Can Selection Build Complexity

But even if we agree that natural selection does work in nature, how much work can it really do? Sure, selection can change the beaks of birds, or the flowering period of plants, but can it build complexity? What about intricate traits like the tetrapod limb; or exquisite biochemical adaptations like blood clotting, which involves a precise sequence of steps involving many proteins; or perhaps the most complicated apparatus that ever evolved—the human brain?

We are at somewhat of a handicap here because, as we know, complex features take a long time to evolve, and most of them did so in the distant past when we weren't around to see how it happened. So how can we be sure that selection was involved? How do we know that creationists are wrong when they say that selection can make small changes in organisms but is powerless to make big ones?

But first we must ask: what's the alternative theory? We know of no other natural process that can build a complex adaptation. The most commonly suggested alternative takes us into the realm of the supernatural. This, of course, is creationism, known in its latest incarnation as "intelligent design" (ID). Advocates of ID suggest that a supernatural designer has intervened at various times during the history of life, either instantly calling into being the complex adaptations that natural selection supposedly can't make, or producing "miracle mutations" that can't occur by chance. (Some IDers go further: they are the extreme "young Earth" creationists who believe that Earth is about 6,000 years old and that life has no evolutionary history at all.)

In the main, ID is unscientific, for it consists largely of untestable claims. How, for example, can we determine whether mutations were mere accidents in DNA replication or were willed into being by a creator? But we can still ask if there are adaptations that could not have been built by selection, and therefore require us to think of another mechanism. Advocates of ID have suggested several such adaptations, such as the bacterial flagellum (a small, hair-like apparatus with a complex molecular motor, used by some bacteria to propel themselves) and the mechanism of blood clotting. These are indeed complex features: the flagellum, for instance, is composed of dozens of separate proteins, all of which must work in concert for the hair-like "propeller" to move.

IDers argue that such traits, involving many parts that must cooperate for that trait to function at all, defy Darwinian explanation. Therefore, by default, they must have been designed by a supernatural agent. This is commonly called the "God of the gaps" argument, and it is an argument from ignorance. What it really says is that if we don't understand everything about how natural selection built a trait, that lack of understanding itself is evidence for supernatural creation.

You can probably see why this argument doesn't hold water. We'll never be able to reconstruct how selection created everything—evolution happened before we were on the scene, and some things will always be unknown. But evolutionary biology is like every science: it has mysteries, and many of them get solved, one after the other. We now know, for instance, where birds came from—they weren't created out of thin air (as creationists used to maintain), but evolved gradually from dinosaurs. And each time a mystery is solved, ID is forced to retreat. Since ID itself makes no testable scientific claims, but offers only half-baked criticisms of Darwinism, its credibility slowly melts away with each advance in our understanding. Furthermore, ID's own explanation for complex features—the whim of a supernatural designer—can explain any conceivable observation about nature. It may even have been the creator's whim to make life look as though it evolved (apparently many creationists believe this, though few admit it). But if you can't think of an observation that could disprove a theory, that theory simply isn't scientific.

How, though, can we refute the ID claim that some traits simply defy any origin by natural selection? In such cases the onus is not on evolutionary biologists to sketch out a precise step-by-step scenario documenting exactly how a complex character evolved. That would require knowing everything about what happened when we were not around— an impossibility for most traits and for nearly all biochemical pathways. As the biochemists Ford Doolittle and Olga Zhaxybayeva argued when addressing the ID claim that flagella could not have evolved, "Evolutionists need not take on the impossible challenge of pinning down every detail of flagellar evolution. We need only show that such a development, involving processes and constituents not unlike those we already know and can agree upon, is feasible." And by "feasible," they mean that there must be evolutionary precursors of each new trait, and that evolution of that trait does not violate the Darwinian requirement that each step in building an adaptation benefits its possessor.

Indeed, we know of no adaptations whose origin could not have involved natural selection. How can we be sure? For anatomical traits, we can simply trace their evolution (when possible) in the fossil record, and see in what order different changes took place. We can then determine whether the sequences of changes at least conform to a step-by-step adaptive process. And in every case, we can find at least a feasible Darwinian explanation. We've seen this for the evolution of land animals from fish, of whales from land animals, and of birds from reptiles. It didn't have to be that way. The movement of nostrils to the top of the head in ancestral whales, for example, could have preceded the evolution of fins. That could be the providential act of a creator, but couldn't have evolved by natural selection. But we always see an evolutionary order that makes Darwinian sense.

Understanding the evolution of complex biochemical features and pathways is not as easy, since they leave no trace in the fossil record. Their evolution must be reconstructed in more speculative ways, trying to see how such pathways could be cobbled together from simpler biochemical precursors. And we'd like to know the steps in this cobbling, to see if each new one could bring improved fitness.

