We know that a very diverse array of plankton existed by 800 Ma, because they are known as fossils. Acritarchs are spherical microfossils with thick and complex organic walls. They are probably dinoflagellates that spent most of their life floating in the plankton. But many amoebalike protists do not have cell walls made of cellulose and so do not preserve well. It's possible that while the surface layers of Proterozoic oceans had huge numbers of floating plankton, Proterozoic seafloors were crawling with successful populations of protists consuming the rich food supplies available in bacterial mats.
Metazoans are most likely a clade, that is, they all descended from one kind of protist. All metazoans originally had one cilium or flagellum per cell, for example. Metazoans also share the same kind of early development. They form into infolded balls of internal cells which are often free to move, and are covered by outer sheets of cells that form an external coating for the animal: a skin, if you like.
The first metazoans were soft-bodied, and we have no fossil record of them. But we can look at the tremendous variety of living animals and at the geologic record to try to reason out what the first metazoans might have looked like and what they might have done.
There are only three basic kinds of metazoans: sponges, cnidarians, and worms. One can imagine scenarios for the divergence of the three great metazoan groups. All of them solved the problem of developing to larger size and complexity, but in different ways.
Sponges evolved by extending the choanoflagellate way of life to large size and sophisticated packaging. They continued to pump water (and the oxygen and bacteria they take from it) through their tissues, in internal filtering modules.
Cnidarians (or coelenterates), including sea anemones, jellyfish, and corals, are built mostly of sheets of cells, and they exploit the large surface area of the sheets in sophisticated ways to make a living. They consist of a sheet of tissue, with cells on each surface and a thickening layer of jellylike substance in the middle. The sheet is shaped into a baglike form to define an outer and an inner surface. A cnidarian thus contains a lot of seawater in a largely enclosed cavity lined by the inner surface of the sheet. The neck of the bag forms a mouth, which can be closed by muscles that act like a drawstring. A network of nerve cells runs through the tissue sheet to coordinate the actions of the animal.
In most cnidarians the outer surface of the sheet acts simply as a protective skin. The inner surface is mainly digestive, and it absorbs food molecules from the water in the enclosed cavity. Because cnidarians are built only of thin sheets of tissue, they weigh very little, and can exist on small amounts of food. They can absorb all the oxygen they need from the water that surrounds them, and they absorb all their food molecules too. Digestive cells lining the cavity then leak powerful enzymes into the water inside the animal. The prey is broken down by these enzymes, and cnidarian absorbs the food molecules through the inner living.
Cnidarians have nematocysts or stinging cells that are set into the outer skin surface. The toxins of some cnidarians are powerful enough to kill fish, and people have died after being stung by swarms of jellyfish. Nematocysts are usually concentrated on the surfaces and the ends of tentacles, which form a ring around the mouth. They provide an effective defense for the cnidarian, but they are also powerful weapons for catching and killing prey, which the tentacles then push into the mouth for digestion in the cavity.
Hardly any sponges can tackle food particles larger than a bacterium, though there are a few exceptions. Living cnidarians routinely trap, kill, and digest creatures that outweigh them many times.
The third and most complex metazoan group is the Bilateria. It contains all other metazoans, including vertebrates. Here I shall simply call them worms. Worms consist basically of a double sheet of tissue that is folded around with the inner surfaces largely joined together to form a three-dimensional animal. In contrast to sponges and cnidarians, worms have evolved complex organ systems made from specialised cells.
All sponges and most cnidarians are attached to the seafloor and depend on trapping food from the water. But many worms, including the most simple group, flatworms, are mobile scavengers and predators. Worms creep along the seafloor on their ventral (lower) surface, which may be different from the dorsal (upper) surface. They prefer to move in one direction, and a head at the (front) end contains major nerve centers associated with checking and testing the environment.
Probably the mobility of worms on the seafloor led to the differentiation of the body into anterior and posterior (head and tail) and into dorsal and ventral surfaces, as the various parts of the animal encountered different stimuli and had to be able to react to them. A well-developed nervous system coordinates muscles so that a worm can react quickly and efficiently to external stimuli.
