[bogrollben]

good point. but this means that step 1 has to immediately be appetizing to be of evolutionary benefit. (arguably it could be). It also conflicts a little bit with the observed behavior that the spider is shaking the web, trying to get attention to the bigger spider. If the true benefit was escape, I would think not-attracting attention would be the resultant behavior.

Just my opinion. I guess we don't know for sure. There's probably a way for some biologist to figure that out for certain - would make an interesting postdoc.

[bonobo]

Step 1 doesn't need to be that appetizing if the cost of building the decoy is low. Any benefit, as little as it may be, is already an advantage if it covers that cost.

About the attention-seeking behavior, we can compare it to the same behavior geckos adopt when they find themselves in danger: they start shaking their tails vigorously to attract the attention of the predator to the tail (geckos are known for being able to detach their tails in these cases, and once detached the tail keep shaking vigorously, potentially attracting the predator giving the gecko a chance to escape). In both cases, the spider and the gecko seem to only engage in this behavior once they already have the predator's attention, so trying to escape without shifting the predator's attention to the decoy could actually put the attention on them.

[ceejayoz]

> but this means that step 1 has to immediately be appetizing to be of evolutionary benefit.

It doesn't necessarily have to be appetizing. Any amount of confusion, hesitation, etc. it can cause is potentially beneficial. Even if only one in a thousand predators hesitates or attacks the wrong spot one in a thousand times when it sees a decoy, that's going to have an effect over millions of years.