The Cambrian explosion or Cambrian radiation was the seemingly rapid appearance of most major groups of complex animals around 530 million years ago, as evidenced by the fossil record.^[1][2] This was accompanied by a major diversification of other organisms.^[3] Before about 580 million years ago, most organisms were simple, composed of individual cells occasionally organised into colonies. In the following 70 million to 80 million years, the rate of evolution accelerated by an order of magnitude,^[4] and the diversity of life began to resemble today’s.^[5]

The Cambrian explosion theory has generated extensive scientific debate. The seemingly rapid appearance of fossils in the “Primordial Strata” was noted as early as the mid 19th century,^[6] and Charles Darwin saw it as one of the main objections that could be made against his theory of evolution by natural selection.^[7]

The long-running puzzlement about the appearance of the Cambrian fauna, seemingly abruptly and from nowhere, centers on three key points: whether there really was an “explosion” of complex organisms in the early Cambrian; what might have caused such rapid evolution; and what it implies about the origin and possible evolution of animals. Interpretation is difficult due to a limited supply of evidence, based mainly on an incomplete fossil record and chemical signatures left in Cambrian rocks.

History and significance

Geologists as long ago as Buckland (1784–1856) realised that a dramatic step change in the fossil record occurred around the base of what we now call the Cambrian.^[6] Charles Darwin considered this sudden appearance of many animal groups with few or no antecedents to be the greatest single objection to his theory of evolution: indeed, he devoted a substantial chapter of The Origin of Species to this problem.^[7]

American palæontologist Charles Walcott, who extensively studied North American fossil animals, proposed that an interval of time, the “Lipalian”, was not represented in the fossil record or did not preserve fossils, and that the ancestors of the Cambrian animals evolved during this time.^[8]

The intense modern interest in the subject was sparked by the work of Harry B. Whittington and colleagues, who in the 1970s re-analysed many fossils from the Burgess Shale (see below) and concluded that several were complex but very different from any living animals.^[9] Stephen Jay Gould’s popular 1989 account of this work, Wonderful Life,^[10] brought the matter into the public eye and raised questions about what the explosion represented. While differing significantly in details, both Whittington and Gould proposed that all modern animal phyla had appeared rather suddenly. But other analyses, some more recent and some dating back to the 1970s, argue that complex animals similar to modern types evolved well before the start of the Cambrian.^[11][12][13]

Difficulty of dating the Cambrian

It has been difficult to work out the chronology of the early Cambrian. Absolute radiometric dates for much of the Cambrian, obtained by detailed analysis of radioactive elements contained within rocks, have only rather recently become available, and for only a few regions.^[14]

Relative dating (A was before B) is often good enough for studying processes of evolution, but this has also been difficult, because of the problems involved in matching up rocks of the same age across different continents, particularly around the internationally-defined Precambrian/Cambrian boundary section.^[15] (the most common technique uses widespread but short-lived fossil species to identify rocks of similar ages)

Therefore dates or descriptions of sequences of events should be regarded with caution until better data become available.

Types of evidence

Body fossils

Body fossils preserve significant parts of organisms and are therefore the most informative type of evidence. Unfortunately they are increasingly rare as one looks further back in time, among other reasons because the rocks in which they are buried are usually covered by more recent rocks and because they may have been eroded before being covered by later rocks. One recent study concluded that “parts of the fossil record are clearly incomplete, but they can be regarded as adequate to illustrate the broad patterns of the history of life.”^[16] But there is evidence that some types of animals or parts of animals are relatively likely to be preserved as fossils in some environments and times, and extremely unlikely to be preserved in other environments and times. Part of this is due to changes in the chemistry of the oceans, which were partly caused by the on-going evolution of life, and these changes were most significant before the start of the Cambrian – for example any increase in the marine biomass would reduce the concentration of carbon, and the appearance of sponges reduced the concentration of silicon.^[17]

Another limitation in the discovery and use of body fossils is the lack of preservation of large portions of the body. In most cases the sole anatomical features that are fossilized are the highly mineralised body parts containing high proportions of silica (sponges' skeletons), calcium carbonate (the shells of bivalves, gastropods and ammonites and exoskeletons of most trilobites and some crustaceans) or calcium phosphate (the bones of vertebrates). The majority of animal species living now are unlikely ever to leave fossils, since they are soft-bodied invertebrates such as worms and slugs. Of the more than 30 phyla of living animals, two-thirds of these have never been found as fossils.^[18]

A fossil of Marrella from the Burgess Shale lagerstätte. The animal was under 2 cm long but the fine-grained shale has preserved a very detailed image of it.

