Search

Evound

Evound

Teaching
Textbook
Software
Videos
Posts
UW Website

Light Dark Automatic

Posts

SimBio Workshop

Presentation Deck
Darwinian Snails Assignment
Pig Check Assignment
Mendelian Pigs Assignment

Home
Posts
SimBio Workshop
Presentation Deck

Presentation Deck

Click on a slide to zoom in and/or read the caption.

Thanks for visiting.

SimBio has virtual labs on a variety of topics in evolution, including natural selection, population genetics, and tree thinking. In this presentation, I will discuss how I use two of the labs as platforms for group independent research projects in a non-major's course at the University of Washington.

SimBio’s evolution line gives students the opportunity to learn evolution by doing virtual experiments. Because they are built to support open-ended investigation, many of SimBio's virtual labs provide opportunities to go beyond the questions covered in the formal instructions.

First, I'll discuss how I use the Darwinian Snails lab to support independent research. Note that this lab is available in two versions---an older workbook-style version and a newer tutorial version, with onscreen instructions and questions. I love the tutorial version, but for this class I use the workbook version---because it includes a more open-ended experimental sandbox. The lab is inspired by Robin Seeley's research on an intertidal snail called the flat periwinkle. Seeley found that after decades of exposure to predation by invasive crabs, the snails had thicker shells with lower spires. She suspect that the populations had evolved by natural selection.

As with all of SimBio's virtual evolution labs, the Snails lab is build around an individual-based model of an evolving population. Because it's a model, not merely an animation, something different happens every time you run it. If you haven't seen the Darwinian Snails lab, I encourage you to download SimBio's demo package---which includes fully-functional versions of everything they make--and play with it yourself. This screenshot is from the first section, where students get to play crab and try to keep themselves fed by eating snails. Thinner-shelled snails require fewer clicks to eat. Once they've seen the population evolve across generations, they can change the parameters of the model (using the checkboxes) to learn the conditions under which the population does or does not exhibit adaptive evolution.

The last section of the workbook version lets students design and run their own experiments. This feature was inspired by Geoffrey Trussell's work. Trussell realized that there is an alternative explanation for the changes in shell thickness that Seeley documented. Perhaps individual snails can smell crabs, or other snails that have been injured by crabs, and grow thicker shells as an inducible defense. This section challenges students to collect data to distinguish between the hypotheses that shell thickness changed via evolution by natural selection versus phenotypic plasticity. Students have two source populations---a thin-shelled one unexposed to crab and a thick-shelled one long exposed to predation. They have four experimental tanks. They can move snails among the source population and the tanks. And they can add ordinary crabs or crabs whose claws are bound with rubber bands---so that the snails can smell the crabs, but are not actually at risk of being eaten.

This screenshot shows one experiment students might devise. I've stocked two experimental tanks with thin-shelled snails and two with thick-shelled snails. One tank in each pair also has crabs with bound claws. The histograms show the distribution of shell thickness in each population before I run the simulation.

This screenshot shows the same populations after I've run the simulation for for about three snail generations. Notice that the snails in Tanks 1 and 2, whose ancestors were thin-shelled, have thinner shells versus the snails in Tanks 3 and 4, whose ancestors were thick-shelled. This indicates that the two source populations are genetically distinct, and is consistent with the evolution-by-natural-selection hypothesis. Also notice that within each pair, the tank with (bound) crabs has thicker shells, on average, than the tank without crabs. This is consistent with the phenotypic-plasticity hypothesis.

The results from my experiment match what Trussell found with real flat periwinkles.

After my students had completed the exercises in the Darwinian Snails workbook---including designing experiments to distinguish between evolution versus phenotypic plasticity---I put them in groups for an additional challenge. Note that you can see the full text of the assignment under Darwinian Snails Assignment in the workshop menu.

I gave them an example...

... including an experimental design...

... a predicted outcome if my guess is right...

... and a predicted outcome if my guess is wrong.

Here are the results from a single replicate of my experiment.

I pointed out that I expected them to do more than one replicate and aggregate the results.

