I had an idea just now. I remember, in the past, when I was younger, hearing stuff about how squirrels bury lots of acorns, etc., but only dig up a few, and presumably forget where many of them are buried. I've devised an experiment.
Let's say that we have a population of squirrels that typically digs up 30% of its buried nuts. The experiment would be to dig up half of a given squirrel's buried nuts when it's not looking (do squirrels hibernate?).
We fully expect the squirrel to dig up some spots where it originally buried nuts, and find them missing.
If the squirrel successfully digs up only 30% of the remaining nuts, then it probably only remembered 30% of the original.
If the squirrel successfully digs up 60% of the remaining nuts (or twice whatever its original percent was), then it probably remembered all locations, but only needed a certain number of buried nuts.
Experiments about this have probably already happened, but I just happened to think of it, and I haven't been blogging enough!