The same locomotion that gave a worm a front-to-back axis also gave it bilateral symmetry. Any other shape would have produced an animal that could not move forward efficiently.
The head usually features the food intake, a mouth through which food is passed into and along a specialised one-way internal digestive tract instead of being digested in a simple seawater cavity. No sponge cell or cnidarian cell is very far away from a food-absorbing (digestive) cell, so these creatures have no specialised internal transport system. But the digestive system of worms needs an oxygen supply, and the nutrients absorbed there have to be transported to the rest of the body. Worms therefore have a circulation system, and the larger and more three-dimensional they are, the better the circulation system must be.
The genetic programming that builds an animal works like efficient computer programming. For example, one could instruct a computer to draw a flower, specifying the size, shape, and position of each petal. Given that petals typically have much the same size and shape, however, one could use one shape and size for every petal, and simply tell the computer to move the pen to the right place before drawing the same petal each time.
Structural genes build each piece of the animal, and regulatory genes make sure the piece is built in the right place at the right time. Thus a set of regulatory genes could be used in combination with a set of "segment" genes to build all the segments along a growing worm. The same sort of regulatory genes could easily be used to build a legs on, say, a millipede or a crab, by calling on a "leg" gene the appropriate number of times instead of a "segment" gene. By calling on slight modifications of the "leg" gene as growth developed, regulatory genes could build an animal whose legs were different along its length (as in insects), or build a vertebrate with different bones along the length of a backbone.
Developmental geneticists have now identified regulatory genes that control which way up an animal is formed, which is front and back, and how the animal varies along its length or around its edges. The most thrilling discovery is that much the same control box is used throughout the metazoans. Sets of genes that sit close to one another in the nuclear DNA perform much the same job in developing animals, but because they call in a variety of structural genes in a variety of patterns in time and body areas, the results in terms of anatomy are vastly different.
Hox genes are such clusters. Sponges have one set of Hox genes (and are simple in structure), whereas mammals have 38 sets in four clusters, and goldfish have 48 in 7 clusters. Hox genes control the growth of nerve nets, segments, and limbs throughout metazoans, and their evolution accompanies the divergence in anatomy and physiology and ecology and behavior that gave us all the variety of living animals. Hox genes provide separate, but complementary evidence to accompany the fossil record; however, it is important to remember that we can only study the genes of surviving groups of animals, not those from the 95% of species that have become extinct.
Protists don't need Hox genes, because they don't divide cells in precise patterns to form a multicellular adult. Hox genes evolved in early metazoans, and provided the genetic tool kit to build viable complex animals. Presumably, Hox genes control the lay-out of a sponge that gives efficiency of water currents passing through the body. In the simplest worms, Hox genes lay out the nerve nets that allow the worm to sense the environment all along the body. One can easily imagine that the earliest metazoans, wherever, whenever, and however they evolved, would quickly radiate into a great variety of body shapes and structures, with natural selection acting equally quickly to weed out the shapes that were poor adaptations, and leaving a scrapbook of successful prototypes that proliferated.
Thousands of these fossils have now been collected worldwide in dozens of different localities. Almost all the fossils occur between 565 and 543 Ma, with the highest abundance and diversity during the last few million years from 550 Ma to 543 Ma. After that the Ediacaran animals seem to have become extinct. Most of them probably left no descendants; others gave rise to some of the Cambrian animals that followed.
There are a few Ediacaran sponges, but most Ediacaran fossils are cnidarians of some sort. Jellyfish and other cnidarians floated just like their living relatives. Colonies of sea pens were attached to the seafloor. Sea pens look like plants, but are cnidarians that capture and eat floating animals in the water. Dickinsonia is a very large flattened animal, up to 45 cm long, and there is some debate whether it is a very unusual worm or a very unusual cnidarian. Other Vendian fossils are worms that patrolled the seafloor. Some squirmed through the surface sediment; others walked on the tufts of bristles located on their body segments. Shallow fossil burrows in the sediment show that some worms were a centimeter across and were deposit feeders, leaving fecal pellets behind them. Other smaller worms left trails on the surface as they wriggled across the sediment. Since Vendian animals were soft-bodied and unprotected, there may have been no large carnivores on the seafloor.
