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

I once disturbed a huge, black knobbly scorpion that came running at me with its tail-sting erect, while I backed off in the opposite direction, gibbering foolishly. My Omani companions laughed and told me that its sting (in the tail of course) was relatively mild. A few hours later I lifted up a rock slab and nestled beneath it was a small yellowish scorpion with a flattened, side-wound sting. I was about to poke at it with my geological hammer when my companions tugged at my arm. ‘Don’t touch, it’s a real killer!’ The smaller, insignificant-looking creatures can often be the most deadly. The spike on Limulus’ tail and the sting on the scorpion’s are closely related structures, and indeed both animals belong to the same great arachnid group. The scorpion learned the trick of arching over the very end of its body to dart poison into enemy or prey. Encased within an external skeleton (cuticle) that prevents evaporation with outstanding efficiency, the scorpion has been able to live far away from water; some species specialise in surviving in the driest places on earth. Scorpions started out as sub-aqueous creatures like Limulus, and only later did they acquire the skill of living on land. Back in Devonian times, 400 million years ago, their relatives, the sea scorpions (eurypterids) were the largest invertebrate predators ever to have lived, some as long as a man. There are fossils from the Carboniferous age that really do look like living scorpions, at least to the non-specialist. They, too, have faced out extinction events that have blasted greater and more glamorous animals. The scorpion is built into mythology as a sign of the zodiac; it features on Roman mosaics, and in the Bible (1 Kings 12:11): ‘my father hath chastised you with whips, but I will chastise you with scorpions’. So one could argue that the scorpion’s connection to human culture is more pervasive than that of the horseshoe crab. My choice of organisms has been guided by the place they occupy in the tree of life, rather than by their innate charisma or significance in folklore and culture. The horseshoe crabs earn a special place in this natural history because their relatives root down to the beginning of the diversification of animals. The early Permian Palaeolimulus is clearly a horseshoe crab, for all that it predated the great extinction that put paid to most species at the end of the Palaeozoic Era, 250 million years ago. Limulitella is present in 242-million-year-old (Triassic) strata dating after the great trauma, evidence of their survival. Those distinctive tracks left by the tips of the legs, and the trail of the tail, that I observed in Delaware have been found as fossils even in the absence of the body itself. There were horseshoe crabs crawling among the coal swamps of the Carboniferous (Pennsylvanian), a little better segmented perhaps than the beasts on Delaware Bay, but carrying the distinct signature of their ancestry. In 2009 my Canadian colleague, David Rudkin, announced the discovery of the oldest typical horseshoe crab in rocks of Ordovician (approximately 450 million years) age, thus taking these simple arthropods back before all the major extinction events that have rocked the Phanerozoic biosphere. Whatever the magic ingredient for survival is, the horseshoes clearly have it in spades. ‘The meek shall inherit the earth’ may be an appropriate motto for their longevity.

      3. Gustav Vigeland’s sculpture of a survivor, the scorpion, Frogner Park, Oslo, Norway.

      Let me describe these early days in more detail. I have already remarked that the horseshoe crabs had set out upon their distinctive path before the first land plants had advanced upon harsh and barren shores; although recent discoveries suggest that a few simple plants may have already ventured onto mud flats. These had probably not yet been followed by insects or spiders. There were fishes already, of a primitive cast, but they had no ambitions then to invade the land, although a few species may have nudged into waters that were not fully marine. The seas abounded with trilobites, which occupied every ecological zone from shallow shores to ocean deeps. These prolific arthropods must have been many times more abundant than the early relatives of the horseshoe crabs. They evidently had an advantage at the time. This may have been the evolution of their robust dorsal ‘shell’ of calcite, which allowed them to develop spines, armour and a tough anchor for muscles, as well as an ability to roll up into tight, impregnable balls when threatened. Trilobites soon learned an array of different feeding habits; some were predators, some ate soft mud, others swam in the open seas. They died out some 255 million years ago. By contrast, the relatives of Limulus may have stayed conservatively on the sea floor as scavengers and predators. The horseshoe crabs on Delaware Bay have an exoskeleton of chitin, which is a natural polymer that is quite tough and flexible, although no substitute for stony calcite. But the horseshoe crab has turned what might have been a weakness to advantage by developing an exceptional immune system. Survival in the long term may depend on more subtle features than armour alone.

      It is possible to trace the horseshoe crab story still further back, into the Cambrian Period more than 500 million years ago, to a time when many of the major types of animals converge towards their common ancestors. The Cambrian was an interval of unprecedented evolutionary activity and I shall describe its special features in more detail in the next chapter. Early relatives of Limulus have been identified in Cambrian strata, but they include some species that look a little different from their hardy survivors. Some of them also look more like early trilobites, such as Olenellus, a form with a big head-shield surrounded by a narrow rim; one might expect a family resemblance if they are indeed closer to a common origin. There are important differences, too. Where the limbs of trilobites are known, they are similar all along the length of the animal: the paired limbs are each split into a walking leg carrying a comb-like branch near its base that in all likelihood functioned as a gill. They are not subdivided into different ‘packages’ in different parts of the body separating walking and feeding appendages in front from gills behind, as they are in Limulus. To add to this, all trilobites had typical ‘feeler’ antennae near the front of the head, and none had the strange chelicerae. This may not be so important, since having antennae seems to have been a general property of primitive arthropods. It might just be that the trilobites still retain this one characteristic more primitive than Limulus and its allies, but they could yet have descended from a common ancestor. Trilobites are abundant fossils on account of their easily preserved calcite hard parts. Fossils of unmineralised animals are altogether more unusual. Spectacular recent discoveries of fossils of soft-bodied animals preserved within Cambrian rocks have been made in China and Greenland. These have revealed an almost embarrassing variety of undoubted arthropods early in the Cambrian. Some of them might seem to bridge the differences between Limulus and its allies and the trilobites, but for every feature that points one way, there seems to be another that suggests something else. For more than a decade now palaeontologists have argued about how these fossils should be classified, and about the only thing they all agree upon is that the Cambrian threw up many animals with curious combinations of characteristics that were probably winnowed out by subsequent evolution. It is not, perhaps, so surprising that ‘mixed up’ animals lived at this Cambrian time, because all the arthropods were not genetically far apart then – they would have had the subsequent 500 million years to box themselves into more separate evolutionary compartments. In these early days the destiny of one animal to become a crustacean, say, and another a chelicerate was not easy to anticipate.

      When scientists are confronted by conundrums of this kind, they usually turn to computers. There are now sophisticated computer programs that deal with the problems of determining relationships between animals. They work by identifying the particular arrangement of the creatures analysed on a branching tree that most succinctly accounts for the features they share with their fellows. The most significant resemblances in morphology should result in organisms being classified together on a single branch. Like so many computer methods, the inner workings of the process are staggering in their complexity, so that for a big problem like analysing the Cambrian arthropods millions of potential arrangements of trees embracing the animals under study will be inspected and rejected. My own appreciation of what goes on inside these machines is thoroughly naïve, and I cannot suppress a vision of thousands of cards being shuffled into piles like a supercharged game of Patience until the answer ‘comes out’. The end product is a diagrammatic tree (technically, a cladogram) that can look enticingly simple. I should add that the way the summary ‘tree of life’ on the endpapers is drawn is not like a cladogram, but it does incorporate the results of many individual cladistic analyses. Like all computer methods, the latter are subject to the familiar caveat of RIRO (Rubbish In Rubbish Out), but the fact that they have been so widely used indicates