Lighting the Cambrian Fuse

The evolution of the genetic tool kit was a key ingredient in the Cambrian explosion. But once it had evolved, the explosion didn’t happen right away. The earliest animals that possessed the genetic tool kit probably lived tens of millions of years before the first record of the Cambrian explosion 535 million years ago. If animals already had so much evolutionary potential within them, something must have prevented them from taking off.

The tool kit of these early animals may have been like a fuse waiting for a match. Before the Cambrian, the oceans weren’t a propitious place for animal evolution. Big, active animals like the ones that appeared in the Cambrian explosion need a lot of energy, and to generate it they need to absorb oxygen. But the chemistry of rocks formed on the seafloor during the Precambrian suggests that there wasn’t much oxygen for them to take up. Photosynthetic algae and bacteria at the ocean’s surface were releasing oxygen in abundance, but little of it was getting very far below. Oxygen‑breathing scavenger bacteria on the surface fed on these photosynthesizers after they had died, leaving the rest of the ocean oxygen‑starved.

Around 700 million years ago oxygen levels began to rise, eventually reaching perhaps half the concentration found today. The rise of oxygen has been linked to the breakup of a supercontinent at the time. As it disintegrated, more carbon may have been carried down into new ocean basins, leaving behind more free oxygen in the atmosphere. Some of that extra oxygen then managed to build up in the ocean as well.

After the oxygen levels rose in the oceans, the planet as a whole appears to have gone through some violent times. Ice ages overwhelmed the earth, according to Harvard geologist Paul Hoffman, bringing glaciers close to the equator. They melted away only after volcanoes released enough carbon dioxide to warm the atmosphere. As life became isolated in refuges during the global ice ages, evolution may have taken place at a faster pace, creating new species with new adaptations. Because animals already had their complex genetic circuits in place, they could respond to this evolutionary pressure by flowering into all the forms of the Cambrian explosion.

Genes and physical conditions may have started the Cambrian explosion, but ecology may have determined how big it eventually became. Among the new kinds of animals that emerged in the early Cambrian were creatures that could, for the first time in the history of Earth, eat algae. These invertebrates used feathery appendages to catch their food and became enormously successful. (Today, vast armies of fairy shrimp, water fleas, and other algae eaters continue their success.) Once algae eaters began to thrive, they spurred the appearance of large, fast‑swimming predators, which in turn could have been devoured by larger predators still. The ocean’s food web quickly wove itself into a complex tangle.

The new pressures of grazing and hunting may have spurred even more diversification, not just among animals but in algae as well. The most common group of algae found in the early fossil record is known as the acritarchs. Before the Cambrian, acritarchs were small and nondescript, but during the Cambrian explosion they suddenly evolved spikes and other ornaments and took on much bigger forms. They were probably evolving defenses against the grazers, making themselves harder to swallow. The grazers evolved ways to get around these defenses and evolved defenses of their own–spikes, shells, and plates of armor–against their predators, which in turn had to find new ways to attack them, evolving claws and crushing teeth and drills and sharper senses. The Cambrian explosion became a fire that could feed itself.

The Party’s Over

Yet within a few million years the fire had gone out. Paleontologists recognize only one new phylum that evolved after the Cambrian explosion–bryozoans, colonial animals that form mats on the ocean floor. That’s not to say that animals have not changed since then. While the first vertebrates were all lamprey‑like creatures, they’ve since evolved into a staggering variety, from the snowy egret to the tree kangaroo, the hammerhead shark, the vampire bat, and the sea snake. But all of these animals have two eyes, a brain housed in a skull, and muscles surrounding their skeletons. Evolution may be a creative force, but it does not have infinite scope. In fact, it works under tight constraints and can fall into many different traps.

