Out of Thin Air: Dinosaurs, Birds, and Earth's Ancient Atmosphere (2006)

The origin of animals is the start of this story. Its timing has been hugely controversial, with two distinct lines of evidence giving quite different views on the timing of the first diversification of animal phyla, that seminal period when some first animal ancestor evolved into many kinds of animals. One of these lines comes from the pattern of appearance of animal fossils in rocks, the second from molecular clock studies on extant animals. This chapter will look at these records for clues to one of the greatest of all paleontological mysteries: what caused the rapid diversification of animal body plans in the Cambrian Explosion?

When did the first animals diversify into the numerous animal phyla? Since the time of Darwin it has been known that fossils of animal life seem to appear suddenly in the fossil record. Over short stratal intervals, sedimentary rocks seemingly bereft of fossils were found to be overlain by rocks with a profusion of highly visible fossils, the most common being trilobites. Trilobites are fossil arthropods: the familiar animals such as insects and crustaceans are unified into a single phylum through the shared presence of hard, jointed exoskeletons, and trilobite fossils are the remains of highly evolved and complex animals. This observation was highly vexing to Darwin (and hugely comforting to his critics), as it seemed to fly in the face of the then newly proposed

theory of evolution. Today, we know that animal life did appear comparatively rapidly in the fossil record, and new dating techniques now put the time of the first complex fossils at slightly older than 540 million years ago, with the first trilobites appearing in the record some single-digit millions of years after that.

In contrast to this slightly more than half-billion-year-old age for the first animals, molecular studies of extant animals looking at the age of divergence of basic animal lineages suggested a far older origin of animals. One influential study by Charles Wray and his associates at the American Museum of Natural History dated the divergence of animals from protozoan (single-celled organisms, as compared to animals, which are all composed of many cells) ancestors at about a billion years ago. In this latter view, animal phyla diverged early but remained at a very small size and invisible to the fossil record for a half billion years. Both views agreed that the appearance of animals in the fossil record was a significant event, which has been called the Cambrian Explosion. To paleontologists, the Cambrian Explosion marked the first evolution of animals. To molecular geneticists, it marked the first evolution of animals large enough to leave remains in the rock record. The controversy raged through the 1990s, to be resolved in the early years of the twenty-first century when new molecular studies, using more sophisticated analyses, essentially confirmed the younger date for the origin of animals that had been championed by paleontologists. There is now agreement that animal life on Earth did not predate 600 million years ago and might be closer to 550 million years in age.

The Cambrian Period is now dated from 544 million to about 495 million years ago. Over those roughly 50 million years, the vast majority of animal phyla first appeared. All specialists agree that this is the most important event in the entire history of animal life, superceded in importance only by the first appearance of life on Earth, perhaps, in the context of the entire history of life on our planet.

The Cambrian Explosion has thus been a “hot” topic in science. A vast library of books, technical articles, and popular accounts of the Cambrian Explosion exists. The paper devoted to these many pages has probably destroyed whole forests, which is a bit ironic, since the Cambrian period occurred long before vascular plants colonized the

land, let alone evolved into trees and forests. With so much written by so many, what more can be said? With new and better estimates of the oxygen levels back then, the topic is ripe for reexamination.

The levels of carbon dioxide and oxygen postulated for the Cambrian Explosion interval are shown in the figure beginning this chapter, taken from the Berner curve introduced in the preceding chapter. According to this curve, oxygen levels soon after the start of the Cambrian Explosion 544 million years ago were about 13 percent (compared to 21 percent today) but then fluctuated. During this time carbon dioxide levels were far higher than they are today, tens of times higher in fact, and such high levels would have produced a greenhouse effect. Even with the drop in carbon dioxide levels at the end of the Cambrian around 495 million years ago, temperatures at this time would have been higher than in the present-day. Since less oxygen is dissolved in seawater with higher temperature, the already anoxic conditions of the oceans, due to the low atmospheric oxygen of that time, would have been exacerbated.

To explain the conditions, flora, and fauna of the Cambrian, imagine that we have journeyed back in time to Earth circa 522 million years ago. To really do this right we need not only a time machine but also a spaceship, in order to better see the position of the continents, for continental position and geological processes accruing from plate tectonic processes had a determining effect on subsequent biotic history.

The first thing we notice as we pass over the land surface is that there is so little vegetation. Low traces of green can be seen in the wetter areas but most of the land surface is bare rock. It looks like the areas around glaciers at high altitudes in our present-day world, but even as we cross the equatorial regions the starkness of the place is apparent. Plant life is limited to moss and vast slicks of plant-like, photosynthesizing bacteria. There is no rich organic soil. There are no trees, no

bushes, no flowers at all. There are no vast grasslands, nor are there any animals.

Even the nature of the river systems seems bizarre. While huge rivers do cascade out of the mountains, as they make their way toward the coastline of the Cambrian Ocean, they remain low and anastomostic, a series of braided streams passing over the gravel of the land surface. Nowhere do we see the giant meandering rivers of our time. There are no lazy river bends, no oxbow lakes, no point bars that future fishermen will stand on casting for trout. There are not even trout, for that matter, or any other visible creature at all in these rivers. No bugs, no minnows, no amphibians, no wading birds, no dragonflies, or anything else so familiar to those who revere and visit the grand rivers of the Planet Earth we know today. The great mountains are being eroded, as they will be through time, but the sedimentary refuse resulting from erosion is passed to the sea in rivers that have been the same now for the 4 billion years since Earth came into existence. That will soon change; riverbanks will, some tens of millions of years hence, become colonized by rooted plants, and rivers will begin to meander over vast floodplains and river valleys. That will come, but not on this day.

