Norway rats have been shown to depend on short-term spatial memory to find food on a radial arm maze (RAM), but what locomotor search tactics are involved in using this memory effectively? Four experiments distinguished tactics of distance minimizing, central-place search, trail following, thigmotactic search, and random search by using different configurations of a RAM placed flat on the floor of an arena. These search tactics make similar predictions on an elevated RAM but predict different outcomes on a floor RAM because the rats are free to approach the food from any direction. After initial trials dominated by exploration, rats traveled along arms to food, even when the resultant distance was up to three times the minimum distance. Withno food present, rats also traveled along arms; with no arms up to present, they traveled along walls to food. It appears that both maze arms and arena walls engage mechanisms related to trail following in rats.

rats in a cool environment (25ЊC) acquired an operant response when rewarded with a 20-s-long warming of the platform (from 25 to 36ЊC) on which they lay. In Experiment 1, the head-turning response was learned by pups at all ages. When the contingency was reversed so that pups were reinforced for turning to the side opposite that correct during training, the original response extinguished for 1-day-olds, but not for 5-or 11day-olds. In Experiment 2, the rewarded side was randomly selected for each trial. One-dayolds perseverated in turning to the side correct on that trial while the reinforcer remained on, but 5-and 11-day-old rat pups did not. We conclude that 1-day-old pups were more responsive to the change in experimental contingency in Experiment 1 due to this thermotaxic behavior.

Hoffman, Flory, and Alberts (1999) demonstrated that 1-, 5-, and 11-day-old rats in a cool environment (25ЊC) acquired an operant head-turning response when rewarded with a 20-s warming of the platform on which they lay. In the current experiment 5-and 11day-old rats in a hot environment (40ЊC) acquired the head-turning response when rewarded with a 20-s cooling of the platform on which they lay, but 1-day-olds did not. The concept of ontogenetic adaptation helps us interpret these results: Neonatal thermotaxis constrains the 1day-olds from learning a novel operant response for a cool reinforcer in a hot environment. Because the thermotaxis wanes from birth, it is not as strong in 5-and 11-day-old pups that are thus able to learn the operant for a cool reinforcer.

Hoffman, Timberlake, Leffel, and Gont (1999) concluded that the tactic of effective trail following (in the form of arm and wall travel), rather than distance minimizing, central-place search, or random search, best characterized the locomotion of rats on a radial arm maze placed flat on the floor of an arena (a floor RAM). The present experiments analyzed further the stimulus control and function of arm and wall travel. Experiment 1 showed that arm travel was controlled more by the edge of a maze arm than by its surface. Experiment 2 showed that rats with whiskers clipped on one side traveled along arms less and along walls more than did intact rats. Experiment 3 showed that maze arms increased search effectiveness and decreased suppression of locomotion by bright light and a novel environment. The results support the hypothesis that arm and wall travel are based on mechanisms of trail following, which, in natural settings, contribute to food finding and regulation of social relations and fear.Hoffman, Timberlake, Leffel, and Gont (1999) examined the potential contributions of four locomotor tactics to effective food gathering by rats on a radial arm maze (RAM). The tactics were distance minimizing, centralplace search, trail following, and random (meandering) search. These tactics, in conjunction with a win-shift rule, make identical predictions about locomotor behavior on an elevated RAM, so Hoffman et al. placed the arms and center platform of a six-arm RAM flat on the floor of a large arena. Because the rats were free to approach the food cups from any direction, on or off the maze, the predictions of these locomotor tactics could be contrasted.In a standard-arm RAM on the floor, rats could distance minimize by traveling from the central platform to a cup and then in a circular route (a series of flat arcs) between cups. They could show central-place search or trail following by moving from the center to each cup and back. The rats also could search randomly, or along walls to food cups (thigmotactic search), or could fail to find food cups in an efficient manner. After exploring freely on and off maze arms during the initial trials, the rats pred ominantly traveled along maze arms to visit the food cups. Arm travel dominated even when the distance required to search all the cups was three times that of a direct route and even when the food cups were never baited. When no maze arms were present, the rats followed the arena walls between cup visits. The pervasiveness ofarm travel in the presence of a variety of more efficient routes to food (see also Roche & Timberlake, 1998) suggests that arm travel is controIled in part by species-typical perceptual-motor mechanisms related to the extensive trail following shown by colonies of wild rats (Calhoun, 1962;Telle, 1966). This trail-following hypothesis raises several empirical and theoretical issues. The first question is, What stimulus characteristics control arm (and wall) travel? Because Hoffman et al. (1999) failed to show a reliable effect of odor in guidi...

No abstract