EFFECTS OF PREFEEDING AND D AMPHETAMINE ADMINISTRATION UNDER A TWO COMPONENT MULTIPLE FIXED RATIO FIXED INTERVAL SCHEDULE IN PIGEONS By VANESSA MINERVINI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2015
Â© 2015 Vanessa Minervini
To my dog Semper for sitting by my side while I wrote this
4 A CKNOWLEDGMENTS This research was supported by the Department of Psychology, and Dr. Drake Morgan generously provided the d a mphetamine . I would like to thank all the members of my committee, Drs. Marc Branch, Jesse Dallery, Drake Morgan, and Neil Rowland , for always providing helpful comments and suggestions. I am support and guidance has been invaluable, and it has been an honor to be his last graduate student. Additionally, I am fortunate for the opportunity to have worked closely with Neil Rowland on a number of projects in his laboratory. I also would like to thank my former labmates Brian Kangas, Anne Macaskill, and David Maguire. Each of them taught me skills, without which I would have never been able to complete this project. I thank each of them for effectively shaping my behavior and for their continued contributions even long after their departures from the Branch Lab . Finally, I would like to th ank my army of undergraduate research assistants for helping with general animal husbandry and data collection. Yi Yang, Luis Otero Valles, Ellie Tsiskakis, Molly Hankla, and Julie Guitelman deserve special recognition for their dedication to the research. They were given detailed spreadsheets for data analysis and were asked to show up on numerous weekends. They never complained and never let me down .
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 7 LIST OF FIGURES ................................ ................................ ................................ ......................... 8 ABSTRACT ................................ ................................ ................................ ................................ ... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ...................... 12 Drug Effects and the Role of the Environment ................................ ................................ ...... 12 Behavioral Momentum Theory ................................ ................................ ............................... 15 Drugs as Disruptors ................................ ................................ ................................ ......... 17 Other Effects of d Amphetamine ................................ ................................ .................... 19 Drug Tolerance ................................ ................................ ................................ ....................... 22 Reinforcement Loss and Behavioral Tolerance ................................ ................................ ...... 22 Summary ................................ ................................ ................................ ................................ . 33 2 GENERAL METHOD ................................ ................................ ................................ ................ 36 Subjects ................................ ................................ ................................ ................................ ... 36 Apparatus ................................ ................................ ................................ ................................ 36 General Procedure ................................ ................................ ................................ .................. 37 Pretraining ................................ ................................ ................................ ....................... 37 Baseline Training and Baseline Conditions ................................ ................................ .... 37 Drug Preparation ................................ ................................ ................................ ............. 38 3 PHASE 1: PREFEEDIN G AND ACUTE D AMPHETAMINE ADMINISTRATION ............ 39 Rationale and Purpose ................................ ................................ ................................ ............ 39 Method ................................ ................................ ................................ ................................ .... 39 Prefeeding ................................ ................................ ................................ ........................ 39 Acute Drug Administration ................................ ................................ ............................. 39 Data Analysis ................................ ................................ ................................ ................... 40 Results and Discussion ................................ ................................ ................................ ........... 41 4 PHASE 2: CHRONIC ADMINSTRATION AND TOLERANCE ASSESSMENTS ............... 53 Rationale and Purpose ................................ ................................ ................................ ............ 53 Method ................................ ................................ ................................ ................................ .... 54 Subjects ................................ ................................ ................................ ............................ 54 Dose Selection ................................ ................................ ................................ ................. 54
6 Chronic Administration and Tolerance Probing ................................ .............................. 55 Post Chronic Saline Administration and Probing ................................ ........................... 55 Data Ana lysis ................................ ................................ ................................ ................... 56 Results and Discussion ................................ ................................ ................................ ........... 58 5 GENERAL DISCUSSION ................................ ................................ ................................ ....... 101 LIST OF REFERENCES ................................ ................................ ................................ ............. 106 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 117
7 LIST OF TABLES Table page 3 1 For all pigeo ns, the number of determinations for each dose during acute d amphetamine administration. ................................ ................................ ............................. 48 4 1 Number of injections each dose was administered to individual pigeons across Phase 2 conditions. ................................ ................................ ................................ ....................... 79 4 2 For each pigeon, the number of intervals in which zero responses occurred during acute and chronic administration. These data correspond to the index of curvature analysis shown in F igure 4 14. ................................ ................................ .......................... 80
8 LIST OF FIGURES Figure page 3 1 Overall response rate, expressed as pecks per seconds, as a function of prefeeding magnitude (left c olumn) and d am phetamine dose (right column). ................................ .. 49 3 2 Response rate, expressed as a percent of control response rate, as a function of prefeeding magnitude (left column) and d am phetamine dose (right column). ................ 50 3 3 Overall reinforcement rate, expressed as reinforcers per min, as a function of prefeeding magnitude (left column) and d am phetamine dose (right column). ................ 51 3 4 Reinforcement rate, expressed as percent of control reinforcement rate, as a function of prefeeding magnitude (left column) and d am phetamine dose (right column). ............ 52 4 1 Initial effects of the chronic dose, repeatedly administered for 30 sessions, on FR (filled symbols) and FI (open symbols) response rates (left panels) and reinf orcement rates (right panels). ................................ ................................ ................................ ............. 81 4 2 Response rates for FR (left column) and FI (right column) components as a function of d amphetamine dose. ................................ ................................ ................................ ..... 82 4 3 Total latencies for FR (left column) and FI (right column) components as a function of d amphetamine dose. ................................ ................................ ................................ ..... 83 4 4 Representative cumulative records (cumulative pecks over continuous time) for Pigeon 106 showing the effects of 3.0 mg/k g acutely and at two point s during chronic administration . ................................ ................................ ................................ ...... 84 4 5 Representative cumulative records for Pigeon 106 showing the effects of doses other than 3.0 mg/kg acutely and during chronic a dministration. ................................ ............... 85 4 6 Representative cumulative records for Pigeon 113 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. ................................ .................. 86 4 7 Representative cumulative records for Pigeon 113 showing the effects of doses other than 3.0 mg/kg acutely and during chronic administration. ................................ ............... 87 4 8 R epresentative cumulative records for Pigeon 400 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. ................................ .................. 88 4 9 Representative cumulative records for Pigeon 400 showing the effects of doses other than 3.0 mg/kg acutely and during chronic administration. ................................ ............... 89 4 10 Representative cumulative records for Pigeon 525 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. ................................ .................. 90
9 4 11 Representative cumulative records for Pigeon 525 showing the effects of doses other than 3.0 mg/kg acutely and during chronic administ ration. ................................ ............... 91 4 12 Representative cumulative records for Pigeon 674 showing the effects of 4.2 mg/kg acutely and at two points during chronic administration. ................................ .................. 92 4 13 Representative cumulative records for Pigeon 674 showing the effects of doses other than 4.2 mg/kg acutely and during chronic administration. ................................ ............... 93 4 14 Mathem atical index of curvature for individual FIs and the whole session during acu te and chronic administration. ................................ ................................ ...................... 94 4 15 Dose response functions for FR and FI response obtained after withdrawal of the chronic dose (denoted by filled triangles). ................................ ................................ ......... 95 4 16 Dose response functions for FR and FI latencies obtained after withdrawal of the chronic dose (denoted by filled triangles). ................................ ................................ ......... 96 4 17 Area under the curve (AUC) measures expressed as a proportion of acute saline or ................................ ................................ ................................ ......................... 97 4 18 Dose response functions for F R (left) and F I (right) reinforcement rates ......................... 98 4 19 Mean FR (top panels) and FI (bottom panels) and F I (top panels) response rates. ........... 99 4 20 Means for the relative difference in FR and FI responding in 10 sessions. ..................... 100
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment o f the Requirements for the Degree of Doctor of Philosophy EFFECTS OF PREFEEDING AND D AMPHETAMINE ADMINISTRATION UNDER A TWO COMPONENT MULTIPLE FIXED RATIO FIXED INTERVAL SCHEDULE IN PIGEONS By Vanessa Minervini May 2015 Chair: Jesse Dallery Major: Psy chology The purpose of the present experiment was to determine environmental factors that modulated the effects of d controlled behavior. The first phase of the study aimed to assess the effects of d amphetamine when admi nistered acutely, whereas the second phase emphasized the effects of chronic d amphetamine administration and the degree to which tolerance developed. The acute effects of a series of doses (0.1 10.0 mg/kg) on response rate were assessed and compared to the effects of presession feeding, and we determined that, under these conditions, d amphetamine did not produce effects similar to those intended to decrease hunger . Next, a single dose was selected for repeated administration prior to 30 consecutive ses sions on the basis that it decreased response rates in both components , but decreased the rat e of reinforcement in the fixed ratio component proportio nally more than in the fixed interval component. With daily administration of the selected dose continuin g, w e then reassessed the effects of the series of doses by administering a probe dose from the series every fifth session. Overall, tolerance developed in the fixed interval component for all pigeons and in the fixed ratio component for all but one pigeon , although the way in which it developed differed across the two components . W e also observed sensitization in the fixed ratio component for two
11 pigeons and correlates such as dose order, relative response rate, and patterns of initial disruption of respo nding were identified. Overall, our results indicated that the development of tolerance was not related to an attenuation of appetite suppression, nor did it depend solely on rate of reinforcement or loss of reinforcement. Rather the latter variables lik ely interact to alter normal response reinforcer relations maintained by the fixed ratio and fixed interval contingencies . Precise behavioral mechanisms remain to be determined by parametrically manipulating across a series of cumulative, integrated exper iments t he relevant variables in future research.
12 CHAPTER 1 INTRODUCTION Drug Effects and the Role of the Environment P sychoactive drugs have ph ysiological mechanisms of action at the level of receptor sites and neurotransmitters , and the effect s of a d rug on behavior were thought to depend solely on these neurobiological mechanisms of action prior to a discovery by Dews (1955). In his seminal article, Dews demonstrated that environmental manipulations can reveal interactions between drug and behavior t hat ultimately determine the behavioral effects. In his study, pigeons responded under a fixed interval (FI) 15 min schedule of food reinforcement and a fixed ratio (FR) 50 schedule of food reinforcement. These schedules engendered distinct baseline res ponse rates as well as patterns of responding. In the FI schedule, the first response to occur after 15 min e lapsed produced a reinforcer. Overall r esponse rates were relatively low and the pattern of response rates were low ea rly in the interval but accelerated across the remainder of the interval. In the FR schedule, the 50 th response produced a reinforcer, response rates were high, and the pattern of responding was a post reinforcement pause followed by constant, high rate of responding (termed a pattern ). Pentobarbital , a central nervous system depressant, then was administered prior to some sessions. Low doses elevated respon se rates for both schedules and higher doses decreased responding, with FI respond ing much more sensitive to the rate decreasing effects of pentobarbital. At two of the moderate doses tested (1.0 and 2.0 mg), FR response rates remained elevated, whereas FI response rates were decreased markedly. Therefore, environmental variables (e.g ., schedule of reinforcement) had to be taken in to account in order to characterize fully and predict the effects of pentobarbital on behavior .
13 A major aim of subsequent studies (e.g., Dews 1956, 1957, 1958 ; Morse & Herrnstein, 1956 ) was to determine pa rticular features of the schedules that contributed to the differential drug effects observed in the original experiment by Dews (1955) . In the original study, the type of schedule, control response rates, response patterning, and reinforcement rates all differed. Dews (1958) found that pentobarbital decreased , rather than increased, FR responding, but in this study the FR response requirement was 900 (which was substantially greater than the FR 50 used in the original study). Thus, the effects of the dr ug seemed to depend on the way in which responding was controlled by the schedule, not just the type of schedule. An additional aim of subsequent work was to test a variety of other pharmacological compounds, such as ce ntral nervous system stimulants, and Dews (1958) found that d amphetamine had effects opposite to those of pentobarbital. d A mphetamine increased response rates on FI 15 min and FR 900 schedules but decreased response rates on an FR 50 schedule and on a variable interval (VI) 1 min schedule . Using a multiple schedule, in which two or more schedules of reinforcement alternate with each schedule signaled by a distinct discriminative stimulus, became a preferred method to show that the rate increasing and rate decreasing drug effects even coul d be apparent in a single experimental session. Using a multiple FR 33 FI 5 min schedule, Smith (1964) demonstrated d amphetamine , while FR response rates dose dependently decreased. T he se within session effect s of d amphetamine necessarily depended on which of the two schedules FR or FI was presently signaled. In a number of replications, the effects of d amphetamine (and other stimul ants) on scheduled controlled behavior were related to baseline respo nse rates in a very orderly way. Responding on schedules that engendered low and high baseline response rates was incre ased and decreased,
14 respectively (e.g., Carney, 1 977; Gibson, 1967; Gonzalez and Goldberg 1977; Leander & McMillan, 19 74; McKearney, 1974; McMillian, 1969 , 1973 ; Owe n, 1960; Tilson & Sparber, 1973 ; Will & Checchinato, 1973 ), and t his phenomenon generated was known as the rate dependency hypothesis (for review see Sanger & Blackman, 1976 ) . A problem with a rate dependenc y interpretation is that rate of responding often is confounded by rate of reinforcement. Schedules that engender high er rates of responding typically have dense r reinforcement frequency , but reinforcement frequency typically is much leaner in schedules t hat engender low er rates of responding (e.g., Catania & Reynolds, 196 8 ). When reinforcement rate is held constant, however, results have shown that the effects of d amphetamine on responding still seemed to be dependent on control rates of responding, in dependent of schedule type or the reinforcement rate maintaining responding (Eckerman & Edwards, 1978; Luck i & DeLong, 1983, McPhail & Gollub, 1975; Sanger & Blackman, 1975; Stitzer & McKearney, 1977). Additionally, Lucki and Delong (1983) observed equal control response rates for random ratio schedules differing in rate of reinforcement and obtained no differences with respect to the effects of d amphetamine . Analyses of FI responding itself also seemed to provide further support for the rate dependency hypothesis . In the early portions of an FI, response rates are low but progressively increase as the interval elapses. Particular doses of d amphetamine increase d the low rates observed early in the interval and decrease d high er rates that occur red near reinforcer delivery at the end of the interval ( e.g., Bignami & Gatti, 1969; Dews, 1978 ; Odum, 2000 ; Odum, Lieving, & Schaal, 2002 ) . As outlined by Bran ch and Gollub (1974), these findings were actually contrived by the way in which the FI data were ana lyzed. Response rates in the early segments of the FI typically were derived from averaging all the FIs in a single session (i.e., multiple
15 repetitions). When the FIs were examined in more detail (an interval by interval analysis), the average rates we re not representative of performance in each FI instance. Branch and Gollub showed that averaging the data may be an inappropriate analysis because often the responding was bimodal: there was a period of either no responding or very low rates followed by a period of rapid respon ding (similar to the break and run pattern characteristic of FR responding). Averaging segments of the intervals across a single session allowed for periods of no responding to be averaged with periods of rapid responding, ultima tely producing intermediate rates and an he effects of d amphetamine were to increase the probability of responding early in the interval, but their analyses did not depict the rate decreasing effect towards the end of the interval. Given several others findings that also cannot be accounted for by the rate dependency hypothesis (e.g., Barret, 1976; Gonzalez & Byrd, 1977; McKearney, 1970, 1974; McKearney & Barret, 1975), Branch (1984 , 200 6 ) suggested that rather than focusing on the c orrelation between control response rates and drug effects, research should focus on behavioral mechanisms of drug action to determine how a drug interacts with factors that are known to normally con trol responding (i.e., under no drug conditions) . More r ecently, Pi tts (2014) emphasized this framework as well . U nderstanding how drugs modify ongoing relations between behavior and the controlling variables of behavior first requires an in depth and precise understanding of the controlling variables (for ex ample, reinforcement, reinforcement magnitude, reinforcement frequency, response reinforcer contingencies, stimulus reinforcer contingencies ), but it will be ultimately more informative in terms of predicting and controlling drug effects . Behavioral Moment um Theory One longstanding subarea of research that has earned much empirical support, proving to be extremely successful at achieving prediction and control of behavior, is known as beh avioral
16 momentum (Nevin, 1974). The behavioral processes involved in behavioral momentum (persistence of behavior in the face of disruption) have been extensively studied and are relatively well understood. Therefore, a behavioral momentum approach may be a useful for evaluating and interpreting drug effects. The concept of b ehavioral momentum was developed as a way to index response strength , and it is analogous to momentum in physics (more specifically, classical mechanics) . In physics, is the product of its velocity and mass. In the experimental a nalysis of behavior, steady state baseline response rate can be viewed as analogous to velocity and subsequent resistance to change in rates in the presence of disruptor variables (e.g., pre session feeding, extinction) as analogous to mass (Nevin, Tota, T orquato, & Shull, 1990). Using behavioral momentum as an index of response strength is considered superior to other measures because behavioral momentum theory holds that response strength can be indexed independently of baseline levels of responding. F or example, baseline response rates can depend on the type of reinforcement schedule, the parameters of that schedule, or both. Whereas response reinforcer contingencies (i.e., operant processes) determine baseline rates of responding, resistance to chang e is assumed to be determined primarily by stimulus reinforcer pairings (i.e., Pavlovian processes). Behavioral momentum theory asserts that behavior in the context of higher reinforcement rates, by virtue of stimulus reinforcer pairings, will be more res istant to change in the presence of disruptor variables. The generality of this prediction has been demonstrated in a number of studies and species (see Nevin & Grace, 2000; Nevin & Shahan, 2011). Behavioral momentum typically is examined by arranging a mu ltiple schedule, and once baseline responding is stable, disruptor variables are introduced . For example, Nevin et al.