Our craft passes over the last of the coast plain and we see the sea. Here at last something looks familiar: waves and currents that look no different from those seen in our world. But once again, when we descend for a closer look, the alienness of this world becomes apparent. We park our craft on the seashore, step out, and gasp in air that is not the thin air of the high mountains but air that gives our lungs no traction. It is as if we have stepped out onto Mount Everest in terms of the amount of oxygen but the seashore in terms of air pressure. This is not thin air, it is low-oxygen air. The air pressure is the same as we experience at sea level in the present, but the amount of oxygen making up the air is lower. With our oxygen masks in place, we walk the sandy shore. The sand is composed entirely of minerals. There are no flakes of seashells, bits of coral, or skeletons of planktonic creatures like foraminifers or coccoliths. It is silica-rich sand bereft of organically precipitated content.

The beach is huge. We walk from the waves’ edge to the high-tide

land, marked still by foam and the wet mark of the recent high tide. The tidal change here is at least 15 meters of vertical distance between the high-tide line and the ocean’s level, a tidal change that can be found only in the Bay of Fundy, Canada, in our own world. We look up at a noticeably dimmer sun, dimmer because all stars grow more energetic through time, and to the east we see a half moon just risen, dim as well in the afternoon sunlight but far greater in size than it will be in our time. The moon is clearly closer to Earth in this long ago time, and that would explain the great tidal range.

We walk to an outcrop of hard basalt that makes one of the headlands of this broad sandy beach, and even in areas of the outcrop where the low tide still washes up, there are few of the familiar features of our world. There is no upper tide assemblage of Littorina (the common seashore periwinkle snails), or barnacles beneath the snails, or lines of mussels beneath the barnacles—no intertidal zonation at all, in fact. But it is not dead, this rock. There is a variety of algae clinging to it, reds and browns that wash in the rushing current, not all that different from the kelps of our world, except for the fact that no animals live amid them.

It is time to look at what lives in the sea. This being a thought experiment, we suit up with appropriate diving gear, clean our masks, don flippers, and dive downward. We are not the first to take this trip. In his 1998 book, The Crucible of Creation, Simon Conway Morris took readers on a submersible trip into the Cambrian Ocean at the site of what would become the Burgess Shale and in so doing described the vast assemblage of organisms that would become the fossil finds of that most important of all fossil deposits, our best view into the world at the height of the Cambrian Explosion. But the Burgess Shale was deposited some 510 million years ago. Here we are in the world of 522 million years ago, almost 12 million years before the Burgess. Twelve million years of evolution is a long time for animals to diversify. We are at the beginning of things, and the place we dive into will come to be known as Chengjiang, China, and, like the Burgess Shale (if not so famously), it will yield fossils with soft parts, giving us a window into the start of the Cambrian Explosion.

We descend through the surf zone and make our way offshore into

deeper water, and, on a muddy bottom still somewhat influenced by wave action at lowest tide, we scan the bottom for life. And life there is—in abundance. The first noticeable organisms are numerous small sponges, most attached to cobbles on the bottom (but with a few rolling slightly with the slight current on the bottom, among the small ripples found in the bottom sediment). A quick census shows at least 30 different kinds, probably each a separate species. Most are the familiar demosponges of our time, but a few hexactinellid, or glass sponges, are found in this undersea forest of sponges. Other phyla are seen as well. There are a few anemone-like creatures, and floating with us in the water column are jellyfish and some ctenophores.

We descend onto the bottom itself and sift through the dark sediment. Very quickly we find a diversity of small worm-like forms. Most are priapulids and one, which will eventually be named Maotionshania, is particularly abundant. We search further and find buried clam-like creatures that on closer inspection are found to be inarticulate brachiopods looking much like the still extant Lingula. These have long and mobile tethers, pedicles that help them dig back into the sediment when we release them back onto the sea bottom. Other inarticulate brachiopods are found encrusted on rocks. Among them are small numbers of another shelled invertebrate, the tube-shelled hyoliths, with their strange arms called “Helens” in honor of the daughter of their discoverer, Charles Wolcott.

Small worms, sponges, bivalve brachiopods, and the tiny hyoliths, these are the minor elements of the fauna. The rest of the fauna here belongs to one phylum, and it is present in a staggering diversity. It is the arthropods that dominate this sea bottom, totally eclipsing the other invertebrates in all of number, diversity, and size. This is truly a world of segmented “bugs.”

They are everywhere. The most common of all are small, ostracod-like bradoriids and tiny bivalved arthropods about 1 millimeter long. There are lots of other bigger arthropods. Some, such as Naraoia, that will survive into Burgess time show little or no exoskeleton. As we approach, some hurriedly roll up like pillbugs for protection. Other arthropods that will be found in the Burgess Shale are here as well, including the bivalved Canadapsis. Large forms with stalked

eyes, to be called Vetulicola and Chuandianella, watch us carefully, while another form with a short head shield, bulbous eyes, and large spines on its tail region scuttles away, crab-like. Amid these strange arthropods are the more familiar trilobites, but, unlike in later Cambrian deposits, they are fewer in number and diversity than their more esoteric cousins. Most are the familiar redlichiacean trilobites with numerous segments, forms that Darwin and his contemporaries considered the oldest animals on Earth and forms that are so striking because of their many segments. The trilobites here are heavily armored compared to the many other thinner carapaced arthropods among them, and these other poorly skeletonized forms will not commonly fossilize. At all localities save this one and a few others, which because of special circumstances will preserve soft parts, only the trilobites will be found, giving the false impression that they were the most common members of the fauna. Here we see that they are only a minor part of the fauna.

While most of this arthropodan assemblage is fairly small in size, we see bigger animals too. There are several arthropod forms that look like sea scorpions and some that have flattened oval-shaped bodies such as the enigmatic Saperion. All of these are somewhat intermediate in size, and now we search for the top carnivore of this ecosystem—it too is an arthropod—and we do not have long to wait. Swimming lazily through the water, some meters above the teeming bottom, we see a meter-long Anomalocaris, famous from the Burgess and here as well, showing its antiquity. It settles downward onto the bottom, its large paddlelike tail slowing as it lands with its many walking legs taking up the shock. With large claws slashing, it begins to feed on the many smaller, fleeing arthropods. The Anomalocaris takes note of another not as big but still substantial invertebrate on the bottom, a heavily armored lobopod, also an arthropod but one that is very rare today. This strange animal looks like a cross between an annelid worm and an arthropod and seems to be related to the still-living onycophorans of our world. The phosphate plates on the lobopods provide some protection but soon it too is killed, and the Anomalocaris centers itself over the body and begins feeding with its peculiar, circularly plated mouth.

