Testing Natural Selection: Part 1

 
The latest issue of Scientific American has an interesting article by H. Allen Orr entitled Testing Natural Selection.

Biologists working with the most sophisticated genetic tools are demonstrating that natural selection plays a greater role in the evolution of genes than even most evolutionists had thought.

Orr is an adaptationist. His perspective on evolution focuses on natural selection as the predominant mechanism. He tends to dismiss all other mechanisms as either uninteresting or unimportant.

I though it might be interesting to compare what a pluralist might say about some of the things in the article. It's one way of highlighting the difference between the two points of view.

Naturally, as a pluralist, I disagree with some statements. My main beef, however, is with the growing tendency to over-emphasize natural selection as we approach the 200th anniversary of Darwin's birth and the 150th anniversary of publication of On the Origin of Species. I think it's possible to describe the differences between evolution in the eighteenth century and evolution in the 21st century without diminishing Darwin's contributions.

Orr begins his article by describing natural selection. He explains that there are several kinds of mutations ...

Most important, we know something about the effects of mutations on fitness. The overwhelming majority of mutations are harmful—that is, they reduce fitness; only a tiny minority are beneficial, increasing fitness.

That's not exactly how I would put it. I would have added that there's a third type of mutation that is neither harmful nor beneficial—neutral mutations.

Furthermore, I would have explained that the frequency of these three different kinds of mutations can vary considerably from one species to the next depending on the organization of the genome. In animals and plants, for example, most of the DNA does not seem to be essential so that the overwhelming majority of mutations are neutral and a smaller number—those that interfere with an essential function—are deleterious. A few mutations can be beneficial.

Orr goes on to say ....

Most mutations are bad for the same reason that most typos in computer code are bad: in finely tuned systems, random tweaks are far more likely to disrupt function than to improve it.

I would not use this analogy because it emphasizes something that I think is false; namely that organisms are "fine tuned systems." I tend to think of them as sloppy Rube Goldberg machines and not as well-tested computer code.

I would say that most mutations in essential regions of the genome are deleterious because random hits in DNA are more likely to make things worse than to make things better. The distinction is subtle, but important. Many adaptationists use language implying that living organisms are almost perfectly adapted to their present environment.

In the next section, Orr describes the advances of population genetics and its influence on how we understand natural selection. I would have described how population genetics led to an understanding of all type of evolution, and not just natural selection. Here's what Orr says,

Population geneticists have also provided insight into natural selection by describing it mathematically. For example, geneticists have shown that the fitter a given type is within a population, the more rapidly it will increase in frequency; indeed, one can calculate just how quickly the increase will occur. Population geneticists have also discovered the surprising fact that natural selection has unimaginably keen “eyes,” which can detect astonishingly small differences in fitness among genetic types. In a population of a million individuals, natural selection can operate on fitness differences as small as one part in a million.

I would have said that the growth of population genetics in the early part of the 20th century led to the recognition of random genetic drift as an important mechanism of evolution. Models were developed to explain how natural selection affected the increase in frequency of a beneficial allele and how neutral alleles could also increase in frequency even though they were invisible to natural selection.

The population geneticists also discovered that harmful alleles could become fixed by accident, although that turns out to be a rare event. More importantly, they discovered that natural selection has a stochastic component. Beneficial alleles will only become fixed part of the time. The probability depends on the fitness advantage. For example, if an allele has a fitness advantage of 10% then it will only become fixed 20% of the time. In 80% of cases when such an allele arises in a population it will be lost by random genetic drift before it becomes fixed.¹

As the fitness advantage diminishes, the probability of fixation becomes lower and lower so that alleles with small fitness advantages (<1%) will hardly ever change the species. That's what population geneticists discovered about natural selection.

The probability of fixation of neutral alleles (or nearly neutral alleles) is very low but since there are so many more of them than beneficial alleles, much of evolution is characterized by changes due to random genetic drift.

The next section is "How Common Is Natural Selection?". This is where Orr asks the key question ...

One of the simplest questions biologists can ask about natural selection has, surprisingly, been one of the hardest to answer: To what degree is it responsible for changes in the overall genetic makeup of a population? No one seriously doubts that natural selection drives the evolution of most physical traits in living creatures—there is no other plausible way to explain such large-scale features as beaks, biceps and brains. But there has been serious doubt about the extent of the role of natural selection in guiding change at the molecular level. Just what proportion of all evolutionary change in DNA is driven, over millions of years, by natural selection—as opposed to some other process?

