Reward devaluation in a two-link chain schedule: effects of reward density in the proximal link and food restriction protocol in rats

Abstract

Food restriction has been shown to affect responding for reward and has been manipulated to devalue reward. Reward density has been shown to alter responding in both first-order schedules and chain schedules. Devaluation has differential effects in first-order schedules and the links of chain schedules. The objective of the current study was to analyze the reward devaluation effect in a two-link chain schedule by manipulating food restriction and reward density in the proximal link; these variables had previously not been studied systematically. The chain schedule required rats to lever press according to a random interval 120-s schedule to turn on a light conditioned stimulus (CS). In the presence of the CS rats were rewarded for pulling a chain; the chain was active on a variable ratio (VR) 5 schedule of reinforcement for some rats, while the chain was on continuous reinforcement (CRF) for the other rats. Food restriction protocols were either 1-hr daily free feeding (1-hr feeding) or a daily ration to maintain weight at 80% of free feed controls (ration). Devaluing the CS while in the isolated proximal link was done by exposing trained rats to sessions of chain pulling in the presence of the CS but withholding reward. In subsequent lever press sessions without the CS, decreased responding was seen in rats that had undergone the devaluation procedure, but only in rats on the 1-hr feeding protocol. This devaluation effect was found in both the CRF and VR 5 schedules. Subsequent tests of lever pressing for the CS also revealed decreased responding. Post-test CS discrimination sessions demonstrated CS devaluation that has not been demonstrated previously. These results demonstrate that the isolated devaluation of the proximal link in a heterogeneous two-link chain schedule can result in decreased responding in the isolated first link. This effect is dependent on sufficient food restriction and demonstrates a measurable devaluation effect which is independent of the reward being present. This study also demonstrated that isolated devaluation of the second link can be subsequently measured in a post-test CS discrimination re-acquisition session regardless of food restriction, which had no significant effect in the session. The devaluation effect has not been demonstrated in a re-acquisition session before, rather responding in previous re-acquisition sessions were only affected by food restriction.