When life goes through a burst of evolutionary transformations, the new species start seeking out ecological niches. In Lake Victoria, the cichlids evolved to scrape algae off rocks, eat insects, and take advantage of other food in the lake. The first algae scrapers may not have done a very good job, but without any other algae scrapers around, not very good was good enough. As these cichlids evolved, they created new ecological niches for even more species of cichlids: the predators that swallow them whole, or the scale rakers, or the egg stealers. Life creates new niches, but it probably can’t create an infinite number of them. Sooner or later, species will begin competing with one another for them instead. Some will win and others will lose. In the older lakes of Malawi and Tanganyika, the cichlids have had millions of years more time to evolve, but they have not invented any ecological niches beyond those found in young Lake Victoria.

The Cambrian explosion may have come to an end when its ecosystem filled up in much the same way, only on a far grander scale. In the Cambrian explosion, big mobile predators appeared on Earth for the first time, as did burrowers and algae grazers. It’s possible that these animals filled all the possible ecological niches and evolved to be so proficient that they shut out newcomers. Without an opportunity to explore new designs, new kinds of animals can’t establish themselves.

Sometimes an explosive burst of evolution may stop because the genetic complexity it creates blocks its own path. The earliest animals were extremely simple, with only a handful of different kinds of cells, assembled by relatively few developmental genes. By the end of the Cambrian explosion, their descendants had evolved many different types of cells and used a complicated network of interacting genes to build their bodies. Often a gene that has evolved to help build one structure was borrowed to build several others. Hox genes, for example, build not only the brains and spines of vertebrates but also their fins and legs. When a gene is responsible for several different jobs, it becomes harder to change. Even if a mutation improves one structure the gene helps to build, it may completely destroy the other ones. The difference between the way evolution operated at the beginning of the Cambrian and at the end may be akin to the difference between trying to remodel a one‑story house and remodeling a skyscraper.

Because evolution can only tinker, it cannot produce the best of all possible designs. Although it has come up with many structures that make engineers coo in admiration, it is often stuck making the best of a bad situation. Our eyes, for example, are certainly impressive video cameras, and yet in some ways they are fundamentally flawed.

When light enters a vertebrate eye, it travels through the jelly and strikes the photoreceptors of the retina. But the neurons in the retina are actually pointed backward. It’s as if we were gazing at our own brain. Light has to make its way through several layers of neurons and a web of capillaries before it finally gets to the nerve endings that can detect it.

Once light strikes the backward‑pointing photoreceptors of the retina, the photoreceptors then have to send their signals back up through the layers of the retina toward the front of the eye. As they travel, the neurons process the signals, sharpening the image. The uppermost layer of retinal neurons connects to the optic nerve, which sits on the top of the retina. The nerve burrows back down through all the neurons and capillaries in order to leave the eye and travel to the back of the brain.

This architecture is, as the evolutionary biologist George Williams has bluntly put it, “stupidly designed.” The layers of neurons and capillaries act like a mask, degrading the light that finally gets through to the photoreceptors. In order to compensate, our eyes are continually making tiny movements so that the shading shifts around the image we see. Our brains can then combine these degraded pictures, subtract away the shading, and create a clear image.

Another flaw arises out of the way the retinal neurons attach to the optic nerve on top of the retina. The optic nerve blocks even more incoming light, creating a blind spot in each eye. The only reason blind spots don’t cause us much trouble is that our brains can combine the images from both eyes, canceling out each blind spot, and create a full picture.

Yet another clumsy element of eye design is the way that the retina is anchored. Because the photoreceptors have delicate, hairy nerve endings, they can’t be cemented firmly in place. Instead, they are loosely joined to a layer of cells lining the wall of the eye called the retinal pigment epithelium. The retinal pigment epithelium is essential for the eye. It absorbs extra photons so that they don’t bounce back at the photoreceptors and blur the image they receive. It also houses blood vessels that supply the retina with nutrients, and as the retina sloughs off old photoreceptors, it can carry away the waste. But the connection between the epithelium and the retina is so fragile that our eyes can’t withstand much abuse. A swift punch to the head can detach the retina, leaving it free to float around the inside of the eye.

Eyes can work perfectly well without taking this shape. Just compare the vertebrate eye to the eye of a squid. Squid eyes are powerful enough to let them track their prey in near darkness. They are spherical and have lenses, just like vertebrate eyes, but when incoming light strikes the inner wall of a squid’s eye, it does not have to struggle through a tangle of backward neurons. Instead, it immediately strikes a vast number of light‑sensitive endings of the squid’s optic nerve. The signals run directly from the nerve endings into the squid’s brain, without having to travel backward over to any intervening layers of neurons.

