Gary A. Lucas

Schedule constraint on the average drink burst and the regulation of wheel running and drinking in rats.

Two experiments compared predictions of a molar-pattern model and a general molar behavior regulation model by requiring rats to wheel run for access to water. In both experiments schedule parameters constrained the baseline average burst length of drinking without constraining total drinking. Five levels of schedule constraint were imposed on time spent per drinking burst (Experiment 1) or the number of drinks per burst (Experiment 2). The results of both experiments supported the general molar behavior regulation view but not the molar-pattern model by showing no increase in total wheel running and no decrease in total drinking under schedule constraint. However, both experiments also showed local effects of drink burst constraint, including a direct relation between the degree of constraint and the local rate of drinking, and an approximation of the temporal distribution of baseline drinking under all degrees of schedule constraint. Most local changes support the view that rats defend the baseline temporal distribution of responding under schedule constraint, though some changes appear related to disruption of local response pattern characteristics.

System-Specific Differences in Behavior Regulation: Overrunning and Underdrinking in Molar Nondepriving Schedules

In two experiments we tested the molar regulation prediction that animals adjust schedule performance to reduce deviations from baseline response totals. Both experiments constrained the baseline drink-burst length under molar nondepriving schedules but allowed rats to continue running without drinking. In Experiment 1, rats were required to run in order to drink. In Experiment 2, water was delivered independently of running by fixed-time (FT) schedules. Under the run-to-drink contingency, rats exceeded their baseline amounts of running (overrunning) but failed to maintain their baseline water intake (underdrinking). The total amount of running that did not lead to drinking approximated baseline running. Under the FT schedules, rats again underdrank, but total running approximated baseline. These results do not support previous studies that have shown molar equilibrium effects under nondepriving reciprocal schedules. We conclude that (a) contingent running may not substitute for independent running; (b) intermittent access to water reduces the total instigation for drinking; (c) molar regulation differs under reciprocal and nonreciprocal schedules; and (d) more dynamic, system-specific regulatory models need to be developed.

Anticipation of future food: suppression and facilitation of saccharin intake depending on the delay and type of future food

A series of studies examined the (Sprague-Dawley) rats tendency to suppress intake of .15% saccharin when it was followed by a second food after 4-, 16-, or 32-min delays. The second foods examined were 32% sucrose, 64% sucrose, lab chow, a Nutrasweet solution, skim milk, and chocolate milk. Saccharin intake was influenced by both the delay and the specific food available. Subsequent analysis showed that saccharin intake before the 4-min delay was an inverse function of the caloric value of the second food. However, saccharin intake before the 16-min delay was better predicted as an inverse function of the hedonic value of the second food. The results suggest that the caloric and hedonic values of a food may influence food selection across different time courses, and that the effective time horizon for the sequential comparison of foods depends on the specific foods that are compared.

Anticipatory contrast as a measure of time horizons in the rat: Some methodological determinants

In three experiments, the time horizon over which the rat evaluates alternative feeding sources was investigated. The time horizon was measured by the suppression of intake of one incentive (a 0.15% saccharin solution) when a preferred alternative incentive (a 32% sucrose solution) was available but delayed. In Experiment 1, we found a direct function between the amount of saccharin intake and the delay time before access to 32% sucrose. Compared with intake for a saccharin-only control, saccharin intake was suppressed before 4-min and 16-min sucrose delays, but not before a 32-min delay. Because previous work (Flaherty & Checke, 1982) had reported suppression before a delay of nearly 32 min, in the subsequent experiments we examined factors that might account for this difference. In Experiment 2, we found that saccharin intake was suppressed before a 32-min delay interval when saccharin and sucrose solutions were presented in a bright-novel test environment but not when the same solutions were presented in the home cage. In Experiment 3, we found that the time between testing and subsequent postsession feeding could also affect the suppression of saccharin intake. Saccharin intake was suppressed when access to 32% sucrose was delayed by 32 min and the test situation was followed by immediate postsession feeding, but not when postsession feeding was delayed by 90 min. These results thus extend estimates of the rat’s time horizon to at least 32 min, but indicate that the effective time horizon can vary, depending on the test situation.

Adjunctive behavior of the rat under periodic food delivery in a 24-hour environment

The behavior of 4 rats living in complex environments was monitored 24 h per day during free-feeding baseline and under conditions of periodic access to food. Under the periodic schedules, the minimum interfood interval (IFI) was increased from 16 to 512 sec in an ascending series. Periodic food produced robust overall increases in investigation of the feeder, drinking, general activity, and rearing, but not in wheel-running. The temporal distribution of behavior within the IFI was similar across subjects and supported the hypothesis that some responses were largely time-locked to the period immediately following eating, while other responses expanded to fill the interval. However, these response differences were not adequately captured by present classification schemes. Finally, the distribution of drinking following a food pellet strongly resembled the distribution of drinking following bouts of feeding in baseline. The results suggest that adjunctive behavior stems from three sources: (1) a simple increase in the number of opportunities for expression of normal preprandial and postprandial behavior, (2) an increase in the preprandial behavior directed toward the site of expected food, and (3) an increase in the postprandial distribution of both site-directed and more general exploratory behavior. These findings suggest that adjunctive behavior is not extraneous, but is an orderly distribution of responses ordinarily related to feeding and foraging for food.

