Re: Teaching a worm

Dear David,

I assume when you say worm, you mean earthworm? The reason I ask is because I worked with C. elegans, a microscopic, non-parastic roundworm. This is why I got your question. In the world of C. elegans research, there are quite a few people interested in the question of can you teach a worm. Learning paradigms have been designed to try and find mutants that cannot learn etc. It may be possible to adapt these tests for earthworms (I am extrapolating, I don't know much about earthworms). Of course, there are also the classical learning experiments with mice and rats. If you are talking about earthworms, then have you thought of the maze idea used for mice and rats? Build a simple maze and put an attractive (to worms) odors/thing at the end of the maze as a reward. See how long the worms take to go through the maze. Do they go through faster after 10 trials? Have they learned the ins and outs of the maze? You'll have to experiment as to what would be an attractive draw for the worms--maybe a nice, moist patch of soil at the end of the maze? You can plot the time it takes for the worms to finish the maze vs. the number of trials. If the experiment works the plot should be a downwards curve that smooths out. You can also test more than one worm and possibly more than one worm at a time. If you want to get really involved, you can isolate worm(s) that are learning challenged, ie., the ones that took a much longer time to learn than others, or you can isolate superworms that learned very quickly. You must explain however that there is a randomness factor involved. In order to get accurate significant results you would have to test thousands of worms, a second grader may not appreciate that chance is an uncontrollable factor.

This is one of the reasons that C. elegans is a good model system to test learning (always providing that you believe worms can learn). They are small enough that they can be tested en mass (>100) for each trial. In the learning paradigms designed for C. elegans, they are exposed to two types of chemicals, one that they like and one that they don't like (in this case it was a garlic extract). C. elegans is chemical sensitive and if you pair either chemical with a food source, you can "teach" the worms to move towards the food source.

You can try that with the earthworms as well (eventually). Pair a noxious substance with a food source and see if the earthworms can associate the one with the other--like Pavlov's dog. However, the more factors you add to your experiments, the more controls you would have to do. So, I suggest that you start simply with as few "problems" as possible. Try to control for as much as you can ie., pick worms that are healthy, the same age and from the same area in the garden. If soil is the reward you plan to use at the end of the maze, make sure the soil hasn't been contaminated with other 'flavours'. The test surface should be cleaned after each trial (I suggest a short maze)--you don't want the trail of one worm to influence the choices of the next worm. And keep track of the worms. You don't want an already trained worm to slither through the maze as an untrained worm and throw off your results.

I hope this helps. If you want more details as to how the experiments are actually done for C. elegans worms, I'll be happy to try and explain. I am answering this assuming that you will be working with earthworms and although the C. elegans experiments may be adaptable for earthworms, it might be easier for your daughter to try the maze idea first. It is highly visible and might produce usable data quickly.

Yours Truly,
Shirley Chan