The Barnes maze: spatial memory on dry land

Opening

Imagine a large, round table, brightly lit from above. Around its edge, evenly spaced, is a ring of holes, each just big enough for a rat or a mouse to slip through. Look into most of them and there's nothing, just a drop into darkness, or a blind end. But under one, and only one, there's a small dark box. A shelter.

On the walls around the room are shapes and posters. The animal is placed in the middle of the table, under the glare. It wants to get out of the light. It wanders to the edge, pokes its nose into one hole, then another, then another. Eventually, it finds the box and disappears into it.

A day or two later, it walks almost straight to the right hole.

A dry-land maze

This is the Barnes maze, a dry-land test of spatial memory that sits alongside the water maze in labs all over the world. It was designed to ask the same kind of question without asking an animal to swim. How did it come about, why did it catch on, and what are its weak points?

How it works

The maze is a circular platform with holes spaced around the rim. One hole leads to an escape box, sometimes called the goal box or escape cage. The others are false-bottomed, or blind-ending, so the animal can look in but not get through. Under bright overhead light, the open table is mildly unpleasant for a rodent, since rodents generally prefer dark, enclosed spaces. The escape box offers relief.

Around the room are visual cues. Just as in the water maze, those cues are the key. The holes all look the same, so the only reliable way to remember which one leads to safety is to remember where it is in relation to the room.

A typical session runs like this. The animal starts in the centre, sometimes under a small cover that's lifted to begin the trial. It explores until it enters the escape box, or until a time limit is reached, in which case it's gently guided to the box. It stays there briefly, then goes back to its home cage. The table is cleaned, and often rotated between trials, so that the animal can't simply follow a smell trail to the right hole. The escape box stays in the same place relative to the room.

Over a few days, a healthy animal gets faster and makes fewer mistakes. A mistake, in this context, is called an error. It's a poke of the nose into a hole that doesn't contain the box. After training, researchers often run a probe trial, removing the box and seeing where the animal searches. That idea is borrowed directly from the water maze.

Where the Barnes maze came from

The maze is named after its inventor, Carol Barnes. She trained in psychology at the University of California, Riverside, and earned her doctorate at Carleton University in Ottawa in 1977. Her research has focused for decades on how the brain changes with age, and she later became a Regents' Professor at the University of Arizona.

The maze first appeared in a landmark paper in 1979, in the Journal of Comparative and Physiological Psychology. Its title was Memory deficits associated with senescence, a neurophysiological and behavioral study in the rat. Barnes tested thirty-two old rats, aged twenty-eight to thirty-four months, and thirty-two mature adults, aged ten to sixteen months.

The rats were first tested on a circular platform, which favoured the use of spatial cues for its solution. And the old rats showed poorer memory for the rewarded place.

That alone would have been a useful finding. But Barnes also recorded electrical responses from the rats' hippocampus, through electrodes implanted in a pathway called the perforant path and in a region called the fascia dentata. She delivered short bursts of high-frequency stimulation, which strengthen the connections between nerve cells. This strengthening is known as long-term potentiation, and it's widely thought to be one of the ways memories are stored.

After a single session, young and old rats responded alike. After three sessions, the young rats kept the strengthened connections for at least fourteen days. The old rats' responses declined after the first session. And here's the striking part. The amount of synaptic strengthening was statistically correlated with how well the rats performed on the circular platform task, both within and between the age groups.

So the very first Barnes maze study did more than test memory. It linked ageing, a cellular mechanism of memory, and behaviour, all in one set of animals. Barnes also found that the stimulation's after-effects selectively impaired the old rats' spontaneous alternation on a T-maze. Which is a nice reminder of how these classic tests connect to one another.

Why it became popular

So why has the Barnes maze become so popular? A big part of the answer is stress. Swimming is hard work for a rodent, and being placed in water is stressful, especially for mice. A 2014 review by Cheryl Rosenfeld and Sherry Ferguson describes the Barnes maze as considered less stressful than water mazes. The motivation is mild, usually just bright light, and sometimes a noise as well.

That makes it attractive for animals that don't swim well, for older animals, and for mice in general. It's widely used to look for memory problems in genetic models of diseases such as Alzheimer's, and to test whether a drug or a toxic exposure affects learning.

It also produces a rich set of measures. The simplest is latency, the time taken to reach the escape box. Count the errors, and you can draw a learning curve. Add video tracking, and you get the same kinds of measures used in the water maze. Distance travelled, speed, time spent in the correct quadrant, time spent moving or resting.

And then there's strategy. Researchers commonly sort each trial into one of three types. A random search, where the animal checks holes with no obvious pattern. A serial search, where it works its way systematically around the rim, hole after hole. And a direct or spatial search, where it heads straight for the target or very close to it.

That distinction matters. An animal using a serial search will eventually find the box every time, even if it has no idea where the box is. Its latency might improve simply because it gets quicker at running around the edge. Only by looking at the path can you tell whether it's really using spatial memory.

The strategy view has turned up some interesting differences. In one study of degus, a South American rodent, females during training more often used a spatial strategy, while males tended towards serial or random searching. And in females, spatial retention was reported to depend on the phase of the oestrous cycle. Findings like these are one reason researchers are encouraged to test both sexes, and not just males.

Practical details differ between species. Rosenfeld and Ferguson point out that cues outside the maze work well for rats, but smaller wild rodents may need cues on the maze itself, with a visual barrier around it.

Its limits

And what are its limits? The same mildness that makes the Barnes maze gentle can also make it unreliable. If the aversive stimulus is weak, some animals simply don't bother to escape. They'd rather keep exploring the table, wandering past the target hole, sniffing, and coming back to it later. That can make a curious animal look forgetful. It's a well-known criticism of the maze, and a reason why some labs add a sound, or end a trial as soon as the animal reaches the right hole, rather than waiting for it to climb in.

Odour is another issue. Animals leave scent on the table, and the escape box itself has a smell. Cleaning and rotating the table help, but they need to be done carefully and consistently.

There's also a question of time. Rosenfeld and Ferguson note that both Barnes and water mazes can be time-consuming. Typically four to seven trials are needed to detect improved learning and memory, or differences between groups.

And as with other spatial tests, the background strain of mouse matters. Strains differ a lot in natural anxiety and in how they explore, so a difference in performance might come from temperament rather than memory.

Finally, because the animal can explore so freely, scoring by eye is hard. Counting errors means watching every nose-poke into every hole. Automated tracking has made this far more consistent, but it depends on the camera seeing the animal clearly against holes and shadows.

Back to the table

Back to that bright round table, with its ring of identical holes and one dark shelter. A rat walks almost straight to the right one, guided by the posters on the walls. It's the same question the water maze asks, where is safety, but answered on foot. Carol Barnes designed it to study memory in ageing, and decades later, it's still helping researchers tell the difference between remembering a place and simply searching until you find it.

You'll find the papers we mentioned, and a transcript, at water-maze.com.