The Generation and Maintenance of Schedule-Induced Polydipsia in Normal Male Rats Without Weight Reduction

The Generation and Maintenance of Schedule-Induced Polydipsia in Normal Male Rats Without Weight Reduction

Physiology & Behavior, Vol. 62, No. 6, pp. 1385–1390, 1997 © 1997 Elsevier Science Inc. All rights reserved. Printed in the U.S.A. 0031-9384/97 $17.00...

276KB Sizes 0 Downloads 43 Views

Physiology & Behavior, Vol. 62, No. 6, pp. 1385–1390, 1997 © 1997 Elsevier Science Inc. All rights reserved. Printed in the U.S.A. 0031-9384/97 $17.00 1 .00

PII S0031-9384(97)00367-3

BRIEF COMMUNICATION

The Generation and Maintenance of Schedule-Induced Polydipsia in Normal Male Rats Without Weight Reduction JAMES T. TODD,*1 LISA A. CUNNINGHAM,† ALISON A. JANES,‡ JOSEPH MENDELSON§ AND EDWARD K. MORRIS\ *Department of Psychology, Eastern Michigan University, Ypsilanti, MI 48197; †Genetic BioSystems, San Diego, CA, ‡School of Pharmacy, §Department of Psychology, and \Department of Human Development, University of Kansas, Lawrence, KS 66045 Received 17 April 1997; Accepted 16 June 1997 TODD, J. T., L. A. CUNNINGHAM, A. A. JANES, J. MENDELSON AND E. K. MORRIS. The generation and maintenance of schedule-induced polydipsia in normal male rats without weight reduction. PHYSIOL BEHAV 62(6) 1385–1390, 1997.—Experiment One demonstrated that two normal male Sprague-Dawley rats (;60 days old) with free access to food and two control rats whose weights were held constant by dietary restriction acquired schedule-induced polydipsia (SIP) in daily 33–35 min sessions of fixed-time 60-s food delivery. Three of the rats showed rapid acquisition of SIP; the fourth acquired SIP more slowly and consumed less per session the other three rats. After a 36 – 40 day period without sessions, the constant-weight rats showed a 37% decrease in overall consumption due to reduced drinking bout length. The SIP of the free-feeding rats was not affected by the interruption. After 90 –100 periodic food delivery sessions, all subjects consumed an average of 11.2–12.2 mL per session compared with 1.8 – 4.8 mL per session in baseline sessions with massed food presentations. Experiment Two replicated the acquisition phase of Experiment One using two non-weight-reduced rats of the age and size of those typically used in SIP studies (;30 weeks old). Both acquired SIP, although one showed only a small average increase in consumption per session over baseline (2.8 mL/session under periodic food vs. 0.8 mL following massed-food presentations). Before weight reduction, the stronger drinker consumed ;8.8 mL per session compared with an average of 0.6 mL per session in baseline. After weight reduction, both exhibited strong SIP (18 –19 mL per session in the final five sessions). This study demonstrates that weight reduction is not a necessary condition for the generation and maintenance of SIP in rats. © 1997 Elsevier Science Inc. Schedule-induced polydipsia Rats

Adjunctive behavior

Weight reduction

SCHEDULE-INDUCED polydipsia (SIP) is a stereotyped pattern of post-food drinking exhibited during the periodic delivery of small quantities of food to rats and some other organisms (3,4,5). The combined effect of numerous post-food drinks is excessive overall consumption. Because SIP does not appear to be maintained by reinforcement contingencies or basic respondent processes, it is sometimes classified as a type of nonoperant, nonrespondent ‘‘adjunctive’’ behavior (30). Reflecting the view that excessiveness is its most important feature (8,9), SIP is typically defined as greater overall water consumption during sessions of periodic food delivery than in equally long sessions without periodic food delivery (20). Because SIP is typically generated on food delivery schedules, standard practice includes diet restriction to 1