It is time to go. As we head back toward the seashore we see one more animal. It is not an arthropod, but it does have a segmented side.

It is a primitive fish-like creature, one of several that live here. Some look like eels, some like hagfish. They are the first chordates, or vertebrates. They are our ancestors.

The distribution of fossils from the fantastic deposits in Chengjiang, China, has given us a new window into the origin of the animal phyla on Earth. To continue the analogy, this is a window to a floor lower than that of the Burgess Shale. The approximately 12 million years separating the age of these two deposits thus gives us a new view of how animals diversified. Because both Chengjiang and the Burgess preserve soft parts and skeletonized animals, we have a good picture of what was there and in what relative abundance. Without this added view yielded by the preservation of soft parts, we would never be sure about the relative abundance of various kinds of animals, for perhaps there was a huge abundance of creatures like soft worms and jellyfish, forms that had no skeletons. Thus we are surprised at what appears to be a clear view of the nature of the fauna at both sites. Over 50,000 fossils have been collected from the Burgess Shale (and a lesser number from Chengjiang). In their summary of the Burgess fauna, Derek Briggs, Doug Erwin, and Fred Collier, in their 1994 book The Fossils of the Burgess Shale, list a total of 150 species of animals. Almost half are arthropods or arthropod-like. But an even more interesting number relates to the number of individual fossils. Well over 90 percent of all Burgess fossils are arthropods, followed in number by many fewer sponges and brachiopods. Like the earlier Chengjiang, the Burgess sea bottom was dominated in the kinds and numbers of animals by arthropods. Arthropods are among the most complex of all invertebrates and yet in these almost earliest of fossil deposits in the time of animals they are diversified and common.

Our visit back to the Cambrian leads to an inescapable conclusion: in sheer numbers of individuals and species (described as diversity) and

in sheer numbers of different kinds of body plans (described as disparity), the arthropods were the most successful of Cambrian animals. How much of this success was due to their principal body plan characteristic—segmentation?

Segmented animals are the most diverse of all animals on the planet today, and most are arthropods. All arthropods, including the highly diverse insects, show repeated body units and body regions based on groupings of individual segments that have specific functions for the animal. The feature uniting the group is the presence of a jointed exoskeleton that encloses the entire body. This exoskeleton even extends into the gut. The exoskeleton cannot grow, so it must be periodically molted and another, slightly larger, one produced. The body has a well-differentiated head, trunk, and posterior region in varying proportions. Appendages are commonly specialized. On terrestrial arthropods the appendages are usually single and enormous, but the marine forms generally have two branches or parts per appendage, an inner leg branch and an outer gill branch, and are thus termed “biramous.”

Arthropods are not alone in being segmented. All annelids are segmented, and some members of generally nonsegmented groups, such as the monoplacophoran mollusk, show segmentation. It appeared early in the history of animals and indeed the Cambrian trilobite fossils, the most common of the earliest preserved animal, show segmentation.

Reconstruction of Marella, a Cambrian arthropod from the Burgess Shale. As can be seen here, each segment behind the head bears a pair of large gills. The area of gills to overall body volume is thus very large in this kind of body plan.

The exoskeleton encloses the soft parts like a suit of armor and that may be its major function: protection. But the consequences of this kind of skeleton are huge: there can be no passive diffusion of oxygen across any part of the body. To obtain oxygen the first arthropods, all marine, had to evolve specialized respiratory structures or gills. The combination of segmentation and an exoskeleton used by the arthropods was clearly a design winner: the arthropods, all with this design, have more species today on Earth than any other phylum. Some (perhaps most) of that success must be due to their characteristic segmented body. In his 2004 book, The Origin of Phyla, James Valentine reflected on what is a major evolutionary puzzle: why were there so many kinds of arthropods in the Cambrian and such large populations of individuals belonging to the many kinds of this group then present? It is worthwhile to look at what he has written on this subject:

Hence we have an interesting puzzle. What we call arthropods are composed of what appear to be many separately evolving groups that have, through convergent evolution, produced body plans of great diversity save for one aspect: all have limbs on each segment that are biramous—each appendage carries a leg of some sort and a second appendage, a long gill.

Why would early primitive (or basal) animal groups opt for segmentation? Perhaps this is the wrong word, for Valentine and others have noted that the arthropods are not so much segmented—which at least in annelids consists of largely separated chambers for each seg-

ment of the body—as repeated. Valentine proposes that this striking body plan arose in response to locomotory needs, stating:

There is no doubt that this type of body plan is an adaptation that aids locomotion. But here we can depart from Valentine and suggest that the main function of this kind of body plan is to allow repeated gill segments, each small enough to be held in optimal orientation beneath the segments. The flow of water across these gills, while at first glance appearing to be passive, may actually represent a pumping gill.

Look at trilobite morphology. While the upper surfaces of the trilobite carapace is commonly preserved, the underside, bearing appendages and gills, is rarely preserved.

Trilobites have long been known to have a curious food acquisition system. The same appendage terminations mark the fusion of the walking legs and gills and end in a blunt paddle structure known as a gnathobase. Paleontologists have long surmised that a forward-moving current created by limb movement would move food material to the mouth underneath the animal. But such a current would also serve to bring new water across the gills, and with the carapace of the animal sitting overhead like a roof or tent, the trilobite or other arthropod could build a directed water current defined by the body above and the sediment or sea bottom below. I propose that oxygen acquisition purposes primarily led to this body plan and that it was only secondarily co-opted for food acquisition. The environmental conditions of oxygen levels far lower than those of today would have provided the stimulus to begin this evolutionary pathway leading to the arthropod body plan.