17 (1990) arranged a multiple schedule consisting of two identical variable interval (VI) components, in which pigeons pecked illuminated k eys differing in color for access to grain. One of the VI components also was accompanied by response independent reinforcer deliveries on a variable time (VT) schedule. The added VT schedule weakened the response reinforcer contingency, such that baseli ne response rates were lower in the presence of the green key stimulus for that schedule. The overall rate of obtained food, however, was much greater relative to the other VI component (red key stimulus). Behavioral momentum was assessed with pre sessio n food access, extinction, and a combination thereof. The results showed that resistance to all three disruptor variables was greater in the component with the added VT food (i.e., in the presence of the green key) despite differences in baseline respondi ng. That is, decreases in rates relative to baseline rates were greater in the component with less stimulus reinforcer pairings. Drugs as Disruptors Given that baseline response rates have been demonstrated to be strongly correlated with drug effects (at least under some conditions) yet are purportedly irrelevant in determining behavioral momentum, one interesting approach is to use a behavioral momentum approach to evaluate drug effects. Under some conditions, drugs may produce effects that are compatib le with predictions made by behavioral momentum theory. Hoffman, Branch, and Sizemore (1987) arranged a multiple FR 5 FR 25 FR 125 schedule. Administration of cocaine produced the greatest and least disruption of responding in the FR 125 and FR 5 compone nts, respectively. That is, the most resistance to change was evident in the component associated with the richest rate of reinforcement (FR 5). With respect to d amphetamine , however, some studies have demonstrated that pharmacological disruptors produ ce different effects than those of traditional disruptors. Cohen (1986) conducted an experiment in which rats responded under a two component multiple
18 schedule with differing rates of reinforcement arranged in each fixed interval (FI) component: FI 30 s and FI 120 s. To evaluate the extent to which response rates were altered, he then administered a range of d amphetamine doses prior to some sessions. These results were then icated that d amphetamine produced comparable decreases in response rates under both FI components of the multiple schedule, whereas the proportion of baseline response rates was greater in the FI 30 component for all rats when extinction was implemented ( i.e., response rates were reduced more in the F I 120 component). Harper (1999) also aimed to assess whether behavioral momentum theory could account for drug induced disruptions in responding. He compared the effects of acute d amphetamine administrati on to those of inter component multiple variable interval (VI) variable interval plus variable time (VT) schedule (much like that arranged by Nevin et al. ) . Components, however, were arranged on separate lev ers, and a light above each lever indicate d which component was in effect . Alt hough inter component food delivery produced less disruption in the VI 30 VT 30 component relative to the VI 30 component, that is, producing the standa rd behavioral momentum result, the results from d amphetamine administration were inconsistent across subjects. Some rats showed no differences in responding between components and some showed a greater change in responding in the component associated wit were consistent with those reported by Cohen (1986) , and these studies indicated that the effects of d amphetamine could not be accounted for fully by behavioral momentum theory. Harper n oted one potential explanation: during d amphetamine administration rats tended to respond
19 more on the unsigna l ed lever (i.e., the currently inactive lever) than they did during baseline, indicating that d amphetamine degraded stimulus control by the light s . W ith little to no stimulus control, the n behavioral momentum determined by stimulus reinforcer pairings presumably would not be apparent because the Pavlovian pairings would not functionally occur . Additionally, the temporal pattern of behavior also may come under the control of stimulus reinforcer relations that govern resistance to change (Nevin & Grace, 2000). Therefore, d amphetamine patterning rather than merely a disruption of overall response rates (e.g., Branch & Gollub, 1974). Other Effects of d A mphetamine Cohen and Branch (1991) suggested that d amphetamine might enhance the effectiveness of stimulus food pairings in establishing that stimulus as a conditioned reinforcer. Quick and Shahan (in press) provided evidence that seemed to be consistent with that notion. In their study, r ats were treated with d amphetamine before sessions in one context (e.g., the right lever was active and illuminated, the house light flashed, and a t one sounded intermittently) and saline before sessions in another context (e.g., the left lever was active and illuminated, the house light was constantly illuminated, and a tone constantly sounded). The schedule of reinforcement was identical in both co ntexts as was the rate of reinforcement obtained. When extinction was implemented, more resistance to extinction was evident in the d amphetamine context relative to the saline context . They reinstated responding during extinction by providing response i ndependent food deliveries and found that the degree of reinstatement was greater in the d amphetamine context. These results indicated that d amphetamine enhanced the value of a particular context, and these results could not be a ttributed to any discri minative stimulus
20 properties of d amphetamine because drug administration was discontinued in the extinction and reinstatement sessions. d A mphetamine may also alter the effectiveness of food as a primary reinforcer. Establishing operation (EO ) and aboli shing operation (AO ) are terms proposed by Michael (1982, 1993). EOs and AOs have two effects: altering the curr ent value of a reinforcer and altering behaviors that have been established by response dependent presentation of that reinforcer. Food depri vation is an example of an EO in that it will increase the effectiveness of food as a reinforcer and evokes responses that have been associated with obtaining food in the past. Conversely, satiation is an example of an abolishing operation in that it wil l decrease the effectiveness of food as a reinforcer and abate responses that previously have been associated with obtaining food. Owen (1960) documented that for rats responding under an FR 30 schedule, satiation and d amphetamine administration produced similar cumulative response records : longer and more frequent post reinforcement pausing occurred . In his study, satiation was observed by extending the session duration until responding had ceased for 60 min . In the present study, e xperimenter administ ered pre session feeding was used to study satiation because it allows for better control over satiation variables . It also can be used for discrete trial tasks in addition to free operant tasks. Thurby and Samanin (1981) arranged a straight runway and mea sured response latencies and run times to assess reinforcer value . They found that 30 min access to food prior to the session increased both latencies and run times , but d amphetamine administration had no effect on runway performance. Because d am phetamine has been shown to increase general locomotor activity in other studies (e.g., Wentink, 1938) , limited conclusions can be drawn with respect to reinforcer effectiveness based on latency and run time.
21 Another measure of value is choice, or prefer ence, in a concurrent schedule of reinforcement. For example, d amphetamine larger, delayed reinforcers over smaller, sooner reinforcers ( Krebs & Anderson, 2012; Pitts & McKinney, 2005; Tanno, Maguire, Henso n, & France, 2014; Wade, de Wit, & Richards, 2000). In other words, d amphetamine decreases what is referred to as impulsive choice, which is the same effect that decreasing food deprivation tends to ha ve on choice (Christensen Szalanski, Goldberg, Anders on, & Mitchell, 1980). Simon, Gilbert, Mayse, Bizon, and Setlow (2009) presented rats with a choice between a 1 pellet reinforcer and a 3 pellet reinforcer. The larger reinforcer was not simply presented after a delay; instead, it was accompanied by a ce rtain probability of shock, making it the risky alternative. When d amphetamine was administered prior to sessions, it dose To determine whether this finding was due to the anorectic effects of the drug, Simon et al. then provided rats with 1 hr or 24 hr pre session access to food. Interestingly, these manipulations failed to affect preference for smaller safe versus larger risky reinforcers. Thus, decreased hunger or decreased motiv ation for the larger reinforcer did not seem responsible for the obtained results in this type of choice paradigm. Rather, d a mphetamine in this type of paradigm may exert its effects primarily through decreasing sensitivity to reinforcer amount , and this behavioral mechanism of drug action would be entirely different than satiation mechanisms . That is, it reduced the control of behavior by amount (e.g., Maguire, Rodewald, Hughes, & Pitts, 2009), and then behavior came more under the con trol of the shock contingencies . The type of operant task employed may influence the degree to which the many effects of d amphetamine described above are apparent. Therefore, in keeping with the essence of suggestions by Branch (1984) and Pitts (2014) , we must turn to ide ntifying behavioral
22 mechanisms of d amphetamine to understand more fully its effects and account for differences across studies . Drug Tolerance The development of drug tolerance is one of the criteria for substance use disorder (American Psychiatric Associ ation, 2013), but because, as noted earlier, drug effects, including tolerance, are not based entirely on pharmacologic processes (e.g., Demellweek & Goudie, 198 2 ; Schuster, Dockens, & Woods, 1966), identifying behavioral mechanisms of tolerance is essenti al to further the understanding of drug action and therefore perhaps addiction (i.e., substance use disorder) and the treatment of addiction (among other psychological and medical conditions). Tolerance refers to diminished sensitivity to the effects of a drug with repeated use. G iven the importance and implications of drug tolerance wi th respect to drugs of abuse and drug addiction, as well as to prescription drugs with therapeutic effects , surprisingly little is known about factors that determine whethe r tolerance develops . d Amphetamine is both abused ( see Clemow & Walker, 2014) as well as prescribed as a long term treatment for some medical conditions (e.g., Alder, Alperine, Leon, Faraone, 2014; De la Harran Arita & Garcia Garcia, 201 4 ; Heal, Smith, G osden, Nutt, 2013; Yanovski & Yanovski, 2014) . A primary goal should be to elucidate how the initial effects of a drug interfere with the normal relations between the behavior and its consequences. This same approach also can be applied in understanding the effects of chronic drug administration (i.e., repeated administration). One effect of chronic administration can be that tolerance develops (e.g., Branch & Dearing, 1982) , but chronic administration alone is not always sufficient. Reinforcement Loss and Behavioral Tolerance Tolerance to behavioral effects of drugs can develop differently depending on the way in which the initial effects of the drug interfere with the normal relations between the behavior and
23 its consequences. Schuster et al. (1966) i dentified b ehavioral factors that can be important determinants of the chronic effects of drugs. In their study, rats were train ed to respond under a multiple schedule of food reinforcement with FI and differential reinforcement of low rates (DRL) compone nts. When d amphetamine was administered acutely (i.e., initially), they observed dose dependent increases in responding for both the FI and DRL components. This increase in responding did not affect the rate of reinforcement in the FI component but decr eased the rate of reinforcement in the DRL component. Next, d amphetamine was administered chronically for 30 sessions. FI responding was elevated and remain ed elevated across those 30 sessions; DRL responding was elevated but eventually decreased, retur ning to control levels , so whether tolerance developed depended on the reinforcement schedule . Additionally, Shuster et al. used a shock avoidance schedule to demonstrate that tolerance did not develop to drug induced increases in responding that facilita ted reinforcement rate. Taken collectively, these findings le d Schuster et al. to propose the reinforcement loss hypothesis, which asserts that tolerance to a drug effect develops only if the initial effect of the drug is such that behavior becomes incomp atible with the prevailing contingencies of reinforcement. If the drug does not alter rate of reinforcement ( e.g., in the FI component) or enhances rate of reinforcement ( e.g., in shock avoidance) , then tolerance does not develop. Because Schuster et al . (1966) used a multiple schedule, they were able to disentangle evidence of behavioral tolerance from that of physiological tolerance. B oth the presence and the absence of tolerance were apparent in a single animal in a 60 min session. If physiological factors (e.g., changes in drug metabolism, changes in receptor sensitivity) alone determined tolerance, then one would expect those types of factors to remain relatively constant throughout the session rather than depend on the component of the multiple s chedule. T herefore, t he
24 development of differential tolerance observed by Schuster et al. was determined , at least in part, by behavioral factors. In outlining the reinforcement loss hypothesis, Schuster et al. describe d that tolerance can be viewed as a functional mechanism. The organism learns to adjust its behavior in order to comp ensate for the loss of reinforcement experienced in the presence of drug effects . The reinforcement loss hypothesis can account for the development of tolerance observed in a number of empirical studies (for reviews see Corfield Sumner & Stolerman, 1978; Wolgin, 1989). There seems to be just as many other studies in which findings are not substantiated by the reinforcement loss hypothesis, though (e.g., Dallery & Lancaste r, 1999; Harris & Snell, 1978; Hoffman et al., 1987; Minervini & Branch, 2013; Rees, Wood, & Laties, 1987; Smith, 1986; Weaver, Dallery, & Branch, 2010). For example, Hoffman et al. (1987) found that tolerance to the effects of cocaine developed in the F R 5 and FR 25 components but not the FR 125 component of a multiple schedule, despite the fact that reinforcement loss occurred in the FR 125 component as well. Smith (1986) emphasized that role of the context, or relative reinforcement loss, in determin ing the development of tolerance , rather than solely loss of reinforcement. Rats responded on a multiple schedule with DRL and random ratio (RR) components. When administered acutely, 1.0 mg/kg of d amphetamine increased DRL responding and decreased RR r esponding, thereby decreasing the rate of reinforcement obtained in both components. There was a proportionally greater loss of RR reinforcement, however. When this same dose was administered chronically, tolerance occurred in the RR component (i.e., res ponse rates recovered) but not in the DRL component (i.e., response rates remained elevated). When the RR
25 component was removed from the session and a multiple schedule no longer was in place, then tolerance developed rapidly to the drug induced increases in DRL responding. Presumably the dissociation between the effects of d amphetamine on DRL responding in a single but not multiple schedule materialized because the loss of reinforcement no longer was relative to an additional context, that is, the RR c omponent. In the multiple schedule, under baseline (i.e., no drug) conditions, a lower proportion of total reinforcers earned per session were coming from the DRL component relative to the RR component. Therefore, during chronic drug administration, res ponding in the DRL component may have been less sensitive to reinforcement loss , whereas responding in the RR component was adjusted to compensate for and recuperate the proportionally greater reinforcement loss . Dallery and Lancaster (1999) also suggested that relative changes of the reinforcement context, rather than an overall reduction in reinforcement frequency, predicted the extent to which tolerance developed to d amphetamine Each session , r ats were exposed to a multiple sch edule consisting of five VI components. The components rang ed from 8 to 250 s , and because each component was in effect for 8 min , the total and relative number s of reinforcers earned were greater in the richer VI components (with small schedule parameter s) than in the leaner . Acute administration of d amphetamine produced considerable decreases in responding for all five VI schedules but more reinforcers were lost in the richer VI components . Consistent with findings reported by Hoffman et al. (1987) an d Smith (1986), Dallery and Lancaster reported that the greatest degree of tolerance occurred in the two richest VI schedules when d amphetamine was administered chronically. In the studies described above, it is important to note that rate of responding a nd rate of reinforcement are not completely independent features of the schedule (e.g., Baum, 1993) . On
26 FR and RR schedules (i.e., Hof fman et al., 1987; Smith, 1986) the reinforcement rate is directly proportional to response rate. On VI schedules (i.e. , Dallery & Lancaster, 1999) changes in response rates affect the overall obtained rate of reinforcement at least until they reach some minimum value, at which point any subsequent increases in response rate have an impact on the rate of reinforcemen t. On DRL schedules (i.e., Schuster et al., 1966; Smith, 1986) small increases in response rates could produce great decreases in reinforcement rates. For example, a rat that responds once every 19 s on a DRL 20 s schedule would never contact a reinforce r but would contact all of the programmed reinforcers were it to respond once every 21 s. Continuing with the DRL 20 s schedule as an example, if response rates were to decrease further, say to one response every 40 s, then reinforcement rate would decrea se as well (the rat would contact about half of the programmed reinforcers). The only contingency specified by an FI schedule is that one response must occur after the interval elapses . O rganisms typically make many responses prior to the scheduled rein forcer , and even though the response requirement for larger FI parameters is identical to that for smaller FI parameters, the number of responses emitted usually increases as the FI parameter is increased (Ferster & Skinner, 1957). Therefore, if some inde pendent variable was implemented and found to decreases FI response rates substantially, the rate of reinforcement could remain completely unchanged so long as the animal responded at the minimal rate, specified by t he FI parameter. Because even marked de creases and increases in responding can have little to no effect of rate of reinforcement, FI schedules allow one to identify tolerance to the effects of a drug on response rate relatively independent of changes in rate of reinforcement. Additionally, be cause low er doses of d amphetamine , administered acutely, can increase responding whereas high doses can decrease responding on FI schedules, one can evaluate
2 7 whether tolerance to either or both of these effects is evident following chronic administration and whether tolerance depends on the initial effects of the dose selected for chronic administration. It may be the case that the degree of change or the direction of change in response rates induced by acute drug administration may determine the extent t o which tolerance ultimately develops. Tilson and Sparber (1973) evaluated the effects of chronic d amphetamine administration when the acute effect was either to decrease or increase response rates. One group of rats responded on an FR 30 schedule and re ceived 1.0 mg/kg prior to daily sessions. An acute dose response function was not determined, but the first administration (of chronic administration) produced a 60% decrease in response rates. A second group responded on an FI 75 s schedule, and after an acute dose response function was determined, a smaller dose 0.16 mg/kg was selected for chronic administration because this was the dose that reliably increased response rates at the beginning, middle, and end of the interval for all subjects. Toleran ce to the rate decreasing effect (for the FR group) and increasing effect (for the FI group) was evident in some rats after only four levels within one week . The tolerance t hat developed was comparable for FR responding and FI responding, but two different doses were used in two different groups of rats , complicating the interpretation . Brocco and McMillan (1983) arranged a multiple FR 30 FI 5 min schedule and found that FI response rates were less sensitive than FR response rates to the rate decreasing effects across a range of doses d amphetamine . T he chronic dose was administered following experimental sessions, such that within session responding was not disrupted. T ole rance was comparable for both of the components, and these results, as well as those reported by Tilson and Sparber (1973), were inconsistent with the reinforcement loss hypothesis (Shuster et al., 1966) .