Beginning rather suddenly, the fossil record contains skeletons: shells and other pieces of mineral that were formed biochemically by animals. Humans have one kind of skeleton, an internal skeleton or endoskeleton, where the mineralization is internal and the soft tissues lie outside. Most animals have the reverse arrangement, with a mineralized exoskeleton on the outside and soft tissues inside, as in most molluscs and in arthropods. The shell or test of an echinoderm is technically internal but usually lies so close to the surface that it is external for all practical purposes. The hard parts laid down by corals are external, but underneath the body, so that the soft parts lie on top of the hard parts and seem comparatively unprotected by them. Sponge skeletons are simply networks of tiny spicules that form a largely internal framework.
There is incredible variety in the type, function, arrangement, chemistry, and formation of animal skeletons; biomineralization is a whole science in itself. With the evolution of hard parts, the fossil record became much richer, because the mineral components of animals resist the destructive agents that affect soft parts of bodies.
Almost as soon as geologists realized that fossils marked time periods in earth history, they also recognized that the quality of the fossil record depended on the style, structure, and composition of the hard parts of the organisms that were preserved . For about a century, in fact, many geologists believed that there was no fossil record before hard parts evolved. The evolution of hard parts defines the beginning of an era in Earth history, the Paleozoic Era, and the beginning of its oldest subdivision, the Cambrian Period. In contrast, Precambrian time was first seen as a time of no life, and then as a time of soft-bodied, mainly bacterial life. Even today, the base of the Cambrian is defined at a time when major new fossils appear in the record.
Why did hard parts evolve in the first place, and why did they evolve when they did? What difference do hard parts make to the biology of an animal?
Worms are soft-bodied. Sponges are sponges, whether they have tiny mineral spicules forming an internal skeleton, or a soft protein like that in bath sponges. But many metazoan groups have skeletons that are such an integral part of their body plans that they only exist as such when they have hard parts. Although there are molluscs without shells (slugs and squids, for example), it seems impossible to be a clam without a shell. Shells are so important to clams that if a clam evolved to be shell-less, its basic biology would be so changed that we would call it something else. An arthropod without a skeleton is basically a worm (unless it's a caterpillar). Thus any worm that evolved hard parts by definition evolved into some other major group.
Therefore, the evolution of hard parts implies the appearance of new kinds of animals on Earth. When one animal group is radically different from any other and is also considered to be a clade, evolved from some single ancestral species, it is a phylum, defined by its own particular body structure, ecology, and evolutionary history. Mollusca and Arthropoda are familiar phyla that must once have had a common ancestor, but that ancestor wasn't a mollusc or an arthropod. There are arguments about the number of phyla living in the world today, mainly because of the bewildering variety of wormlike creatures, but most people would count about 30 phyla of living animals.
Worms contribute to the fossil record, especially by leaving trace fossils of burrows and trails, and there are various small, puzzling groups of early shelled animals that became extinct without leaving descendants. But for paleontological purposes, only nine or ten phyla are or have ever been important in terms of hard parts. It is stunning to realize that all but Bryozoa are known from Cambrian rocks, and all but Bryozoa and Chordata are known from Early Cambrian rocks. Yet only two (Cnidaria and Porifera) are found in Ediacaran rocks.
At face value, these facts suggest that a spectacularly rapid burst of innovation at the beginning of Cambrian time produced most of the major body plans of animals. Each phylum is radically different from any of the others, so each must have followed a different evolutionary pathway. Even the hard parts they evolved are very different from one another. Sponges evolved an internal skeleton of fine silica needles. Molluscs and most brachiopods evolved an external shell made of calcium carbonate, but the two phyla used different minerals, and different crystal structure. Some brachiopods used calcium phosphate for their shells. Arthropods evolved chitin, but different groups of arthropods impregnated the chitin with calcium phosphate or with calcium carbonate. Echinoderms have an internal skeleton just under the skin, made up of small separate plates of calcium carbonate, each one a single calcite crystal.