The Cambrian fossil record includes an unusually high number of lagerstätten which preserved the fossils' soft tissues in extremely fine detail, allowing a very informative study of animals that normally would not have left fossils. The fine detail of the deposits has allowed paleontologists to examine the internal workings of animals which in other sediments are only represented by shells, spines, claws, etc. The most significant Cambrian lagerstätten are: the early Cambrian Maotianshan shale beds of Chengjiang (Yunnan, China) and Sirius Passet (Greenland)^[19]; the middle Cambrian Burgess Shale (British Columbia, Canada)^[20]; and the Upper Cambrian Orsten (Sweden) fossil beds.

While lagerstätten are superior to most fossil beds in preserving fine anatomical detail, they are far from perfect. The majority of then-living animals are probably not represented because lagerstätten are restricted to a narrow range of environments (e.g. where soft-bodied organisms can be preserved very quickly such as by mudslides), and the exceptional events that cause quick burial make it difficult to study the normal environments of the animals.^[21] In addition, the known lagerstätten cover only a very limited period of time within the Cambrian, and none covers the crucial period just before the start of the Cambrian. Because normal fossil beds are very rare and lagerstätten even rarer, both are very unlikely to show the first occurrence of any type of organism.^[22]

Trace fossils

Trace fossil of the type called Cruziana, possibly made by a trilobite.

Trace fossils consist mainly of tracks and burrows on and under what was then the seabed.

Trace fossils are particularly significant because they represent a data source that is not limited to animals with easily-fossilized hard parts. Also many traces date from significantly earlier than the body fossils of animals that are thought to have been capable of making them.^[23] Whilst exact assignment of trace fossils to their makers is generally impossible, traces may provide the earliest physical evidence of the appearance of moderately complex animals (comparable to earthworms).

Geochemical observations

The ratios of three major isotopes, ⁸⁷Sr / ⁸⁶Sr, ³⁴S / ³²S and ¹³C / ¹²C, undergo dramatic fluctuations around the beginning of the Cambrian.^[24] This chemical signature in the rocks of the Precambrian/Cambrian boundary is difficult to interpret, and may be related to continental break-up, the end of a “global glaciation”, or a catastrophic drop in productivity caused by a mass extinction just before the beginning of the Cambrian.

Carbon has 2 stable isotopes, carbon-12 (¹²C) and carbon-13 (¹³C). Causes often suggested for changes in the ratio of ¹³C to ¹²C found in rocks include:^[25]

• A mass extinction. Chemistry is largely driven by electro-magnetic forces, and lighter isotopes such as ¹²C respond to these more quickly than heavier ones such as ¹³C. So living organisms generally contain a disproportionate amount of ¹²C. A mass extinction would increase the amount of ¹²C available to be included in rocks and therefore reduce the ratio of ¹³C to ¹²C.
• A methane “burp”. In permafrosts and continental shelves methane produced by bacteria gets trapped in “cages” of water molecules, forming a mixture called a clathrate. This methane is very rich in ¹²C because it has been produced by organisms. Clathrates may dissociate (break up) suddenly if the temperature rises or the pressure on them drops. Such dissociations release the ¹²C-rich methane and thus reduce the ratio of ¹³C to ¹²C as this carbon is gradually incorporated into rocks (methane in the atmosphere breaks down into carbon dioxide and water; carbon dioxide reacts with minerals to form carbonate rocks).

Comparative anatomy

Cladistics is a technique for working out the “family tree” of a set of organisms, and has most often applied to evidence from comparative anatomy (features of the bodies of organisms). In this kind of analysis it is possible to include both living and fossilized organisms and work out their evolutionary relationships. Sometimes one can conclude that group A must have evolved before groups B and C, because B and C have more similarities to each other than either has to A. On its own this method can say nothing about when A evolved, but if there are fossils of B or C dating from X million years ago, then A must have evolved more than X million years ago.