28 That is, tolerance developed to the rate decreasin g effects despite an experience with no loss of reinforcement during chronic administration (i.e., Brocco & McMillan, 1983) and to rate increasing effects despite that behavior was not altered in a way that disrupted contact with FI reinforcement contingen cies (i.e., Tilson & Sparber, 1973) . Harris, Snell, and Loh (1979) arranged a multiple FR 30 FI 2 min schedule under which rats responded , and their effects appeared to be more in line with the p redictions of the reinforcement loss hypothesis . T hey ob served tolerance to the initial effects of d amphetamine on FR respond ing (i.e., the dose response function shifted up and to the right). FI responding was still decreased to the same extent by all three doses tested , and they reported that the decreases in FI responding were not accompanied by decrease s in reinforcement rate. The chronic drug regimen was highly idiosyncratic in their study , though . Typically during chronic administration a single dose is repeatedly administered prior to consecutive sess ions (e.g., Branch & Dearing, 1982), but Harris et al. (1979) added d amphetamine available in the home cage , presumably leading to different amounts of drug ingested by different animals , making it difficult to determine how ind ividuals were affected . They then altered the concentration of d amphetamine across weeks. For the most part, within session responding during the chronic regimen remained relatively undisturbed except that FR responding was greatly reduced or eliminated when the highest concentration of drug was in the water . Then, b efore redetermining the dose response functions (subcutaneously administered doses were given prior to sessions) , d amphetamine was removed from the water bottles for 12 days . The tolerance observed in their study is somewhat hard to interpret given that FR response rates (and reinforcement rates) were eliminated for approximately one week of oral d amphetamine ingestion and subsequently were at control levels for 12 days (no drug ingestion) ,
29 whereas no changes in FI response rates or reinforcement rates took place prior to redetermining the dose response functions. Although oral ingestion and subcutaneous injection of d amphetamine did appear to somehow interact in a schedule dependent mann er, it is not clear that those results can be attributed to differences in reinforcement rates and consequent recovery of responding to compensate for drug induced losses. Subsequent work by Harris and Snell (1980) attempted to addr ess some of those li mitations. Rats received a test dose of subcutaneous d amphetamine prior to a session, and this dose then was administered prior to seven consecutive sessions in order to see if the rate decreasing effects of this dose on responding would change as a resu lt of repeated administration . Some rats responded under a multiple FR 30 FR 5 min schedule, some under just FR 30, and the rest under just FI 5 min. in either component of the multiple schedule or eith er of the two simple schedules. It is unfortunate that the authors did not extend chronic administration or test other doses , as a careful look at their data suggests patterns may have been developing. Specifically, i n the multiple schedule, FR response rates were slightly higher in five instances (sessions) than they were initially when the test dose was administered; FI response rates remained unchanged from the effects of the test dose. For the simple schedules, a slight increasing trend in FI respon ding suggests that tolerance to the effects of d amphetamine may have been emerging, but FR responding remained at low levels with no trend. These data offer some support for the idea that tolerance may develop differentially for FR and FI responding and that the relative context likely is important. The test dose decreased control FR response rates by about 50% (from 1.1 per s to a little over 0.5 per s) when the FR was a component of the multiple schedule. In the simple FR schedule, the test dose also decreased response rates to 0.5 per s, but the control response rates
30 were much higher (almost 2 per s). For FI responding, control rates were reduced from 0.4 to 0. 2 per s in the multiple schedule and from 0.3 to 0.1 in the simple schedule. The data we re presented based on the mean performance of the 4, 3, and 5 rats from each condition, so whether the degree of disruption in rates is meaning ful (i.e ., representative of individual subject performance) remains to be determined. Reinforcement rate data were not presented by Harris and Snell (19 80), but it remains possible that there is an interaction between the degree of response rate disruption and degree of change in reinforcement rate that determines the development of tolerance. C onsistent with Sch uster et al. (1966) and Smith (1986), one might predict that given the context of the FR component, in which there was a proportionally greater decrease in responding and proportionally greater reinforcement loss, FI tolerance would not develop in a multip le schedule even if Harris and Snell (1980) had extended chronic administration to include more sessions . Conversely, if there were no change in overall response rate despite changes in reinforcement rate (either initially or during chronic administration ), then this would suggest that control response rates and patterning may be important determinates of tolerance. Again, because Harris and Snell (198 0) did not report reinforcement rate data, these accounts are speculative. When there no longer was a r elative reinforcement context, p erhaps the recovery of FI but not FR responding was more dependent on the distinct control response rates and patterning engendered by the schedules as opposed to sensitivity to reinforcement loss . Under FI contingencies, the probability of a response producing a reinforcer increases as the time since the last reinforcer increases. Under FR contingencies, no such relation exists the probability of reinforcement increases only as the number of responses emitted increases. Therefore, when the effect of a drug is to suppress FI responding as much as (or even more than) it suppresses FR
31 responding, relative to the distinct control rates maintained by each schedule, then the probability of contacting a reinforcer is substantia lly greater under the FI contingencies so long as the organism makes an occasional response between periods of non responding . complete suppression of the behavior being s tudied, this may prevent or at least delay the development of compensatory mechanisms which result in tolerance. This is particularly true if the compensatory mechanisms are behavioral in nature and require reinforcement for their strengthening and mainten Therefore, even when reinforcement rates are not disrupted under FI schedules, tolerance to the rate decreasing effects of a drug might occur simply because the responses are more likely to be reinforced, strengthened, and maintained. U nder FR schedules, tolerance usually does not develop when the dose selected for chronic administration is too high (i.e., the acute effects of this dose eliminate responding), but tolerance can develop to initially rate eliminating effects of high doses w hen the dose selected for chronic administration is relatively lower (Bowen, Fowler, & Kallman, 1993; Branch, Wilhem, & Pinkston, 2000; Minervini & Branch, 2013; Pinkston & Branch, 2004; Stafford & Branch, 1996). While reinforcement loss can generally bu t not always account for tolerance that develops under FR or DRL contingencies, the behavioral determinants of tolerance under FI contingencies seem to be much more complex than reinforcement loss (or even relative reinforcement loss). T olerance can occur not only to effects of the chronic dose, but to other doses as well, even when the chronic dose has an opposite effect on behavior than the other doses. For example, Schama and Branch (1994) showed that pigeons developed cocaine tolerance to the rate inc reasing effects of a dose that was administered chronically and to the rate decreasing effects of higher doses. Weaver and Branch (2008) reported that p igeons
32 developed cocaine tolerance to rate decreasing effects during chronic administration and to the rate increasing effects of lower doses. This may be because, under FI contingencies, the control patterns of responding as well as the way in which those patterns may be altered by different doses are more complex. Katz, Dworkin, Dykstra, Carter, and Wi tkin (1990) aimed to address whether the size of the dose selected for chronic d amphetamine administration modulated the development of tolerance differentially in FR and FI components of a multiple schedule. First , dose response functions were obtained for four pigeons. For two pigeons, the dose selected for chronic administration was 1.0 mg/kg. Acute administration of this dose produced only a slight decrease in FR response rates and no effect on FI response rates in one pigeon . Following chronic adm inistration, dose effect functions were re determined; the functions shifted up and to right for the FR component but were unchanged for the FI component. F or the other pigeon acute administration produced a moderate increase in FI response rates with no effect on FR response rates. Tolerance was observed for both components, but the changes in the dose response functions were different. For the FR component the function shifted up and for the FI component there was a general flattening of the function. Two additional pigeons received chronic administration of 5.6 mg/kg. Acutely, this dose produced large decreases in response rates, nearly eliminating responding in FR and FI components for both pigeons. Following chronic administration, the FR dose re sponse functions shifted marginally to right, indicating that tolerance primarily was restricted to the effects of the chronic dose. For the FI component, flattening of the dose response functions was observed . For one pigeon, the flattening of the funct ion provided evidence that tolerance could develop the rate increasing effects of some doses in the absence of repeated exposure to that particular effect.
33 For the other pigeon, the flattening of the function indicated that tolerance had developed the rat e decreasing effects of some doses, but the function also had shifted down. That is, when small doses of d amphetamine were administered f ollowing chronic administration, they produced lower response rates relative to the acute effects. T his pigeon diffe red from the other pigeon that received the 5.6 mg/kg in two ways. Baseline response rates were approximately twice as high, and rate increasing effects were not observed at any dose during acute administration. Similar downward shifts following chron ic d amphetamine administration also were reported by Branch (1979) . T he chronic dose administered to the monkeys in his study was one that substantially increased FI response rates , however. Dose specific tolerance to the rate increasing effects of the chronic dose undoubtedly had developed , but Branch noted that interpreting the downward shifts of the entire dose response function depe nded on the way in which tolerance was defined. He concluded that it was more appropriate to compare the effects of t he doses to the effects of saline rather than to the acute effects of the doses . When saline was were lower th an in baseline (i.e., prior to chronic drug exposure) . Thus, even though dose specific tolerance had developed, the chronic dose was not without effect entirely. Additionally , Branch demonstrated that changes in temporal patterns of responding were not necessarily accompanied by changes in overall response rates (and vice versa) . Summa ry C onflicting reports in the literature regarding the development of tolerance under FI contingencies may stem , in part, from problems with the dependent variable selected for measurement of tolerance . Acute d amphetamine administration alters the patter n of FI responding in very orderly ways that can be concealed when whole session analyses are
34 conducted ( e.g., Branch & Gollub, 1974). A ssessing changes in averaged, overall response rates might not be the most complete measure of tolerance to the effects of d amphetamine on FI responding , either . The existing literature offers much to be desired in terms of cohesive, comprehensive explanations of drug tolerance. Differen t results across studies seemed to be either disregarded by the authors or too easi selection of the chronic dose) . If one obtains results that are inconsistent with other reports in the literature , then subsequent experiments need to be more refined with potential c ontrolling variables need ing to be manipulated more systematically, such that knowledge is cumulative . The relation between control patterns of responding, acute drug effects, and reinforcement rate on FI schedules needs to be more fully characterized and compared to that of FR schedules. The present study was not intended to accomplish such a massive feat on its own, of course, but rather to gain a better understanding about necessary and sufficient environmental variables in modulating drug action the degree to which tolerance develops . To accomplish this , we arranged a 2 component multiple FR 30 FI 5 min schedule, under which pigeons responded for grain reinforcement. We were interested primarily in how response rate, rate of reinforcement, and contin gencies of reinforcement may interact to modulate the acute drug effects as well as the development of tolerance. T herefore, t he present study consisted of two phases. The aim of Phase 1 was to evaluate the effects of acutely administered d amphetamine o n response rate compared to effects of presession feeding . The aim s of Phase 2 were to evaluate the development of tolerance in the FR and FI component s during repeated administration of a single dose of d amphetamine , assess whether tolerance had develop ed to the effects of lower and higher doses, and determine whether tolerance was reversible following repeated saline
35 administration. A novel contribution of the present research was the moment by moment measurement that permitted analyses to be conducte d at many temporal scales.
36 CHAPTER 2 GENERAL METHOD Subjects All procedures and details of animal husbandry were approved by the University of Five experimentally naÃ¯ve adult male White Carnea u pigeons (numbered 106, 113, 400, 525, and 674) were obtained from Double T Farms, Glenwood, Iowa. They were maintained at 85% of their free feeding body weight via post session feedings consisting of mixed grain (Purina ProGrainÂ®) and pellets (Purina Pig eon Chow Checkers Â® ) in equal proportions. Outside of daily sessions, the pigeons were individually housed in a temperature and humidity controlled colony room with a 16:8 hr light:dark cycle. In their home cages, the pigeons freely had access to water and health grit. Apparatus The experiment was conducted in five pigeon operant conditioning chambers (interior dimensions 35 x 30 x 35 cm). The front panel of the chambers contained a houselight positioned 30 cm from the chamber floor, and 7 cm below the ho uselight were 3 horizontally positioned keys. The keys measured 2.5 cm in diameter and were positioned 8 cm from the walls and 5.5 cm apart. Only the center key illuminated white or red was used; its activation required a minimum force of 0.098 N. A 30 ms tone, produced by a Mallory SonalertÂ®, accompanied each successful response to the key. Located 10 cm directly below the center key and 11 cm from the chamber floor there was a 5 cm by 5.5 cm aperture thro ugh which access to a solenoid operated hopper fill ed with buckwheat, milo, and hempseed could be provided. During reinforcer presentations, the aperture and raised hopper were illuminated for 3 s , and all other lights in the chamber were extinguished. In one of the chambers, an additional 5 cm by 5.5 cm a perture for pellet delivery was located 3.5 cm to the left of the grain aperture, but it served no function in
37 this experiment. White noise at approximately 95 dB masked extraneous sounds in the rooms containing the chambers. In an adjacent room, computers running EC BASIC software (Palya & Walter, 1993) controlled experimental events and recorded data, and cumulative response recorders provided live time data collection. General Procedure Pretraining After the pigeons reliably approached and ate from the r aised hopper, responses to the center key (illuminated white) were shaped by reinforcing successive approximations. Once key pecking was established, the next session consisted of an FR 1 schedule , with the center key illuminated white , and lasted until the pigeon earned 60 reinforcers. Baseline T raining and Baseline Conditions To arrange a two component multiple FR FI schedule, we introduced a second key color red. A 5 min blackout occurred at the start of all sessions, and t he white key and the red ke y always signed the FR and FI components, respectively. There were 30 s inter component intervals (ICIs), during which all lights in the chamber were extinguished. The components strictly alternated after each food presentation , and sessions ended after 20 presentations of each component type . A 1 min limited hold was in effect for both components. That is, i n order to earn a reinforcer (i.e., 3 s access to grain), the pigeon had to complete the FR response requirement within 60 s in the presence of the white key and emit a single peck within 60 s of the FI timer elapsing in the presence of the red key. If the pigeon contacted the limited hold contingency, then a reinforcer was not delivered and the 30 s ICI was initiated. T o achieve the targeted te rminal schedule F R 30 FI 5 min t he FR and FI requirements in each component were increased gradually across seven sessions. The progression was as follows: FR 2 FI 1 min; FR 5 FI 2 min; FR 8 FI 3 min; FR 12 FI 4 min; FR 20 FI 5 min; FR 25
38 FI 5 min; FR 30 FI 5 min. Each pigeon then completed 50 baseline sessions at the targeted terminal schedule such that responding was stable for at least 30 sessions prior to the start of any subsequent experimental manipulations. Drug P reparation d A mphetamine sulfate, g enerously provided by the laboratory of Dr. Drake Morgan in the Department of Psychiatry at the University of Florida, was dissolved in 0.9% saline solution. Doses were delivered in mg/kg, expressed in terms of the sulfate , and administered intramuscularl y (pectoral muscle) immediately prior to a session. During daily administration, the injection site alternated between sides of the breast. Dose volume was determined by 0.1% feeding body weight. For example, Pigeon 106 was maint ained at 514 g and always received a 0.51 mL injection.
39 CHAPTER 3 PHASE 1: PREFEEDING AND ACUTE D AMPHETAMINE ADMINISTRATION Rationale and Purpose Anorectic effects of d amphetamine have been well established (e.g., Caul, Jones, & Barrett, 1988; Kornbl ith & Hoebel, 1976; Milloy & Glick, 1976; Wellman, Davis, Clifford, Rothman, & Blough, 2008), but whether pre session feeding and d amphetamine have similar effects on the effectiveness of food as a reinforcer has not yet been well established. Therefore, the purpose of this phase was to compare the effects of prefeeding with those of d amphetamine administration from a behavioral momentum perspective to see if d amphetamine function ed as an abolishing operation (AO) to exert its effects on food maintained responding . Method Prefeeding M omentum tests were conducted using pre session feeding , during which the pigeon consumed Purina ProGrainÂ® in the homecage starting 45 min prior to the session . Every pigeon always finished the entire meal within the allotte d time. This pre feeding took place every third session provided that the pigeon was at its 85% free feeding weight . Pigeons received three different meal magnitudes , and each meal magnitude was assessed twice before introducing the next magnitude. For a ll pigeons, t he first meal magnitude was 25 g, and the second meal magnitude was 50 g. For Pigeons 106 and 113, the third meal magnitude was 12.5 g, and for Pigeons 400, 525, and 674 it was 75 g. Acute Drug A dministration Pigeons received an intramuscu lar (pectoral) injection of d amphetamine immediately prior to every fifth se ssion. Doses were administered in either an ascending order (Pigeons 106 and 113) or descending order (Pigeons 400, 525, and 674) . The series consisted of 0.0 (i.e.,
40 saline), 0. 1, 0.3, 1.0, 3.0, and 5.6 mg/kg. After all the probe dose s from the series had been administered, the entire series was repeated, yielding at least two determinations of effects of each dose. Additional doses were administered to individual pigeons as needed to complete the curve. For each p igeon the number of times a doses was administered is shown in Table 3 1. Data A nalysis Responses per second in each component were determined for each session. Overall response rate was defined as the total number of FR and FI responses emitted in the presence of the corresponding discriminative stimuli divided by seconds each key color was illuminated per session. The response rates were then averaged for the two determinations at each meal magnitude as well as f or the number of determinations at each d amphetamine dose. Next, the response rate s were reanalyzed as a percent of control responding to normalize the absolute differences in response rates and assess relative disruption of rates, independent of baseli ne levels . Control values were average s taken from the six total baseline sessions that immediately preceded a momentum test (i.e., prefeeding) , or from each session that imme diately preceded a drug probe (i.e., acute administration). T otal number of con trol sessions v aried across individual pigeons, but can be derived from the information in T able 3 1. Additionally, we evaluated the extent to which prefeeding and d amphetamine administration affected the rate of reinforcement , expressed as the number o f reinforcers earned per minute for the FR and FI components . That is , reinforcement rate was the total number of FR and FI components (out of a possible 20 each) ending in reinforcer delivery divided by the total minutes each key color was illuminated per session.