Negative anticipatory contrast and preference conditioning : flavor cues support preference conditioning, and environmental cues support contrast

In 2 experiments, access to a 0.15% saccharin solution was followed on alternating days by access to a 32% sucrose solution and the same saccharin solution. In Experiment 1, rats increased both intake of and preference for a flavored saccharin solution that predicted sucrose, but neither effect was found using a predictive odor cue alone. Experiment 2 replicated the predictive flavor results but showed suppression of saccharin intake when environmental cues predicted sucrose. When both flavor and environment predicted sucrose, saccharin intake did not change, but preference for the predictive flavor increased. Discriminative taste cues appear to facilitate the development of preference conditioning, but environmental cues favor negative anticipatory contrast effects. Also, preference conditioning and contrast may develop concurrently and compete for expression.

Periodic water, interwater interval, and adjunctive behavior in a 24-hour multiresponse environment

To examine the generality of the interreward response effects shown by rats under periodic food delivery, we presented .10 ml of water at minimum interwater intervals that ranged from 8 to 512 sec. Use of a 24-h multiresponse environment allowed evaluation of interdrink responses with respect to their excessiveness, patterning, and functional relationship to the interwater interval. In contrast to the extensive activity-inducing effects of periodic food, the only major excitatory effect of periodic water was increased attention to the water source. Although there were a few bitonic and direct relationships between interwater interval and changes in responding, the great majority of functions were inverse or inconsistent. Further, unlike the increase in drinking under periodic food, total eating decreased under periodic water. The major similarity with food reward was the apparent separation of interreward behavior into three general classes of reward-appropriate foraging responses: area-restricted search after reward, more general search (and waiting), and focal search preceding the next reward delivery.

Interpellet delay and meal patterns in the rat

Rats typically eat in short discrete meal periods separated by long intermeal intervals. The present study measured meal patterns and total intake of food and water when the rate of access to food pellets within the meal was reduced by delaying the time between pellet deliveries. In contrast to studies that reduce the caloric density of food, simple interpellet delay produced a marked reduction in meal size at the shorter (32 sec or less) interpellet delays. However, longer delays (up to 128 sec) produced no further change in average meal size. The results suggest that meal size is determined (1) by a positive feedback component with a relatively short decay time probably based on gustatory stimulation from feeding, and (2) by a negative feedback component capable of integrating total intake across delays of up to 1 hour. Increasing the delay between pellets appeared to interfere with the positive feedback component, but not with the negative feedback component.

Some effects of reinforcer availability on the pigeon’s responding in 24-hour sessions

Restrictions on food availability produced by schedules of reinforcement were examined in three homing pigeons continuously housed in operant chambers. Total daily access to food was free to vary and depended on the subject’s contact with the schedule in effect. Experiment 1 varied reinforcer duration within a continuous reinforcement schedule in order to provide a description of the pigeon’s feeding pattern under minimal constraints. In Experiments 2 and 3, access to food was contingent on responding in fixed-interval schedules, and limits on availability of food were varied by changing the duration of reinforcement (Experiment 2) or the frequency of reinforcement (Experiment 3). In all three experiments, a decline in the scheduled availability of food produced an increase in both the overall response rate and the local response rate. In addition, the distribution of responding across the day followed a diurnal rhythm typical of the pigeon’s unconstrained pattern of food intake. These effects are consistent with previous studies showing an inverse relationship between instrumental response rate and reinforcer availability in the absence of fixed deprivation, and support the interpretation that this inverse relationship results from constraints imposed on preferred patterns of intake. The data on the local distribution of responses were consistent with an extension of the response-deprivation hypothesis (Timberlake & Allison, 1974) to local response patterning.

US duration and local trial spacing affect autoshaped responding

The acquisition and maintenance of signal-directed pecking was examined in week-old Leg-Horn chicks responding to a keylight stimulus paired with heat. In contrast with previous studies using pigeons with food as the US, both speed of acquisition and asymptotic level of keypecking were a direct function of US duration. Experiment 2 examined responding using a within-subject design to isolate the effects of trial spacing on performance during the immediate trial from the effects on performance during a following trial of fixed length. These comparisons revealed a significant effect of intertriai interval (ITI), with less responding after shorter intervals. The effect of different temporal spacing was apparent in responding on the immediate trial, but not on the following trial. These local ITI effects were better predicted by a recent autoshaping model based on relative waiting time (Jenkins, Barnes, & Barrera, 1981) than by a model based on relative US expectancy (Gibbons & Balsam, 1981). However, neither model predicted the effect of US duration. A reexamination of the US-duration literature suggested that the diversity of previous findings is consistent with the assumption that conditioned responding is an inverted U-shaped function of US duration.