Neutral theory and non-Darwinian evolution for newbies, Part 1

The Darwinian view:

Natural selection is daily and hourly scrutinising, throughout the world, the slightest variations; rejecting those that are bad, preserving and adding up all that are good; silently and insensibly working, whenever and wherever opportunity offers

Charles Darwin

Software that models the Darwinian view are genetic algorithms like Avida, Tierra, Ev, Steiner, Geometric and Cordova’s remarkable algorithm. Winston Ewert discusses these here.

By way of contrast, I expressed my view as:

Evolution in the wild slowly lets go of the good, and keeps adding up the bad.

Software that models this view is Mendel’s Accountant. This view is also unwittingly modeled in most textbook population genetics — one just has to be willing to see it. Why is it that the most prominent software simulation “proofs” of Darwinian evolution are not biologically realistic, like Steiner or Avida? I pointed out the humorous fact here that evolutionists seem awfully reluctant to transform the math models of textbook population genetics into publicly available simulations of evolution in deep time, while instead they would prefer to highlight Steiner and Avida as examples of the success of Darwinism.

The most comprehensive software model of textbook population genetics for important evolutionary questions was Mendel’s Accountant. It was written by 10 or so creationists (some from top tier science institutions like Cornell, MIT and Princeton), but Darwinists refuse to acknowledge it by saying it is unrealistic, whereas they celebrate Avida and Steiner which have no basis in biological reality whatsoever! With all the nasty comments about Mendel’s Accountant, the critics seem unwilling to actually write and publish software that has the “right” parameters, and not even that, they can’t point you to an existing simulation that is in material disagreement with Mendel’s Accountant.

The claim of this essay is that real evolution of populations as modeled by textbook population genetics does not conform to the Darwinian view, but rather the non-Darwinian view. Shockingly this view is well summarized by this statement:

Most molecular evolution is neutral. Done.

PZ Myers
Former #1 Science Blog

PZ provided a succinct definition of non-Darwinian neutral theory. Done.

Neutral molecular evolution means that most molecular “traits” (a single nucleotide, a codon, a gene, etc.) are not subject to being selected for or against. Their persistence in the population is a matter of blind luck. We have to be very careful here — just because a “trait” is neutral does not mean it has no functional utility. Many organisms have had functional systems (eyes, wings, legs, stomachs, blood, etc.) selected against. So functionality isn’t necessarily correlated with selective advantage. Darwinists proudly call this Survival of the Sickest. I just have to point out that “beneficial” doesn’t necessarily mean functional, sometimes the opposite is true as described by Behe’s Rule.

What PZ may not realize is that if most molecular evolution is neutral AND most mutations erode function, then real evolution must inevitably lose functional complexity, not build it. Famed applied geneticist John Sanford of Cornell calls this principle of deterioration Genetic Entropy.

So how do we know that most molecular evolution is free of selection? First, it is helpful to understand how we mathematically model how traits are strongly or weakly selected for or against or not at all. If we say a trait is strongly selected against, we model this by saying individuals with that trait will tend to have few if any offspring. We can call such traits “deleterious” traits. Conversely if an individual has a selectively favorable trait, we model this trait by saying individuals with that trait will tend to give the individual lots of offspring. We can call this trait a “beneficial” trait. Strength of selection then is measured in terms of how a trait will TEND to increase or decrease the number of offspring an individual may have.

For example, molecular traits that make someone not want to have tons of kids are traits that are “deleterious” whereas molecular traits that make someone want to have tons of kids are traits are “beneficial”. Superficially, we might suspect Richard Dawkins (1 kid) has on balance more deleterious traits than Octomom (14 kids). Poetic justice I suppose.

Given the description of how we define strength of selection in terms of reproduction, we’re now in a position to understand the beginnings of neutral theory and how evolution as a matter of principle had to be non-Darwinian. To illustrate why without all the bewildering math, I’ll start with a simplistic model with a steady population of two individuals (one male, one female). This is purely a pedagogical model, but one than can be extend and refined to more realistic models.