Molecular phylogenetics

Molecular phylogenetics attempts to reconstruct the relationships between organisms by comparing details of their biochemistry, such as their DNA. In other words, it applies the analysis techniques of cladistics to biochemical rather than anatomical features. It provides an alternative line of evidence about evolution in the Cambrian and Precambrian, although the need for calibration against the fossil record means it is not entirely independent. Further, since the “clocks” measure molecular evolution, a period of rapid evolution is indistinguishable from a longer period of slow change, so it is unwise to rely on molecular phylogeny for estimates of dates^[26].

Evidence in rocks

This lists the main items in order of the time when the relevant rocks were formed, because timing is the central issue in the Cambrian explosion – but remember that dating rocks from the Cambrian and earlier rocks is very difficult. The survey also starts well before the start of the Cambrian and finishes in the early Ordovician, because some scientists think that the diversification of animal life started before and finished after the Cambrian.^[27]

It covers body fossils, trace fossils and geochemical evidence, because these are all found in rocks which can be dated at least approximately. Arguments based on molecular phylogenetics will appear in a separate section, because this type of evidence is much harder to date with confidence.

Explanation of a few scientific terms

To avoid becoming even longer this article uses some scientific terms, and this is a good place for some simple explanations.^[28]

Phylum is the highest level in the Linnean system for classifying animals. Phyla can be thought of as groupings of animals based on general body plan.^[29] Despite the seemingly different external appearances of organisms, they are classified into phyla based on their internal organizations.^[30] For example despite their obvious differences spiders and crabs both belong to the phylum Arthropoda; but earthworms and tapeworms, although similar in shape, are members of the Annelida and Platyhelminthes respectively.

But the word "phylum" does not describe a fundamental division of nature (not like the difference between electrons and protons). It simply refers to a very high level in the classification system created by Linnaeus in the 18th century to describe all the animals which are alive to-day. This system is not perfect even for modern animals: different books quote different numbers of phyla, mainly because they disagree about the classification of a huge number of worm-like species. Classification systems based on living organisms, including Linneus', do not accommodate extinct organisms well, or even at all.^[18][31]

Triploblastic means consisting of 3 layers, which are formed in the embryo (quite early in the animal's development from a single-celled egg to a larva or juvenile form). The innermost layer forms the digestive tract (gut); the outermost forms skin; and the middle one forms muscles and all the internal organs except the digestive system. Most types of living animal are triploblastic – the best-known exceptions are Porifera (sponges) and Cnidaria (jellyfish, sea anemones, etc.).

Bilaterian means having 2 sides; this implies that they also have top and bottom surfaces and, perhaps more importantly, distinct front and back ends. All known bilaterian animals are triploblastic, and all known triploblastic animals are bilaterian except for echinoderms (but sea cucumbers do have distinct front and back ends; and echinoderm larvae have 2 sides). Porifera (sponges) and Cnidaria (jellyfish, sea anemones, etc.) are radially symmetrical (like wheels).

Coelomate means having a body cavity (coelom) which contains the internal organs. Most of the phyla featured in the debate about the Cambrian explosion are coelomates: arthropods, annelid worms, molluscs, echinoderms and chordates (which includes us vertebrates) - the non-coelomate priapulids are an important exception. All coelomate animals are triploblastic, but some triploblastic animals do not have a coelom (e.g. flatworms; their organs are surrounded by unspecialized tissues). Some bilaterian animals are not coelomates (e.g. flatworms). Echinoderms are coelomates; living species do not look bilaterian (they are radially symmetrical, although sea cucumbers) have distinct front and rear ends), but the earliest echinoderms are still poorly understood and some may have been bilaterally symmetrical.^[32]

Decline of stromatolites over 1 billion years ago

Modern stromatolites in Shark Bay, Western Australia.

Stromatolites are not organisms, they are stubby pillars of sediment built by photosynthesizing microorganisms, especially cyanobacteria. They are now restricted to hostile environments such as extremely salty lagoons, because in less hostile environments they are eliminated by grazing and burrowing invertebrates.