Although advocates of ID claim a supernatural hand behind these pathways, dogged scientific research is beginning to give plausible (and testable) scenarios for how they could have evolved. Take the blood-clotting pathway of vertebrates. This involves a sequence of events that begins when one protein sticks to another in the vicinity of an open wound. That sets off a complicated cascade reaction, sixteen steps long, each involving an interaction between a different pair of proteins and culminating in the formation of the clot itself. Altogether over twenty proteins are involved. How could this possibly have evolved?

We don't yet know for sure, but we have evidence that the system could have been built up in an adaptive way from simpler precursors. Many of the blood-clotting proteins are made by related genes that arose by duplication, a form of mutation in which an ancestral gene, and later its descendants, becomes duplicated in full along a strand of DNA because of a mistake during cell division. Once they arise, such duplicated genes can then evolve along separate pathways so that they eventually perform separate functions, as they now do in blood clotting. And we know that other proteins and enzymes in the pathway had different functions in groups that evolved before vertebrates. For example, a key protein in the clotting pathway is called fibrinogen, which is dissolved in blood plasma. In the last step of blood clotting, this protein gets cut by an enzyme, and the shorter proteins (called fibrins) stick together and become insoluble, forming the final clot. Since fibrinogen occurs in all vertebrates as a blood-clotting protein, it presumably evolved from a protein that had a different function in ancestral invertebrates, who were around earlier but lack a clotting pathway. Although an intelligent designer could invent a suitable protein, evolution doesn't work that way. There must have been an ancestral protein from which fibrinogen evolved.

Russell Doolittle at the University of California predicted that we would find such a protein, and, sure enough, in 1990 he and his colleague Xun Xu discovered it in the sea cucumber, an invertebrate sometimes used in Chinese cooking. Sea cucumbers branched off from the vertebrate lineage at least 500 million years ago, yet they have a protein that, while clearly related to blood-clotting proteins of vertebrates, is not used to clot blood. This means that the common ancestor of sea cucumbers and vertebrates had a gene that was later co-opted in vertebrates for a new function, precisely as evolution predicts. Since then, both Doolittle and cell biologist Ken Miller have worked out a plausible and adaptive sequence for the evolution of the entire blood-clotting cascade from parts of precursor proteins. All of these precursors are found in invertebrates, where they have other, nonclotting functions, and were evolu-tionarily co-opted by vertebrates into a working clotting system. And the evolution of the bacterial flagellum, though not yet fully understood, is also known to involve many proteins co-opted from other biochemical pathways.31

Hard problems often yield before science, and though we still don't understand how every complex biochemical system evolved, we are learning more every day. After all, biochemical evolution is a field in still its infancy. If the history of science teaches us anything, it is that what conquers our ignorance is research, not giving up and attributing our ignorance to the miraculous work of a creator. When you hear someone claim otherwise, just remember these words of Darwin: "Ignorance more frequently begets confidence than does knowledge: it is those who know little, and not those who know much, who so positively assert that this or that problem will never be solved by science."

It appears, then, that in principle there's no real problem with evolution building complex biochemical systems. But what about time? Has there really been enough time for natural selection to create both complex adaptations as well as the diversity of living forms? Certainly we know that there was enough time for organisms to have evolved—the fossil record alone tells us that—but was natural selection strong enough to drive such change?

One approach is to compare the rates of evolution in the fossil record with those seen in laboratory experiments that used artificial selection, or with historical data on evolutionary change that occurred when species colonized new habitats in historical times. If evolution in the fossil record were much faster than in laboratory experiments or colonization events—both of which involve very strong selection—we might need to rethink whether selection could explain changes in fossils. But in fact the results are just the opposite. Philip Gingerich at the University of Michigan showed that rates of change in animal size and shape during laboratory and colonization studies are actually much faster than rates of fossil change: from 500 times faster (selection during colonizations) to nearly a million times faster (laboratory selection experiments). And even the fastest rates of evolution in the fossil record are nowhere near as fast as the slowest rates seen when humans practice selection in the laboratory. Further, the average rates of evolution seen in colonization studies are large enough to turn a mouse into the size of an elephant in just 10,000 years!

The lesson, then, is that selection is perfectly adequate to explain changes that we see in the fossil record. One reason why people raise this question is because they don't (or can't) appreciate the immense spans of time that selection has had to work. After all, we evolved to deal with things that happen on the scale of our lifetime—probably around thirty years during most of our evolution. A span of ten million years is beyond our intuitive grasp.

Finally, is natural selection sufficient to explain a really complex organ, such as the eye? The "camera" eye of vertebrates (and mollusks like the squid and octopus) was once beloved by creationists. Noting its complex arrangement of the iris, lens, retina, cornea and so on—all of which must work together to create an image—opponents of natural selection claimed that the eye could not have formed by gradual steps. How could "half an eye" be of any use?

