linking back to brembs.net






My lab:
lab.png

We were discussing the different learning experiments Drosophila had been subjected to the other day. There is rather little on color learning and most of that is not controlling for intensity differences the colors may have for the fly. One of the questions that were raised was whether Drosophila had ever been conditioned clasically at the torque meter. That is, the entire surrounding of the fly illuminated in one color (say green) for three seconds and the aversive heat switched on, followed by three seconds of another color (say blue) and the heat switched off during training. To test if the fly had learned to avoid one color (or prefer the other), the fly was allowed to control the vertical position of four identical stripes in its environment with it's yaw torque, such that flying towards one pair of stripes illuminated the environment green and towards the other resulted in blue illumination. Just as shown at time 2:35 of this video:


In fact, these experiments had been done, but never published. I did them in November and December of 1998, during my PhD thesis. Because there might be someone around who would liek to know if the flies can do this, I'm publishing the results here. The performance of the flies is measured in 'Performance Indices' which are +1 if the fly chose only the 'cold' color and -1 if it chose only the 'hot' color. The PI is zero if the fly did not show any preference. Note that during training, the fly's behavior has no effect and thus there are no performance indices in this phase of the experiment. It is also important to point out that the stripes were slowly rotated around the fly during training, such that the switch between color and heat on/off would occur exactly between two stripes, just as in the closed-loop test afterwards.

classcollearn_small.png

These experiments were performed in parallel with the - at the time - already published experiments using visual patterns (T-shaped), to see if there was any difference in the performance of the flies.

classpatlearn_small.png

As you can see, the performance hardly differed between the two groups at all. Drosophila can learn to distinguish these colors fine, even using classical conditioning. However, also in these colors, differences in intensity were not controlled. In fact, we know that these two colors used here also differ in intensity.
Posted on Thursday 27 October 2011 - 19:46:07 comment: 0
{TAGS}

Render time: 0.0621 sec, 0.0057 of that for queries.