Stromatolites are an important part of the fossil record for about the first 3 billion years of life on earth, peaking about 1250 million years ago, but after then they declined in abundance and diversity, and by the start of the Cambrian had fallen to 20% of their peak. The most widely-supported explanation is that stromatolite-building organisms were the victims of grazing animals, which would imply that sufficiently complex animals were common over 1 billion years ago.^[11][12] This connection is supported by the facts that: stromatolites declined again when the abundance and diversity of marine animals increased in the Ordovician evolutionary radiation; and stromatolite abundance increased after the end-Ordovician and end-Permian extinctions decimated marine animals, but fell back to earlier levels as marine animals recovered.^[33]

Increase in abundance and spininess of acritarchs

Acritarchs include the remains of a wide range of quite different kinds of organisms - ranging from the egg cases of small metazoans to resting cysts of many different kinds of chlorophyta (green algae). They first appear in rocks about 2 billion years old, but about 1 billion years they started to increase in abundance, diversity, size, complexity of shape and especially size and number of spines. Their populations crashed during the Snowball Earth episodes, but they reached their highest diversity in the Paleozoic era. Their increasingly spiny forms in the last 1 billion years probably result from the need for defense against predators, especially predators large enough to swallow them or tear them apart. Other groups of small organisms from the Neoproterozoic era also show signs of anti-predator defenses.^[34]

Further evidence that predation, or at least herbivory, on plankton first appeared around this time comes from a consideration of taxon longevity. The abundance of planktonic organimsms that evolved between 1,700 and 1,400 million years ago were limited by nutrient availability - a situation which limits the origination of new species because the existing organisms are so specialised to their niches, and no other niches are available for occupation. Around about 1,000 million years ago, species longevity fell sharply, suggesting that predation pressure, probably by protist herbivores, became an important factor. Predation would have kept populations in check, meaning that some nutrients were left unused, and new niches were available for new species to occupy.^[35]

Trace fossils 1 billion years ago?

Trace fossils found in rocks about 1 billion years old in India may represent marks of creatures moving across and below soft surfaces. The organisms making the traces were clearly not exploiting deep sediments, but only the layers immediately below the mat of cyanobacteria that covered the seabed. The researchers concluded that the burrows were produced by the peristaltic action of triploblastic metazoans up to 5 mm wide—in other words by animals about the diameter of earthworms, about as complex and possibly coelomates.^[36] But other researchers have dismissed this and other purported finds of trace fossils older than about 600 million years ago, usually on the grounds that they were produced by physical processes rather than by organisms.^[37]

Cryogenian glaciations

The Cryogenian Period between 750 and 600 million years ago was cold, with a few major glaciations:^[38]

• The Sturtian, for which evidence was found in South Australian deposits, occurred about 720 million years ago.
• The Changan (glacial deposits found in China)
• The Tiesiao (glacial deposits found in China) ended before 633 million years ago.
• The Nantuo (glacial deposits found in China) began later than 633 million years ago and is probably equivalent to the Marinoan glaciation in South Australia, which is dated at 630 million years ago.

Doushantuo Formation

The Doushantuo Formation in China contains one of the oldest known lagerstätten. These rocks range from about 635 million to about 551 million years ago, but their animal fossils are mostly less than 580 million years old, predating by perhaps 5 million years the earliest of the 'classical' Ediacaran faunas (see below) from Mistaken Point, Newfoundland.^[39] Doushantuo fossils are all marine, microscopic and highly preserved. They include algae, giant acritarchs and what may be phosphatised embryos of bilaterian animals; but some scientists think the “embryos” are fossils of giant sulfur-metabolising bacteria like Thiomargarita, which is so large that it is visible to the naked eye.^[40]

Vernanimalcula interpreted as an early coelomate. Note that some paleontologists think this “fossil” is a result of purely mineral processes.

