Lab 4 -- Evolutionary Forces


Population genetics refers to the most basic processes of evolution, the change in frequencies of alleles within populations. Since alleles may have different effects on the outward appearance of an organism, or phenotype, changes in the frequencies of alleles can result in changes in the appearance of populations. Changes in allele frequencies within populations are thought to be responsible for most patterns of evolutionary change, when magnified by the accumulation of time.

Modern evolutionary biologists recognize that a number of forces can alter allele frequencies within populations. These include:

Mutation -- the spontaneous change of one allele into another

Gene Flow -- the influx/outflow of alleles from/to other populations

Genetic Drift -- the alteration of allele frequencies via chance

Selection -- the fact that certain genotypes (combinations of alleles) have a relatively higher chance of survivorship or fecundity than other genotypes, or higher fitness. It is important to remember that fitness is a combined result of the genotype’s phenotypic expression and the environment.

In the absence of these forces, the Hardy-Weinberg theorem demonstrates that allele frequencies should remain at an equilibrium. This set of computer exercises will focus on how mutation, genetic drift, gene flow and selection can result in evolution of allele frequencies. In particular, we will consider what characteristics of populations determine which of these forces has the greatest effect. We will also explore how these forces of evolution can produce differentiation among populations.

About the program -- Populus is a program developed by evolutionary biologists at the University of Minnesota. It is distributed free of charge (ask your instructor if you'd like your own personal copy). The program has been installed on the PC network on campus. Access it by typing X at the first menu, and waiting for the DOS prompt. At the prompt type "Pop34" and <Return> to start the program.

1. Populus is a large set of different computer simulations arranged in menus. You can "navigate" between these menus using the arrow keys to highlight menus and programs and pressing <enter>. Don't try to use the mouse -- it doesn't work for this program. If you want to leave any program or menu press <ESC>, the "escape" key, in the upper left corner of your keyboard. In most cases, the bottom of the screen has instructions for how to proceed.

2. When you first start a simulation a description will appear on the screen. These descriptions are very general to begin with, and become more complex as you read on. Press <PAGE UP> and <PAGE DOWN> to move around in these descriptions. Press <ENTER> to start the simulation.

3. Many simulations allow you to compare the results of the previous run with the current run. After you enter a program, press <F4> to activate this option. Turn this option off by pressing <F4> again.

4. Help screens explaining the options of each simulation or plot can be accessed by pressing <F1> at any time. <F2> will return you to the description of the simulation.

Directions for working through the simulations are in BOLD.



Selection is what we usually think of as the major force causing evolutionary change. Simply put, selection occurs when certain genotypes have higher propensities for survival and/or reproduction than others. These differences are usually defined as fitnesses, the average number of offspring for an individual of each genotype (if an individual doesn't survive, it has zero offspring). In genetic models, fitnesses are usually expressed as relative fitness,

relative fitness = # of offspring for genotype/ # of offspring for most fit genotype


This means that relative fitnesses range between 0 and 1. Consider the following example:

Genotype # of offspring Relative fitness

----------- ---------------- ------------------

AA 18 18/20 = 0.9

Aa 20 20/20 = 1.0

aa 19 16/20 = 0.95

The above is a case of where the heterozygote is higher in fitness than the two homozygotes. What do you predict would happen if both alleles, a and A, were present in a population?

Choose "Selection . . ." in the first menu. Then, choose "Autosomal Selection" and read the first two introductory paragraphs. Press <Enter> to get the input screen. Press F4 to activate the "Plot the last data" option.

The 'default' settings for this simulation illustrate a case in which the heterozygote Aa has a higher fitness than either homozygote, as in the example above.

Press <enter> to start the simulation.

You will be able to view the results for this simulation in three graphs:

(average fitness of the population) vs. p

You can cycle between these graphs by pressing the space bar. Try it.

1. What happens to the allele frequency of A (p) under these conditions of selection?




2. Press the space bar to see the D p vs. p graph. Can you predict what (the equilibrium frequency of A) is under these conditions? Explain why this is a stable or unstable equilibrium from this graph.





3. Press the space bar to see the vs. p graph. What is the relationship between and ?


4. Press <ESC> and alter the relative fitnesses of the two homozygotes, keeping WAa = 1. How does this affect the equilibrium frequency of the A allele ()? (Note that the results from the previous conditions are shown in black). Experiment with the relative fitnesses of the two homozygotes until you can write a verbal description of their relationship to .







5. Enter the following relative fitnesses: WAA = 0.8, WAa = 0.5, and Waa = 1. Describe the result below.






6. Press the space bar to see the D p vs. p graph. What are the equilibrium points? Are they stable equilibria?





7. Press the space bar to see the delta-p vs. p graph. Under these conditions, is always at the maximum value of delta-p? How does this affect how you think about the outcome of natural selection with respect to adaptation?





