil but also on what one could call a vertebrate. This had long been disputed territory, and the current view promulgated by Janvier was that the vertebrates consisted of animals with a cranium and vertebral elements; those which just possessed the cranium were then referred to as craniates and not considered vertebrates. It was, however, a messy division. What Aldridge and his collaborators required was a chink in Janvier's armor, and they found that chink by appealing to “common usage,” which placed Janvier's craniates in with the vertebrates. The argument, which came from developmental biology, drew upon a shared feature present in the embryos of all vertebrates: the neural crest, a feature from which vertebrae, cranium, and other defining elements emerged. Aldridge remained convinced that the conodonts were anatomically close to the hagfish.
One important hurdle to winning this argument seemed to be to persuade Janvier, but he and Peter Forey of the Natural History Museum in London were not convinced, and certainly not by Sansom's work. They argued that the similarities between the structure of conodont elements and enamel were superficial; they looked similar but there were serious differences. They wanted dentine: “There appears to be a total absence of dentine, which is unexpected if conodonts are vertebrates. Dentine is universal in vertebrates and is thought to be the most primitive of vertebrate hard tissues.” They were not alone; others also doubted Sansom and his colleagues.19
Most vociferous among the vertebrate paleontologists objecting to the conodont vertebrate was Sue Turner, a fossil fish specialist at Queensland Museum in Australia. She and Alain Blieck were two of the organizers of a major international symposium that both commemorated the work of Walter Gross and discussed the problem of “Palaeozoic microvertebrates” at Walliser's university in Göttingen in August 1993. For Paul Smith, Sansom, and Purnell, this was their first opportunity to present their views on the conodont animal to a traditional vertebrate paleontology community, and they soon discovered they had driven headlong into a brick wall. Paul Smith later recalled, “One lunchtime we realized that things were a bit quiet around the conference venue, only to discover that an impromptu meeting had been called to discuss how this community could counter the conodont hypothesis. It was at a reception at the end of that meeting, in Otto Walliser's house, that Sue Turner told us that ‘no evidence we could ever provide’ would convince her that conodonts were vertebrates.” When Nature's Henry Gee heard this, he christened the Birmingham-Leicester team, the “the Pommie Bastard Conodont Group,” an epithet these workers then used ironically to identify themselves. Gee suggested that instead of attempting to overcome the seemingly immovable obstacle of at least some in vertebrate paleontology, they concentrate on convincing the broader and less politically organized biological community. Turner, Blieck, and a number of others at this meeting would now hold their resolve indefinitely. They would be joined by conodont workers who, while not working directing on the biology of the animal, were nevertheless unconvinced by the British animal.
Sansom, with Paul Smith and Moya Smith, continued his assault on Janvier's stronghold. This time they focused on Chirognathus, a conodont genus with delicate elements that Branson and Mehl had described as fibrous, alluding to their frayed, wood-like structure. At one time these fossils were considered fundamentally different from other kinds of conodont, but Lindström and Ziegler had shown this structure simply resulted from a lack of white matter. Branson and Mehl had described the basal filling as bone-like and with Stauffer they marveled at the survival of fossils with such “delicate, sharp-pointed, hand-like crests with the fibrous structure.” Sansom and his co-workers found that the base of these fossils did not contain the globular cartilage he had seen in the previous study but a form of dentine. This diversity of tissue was not unknown in early fish remains, and in 1994 they concluded, “It seems that both conodonts and agnathans experimented with many tissue types and tissue associations in the early history of vertebrate biomineralization.” The discovery of dentine was, they believed, particularly significant. It seemed to offer “additional and conclusive evidence of the vertebrate nature of conodonts.”20
Anne Kemp and Bob Nicoll responded to the earlier paper by searching for collagen in the conodont elements. Collagen occurs in bone, cartilage, and dentine but is not – in living animals – found in enamel. Using a stain that reacts to the presence of collagen, they investigated the organic components within the elements and found the lamellae making up the crown of the conodont element took the stain, but not the white matter or basal tissue. This suggested that the lamellae did not consist of enamel and that the white matter and basal tissues were not bone or dentine. These results flatly denied those of Sansom and his co-workers. Kemp and Nicoll continued to favor amphioxus as the most appropriate comparator organism and saw no evidence of vertebrate affinity.21 The British, however, doubted that material this old could preserve chemically active collagen capable of responding to the stains used by biologists working on living animals.
Purnell now stepped in, supporting Sansom and his colleagues with what he believed was unequivocal evidence that conodont elements did function as teeth; food was “crushed” and “sheared” and had left the marks to prove it.22 From studying completely preserved apparatuses, it was now possible for him to know which surfaces came together in processing food, and in yet another short paper, rapidly published in Nature in April 1995, he reported how he had looked for signs of wear like those found in mammal teeth. Purnell knew that such evidence did not simply support the tooth theory, it opened up entirely new possibilities for understanding the animal as a living entity. Mammal teeth, he wrote, show three types of wear: polishing, scratching, and pitting. He found these same distinctive features on his tiny conodont elements. His platform element showed pitting but no scratching and suggested a crushing mechanism. A blade element showed polishing, which suggested that area was not exposed to abrasive food. Other elements did show scratching, but to variable degrees, suggesting exposure to different kinds of food. And most interestingly, but problematically, these wear patterns were found on young conodont elements too, meaning that the elements were in use while they were still growing. The elements were, it seemed, periodically enclosed for rep
air but spent most of their time exposed and functioning.
With a little orchestration behind the scenes, the very next paper in this issue of Nature revealed Gabbott's Soom Shale animal. Remarkably, the animal had preserved in fine detail evidence of muscle-fiber and its component fibrils. Preservation of muscle around the eye suggested more advanced “encephalization” or “brain” development. This, together with other preserved features, indicated that the conodont animal might have been more advanced than the hagfish or the lamprey. Although anatomically similar to the Scottish animals, Promissum possessed a rather different apparatus, indicating that it belonged to a different group of conodont animals. It had, of course, come from much older rocks, and its preserved muscle tissue was now the oldest in the vertebrate fossil record. It was something like the muscle fiber found in amphioxus, but more like that found in slow agnathan fish. It suggested that the animal could cruise rather than race.23