Rotation of water in the Morris water maze interferes with path integration mechanisms of place navigation.

The idea that place navigation in the Morris water maze is implemented by path integration between locations determined by landmark sighting was investigated in a 200-cm-diameter pool in which circular (7.2 degrees/s) motion of water could be induced by tangentially arranged water jets. The rats were trained at 8 trials per day to navigate to an erectable platform which was raised after the rat had spent a criterion time in the target annulus (30 cm in diameter) in the midpoint of the NW quadrant. Asymptotic escape latency of 7 s was reached after 9 days in moving water (n = 8) and after 6 days in stationary water (n = 8). The group overtrained for 13 days in stable water performed well even after it was transferred to moving water. Changing the sense of rotation of water from counterclockwise to clockwise did not affect the asymptotic performance. The above findings show that overtrained rats rely on landmark sighting rather than on path integration. The influence of water movement reappeared when place navigation to a new target (SW) was examined in alternating 2-s periods of light (L) and darkness (D). On the first day, the latencies were 15.2 +/- 1.2 and 22.8 +/- 1.9 s in stable and moving water, respectively, but dropped to 10 s on the following day. The tracks generated in the L period were more tortuous than those generated in the D period and this difference was more pronounced in moving than in stable water. It is concluded that path integration mechanisms supporting navigation during intervals of darkness are impaired in moving water but that this impairment disappears in overtrained animals.