8. Set WAA = 1, WAa = 1 and Waa = 0.7. Choose the single frequency option and set the initial frequency at 0.01. What is ?


Is the deleterious allele eliminated after 200 generations? Explain why D p is so small when p is close to 1.










Choose "Genetic Drift . . ." from the Main Menu, and the "A Monte Carlo Model." Read the introductory screens. Press <Return> to see the parameter screen.

In the last model, we assumed that populations were very large -- in fact, we assumed that they were infinitely large. That was because populations of finite size are subject to a second force of evolution, genetic drift. In order to understand how genetic drift works, we will again look at how the frequency of alleles change in single populations, but first we will assume that there is no selection acting on these alleles. Looking at such neutral alleles will allow us to see the effects of genetic drift when it alone is acting.

1. Change N = 200 and runtime to 100 generations. Press <Return> to start.

Note that the graph shows how drift affects the frequencies of alleles at 6 independent loci (each a different color of line). Note how frequencies may move up or down each generation. Note also that genetic drift will stop when the frequency of the allele is either 0 or 1. This means that one allele has been lost from the population -- in the absence of mutation, an allele lost from a population cannot be recovered. We refer to allele frequencies as reaching fixation at 0 or 1 -- a particular allele is fixed at frequency of 1, and lost at frequency of 0.

Did allele frequencies show any tendency to move up or down?

Note how many alleles fixed at 0 and 1 -- in either case, an allele is lost and thus there is a decrease in genetic variation.

2. Run the simulation again under the same conditions. Did you get the same answer? Explain why or why not.





3. Run the simulation with N=100. Did you get a different answer? Do the simulation again to see if you get a similar answer.






4. Based on these results, you might hypothesize that the larger population size is, the slower alleles are lost due to genetic drift. Test your hypothesis with an experiment. Run simulations 5 times each for the following population sizes, taking note each time of the number of alleles fixed after 100 generations: 25, 100, 200. Calculate the mean and 95% confidence intervals and see if they overlap for each pair of population sizes. [Remember that the 95% confidence interval for a mean is the range between mean ± (2*standard error)]




Under what conditions in real populations do you think genetic drift would be most important? Do these conditions occur very often?





Consider the following question in light of the results of this simulation: If genetic drift causes a population to become less diverse, how does it affect the degree of similarity between populations? Can this be thought of as a source of diversity?













Choose "Drift and Selection" from the "Drift" menu. Read the introduction. Press <Return> to see the parameter screen.

In real populations, both selection and genetic drift can occur. But what influences whether genetic drift or selection is the stronger force? Does genetic drift ever overwhelm the force of selection? [Note that the strength of selection is seen in the differences between fitnesses, and the strength of drift is determined by the population size.]

1. Change N = 500, wAA = 0.8, wAa = 1.0, waa = 0.8, and # of generations to 100. What do you expect to occur with these genotype fitness? Is your expectation fulfilled?



2. Change N = 250. How is the result different from the last simulation?



3. Change N = 50. How is the result different from the last simulation? Is the allele lost? Repeat the simulation under these conditions several times.


4. Change N = 4. How is the result different from the last simulation? Does the allele frequency reach fixation? Repeat several times.


5. Change N = 500, p = 0.1, wAA = 1, wAa= 1 and waa = 0.9. Do you get a predictable result?



6. Change N = 25 and run several times. What do these simulations suggest can be the effect of small population size on ability of advantageous alleles to sweep to fixation?









Choose "Selection . . ." in the first menu, then "Selection and Mutation" in the second menu.

Turn on "Last data plot" by pressing F4.

This simulation looks at the maintenance of polymorphisms in populations due to a balance between recurrent mutation and selection. The fitness of genotypes is described as follows:

WAA = 1

WAa = 1- hs

Waa = 1-s

[If h = 0, then the A allele is dominant. If h = 1, then the a allele is dominant. If h= 0.5, the heterozygote has intermediate fitness.]


1. Change the default conditions so that h=0 and begins with a single frequency of 0.1. In the absence of mutation, we would expect the a allele to be removed from the population eventually by selection.

What is the effect of increasing the strength of selection against the a allele on its equilibrium frequency (Note that the program plots the frequency of the A allele)? This is easiest to see in the graph relating p vs. D p (Hit the space bar to change graphs).


2. What is the effect of increased rates of mutation on the equilibrium frequency of a?




On to Lab 5 - Animal Behavior

Back to Lab 3 - Mendelian Genetics

Back the BIO 136 Homepage