One Doushantuo fossil from about 580M years ago, Vernanimalcula (0.1 to 0.2 mm in diameter), has been described as a possible adult triploblastic coelomate bilaterian, in other words about as complex as an earthworm or a mollusc;^[41] others think it was more probably created by non-biological rock-forming processes;^[42] but the team that discovered Vernanimalcula have defended their conclusion that it was an animal, pointing out that they found 10 specimens of the same size and configuration, and stating that non-biological processes would be very unlikely to produce so many specimens that were so alike.^[43]

The Gaskiers glaciation, known from glacial deposits in Newfoundland and Massachusetts, is later than the earliest Doushantuo fossils although it is regarded as the last of the Cryogenian series of glaciations.^[38]

The most recent Doushantuo rocks show a sharp decrease in the ¹³C/¹²C carbon istope ratio. Since this change appears to be worldwide but its timing does not match that of any other known major event such as a mass extinction, it may represent “possible feedback relationships between evolutionary innovation and seawater chemistry” in which metazoans (multi-celled organisms) removed carbon from the water, this increased the concentration of oxygen, and the increased oxygen level made possible the evolution of new metazoans such as Namapoikia (see below).^[39]

Ediacaran organisms

Dickinsonia costata, an Ediacaran organism of unknown affinity, with a quilted appearance.

Fossil of Spriggina, one of the Ediacaran biota and possibly a trilobite

Main article: Ediacaran biota

Strange-looking fossils were found first in the Ediacara Hills in Australia and then in marine sediments from many parts of the world including Charnwood Forest (England) and the Avalon Peninsula (Canada), with dates between 610 million and 543 million years ago (right up to the start of the Cambrian). Most of the Ediacaran biota were at least a few centimeters long, significantly larger than previous finds. The Mackenzie Mountains of northwestern Canada contain 3 distinct assemblages (sets) of Ediacaran fossils: (1) the oldest, dating between 610M and 600M years ago, before the last of the Cryogenian glaciations, are the smallest and least diverse; (2) the middle group, from about 575M to 549M years ago, is found world-wide and includes at least nine genera of disc-like fossils; (3) the last, from 549M to 543 M years ago, includes the full diversity of discs, fronds and apparently segmented forms.^[44]

Many were unlike anything that appeared before or since, resembling discs, mud-filled bags, or quilted mattresses – one palæontologist proposed that the strangest organisms should be classified as a separate kingdom, Vendozoa.^[45] The earliest known body fossils of complex organisms are of one of these strange organisms, Charnia, from about 580 million years ago.^[46]

But some were possibly early forms of the phyla at the heart of the debate about the "Cambrian explosion": Kimberella may have been a mollusc (see below),^[47][13] and is one of the rare Ediacaran fossils whose mode of feeding may be known, enabling easier comparison with Cambrian forms; Arkarua was possibly an echinoderm, although it lacked a feature present in later echinoderms (stereom, a unique crystalline form of calcium carbonate from which their skeletons are built);^[48] Spriggina was possibly a trilobite and therefore an arthropod,^[49] but its body segments seem to be offset across the midline rather than being symmetrically paired as as they are in all known arthropods;^[50] Parvancorina is perhaps a more promising example of an early arthropod.^[51] However, such fossils lack any evidence of legs or a complex digestive system.

Cloudina is a small animal (diameter 0.3 mm to 6.5 mm; length 8 mm to 150 mm) which looks like a rather loose, wobbly stack of cones, sharp end downwards. It has been suggested that Cloudina is a stem group polychaete worm, but there is still much debate about how to classify it.^{[52][53] [54]} More importantly it was one of the earliest animals to have a calcareous shell, i.e. hard parts in the palæontologists’ sense. In some locations, up to 20% of Cloudina fossils contain predatory borings ranging from 15 to 400 µm in diameter. Some tubes had been bored multiple times, indicating that Cloudina could survive attacks (predators do not attack empty shells). The rather similar shelly fossil Sinotubulites, which appears in the same fossil beds, was not affected by borings. In addition, the distribution of borings suggests selection for size. This evidence of selective attacks by predators shows the possibility of speciation in response to predation, which is often suggested as a potential cause of the Cambrian explosion.^[55]

In 2002 another mineralized metazoan, Namapoikia, was found in rocks about 549 million years old, i.e. about 6 million years before the start of the Cambrian. Namapoikia was up to 1m (39in) in diameter and was probably a cnidarian (group which includes jellyfish and sea anemones) or a poriferan (i.e. a sponge).^[56]

It is generally agreed that at least the vast majority and possibly all of the "classic" Ediacaran biota (the organisms that looked most different from any of to-day’s animals) became extinct some time before the start of the Cambrian.^[57][58] One Cambrian discovery may be a fossil of Swartpuntia, a genuine "Vendobiont".^[59] Other finds have been reported as "Vendobionts" that survived into the Cambrian, ^[60][61][62] but it appears that these are not "Vendobionts" after all and some are probably colonies of microbes.^[63][64]

Mollusc-like animals 555 million years ago

Fossil of Kimberella, a triploblastic bilaterian and possibly a mollusc.

