Although swim speed is not a direct measure of learning, it plays a key role in interpreting behavior in Morris water maze studies. HVS Image systems give both the average speed (calculated by dividing the path length by the latency or trial duration) and the active speed (excluding any moments of immobility or floating, giving a clearer picture of intentional behavior), measured in centimeters per second – giving objective measures that allow comparison between subjects or conditions.

Some of the most commonly used water maze measures (latency, path length and platform crossings) are strongly affected by swim speed, so it’s important not to ignore it, especially if relying largely on these measures rather than the more sophisticated analyses available in the HVS Image system.

Slower swim speeds can help distinguish between cognitive and non-cognitive explanations for poor performance, and can point to issues that are unrelated to learning, such as motor deficits in stroke models or cerebellar lesions.

Key uses include:

• Interpreting measures used in acquisition and probe trials
• Controlling for motor differences
• Controlling for motivational differences
• Detecting sensorimotor or neurological impairments
• Screening for outliers or non-task related behavior
• Excluding trials or subjects that do not meet the criteria for your study (including screening out unsuitable subjects in pre-training sessions).

When interpreting commonly used measures such as latency or path length in learning trials, and platform crossings in probe trials, it’s important to check that these are not confounded by differences in swimming ability. If one group is significantly slower than another, longer latencies in learning trials, and lower numbers of platform crossings in probe trials, may reflect motor or motivational issues rather than cognitive ones, for example with aged, lesioned, or sedated subjects.

Swim speed is also valuable in screening for outliers or for non-task-related behavior. A swim speed lower than 5 cm per second for an extended time may indicate floating, which could be due to stress, illness, or disengagement from a task. High variability in swim speed across trials might indicate inconsistent engagement or fatigue, rather than cognitive issues.

In some circumstances, you may wish to exclude subjects whose average swim speed is below a certain level, so as to include only subjects who meet the criteria for the study. This can make sure that results are not invalidated by including trials that appear to show cognitive impairments when in fact there is a different factor reducing swim speed, which in turn reduces performance on the task. In the HVS Image system the default value for ‘slow’ is 5 cm per second, and you can set this to whatever speed is appropriate to your study.

In other circumstances, a slow speed may be an informative expression of the condition being studied, for example in aging and neurodegeneration such as Alzheimer’s models; pharmacological interventions such as sedation or motor side effects; neurodevelopmental disorders such as autism, hyperactivity, hypoactivity; stress and anxiety related studies; and lesion models, e.g. cerebellum, motor cortex, vestibular system.

When to include or exclude slow swimmers:

So in some cases, it’s scientifically important to include slow swimmers, and in other cases they need to be excluded to maintain data quality.

If you’re studying aging, neurodegenerative diseases, or injury models, then swim speed may be a variable of interest. In Alzheimer’s models, for example, slower speed might reflect disease progression or motor-cognitive interaction. Likewise, if you’re assessing motivation, apathy, or fatigue, low swim speed will be of behavioral interest, and slow swimming subjects should be retained in analysis, for example, in chronic stress studies or mood disorder models. It may also be important in rehabilitation or recovery paradigms, where you may wish to track both navigation and swim speed over time, in order to observe any gradual improvements in both cognitive and physical performance.

There are other cases where you may wish to exclude animals with very low swim speeds. For example, if a subject is barely moving due to illness, sedation, extreme stress, or lack of motivation, behavior no longer reflects learning or memory performance, so including that data when investigating spatial memory or learning could falsely suggest cognitive impairment when that’s not the real issue. In motor or sensory impairment models (e.g., models of Parkinson’s disease, spinal cord injury, visual deficits), animals may have physical limitations that affect their water maze performance, regardless of their cognitive ability – a very low swim speed may make it hard to tell whether poor performance is due to cognitive deficits or physical inability.

The average speed does not, of course, reveal fluctuations in speed during the trial. In the HVS Image system, you not only see the average for the whole trial and the average active swim speed, but also the speed in different parts of the trial by using time slice analysis, to reveal rapid bursts or pauses.

Swim speed in time slice analysis may reveal:

• Pauses during scanning or focused search, especially in early learning or probe trials.
• High-anxiety subjects freezing or hesitating near walls, then swimming rapidly when they re-engage.
• Inconsistent swim patterns caused by some drugs or treatments, e.g. stimulants or sedatives.
• Stop-start swimming in subjects with mild motor or sensory deficits, as the subject compensates for control or balance issues, or in unengaged subjects who may stop intermittently.
• Reduced speed near the remembered platform site in probe trials, in subjects with spatial memory.