So why is segmentation needed for a large gill surface? The same amount of gill surface area could be evolved and exist as one or a small number of large sheets attached to the body. But such large structures would easily fold on themselves and could also become problems during swimming by increasing drag.

How did this kind of repeated body part system come about at all? I propose that an important group of regulatory genes known as Hox genes were themselves co-opted in order to provide repeated gills.

A rich new topic of research spanning the last two decades has been at the interface of development and evolution. There are rules to the formation and growth of embryos as the genetic code of the genome becomes expressed as the protein and protoplasm of the growing individual. Newly discovered “regulatory” gene complexes, such as one called the Hox gene complex, are increasingly viewed as being nearly as important as the genetic code itself. As for segmentation, biologist S. A. Newman has postulated that it is a consequence of developmental regulation and pattern, not a feature evolved specifically for function. In other words, segments initially may have come about not because they work better in the day-to-day life of an organism with them but because they made growth and the many morphological changes occurring in these animals from embryo to adult take place more smoothly. Once in place, however, the presence of segments was molded by evolution for use in specific functions such as locomotion, respiration, feeding, and reproduction through evolutionary processes placing specific locomotory, respiratory, and reproductive appendages on various segments.

Do animals with segments really have an advantage in respiration over those that are nonsegmented? Let’s look at a measure of comparing respiratory structures, starting with passive systems. In a passive system, as mentioned in Chapter 1, the oxygen-carrying medium, in this case seawater, comes in contact with the respiratory epithelium. Oxygen diffuses across the membrane into the circulatory fluid and is moved to the various parts of the body and its cargo of cells, all needing oxygen. Carbon dioxide is exchanged across the same membrane out of the body. Three factors control the rate at which oxygen can be extracted: its concentration in the water, the rate at which it can be

diffused over the membrane (largely a function of the thickness of the membrane), and the volume of new water that comes in contact with the gill surface (largely a function of the size of the gill and the rate of water movement over the gill surface).

The simplest way to increase the volume of new water in contact with the gill surface is to increase the size of the gill, and one way to compare oxygen-extraction efficiencies is to compare relative gill sizes for a given volume of the animal whose gill it is. Thus, a simpleminded way of comparing invertebrates in terms of oxygen-extraction efficiency is to measure gill surface and animal volume and then compare them, assuming constant thickness of the respiratory surfaces. In other words, we are assuming that all gills are equally efficient for unit area, an assumption that is probably not completely correct but close enough to allow an adequate comparison.

Trilobites turn out to have a very high gill to body volume ratio, based on as yet unpublished observations made in the course of writing this book, compared to most modern-day crustaceans, especially all crabs. We really cannot determine whether trilobites and other segmented arthropods have a respiratory advantage over other, nonsegmented animals, although clearly their segmented body plan has been extremely successful. Let’s also look at the other major Cambrian animals and their respiratory successes.

Many of the animals present in the Cambrian had no specialized respiratory structures, and they generally shared a common characteristic: they were covered with epithelium that allowed for the passage of water directly from the sea into the body of the animal. For them the entire body was a gill. Examples include many soft worm-like forms, such as nemerteans and the many varieties of the jellyfish-sea anemone clan, the cnidarians. The cnidarians were and are composed of but two cell layers, and they all show a gut cavity that also allows respiration. Adding to this basic structure of the cnidarians are many tentacles and in larger forms, such as the large sea anemones, there is an interior gut that shows complex invaginations for digestive purposes. Thus, these

animals have a very high surface area to volume ratio. Cnidarians prove themselves to be characterized by the highest surface area to volume ratios of any animal yet examined. While there are few fossils of cnidarians from the Burgess Shale or Chengjiang deposits, those that we have (such as the numerous examples of pennatulaceans, or Sea Pens) demonstrate the high surface to volume ratio of this phylum.

Sponges are another animal found in abundance in the early Cambrian deposits. Like the cnidarians, sponges show no respiratory structures, nor would we expect any. With a body plan built around a series of sacs (like the cnidarians but with even less organization: there are no true tissues in a sponge), all sponges show a very high surface area to volume ratio. In fact, sponges are like agglomerations of numerous, single-celled organisms, with each cell essentially in contact with seawater. But even with this advantage, sponges show an even more efficient way of gaining oxygen. Their main feeding cell, called a choanocyte, causes large volumes of water to pass through the sponge. Some sponge specialists have suggested that a sponge passes as much as 10,000 times its volume in seawater through its body each day. This puts the sponge in the category of having functional “pump gills,” since the feeding cell forces water past the cells at a phenomenal rate. Consequently, sponges are capable of living in extremely low-oxygen conditions.

So if most soft-part-covered animals, except for the very large ones, do not seem to require specialized respiratory structures, what of those with hard parts? As noted earlier in this chapter, the major groups of animals with hard parts in the Cambrian were the huge tribe of arthropods, followed in numerical importance (in most Cambrian marine strata) by brachiopods and then by a smaller number yet of mollusks and echinoderms. We have already discussed the respiratory system of arthropods, particularly trilobites, and will discuss these less numerous groups next.

First, Cambrian echinoderms make up a weird assemblage of small boxlike animals. All echinoderms have an outer layer of living flesh. The oxygen needs of this layer are taken care of by direct diffusion of oxygen from seawater and by the layer’s formation with gill-like elements emanating from the flesh in some modern forms, such as sea urchins, starfish, and sea cucumbers. All of these echinoderms also have

thin-walled tube feet, used for locomotion but also serving for respiration, as oxygen can diffuse across the tube feet and then pass inward through small holes in the skeleton (the ambulacra) to the interior of the box. There were many curious forms related to the more familiar kinds in our world. All of these echinoderms surely were covered with a thin epidermis as well, so that only their interiors needed enhanced respiratory techniques. All had hard skeletons of outer plates that surely inhibited oxygen uptake, thus requiring some form of adaptation for respiration for the interior. This need for interior respiration is somewhat alleviated by the fact that most echinoderms have very little flesh in their boxes. A sea urchin, for instance, while having a voluminous space in its spherical skeleton, actually has little flesh within; our measurements of Puget Sound urchins show that only about 20 percent of the volume within their box is composed of living protoplasm. The rest is seawater or the unique water vascular system that enables the tube feet.

