ZOOL 304, Class Notes

Chapter 3, Jan. 27 - Feb. 10.  

Text reading:  Chapter 3.

Notes for chapter 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17

304 index page

Introductory comments.

In the previous chapter (Chapter 2), we examined adaptive evolution.  In this chapter (Chapter 3), we consider neutral evolution.  In the next chapter (Chapter 4) we learn about genetic processes that affect genetic variation within populations.  Ultimately, all evolution must be explained in terms of these processes.  These chapters present the core concepts for understanding the mechanisms of evolution.

304 index page

Adaptive evolution vs. neutral evolution

By definition, neutral evolution modifies traits in ways that do not affect fitness. In the context of evolution, "neutral" means "NOT adaptive".

Evolution of some form is inevitable whenever there is variation both in reproductive success and in heritable traits.  

Neutral evolution is evolutionary change which is NOT based on fitness differences.

Like fitness (below), neutrality is a relative term.  Traits are not neutral; trait variation is neutral, if the trait variants do not differ from one another in reproductive success.  Evolution is neutral if the change from one variant to another is not caused by a statistically significant difference in reproductive success.

NOTE:  The word "trait" may refer EITHER to some characteristic feature that is variable (e.g., "eye-color") OR to a particular state or condition of a feature (e.g., "blue-eyed trait" vs. "brown-eyed trait").  I usually use "trait" to refer to a feature, while I use a term like "trait variant" or "character state" to refer to a particular condition or measurement of that feature.  The term "trait" may also be used to refer either to phenotype or genotype.  In most cases, context is adequate to determine the intended usage.  

Since causation is associated with correlation (effects should be correlated with their causes), we conclude that adaptive evolution implies a correlation between some particular heritable variation and variation in reproductive success, as explained in Chapter 1 (pp. 9-10).  Absence of such correlation implies that trait variation does not affect reproductive variation, and hence that evolutionary change in the trait is neutral.  Hence our textbook defines the difference between adaptive evolution and neutral evolution on the basis of presence or absence of correlation between heritable variation and variation in reproductive success.  

It should be understood that this definition refers to a statistically significant correlation, preferably one which is predictable, at least in principle (i.e., one based on a causal relationship between the qualities of a trait and the effects of that trait on reproductive success, whether or not the actual mechanism has been demonstrated). 

This somewhat picky point is similar to the distinction made in probability, between the abstract or ideal expectation of 50 per cent heads in a series of "fair" coin flips and the realized probability in some actual series of flips, which usually differs from 50:50 by a statistically-insignificant amount.  In the process called genetic drift, the fact of drift is itself a realized correlation between some heritable variation and reproductive success.  But that correlation occurred by chance; the heritable variation did not cause the variation in reproductive success.  Although drift and other aspects of neutral may have important and biologically significant consequences, that significance should be clearly distinguished from the measure of statistical significance by which we recognize the difference between true correlation and chance correlation. 

In adaptive evolution, causation flows from heritable variation to variation in reproductive success to change in the heritable trait value.  In neutral evolution, causation flows from chance association between reproductive success and heritable variation to change in heritable trait value.  This is most easily understood in the case of genetic drift.

304 index page

Randomness.

The mathematics of probability yields some strange and often counter-intuitive expectations.  The general difficulty most people have in understanding probability is powerfully demonstrated by huge profits earned by casinos and other gambling establishments.  The following are some general points with some significance for understanding stochastic processes in evolution.

304 index page

Comments on Chapter 3

304 index page


Fitness defined

Chapter 3 introduces the word fitness as a substitute for "reproductive success".  This makes for more concise word use, and we shall adopt this use.

However, care is in order.  Many words in evolutionary biology, such as "adaptation" and even "evolution" itself, are typically used rather casually, often with misplaced confidence that the intended meaning will be apparent from context.  However, the word fitness is often used with a very precise, mathematically-defined meaning, where it is often designated by the quantitative variable w.  (Sewall Wright, who introduced w as a designation for fitness, is said to have explained that w stands for "worth".)  

Fitness is often easier to define and apply in a mathematical model than it is to measure or demonstrate in an actual population.  And there are many distinct mathematical definitions for fitness.  

Whenever you encounter the term fitness, think "relative fitness", think differences in reproductive success.