A fossil bed in Russia contains a few layers of volcanic ash which have been dated by radiometric methods (uranium-lead ratios in zircons) to a little over 555 million years ago. The fossils found there include Kimberella, the oldest well-documented triploblastic bilaterian. Kimberella was 3 mm to 100 mm long and very like a mollusc: its body was metameric (built as a series of repeated “modules”) but without visible segmentation; it had a single broad, muscular foot and a single shell (not mineralized but fairly stiff). So far Kimberella fossils show no sign of a radula (toothed chitinous “tongue”, which is the signature feature of modern molluscs except bivalves), but radulae are very rarely preserved in any fossil molluscs. However the rocks around the Kimberella fossils bear scratches which are very similar those made by the radulas of grazing molluscs. Researchers concluded that “This is important evidence for the existence of large triploblastic metazoans in the Precambrian and indicates that the origin of the higher groups of protostomes lies well back in the Precambrian.”^[47][13]

Change in carbon isotope ratios at Ediacaran-Cambrian boundary

Carbon has 2 stable isotopes, carbon-12 (¹²C) and carbon-13 (¹³C). At the boundary between the Ediacaran and Cambrian periods the ratio of ¹³C to ¹²C drops sharply, and then is unusually erratic until the mid-Cambrian. There is no easy explanation for the rapid variation of the ratio in the first half of the Cambrian, and at present it is impossible to decide between the two widely-supported explanations for the sharp drop at the Ediacaran-Cambrian boundary, a mass extinction or a methane “burp”.^[25]

Ediacaran and Early Cambrian diversification of trace fossils

The earliest Ediacaran fossils (Assemblage 1 above), 610-600M years ago, contain only cnidarian resting traces. Around 565M years ago (Ediacaran Assemblage 2 above) more complex trace fossils appear, which require a body plan with a hydrostatic skeleton against which muscles pull, i.e. more complex body structures than those of cnidarians or flatworms.^[44]

Around the start of the Cambrian (about 543 million years ago) many new types of traces first appear, including well-known vertical burrows such as Diplocraterion and Skolithos, and traces normally attributed to arthropods, such as Cruziana and Rusophycus. The vertical burrows indicate that worm-like animals acquired new behaviors and possibly new physical capabilities. If traces such as Cruziana and Rusophycus were produced by arthropods, that would indicate that arthropods or their immediate predecessors had developed exoskeletons, although not necessarily as hard as they became later in the Cambrian.^[37]

Small shelly fauna

Fossils known as “small shelly fauna” have been found in many parts on the world, and date from just before the Cambrian to about 10 million years after the start of the Cambrian (the Nemakit-Daldynian and Tommotian ages; see timeline). These are a very mixed collection of fossils: spines, sclerites (armor plates), tubes, archeocyathids (sponge-like animals) and small shells very like those of brachiopods and snail-like molluscs – but all tiny, mostly 1 to 2 mm long.^[65]

Early Cambrian trilobites and echinoderms

Fossilized trilobite, an ancient type of arthropod

The earliest Cambrian trilobite fossils are about 530 million years old, but even then they were quite diverse and world-wide, which suggests that these arthropods had been around for quite some time.^[66]

The earliest generally-accepted echinoderms appeared at about the same time, although it has been suggested that some fossils from the Ediacaran period were echinoderms (see above). The early Cambrian Helicoplacus was a cigar-shaped creature up to 7 cm long that stood upright on one end. Unlike modern echinoderms it was not radially symmetrical with the mouth at the center, but had a spiral food groove on the outside along which food was moved to a mouth that is thought to be located on the side.^[67]

Sirius Passet fauna

Sirius Passet is a lagerstätte in Greenland which was formed about 527 million years ago. Its most common fossils are arthropods, but there is only a handful of trilobite species. There are also very few species with hard (mineralized) parts: trilobites, hyoliths, sponges, brachiopods, and no echinoderms or molluscs.^[68]

One of the arthropods, Pauloterminus, has a bivalve-like carapace.