Food deprivation

reduce the weight of experimental subjects to 70 –90% of their pre-experimental free-feeding (ad lib) weights. Correspondingly, a common feature of experimental reports and reviews on SIP is a reference to the necessity or importance of weight reduction (or the food deprivation operation) for producing SIP (5,6,15,20,21,27). Although weight reduction by food deprivation is apparently widely believed to be crucial for the development of SIP, its necessity for establishing SIP under standard experimental conditions (i.e., daily sessions of intermittent food deliveries) has not been determined. Diminished or eliminated SIP at recovered ad lib weight (4,21,27) does not demonstrate that weight reduction is required to establish SIP. In contrast, the acquisition of SIP by rats with less than 10 –30% weight reduction (17,19), or during weight

To whom correspondence should be addressed. E-mail address: PSY [email protected]

1385

Periodic reinforcement

1386

TODD ET AL.

recovery (4,27,29), suggests that weight reduction might be less important than is generally believed— but does not show that it is entirely unnecessary. In the few reported experiments in which no weight reduction was programmed, SIP either did not develop or the studies are procedurally difficult to compare with typical studies of SIP. For example, SIP was not reported in non-weightreduced rats after 27 atypically short (10 min) sessions of fixedinterval 60-s food availability (27). These sessions were one-third to one-sixth the length of standard sessions and may have provided insufficient time for SIP to develop. Similarly, researchers were unable to generate unequivocal SIP in non-weight-reduced subjects using reinforcers such as brain stimulation (2), running wheel access (22,28), and visual displays [for monkeys, (10)]. Elevated drinking and other activities have been reported in humans using food and nonfood deliveries (1,9,11,25). The existence of SIP in humans is disputed, however, because of procedural differences between human and nonhuman studies (7,18). Because the necessity of weight reduction in the generation and maintenance of SIP has not been determined, the purpose of this set of experiments was to examine the development of SIP in normal, non-weight-reduced rats using standard experimental procedures. Previous research that showed a direct relationship between weight reduction level and SIP suggested that SIP in nonweight-reduced rats would be weaker than SIP generated under weight reduction (12,21,27,29). Thus, several measures of drinking were recorded to ensure that even weak SIP could be detected (i.e., drinking bout duration, post-food latency, bout probability, and overall quantity consumed). Experiment One examined SIP during sessions of fixed-time 60-s (FT 60-s) food delivery in non-food-deprived rats and in rats held at a constant weight (also, technically speaking, not-weight-reduced). Because Experiment One employed atypically young rats (;60 days), Experiment Two examined SIP in free-feeding rats of a more typical experimental age (;30 weeks). METHOD

Experiment One: SIP in Constant-Weight and Free-Feeding Rats Subjects. Four male Sprague-Dawley rats acquired at 23 days of age served as subjects. The two lightest rats (68 and 73 g) were designated 1CW and 1FF, respectively (for constant weight and free feeding, respectively); the two heavier rats (90 and 108 g) were designated 2CW and 2FF. For 12 days, the rats were housed in pairs in clear polycarbonate cages (26 cm wide, 48 cm deep, and 20 cm high) with shredded pine bedding and free access to Purina #5001 maintenance ration (Ralston Purina, St. Louis, MO) and water. Afterward, they were housed individually. To facilitate the acceptance of the food pellets (Bio–Serv, Frenchtown, NJ) used in the experimental sessions (BioServ 45 mg; #00211), 50 –100 pellets were placed daily in each cage until the start of the experiment. Experimental sessions began when the constant-weight member of each pair reached ;200 g, as determined by daily presession weighing—34 days for 1CW and 1FF; 42 days for 2CW and 2FF. After the third session of the experiment, the daily ration of rats 1CW and 2CW was adjusted to keep their weights a close to 200 g as possible without causing a weight reduction. Each showed an average daily weight gain of ;0.5 g with occasional small weight reductions (,1% typical). Rats 1FF and 2FF continued to have unrestricted access to food in their home cages and gained weight normally. By the end of the experiment, 1CW and 2CW weighed 254 and 255 g; 1FF and 2FF weighed 696 and 542 g. Apparatus. All sessions were conducted in standard operant chambers for rats with the response levers removed. A stainlesssteel drinking tube with a 2 mm aperture extended 3 cm into the