More common than echinoderms were mollusks. Most during the Cambrian were small in size. Each of the major classes of mollusks (gastropods, bivalves, and cephalopods) is found in Cambrian strata. The most common mollusks were monoplacophorans, a minor class today but common in the Cambrian. They had a limpet-like shell and a snail-like body with a broad, creeping foot. Most interestingly, alone among mollusks of the time, they showed a body organization that suggests segmentation. From looking at muscle scars on the fossil shells and from comparing anatomy of the still living forms, it looks as if the Cambrian monoplacophorans had multiple gills. Modern-day gastropods have a single pair of gills or sometimes a single gill. But the Cambrian monoplacophorans, which lived a very snail-like existence in all likelihood, found it necessary to have multiple gills. The respiratory adaptations of the mollusks are distinctive enough to warrant additional discussion, below.

Second in abundance to the arthropods were brachiopods, a phylum related to bryozoans that are routinely mistaken for bivalved mollusks. Yet while the shells of bivalves and brachiopods show a superficial similarity, the internal anatomies of the two groups are radically different. The major feature of a brachiopod is a feeding organ known as a

Micromitra, a Burgess Shale brachiopod. Although looking like a clam, its internal anatomy is very different from any mollusk. There is a large organ called a lophophore that, in shower-curtain fashion, separates the interior of the shell into two compartments. As water is forced over this net-like structure, both food and oxygen are extracted. Thus, all brachiopods are examples of creatures using the pump-gill respiratory system. The spines on the outside of this species are not part of the respiratory system.

lophophore, composed of a large loop with numerous long, thin fingers producing a delicate fan-like shape in the shell. This organ filters seawater for food, and, as it is filled with a body fluid and is very thin, it also serves as an exquisite respiratory organ. Like the body of an echinoderm, the interior of a brachiopod shell is almost all water. There is very little flesh, which is in contact with a steady flow of seawater.

The brachiopod lophophore creates several currents of seawater that pass into the sides of the shell, move across the lophophore, and are then sent out the front of the shell. This constant stream of new water entering a brachiopod has the same effect as the current passing through a sponge. The small volume of flesh to great surface area of the lophophore, coupled with the steady flow of water (many times the volume of the interior of the shell), makes the brachiopod consummately adapted for a world of low-oxygen. While the lophophore is explained as primarily an adaptation for feeding, its respiratory function may, in fact, have been a primary aspect of its evolution and shape in the first place. By building a shell and then creating the water currents passing over the lophophore, the brachiopods appear to have been among the first examples of an animal using a “pump gill.” Sponges also pump but they do not have an internal organ (nor indeed

do they need one) to deal with oxygen extraction. But the brachiopod shell design allows a clever combination of feeding and respiration in the same organ. This design appears to be a classic example of a body plan developed either partially or mainly to deal with life in a low-oxygen environment.

Following arthropods, mollusks are the most diverse groups of animals on Earth today, and surely it was the same in past times. In terms of body plan they are very different from arthropods. Only a few groups of mollusks show any sort of body part repetitiveness; mollusks have a single unsegmented body, and none are jointed. The dorsal body wall is cloaked by a large flap of tissue known as the mantle, and this tissue secretes the calcium carbonate shells that are found in the majority of mollusks, while the ventral surface is primitively shaped as a broad, flat “foot” that allows locomotion. The mouth is anterior and the anus posterior, and most have an open circulatory system. There are exceptions even to this, however, the most important being the cephalopods, a group of mollusks that we will look at in detail below.

Mollusks are known from the early Cambrian, and they had probably already differentiated into many separate groups by the earliest Cambrian time. The group appears to have evolved from some type of flatworm. Early in the history of mollusks, however, the evolution of the dorsal shell created the same problem that was faced by arthropods: with a shell covering, large areas formerly used for passive oxygen uptake were no longer available. The standard explanation is that, like the arthropods, the evolutionary response to the new barrier to diffusion created by the shell was to create a gill system. Here we can propose a novel scenario: that the shell evolved to increase the efficiency of the respiratory process rather than solely as a protective organ. Like so many pathways in evolution, the mollusk shell evolution pathway clearly involved multiple uses of the evolved structure, and protection was a consequence of shell formation, protection not only from predators but also perhaps more importantly in the latest Precambrian pro-

tection from harmful ultraviolet radiation or desiccation in intertidal settings. This idea can be formalized here:

Because no representatives of the earliest mollusks still exist, there has been much speculation about what the “prototypical” mollusk may have looked like. This creature has even received a name: HAM, for Hypothetical Ancestral Mollusk. HAM is depicted in every invertebrate zoology and paleontology text. The creature is cephalized with a distinct head; there is the broad foot, internal viscera, and capping dorsal shell. Respiration is through the paired gills, located in a cavity at the back of the animal.

While the HAM model is but a theoretical construct, its overall morphology is at odds with the earliest mollusks found in the Cambrian, most belonging to lineages now extinct. Most of the earliest forms were minute, and in general their preservation is very poor, often simply as phosphate casts or molds in rock. Such preservation makes detailed anatomical reconstruction very difficult. Patient work by paleontologists Bruce Runnegar and John Pojeta has succeeded in classifying many of these earliest fossils, and the surprise is that the earliest do not resemble HAM in many ways. The largest group belongs to the monoplacophoran class, which may be ancestral to all subsequent mollusks. Long thought to have gone extinct at the end of the Permian, the discovery of living monoplacophorans in deep sea settings in the 1950s led to a much greater understanding of the life of the early mollusks. The living forms confirmed what muscle scars found on the interior of the earliest monoplacophorans fossils asserted—there was more than a single pair of gills. In fact, multiple pairs of muscles lined the entire length of the interior of the shell, leading to

the conclusion that these early forms showed an evident segmentation or at least a repeat of the gill-blood vessel system. Since it is only the gills and supporting blood and filtering systems that show this repeated pattern, it can be surmised that, as in arthropods, this repeated pattern was an adaptation for the increased respiratory surface area of the gills. Today, a somewhat similar pattern of repetition, extending even to the shell, is found in the chitons, today commonly found on intertidal beaches. The point here is that the earliest mollusks seemed to have been designed with multiple sets of gills, necessary in a low-oxygen world of the Cambrian.

