[_ Old Earth _] Is natural selection without a doubt true?

[coelacanth]

Dave Slayer, was that helpful?

BTW, I hope your name is not meant to imply that you like to kill people named "Dave", in which case I feel I must inform you that my name is most definitely NOT Dave ;)

 

[Barbarian]

I am interested, however, in finding out what your simulation is. Please go on

OK, two six-sided dice, different colors. They will be used to determine mutations and what the mutations are.

First simulation: Random change, no natural selection.
On a sheet of graph paper, draw twelve columns of boxes, six squares per column. Put "5" in every box.

Pick a simple rule for fitness. (for example, the higher the sum of the boxes, the higher the fitness)

1. compute fitness of each column (an individual in the population)

2. compute average fitness of the population and record it.

3. For each column, roll the two dice. One specifies which box will mutate, the other specifies the number you put in.

4. When finished, again calculate and record average fitness.

5. Randomly pick two columns (the survivors) and have them reproduce.
a. their offspring will be identical with one mutation, determined by the dice
b. six offspring each.

6. repeat from 4. for at least 12 steps. Graph fitness over time.

Now, do the same thing, except pick the two columns with the highest fitness each time.

Compare the two graphs.

 

[coelacanth]

I like that set-up, Barbarian. Where did that simulation come from?

 

[Barbarian]

I teach biology. My invention. If it flies well, I ask students to take the natural selection graph, and continue it, but with slightly changed rules. It nicely simulates how a change in environment causes evolution, and how a static environment favors stabilizing selection.

I'm sure you can see how two different fitness tests run at the same time, could be used to show disruptive selection. Two different populations will evolve.

 

[coelacanth]

I teach biology. My invention. If it flies well, I ask students to take the natural selection graph, and continue it, but with slightly changed rules. It nicely simulates how a change in environment causes evolution, and how a static environment favors stabilizing selection.

I'm sure you can see how two different fitness tests run at the same time, could be used to show disruptive selection. Two different populations will evolve.

Kudos! Do the students like it?

For disruptive selection would you just set 1 and 6 as both being most fit, and least fit being 3 and 4?

 

[Barbarian]

It's an interesting exercise. The bright and questioning ones love it. Some think it's a lot of work.

For disruptive selection, for example, one test would be "add a point only for even numbers" and the other "add a point for every odd number."

Then graph the Y axis as the number of odd numbers. Using a scatterplot showing all the columns, students can see the population diverge into two separate ones. (each time you pick the top 2 for both categories)

At the point where there is no overlap between populations, you have speciation.

Edit: On reflection, I can see that your rule would work, too.