Darwin brilliantly addressed, and rebutted, this argument in The Origin. He surveyed existing species to see if one could find functional but less complex eyes that not only were useful, but also could be strung together into a hypothetical sequence showing how a camera eye might evolve. If this could be done—and it can—then the argument that natural selection could never produce an eye collapses, for the eyes of existing species are obviously useful. Each improvement in the eye could confer obvious benefits, for it makes an individual better able to find food, avoid predators, and navigate around its environment.

A possible sequence of such changes begins with simple eyespots made of light-sensitive pigment, as seen in flatworms. The skin then folds in, forming a cup that protects the eyespot and allows it to better localize the light source. Limpets have eyes like this. In the chambered nautilus, we see a further narrowing of the cup's opening to produce an improved image, and in ragworms the cup is capped by a protective transparent cover to protect the opening. In abalones, part of the fluid in the eye has coagulated to form a lens, which helps focus light, and in many species, such as mammals, nearby muscles have been co-opted to move the lens and vary its focus. The evolution of a retina, an optic nerve, and so on, follows by natural selection. Each step of this hypothetical transitional "series" confers increased adaptation on its possessor, because it enables the eye to gather more light or form better images, both of which aid survival and reproduction. And each step of this process is feasible because it is seen in the eye of a different living species. At the end of the sequence we have the camera eye, whose adaptive evolution seems impossibly complex. But the complexity of the final eye can be broken down into a series of small, adaptive steps.

Yet we can do even better than just stringing together eyes of existing species in an adaptive sequence. We can, starting with a simple precursor, actually model the evolution of the eye and see whether selection can turn that precursor into a more complex eye in a reasonable amount of time. Dan-Eric Nilsson and Susanne Pelger of Lund University in Sweden made such a mathematical model, starting with a patch of lightsensitive cells backed by a pigment layer (a retina). They then allowed the tissues around this structure to deform themselves randomly, limiting the amount of change to only 1 percent of size or thickness at each step. To mimic natural selection, the model accepted only "mutations" that improved the visual acuity, and rejected those that degraded it.

Within an amazingly short time, the model yielded a complex eye, going through stages similar to the real-animal series described above. The eye folded inward to form a cup, the cup became capped with a transparent surface, and the interior of the cup gelled to form not only a lens, but a lens with dimensions that produced the best possible image.

Beginning with a flatworm-like eyespot, then, the model produced something like the complex eye of vertebrates, all through a series of tiny adaptive steps—1,829 of them, to be exact. But Nilsson and Pel-ger could also calculate how long this process would take. To do this, they made some assumptions about how much genetic variation for eye shape existed in the population that began experiencing selection, and about how strongly selection would favor each useful step in eye size. These assumptions were deliberately conservative, assuming that there were reasonable but not large amounts of genetic variation and that natural selection was very weak. Nevertheless, the eye evolved very quickly: the entire process from rudimentary light-patch to camera eye took fewer than 400,000 years. Since the earliest animals with eyes date back 550 million years ago, there was, according to this model, enough time for complex eyes to have evolved more than fifteen hundred times over. In reality, eyes have evolved independently in at least forty groups of animals. As Nilsson and Pelger noted dryly in their paper, "It is obvious that the eye was never a real threat to Darwin's theory of evolution."

So where are we? We know that a process very like natural selection— animal and plant breeding—has taken the genetic variation present in wild species and from it created huge "evolutionary" transformations. We know that these transformations can be much larger, and faster, than real evolutionary change that took place in the past. We've seen that selection operates in the laboratory, in microorganisms that cause disease, and in the wild. We know of no adaptations that absolutely could not have been molded by natural selection, and in many cases we can plausibly infer how selection did mold them. And mathematical models show that natural selection can produce complex features easily and quickly. The obvious conclusion: we can provisionally assume that natural selection is the cause of all adaptive evolution—though not of every feature of evolution, since genetic drift can also play a role.

True, breeders haven't turned a cat into a dog, and laboratory studies haven't turned a bacterium into an amoeba (although, as we've seen, new bacterial species have arisen in the lab). But it is foolish to think that these are serious objections to natural selection. Big transformations take time—huge spans of it. To really see the power of selection, we must extrapolate the small changes that selection creates in our lifetime over the millions of years that it has really had to work in nature. We can't see the Grand Canyon getting deeper, either, but gazing into that great abyss, with the Colorado River carving away insensibly below, you learn the most important lesson of Darwinism: weak forces operating over long periods of time create large and dramatic change.

This page intentionally left blank

Was this article helpful?

0 0

Post a comment