While the monoplacophorans flourished in earliest Cambrian strata and have survived to the present-day, other groups of mollusks soon evolved from the monoplacophorans. In them a reduction is seen in the number of paired gills, with ancestral gastropods and bivalves reducing the number of gills to a single pair and with primitive cephalopods maintaining perhaps two pairs, based on the anatomy of the modern-day Nautilus. So why was there a reduction of gills? As the reduction of the number of paired gills in mollusks occurred during the Cambrian, long before the models of Berner and others suggest that oxygen levels began rising, how can this observation be reconciled with the low-oxygen levels thought to have been present in marine environments during the earliest times of the Cambrian? Here it is proposed that the cephalopod shells evolved by the ancestral stocks provided a geometry that, like the monoplacophorans, allowed the formation of currents of water to be passed over the gills at a pressure that greatly increased the volume of water the gill surfaces would encounter in any given unit of time. In essence, the formation of these cephalopod shells, coupled with a ciliary system creating a unidirectional water flow across the gill surface, essentially increased the functional size of the gills. With this shell-ciliary system, a single pair of small gills became functionally equivalent to a much larger gill area of an organism with a “passive” gill system. Shells became part of the respiration organ in creating enclosed spaces or functional tubes for pressurized water to flow over the gills. This system also produced an effective method of passing used water out of the system, and the water flow also functioned to take out fecal and waste material at the same time.

Once in place, this new molluscan “shell-pump” form of respiration elaborated in several directions. The old passive method of multiple gills survived in the monoplacophorans and chitons, but neither group was ever very successful. But the pump gill forms showed a vastly different history, and here it is proposed that this huge success in terms of both species-level diversity and variety of form is directly related to the evolution and exploitation of this new kind of respiratory system. Three distinct elaborations on the pump gill respiratory system evolved in the Cambrian, and then each proceeded to undergo spectacular evolutionary flowerings producing numerous species. One group of mollusks rotated the body so that the gills faced forward into a water flow directed through the front of the shell’s aperture and over the gills and exiting at the back of the aperture. These became the gastropods. A second group combined respiration and feeding by enlarging the gills even more with a pair of enclosing shells, using the bivalved shell morphology method of partitioning water flow across the gills in a fashion analogous to the brachiopods. This group became the bivalves. The third group left the gills in the back of the body but vastly increased the strength of the water flow over the gills by changing from a ciliary source of incurrent water propulsion to one caused by muscular pumping of the entire body, with each pump drawing water in across the gills and then using a return stroke forcibly ejecting the water back out of the shell. These became the cephalopods. In all three, shell form became a compromise between strength against predation and optimal shape for the channeling of interval water currents across the gill surface.

Let’s look at the evolution of the cephalopods, which today include the familiar octopus and squid, in more detail. We know the when of their origin and even the why—the world was ready for the existence of highly mobile carnivores to prey on the abundant and slow-moving Cambrian arthropods—a stock of animals succeeded in evolving a body plan that could move fast despite the scarcity of oxygen.

While it has long been known that the cephalopod shell is a buoyancy organ, here is a new hypothesis: the nautiloid shell (and the wonderful shells of their future descendents, the ammonoids, creatures

discussed in many future chapters) was first evolved as a respiratory organ in response to the low-oxygen content of the Cambrian oceans—and still functions like this today in what is the last remaining externally shelled genus.

All of this wonderful majesty of design came about because a late Cambrian mollusk discovered a new way of beating the problem of low oxygen: it evolved a new kind of molluscan respiratory pump, so that muscular action, instead of cilia, created a high-pressure volume of water that passed over the gills. After being mined for its oxygen and laden with carbon dioxide from respiration, all that water had to be expelled from the animal with as much force as it was brought in, if the same water was not to be recirculated. What better way than to jet it out through a tube! What a surprise it must have been when each jet of water jerked the entire animal backward. When predators came, an even more vigorous jet of water would jerk the earliest cephalopod (or was it still a monoplacophoran?) out of harm’s way.

Natural selection then honed the system. But the shell was a heavy burden to bear, and thus a solution for neutral buoyancy was selected. When all was combined, in the late Cambrian, a new kind of animal was set loose. This was a hugely important episode in the history of life, as the cephalopods remained the dominant carnivores in the sea from that point until 65 million years ago when the ammonites, descendants of the cephalopods, were killed off by the Chicxulub asteroid strike, a time interval of about 450 million years. And as we now know, extinction of the ammonites was not the end of the cephalopods, for in large areas of the ocean, such as the pelagic or midwater regions, cephalopods remain the dominant carnivores, outcompeting fish in these dim regions of the sea. As we will see, cephalopods will be obvious and common players in the local marine ecosystems.

The cephalopods are the pinnacle of invertebrate evolution in size and intelligence and rival the arthropods in complexity. Yet we are faced with a dilemma: the cephalopods have so radically departed from the characteristic body plan of head, creeping foot, and simple covering shell in both soft parts and hard parts that major explanations must be offered of how—and why—such radical changes occurred. When the major transition from a bottom-living ancestor to the swimming cephalopod, which first appeared in the late Cambrian (at about the same time as the Burgess Shale fauna, in fact), is discussed, the explanation is that the evolution of the buoyancy organ shell came first, followed by soft-part adaptations that would have perfected swimming—the jet propulsion system still used by squids today.