41 Results and Discussion Figure 3 1 shows overall response rate s as a function of prefeeding magnitudes (left column) and d amphetamine doses (right column) . Individual pigeons are shown by row, and FR and FI components are denoted by filled circ les and open squares, respectively. For all pigeons, control FR rates were higher and stayed higher than FI rates across at all three meal magnitudes tested. In addition, for four of the pigeons, the slope of the function for FI is steeper than that for the FR. A ll pigeons FR rates were maintained well above control FI rates even at the largest meal magnitude. At first glance, it is tempting to conclude that prefeeding and d amphetamine produced similar effects because FR rates exceeded FI rates across the range of doses administered. In order to compare the effects properly , however, Figure 3 2 must be examined in conjunction with Figure 3 1. As predicted by behavioral momentum theory (Nevin, 1974) , behavior in the component with the richer reinfor cement history will be more resistant to change , by virtue of stimulus reinforcer relations respondent processes i ndependent of absolute diff erences in baseline responding. This sometimes is referred to as In the FR component, p igeons earned a reinforcer by emitting 30 pecks , and pigeons typically responding at about 3 pecks per second (Figure 3 1 ). It took them, on average, 10 s to complete the ratio requirement. Conversely, in the FI component reinforcers were spaced by at mi nimum 5 min. Figure 3 2 shows the response rates expressed as a percent of control responding , and it indeed reveal ed that responding in the FR component had greater relative resistance to the effects of prefeeding for all but Pigeon 113. The same was not true when d amphetamine doses were administered. At higher doses that produced decreases in both FR and FI response rates (the descending limbs of the dose response functions, Figure 3 1), there did not appear to be differences in relative decreases ( relative to control levels, Figure 3 2). At lower doses, baseline response rates were elevated in
42 the FI component for Pigeons 106, 674, and, to a lesser extent, 400 . At these doses, FR rates either remained at control levels or decreased only marginall y. There were some doses at which FI responding was less sensitive to the rate decreasing drug effects (i.e., more resistance than FR responding), such as 1.7 mg/kg for Pigeon 113 and 3.0 mg/kg for Pigeons 106 and 525. The effects obtained with d amphet amine were not predicted by behavioral momentum theory. From this we can conclude that d amphetamine does not seem to exert its effects on operant behavior by diminishing food as a reinforcer. Although amphetamines have known anorectic pr operties at least under some conditions (e.g., Caul et al., 1988; Demellweek & Goudie, 1982; LeSage, Stafford, & Glowa, 2004; Owen, 1960 ), our study in a very straightforward manner demonstrated that the behavioral mechanism of drug acti on for food maintai ned operant responding was not related to satiation like processes . In fact, even under ad libitum feeding conditions , the mechanisms of amphetamines may not be directly related to reducing the value of food (LeSage et al., 2004). A possible behavioral mechanism that may account for the present findings is that d a mphetamine reduce d sensitivity to the rate of reinforcement ( e.g., Jimenez Gomez & Shahan, 2007; Ta, Pitts, Hughes, McLean, & Grace , 2008). If the duration of the components (and therefore rein forcement rate) no longer became a controlling variable of responding under d amphetamine administration then this could explain why FI rates were more disrupted than FR rates following prefeeding but were less disrupted than FR rates under the effects of d amphetamine. This was not true at every dose for all pigeons, however, so pe rhaps the effect would have been more pronounced if the reinforcement rates were made to be more similar. Our study may have concealed sensitivities to reinforcement rates beca use the rates were so disparate (i.e., roughly once per 10 s in the FR versus once per 5 min in the FI component). This notion
43 could be examined in future studies and behavioral momentum procedures can be used as a tool to evaluate suspected behavioral mec hanisms of drug action, such as decreased control by rate of reinforcement. A lternative ly, the nature of the contingencies, rather than the duration of the components (and therefore reinforcement rate ) , may have differentially affected relative resistance . Typically momentum is assessed using multiple schedules with alternating VI VI components in order to keep reinforcement rates constant and allow response rates to vary . Conversely, r atio schedules usually are not used to assess momentum given that cha nges in reinforcement rate (i.e., the stimulus food pairings that determine resistance to change) are directly proportional to response rates. As suggested by Branch (2000), local contingencies could be important determinants of resistance to change in so much as they are important determinants of high response rates typically maintained by ratio schedules . That is, under ratio contingencies the p robability of a response being reinforced depends only on the number of responses made , so, under these cont ingencies, most of the reinforced inter response times (IRTs) are short, few are long . Under interval schedules, by contrast , the probability of a response being reinforced increases with the time since last response. IRT analyses indicate that most are relatively long (few are short) and reinforced IRTs tend to be longer than overall average IRTs (e.g., Baum, 1993; Catania, Matthews, Silverman, and Yohalem, 1977). Nevin , Grace, Holland, and McLean (2001 ) arranged a direct test of whether the differe nt response rates and IRT distributions observed under VR and VI contingencies determine momentum. The rate of reinforcement was held constant by yoking the VI parameter values to the inter reinforcer intervals that occurred in VR component. Under these conditions, behavioral momentum theory predicts that disruption in responding should be equal for both components
44 because the frequency of reinforcement was identical. Results showed that b aseline rates in the VR component were approximately twice as high as those in the VI component, but VI responding was more resistant to disruption. Thus, momentum appeared to be related to the way in which the contingencies were arranged rather than strictly the rate of reinforcement. Nevin et al. (2001) postulated th at interval schedules may establish greater resistance to change because response rates can decrease without greatly decreasing the rate of reinforcement. Interestingly, a much earlier study conducted by Eckerman and Edwards (1978) showed similar finding s when d amphetamine was administered prior to sessions. Pigeons responded under a multiple FR FI schedule, with FI intervals yoked to the inter reinforcement intervals from the FR component, thus rate of reinforcement was equivalent. Consistent with Ne vin et al. (2001), they found ratio schedules maintained higher rates of responding under control conditions, but responding under the interval schedule was much less disrupted by drug administration. They attributed their results to rate dependency effec ts of d amphetamine , that is, that higher rates will be decreased proportionally more than lower ones . These findings not only are consistent with, but also extend, the conclusions of Nevin et al. (2001) . T he way in which a disruptor ( e.g., d amphetamine ) impacts the typical relation between responding and response contingencies in a given context can determine resistance to change. This interaction becomes especially evident when reinforcement rates are controlled across contexts. The r einforcement rat e was not held constant , rather was quite different, across the FR and FI components in our study, and Figure 3 3 shows the obtained overall reinforcement rate s in each component . Between the control conditions of prefeeding and the control conditions of acute drug administration, the rate of reinforcement remained relatively unchanged for all pigeons. In the FI component, the reinforcement rate was comparable for all pigeons (0.19
45 reinforcers per min ) . The rate of reinforcement in the FR component was s ubstantially richer than in the FI component (it ranged from 4.5 reinforcers per min [Pigeon 113] to 7.3 reinforcers per min [Pigeon 400] ). Under prefeeding conditions, the rate of reinforcement in the FR component always exceeded that in the FI component . This also was true for two pigeons (400 and 674) under d amphetamine administration, but for the other pigeons (106, 113, and 525) the rate of reinforcement was comparable in both components at the highest doses tested. Because the maximum programmed rate of reinforcement in the FI component could not exceed 0.2 reinforcers per min independent of response rate and the rate of reinforcement in the FR component necessarily varied with response rate, Figure 3 4 shows the reinforcement rates expressed as p ercent of control for each component. Prefeeding did not affect the rate of reinforcement differentially across components for Pigeon 106 or 400 . For Pigeons 525 and 674 reinforcement rates were undifferentiated, except at the highest magnitude (75 g), a t which FI reinforcement rates were less affected than FR reinforcement rates. Reinforcement rates decreased less in the FI component for Pigeon 113 at the highest (50 g) and moderate (25 g ) magnitude, too. For all pigeons during drug ad ministration FI r einforcement rates overall appea red to be less disrupted than FR reinforcement rates. This is especially evident if the descending limbs of the curves shown in Figure 3 4 are compared to the descending portions of the response rate curves shown in Figures 3 1 and 3 2. The overall picture is one in which behavior in the component where reinforcement rate decreases were more pronounced (i.e., the FR component) was less resistant to disruption than that in the FI component. Thus when rates of reinforcement are unequal across component s the degree to which the contingencies permit changes in responding to alter reinforcement rates could contribute to resistance to the effects of disruptors. To the extent that drug effects can be considered di sruptors, t his h as important
46 implications with respect to the development of drug tolerance. Minervini and Branch (2013) reported that greater tolerance to the rate decreasing effects of cocaine developed w hen reinforcement rate s were higher and responding was less disru pted by acute drug administration. These effects were not associated with higher levels of baseline responding, but because FR schedules were used in their study, the independent and combined contributions of reinforcement rates and relative resistance co uld not be separated. Other studies also have found inconsistencies with behavioral momentum predicting the acute effects of drugs, however. For example, Cohen (1986 ) reported that the effects of extinction on resistance to change under FI 30 and FI 120 components were consistent with behavioral momentum theory greater persistence in the richer component . When d amphetamine was used as a disruptor, however, it did not produce consistent effects, and sometimes even produced greater persistence in the leaner component (i.e., FI 120). P harmacological disruptors can produce different effects than those of traditional disruptors, such as prefeeding and extinction (e.g., Harper, 1999; Heyman, 1983 ). The results of the present study are consistent with t hose findings as well. Rather than necessarily considering this as a limit of the scope of behavioral momentum theory, one might consider it a potential opportunity for identifying behavioral mechanisms of drug action (e.g., Pinkston, Ginsburg, & Lamb, 2 009, 2014) . If the disruption produced by a specific pharmacological compound is found to resemble that produced by extinction , for example, but not by presession feeding then that would seem to suggest that a behavioral mechanism could be a reduction in sensitivity to certain consequences. This then could be further examined by comparing drug induced versus extinction induced disruption to qualitatively different reinforcers (e.g., types of food, drug reinforcers, or access to littermates , etc. ) . Altern atively, responding could be brought under the
47 control of negative reinforcement contingencies differing in reinforcement rate, and then the disruption effects of the drugs could be compared to the disruptive effects of escape extinction. From the standp oint of detecting and characterizing environment behavior mechanisms, it may prove useful to i dentify and develop an understanding of factors as well as some parametric limits of those factors that modulate whether pharmacological disruptors produce effect s similar to those of traditional disruptors.
48 Table 3 1. For all pigeons, the number of determinations for each dose during acute d amphetamine administration. Dose (mg/kg) Pigeon 106 Pigeon 113 Pigeon 400 Pigeon 525 Pigeon 674 0.0 2 2 2 2 2 0.1 2 2 2 2 2 0.3 2 2 2 2 2 1.0 2 3 2 2 2 1.7 3 2 2 2 2 2.4 3 3.0 2 3 2 4 5 4.2 3 3 5.6 3 2 2 2 4 10 2
49 F igure 3 1. Overall response rate, expressed as pecks per seco nds, as a functi on of prefeed ing magnitude (left column) and d amphetamine dose (right column). FR rates are denoted by filled circles and FI rates are denoted by open squares. Each row represents an individual pigeon. Note x and y axes are logged and y axis scales di ffer .
50 Fi gure 3 2. Response rate, expressed as a percent of control response rate, as a function of prefeed ing magnitude (left column) and d amphetamine dose (right column). FR rates are denoted by filled circles a nd FI rates are denoted by open squares. Each row represents an individual pigeon. Note x axes are logged .
51 Figure 3 3. Overall reinforcement rate, expressed as reinforcers per min, as a function of prefeeding magnitude (left colum n) and d amphetamine dose (right column). FR reinforcement rates are denoted by filled circles and FI reinforcement rates are denoted by open squares. Each row represents an individual pigeon. Note x and y axes are logged and y axis scales differ.
52 Figure 3 4 . Reinforcement rate, expressed as percent of control reinforcement rate, as a function of prefeed ing magnitude (left column) and d amphetamine dose (right column). FR rates are denoted by filled circles and FI rates are denoted by open squares. Each row represents an individual pigeon. Note x axes logged .
53 CHAPTER 4 PHASE 2: CHRONIC ADMINSTRATION AND TOLERANCE ASSESSMENTS Rationale and Purpose The reinforcement loss hypothesis provides an account for the deve lopment of behavioral tolerance as distinct from physiological tolerance : t olerance should develop to the effects of a drug when responding becomes incompatible with the prevailing contingencies of reinforcement (Schuster et al; 1966; Schuster, 1978). Th e results of previous studies are mixed with respect to the development of tolerance to the effects of d amphetamine on FI responding (e.g., Harris & Snell, 1980; Katz et al., 1990; Schuster et al., 1966; Tilson & Sparber, 1973) and cannot be attributed to reinforcement loss in many instances (e.g., Branch, 1979). At present, it is difficult to identify narrowly the variables that modulate whether and the degree to which tolerance develops , as several features differed across studies, such as components of the multiple schedule, schedule paramet ers, duration of chronic drug administration, dose selected for chronic drug administration, baseline response rates, and acute drug effects (to name a few). Moreover, when findings were replicated (or partially rep licated) across studies, o ther possibly relevant variables (e.g., response patterning, rates of reinforcement) were not the primary focus so were not reported , making it impossible to ascertain whether potential confounds existed across studies . Overall t he literature seems to be demonstrations . The aim s of this phase of the experiment were to identity environmental factors that modulated (a) the development of tolerance to rate decreasing effects of d amphetamine during repeated administration (b) the ext ent to which tolerance was evident when the effects of higher and lower doses were assessed and (c) the extent to which the effects of repeated administration were attenuated following withdrawal of the chronic dose. Accomplishing these aims and contribu ting to the existing literature hinged primarily on the deliberate selection of a dose to be
54 administered chronically. The chronic dose was selected based on its acute effects. Specifically, this dose differentially affected responding under FR and FI co mponents of a multiple schedule in terms of absolute response rates, relative response rates, and relative reinforcement rate. To this end, we arranged to have the opportunity to examine the development of tolerance by reinforcement, thus providing a mor e complete look than most previous work. Method Subjects The same five pigeons used in Phase 1 served as the subjects for Phase 2. Phase 2 was initiated immediately upon completion of Phase 1. Dose S election Based on the acute dose response functions o btained in Phase 1 (Figures 3 1 through 3 4 ) and visual analysis of cumulative response records, a dose was selected for repeated administration. For individual pigeons, we selected the largest dose from the series that satisfied all of the following crit eria: (a) response rates were diminished in both the FR and FI components but not reduced to zero (b) the absolute response rate w as lower for the FI component (c) the response rates in the FR and FI components relative to corresponding control rates were similar and ( d ) the rate of reinforcement in the FI component was decreased proportionally less than the rate of reinforcement in the FR components. For Pigeons 106, 113, 400, and 525 this dose was 3. 0 mg/kg, and for Pigeon 674 it was 4.2 mg/kg. Note th at the 5.6 mg/kg dose satisfied the criteria for dose selection with Pigeon 674. We opted not to use this dose based on previous anecdotal observations in our laboratory . After only 14 sessions of chronic administration of 5.6 mg/kg, a nother pigeon (no t one of the five subjects from the present experiment) unexpectedly was found deceased (reasons unknown) , so we erred on the side of caution by proceed ing with the 4.2 mg/kg dose for Pigeon 674 .
55 Chronic A dministration and Tolerance Probing To b egin the study, p igeons received the chronic dose, injected intramuscularly, prior to every session for 30 consecutive sessions. As chronic administration then continued, t o determine whether and the extent to which tolerance developed , we reassessed the effect of the series of doses by occasionally substituting higher and lower doses for the chronic dose. Probe doses were administered every fifth session and in the same order as acute administration. On sessions intervening probes, pigeons received their chronic dose. After two determinations of each dose from the series, we sometimes replicated the effects of doses at critical points in the curve. Table 4 1 shows the total number of determinations at each dose as well as the total injections received of the chr onic dose. All pigeons first received 30 days of repeated administration, but the number of chronic doses during tolerance probing could vary as the chronic dose was administered four times for every probe dose and depended on the number of administration s of probe doses. Rarely (at least once but no more than three times for every pigeon) an apparatus failure caused a session to terminate abruptly and the data to be stored erroneously, so some sessions had to be repeated during tolerance probing . I n most instances, the apparatus failure s occurred i n sessions intervening between probe doses (i.e., the chronic dose was administered); but if it occurred for a probe dose, then four additional chronic doses were administered prior to repeating that probe dose. Post Chronic Saline Administration and Probing Upon completing the dose response functions, administration of the chronic dose was withdrawn, and p igeons instead received a saline injection prior to every session for 30 consecutive sessions. Then, to determine whether the effects observed under chronic administration were reversed or attenuated, we again assessed the effect of the series of doses .