Looking at the variety of minerals involved, it's clear that there wasn't some simple chemical change in the oceans (an increase in phosphate, for example) that triggered the invention of skeletons. Yet the evolutionary event was global, so it was probably triggered by some global biological or ecological factor. The event has come to be called the Cambrian Explosion. We need to know why these different creatures evolved different types of skeletons, and why they did it so rapidly. The radiation of the phyla of metazoans was apparently so rapid that we still have no idea of the order of branching. A symposium in 1998 heard several, radically different proposals for metazoan cladograms, and the organizers later reported that ten more years' research would help to begin to solve the question.
However, perhaps the Cambrian Explosion was not so dramatic. The ancestors of Cambrian animals may have lived in Ediacaran times but have not yet been discovered. The earliest metazoans may have been microscopic and soft-bodied, for example, shaped more like the larvae of their descendants than the adults. Perhaps they were tiny, soft-bodied creatures in the plankton of Slushball Earht. If so, metazoans could have had a long history that would not be fossilized. They may well have been very different from one another, as a result of 150 m.y. or so of evolution within soft parts, after the Hox gene system began to generate diversity. Then suddenly, around 550 Ma, many of them suddenly gained hard parts and large size, and "exploded" into the fossil record. In doing so, they evolved the larger features that allow us to identify them anatomically and ecologically as the metazoan phyla that still survive.
In this scenario, what evolved in the Cambrian explosion was ecological molluscs and ecological arthropods. Their ancestors may have had important differences in the DNA of important genes, but they had all been ecologically similar. This would make the Cambrian explosion an ecological event rather than a phylogenetic one. However, it is also possible that all the divergent evolution that formed the multiple metazoan lineages occurred just before and just into the Cambrian, making the Cambrian explosion genetic as well as ecological, and dramatic indeed.
The truth is probably somewhere in between (as it often is). Increasingly, molecular evidence (only available from the living survivors of the Cambrian explosion, remember!) suggests considerable previous divergence among soft-bodied metazoans. Some estimates place a major metazoan radiation well before the Ediacaran.
For some purposes, it doesn't matter whether the Cambrian explosion genuinely represents the evolution of new phyla, or whether it represents the evolution of new characters within pre-existing phyla. What is clear is that the evolution of morphology that reflects new adaptation and novel ecology occurred very rapidly indeed at the beginning of the Cambrian, and that needs explanation. We'll review the evidence, and then look for explanation of the Cambrian Explosion.
Echinoderms familiar to us today are sea urchins and starfish, but peculiar plated echinoderms are found in very early Cambrian rocks. The helicoplacoids looked rather like twisted deflated footballs covered with small plates. They were attached to the seafloor by one pointed end, and three food grooves led to a mouth halfway up on one side. Most echinoderms today are symmetrical, many of them with almost perfect fivefold patterns of arms and plates, but the first known echinoderms were distinctly unlike that, showing that fivefold symmetry is a derived character that evolved later in echinoderm history.
Brachiopods are relatively abundant Cambrian fossils, creatures that had two shells protecting a small body and a large water-filled cavity where food was filtered from seawater pumped in and out of the shell. Brachiopods lived on the sediment surface or burrowed just under it.
These animals are large, and they are easily assigned to living phyla. For the first time, the seafloor would have looked reasonably familiar to a marine ecologist. Trilobites probably ate mud, and echinoderms and brachiopods gathered food from seawater. Yet some ecological puzzles remain. There are no obvious large predators among these earliest skeletonized Cambrian fossils, no obvious grazers unless trilobites ate algae, and no swimmers, only floating plankton.
These soft-bodied animals are found in Early and Middle Cambrian rock formations in China, Canada, Greenland, and Poland, and in Utah. I shall call them all the Burgess fauna, after the Burgess Shale in Canada where they were first discovered.
More than half the Burgess animals burrowed in or lived freely on the seafloor, and most of these were deposit feeders. Arthropods and worms dominate the Burgess fauna. Only about 30% of the species were fixed to the seafloor or lived stationary lives on it, and these were probably filter-feeders, mainly sponges and worms. Thus, the dominance of most Cambrian fossil collections by bottom-dwelling, deposit-feeding arthropods is not a bias of the preservation of hard parts: it occurs among soft-bodied communities too. Trilobites are fair representatives of Cambrian animals and Cambrian ecology.