chamber 5 cm to the left of a recessed food cup. Drinking-tube licks were recorded by electrical contact drinkometers. To permit contact with only the tip of the drinking tube, each tube shaft was partially covered by a 2.5-cm inner diameter end-cap for PVC pipe. Procedure. All sessions were ;35 min in length and were conducted daily except during a 36 – 40 day ‘‘hiatus’’ for emergency asbestos removal and three 3–5 day interruptions. Three conditions were employed: 1) Massed food baseline (MFBL)— each rat was placed in the experimental chamber where it found 30 food pellets waiting in the food cup. These sessions were conducted to assess the drinking associated with nonscheduled food consumption; 2) No-food baseline—these were identical to the MFBL sessions except that no food pellets were available; 3) Fixed-time 60-s food delivery (FT 60-s)—a pellet was delivered every 60 s independent of the rat’s responding. The first pellet arrived about 2 min after the rat was placed in the chamber; the rat was returned to its home cage within 5 min of the last pellet delivery. The conditions were, in order: a) five MFBL sessions, b) 61–74 FT 60-s sessions, c) 13 post-hiatus FT 60-s sessions, d) five additional MFBL sessions, e) one no-food baseline session, and f) six final FT 60-s sessions (see Fig. 1). In the nine FT 60-s sessions preceding the hiatus, rats 1CW and 1FF were given 0, 30, 60, or 90 pellets immediately before each session. This test of the effects of presession feeding on SIP was not conducted with the other subjects because of the hiatus. RESULTS

Figure 1 shows the number of intervals containing a drinking bout of at least 1-s duration and the quantity of water consumed per session by each rat in Experiment One. After ;10 sessions, the quantity consumed by rats 1CW, 2CW, and 1FF under FT 60-s food delivery (8 –20 mL/session) reliably exceeded (with three exceptions for Rat 1FF) the maximum amount consumed during any baseline session (0 – 8 mL/session). (Table 1 shows mean consumption in the final 5 sessions of each condition). The probability of postfood drinking bouts rose to 0.80 (24 out of 30) or higher after ;10 sessions and to 0.90 or higher in rats 1CW and 2CW after 15 sessions. The presession feedings given to rats 1CW and 1FF at the end of the first FT 60-s condition had no apparent effect. Although Rat 2FF exhibited a slower acquisition and lower overall consumption than the other subjects during the first FT 60-s condition, after 8 sessions it reliably consumed more each session (except in session 35) than it had in any MFBL session. By session 70, its postfood drinking probability was at least 0.66. The hiatus did not appreciably alter the bout probability, overall consumption, or acquisition trend (2FF) in the free-feeding rats. In contrast, the quantity consumed per session by both constantweight rats decreased by ;37%. Because bout probability did not decrease, the reduction in quantity is attributable to reduced bout durations. (The average bout durations for rat 2CW in the final five prehiatus sessions after the first, fifteenth, and thirtieth pellet of each session were 22, 14, and 10.7 s, respectively, and 14, 9.4 and 6.6 s in the first five posthiatus sessions). By the final five FT 60-s sessions, all subjects consumed similar quantities of water (11.2–12.2 mL/session). However, the free-feeding rats showed slightly lower drinking probabilities than the constant-weight rats (96 and 84% for 1FF and 2FF respectively; 98 and 99% for 1CW and 2CW). Figure 2 contains cumulative records which show the temporal features of the SIP exhibited by subjects 1CW and 1FF in the first session that each exhibited post-pellet drinking (left), during acquisition in the first FT 60-s condition (middle), and after the development of strong SIP (right). Drinking bouts typically began within 5 s of each pellet delivery regardless of probability. (The

POLYDIPSIA WITHOUT WEIGHT REDUCTION

1387

FIG. 1. Number of drinking bouts and total quantity of water consumed per session by all subjects in each condition of Experiment One. The massed food baseline conditions are designated ‘‘BL’’; the fixed-time food delivery sessions are designated ‘‘FT 60-s.’’ The final (sixth) session of the second baseline condition (BL) was a no-food baseline session and is indicated by a shaded dot. Omitted data points indicate sessions in which data were spoiled due to experimenter error or equipment failure.

