The Buzz on Honeybee Cognition and Memory

I'm currently working on a paper regarding learning and memory in honeybees. Actually I've been working on it for about 3 years. It got put on the backburner when I entered graduate school, but I finally decided its time to get this thing published. So, why not give you a little preview of the remix?

The paper's topic is a common and well-studied phenomenon in humans, the serial position effect. In the 1960s, Sperling conducted studies on human sensory storage, demonstrating that people have an extremely accurate memory for visual stimulation although the duration is brief. Essentially, people have a very detailed short term memory, the accuracy of which decays rapidly. For example, consider this list of words:

Hippopotamus
Escalade
Buffer
Photocopy
Membrane
Parrot
Device
Headphones
Notepad
Stop Sign
Glove
Leg
Email
Tupperware
Soda
Needle
Cardboard
Medicine
Car
Tea
Wave
Cheese
Studio

OK, now that you've read through the list a couple times, go below the fold (and cover up the list with your hand or some paper. Don't cheat, or you'll miss the effect!)

You're covering the list right?

OK, now, what items from the list can you remember with the best accuracy? Write them down, as many as you can remember. Now that you wrote them, look at the list again. How many did you remember, and did you notice a pattern?

In humans, the serial position effect refers to the tendency for people to remember the first and last items presented in a list. Items presented first (the primacy effect) and last (the recency effect) tend to be remembered with greater frequency than those in the middle, therefore an item's position in a sequence affects its ability to be recalled. The reasoning behind this has to do with the properties of memory storage: short term storage (STS) and long term storage (LTS). New data must spend some time in STS before it can be encoding permanently in LTS, and the capacity of STS is quite limited while LTS is considered nearly infinite.

When personal telephone numbers became common, the government and companies quickly learned to limit telephone numbers to seven digits. This is due to the fact that studies were showing that the limit of human short term memory was seven items, plus or minus two. When a person begins hearing or seeing a list of items to remember, short term memory is mostly empty and more cognitive resources can be spent memorizing the first items which enter STS. There will exist very little interference for these items to pass from STS to LTS. But as the STS begins to be loaded to capacity (7+/-2) and the items keep on coming, interference begins to inhibit these items from being transfered to LTS. The items at the end are remembered well, since this interference ends and the items have less "competition" for the resources needed to shuttle them into LTS. This data is "fresh in your mind," so to speak.

This results in a curvilinear representation of recall data, where the percent recalled is greater at the beginning and end.

So what does this have to do with honeybees? Animal studies have demonstrated the serial position effect in a variety of non-human animals (moneys, rats, and birds). Yet no one had explored it in a invertebrate species, such as honeybees. This naturally comes from the assumption that as honeybees are much more neurologically simpler, complex memory phenomena are beyond their grasp. But consider how honeybees must find nectar and pollen in their environment, and remember it well enough (by the position of the sun) to relay this memory to other members of the hive. They "dance" out the location of the food source, other bees take notice and encode this data, and go directly to the food source. Its quite amazing.

So, as honeybees are able to encode and remember complex positional data, could they also demonstrate the serial position effect? We tested this by capturing a bee (in a matchbox) and placing a small drop of non-toxic red paint on its back to identify it. We then took it to an enclosed experimental apparatus with five targets, each had a drop of sucrose (exact same amount). We placed it on a target (which was considered the first) and forced the choice of the next target by removing its cover one at a time. The sequence it visited the targets was recorded. We then allowed it to fly back to the nest from the apparatus, so it could encode the positional data and store its sucrose. It was only a matter of time before the honeybee returned to the apparatus (which had another drop of sucrose on each target), and then we recorded the sequence which the bee visited the targets. The results (of 25 bees record this way) showed a clear preference for the 1st and 5th targets (first and last) it was trained on. The paper in progress will explore more details of the experimental design (don't want to give away too much!) but it is exciting that the effect is reproduced in honeybees. This seems to suggest that memory is quite similar across species and that highly organized and complex neurological structures aren't necessarily required for memory.