So here is the current concept. Long study of the living Nautilus has demonstrated that the iconic chambered shell does indeed function as a buoyancy organ, as first argued by Robert Hooke in the 1600s. The chambered portions are filled mainly with gas but with small volumes of water either present as free liquid at the bottom of each chamber or entrained in felt-like, water-containing membranes lining the interior of each septum and the siphuncle, the long calcareous (but porous) tube that pierces each chamber and contains a strand of living tissue connected to the back of the Nautilus’s soft parts. To make a new chamber, this posterior part of the body moves forward in the shell a short distance and the newly created space is filled with liquid, a blood filtrate. The Nautilus secretes a thin calcareous partition in front of this space and then progressively thickens (and thus strengthens) this new chamber. When sufficiently strong, an epithelium lining the siphuncle (which has been constructed of both hard and soft parts simultaneously with the secretion of the new chamber) begins to actively transport sodium and chloride ions out of the liquid filling the new chamber.

The chamber liquid in this way becomes progressively fresher compared to the blood in the siphuncular tube, thus producing an osmotic gradient. The chamber liquid, now relatively fresh compared to the blood, flows into the siphuncular tube and is passed by the blood vessels into the body of the Nautilus where it is excreted from the body. The removal of this liquid lowers the overall density of the animal with

its heavy shell. It was long thought that, under pressure, the animal secreted the gas found in the chambers but this is not the case. Chamber pressure is never more than one atmosphere, regardless of the depth of the nautilus (and these animals have been caught at a depth of 600 meters, where the ambient pressure pushing against the shell is 60 atmospheres). This gas enters the shell by simple diffusion in response to the near vacuum conditions caused by liquid emptying (nature does indeed abhor a vacuum), but the gas plays no functional role. If liquid removal can balance the density increases caused by shell formation at the aperture and by growth of the soft parts (also denser than surrounding seawater), neutral buoyancy can be maintained.

So this is the major adaptation first seen in the Cambrian: A mollusk must have somehow built a calcareous wall at the back of its shell but left a strand of tissue within, which then had to evolve into an epithelial pump. So (the story goes) some poor late Cambrian monoplacophorans find themselves getting lighter and lighter and at some point float off the surface of the ocean bottom, off for an uncontrolled ride. Eventually, evolutionary forces shape the soft parts into tentacles and more importantly, shape a propulsion system created by the evolution of a jet of water forced through a tube-like funnel. Since the entire shell and animal is of neutral buoyancy, even a feeble squirt through the funnel would cause the animal to move off the bottom, presumably out of harm’s way.

At any rate this is how the story has been understood. But in recent years many have become increasingly skeptical of this hypothesis, and eventually three lines of evidence have changed our views of the reasonableness of this scenario. First was the realization that the Nautilus does not use its buoyancy system for propulsion. For years, paleontologists thought that nautiluses, and by extension the many fossil ammonoids and Nautilus species, undertook nightly vertical migrations from deep to shallow water and did this by changes in buoyancy. Here a balloon analogy colored our view of things. A hot air balloon rises when new, heated air is vented into the balloon or if ballast is thrown out. In either case the density of the balloon is lessened, and the balloon rises in the sky. The balloon is brought back to Earth by venting the hot air. But much research showed that a powered dirigible is a

Nautiloid cephalopods. This group evolved the consummate pump gill, and probably because of this became the largest of all invertebrate animals in the sea, with forms capable of thriving in water too low in oxygen for most other animals. The path of water into the animal, across its gills, and then out the funnel beneath the body allowed respiration and locomotion to be superbly combined.

better model for the nautilus. In the grandiose Zeppelins of the early twentieth century, the balloon was kept at neutral and largely unchanging buoyancy in air and it rose or descended through its use of powered engines. So too with the nautilus. It was indeed discovered that the nautilus undertakes nightly vertical migration, but changes in buoyancy through trim of the shell are not involved. Instead, the strong swimming action of water jetted through a tube beneath the shell powers the animal upward and also pushes it back down to depth.

A second line of discovery dealt with the jet power itself. It is enormously strong, and as scientists made further measurements of the system, it was discovered that a large volume of water was constantly being pumped through the front of the shell, even when the animal is at rest and motionless. There is a simple reason for this—the propulsion system is an offshoot of the respiration system. All of this water that is eventually destined to jet through the locomotion tube, called the hyponome in the Nautilus and Allonautilus, first passes over two pairs of large and complex gills in the back of the mantle cavity. The design allows a dual function—respiration and locomotion—from the same energy expended to draw water in and then force it out.

Finally, in the 1990s it became clear that the Nautilus, Allonautilus, and other cephalopods as well are very efficient at respiration, even in seawater that has oxygen levels lower than that of normal seawater. New research now shows that the Nautilus commonly visits low-oxygen environments. A nautilus can be removed from water for a half-hour with little ill effect. Squid have recently been observed to enter low-oxygen water masses in the Gulf of Mexico with impunity, places where fish cannot go. The system that allows movement has made these animals champion respirers. These three challenges to the earlier view of Nautilus’s evolution being driven by the need for neutral buoyancy has led me to a new model for Nautilus’s Cambrian evolution.

Cephalopods thus came about following the evolution of a marvelous and efficient gill, one that allows them even today to visit anoxic water masses and still harvest whatever oxygen that can be found there. This was the secret of their success, and later in time, when oxygen levels rose, they became even more efficient. They always had a better respiratory system than their prey and competitors, and the doubling up of the respiratory system with locomotion sealed their success. Eventually most lost their shells, but the respiratory system of the cephalopods remains supreme and will ensure their existence far into the future.

Having looked at the respiratory structures of the most populous Cambrian animals, it is time to take a look at a small and insignificant group of fossils found at Chengjiang and later at the Burgess Shale—small, fish-shaped animals that ultimately gave rise to us. The most famous of these was named Pikaia, an animal featured at length in Steven Jay Gould’s book about the Burgess Shale animals called Wonderful Life. The origin of our own phylum is, of course, of intense interest and has been the subject of acrimonious debate for decades, caused in no small way by the dearth of fossils from the time of presumed chordate, or vertebrate, ancestry. Data pertaining to the origins of our phylum come from comparative anatomy and development of living representatives of various phyla seemingly related to us chordates, from DNA studies,

and more recently from the interesting new fossils found in Early Cambrian strata in Chengjiang, China.