average latency between the first food pellet delivery per session and drinking by all subjects in the five immediate pre- and posthiatus sessions was 2–3 s). These records show that the post-pellet drinking pattern characteristic of SIP in rats occurred before consumption was clearly excessive. Rat 1CW, for instance, exhibited post-pellet drinking in the first FT 60-s session. METHOD

Experiment Two: Acquisition of SIP by Standard-Age Free-Feeding Rats Subjects and apparatus. Two experimentally naive male Sprague-Dawley rats (5FF and 6FF) were ;30 weeks old and weighed TABLE 1 AVERAGE QUANTITY CONSUMED PER SESSION IN THE FINAL FIVE SESSIONS OF EACH CONDITION OF EXPERIMENT ONE

573 and 545 g at the beginning of the study. Each had been housed individually with continuous access to food and water in the type of cage used in Experiment One. This experiment used the same feeding regimen and apparatus as Experiment One. Procedures. Five daily MFBL sessions were followed by 107– 139 sessions of FT 60-s food delivery. For the first 87 (5FF) or 127 (6FF) sessions (with two 1-day exceptions), each subject was allowed free access to food in the home cage; following sessions 30 and 55, only 5 g of ration was given. This diet restriction resulted in temporary weight loss of 2.8 – 4.9%. After session 87 (5FF) or 127 (6FF), the subjects were given 5 g per day of maintenance ration to determine if each was capable of acquiring strong SIP under weight reduction. (The weight of 5FF decreased 27% from 679 to 499 g; 6FF’s weight decreased 22% from 609 to 477 g) The only handling difference between Experiments One and Two was that the rats were weighed only before each daily session. RESULTS

Subject Condition

1CW

2CW

1FF

2FF

Baseline 1 FT 60-s 1 Posthiatus Baseline 2* FT 60-s 2

2.4 16.8 13.0 1.8 11.8

0.8 13.3 10.2 1.8 11.8

3.2 11.2 10.4 3.2 12.2

1.6 7.2 10.2 4.8 11.2

Note: All measurements are in milliliters (mL). * No-food baseline session not included.

Figure 3 shows that both subjects in Experiment Two exhibited SIP under free-feeding conditions: Rat 5FF consumed an average of 2.6 mL per session in the five sessions before the first 5 g feeding compared with an average of 0.8 mL per session in baseline. Corresponding measures for rat 6FF were 8.8 mL (under FT 60-s) and 0.6 mL (under MFBL). The 5-g feedings immediately increased the probability and quantity consumed in both subjects; the quantity consumed increased by a factor of four in rat 5FF and by about 50% in 6FF. Although the effects of the 5 g feedings seemed to persist for several days, the overall acquisition trend appeared unaffected. The final weight reduction condition

1388

TODD ET AL.

FIG. 2. Cumulative records of drinking by subjects 1CW and 1FF in the first session that each exhibited post-pellet drinking (left), during acquisition in the first FT 60-s condition (middle), and after the development of strong SIP (right). Downward deflections of the pen indicate FT 60-s food deliveries. The pen reset after 450 licks.

increased overall consumption and bout probability in both subjects— by the final five sessions of the experiment, 6FF consumed an average of 18 mL per session compared with 11 mL in the final five sessions of the free-feeding condition. Consumption by 5FF in corresponding sessions increased from an average of 2.6 mL per

session to 19 mL. Figure 4 contains cumulative records of drinking by rats 5FF and 6FF. The records for 5FF correspond to the first FT 60-s session containing post-pellet drinking (left), the session immediately after the first 5 g feeding (middle), and a FT 60-s session during weight reduction (right). The records for 6FF cor-

FIG. 3. Number of drinking bouts and total quantity of water consumed per session by all subjects during Experiment Two. The massed-food baseline condition is designated ‘‘BL.’’ All other sessions consisted of FT 60-s food delivery. The designation, ‘‘5 g,’’ on the graph indicates the point at which a 5 g postsession feeding occurred. Omitted data points indicate sessions in which data were spoiled due to experimenter error or equipment failure.