To understand the flaws of the vertebrate eye (as well as its strengths), evolutionary biologists look back at its origins. The best clues to the early evolution of vertebrate eyes come from lancelets, the closest living relatives to vertebrates. The nerve cord of a lancelet is actually a tube, and the neurons that line it have hairlike projections called cilia that project into its hollow center. Some of the neurons at the very front of the tube serve as a light‑sensing eyespot. Like other neurons in the lancelet, these light sensors are pointed inward, which means that they can only detect light that strikes the opposite side of the lancelet’s transparent body and pass into the hollow tube.

Just in front of these light‑sensing neurons, the neural tube ends. The cells that line the front of the tube have dark pigment inside them, which scientists suspect acts as a shield, blocking out light striking the front of the lancelet. Since light can’t hit its eyespot from all directions, the lancelet can then use it to orient itself in the water.

Thurston Lacalli, a biologist at the University of Saskatchewan, has discovered some remarkable similarities between the structure of the lancelet eyespot and a vertebrate embryo’s eye. A vertebrate brain forms initially as a hollow tube, just like the one in lancelets, with inward‑pointing nerve cells. Eyes develop at the front end of this tube, as its walls project outward into a pair of horns. At the tip of each horn, a cup forms. On this cup’s inner surface, the retinal neurons establish themselves, their nerve endings still pointing inward. On the outer surface, the pigmented cells take hold.

If you slice this eyecup and look at the arrangement of cells, you’ll find it has the same topography as the lancelet eyespot. The retinal neurons still point inward, toward the center of the neural tube, as the light‑sensing neurons do in lancelets. The rods and cones of the retina are highly evolved versions of the lancelet’s cilia. As the tube changes shape in vertebrates, they end up pointing toward the wall of the eye. Moreover, the retinal neurons in a vertebrate embryo are still positioned between the pigmented cells and the optic nerve in the same way as they are in the lancelet head.

These connections are clearest when a vertebrate is still a new embryo. The more it develops, the harder it becomes to see the similarities. The walls of the cup grow so thin that eventually the cells on the inner edge and the outer edge press against each other. In the process, they create the peculiar, delicate connection between the retina and the pigmented epithelium.

These similarities between vertebrate embryo eyes and lancelet eyespots offer clues to how our eyes acquired their strange shape. The eyespot of a lancelet‑like ancestor evolved into a pair of cup‑shaped light detectors branching off the nerve tube. Their cup shape allowed them to capture more light than a flat eyespot. It gradually curved into a spherical eye that could begin to form images on its retina. But because the vertebrate eye was built on a basic lancelet design, it was stuck with retinal neurons pointing away from the incoming light.

The structure of eyespots in the ancestors of vertebrates constrained the forms that they could later take. Evolution could only make the best of a lancelet‑like anatomy and the rules of development that built it. In order to turn an eyespot into a true eye, we have to put up with blind spots, detached retinas, and degraded light. Yet the advantages of having any ability to form images outweighed the unavoidable shortcomings of the eye’s design.

Once the vertebrate eye had evolved, complete with lens, jelly, and backward retina, many lineages evolved new versions that work better in their own environment. For example, three different lineages of fishes have each evolved double eyes. Their eyes have two pairs of lenses rather than one; when these fishes float at the water’s surface, one pair of eyes gazes up into the air, while the other looks down into the water. The upward‑pointing eyes are shaped to focus light as it passes through air, while the other pair is designed to handle the optics of water.

Meanwhile a few vertebrates on land–most notably birds and primates like ourselves–have evolved extraordinarily powerful vision. They developed a dense patch of photoreceptors in a small region of the retina called the fovea; the neurons that would normally block the path of incoming light to the fovea are pushed to the sides. And yet, despite all these innovations, the backward retina endures. Thanks to 530 million years of evolutionary constraints, our children will never be able to see like squid.