Like reproductive success, fitness is best understood as a relative term.  That is, it only makes sense to measure the fitness of each individual or genetic variant in relation to some standard.  A standard fitness value of 1 is commonly assigned as a measure either for the average reproductive success of one prevailing variant (the "wild-type") or else for the average reproductive success of an entire population.

The standard value of 1 for average fitness is not assigned arbitrarily.  It is important to appreciate the reason:  Over the long term, the average reproductive output of each individual in a stable population must necessarily be 1.  If it were greater than one, the population would increase indefinitely (and overrun the world).  If it were less than one, the population would shrink until it eventually vanished into extinction.

Fitness is also context-dependent.  Thus measurements of fitness only make sense for a particular population with a particular gene-pool, existing at a particular time within a particular set of environmental conditions.

This is the real reason why it makes little sense to speak of adaptive evolution leading to progress.  Even though selection always, inevitably, tends to preserve the more-fit among different available variants, there is no context-independent quality which "advances" or "improves".  (As noted just above, any population that survives over the long haul must have an average fitness of 1.)

An emphasis on context-dependence of fitness is important for clear thinking.  Here's another way to express the concept.  In spite of everyday usage, traits do not cause fitness.  Rather, in some specified context, differences among traits are responsible for relative differences in fitness.  

There's an analogous concept in genetics.  There is no such thing as a gene which causes a trait.  Rather, in some specified genetic and environmental context, differences between genes are responsible for trait differences.

304 index page

Characteristics of neutral variation

After defining fitness (see above), the textbook presents an example of experimental evolution in E. coli.  This example is intended to emphasize the reality of neutral genetic variation.  Such neutral genetic variation can take two forms:

Two significant points should be noted:

How can genetic variation be neutral?

Molecular biology provides the basis for understanding several forms of neutral variation, which may involve any of the following.

Some DNA variation that is expressed as phenotypic variation may nevertheless be neutral if it is not be correlated with reproductive success.  Like molecular variation that does not affect phenotype, phenotypic variation is also neutral if it does not affect fitness.

What are the causes of random evolutionary change?

Neutral evolution involves genetic variation that is not correlated with fitness.  With no such correlation, such evolutionary change is often described as random.  

Note that the word "random" is hard to define satisfactorily.  

Mutation introduces new alleles at random.  

When used to describe mutation, "random" means that the origin of an allele is unrelated to its particular effect on fitness.  But note that mutations are certainly NOT random in the sense of "every possible mutation is equally probable".  Some specific types of mutation are much more likely than others.  Some gene and chromosomal locations are much more prone to mutation than others.  And quite a few genetic mechanisms function in ways which influence the type and location of mutations (for example, by concentrating variation in certain regions such as the genes which create antibodies).  The "randomness" of mutation means simply that no known mechanism is capable of causing specifically advantageous mutations.

Mutations come in many varieties.  These include "point mutations" affecting individual nucleotides; insertions, deletions, inversions, translocations, and duplications of various sizes (ranging from a few nucleotides to significant pieces of chromosomes); and wholesale duplication of the entire genome (leading to polyploidy).  Evidence for gene duplication is provided by the existence of multigene families.

This is not the place to discuss the evolutionary significance of various types of mutation.  But it should be noted that certain mutational variants may be neutral at the time of their origin and yet nevertheless become profoundly significant for subsequent adaptive evolution.  This is especially evident in the case of gene duplications.

Once mutations are introduced into a population, various mechanisms determine whether or not the new alleles will increase in frequency.  

Random processes in population genetics are all of those processes by which alleles vary in reproductive success in ways uncorrelated with fitness.  

304 index page

Genetic drift

"Genetic drift" refers to the net accumulation of neutral genetic change due to statistical fluctuations in allele frequencies.  This change due to drift has two principal effects, reducing genetic variation within any single population and accumulating genetic variation among related populations.  

Genetic drift has some considerable significance for evolutionary biology, especially for understanding and interpreting variation which may be observed at the molecular level.  

See summary of Key Points about Genetic Drift.

For more extensive discussion, see genetic drift and neutral evolution.

Notes for chapter 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17

304 index page

Comments and questions: dgking@siu.edu
Department of Zoology e-mail: zoology@zoology.siu.edu
Comments and questions related to web server: webmaster@science.siu.edu


SIUC / College of Science / Zoology / Faculty / David King / ZOOL 304
URL: http://www.science.siu.edu/zoology/king/304/ch03.htm
Last updated:  28 January 2005 / dgk