I've skipped blogging about the last two preparations in my seemingly endless quest towards establishing operant learning in isolated leech ganglia, because they both were very frustrating: in contrast to the previous preps, I was not able to get any of the preps to generate any crawling at all in the first ten trials. One was an experimental preparation, where I stimulated the anterior ganglon with every crawling in the following ten trials and the other one was an unstimulated control.

So after just a few weeks, the data seem to suggest that contingent stimulation of the anterior ganglion, irrespective of whether crawling or swimming was the chosen pattern, increases the probability of swimming patterns compared to control preparations:



exp - experimental; cont - controls; pt - pre-test; tr - training; te - test. Each of the three phases lasted ten trials with a 3 minute inter-trial-interval. N is about 6 in each group.

Below are the graphs for the four different groups which got pooled in the graph above, abbreviations as above. All bars are means of the trials where swimming or crawling patterns could be unambiguously identified. Whiskers depict standard errors of the mean. N is about three in each group.



I think there's some interesting plasticity in the system, but so far I can't really figure out what sort. It seems to be dependent on contingent stimulation, but it's not obviously rewarding or punishing. Rather, the continget stimulation seems to tell the animal to swim, no matter what it did when it received the instruction, as long as it did something rather than nothing.

I guess I'll use my last week here to just fill up these data, unless I can come up with some more exciting experiments until tomorrow...