POLYDIPSIA WITHOUT WEIGHT REDUCTION

1389

FIG. 4. Cumulative records of drinking tube licks by subjects 5FF and 6FF. The records for 5FF show the first FT 60-s session containing post-pellet drinking (left), the FT 60-s session immediately after the first 5-g feeding (middle), and an FT 60-s session during the weight reduction condition (right). The records for 6FF show the first session with post-food drinking (left), a session prior to the first 5-g feeding (middle), and a session prior to the weight reduction condition (right). Downward deflections of the pen indicate FT-60-sec food deliveries. The pen reset after 450 licks.

respond to the first session with post-food drinking (left), a session prior to the first 5 g feeding (middle), and a session prior to the weight reduction condition (right). These records show that the characteristic postfood topography of SIP appeared before excessive consumption (in the first FT 60-s session for 6FF) and that bouts typically began within 5 s of pellet delivery. DISCUSSION

The results of Experiments One and Two demonstrate that weight reduction is not necessary for establishing or maintaining SIP in rats under standard experimental conditions. In all but one subject (5FF), the overall consumption by the subjects was comparable to levels reported (8 –22 mL/session) in procedurally similar experiments using rats reduced to 80 – 85% of their preexperimental free feeding weights (21,27,29). The finding that one standard-age subject (6FF) acquired SIP as rapidly as one of the younger subjects suggests that age and weight might not be reliably related to SIP acquisition (although the small number of subjects used prevents a firm conclusion in this regard). The functional independence of certain aspects of SIP (e.g., bout probability and duration) was demonstrated when the posthiatus probability of drinking remained unchanged in subjects which nevertheless showed a 37% decrease in overall consumption due to reduced bout durations. This decrease in posthiatus consumption appears consistent with the decay of a sensitized response over a period without elicitations (32). The lack of a similar effect in the free-feeding rats complicates this interpretation, however. Although the acquisition of SIP by the constant-weight rats was expected due to their relative weight reduction (21,29), the SIP literature suggested that SIP would not develop in non-weightreduced rats. This conclusion appears to be based primarily on the finding that previously weight-reduced rats stopped drinking excessively while regaining their preexperimental free-feeding weights (4,27). The present experiments, in contrast, examined the acquisition phase of SIP in young, growing subjects (which may have been effectively in a state of constant food deprivation) or standard-age subjects with relatively stable weights. Thus, a crucial factor in the development and maintenance of SIP might not

be the weight reduction level, per se, but changes in weight relative to overall growth and dietary requirements. Also relevant are non-weight-related factors that affect the consumption of the food pellets in the experimental setting. It is possible that the introduction of the pellets in the home cage before the experiment might have produced a kind of selective food deprivation during the experiment which enhanced the reinforcing value of the food pellets. This issue remains open, however, because such selective deprivations are an artifact of all procedures which manipulate the availability, scheduling, and type of food. Several implications arise from the present findings. The most fundamental is that it is no longer appropriate to state or imply that weight reduction by food deprivation is a necessary condition for SIP. Thus, theories which consider weight reduction to be a necessary condition for the acquisition of SIP may require modification. Activation (26) or frustration (24) theories, for instance, would have to allow either for the possibility that the central state mediating SIP exists in the absence of weight reduction or view the central state to be just one of a number of factors sufficient to engender SIP. Such modifications, would probably be not be major; classic mediational accounts already provide a model for explaining responding in the absence of drive (23). Moreover, both mediational and descriptive accounts will have to consider a more complete analysis of the roles and interactions of absolute and relative body weights, rates of change of body weight, growth rate, and perhaps other age-related and dietary factors. On a more procedural level, conducting studies of SIP without weight reduction might ease stress on subjects and eliminate the need for time-consuming monitoring of subjects’ weights. As a baseline for other measures, SIP without weight reduction might be more sensitive to certain variables than the apparently asymptotic SIP sometimes produced under weight reduction. A second set of implications arising from the present findings involves the issue of whether SIP should be conceputualized as a distinct molar phenomenon defined by excessiveness [i.e., a ‘‘third class’’ of ‘‘adjunctive’’ behavior (6,7,20,30)] or as a series of discrete drinking bouts elicited by individual food deliveries, thus making excessiveness an artifact of repeated occurrence (16). The