Compared to other phyla, ours is very different in several anatomical characteristics. Most animal phyla that show bilateral symmetry have a nerve chord running the length of the body, but this nerve cord lies beneath the gut. In chordates, it is dorsal to the gut. Getting to this reversal of anatomy required some major evolutionary rearranging. One of the oldest and most elegant hypotheses came from the evolutionist W. Garstang, who noted the similarity between the presumed anatomies of ancestral vertebrates (and the fish-like but boneless Lancelet, Amphioxus, which is often used as a model of what the first chordates might have looked like) and the larval stage of the common tunicate, commonly called a sea squirt. While tunicates are sessile filter feeders and look nothing at all like any sort of living chordate, their larva strongly resemble small fish. It was thus theorized by Garstang, followed later by a slew of anatomists, that true vertebrate chordates arose from the larva by a process called paedomorphism—where evolution causes larval characteristics to appear in adults. Later DNA studies have supported this view. We now have a pretty good idea about the “tree” of evolution leading to us chordates: our nearest nonchordate ancestor appears to be the phylum Urochordata, the phylum containing the sea squirts, as so long ago suggested by Garstang.

Is there a connection between this phylogeny, or proposed evolutionary pathway, and levels of oxygen in the latest pre-Cambrian, when this split of the tunicate group into tunicates and vertebrates probably occurred? There has never been any published suggestion that oxygen levels had anything to do with the evolution of the phylum Urochordata. Let’s change that now:

To support this hypothesis we need to look in some detail at the anatomy of the tunicates, present-day representatives of the Urochordata. A tunicate looks like anything but a vertebrate at first glance. Tunicates are sessile filter feeders with two prominent “chimneys” on their topside. Water is sucked into the first tube, forced across a large gill/feeding filter, and then forcibly jetted out of the second tube. The pump itself is powered by muscular action of the outer body wall. Some tunicates move several thousand times their body volume of water through this system daily. According to new, unpublished calculations, this makes tunicates among the champion pump respirers in all the animal kingdom.

Almost the entire inside of the tunicate’s sack-like body is filled with an enormous gill structure. It is highly subdivided, creating a large and intricate surface area for respiration. Blood moves through the interior of these gills, allowing a highly efficient respiratory exchange and is moved through a circulatory system by a heart. Tunicates have no blood pigment, and their inactive life style requires little oxygen. At first glance, their gill system seems to be a case of overkill—there is far more potential than need. In our highly oxygenated oceans it has been estimated that tunicates extract only 10 percent of the available oxygen, which is all they need. If ever there was a design that can handle very low oxygen, this is it.

The respiratory organ inside the adult tunicate has a series of slits within it, and during development these same slits are formed in the front of the tadpole-like larval stage. The larva settles on the bottom and metamorphoses into the sessile adult. The gill slit system of the larva is essentially a perfect preadaptation for the gill system found in fish because the gill structure in the adult tunicates is of a morphology that is fortuitously and easily changed through evolution into the familiar fish gill. The fossil record from Chengjiang contains fully evolved and recognizable tunicates, indicating that tunicates were certainly there well back in the Cambrian.

Although some scientists speculate that the pump may have evolved to capture food when the body was small enough to respire more passively, rather than for respiration, our perspective emphasizes the significance of low-oxygen levels in driving evolutionary responses

to respiratory requirements. Here is the new reasoning behind Hypothesis 3.3. As we have seen, the very low oxygen of the late Precambrian stimulated many morphological ways to get around the obstacle of low oxygen. What little oxygen that was present had to be extracted from the water, and to do this, animals developed a host of morphological solutions. The solution of the tunicate was one of the most elegant and effective. It was also of a design that could be modified for use in a bilaterally symmetrical, swimming creature—the ancestor of our group. Thus, we chordates owe our existence to a solution to the low-oxygen problem confronted by the founding animals involved in the Cambrian Explosion.

The Cambrian Explosion was a time of rapid diversification. But near the end of the period, that push toward ever-greater numbers of different types of animals stopped and diversity began to drop. A mass extinction, perhaps the first ever encountered by animals, descended on life; it’s quite clear there was a series of low-oxygen events in the sea. The trilobites were seriously affected: whole families disappeared, and the survivors and newly evolved forms of the first part of the succeeding Ordovician had a different morphological look. Also killed off were many of the exotic arthropods that had lived during the period, in-

Reconstruction of Pikaia, one of the earliest chordates.

cluding Anomalocaris. It’s a shame that there is not a Burgess Shale–like deposit in the earliest Ordovician rocks so that we could really take a census of how many soft-part animals known from the Cambrianaged Burgess Shale made it through this biotic crisis. It may be that the end-Cambrian event was every bit as destructive as any of the Big Five—but that it has remained underestimated because fewer animals at the time had skeletons and were thus less likely to have been preserved. Whatever its destructiveness, one thing seems clear: the extinction appears to coincide with a rapid drop in oxygen, and this drop may be related to fundamental changes in the carbon cycle. As noted in Chapter 2, sudden drops in oxygen were mass extinction instigators. But the extinction was also an instigator of future diversity, if the relationship noted in Chapter 2 about low oxygen stimulating new species formation is correct.

It would be fascinating to compare the respiratory structures in those animals going extinct at the end of the Cambrian to those that survived. My suspicion is that the survivors were more efficient respirers than those that died out. This is research for the future, however.

The oxygen drop at the end of the Cambrian was short lived. There was a rebound in atmospheric oxygen levels, and with the rebound came a higher diversity of life in the succeeding Ordovician period, coming, perhaps, from a variety of new body plans evolved in the crisis of low oxygen at the end of the Cambrian. This Ordovician expansion of life is the subject of the next chapter.