F.J. van der Staay
Institute for Neurobiology, Troponwerke, Köln, Germany
The Morris water escape task is one of the most frequently used experimental paradigms to assess disturbances of cognitive functions as a consequence of aging, genetic manipulations (transgenic animals), of specific brain lesions, of experimental induced infarcts and to characterize properties of putative cognition enhancing compounds. The standard water escape task, in which a rat or mouse is required to localize a submerged platform, measures predominantly spatial reference memory (RM). The RM holds trial-independent information about, for example, the position of the escape platform in the water tank.
Repeated acquisition procedures have been designed to assess an additional memory component, called working memory (WM). Within a daily training session of the repeated acquisition paradigm, each of four start positions is used randomly in every series of four trial pairs. From one daily session to the next, the escape platform is positioned in a different quadrant and each position is used once in a series of four consecutive sessions. Rats are randomly started from each of four starting positions on the first trial of a pair and from the same start position on the next trial of the pair. Successful repeated acquisition is demonstrated when subjects have shorter latencies to find the platform during the second trial of a pair than during the first trial (one trial learning), i.e. they show an improved WM performance.
In a series of experiments, we tried to determine a) how aging affects the RM and WM performance of rats in the Morris task, b) which characteristics of the experimental environment affect the water escape RM learning, and c) whether rat and mouse strains differ in this task.
The most commonly used measure to describe learning performance in the different versions of the Morris water escape task is the latency to escape onto the platform'. Our experiments show that the escape latency can severely be biased by the animal's swimming speed. Thus, if escape latency is used as the measure to describe the animal's learning, then it is mandatory to show that this measure shows high correlation with other measures such as the distance swum to reach the platform. Otherwise, the escape latency alone does not provide a valid measure of learning in the Morris water escape task. In those cases where the escape latency is unrelated to other indices of spatial learning, learning is best described by considering additional parameters, such as the distance swum to find and escape onto the submerged platform' and the swimming speed.