Halkieria has features associated with more than one phylum, and is discussed below.

Reconstruction of Kerygmachela from Sirius Passet, viewed from the top, with the head to the right. The shaded areas on the lobes (flaps on the sides) are thought to have functioned as gills.

The strangest-looking animals from Sirius Passet are Pambdelurion and Kerygmachela. They are generally regarded as anomalocarids because they have long, soft, segmented bodies with a pair of broad fin-like flaps on most segments and a pair of segmented appendages at the rear. The outer parts of the top surfaces of the flaps have grooved areas which are thought to have acted as gills. Under each flap there is a short, fleshy leg. This arrangement suggests the animals are related to biramous arthropods. Both were apparently blind, as the fossils show no trace of eyes. Kerygmachela had a small conical mouth flanked by robust, unsegmented appendages which had short spines on the front edge and were tipped with longer spines. The spiny front limbs suggest that it may have been a predator, but its small mouth suggests it would have been restricted to very small prey. Pambdelurion lacked trailing appendages but had a more typically anomalocarid-style mouth, a relatively large ring of crushing plates under the front of its head. Its mouth was flanked by a pair of thick, segmented appendages slightly longer than the swimming flaps and equipped with a flexible spine on each segment.^[69]

Chengjiang fauna

There are several Cambrian fossil sites in the Chengjiang county of China’s Yunnan province. The most significant is the Maotianshan shale, a lagerstätte which preserves soft tissues very well. The Chengjiang fauna date to between 525 million and 520 million years ago, about the middle of the early Cambrian epoch, a few million years after Sirius Passet and at least 10 million years earlier than the Burgess Shale.

The Chengjiang sediments provide what are currently the oldest known chordates, the phylum to which all vertebrates belong. The 8 chordate species include Myllokunmingia, possibly a very primitive agnathid (jawless fish) and Haikouichthys, which may be related to lampreys.^[70] Yunnanozoon may be the oldest known hemichordate (a phylum closely related to chordates).^[71]

Vetulicola is a small swimming animal with a carapace covering the front half of its body. Its classification is uncertain: it has paired openings connecting the pharynx to the outside, which may be primitive gill slits; because of these, some researchers argue that it is a deuterostome (“super-phylum” which includes chordates) and possibly even a larvacean (urochordate which remains free-swimming throughout its life); but others classify it as an arthropod.^[72][73][74]

Reconstruction of Anomalocaris saron, viewed from the top with the head to the right. The shaded patches at the bases of the flaps are thought to have acted as gills.

Anomalocaris was a mainly soft-bodied swimming predator which was gigantic for its time (up to 70 cm = 2¼ feet long; some later species were 3 times as long); the soft, segmented body had a pair of broad fin-like flaps along each side, except that the last 3 segments had a pair of “fans” arranged in a “V” shape. Unlike Kerygmachela and Pambdelurion (see above), Anomalocaris apparently had no legs, and the grooved patches which are thought to have acted as gills were at the bases of the flaps, or even overlapping on to its back. The two eyes were on relatively long horizontal stalks; the mouth lay under the head and was a round-cornered square of plates which could not close completely; and in front of the mouth were two jointed appendages which were shaped like a shrimp’s body, curved backwards and with short spines on the inside of the curve. Amplectobelua, also found at Chengjiang, was similar, smaller than Anomalocaris but considerably larger than most other Chengjiang animals. Both are thought to have been powerful predators.

Hallucigenia looks like a long-legged caterpillar with spines on its back, and almost certainly crawled on the seabed.^[68]

Nearly half of the Chengjiang fossil species are arthropods, few of which had the hard, mineral-reinforced exoskeletons found in most later marine arthropods; only about 3% of the organisms known from Chengjiang have hard shells, and most of those are trilobites (although Misszhouia is a soft-bodied trilobite). Many other phyla are found there: Porifera (sponges) and Priapulida (burrowing “worms” which were ambush predators), Brachiopoda (these had bivalve-like shells, but fed by means of a lophophore, a fan-like filter which occupied about of half of the internal space), Chaetognatha (arrow worms), Cnidaria (jellyfish, sea anemones), Ctenophora (comb jellies), Echinodermata (starfish, sea urchins, etc.), Hyolitha (enigmatic animals with small conical shells), Nematomorpha (horse hair worms, parasites which are typically about 1 m long and 1 mm to 3 mm in diameter), Phoronida (horseshoe worms which live in chitinous tubes and feed by means of a lophophore), and Protista (single-celled animals).^[75]

