Teaching Evolution with Models

In 2005 only a quarter of the U.S. adult population subscribed to the idea that modern-day organisms evolved through “natural causes.”¹ Some people, to be sure, believe in a literal interpretation of the Bible, but for many more it is simply not conceivable that the extraordinary complexity and interdependence we observe in living things could be anything other than the result of intentional design. To most of us it is quite literally incredible that random change, accompanied by variance in reproductive fitness ascribable to inherited traits, can produce the same outcome without intentionality and with no external intervention.

Yet a simple computer model can demonstrate how evolution occurs. Imagine a simplified model of a plant that needs only one thing—water. If just enough water is present the plant will grow and produce seeds, which will germinate and eventually produce other plants, which in turn will make their own seeds and offspring. If there is a bit too much or too little water, the plant will be sickly and produce fewer seeds or no seeds at all. In extreme cases—if the plant gets much too much or far too little water—it will die without producing any offspring at all.

While this model is easy to build on a computer, it has one major flaw—it is unstable. If the birth rate exceeds the death rate on average, the number of plants will grow without limit; if the inequality is reversed, they will die off. Luckily, the problem is easy to fix. We need to ensure that fewer plants grow to maturity when they are “overcrowded”—which is, of course, what happens in nature.

Where’s the evolution?

So far, so good, but where does evolution come in? Imagine that our model plants come in different varieties. Some are adapted to more water, some to less. For simplicity, let’s assume that there are 10 varieties of plant. Level 1 is adapted to live in very dry climates, and level 10 in very wet ones. Note that the two extreme varieties are likely to look quite different. For instance, a level 1 might have a very long taproot like a dandelion, while a level 10 might have shallow roots that spread out laterally. But nearby varieties don’t differ much at all. A level 4 plant looks almost the same as a level 5.

Now we know that offspring don’t always look exactly like their parents; even the littermates of purebred dogs show some variation. How do we add this important feature of the real world into our model? Easy! Imagine that when a plant produces seeds the offspring sometimes shift levels. So a level 5 plant, for instance, will mostly produce level 5 offspring, but every once in a while it will make a level 4 or a level 6 plant. Assuming that the original plant was in a level 5 environment—that is, that it was getting just the right amount of water—its level 4 and level 6 offspring will be slightly disadvantaged. They will probably die young and have fewer seeds on average, and, therefore, fewer offspring. And if they have offspring that are still more maladapted to the environment—level 3 or level 7, say—those will never go to seed and will produce no offspring at all. So the population of plants will stay more or less at level 5, with the occasional 4 or 6, which may well go unnoticed since they look so much like the level 5’s. Still no evolution? Wait, we’re getting there. Just one little thing to add to the model… Environmental pressure

Environments are not eternal. Weather patterns change. Some rivers dry up, others flood their banks. Ponds fill in and become marshes and wetlands and eventually dry land. What will happen in our model if we vary the environment? Try this mental exercise. To that computer model in your head add a slider with a range from 1 to 10, corresponding to different amounts of water, each suitable to the different levels of plant. At the outset the slider is set to 5 and the model starts out with a hardy population of level 5 plants “growing” on your mental screen, with the occasional 4 and 6 mixed in, as described above. What will happen if you move the slider to 10, causing the environment to become a lot wetter? Think about that before reading further.

The answer is that it depends on how fast you move the slider. If you move it too fast, the entire population of plants will die off. But if you move it slowly enough the plant population will have time to adapt and will eventually evolve from level 5 to level 10. Relying only on the fact that the plants aren’t all the same at each generation, and that those that are more adapted to the environment are likely to have more offspring, the population of plants will—over many generations—adapt even to enormous changes in the environment. And it does this entirely through natural processes!

Of course, in a way we cheated, didn’t we? We created all those different levels of plants, specifically designed to thrive in all those different environments, before we even ran the model. It evolved, all right, but it evolved into something that was there to begin with! Point well taken. The fact that our model evolves doesn’t prove evolution by natural selection, it simply illustrates it. And that, with support from the National Science Foundation, is what we have set out to do.

Evolution Readiness

In a recent project called Evolution Readiness, we are building the model described above, among others. Starting in the fall, we plan to try it with fourth grade students at schools in Massachusetts, Missouri, and Texas. For example, we will challenge the students to make a plant population evolve by gradually altering its environment, and we will monitor their actions to see what they do. Since the roots of the plants are not visible “in the wild,” the student must move a plant into a virtual laboratory in order to see its roots. If we observe students doing this spontaneously, we can reliably infer that they are looking for differences between the plants and associating those differences to changes in the environment.

This spring our collaborators at Boston College will present fourth graders at each participating school with an assessment designed to measure their understanding of evolution and some associated “big ideas,” such as natural selection. Next year and in 2011, we will compare this baseline data to the performance of students taught by the same teachers, who have used our models and curricular materials. This summer we will host a three-day workshop for a dozen teachers plus administrators and technical supervisors from the three participating schools. We will keep in touch with the teachers through an online course.

The Evolution Readiness project is a challenging one. Although the individual concepts we cover— adaptation, variability, and inheritance—are included in state and national science standards for the elementary grades, they are rarely integrated and presented as a mechanism for evolution. We will be working very closely with participating teachers over the next two years to achieve that goal.

¹ “Public Divided on Origins of Life,” Pew Forum on Religion and Public Life (2005).