We've discussed this distinction between molecular changes and physical traits many times. One of the most annoying characteristics of adaptationists is that they insist on relegating other mechanisms of evolution to the level of DNA sequences but refuse to consider anything but natural selection when it comes to visible phenotypes. There is no justification for this assumption. Many physical traits can be neutral or even deleterious. They were not fixed by natural selection.²

What Orr says is simply not true. There are many biologists who seriously doubt that natural selection drives the evolution most physical traits, even though such pluralists readily agree that most adaptions are due to natural selection. Random genetic drift is a plausible way to explain many physical traits.

Until the 1960s biologists had assumed that the answer was “almost all,” but a group of population geneticists led by Japanese investigator Motoo Kimura sharply challenged that view. Kimura argued that molecular evolution is not usually driven by “positive” natural selection—in which the environment increases the frequency of a beneficial type that is initially rare. Rather, he said, nearly all the genetic mutations that persist or reach high frequencies in populations are selectively neutral—they have no appreciable effect on fitness one way or the other. (Of course, harmful mutations continue to appear at a high rate, but they can never reach high frequencies in a population and thus are evolutionary dead ends.) Since neutral mutations are essentially invisible in the present environment, such changes can slip silently through a population, substantially altering its genetic composition over time. The process is called random genetic drift; it is the heart of the neutral theory of molecular evolution.

As I've already pointed out, random genetic drift was discovered in the 1920s and it was incorporated into the first version of the Modern Synthesis in the 1940s. It dropped out of favor when the synthesis hardened at the time of the Darwin centennial in 1959.

Random genetic drift was revived in the late 1960's with the discovery of neutral alleles. Drift is the way in which selectively neutral alleles become fixed in a population. Random genetic drift and neutral theory are not synonyms.

As I indicated above, since the vast majority of animal and plant genomes is non-essential, it stands to reason that the vast majority of alleles will be neutral. Thus at the molecular level, at least, random genetic drift must be the dominant mechanism of evolution.

By the 1980s many evolutionary geneticists had accepted the neutral theory. But the data bearing on it were mostly indirect; more direct, critical tests were lacking. Two developments have helped fix that problem. First, population geneticists have devised simple statistical tests for distinguishing neutral changes in the genome from adaptive ones. Second, new technology has enabled entire genomes from many species to be sequenced, providing voluminous data on which these statistical tests can be applied. The new data suggest that the neutral theory underestimated the importance of natural selection.

Hmmm ... I could see where this was going even before I read it. Orr is about to quote the infamous work of Drosophila geneticists who have devised complicated tests to show that some synonymous mutations might confer a selective advantage in one species but not in another closely related species. Some of the papers claim that many alleles in coding regions are not neutral even thought they don't change the amino acid. There's no question that this is true in some cases.

It's also true that mutations altering the amino acid are sometimes beneficial, and therefore selected. However, if you align the amino acid sequences of a given gene from hundreds of species and map them on to the structure of the protein it becomes readily apparent that most substitutions cannot have a significant effect on the function of the protein. They must be neutral, or nearly neutral. As a matter of fact, in most proteins it is difficult to find any clearly beneficial alleles present in one species and not in the others.

In one study a team led by David J. Begun and Charles H. Langley, both at the University of California, Davis, compared the DNA sequences of two species of fruit fly in the genus Drosophila. They analyzed roughly 6,000 genes in each species, noting which genes had diverged since the two species had split off from a common ancestor. By applying a statistical test, they estimated that they could rule out neutral evolution in at least 19 percent of the 6,000 genes; in other words, natural selection drove the evolutionary divergence of a fifth of all genes studied. (Because the statistical test they employed was conservative, the actual proportion could be much larger.) The result does not suggest that neutral evolution is unimportant—after all, some of the remaining 81 percent of genes may have diverged by genetic drift. But it does prove that natural selection plays a bigger role in the divergence of species than most neutral theorists would have guessed. Similar studies have led most evolutionary geneticists to conclude that natural selection is a common driver of evolutionary change even in the sequences of nucleotides in DNA.

Pluralists disagree. We still think that random genetic drift is by far the dominant mechanism at the molecular level and that it even plays a significant role at the level of visible phenotypes.

In addition, we like to remind adaptationists that most beneficial alleles are eliminated by random genetic drift before they ever become fixed in a population.

---

1. Many biologists, and most evolutionary psychologists, do not understand this important point. They think that all they have to do is identify some (real or imagined) benefit and it will automatically take over the population no matter how small the benefit.

2. I know that Orr said "most" physical traits and not "all" physical traits. It's a distinction without meaning since the percentage of non-adaptive changes that adaptationists are willing to admit, grudgingly, is not much different than zero.