56 A dose from the series was substituted for saline every fifth session, the order of the doses remain ed constant, and saline continued to be administer ed on intervening sessions. See Table 4 1 for the number of determinations at each dose for individual pigeons. Data Analysis The primary dependent variables were overall response rates , reinforcement rat e, and total latencies for both component s . Overall response rate s were defined as the total number of FR and FI responses emitted in the presence of the corresponding discriminative stimuli divided by seconds each key color was illuminated per session . Similarly, reinforcement rate was the total number of FR and FI components (out of a possible 20 each) ending in reinforcer delivery divided by the total minutes each key color was illuminated per session. A latency was d efined as the time from presentati on of a discriminative stimulus until the first response was emitted. If a pigeon failed to respond in any of the 40 components per session (20 FR and 20 FI), then the latency was set equal to a maximal value. This value was the limited hold (i.e., 60 s) for the FR and the limited hold plus the interval parameter (i.e., 360 s) for the FI. The 20 FR and 20 FI latencies were summed for each session. Dose response functions for rates and latencies in both FR and FI schedule components were constructed. W e used the mean effects of at least two determinations of each probe dose during chronic administration (i.e., when probes were substituted for the chronic dose) and did the same for post chronic administration (i.e., when probes were substituted for salin e). The axis because the controls were the effects of the chronic dose and the effects of saline (i.e., post chronic) on sessions that took place between each probe dose. As such, the means at these va lues were based on many determinations (excluding the 30 consecutive session prior to the start of probing). Table 4 1 provides the number of determinations at each dose for individual pigeons.
57 Original c opies of the cumulative response records generated by a mechanical recorder were reconstructed electronically using the time stamped raw data stream of within session events. This allowed the response pen to be reset after a component ended and be colored distinctly for FR and FI compone nts . Both of these techniques were incorporated to aid in visual analysis of response patterning. To quantify response patterning in the FI component, we determined the mathematical index of curvature (Fry, Kelleher, & Cook, 1960 ), which is a measure of the relative distribution of responses across FIs. FIs were divided into ten 30 s bins, and the following equation from Fry et al. was used to determine the index of curvature for each of the 20 FI repetitions per session: R 1 is the total number of responses that occurred in the first bin, R 2 is the total number of responses that occurred in the first and second bins, R 3 is the total number of responses that occurred in the first, second, and third bins, and so on until R 10 , which is the total number of responses that occurred in all bins. The possible range of the index is 0.9 to +0.9, where an index of 0.9 indicates that all responses occurred in the first bin (i.e., response rate slow ed across the interval) , an index of 0 indicates that an equal number of responses occurred in each bin (i.e., constant rate across the interval) , and an index of +0.9 indicates that all respons es occurred in the last bin (i.e., response rates accelerated across the interval) . In addition the determining the index for each of the 20 FIs per session, we also determined an overall index for the session (note that this is what typically is reported in the literature) . To do so, the number of responses that occurred in each bin was summed across the 20 FI repetitions in each session . Then, the equation delineated above was used except that this time R 1 was the total number of responses that occurred in only the first bin s for all 20 FIs, R 2 was the total number of responses
58 that occurred in the first two bins for all 20 FIs, R 3 was the total number of responses that occurred in the first three bins for all 20 FIs, etc. Additional analyses included qu anti fication of dose effect changes. Both response rate and latency measures were plotted as a function of dose to obtain dose response functions, from which area under th e curve (AUC) measures could be determined . In order to normalize the values for abs olute rates that differed across components and pigeons, AUC measures were expressed as a proportion of the no effect value. To calculate this, values acro ss doses were set equal to the value for the acute saline for individual pigeons. That is, hypothet ical po ints that revealed no change in responding by any dose were used . The areas under the three actual dose response curves (i.e., acute, chronic, post chronic) then were divided by the hypothetical no change Therefore, a proportion closer to 1 .0 indicate should indicate less sensitivity to the drug effects under acute administration of the doses. Because d amphetamine can produce increases in rates and decreases in latency , a proportion of exactly 1.0 could indicate a function having both asce nding and descending limbs, a common divisor (that is for the chronic and post chronic functions as well) seemed to better describe instances of vertical shifts . This was usef ul in attempting to quantify the extent to which changes observed under chronic administration were reversible. In other words, we needed to know if the post chronic functions more resembled the chronic functions or the acute functions. Results and Discus sion Figure 4 1 shows response rates (left panels) and reinforcement rates (right panels) under FR ( filled symbols) and FI (open symbols) components across the initial 30 sessions of repeated administration of the chronic dose . The horizontal solid and da shed lines represent FR and FI response rates, respectively, in the session immediately preceding the start of chronic
59 administration . The rate decreasing effects on response rate and reinforcement rate that were evident on the first session of chronic ad ministration further replicated the effects obtained under acute administration of this dose. With respect to reinforcement rates, the FI reinforcement rates recovered fully by the first few sessions for most pigeons and approximately by the tenth session for Pigeons 400 and 674. Following recovery to control levels, reinforcement rates were maintained for the remainder of chronic administration. FR reinforcement rates never recovered fully during chronic administration , providing clear evidence for sche dule dependent tolerance . With respect to response rates, a cross the first 4 to 5 sessions, an increasing trend in FR responding was evident for all but Pigeon 400. What looked like an initial trend towards tolerance did not persist , however . FR respon se rates were maintained at levels comparable to the maximum point of the early trend for Pigeons 113 and 674 . FR response rates decreased for Pigeons 106 and 525 following the early trend, and rates were maintained at the same level as the effect in the first session (Pigeon 525) or well below the effect in the first session (Pigeon 106). . FI responding also returned to ward control levels very early (2 4 sessions) , again, for all but Pigeon 400. For Pigeons 113, 525, FI rates were variable across the 30 sessions, with the rate in the last session not differing from the rate in the first session , although recall that reinforcement rate had recovered . I rates were initially variable but progressively became more stable, with the stabilized rates higher than rate in the first session and nearly at control levels. For Pigeon 400, FI rates decreased initially but then started to approach control levels, w demonstrate the most tolerance observed for any pigeon. They reached control level by the third session, and although some variability was evident, they remained high. Ov
60 approached and met control levels even if they were not maintained at control levels. Pigeon 674 was the only pigeon whose FI rates excee ded control levels in some sessions. Figure 4 2 shows the dose response functions for FR (left column) and FI (right column) response rates that were obtained when a probe dose was substitut ed for the chronic dose (open squares). The filled circles indi cate the acute dose response effects determined in Phase 1 (see right column of Figure 3 1, where acute dose response functions for FR and FI are both shown in a single panel per individual pigeon). Note that unlike Figure 3 1 (as well as Figure 4 1), the y axes in Figure 4 2 are not logged given that FR and FI response rate data are shown in distinct panels. For Pigeon 106, in the FI component (right column), all of the probe doses, except 0.1 mg/kg, produced response rates that were increased relative to the effects of saline, but the function was flattened. Acutely, 0.3, 1.0 and 1.7 mg/kg comprised the ascending limb of the function, whereas 3.0 and 5.6 mg/kg comprised the descending limb. During chronic administration approximately the same rate of responding was observed at each of the s e doses . Tolerance to was most evident at 5.6 mg/kg , which previously eliminated responding. The effects of chronic administration were substantially different on FR responding (left column). A ll the data points of the chronic dose response function fe ll along the x axis . The dose response function does not entirely portray that either few or zero responses were emitted each session, and that this responding appeared to be unsystematic because but it was not determi ned differentially for (a) any one probe relative to each of the other probes (b) any one probe relative to 3.0 mg/kg or (c) all probes relative to 3.0 mg/kg. The responses that did occur never resulted in reinforcement. In some sessions, the total numbe r of responses emitted even exceeded 30 but
61 were distributed across the session such that limited hold contingency always ended the FR components . Pigeons 106 and 113 were the only two subjects that showed higher response rates under 3.0 mg/kg (the chron ic dose) compared to when saline was substituted as a probe. For Pigeon 106 that occurred only in FI component but occurred in both components for Pigeon 113. Those two pigeons also were the only two which showed a clear decrease in rate following saline administration as chronic administration continued. The tolerance to the effects of 5.6 mg/kg for Pigeon 106 in the FI component and for Pigeon 113 in both components was, by far, the most dramatic instances of tolerance obtained in the present experimen t. These two pigeons also were the only ones receiving ascending series of doses. Based solely on present data, there is no way to determine whether the effects of saline, the magnitude of tolerance, and the order of the series are coincidental or import ant features. A comparison between these two subjects shows, at the very least, that the effects of chronic administration on FI responding were reproducible under similar conditions, despite the anomalous effect on FR responding for Pigeon 106. Pigeon 106 and Pigeon 400 had similar changes in FR responding when 3.0 mg/kg was administered repeatedly (Figure 4 1), but for Pigeon 400 that sensitization like effect was attenuated when probe doses were administered in place of 3.0 mg/kg. F irst recall that P igeon 400 always received the series of doses in a descending order, then note that the bottom range of effects is zero (or nearly zero) in the FR component at doses 5.6, 3.0, and 1.7 mg/kg . When the 1.0 mg/kg dose first was administered as a probe , FR re sponding was revived and continued to recover across repeated administration of 3.0 mg/kg between probe doses. Although Figure 4 2 does not show this across session recovery of responding, the upper range of effects at 3.0 mg/kg indicates that eventually tolerance developed , but it was restricted to the effects of just the 3.0
62 mg/kg dose . By comparison, tolerance in the FI component developed to the rate increasing and rate decreasing effects across the range of doses. Not only was tolerance dose specific in the FR component for Pigeon 400 , but there also was evidence of sensitization to the effects of some doses. W hen administered acutely, 0.3 and 1.0 mg/kg did not decrease rates relative to saline levels but during chronic administration these doses dec reased responding relative to saline levels. Similarly, the rate decreasing effects of 1.7 mg/kg were greater during chronic administration. The effects of saline probes during chronic administration did not differ from the effects of saline prior to chron ic administration and, along with no effects produced by 0.1 mg/kg before and during chronic administration, indicated that there was not a downward shift of the entire dose response function. Therefore, there was evidence of sensitization at the moderate doses. What, unfortunately , cannot be determined is whether a saline probe would have revived responding in the manner as occurred after the first administration of 1.0 mg/kg. Because probe doses were administered in descending order, the first determi nation of the saline probe did not occur until after responding was already revived. So it is possible that a downward shift in the entire function would have been evident had we administered a saline probe first. Pigeon 106 and Pigeon 400 responded ve ry similarly prior to probe dosing (i.e., Figure 4 1), and Pigeon 106 received an ascending series of probe dose and no recovery of responding at any dose was evident . We counterbalanced the order (ascending or descending) across subjects , and the two pig eons whose FR responding was eliminated by repeated administration of the chronic dose had been assigned to different orders . Differences in effects typically are not reported between ascending or descending series of probe doses, but the preferred dose f or chronic administration
63 has been one selected specifically because it did not eliminate responding (e.g., Hoffman et al., 1987; Schuster, 1978). Some studies have manipulated the dosing regimen for chronic cocaine administration (e.g., Branch et al., 2 000; Miller & Branch, 2002). In these studies, a different dose of cocaine was administered prior to consecutive sessions such that chronic administration consisted of exposure to repeated cycles of a series of doses (i.e., variable dosing). Branch et al . found that the magnitude of tolerance did not differ substantially for pigeons that received a descending series of variable doses of cocaine compared to those that had received fixed doses of cocaine. Similarly, Miller and Branch reported that the magn itude of tolerance was similar for ascending and descending series of variable doses, but the development of tolerance was more gradual when the descending series was used. Perhaps the order of the probe doses plays a role only when responding has been el iminated. In our study, r esponding was not eliminated by 3.0 mg/kg administered acutely, but future studies could purposefully select a large, rate eliminating dose to (e.g., Miller & Branch, 2004) . Miller and Branch (2004) chronically administered a large dose of cocaine that when administered acutely . If any recovery of responding became apparent during chronic adm inistration, the dose was increased and administered until responding was eliminated for 50 consecutive sessions. For the next 100 sessions , pigeons in the control group continued to receive the chronic dose prior to every sessions, whereas other pigeons (i.e., saline group) received a saline injection in lieu of the chronic dose every fifth session. In this phase, Miller and Branch found that occasional saline administration attenuated the rate eliminating effects of cocaine for some pigeons in the salin e
64 group and that responding remained eliminated for all pigeons in the control group. In the final phase of their study, dose effect functions were redetermined. For the saline group, a shift in the dose response function to the right indicated that tole rance had developed to the effects of cocaine. Conversely, in the control group, sensitization to the effects of cocaine was evident initially (but later dissipated) . Overall, Miller and Branch demonstrated that the occasional opportunity to respond in t he absence of drug promoted the development of tolerance , with which the results from Pigeon 400 were consistent . One interpretation of this is that saline (or low enough dose of drug) allows the organism to re contact the reinforcement contingencies afte r a prolonged period of exposure to the discriminative stimulus but not reinforcement, and the reinstated responding then generalizes to other doses. Another interesting finding in Figure 4 2 is evidence of tolerance in FR and FI components, even when tole rance had not emerged durin g the consecutive 30 sessions (as evident for Pigeons 113 and 525 ). For both pigeons, tolerance was evidenced by less of an effect at the higher doses that decreased rates in the FR and FI components when administered acutely (5 .6 and 10.0 mg/kg). The degree to which tolerance developed did not differ across components, but did across subjects with Pigeon 113 showing greater tolerance than Pigeon 525 . For Pigeon 525, there was a downward shift at the lower doses (0.1 through 1. 7 mg/kg). These doses decreased responding relative to the effects of saline during chronic administration more so than during acute administration. For Pigeon 113, there not only was a shift downward at the lower doses (0.1 to 1.0 mg/kg) but also at sal ine. Th at is, the acute effects of these low doses were comparable to the saline probe ( acute ) , but when the effects were reassessed during chronic administration, they produced incre ases in rates relative to the saline probe (chronic).
65 Pigeon 674 showed dose specific tolerance in the FR component to the extent that the shift in mean and range of effect of probes at 4.2 mg/kg , in conjunction with the increasing trend toward control levels that occurred prior to probe doses (shown in Figure 4 1) , are indic ative of tolerance. Because the 3.0, 4.2, 5.6 mg/kg doses produced variable effects during acute and chronic administration, determinations at these doses were repeated to produce more secure estimates of the means . It was also determined that the order of the determinations for this subject was not predictive of the degree of responding that occurred , except for the effects of 4.2 mg/kg (i.e., the chronic dose), which tended to be lower towards the beginning, and higher towards the end, of probe dosing. This indicated that tolerance which already had developed noticeably under pre probe administration continued to develop further when probe doses were occasionally substituted for the chronic dose . Whereas tolerance was enhanced for Pigeon 674, when pro bes were administered to other subjects, obvious instances of tolerance occurred for the first time. In our study, 30 sessions was selected based on some of our previous work on the development of cocaine tolerance (e.g., Minervini & Branch, 2013) as well as ef fects reported in the literature on chronic d amphetamine administration (1 week to 60 days), but the length of chronic administration could be manipulated such that probing for tolerance takes place prior to any notable changes in, say, rates. Thi s may enable one to test the notion that the FI dose response function flattened relative to the dose administered chronically. If that were the case, then you might expect to see tolerance at doses nearer to the chronic dose earlier than changes that tak e place at the lower doses. Tolerance to the rate increasing effects of amphetamine in the FI component, was evident in every subject in the present study, although individual differences in the degree of effect were
66 present. Rate increases observed u nder acute administration were accompanied by corresponding pause decreasing effects, and these are shown in Figure 4 3. Decreases in FI pausing were attenuated under chronic administration, but overall the pause data were less reflective of the tolerance that had developed . Thus, t he vertical shift in response rate during chronic administration was not the same sort of flattening that occurred for FI changes. To better capture details of responding that session averages failed to capture , cumulative r e sponse records were constructed . Figures 4 4 through 4 13 show two sets of cumulative record s for each pigeon , and each cumulative record was selected because performance in that session was representative of performance at large . The first set shows thr ee panels from top to bottom: the acute effects of the dose that was ultimately selected as the chronic dose for repeated administration, the effects of this dose following the initial 30 consecutive session plus after 25% 50% of probe dose assessments had taken place, and finally, the effects of the chronic dose t he last session prior to the start of pos t chronic saline administration . The second set shows two columns of cumulative records. The left colum n shows the acute of effects saline , as well a s doses lower and higher than the chronic dose that produced appreciabl e change s in performance . The selected doses are shown in ascending order from top to bottom even if the pigeon actually received them in descending order . The right column shows the e ffects of the same doses during chronic administration. Every panel shows 20 repetition s te that a reinforcer was earned. The precision has been highly preserv ed in reconstructing the originals to electronic form, so readers are encouraged to use the zoom feature if interested in inspecting the details more closely.
