Survivors: The Animals and Plants that Time has Left Behind. Richard Fortey

it slid through the major event that secured the removal of the dinosaurs from our planet; like Limulus’ ancestors, Peripatus is made of sterner stuff, not to be seen off by mere global catastrophes. But now there comes a surprise. When we go back yet another 200 million years all the way to the Cambrian Period, to the time of ‘explosive’ evolution at the beginning of complex animal life, there, too, were relatives of velvet worms – they prove to be more common as fossils in Cambrian rocks than they are in rocks laid down in later geological periods. They began their history under the sea, in the cradle of life, like everything else. And they proved to be survivors. They shared their early world with trilobites, and the first relatives of horseshoe crabs, and the distant ancestors of scorpions. So much in biology seems to converge back more than 500 million years ago to the Cambrian ancient sea floor. The ancestors of the velvet worms were yet another kind of animal that later moved onto land – and this happened at least 300 million years ago. Because of their rarity as fossils it is not possible to say whether velvet worms got onto land before or after scorpions; we shall probably never know. Unlike scorpions, they needed to stay in wet, or at least humid environments, but just like those venomous arachnids none of their close relatives managed to survive to the present day beneath the sea. For Peripatus and its relatives going on land was arriving at some sort of haven.

      Probably the best-known onychophoran from the Cambrian is called Aysheaia pedunculata. It was named a century ago by the renowned palaeontologist Charles Dolittle Walcott of the Smithsonian Institution, Washington. It occurs in what is probably also the most famous rock formation of that age, the Burgess Shale of British Columbia, Canada. A locality near Mount Field in the Rocky Mountains discovered by Walcott yielded the first known, diverse fossil fauna of ‘soft bodied’ organisms, that is, those lacking hard mineralised shells, which are the kinds that give us ‘regular’ fossils. The Burgess Shale allows us to see something of the whole panoply of marine life at a seminal time – although admittedly it only samples the larger organisms. The fossils are preserved as silvery films on the surface of the black shale, so that they are subtle casts made by fine minerals before the animals could be scavenged or they fell apart. The exact circumstances of their preservation are still being debated, but it is certain that quick burial and protection from normal decay played an important part. Whatever the cause, Aysheaia is preserved in extraordinary detail.

      4. Cambrian lobopod fossil Aysheaia pedunculata from the Burgess Shale in the Canadian Rockies, British Columbia.

      Comparing Aysheaia with Peripatus reveals that they are of similar size and shape, the former reaching about six cm in length. The fact that differently sized animals of Aysheaia retain the same form as they get larger, implies a simple growth pattern like that of the modern velvet worms. In Aysheaia the fine rings encircling the body are clearly visible, and little prickles are much like the papillae of the living animal; add to that their stumpy conical legs look very alike, and at the tips in the fossil little sickle-like claws can be clearly seen. But there are some differences between this most ancient animal and the creature I helped to dig out of its woody habitat – it would have been astonishing if there were not. Most obviously, there is a pair of gill-like structures on the head end of the fossil. This is hardly surprising since the animal was living under water. There is also no sign of the special slime glands in the fossil. This must have been a later development, which presumably would also have been acquired after the terrestrial invasion. But it would take a hardened sceptic not to believe that these animals were related. Of course in science there are always such sceptics, and the special features of Aysheaia were emphasised by some at the expense of its many similarities to Peripatus, but I believe most students today would accept the onychophoran tag on the Cambrian creature.

      The story got interesting when a second, and much more peculiar-looking Burgess Shale species was assigned to the onychophorans. This animal had been named in 1977 Hallucigenia by the Cambridge palaeontologist Simon Conway Morris, but his original description of the fossil was upside down. Hallucigenia carried paired spikes on its back which Conway Morris had originally interpreted as legs (he later acknowledged his error with good grace), while the true legs were more spindly affairs than those of living velvet worms or, indeed, Aysheaia. The spines arose from hardened plates, which had been found separately as fossils in early Cambrian strata, but had been unfathomable up to that time. The mystery was not fully elucidated until much better preserved, soft-bodied fossils began to be found over the last decade or so in strata cropping out around Chengjiang in Yunnan Province in China (these are known as the shales of the Maotianshan Formation). The new fossils were up to ten million years older than the Burgess Shale examples, and have now proved even more diverse. They include at least six animals that can be assigned to the same group as the velvet worms. One of them carries spikes on its back and was an additional species of Hallucigenia, another one (Paucipodia) was an altogether slimmer affair than its distant living relatives, with only nine pairs of slender legs. One fact was now becoming clear: the relatives of the velvet worm were much more varied in the early days. There were lots of them of several distinct kinds, but they did all share those lobe-like legs, often tipped by little claws. An appropriate term for the whole group, both living and fossil, achieved wide currency during the 1990s – they were ‘lobopods’. Thanks to the special preservation of these Cambrian fossils it was possible to see surprisingly varied and delicate lobopod animals in unprecedented detail. Living velvet worms began to seem more of an evolutionary afterthought.

      The plot thickened still further at this time, for up in Greenland Dr Graham Budd and his colleagues were finding yet more soft-bodied animals in the early Cambrian Buen Formation. These showed certain similarities to onychophorans, like the rings along the body, but the animal named Kerygmachela by Budd had a pair of grasping appendages at the front and was obviously a hunter capable of grasping prey. The lobopods were clearly going to spring yet more surprises.

      The question now arises as to where this curious bunch of animals fits in on the tree of life. I have already described how Cambrian fossil faunas included many kinds of jointed-legged animals or arthropods, such as distant relatives of the horseshoe crab. All these arthropods would have had a tough chitin covering over the body that made the ‘invention’ of hinged joints necessary. Without them, the animals would have been as helpless as a medieval jouster whose articulated armour had rusted into immobility. But with hinges added, arthropods were equipped with a versatile covering that could be recruited to be armour, jaws or toolkit as the occasion demanded. The future walked on spindly legs. Like arthropods, velvet worms and their relatives were, and are, segmented animals. Unlike arthropods they did not have a strong coat made of chitin: no hinges were possible. Their lobopod legs were effective enough in their own plodding way, but they could not be extended into the attenuated pins of a daddy-long-legs. That requires serious mechanical engineering, and the stiffening support of a hard skeleton. On the other hand, some features of internal anatomy seem to be very similar between living onychoporans and arthropods. I could mention the diffuse circulation system and the arrangement of the nerve cords, and some scientists are impressed by the presence of antennae in both kinds of organisms. At least one of the Cambrian lobopods shows evidence of simple eyes. The musculature is differently arranged in lobopods and arthropods, which actually allows the lobopods greater bodily flexibility.

      Their fundamental similarities make it likely that Peripatus and arthropods share a common ancestor. The arthropods seem to be more advanced in several respects: the jointed legs could only have been added when the ‘skin’ acquired its hard outer layer, and sophisticated compound eyes like those of Limulus must surely have been a later development. This is another way of saying that lobopods are probably sited on a lower level on the great tree of life, likely to have been around before the arthropods evolved. There are some scientists who would claim that they are the true ancestors of the arthropods, or even that different kinds of lobopod gave rise to different kinds of arthropods. Partly, this depends on the interpretation of the jawed animal Kerygmachela from Greenland that seems to display something of an amalgam of lobopod and arthropod characteristics. Whatever the final interpretation, these recent discoveries of Cambrian fossils provide another case of neat categories of animal classification blurring at the time of the ‘explosive’ phase of animal evolution. The story also takes us back further in time than we have been before.