1390

TODD ET AL.

molar definition of SIP creates a conceptually difficult situation in which drinking bouts during sessions of nonexcessive overall consumption must be regarded as belonging to a different class of behavior than bouts which contribute to excessive consumption, even if the bouts in the two situations exhibit exactly the same temporal and topographical relationships to food pellet consumption. The conceptual status of the well-defined postfood drinking bouts by rat 5FF in most pre-weight-reduction FT 60-s sessions is thus questionable because overall consumption did not always exceed baseline levels; the response of the other rats in the early FT 60-s sessions is similarly indeterminate. Moreover, not only is there no agreement about how excessiveness is to be determined (31), the conceptual relevance of numerous other within- and between-sessions drinking measures is questionable under the excessiveness criterion because some of these measures (e.g., eatdrink latency) can be unrelated to overall response magnitude. Given the conceptual difficulties described above, as well as the general weakness of the argument that SIP does not resemble known types of repondent behavior (30), a reclassification of SIP might be warranted. As demonstrated here, the acquisition of SIP to resembles the development of a less-well known type of respondent behavior: sensitized elicited responding (30,32). In the present instance, postfood drinking develops gradually with repeated elicitations, exhibiting an increase in the probability of stereotyped drinking bouts and, as shown elsewhere, a correspond-

ing decrease in eat-drink latencies (14). Like sensitized responding, some dimensions of SIP are also diminished by a recovery period without elicitations (32). Thus, ‘‘schedule-induced polydipsia’’ might remain a useful procedural term to describe the effects of periodic food deliveries on water consumption, but the ‘‘induction’’ might be better conceptualized as the sensitization of ordinary food-related (prandial) drinking (13) by repeated food deliveries. This reclassification of SIP as a type of elicited behavior might set the occasion for a reevaluation of the adequacy of overall drinking measures in the analysis of SIP, closer examinations of the molar and molecular relationships between environmental manipulations and water consumption, and a more critical assessment of the correspondences between SIP and other ‘‘adjunctive’’ behaviors. ACKNOWLEDGEMENTS

This work was conducted at the Department of Human Development at the University of Kansas and partially supported by a Public Health Service Predoctoral Training Grant (HD07173-100111) to the first author. Parts of Experiment One are based upon an undergraduate honors thesis project conducted by the second author. The authors thank David G. Born, John C. Colombo, James A. Sherman, and Gina E. Truesdell-Todd for helpful comments and suggestions. Lisa M. Johnson provided the subjects for Experiment Two.

REFERENCES 1. Allen, J. D.; Butler, J. A. The effect of interplay interval on adjunctive behavior in humans in a game-playing situation. Physiol. Behav. 47:719 –725; 1990. 2. Cohen, I. L.; Mendelson, J. Schedule-induced drinking with food, but not ICS, reinforcement. Behav. Biol. 23:37– 44; 1974. 3. Falk, J. L. Production of polydipsia in normal rats by an intermittent food schedule. Science 133:195–196; 1961. 4. Falk, J. L. Conditions producing psychogenic polydipsia in animals. Ann. NY Acad. Sci. 157:594 – 609; 1969. 5. Falk, J. L. The nature and determinants of adjunctive behavior. Physiol. Behav. 6:577–588; 1971. 6. Falk, J. L. The origin and functions of adjunctive behavior. Anim. Learn. Behav. 5:325–335; 1977. 7. Falk, J. L. Schedule-induced behavior occurs in humans: A reply to Overskeid. Psychol. Rec. 44:45– 62; 1994. 8. Falk, J. L.; Tang, M. What schedule-induced polydipsia can tell us about alcoholism. Alcoh. Clin. Exp. Res. 12:577–585; 1988. 9. Fallon, J. H.; Allen, J. D.; Butler, J. A. Assessment of adjunctive behaviors in humans using a stringent control procedure. Physiol. Behav. 22:1089 –1092; 1979. 10. Hudson, R.; Singer, G. Polydipsia in the monkey generated by visual display schedules. Physiol. Behav. 22:379 –381; 1979. 11. Kachanoff, R.; Leveille, R.; McLelland, J. P.; Wayner, M. J. Scheduleinduced behavior in humans. Physiol. Behav. 11:395–398; 1973. 12. Keehn, J. D. Schedule-induced polydipsia, schedule-induced drinking, and body weight. Bull. Psychonom. Soc. 13:78 – 80; 1979. 13. Keehn, J. D.; Jozsvai, E. Topographical differences between scheduleinduced and prandial drinking by rats. Psychol. Rec. 39:247–262; 1989. 14. Keehn, J. D.; Stoyanov, E. The development of adjunctive drinking in rats: Conditioned and unconditioned components. Anim. Learn. Behav. 14:411– 415; 1986. 15. Lamas, E.; Pellon, R. Food deprivation and food-delay effects on the development of adjunctive drinking. Physiol. Behav. 61:153–158; 1997. 16. Lotter, E. C.; Woods, S. C.; Vasselli, J. R. Schedule-induced polydipsia: An artifact. J. Comp. Physiol. Psychol. 83:478 – 484; 1973. 17. Lucas, G. A.; Timberlake, W.; Gawley, D. J. Adjunctive behavior of the rat under periodic food delivery in a 24-hour environment. Anim. Learn. Behav. 16:19 –30; 1988.