67 The primary finding collectively demonstrate d by the cumulative records is that FI responding w as more enduring than FR responding in many instances of drug administration . Under both acute and chronic conditions , t here were , in which few or zer o responses occurred , than FI components . This occurred despite the fact that FR reinforcement rate was considerably greater under baseline and even acute administration (refer back to the right column of Figure 3 3 as well as Figure 4 18 described below). Thus, these findings were inconsistent with a behavioral momentum accou nt of tolerance as discussed by Poli ng, Byrne, Christian, and Lesage (1997) and later summarized nicely by Nevin and Grace (2000): both upon initial administration an d as its effectiveness wanes during the development of These findings also were inconsistent with a reinforcement loss account of tolerance described by Smith (1986). That is, acute administration caused a greater proportion of reinfor cement loss in the FR component (refer back to Figure 3 4), however, responding in this component did not recover more so than FI responding under chronic administration . The disruptive effects of the chronic dose, given acutely, could be characterized in two ways. Responding was disrupted predominantly in the first portion of the session, with the intervals becoming more smooth and eventually fairly close to control responding ( compare : Figure 4 4 top and 4 5 top left; Figure 4 10 top and 4 11 top left; F igure 4 12 top and 4 13 top left). The other pattern of responding was scattered and disrupted across the session ( compare : Figure 4 6 top and 4 7 top left; Figure 4 8 top and 4 9 top left). The recovery of temporal patterning in the FI component seemed rates never returned to rates observed under acute saline (e.g., Pigeons 113 and 400) but by the end of chronic administration they clearly no longer were disrupted pattern wise . Interestingly,
68 these tw o pigeons experienced an overall disruption initially, and t he failure to return to saline levels may have been adaptive; at the end of chronic administration pigeons were earning all of the programmed reinforcers despite less response output. For the o ther three pigeons (106, 525, and 674) , when the acute effect was to impair within session performance towards the start of the session to a greater extent, FI rates did recover to saline levels. The middle panel (Figures 4 4, 4 10, and 4 12) shows the re covery of latencies and scalloping, as well. A gradual return of FR pattern did not occur. Once chronic administration was under way, pigeons either completed all of a ratio or did not complete it . Because the limited hold was only 60 s and d amphetamin e increased latencies to respond, one expectation was that ratio responding would be initiated too late. Thus tolerance would develop as an adjustment to contacting the limited hold while completing the present response requirement . This makes sense in t hat there may be more aversive properties associated with a limited hold following a key peck response rather than some behavior maintained by non programmed reinforcement . As chronic administration continu ed , however, this hardly ever happened (some pro be doses engendered a response pattern like that though). T hey completed more of the 20 possible , but tolerance did not develop by making a few responses at the start of all the ratios. It is hard to make the case then that tolerance under one set of con tingencies developed before the other ; rather, they seem to develop differently. This could be related to a reduction in incompatible behaviors. For example, Pinkston and Branch (2003) showed that when tolerance had developed to the rate decreasing effec ts of cocaine, their movements , indexed by a measure of ambulating , were more constrained (to presumably the vicinity of the key). When tolerance had not developed to the effects of the highe st doses, then there was a sensitization effect o f nd accompanied by complete elimination of key pecking.
69 Wolgin and colleagues (e.g., Salisbury & Wolgin, 1985; Wolgin, 2000; Wolgin & Wade, 1995) extensively have studied behaviors prone to sensitization in rats, such as activity, grooming, and head ticks , and have termed the subsequent reduction of these behaviors as . They showed that d amphetamine produces decrease in milk drinking via stereotypy . When some rats were given an intraoral cannula d amphetamine suppress ed milk intake m uch less than it was suppressed in other rats required to approach and drink from a sipper to access milk. The sipper access group, however, developed more tolerance and faster than the cannula ted rats. Finally, to demonstrate that thi s was due to learni ng processes, a response contingency was put into place: milk was provided through the intraoral cannula only if the rat maintained a still head position for certain amounts of time. They found that rats do develop a substantial degree of tolerance to the stereotyped motor effects of d amphetamine, but their response patterns and distribution of responding looks different than that of saline treated rats. food (like t he cannula was for rats ) , and that could explain why tempor al patterning seemed to recover differentially . If the pigeons were engaged in stereotypy (e.g., preening, walking, flapping) this would not be incompatible with earning reinforcemen t in the pres ence of the red key so long as they stopped to peck a few times. Conversely, s tereotypy in the presence of the white key would directly impair their rate of reinforcement on the FR schedule so it tended to be an all or none response, which generally is co nsistent with FR responding under typical and drug procedures (Marusich, Branch, & Dallery, 2008; Mazur & Hyslop, 1982; Pinkston & Branch, 2003, 2010), but it does not explain how FR tolerance could have developed.
70 To quantify the FI response patternin g seen in the cumulative response records, the mathematical index of curvature (Fry et al., 1960) was determined and these data are shown in Figure 4 14. The solid thick line (for acute doses) and dashed thin line (for chronic doses) indicate the overall index for the entire session and can be compared to the individual data points (filled circles correspond to the solid line and open circles correspond to the dashed line) . Each individual data point denote s the index for a single interval of th at session . To clarify further, the data paths that represent the overall index were not determined merely by obtaining the mean of the individual data points ; in fact the overall index turned out to be unrelated to mean of the individual data points in any orderly way (not shown) . Additionally, note that what we have termed the overall index here actually is the standard index reported in the literature (i.e., . We found that acute administration tended to dec rease the overall index as the dose increased. The degree of decrease was comparable to previous studies in which pigeons responded under an FI 5 min component in the context of other, shorter FI components (Branch & Gollub, 197 4 ) and under an FI 2 min co mponent in the context of an FR 30 component (Katz et al., 1990), which speaks to the generality of the findings. During chronic administration, the overall index flattened across doses, indicative of tolerance to the effect of disrupted FI patterning. K atz et al. reported that this flattening of the overall index was markedly more pronounced than any changes in response rates, but in our study the degree of recovery in patterning seemed related to the degree of tolerance in terms of response rates for in dividual pigeons (shown in the right column of Figure 4 2) . The overall index for chronic saline was lower than that for acute saline in three pigeons (106, 400, and 674), meaning that responding became less accelerated and more constant across the interv al. A comparison between the
71 vertical shifts in response rate at the saline dose for Pigeons 106 and 113 (Figure 4 2) with the ir overall indexes at the saline dose highlights another interesting feature: l ower response rates during chronic administration were not necessarily associated with disrupted patterning. Our study is the first to examine how the distribution of individual indexes for all the FIs in a session changes during chronic drug administration , although Gollub (1964) and Branch and Gollub ( 197 4 ) did conduct several detailed analyses of the repetitions of FI performance compared to average performance of the overall session but not in the context of chronic administration . More often than not, the individual indexes were distributed across a range of values rather than clustered around the overall index. Increases in the overall index during chronic administration stemmed from a narrowing of the range in addition to an upper shift in the distribution . That is, more of the 20 intervals could be characterized as having accelerated rates and there were only two intervals in which marked deceleration was apparent during chronic administration (Pigeon 400 at 0.3 mg/kg and Pigeon 674 at saline). Perhaps separating out the individual indexes for i ntervals ending in reinforcement versus those ending in the limited hold would explain why the bottom of the range is higher under chronic than under acute dose t o that from the chronic 5.6 mg/kg dose (Figure 4 7, bottom panels). Acutely, there were 7 intervals (the third and fourth, twelfth and thirteenth, and the last three intervals at the end of the session) in which responding occurred but did not result in r einforcement , and these might comprise the five individual indexes that were negative for acute 5.6 mg/kg in Figure 4 14. Additionally, the two individual indexes that were slightly above zero for acute 5.6 mg/kg in Figure 4 14 appeared to reflect the res ponse pattern in the first and second intervals of that session (Figure 4 7, bottom left panel). The latencies to the first response of each interval were
72 much longer for most of the repetitions for chronic 5.6 mg/kg (Figure 4 7, bottom right) , thus most of the individual indexes in Figure 4 14 at this dose well exceed zero . If no responses were made in a given interval, then the index of curvature for that particular interval could not be calculated. Thus, fewer than 20 data points were plotted for some doses. Table 4 2 shows instances in which no responses were made. In general, FIs were excluded most often when the high dose (5.6 mg/kg) was administered acutely. When 5.6 mg/kg was administ ered as a probe during chronic administration, responding occu rred in 19 or 20 out of 20 intervals for all pigeons except Pigeon 400 for whom responding occurred in 9 of the 20 intervals. A n analogous pattern emerged for the chronic dose, offering further evidence that tolerance had developed. Note that the perspect ive described by Wolgin (2000), contains features that overlap with views such as reinforcement loss (Schuster et al., 1966) and behavioral momentum (Nevin, 1974). In t he literature on behavioral mechanisms of drug tolerance , reinforcement loss and behav ioral momentum both have been shown to predict and account for drug effects and both have been criticized for lacking scope ( Cohen, 1986; Jimenez Gomez & Shahan, 2007; Pinkston, Ginsburg, & Lamb, 2011, 2014; Poling et al. , 2000; Spealman, Katz, Witkin 1978 ; Quick & Shahan, in press ). At present, the problem with conflicting reports seems to be that too many parameters vary across studies t o create cumulative knowledge. The exact opposite holding too many variables constant may be just as detrimental. The discussion sections across a wide variety of papers, however, fail to incite action. P ortions of findings are each just explained away by distinct predictions of different frameworks. Identifying the boundaries parametrically , not just through demonstra tions of these frameworks ultimately should allow us to see where, if at all, they converge. For example, in the present study neither reinforcement
73 loss nor resistance to change can fully account for the results. Alternatively, it is likely that the fun ction of the controlling variables were differentially altered by d amphetamine . At present, there is not a best way to identify the relative degree to which certain features of the environment contributed. For example, the higher response rates and rich er reinforcement rates seem like desirable conditions that would promote the development of tolerance. T he shorter component duration in the context of the multiple schedule may have offered greater stimulus reinforcer control in the F R component . Holdin g more variables constant may be useful to some extent, but in the case of behavioral momentum, an important determining variable reinforcement contingency was revealed only by loosening the control of other variables that otherwise procedurally obscured t he phenomenon ( i.e. , Nevin et al., 2001). Additional support of the differences observed across FI and FR components can be seen in post chronic changes. Figures 4 1 5 and 4 1 6 show response rates and latencies, respectively. Here the dose response functi ons from the acute and chronic assessments are identical to those shown in Figures 4 2 and 4 3, with the dose response function s obtained under post chronic exposure (i.e., while daily saline injections w ere administered) plotted as well . The post chronic effects are represented by the filled triangles. Generally , the effects of the low doses on rate and pause were reversed or attenuated during chronic saline administration for all pigeons, but in several instances the effects of the higher doses were not (e.g., the 3.0 mg/kg dose for Pigeons 113, 525, and 674, as well as the 4.2 and 5.6 mg/kg doses for Pigeons 525 and 674). The effects on rate and pause seemed to be attenuated more in the FR component than the FI component. For all pigeons , the three low est doses did not produce rates that differed from control responding (i.e., daily saline) in the FR component during post chronic administration (with the vertical shifts for Pigeons 400 and 525 considered, the effects of subsequent doses were
74 not differe nt from saline and were different from the effects during chronic administration). This reversal seemed to be more clearly differentiated from the effects of th o se doses in the FI component , and it seemed like there were orderly within subject differences across the two components . For example, pigeons who showed the same effects during post chronic as they did during acute (especially at the lower doses) in the FR component seemed to show less clear distinction between these doses in the FI component. produce the same rate increasing effects (as they did acutely) and the high doses had th e same rate decreasing effects as they did acutely . Changes in the dose response functions for latencies (Figure 4 1 6 ) were less orderly than for response rates. To provide another evaluat ion of the extent to which the initial effects of the doses were reproducible following post chronic saline administration area under the curve (AUC) meas ures were determined and are shown i n Figure 4 17 . When thes e measures were calculated using proportion of control in each condition (i.e., acute saline, chronic saline, post chronic saline), the results did not appear representative of actual changes in the function, especially for pigeons with shifts in control responding across conditions. Instead, relative AUC was calculated as a proportion of the area as bounded by the effects of acute saline. Changes at the ascending and descending limbs of the dose response functions likely represen t meaningful changes in drug effect that cannot be adequately described by AUC. So, rather than using this analysis to assess the magnitude of tolerance between FR and FI components, these measures were conducted as a way to quantify reversibility of effe cts across conditions. That is, the noteworthy feature of Figure 4 17 rather the degree of similarly across the three conditions for each individual pigeon. For example, the AUC measures for Pigeon 674 FI rates in this figure appeared to indicate that
75 effects obtained during post chronic administration were not in the direction of those obtained during acute administration (i.e., the height of the bars decreased across the three phases , as opposed to the first and third phases having similar heights compared to the second phase ) , but they depict n either the source of th is difference nor changes in tolerance. Those interpretations can be made only by referring to Figure 4 1 5 , where for Pigeon 674, the source of the difference appears to be related to the effects of 1.0 and 3.0 mg/kg. Overall, AUC for rates (FR and FI) showed that the effects were reversible at least at some doses. AUC for latencies did not show a consis tent pattern. Latencies were plotted as total time spend pausing each session and then change, that is reproducible changes. This is true for the response ra te data to some extent as well. The cumulative response records make for a much more compelling case of the development of tolerance Pre vious findings on reinforcement loss as a determinant of tolerance have been mixed. In rats, tolerance did not develo p to the rate increasing effects associated with no change or enhanced reinforcement rates (e.g., Schuster et al., 1966), but other times it depend ed on the context (e.g., Smith, 1986 ; Tilson & Sparber, 1973 ) . Pigeon s and monkey s have shown tolerance to t he rate increasing effects of cocaine even when the effect is enhanced reinforcement rate under FR schedules or no effe ct on reinforcement rate under FI schedules (Branch, Walker, & Brodkorb, 1999; Schama & Branch, 1994). In our study rate of reinforceme nt in the FI component was less disrupted by acute d amphetamine relative to control levels, except at the highest dose where decreases were compara ble in both components (Figure 3 4 ) . Figure 4 1 8 shows dose response function for obtained reinforcement rat e in the FR and FR components. Acute d amphetamine at the highest
76 doses tested for each pigeon resulted in complete reinforcement loss (expectedly, response rates were eliminated or nearly eliminated, which can be seen in Figure 4 2 as well as in the cumu lative response records ) . The chronic dose response functions for FI reinforcement rates shown in Figure 4 1 8 (right column) shows that Pigeons 106, 113, and 674 were able to obtain the maximum reinforcement rate permitted by the FI schedule. For Pigeon s 400 and 525, chronic administration attenuated FI reinforcement loss, but reinforcement rate reached the maximum possible at relatively lower doses (4.2 and 3.0 mg/kg) compared to the other pigeons. These findings seem relevant in explaining the flatten ing of the functions for FR responding in Figure 4 2. Pigeons seemed to be sensitive to the lowe r rate of responding necessary to maintain the maximum reinforcement rate. That is, the shift downward in FI responding at lower doses even when they those ra tes were decreased relative to saline seemed to indicate an adaptive response. The curve flattens at the point where emitting any would not further change the rate of reinforcement. To assess for potential changes in sensitivity to reinf orcement rate, it would have been interesting if we had re conducted the momentum tests by implementing prefeeding at some points during chronic administration. In any event, the notion could be better addressed in future research by also including a drug dose + prefeeding combination as one of the disruptors to see if relative resistance may be related to drug exposure (e.g., pre , during , and post chronic). If this type of study were to be arranged, interpretation of the results could be addressed wit h respect to previous reports of changes in temporal discounting before, during, and after chronic cocaine administration (e.g., Dandy & Gatch, 2009) and chronic amphetamine administration (e.g., Huskinson, Krebs, & Anderson, 2012). Drawing parallels like these may shed more light on the underlying behavioral processes and mechanisms.
77 Contrary to the downward shift that occurred at low doses in curves for the FI component, the downward shift that occurred at low doses in the FR component could not be i nterpreted as adaptive, as slower response rates necessarily decreased reinforcement rates to the same extent. Katz et al. (1990) found that during chronic administration of a high dose (one that initially decreased FR and FI response rates and reinforce ment rates substantially) minimal tolerance was evident and was restricted to the higher doses. The entire dose response function for FI responding tended to flatten, but they did not report any downward shifts in the dose response function for FR respondi ng. Our findings were partially consistent with theirs in that FR tolerance was observed prim arily at higher doses, but contrary to theirs in that FR dose response functions shifted down for all pigeons . When the within study results are, for the most pa rt, orderly, then one could turn to the across study differences that may provide a hint as to what the potential mechanism was. At first glance the differences do not seem significant (i.e., pigeons responding in a multiple FR FI schedule, similar dosing procedures, even some similar dependent variables like baseline response rates). A closer look reveals many small differences, any or all of which could be interacting with the effects of d amphetamine, such as an FI 2 min rather than FI 5 min, context, chronic dose administered post session on probe sessions, 30 s limited hold rather than 60 s , pigeons maintained at 80% rather than 85% weights, etc. Not enough is known about the independent effects of all those variables, much less their interactions w ith each other , and moreover, their independent and combined interactions with d amphetamine. Figure 4 1 9 summarizes FR and FI response rates in baseline and towards the end of acute administration. The means were based on the last 10 session s, and, impo rtantly, session s
78 the sam e, with the range of FR rates between 2 4 per second and FI 0.3 0.7 per second across subjects. Within subjects, FI rates were slightly higher for Pigeons 106, 113, and 400 prior to the start of chronic administration. These data were analyzed further, and the results are shown in Figure 4 20 . This figure shows the ratio of the high response rate engendered by the FR schedule and the lo w response rate engendered by the FR schedule. This was determined by first obtaining the relative ratio of rates fo r each of the 10 sessions. Then, the 10 ratios were averaged. This analysis showed that, just before chronic administration commenced, tw o pigeons differed in a way from the others and this difference was correlated with their FR performance under repeated administration of the chronic dose for 30 consecutive sessions . That is, Pigeons 106 and 400 were responding in the FR component a t a rate that was near 10 fold greater than FI responding , and only those 2 subjects showed a complete cessation of responding in the FR component during repeated administration of the chronic dose . It would be interesting to know more about what happens when response rates are manipulated relative to each other (similar to relative reinforcement context described by Smith, 1986 ). It would be beneficial to test the effects of various disruptors when response rates are manipulated in that manner, as this potent ially could inform us about drug effects. So far some evidence is that Nevin et al. (2001) found a strong correlation between individual subject differences in baseline performance and subsequent disruption: the degree to which VR response rates were mai ntained at a higher rate than VI response rates in baseline was associated with the extent to which responding in the presence of the VI component persisted.