Early Cambrian crustaceans

Crustaceans are one of the three great modern groups of arthropods – the others are chelicerates (spiders, scorpions, horseshoe crabs) and uniramia (the most important uniramians are insects, millipedes, centipedes). Ercaia is a small crustacean from 520 million years ago, found in the Maotianshan shale (a lagerstätte described above).^[76] Small phosphatocopid crustaceans (a group known only in the Cambrian) have been found in the Protolenus Limestone (early Cambrian) of Shropshire, England.^[77]

Burgess Shale

The Burgess Shale was the first of the Cambrian lagerstätten to be discovered (by Walcott in 1909), and the re-analysis of the Burgess Shale by Whittington and others in the 1970s was the basis of Gould’s book Wonderful Life, which was largely responsible for non-scientists' awareness of the Cambrian explosion. The fossils date from the mid Cambrian, about 515 million years ago and 10 million years later than the Chengjiang fauna.

The most common Burgess Shale fossils are arthropods, but many of them are unusual and difficult to classify, for example:

• Marrella is the most common fossil (see picture above), but Whittington’s re-analysis showed that it belonged to none of the known marine arthropod groups (trilobites, crustaceans, chelicerates; well-known modern chelicerates include spiders and scorpions).^[78]
• Yohoia was a tiny animal (7 mm to 23 mm long) with: a head shield; a slim, segmented body covered on top by armor plates; a paddle-like tail; 3 pairs of legs under the head shield; a single flap-like appendage fringed with setae (bristles) under each body segment, probably used for swimming and/or respiration; a pair of relatively large appendages at the front of the head shield, each with a pronounced “elbow” and ending in four long spines which may have functioned as “fingers”. Yohoia is assumed to been a mainly benthic (bottom-dwelling) creature that swam just above the ocean floor and used its appendages to scavenge or capture prey. It may be a member of the arachnomorphs, a group of arthropods that includes the chelicerates and trilobites.^[79]
• Naraoia was a soft-bodied animal (no mineralized parts) which is classified as a trilobite because its appendages (legs, mouth-parts) are very similar.
• Waptia, Canadaspis and Plenocaris had bivalve-like carapaces. It is uncertain whether these animals are related or acquired bivalve-like carapaces by convergent evolution.^[80]

Pikaia resembled the modern lancelet, and was the earliest known chordate until the discovery of the fish-like Myllokunmingia and Haikouichthys among the Chengjiang fauna.

Reconstruction of Opabinia, one of the strangest animals from the Burgess Shale

But the “weird wonders”, creatures that resembled nothing known in the 1970s, attracted the most publicity, for example:

• Whittington’s first presentation about Opabinia made the audience laugh.^[81] The reconstruction showed a soft-bodied animal with: a slim, segmented body; a pair of flap-like appendages on each segment with gills above the flaps, except that the last 3 segments had no gills and the flaps formed a tail; five stalked eyes; a backward-facing mouth under the head; a long, flexible, hose-like proboscis which extended from under the front of the head and ended in a “claw” fringed with spines. Subsequent research has concluded that Opabinia is a lobopod, closely related to the arthropods and possibly even closer to ancestors of the arthropods.^[82]
• Anomalocaris and Hallucigenia were first found in the Burgess Shale, but older specimens have been found in the Chengjiang fauna. They are now regarded as lobopods, and Anomalocaris is very similar to Opabinia in most respects (except the eyes and feeding mechanisms) – see above.
• Odontogriphus is currently regarded as either a mollusc or a lophotrochozoan, i.e. fairly closely related to the ancestors of molluscs (see above).

Molluscs, annelids or brachiopods?

Fossil of Halkieria

Wiwaxia, found so far only in the Burgess Shale, had chitinous armor consisting of long vertical spines and short overlapping horizontal spines. It also had what looked like a radula (chitinous toothed “tongue”), a feature which is otherwise only known in molluscs. Some researchers thin