18. Overskeid, G. Is any human behavior schedule-induced? Psychol. Rec. 42:323–340; 1993. 19. Peterson, M. R.; Lyon, D. O. Schedule-induced polydipsia in rats living in an operant environment. J. Exp. Analysis Behav. 29:493–503; 1978. 20. Roper, T. J. What is meant by the term ‘‘schedule-induced,’’ and how general is schedule-induction? Anim. Learn. Behav. 99:433– 440; 1981. 21. Roper, T. J.; Nieto, J. Schedule-induced drinking and other behavior in the rat, as a function of body-weight deficit. Physiol. Behav. 23:673– 678; 1979. 22. Singer, G.; Wayner, M. J.; Stein, J.; Cimino, K.; King, K. Adjunctive behavior induced by wheel running. Physiol. Behav. 12:493– 495; 1974. 23. Spence, K. W. Behavior theory and conditioning. New Haven: Yale University Press; 1956. 24. Thomka, M. L.; Rosellini, R. A. Frustration and the production of schedule-induced polydipsia. Anim. Learn. Behav. 3:380 –384; 1975. 25. Wallace, M.; Singer, G.; Wayner, M. J.; Cook, P. Adjunctive behavior in humans during game playing. Physiol. Behav. 14:651– 654; 1975. 26. Wayner, M. J. Motor control functions of the lateral hypothalamus and adjunctive behavior. Physiol. Behav. 5:1319 –1325; 1970. 27. Wayner, M. J.; Rondeau, D. B. Schedule dependent and scheduleinduced behavior at reduced and recovered body weight. Physiol. Behav. 17:325–336; 1976. 28. Wayner, M. J.; Singer, G.; Cimino, K.; Stein, J.; Dwoskin, L. Adjunctive behavior induced by different condition of wheel running. Physiol. Behav. 14:507–510; 1975. 29. Wayner, M. J.; Stein, J. M.; Loullis, C. C.; Barone, F. C.; Jolicoeur, F. B.; Rondeau, D. B. Effects of two body weight reduction regimens on schedule dependent and schedule induced behavior. Physiol. Behav. 21:639 – 645; 1978. 30. Wetherington, C. L. Is adjunctive behavior a third class of behavior? Neurosci. Biobehav. Rev. 6:329 –350; 1981. 31. Wetherington, C. L.; Riley, A. L. Differences in food consumption under intermittent and continuous reinforcement schedules of water delivery: Some implications for schedule-induced behavior. Anim. Learn. Behav. 13:331–337; 1985. 32. Wetherington, C. L.; Riley, A. L. Diminution of schedule-induced polydipsia after a long rest period. Physiol. Behav. 37:375–378; 1986.