79 Table 4 1. Number of injections each dose was administered to individual pigeons acros s Phase 2 conditions. Pigeon Dosing Saline 0.1 0.3 1.0 1.7 2.4 3.0 4.2 5.6 10 106 Chronic 30 Chronic dose response 2 2 2 2 2 53 2 Saline 30 Saline dose response 49 2 2 2 2 3 2 113 Chronic 30 Chronic dose response 2 2 2 2 2 2 51 2 Saline 30 Saline dose response 60 2 2 3 3 2 2 2 400 Chronic 30 Chronic dose response 2 2 2 2 2 39 2 Saline 30 S aline dose response 49 2 2 2 2 2 2 525 Chronic 30 Chronic dose response 2 2 2 2 2 65 2 2 2 Saline 30 Saline dose response 54 2 2 2 2 2 2 2 2 674 Chronic 30 Chronic dose respo nse 2 2 2 2 2 2 49 2 Saline 30 Saline dose response 48 2 2 2 2 2 2 2
80 Table 4 2. For each pigeon, the number of intervals in which zero responses occurred during acute and chronic admini stration . These data correspond to the index of curvature analysis shown in Figure 4 14. Pigeon Dose Acute Chronic 106 saline 0 0 0.3 0 0 1.7 0 0 3 2 0 5.6 18 1 113 saline 1 8 1.0 0 0 3.0 0 0 5.6 2 0 400 saline 1 0 0.3 0 1 3.0 2 1 5.6 18 9 525 saline 0 1 1.7 0 0 3.0 2 1 5.6 4 1 674 saline 0 1 1.0 0 0 4.2 3 0 5.6 5 0
81 Figure 4 1 . Initial effects of the chronic dose, repeatedly administered for 30 sessions, on FR ( filled symbols) and FI (open symbols) re sponse rates (left panels) and reinforcement rates (right panels) . The horizontal solid and dashed lines represent control levels of FR and FI responding, respectively, in the session immediately preceding chronic administration. Note y ax e s are logged a nd the scales differ .
82 Figure 4 2. Response rates for FR (left column) and FI (right column) components as a function of d amphetamine dose. Acute administrations are denoted by filled circles, and chronic adminis trations are denoted by open squares. Individual pigeons are shown by row. Errors bars denote range. Note x axis is logged and y axes scales differ .
83 Figure 4 3. Total latencies for FR (left column) and FI (right column) components as a function of d amphetamine dose. Acute administrations are denoted by filled circles, and chronic administration s are denoted by open squares. Individual pigeons are shown by row. Errors bars denote range. Note x and y axes are logged an d y axis scales differ .
84 Figure 4 4 . Representative cumulative records (cumulative pecks over continuous time) for Pigeon 106 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. For this and all f ollowing cumulative records, the red and blue lines show FR and FI components, respectively. The cumulative total was reset to zero after each component.
85 Figure 4 5 . Representative cumulative records for Pigeon 106 showing the effects of doses other than 3.0 mg/kg acutely and during chronic administration. Details as in Figure 4 4.
86 Figure 4 6 . Representative cumulative records for Pigeon 113 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. Details as in Figure 4 4.
87 Figure 4 7 . Representative cumulative records for Pigeon 113 showing the effects of doses other than 3.0 mg/kg acutely and during chronic admin istration. Details as in Figure 4 4.
88 Figure 4 8 . Representative cumulative records for Pigeon 400 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. Details as in Figure 4 4.
89 Figure 4 9 . Representative cumulative records for Pigeon 400 showing the effects of doses other than 3.0 mg/kg acutely and during chronic administration. Details as in Figure 4 4.
90 Figure 4 10 . Representative cumulative records for Pigeon 525 showing the effects of 3.0 mg/kg acutely and at two points during chronic administration. Details as in Figure 4 4.
91 Figure 4 11 . Repre sentative cumulative records for Pigeon 525 showing the effects of doses other than 3.0 mg/kg acutely and during chronic administration. Details as in Figure 4 4.
92 Figure 4 12 . Representative cumulative records f or Pigeon 674 showing the effects of 4.2 mg/kg acutely and at two points during chronic administration. Details as in Figure 4 4.
93 Figure 4 1 3 . Representative cumulative records for Pigeon 674 showing the effects of d oses other than 4.2 mg/kg acutely and during chronic administration. Details as in Figure 4 4.
94 Figure 4 14. Mathematical index of curvature for individual FIs and the whole session during acute and chronic administration. Acute administration is deno te d by black filled circles and solid line, and chronic administration is denoted by open circles and dashed line. The sessions displayed here include only those for which cumulative response records were shown in the preceding figures.
95 Figure 4 1 5 . Dose response functions for FR and FI response obtained after withdrawal of the chronic dose (denoted by filled triangles). Note this figure differs from Figure 4 2 only in that the post chronic data have been added.
96 Figure 4 1 6 . Dose response functions for FR and FI latencies obtained after withdrawal of the chronic dose (denoted by filled triangles). Note this figure differs from Figure 4 3 only in that the post chronic data have been adde d.
97 Figure 4 1 7 . Area under the curve (AUC) measures expressed as a proportion of acute saline or dose response functions presented in Figures 4 1 5 and 4 1 6 . White, gr ey, and black bars represent the time points at which the dose response function were obtained: acute, chronic, and post chronic , resp ectively. Note the y axi s scale and break for the FR Latency panel only (bottom, left) .
98 Figure 4 1 8 . Dose response functions for FR (left) and FI (right) reinforcement rates. Acute, chronic, and post chronic administrations are denoted by filled circles, open squares, and filled triangles, respectively. Errors bars denote range . Note x axis is logged and y axes scales differ .
99 Figure 4 1 9 . Mean FR (top panels) and FI (bot tom panels) and FI (top panels) response rates. The means were based on the overall response rates of the last 10 sessions from baseline (lef t panels) and the last 10 sessions on which no drug was administered during the acute administration. Error bars represent standard deviation.
100 Figure 4 20 . Means for the r elative difference in FR and FI responding in 10 sessions . The abs olute means are presented in Figure 4 1 9 ; the details, otherwise, are the same as described for Figure 4 1 9 .
101 CHAPTER 5 GENERAL DISCUSSION The purpose of the present research was to further an understanding of behavioral determinants of drug ac tion and tolerance . More specifically, we aimed to evaluate the effects of acutely administered d amphetamine on response rate using a behavioral momentum framework when the rate of reinforcement and schedule contingencies differed across contexts, that i s, FR and FI components of a multiple schedule. A second pur pose was to evaluate the development of tolerance in each component during chronic administration of d amphetamine. P revious studies have generated contradictory results. Under arrangements simi lar to those of the present study, tolerance has been observed, sensitization has been observed, and neither tolerance nor sensitization (i.e., no change) has been observed. We were primarily interested in how response rate, rate of reinforcement , and con tingencies of reinforcement may interact to modulate the acute drug effects as well as the development of tolerance . In Phase 1 we found that FR responding was less sensitive than FI responding to the disruptive effects of prefeeding, but more sensitive to the disruptive effects of d amphetamine administration. From this we were able to determine that the effects of d amphetamine on food maintained responding cannot be attributed simply to satiation like, anorectic effects. Additionally, t hese results were not predicted by behavioral momentum theory, as the rate of reinforcement was substantially greater in the FR component. Based on the results of Phase 1, we selected a dose for chronic administration. The primary criteria were that the dose decreased re sponse rates in both components but decreased the rate of reinforcement in the FI component proportionally less than in the FR component. Doing so enabled us to arrange preemptively for any obtained findings to be addressed from a behavioral momentum fram ework as well as a reinforcement loss framework.
102 Across 30 sessions of repeated administration, changes in response rates genera lly were somewhat minimal, with FI response rates more closely approximating control levels of responding than FR response rat es in four of five pigeons. The changes in FI performance, however, did result in recovery of rate of reinforcement, something that did not occur in the FR component. In some respects this supported the role of behavioral momentum in predicting subsequen t tolerance and was consistent with previous findings that the less initial disruption was related to greater tolerance (e.g., Poling et al., 2000). When lower and higher doses were occasionally substituted for the chronic dose in order to obtain dose res ponse functions, instances of tolerance w ere evident in both the FR and FI components for most pigeons. In the FI component, however, tolerance was usually evident across the range of doses assessed for all pigeons. In the FR component, the greatest deg ree of tolerance appeared to be dose specific. Tolerance developed to the effects of the chronic dose, but whether and the extent to which tolerance developed to the effects of lower and higher doses was inconsistent across pigeons. Proportionally greate r reinforcement loss occurred in the FR component, which, according to the literature (Schuster et al., 1966; Smith, 1986) should have resulted in greater degree of tolerance during the 30 consecutive sessions and dose response redeterminations. Our overa ll finding that tolerance developed in both components was similar to previous findings when shorter (2 min) FI components (Katz et al., 1990) and longer (10 min) FI components (Marusich & Branch, 2008) were arranged with an FR 30 component , but the magnit ude of tolerance in each component differed across studies. This suggested that the FI parameter may be relevant when there is the added context of the FR component . Tolerance has not been shown to be parameter dependent when a multiple schedule consists of only FI components , though ( Schama & Branch, 1989) , whereas tolerance has been relatively consistently parameter dependent when only FR
103 schedules have been employed ( e.g., Hoffman et al., 1987; Hughes & Branch, 1991; Minervini & Branch, 2013; Nickel, Al ling, Kleiner, & Poling, 1993; Van Haaren & Anderson, 1994; Yoon & Branch, 2004) . Neither rate of reinforcement nor loss of reinforcement seemed to explain the development of tolerance in the present study or account for differences across studies. An at tenuation of anorectic effects also could not have contributed to the development of tolerance because no such effect was found to exist in Phase 1 . Generally, the effects of d amphetamine on rate and pause during chronic drug administration were reversed or attenuated during repeated saline administration for all pigeons. The effects on rate and pause seemed to be attenuated more in the FR component than the FI component, but overall, the initial effects of the doses were reproducible following a period o f no drug exposure. To characterize FI performance fully, it was important to include some measure of temporal distribution of responding in addition to the rate and pause measures (e.g., Fry et al., 1960; Gollub, 1964). The effects of d a mphetamine on F I performance have been viewed previously as a disruption of response patterning rather than merely a disruption of overall response rates (e.g., Branch & Gollub, 1974) , so we also assessed the development of tolerance by visually analyzing c umulative resp onse records and by calculating mathematical index of curvature . We found that acute d amphetamine administration tended to decrease the o verall index dose dependently and that d uring chronic administration it flattened across doses, indicative of toleran ce to the effect of disrupted FI patterning consistent with previous reports (e.g., Katz et al., 1990 ) . Cumulative records showed that t he recovery of temporal patterning in the FI component seemed to re cover before rates recovered. Even though s ome pige never returned to control (no drug) levels, by the end of chronic administration pattern disruptions no longer were evident . Moreover, FR performance recovered differently than the
104 way in which F I performance recovered. In the FR component, pigeons tended to complete more of the 20 possible, but tolera nce did not develop by making more and more responses in each of the ratios until they achieved the requirement for reinforcement (i.e., 30) . The results from these analyses further strengthene d the finding that the development of tolerance observed in the present study was schedule dependent. To identify behavioral mechanisms, f uture studies might aim to address the role of reinforcement rate and reinforcement loss when the contingencies (i.e. , the schedule type) differ versus when they are e quivalent. That is, one potential mechanism may be that the drug effect is to decrease sensitivity to rate of reinforcement (e.g., Jimenez Gomez & Shahan, 2007) . If this is the case, then pharmacological di sruptors should produce effects contrary to environmental disruptors when rate of reinforcement differs (and the contingencies are the same), an arrangement that is typical is many behavioral momentum studies. This could then be compared to the effects of a pharmacological disruptor when the contingencies differ but rate of reinforcement is held constant. Under these conditions, one would expect to see the role of the contingencies alone, but different rates of reinforcement eventually could be introduce d as well. If the contingencies were deemed to be important determinants of acute and chronic drug effects, as the results of the present student tentatively indicate, a behavioral mechanism still would need to be identified. One such mechanism could b e that the drug effect is to reduce sensitivity to reinforcer amount (Maguire et al., 2009). E ven though the same reinforcer (e.g., 3 s of food) normally may maintain responding comparably across different contingencies , the drug could differentially alte r the function of the reinforcer depending on the contingencies. That is, if d amphetamine decreases sensitivity to amount, decreasing the amount could interfere with normal response reinforcer relations under FR schedules more so than FI schedules. Some
105 evidence to support this notion comes from the behavioral economics literature. Tsunematsu (2000) determined that at increasingly high costs FI contingencies maintained more responding than FR contingencies with equivalent costs ; decreasing the magnitude of a reinforcer is considered an equivalent manipulation to increasing the cost of a reinforcer (Hursh, Raslear, Shurtleff, Bauman, & Simmons, 1988). Accordingly, the present study revealed that both the initial disruption and the subsequent recovery of responding were distinctly related to the contingencies . Finally, if behavioral mechanisms of d responding were found to be related to changes in reinforcer sensitivity, then such a revelation even would successfull y integrate long standing findings from the early state of affairs in behavioral pharmacology research that level of food deprivation interacts with d amphetamine to modulate drug effects (e.g., M eginniss, 1967; Owen & Campbell, 1974 ). I dentifying behavio ral mechanisms of drug effects and subsequent tolerance to those effects necessitates a series of careful experiments , each one building upon its antecedent, be conducted. This research serves as one step towards achieving th at ultimate goal .
106 LIST OF R EFERENCES Alder, L. A., Alperine, S., Leon, T., & Faraone, S. (2014). Clincal effects of lisdexamfetamine and mixed amphetamine salts immediate release in adult ADHD: results of a crossover design clinical trial. Postgraduate Medicine, 126 , 17 24. Ameri can Psychiatric Association. (2013 ). Diagnostic and statistical manual of mental disorders (5 th ed.). Washington, DC: Author. Barrett, J. E. (1976). Effects of alcohol, chlordiazepoxide, cocaine and pentobarbital on responding maintained under fixed inter val schedules of food or shock presentation. Journal of Pharmacology and Experimental Therapeutics, 196 , 605 615. Baum, W. M. (1993). Performances on ratio and interval schedules of reinforcement: data and theory. Journal of the Experimental Analysis of Be havior, 59 , 245 264. Bignami, G., & Gatti, G. L. (1969). Analysis of drug effects on multiple fixed ratio 33 fixed interval 5 min in pigeons. Psychopharmacologia, 15 , 310 332. Bowen, S. E., Fowler, S. C., & Kallman, M. J. (1993). Effects of variation in c hronic dose of cocaine on contingent tolerance as assessed in a milk drinking task. Psychopharmacologia, 113 , 67 75. Branch, M. N. (1979). Consequent events as determinants of drug effects on schedule controlled behavior: modification of effects of cocain e and d amphetamine following chronic amphetamine administration. Journal of Pharmacology and Experimental Therapeutics, 210 , 354 360. Branch, M. N. (1984) . Rate dependency, behavioral mechanisms, and behavioral pharmacology. Journal of the Experimental Analysis of Behavior, 42 , 511 522. Branch, M. N. (1990). Cocaine tolerance: interactions among random ratio and random interval reinforcement schedule parameters and repeated exposure to cocaine. Drug Development Research, 20 , 19 30. Branch, M. N. (200 0). Gaining (on) momentum. Behavioral and Brain Sciences, 23 , 92 93. Branch, M. N. (200 6 ). How r esearch in b ehavioral p harmacology i nforms b ehavioral s cience . Journal of the Experimental Analysis of Behavior, 85 , 407 423. Branch, M. N., & Dearing, M. E. ( 1982). Effects of acute and daily cocaine administration on performance under a delayed matching to sample procedure. Pharmacology, Biochemistry, and Behavior, 16 , 713 718. Branch, M. N., & Gollub, L . (1974). A detailed analysis of the effects of d ampheta mine on behavior under fixed interval schedules. Journal of the Experimental Analysis of Behavior, 21 , 519 539 .