I also gave them additional suggestions for questions they might address. See the next four slides for more details on how to address Questions 2 through 4.

The trick for setting up pairwise matings is to make sure the snails are touching before running the simulation. This encourages them to mate.

This option has proven fairly popular among my students. At least at first glance, it appears require relatively little work to complete a lot of replicates.

This one demands a pair bit of effort to just run the experiment. But I've had at least one group do a nice job with it.

This may be the most conceptually abstract of my suggested questions.

Upon completing their experiments, I asked my students to prepare formal written reports. (I've used real student results with permission in this presentation.)

This group did a nice job estimating the heritability of shell thickness.

This group did a nice job showing that diversity increases with the number of founders. And they thought carefully about graph designs that would show a more complete picture of the data.

This group included at least one member who knew how to use R to produce elegant graphs.

They and I were surprised when the data did not appear, at first glance, to be consistent with the hypothesis. On closer examination of the method they used to analyze their data, we realized that they had inadvertently turned, for example, their ten replicates with 5 founders into a single replicate with 50 founders. In retrospect, while the question this group took on appears straightforward, the analysis is anything but. Variation in the amount of variation is an abstraction of an abstraction. And the teachable moments don't always arise where I expect them!

The second independent research assignment I'll discuss uses SimBio's Mendelian Pigs lab.

The lab is inspired by a true story about coat color in pigs. There are at least 4 alleles of the MC1R gene, with interesting allelic relationships. Note, for example, that allele _W_ is recessive to _B_, dominant to _R,_ and codominant (or incompletely dominant?) with _S._

The lab itself is designed to review and reinforce what the students know about Mendelian Genetics, and then help them start scaling up---from thinking about genotypes and phenotypes among the offspring from a single cross to thinking about allele, genotype, and phenotype frequencies in whole populations. And how these frequencies change across generations. The last section is an open-ended sandbox. The students have four pure-breeding stock populations (each fixed for one of the alleles) and an experimental field they can populate with founders, from the stock pens, in any combinations they chose. They can also add wolves that prefer particular phenotypes to varying degrees. This allows for a wide variety of selective scenarios.

For several years, I've been giving my students a supplementary assignment that challenges them to check whether the virtual pig population in the individual-based model evolves as predicted by an analytical model using Hardy-Weinberg arithmetic. The full text of this assignment, and the software tools needed to complete it---are available under Pig Check Assignment in the workshop menu.

Here's one of the exercises in the Pig Check Assignment. It involves selection against a recessive allele---which should result in a decrease in the allele's frequency.

The green line in this graph shows the analytical prediction. The orange dots show data from 20 students. As long as the students set up the pig simulation and the model parameters correctly, the analytical predictions are gratifyingly accurate.

After they have completed both the Mendelian Pigs set lab and the Pig Check assignment, I put my students into groups for another independent research challenge. Here are five suggested questions. The fifth involves a consistent phenomenon that was puzzling to me---and is a case where I feel like I learned something about population genetics from my own homework assignment! You can find the entire text of the assignment, and the software tools needed, under Mendelian Pigs Assignment in the workshop menu.

This group did a nice job exploring what happens when heterozygotes have higher fitness than either homozygote.

This group did a nice job with a more challenging situation---in which heterozygotes have _lower_ fitness than either homozygote. They found that across a range of parameter combinations the analytic predictions were generally accurate. But they found two rather puzzling cases---marked with blue asterisks here and on the next slide---in which the virtual pigs behaved rather differently than the analytical model led them to expect.

Reading their formal report revealed that the group came close to discovering the challenging notion of an unstable equilibrium. And gave me deeper insight into their thinking than I'd otherwise have gotten.

Thanks for reading through the presentation. Please get in touch if you have questions or suggestions.

And please get in touch with the good folks at SimBio if you have questions about using their materials with your students.

Next

Darwinian Snails GIRP

Last updated on Jun 23, 2021

© 2021 Jon C. Herron

Published with Wowchemy — the free, open source website builder that empowers creators.

Cite