The main delights of the Burgess fauna are the unusual animals which have provided fun and headaches for paleontologists. Aysheaia looks like a caterpillar with thick soft legs. It is called a lobopod because of the strange shape of its limbs.
There are undoubted predators in the Burgess fauna. Priapulid worms today live in shallow burrows and capture soft-bodied prey by plunging a hooked proboscis into them as they crawl by. There were at least seven species of priapulids in the Burgess Shale, including Ottoia. Opabinia is a highly evolved predator. It is long and slim, with a vertical tail fin, so it probably swam about. It has five eyes and one large grasping claw on the front of its head.
Halkieriids are flattened creatures averaging about 5 cm long. They look like flattened worms, with perhaps 2000 spines forming a protective coating embedded into the dorsal surface.
Altogether, the Burgess faunas give us a good idea of the sorts of exciting but extinct soft-bodied creatures that may always have lived alongside the trilobites but were hardly ever preserved.
Early echinoderms had lightly plated skeletons just under their surfaces, and the most reasonable explanation of their first function is support, accompanied or followed by the function of defense.
For other animals, skeletons provided a box that gave organs a controlled environment in which to work. Filters were less exposed to currents, so perhaps they would not clog so easily from silt and mud. A boxlike skeleton would also have given an advantage against predation. Molluscs and brachiopods may have evolved skeletons for these reasons.
Arthropods, and especially trilobites, are strongly armored all over their dorsal surfaces, not just in the head region. Most likely their armor served for the attachment of strong muscles. Muscles pull and cannot push. Worms move by using internal hydraulic systems, as we have seen. On the other hand, walking demands that limbs push on the sediment, and that is very unrewarding if the other end of the leg is unbraced. Arthropods evolved a large, strong dorsal skeleton against which their jointed legs were firmly braced, allowing them to move much more efficiently than worms do. Later, the strong skeleton would have been effective against predators. It's unlikely that trilobites evolved skeletons for defense. Cambrian trilobites had small tails that did not protect them very well when they rolled up. (Later trilobites had larger tails and could roll up to protect themselves almost completely.)
Skeletons seem to have evolved for many different reasons, in many different chemistries, in many different animals, but why did they evolve in a very short geological time and in two abrupt waves? The only common factor is the dramatic invasion of new ways of life on the Early Cambrian seafloor, into ecologies that were impossible without support or sheltering of internal organs or muscular bracing.
Despite all the discussion of skeletons, the Burgess fauna shows that dramatic evolution took place also in animals that did not have strong skeletons. However, many of these animals had outer coverings that were tough, but lightly mineralized: the Burgess arthropods are particularly good examples.
The common factor along successful groups of Cambrian animals is larger body size. All of this suggests that in some way the world had become ready for large animals, and in turn that tells us that the Cambrian event was driven by worldwide ecological factors, but we do not yet know what they were. They could have been related to a change in food supply in the sea, which in turn depends on upwelling, which in turn depends on climatic and geographic patterns on a global scale. We don't yet know enough about Cambrian geography and climate to say anything sensible about these factors, but it's here that the answer probably lies and where future research should be focused.
Whatever the underlying global cause was, some specific mechanisms have been suggested to explain the Cambrian explosion.
Unfortunately, Paine's original observations may have been made in an unusual community. Along other rocky shores (for example, in South Africa) the top predators are seabirds. They keep the shores so clean of herbivores such as limpets, and so well fertilized with guano, that seaweeds take over and choke out the huge diversity of animals that would otherwise live on the rocks. Thus, the action of predators may lower or raise diversity, depending on the ecological food chain of the particular community. There may be no general principle linking predation to high diversity.
Nevertheless, Steven Stanley used Paine's work to suggest that the evolution of predation triggered the Cambrian radiation. Stanley made an intellectual jump to suggest that predators can cause additional diversity in their prey. He argued that if predators first appeared in the Early Cambrian, they may have caused the increase in diversity at that time. Perhaps predators also encouraged the evolution of many different types of skeletonized animals.
