Pool & room
Room cues: where to put them and what to avoid
Your animal is reading the room. Make sure it reads what you meant.
A rat in the water maze is a little navigator, taking bearings from whatever it can see. The original notes asked for "at least 3 different cues, e.g. 3 different posters, (so it can 'triangulate'), always visible from inside the pool", placed well outside the pool. That's the easy part. The trickier part is the cues you never meant to provide: the camera, a lamp, a door that opens, or you, standing in the same place every time.
: If the cues are close to the poolside, the rat could be associating the cue with the platform rather than mapping the room.
Questions people ask
Where should I put the distal cues, and why?
Put several large, distinct cues on the walls around the maze, high enough to be seen over the pool wall, and never move them during a study. Animals use them to work out where the platform or escape hole is. Make the cue set asymmetric so that no two directions look the same. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Room cues mainly give direction, while the apparatus gives distance [Hamilton 2007]. Use plenty of fixed cues [Vorhees & Williams 2006] that are prominent on the walls [Vorhees & Williams 2024]. On the Barnes maze, cues are placed 50–200 cm away and not directly over the escape hole [Gawel 2019]. HVS suggests cues at about twice the pool diameter from the centre, each subtending about 5° [HVS].
Going deeper
The 2× diameter and 5° rule is HVS practice, not a tested standard. Asymmetry is our inference from the need for distinctive cues.
How big and what colour should cues be?
Use big, bold, high-contrast shapes (black and white work best). Rodents see detail poorly and do not see red, so colour differences that you notice may not help them. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Acuity is about 0.5 cycles/degree in C57BL/6 mice, about 1 in pigmented rats [Prusky 2000] and about 0.5 in albino rats [Prusky 2002]. Rodents are dichromats with UV and middle-wavelength cones and no red cone [Peirson 2018; Jacobs 2001].
Going deeper
Illustrative: at 0.5 cycles/degree the finest resolvable stripe pair spans 2°, about 7 cm at 2 m. Make features well above this size.
What uncontrolled cues could my animals be using?
Your animals could be using anything stable that they can see, hear or smell, such as doors, shelves, lights, the camera, curtains, heaters, cages, the experimenter and even a noisy vent. These are fine as long as they never move. Walk round the room at the animal's eye level and list them. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Objects in the room act as cues [Gawel 2019]; plenty of cues that are never moved [Vorhees & Williams 2006]. Odour cues are used on dry mazes, and no odour sources (such as bins) should be in the room [Gawel 2019]. Noise sources matter too [Home Office].
Going deeper
The walk-round checklist and the airflow/sound-as-cue points are our suggested practice, not tested findings.
How do I design my experiment so I don't create accidental cues?
You avoid accidental cues by keeping everything except the intended cues the same on every trial. 1) Use pseudo-random, balanced start positions. 2) Stand in the same place every trial, hidden if possible, in neutral clothing. 3) Keep the room exactly the same each day. 4) Use the same cleaner. 5) Randomise and blind where you can. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Use two cardinal and two ordinal starts so start-to-platform distances are similar, with two starts where the platform is to the left and two to the right; use an unused start for the probe [Vorhees & Williams 2024]. A visible experimenter should wear something neutral [Gawel 2019]. Experimenter sex and odour can affect rodents [Sorge 2014]. Report randomisation and blinding [Percie du Sert 2020].
Going deeper
Check the platform: no shadow, ripple or visible edge from the animal's viewpoint (our suggested check; shadows are not covered by our sources).
Should there be marks or cues inside the pool?
Normally no. Wipe off marks on the inside wall above the water before each session, unless you are deliberately testing proximal cues. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Proximal cues change the type of memory tested [HVS]. Practice is mixed: one protocol allows cues on the pool interior above the water [Bromley-Brits 2011]. On the Barnes maze, an intra-maze wall promoted a serial (non-spatial) strategy [Gawel 2019].
Going deeper
Decide, keep it constant and report it. A seam or scratch can act as a proximal cue: use a seam-free interior [Vorhees & Williams 2024].
Can the tracking camera or lights become a cue?
They can if they change. Keep the camera, PC and lights in the same place and state for every trial, and cover any indicator LEDs the animal could see. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Infrared at 940 nm is used because animals cannot see it while the camera can [Vorhees & Williams 2024]; keep the PC out of sight [HVS].
Going deeper
The LED point is an inference, not a published finding.
Where should I place the cues?
For spatial learning place large, three-dimensional cues far from the pool on room walls or fixed structures, visible to the animals. 1) Avoid cues on or inside the pool wall. 2) Keep the arrangement consistent within a study. 3) Moving cues won't affect tracking unless they cause reflections.
More detail
Place large, distinct cues on the walls of the room around the pool (distal cues), keep them fixed for the whole study, and keep cues off the pool itself. Distal extramaze cues approximate optical infinity and support allocentric, hippocampal-dependent navigation. Pool-wall cues act as local landmarks, which can compensate for impairments and alter learning rates, deficits and comparability. Remove cues stuck inside the pool. Build in a comparison if you change cue arrangements.
Going deeper
In VR, distal cues can be near the pool or large and far away; proximal cues can be moved in forced-strategy trials.
What colours can rats and mice see?
Rats and mice have two-colour vision: they see ultraviolet and green-blue, but not red. They have no red cone, so red looks dim to them, and they can see UV that we cannot. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Rats and mice are dichromats: they have ultraviolet and middle-wavelength (green) cones but no red cone, and their vision is dominated by rods. Mouse rod about 498 nm, M cone about 508 nm, UV cone about 360 nm; 97% rods [Peirson 2018]. Rats have UV and M cones and dichromatic colour vision [Jacobs 2001]. Mice are about 12 times less sensitive than humans at 600 nm [Peirson 2018].
Going deeper
Lux ignores UV and is human-weighted, so lamps with UV output may look brighter to rodents than your meter shows [Peirson 2018].
Should I use coloured cues?
Use black and white or strongly contrasting shapes instead. Red and green of similar brightness may look alike to rodents. Your local ethics committee, licence conditions and vet always take precedence.
More detail
Use cues that differ in brightness, size and shape rather than colour, because rodents see colour poorly and red hardly at all. Rodents lack a red cone [Peirson 2018; Jacobs 2001] and have low acuity [Prusky 2000].
Going deeper
We found no maze test of red/green cue confusion; this follows from the photoreceptor data.
Spotted a mistake, or know a better source? Tell us and we'll fix it and say so on our corrections page.