Suppose further in this model that each of the first parents (call them Adam and Eve) have 3 billion nucleotides in their genome, half of which are “deleterious” and half of which are “beneficial”, and the next generation of kids will generally have about the same number on average of 1.5 billion beneficial and 1.5 billion deleterious traits as their parents for the first generational cycle. These proportions are subject to change with each generation, but this is how the model starts out.

Suppose further there will be no new mutations, all we want to do is see how we can evolve an individual that has only beneficial nucleotides and no deleterious ones from the existing gene pool. Suppose also for every generation that the 2 parents always make 8 kids, and only 2 of the kids reproduce, the other 6 are either killed or precluded from reproducing, thus the population is at a steady effective size of 2 for every generation.

If we said the selection against every “deleterious” trait was maximal, meaning any individual with such a trait would not have offspring, it is clear even if a kid somewhere down the line miraculously had only 1.2 billion deleterious nucleotides and 1.8 beneficial ones, under the assumption of maximum selection against deleterious traits, even that wonder kid would not reproduce, and hence if the best kid must die out, all the kids must die out and the species goes extinct! The only solution then is to suppose selection against deleterious traits must necessarily be weak.

Conversely if we guaranteed that individuals with “beneficial” traits had 1 extra kid for every beneficial trait they had, then each individual would have 1.5 billion offspring! Clearly this would violate the constraint that each pair of parents had only 8 kids. Thus, the only solution is to then assume “beneficial” traits are also weakly selected for in the sense that “beneficial” traits increase reproduction rates ever so slightly. The net result is we can effectively say most traits are not selected for or against, and therefore are neutral, and therefore most molecular evolution must be non-Darwinian as a matter of principle.

We can see then that reproductive capacity (number of kids per parent) and the size of the genome affect how much selection in principle can be focused on individual nucleotides in the genome in a particulate (not corporate) manner. The smaller the reproductive capacity of a species and the larger the genomes, the more dilute selection must be on each nucleotide as a matter of principle.

The bottom line is molecular evolution at the nucleotide level must in principle be mostly neutral which means it must be non-Darwinian in principle. If most molecular evolution is non-Darwinian, this strongly suggests most other kinds of evolution must be non-Darwinian! Why should other levels of evolution higher than the molecular level be exempt from the mathematical considerations just laid out? The view that almost all evolution must be non-Darwinian has been advocated by Masotoshi Nei, a fact I pointed out here.

It might be helpful at this point to draw some important distinctions since the phrase “neutral theory” can be subject to lots of equivocation. Neutral theory generally refers to the views of Kimura and Nei and other evolutionists. When I previously said many creationists were neutralists, I was merely saying many creationists argue the biological world must be free of selection as populations reproduce. This is the view of the ReMine and Sanford who are creationists but who reject Kimura/Nei’s view how complex organisms came to be.

I will try in the future to make a distinction between Kimura/Nei vs. ReMine/Sanford neutral theories. Kimura/Nei and ReMine/Sanford agree that most biological change is non-Darwinian and selectively neutral, but they disagree about what happened in the historical past. ReMine/Sanford argue neutral theory in the operational sense, but not in the historical sense. Kimura/Nei argue neutral theory in both the operational and historical sense.

Part 2 will delve into mutation accumulation, fixation, and random loss of traits by recombination or random accidents.

Hope this helps.


1. I put “beneificial” and “deleterious” in quotes to emphasize the notions of good and bad in the evolutionary world view have gotten so twisted that sickle cell anemia, juvenile diabetes, cystic fibrosis are viewed as “beneficial” traits whereas stomachs, eyes, wings, legs, blood are occasionally “deleterious” — exactly the opposite of how we intuitively view such traits.

2. I put the word “trait” in quotes because traits are often thought of in terms of phenotypic traits, whereas for molecular evolution we are often focusing on evolution at the nucleotide level and so we have to redefine the notion of trait to include nucleotides.

3. Here is the more mathematical description of how selection is modeled.
One Gene. Anyone with better links is invited to post.

4. Any corrections to what I said are welcome. Thanks in advance.