Geerat Vermeij provided critical support for Stanley's idea, suggesting how new predators might indeed cause diversification among prey (at any time). In response to new predators, prey creatures might evolve large size, or hard coverings made from any available biochemical substance, or powerful toxins, or changes in life style or behavior, or any combination of these, all in order to become more predator-proof. And as the new predators in turn evolve more sophisticated ways of attacking prey, the responses and counter-responses might well add up to a significant burst of evolutionary change.
Are the characteristics of early Cambrian fossils consistent with a predation theory? The rules of the predator/prey game may have changed radically as large multicellular creatures evolved. Many Early Cambrian fossils have hard parts that look defensive. For example, some sharp little conical shells may have been spines that were carried pointing outwards on the dorsal and lateral sides of animals, to fend off predators.
More direct evidence comes from little tubes called Cloudina that were the first hard parts ever evolved by any animal. Chinese specimens from Ediacaran rocks occasionally have holes bored into them, presumably by some unknown predator. There are armored and spined Early Cambrian animals, and some Early Cambrian trilobites have healed injuries that may indicate damage by a predator. Defensive structures made of hard parts could therefore have contributed to the increase in the number of fossils in Early Cambrian rocks.
Present evidence suggests that predation played an important part in generating the Cambrian event. It's difficult to be certain, because the only major predators we have discovered are the big soft-bodied predators from the Burgess fauna, and we have no evidence of what they ate: anything they could catch, probably! However, predation does not explain the timing of the Cambrian explosion: why not 100 m.y. earlier, or later? Certainly predation alone cannot account for all the variety of skeletons that we see.
If the oxygen idea is true, however, it could explain much of the Cambrian explosion. Where did the increased oxygen come from? Oxygen is produced in photosynthesis, but usually the plant tissue that is produced at the same time eventually is eaten and digested (using up oxygen), or it rots (using up oxygen). Only if the organic matter is buried does the oxygen stay free in the sea and the atmosphere. What would increase the amount of carbon buried in the seafloor?
Graham Logan and his colleagues pointed out that the evolution of metazoans ("worms," let us say) also involved the evolution of guts, and therefore of feces, usually in the form of compact fecal pellets. If carbon-rich fecal pellets are buried quickly, organic matter is removed from oxidation effectively, and this raises oxygen levels in the sea and in the atmosphere. In this view, then, the rise of metazoans large enough to produce reasonable quantities of fecal pellets was responsible for the rise in oxygen, which then permitted even more (and larger) metazoans) to evolve, and so on, until it became advantageous for those metazoans to evolve skeletons. This is a very attractive idea, especially as "worms" would characteristically produce their fecal pellets in or on the seafloor, where they would be buried easily and quickly.
Even so, it is not clear that oxygen, or predation, or any other single parameter can be identified as the reason for the nature, the scale, and especially the timing of the Cambrian explosion. One could argue (and people have) that the world is full of complex creatures, so complexity must have evolved sometime. Whenever it evolved, it was bound to cause a visible "burst" in the fossil record, but perhaps there was no "trigger" for the Cambrian explosion we see in the record. The first large animals evolved at that time, and it is hardly surprising that they spread rapidly and diversified into many body plans, with different groups evolving hard parts of different chemistry and structure. After the dramatic events early in the Cambrian, the increase in numbers and diversity of fossils later in the period seems anticlimactic. Cambrian fossil collections are not very complex ecologically; they are dominated by trilobites, most of which lived on the seafloor and were deposit feeders. Filtering organisms are very much secondary, and although there are large carnivores, they are represented only by anomalocarids.
The Cambrian explosion is spectacular, but it is not unique; in my view the spectacular diversification of the diapsid reptiles, especially the archosaurs, in the Late Triassic is an analogous case, as is the diversification of the mammals after the end of the dinosaurs. Such radiations stand out from "normal" evolutionary events just as "mass extinctions" stand out from the rest. On a real planet inhabited by real organisms, evolutionary rates are likely to vary in time and space, and evolutionary events are likely to vary in magnitude, duration, and frequency. We should not expect that ideal rules we might propose for an ideal planet would be followed by the natural world; instead, we have to find out from that natural world what the rules actually were.