IVERSEN, I. H. Reinforcement omission and schedule-induced drinking in a response-independent schedule in rats. PHYSIOL. BEHAV. 18(3) 535-537, 1977.
Physiology & Behavior, Vol. 18, pp. 535-537. Pergamon Press and Brain Research Publ., 1977. Printed in the U.S.A.
BRIEF COMMUNICATION Reinforcement Omission and Schedule-Induced Drinking in a Response-Independent Schedule in Rats 1 IVER H. IVERSEN Institute of Pharmacology, University of Zurich (Received 18 February 1976) IVERSEN, I. H. Reinforcement omission and schedule-induced drinking in a response-independent schedule in rats. PHYSIOL. BEHAV. 18(3) 535-537, 1977. - Drinking, food-cup contact, exploration, and grooming were analysed in two food-deprived rats when food pellets were delivered on a fixed-time (FT) schedule of 1 min. The probability of drinking was high immediately after pellet delivery and then decreased to zero with the gradual increase in the probability of food-cup contact, exploration, and grooming. The probability of grooming decreased and the probability of food-cup contact and exploration increased when drinking was absent after occasional feeder operations without pellet delivery (reinforcement omission). This omission effect upon food-cup contact and exploration responses was experimentally simulated by preventing the drinking burst after ordinary pellet delivery. Response interactions Reinforcement omission Frustration Exploration Food-cup contact Fixed-time schedule
IN THE study of schedule-induced drinking [ 4], emphasis has recently been on the direct relationship between absence or presence of food pellets and subsequent absence or presence of drinking. After a brief stimulus associated with pellet delivery, as feeder operation in the absence of pellet delivery, for example, drinking may be absent [ 1,6'] or occur to lesser extent [ 71,8]. The procedure of occasional reinforcement omission has also been used to study the control of reinforcement over the subsequent pattern of reinforced behavior. Pauses in reinforced behavior were frequently reported to be shorter after reinforcement omission than after ordinary reinforcement under different reinforcement schedules [ 11,12]. Thus, after reinforcement omission, schedule-induced drinking may be absent and the probability of explicitly reinforced behavior may simultaneously be increased [ 6]. In a recent study, manipulations of burst durations of schedule-induced drinking proved sufficient to affect pauses in concurrent reinforced behavior [ 5]. The purpose of the present experiment was to extend this relationship to the reinforcement omission effect in schedules involving no explicitly reinforced response. Water drinking, food-cup contact, exploration, and grooming were recorded when food pellets were merely delivered response independently every minute (fixed time 1 min schedule). The changes in
Schedule-induced drinking
Grooming
the pattern of food-cup contact, exploration, and grooming were compared during absence of drinking after feeder operation without pellet delivery (reinforcement omission) and during prevention of drinking after ordinary pellet delivery. METHOD
Animals
Two naive Sprague-Dawley albino rats approximately 4 months old at the start of the experiment were housed in individual living cages with water continuously present. The rats were maintained at 80% of their free feeding body weights throughout the experiment. Apparatus
The experiment was carried out in a 33 cm dia., 15 cm broad running wheel which was, however, locked through out the experiment. A food cup, a drinking tube, and a lever were positioned on a plate which served as a door to the wheel. The 3.0 cm broad, 2.0 cm high, and 1.0 cm deep food cup was mounted 2.0 cm above the running surface and connected through a hole in the door to a Ralph Gerbrands pellet feeder. Contacts with and being at a
1 This research was supported by a traineeship from the European Brain and Behavior Training programme. Preparation of the manuscript was supported by a scholarstipend from the University of Copenhagen. I would like to express my appreciation to Dr. Joseph P. Huston for the use of his laboratory facilities. Reprints may be obtained from the investigator, Ryesgade 34 A, 2200 Copenhagen N, Denmark.
535
IVERSEN
536
distance of 0.1 cm or less from the food cup were recorded with a modified body-capacitance sensitivity system. A 0.8 cm diameter ball-tipped stainless steel tube was mounted on an extension of the armature of a modified Claire and Co. relay and connected to a bottle containing tap water. For as long as the relay was operated, the tip of the tube was accessible in a hole in the door, 2.0 cm from the left edge of the food cup and 4.0 cm above the running surface. The drinking tube was presented or retracted approximately 0.1 sec after onset or offset of the operating current to the relay. A white light above the hole was lit for as long as the tube was accessible. Contacts with the drinking tube were recorded with a drinkometer circuit. Food-cup contact and tube contact were physically compatible responses. (The lever was mounted to the left of the tube hole but was not used in this experiment. Inspection indicated that lever pressing only occurred a few times for each session.) The apparatus was enclosed in a sound-attenuating chamber with white masking noise and houselight present during sessions.
Procedure One 45 mg Campden Instruments food pellet was delivered each minute independently of any response by the rat (FT 1 min) for 14 sessions. The drinking tube was continuously present, and a session was terminated after delivery of 40 pellets. For each of the next 4 sessions, 6 of the 40 feeder operations per session did not result in pellet delivery. For six additional trials per session, the tube was retracted immediately at pellet delivery and first presented at the next feeder operation. For the remaining 28 feeder operations in each session, a pellet was delivered and the tube was accessible throughout the interval, as during the preceding 14 baseline sessions. The 12 test trials were mixed between the regular intervals, with the restriction that at least three intervals with pellet delivery intervened between intervals with no pellet. (The operations necessary to prevent pellet delivery at some feeder operations were performed manually by the experimenter. A thin, soft, and flexible rubber tube from the feeder to the food cup was silently disconnected during intervals preceding feeder operation without pellet delivery and reconnected during the next interval. Observations revealed no disturbance of the performances during this operation.) To avoid any bias in the manual scheduling of feeder operations with and without pellet delivery, a predetermined sequence of events was followed during sessions. Exploration (sniffing, climbing in the locked wheel, and standing on hind legs) and grooming (scratching and licking the body) were recorded by the experimenter by pressing appropriate keys for as long as the behaviors occurred. The times spent in contact with or close to the food cup, in contact with the drinking tube, exploring, and grooming were calculated for 6-sec subintervals within each interval between feeder operations. The probabilities of food-cup contact and tube contact were also analysed for the first three 2-sec intervals immediately after pellet delivery. Data for the first interval in a session, before the first pellet delivery, were not included in the analysis as licking the tube never occurred in this interval. Data are presented for each interval of the last 4 sessions. RESULTS
A licking burst reliably occurred immediately after
retrieval of each pellet by the fifth baseline session for both rats. Food-cup contact and exploration also occurred within each inter-pellet interval whereas grooming occurred within only approximately 40% of the intervals. The response probabilities are shown in Fig. 1 for each condition for Rat 1. Similar data were obtained for Rat 2. Overall response probabilities are shown in Table 1 for each condition and Rat. Within the first few sec after pellet delivery, the food-cup contact probability decreased from 1.0 to near zero simultaneously with a rapid increase in the licking probability to 1.0. (The median latency between pellet delivery and licking onset was 1.8 and 2.0 sec for Rat 1 and Rat 2, respectively. Observations revealed that pellet eating was reduced to mere pellet retrieval with perhaps only one or two chewing movements.) Subsequently, the licking probability gradually decreased with a concomitant increase in the food-cup contact, exploration, and grooming probabilities. In the last 6 sec preceding pellet delivery, the food-cup probability increased to nearly 1.0 with con current low or zero exploration, licking, and grooming probabilities. (Similar response probability profiles were obtained for each rat during the last four baseline sessions.) NO PELLET
PELLET WATER
lO[L 0,8 0.6 0.4 0.2
>-
�
:::J
co