107 Branch, M. N., Walker, D. J., & Brodkorb, G. W. (1999). Attenuation of cocaine induced response rate increases during repeated administration de spite increases in rate of reinforcement. Psychopharmacologia, 141 , 413 420. Branch, M. N., Wilhelm, M. J., & Pinkston, J. W. (2000). A comparison of fixed and variable doses of cocaine in producing and augmenting tolerance to its effects on schedule cont rolled behavior. Behavio u ral Pharmacology, 11 , 555 569. Brocco, M. J., & McMillan, D. E. (1983). Tolerance to d amphetamine and lack of cross tolerance to other drugs in rats under a multiple schedule of food presentation. Journal of Pharmacology and Ex perimental Therapeutics, 224 , 34 39. Carlton, P. L., & Wolgin, D. L. (1971). Contingent tolerance to the anorexigenic effects of amphetamine. Physiology and Behavior, 7 , 221 223. Carney, J. M. (1977). Pargyline induced increases in sensitivity to the effe cts of drugs on operant behavior in pigeons. Psychopharmacologia, 52, 97 102. Catania, A. C., Matthews, T. J., Silverman, P. J., & Yohalem, R. ( 1977 ). Yoked variable ratio and variable interval responding in pigeons . Journal of the Experimental Analysis of Behavior, 28 , 155 161. Catania , A. C., & Reynolds, G. S. ( 196 8 ). A quantitative analysis of the responding maintained by interval schedules of reinforcement. Journal of the Experimental Analysis of Behavior, 11 , 327 383 . Caul, W. F., Jones, J. R., & Ba and body weight: the role of adaptive processes. Behavioral Neuroscience, 102 , 441 450. Christensen Szalanski, J. J., Goldberg, A. D., Anderson, M. E., & Mitchell, T. R. (1980). Deprivation, delay of reinforcement, and the selection of behavioural strategies. Animal Behavior, 28 , 341 346. Clemow , D. B., & Walker, D. J. ( 2014 ). The potential for misuse and abuse of medications in ADHD: a review. Postgraduate Medicine, 126 , 64 81. Cohen, S. L. (1986). A pharmacological examination of the resistance to change hypothesis of response strength. Journal of the Experimental Analysis of Behavior, 46 , 363 379. Cohen , S. L., & Branch , M. N. (1991 ). Food paired stimuli as conditioned reinforcers: effec ts of d amphetamine . Journal of the Exper imental Analysis of Behavior, 56 , 277 288. Corfield Sumner, P. K., & Stolerman, I. P. (1978). Behavioral tolerance. In D. E. Blackman & D. J. Sanger (Eds.), Contemporary research in behavioral pharmacology (pp. 391 448). New York: Plenum.
108 Dallery, J., & Lancaster, J. S. (1999). Individual differences in behavioral tolerance to amphetamine and the economic context of reinforcement loss. Behavio u ral Pharmacology, 10 , 257 269. Dandy , K . L . , & Gatch, M . B. (2009). The effects of chronic cocaine exposure on impulsivity in rats. Behavio u ral Pharmacology , 20 , 400 405. De la Harran Arita , A. K., & Garcia Garcia, F. (2014). Narcolepsy as an immune mediated disease. Sleep Disorders, 2014 , 1 6. Demellweek, C., & Goudie, A. J . (1982). The role of reinforcement loss in the development of tolerance to amphetamine anorexia. IRCS Medical Science, 10 , 903 904. Dews, P. B. (1955). Studies on behavior. I. Differential sensitivity to pentobarbital of pecking performance in pigeons de pending on the schedule of reward. Journal of Pharmacology and Experimental Therapeutics, 113 , 393 401. Dews , P. B. (1956). Modification by drugs of performance on simple schedules of positive reinforcement. Annals of the New York Academy of Sciences, 65 , 268 281. Dews, P. B. (1957). Studies on behavior. III. Effects of scopolamine on reversal of a discriminatory performance in pigeons, Journal of Pharmacology and Experimental Therapeutics, 119 , 343 353. Dews, P. B. (1958). Studies on behavior. IV . Stimul an t actions of methamphetamine. Journal of Pharmacology and Experimental Therapeutics, 1 22 , 137 147 . Dews, P. B. (1978). Studies on responding under fixed interval schedules of reinforcement: II. The scalloped pattern of the cumulative record. Journal of the Experimental Analysis of Behavior, 29 , 67 75. Durgin, A., Porter, L., Bradley, K., Laraway, S., & Poling, A. (2009). Cocaine and automaintained responding in pigeons: Rate reducing effects and tolerance thereto with different durations of food deliver y. Pharmacology, Biochemistry, and Behavior, 93 , 460 464. Eckerman, D. A., & Edwards, J. D. (1978). Schedule dependent effect of d amphetamine on pausing by pigeons . Pharmacology, Biochemistry, and Behavior, 8 , 319 321. Ferster , C. B., & Skinner, B. F. ( 19 57 ). Schedules of reinforcement . New York: Appleton Century Crofts . Fry, W., Kelleher, R. T., & Cook, L. (1960). A mathematical index of performance on fixed interval schedules of reinforcement . Journal of the Experimental Analysis of Behavior, 193 199. Gibson, D. A. (1967). Effects of d amphetamine on multiple schedule performance in the pigeon. Psychonomic Science, 7 , 3 4.
109 Ginsburg, B.C., Pinkston, J.W., & Lamb, R.J. (2011). Reinforcement magnitude modulation of rate dependent effects in pigeons and rats. Experimental and Clinical Psychopharmacology, 19 , 285 294. Gollub, L. R. (1964). The relations among measures of performance on fixed interval schedules. Journal of the Experimental Analysis of Behavior , 7 , 337 343. Gonzalez , F. A., & Byrd, L. D. (1 977). Mathematics underlying the rate dependency hypothesis. Science, 11 , 546 550. Gonzalez, F. A., & Goldberg, S. R. (1977). Effects of cocaine and d amphetamine on behavior maintained under various schedules of food presentation in squirrel monkeys. Jour nal of Pharmacology and Experimental Therapeutics, 201, 33 43. Grace, R.C., Schwendiman, J.W. , & Nevin, J.A. (1998). Effects of unsignaled delay of reinforcement on preference and resistance to change. Journal of the Experimental Analysis of Behavior, 69 , 247 261. Harper, D. N. (1999). Drug induced changes in responding are dependent on baseline stimulus reinforcer contingencies. Psychobiology, 27 , 95 104. Harris, R. A., & Snell, D. (1980). Effects of acute and chronic administration of phenobarbital a nd d amphetamine on schedule controlled behavior. Pharmacology, Biochemistry, and Behavior, 12 , 47 52. Harris, R. A., Snell, D., & Loh, H. H. (1978). Effects of d amphetamine, monomethoxy amphetamines and hallucinogens on schedule controlled behavior. J ournal of Pharmacology and Experimental Therapeutics, 204 , 103 117. Harris, R. A., Snell, D., & Loh, H. H. (1979). Effects of chronic d amphetamine treatment on schedule controlled behavior. Psychopharmacologia, 63 , 55 61. Heal, D. J., Smith, S. L., Gosd en, J., & Nutt, D. J. (2013). Amphetamine, past and present a pharmacological and clinical perspective. Journal of Psychopharmacology, 27 , 479 496. Heyman, G. M. ( 1983 ). A parametric examination of the hedonic and motoric effects of drugs: pimozide and amphetamine . Journal of the Experimental Analysis of Behavior, 40 , 113 122. Hoffman, S. H., Branch, M. N., & Sizemore, G. M. (1987). Cocaine tolerance: Acute versus chronic effects as dependent upon fixed ratio size. Journal of the Experimental Analysis o f Behavior, 47 , 363 376. Hughes, C. E., & Branch, M. N. (1991). Tolerance to and residual effects of cocaine in squirrel monkeys depend on reinforcement schedule parameter. Journal of the Expe rimental Analysis of Behavior, 56 , 345 360.
110 Hursh, S. R. , Rasl ear, T. G., Shurtleff, D., Bauman, R., & Simmons, L. (1988). A cost bene fit analysis of demand for food. J Journal of the Experimental Analysis of Behavior, 50 , 419 440 . Huskinson , S . L . , Krebs , C . A . , & Anderson , K . G. (2012). Strain differences in delay discounting between Lewis and Fischer 344 rats at baseline and following acute and chronic administration of d amphetamine. Pharmacology, Biochemistry, and Behavior, 101 , 403 416. Jimenez Gomez , C., & Shahan, T. A. ( 2007 ). Resistance to change of alcohol self administration: effects of alcohol delivery rate on disruption by extinction and naltrexone. Behavioural Pharmacology, 18 , 161 169. Katz, J. L., Dworkin, S. I., Dykstra, L. A., Carter, R. B., & Witkin, J. M. (1990). Some behavioral effects of repeat ed d amphetamine administrations. Drug Development Research, 20 , 31 41. Kelleher, R. T., & Morse, W. H. (1964). Escape behavior and punished behavior. Federation Proceedings, 23, 808 817. Kelleher, R. T., & Morse, W. H. (1968). Determinants of the specif icity of behavioral effects of drugs. Ergebnisse der Physiologie, Biologischen Chemie, und Experimentellen Pharmakologie, 60, 1 56. Kornblith, C. L., & Hoebel, B. G. (1976). A dose response study of anorectic drug effects on food intake, self stimulation, and stimulation escape. Pharmacology, Biochemistry, and Behavior, 5 , 215 218. Krebs, C. A., & Anderson, K. G. (2012). Preference reversals and effects of d amphetamine on delay discounting in rats. Behavioural Pharmacology, 23 , 228 240. Leander, J. D., & McMillan, D. E. (1974). Rate dependent effects of drugs. I. Comparisons of d amphetamine, pentobarbital and chlorpromazine on multiple and mixed schedules. Journal of Pharmacology and Experimental Therapeutics, 188 , 726 739. LeSage, M. G., Stafford, D., & Glowa, J. R. (2004). Effects of anorectic drugs on food intake under progressive ratio and free access conditions in rats. Journal of the Experimental Analysis of Behavior , 82 , 275 292. Lucki , I., & DeLong, R. E. (1983). Control rate of response or reinf effect on behavior. Journal of the Experimental Analysis of Behavior , 40 , 123 132. MacPhail, R. C., & Gollub, L. R. (1975). Separating the effects of response rate and reinforcement frequency in the rate dependent effects of amph etamine and scopolamine on the schedule controlled performance of rats and pigeons. Journal of Pharmacology and Experimental Therapeutics, 194 , 332 342.
111 Maguire, D. R., Rodewald, A. M., Hughes, C. E., & Pitts, R. C. ( 2009 ). Rapid acquisition in concurren t schedules: effects of d amphetamine on sensitivity to reinforcement amount. Behavioural Processes, 81 , 238 243. Marusich, J. A., & Branch, M. N. (2008). Differences in the behavioral time course of effects of rate increasing and rate decreasing doses o f cocaine in pigeons. Pharmacology, Biochemistry, and Behavior, 89 , 150 159. Marusich, J. A., Branch, M. N., & Dallery, J. ( 2008 ). Limitations to the generality of cocaine locomotor sensitization. Experimental and Clinical Psychopharmacology, 16 , 282 292 . Mazur , J. E., & Hyslop, M. E. ( 1982 ). Fixed ratio performance with and without a postreinforcement timeout. Journal of the Experimental Analysis of Behavior, 38 , 143 155. McKearney, J. W. (1970). Rate dependent effects of drugs: modification by discrimina tive stimuli of the effects of amobarbital on schedule controlled behavior. Journal of the Experimental Analysis of Behavior, 14 , 167 175. McKearney, J. W. ( 1974 ). Effects of d amphetamine, morphine, and chlorpromazine on responding under fixed interval s chedules of food presentation or electric shock presentation. Journal of Pharmacology and Experimental Therapeutics, 190 , 141 153. McKearney, J. W., & Barrett, J. E. (1975). Punished behavior: increases in responding after d a mphetamine. Psychopharmacologi a, 41 , 23 26. McMillan , D. E. (1969) . Effects of d amphetamine on performance under several parameters of multiple fixed ratio, fixed interval schedules. Journal of Pharmacology and Experimental Therapeutics, 167 , 26 33. McMillan , D . E. (1973). Drugs and punished responding. I. Rate dependent effects under multiple schedules . Journal of the Experimental Analysis of Behavior, 19 , 133 145 . Meginniss, R. F. (1967). Interaction of d amphetamine and food deprivation on fixed ratio behavior of pigeons. Psycholo gical Reports, 20 , 355 358. Michael, J. (1982). Distinguishing between discriminative and motivating functions of stimuli. Journal of the Experimental Analysis of Behavior , 37 , 149 155. Michael , J. (1993). Establishing operations. The Behavior Analyst , 16 , 191 206 . Miller, M. L., & Branch, M. N. (2002). Role of dose order in the development of tolerance to effects of cocaine on schedule controlled behavior in pigeons. Psychopharamacology, 163 , 302 309. Miller, M. L., & Branch, M. N. (2004). Tolerance to re sponse disruptive effects of cocaine is facilitated by opportunity to respond in the presence of drug. Pharmacology, Biochemistry, and Behavior, 79 , 43 54.
112 Milloy , S., & Glick, S. D. ( 1976 ). Factors influencing tolerance to d amphetamine induced anorexia in rats. Archives Internationales de Pharmacodynamie et de ThÃ©rapie, 221 , 87 95 . Minervini, V., & Branch, M. N. (2013) . administration of a dose without detected effects on response rate or pause. Journal of the Experimental Analysis of Behavior , 100 , 316 332. Morse , W. H., & Herrnstein, R. J. ( 1956 ). Effects of drugs on characteristics of behavior maintained by complex schedules of intermittent positive reinforcement. Annals of the New York Academy of Scie nces, 56 , 303 317. Nevin, J. A., & Grace, R. C. (2000). Behavioral momentum and the Law of Effect. Behavioral and Brain Sciences , 23 , 73 130. Nevin, J. A., Grace, R. C., Holland, & McLean, A. P. (2001). Variable ratio versus variable interval schedules: R esponse rate, resistance to change, and preference. Journal of the Experimental Analysis of Behavior, 76 , 43 74. Nevin, J. A. (1974). Response strength in multiple schedules. Journal of the Experimental Analysis of Behavior, 21 , 389 408. Nevin, J.A., & Sha han, T. A. (2011). Behavioral momentum theory: equations and applications. Journal of Applied Behavior Analysis, 44 , 877 895. Nevin, J. A., Tota, M. E., Torquato, R. D., & Shull, R. L. ( 1990 ). Alternative reinforcement increases resistance to change: Pav lovian or operant contingencies? Journal of the Experimental Analysis of Behavior, 53, 359 379. Nickel, M., Alling, K., K leiner, M., & Poling, A. (1993). Fixed ratio size as a determinant of tolerance to cocaine: is relative or absolute size important? Beh avioural Pharmacology, 4 , 471 478. Odum, A. (2000). Behavioral pharmacology and timing. Behavioural Processes, 57 , 107 120. Odum, A. L., Lieving, L. M., & Schaal, D. W. ( 2002 ). Effects of d amphetamine in a temporal discrimination procedure: selective chan ges in timing or rate dependency? Journal of the Experimental Analysis of Behavior , 78 , 195 214. Owen, J. E. (1960). The influence of dl , d , and l amphetamine and d methamphetamine on a fixed ratio schedule. Journal of the Experimental Analysis of Beha vior , 3 , 293 310. Owen , J . E ., & Campbell, J . T. (1974). Effects of methamphetamine on fixed ratio responding of food satiated and food deprived rats. Psychopharmacologia, 34 , 223 2 32. Palya, W. L., & Walter, D. E. (1993). A powerful, inexpensive experimen t controller or IBM PC interface and experiment control language. Behavior Research Methods, Instruments & Computers, 25 , 127 136.
113 Pinkston, J. W., & Branch, M. N. (2003). Sensitization to cocaine in pigeons: interaction with an operant contingency. Exper imental and Clinical Psychopharmacology, 11 , 102 109. Pinkston, J. W., & Branch, M. N. (2004). Repeated post or presession cocaine administration: roles of dose and fixed ratio schedule. Journal of the Experimental Analysis of Behavior, 81 , 169 188. Pinks ton, J. W., & Branch, M .N. (2010). Acute and chronic effects of cocaine on the spontaneous behavior of pigeons. Journal of the Ex perimental Analysis of Behavior, 94 , 25 36. Pinkston, J. W., Ginsburg, B. C., & Lamb, R. J. (2009). Examination of reinforcem ent magnitude on the pharmacological disruption of fixed ratio performance. Experimental and Clinical Psychopharmacology, 17, 237 246. Pinkston, J. W., Ginsburg, B. C., & Lamb, R. J. ( 2014 ). Reinforcer magnitude and rate dependency: an evaluation of resist ance to change mechanisms. Behavioural Pharmacology, 25 , 629 636. Pitts, R. C. (2014) . Reconsidering the concept of behavioral mechanisms of drug action. Journal of the Experimental Analysis of Behavior , 101 , 422 441. Pitts , R. C., & McKinney, A. P. (20 05). Effects of methylphenidate and morphine on delay discount functions obtained within sessions. Journal of the Experimental Analysis of Behavior , 83 , 297 314. Poling, A., Byrne, T., Christian, L., & Lesage, M. G. ( 2000 ). Effects of cocaine and morphin e under mixed ratio schedules of food delivery: support for a behavioral momentum analysis. Pharmacology, Biochemistry, and Behavior, 66 , 313 321. Quick, S. L., & Shahan, T. A. (in press). The impact of d amphetamine and SCH23390 on behavioral momentum o f food seeking and reinstatement in rats. Behavio u ral Pharmacology . [Epub ahead of print] Rees, D. C., Wood , R. W., Laties, V.G. (1987). Stimulus control and the development of behavioral tolerance to daily injections of d amphetamine in the rat . Journa l of Pharmacology and Experimental Therapeutics , 240, 65 73. Salisbury, J. J., & Wolgin, D. L. (1985). Role of anorexia and behavioral activation in amphetamine induced suppression of feeding: Implications for understanding tolerance. Behavioral Neuroscien ce, 99 , 1153 1161. Sanger, D. J., & Blackman, D. E. (1975). Rate dependent effects of drugs on the variable interval behavior of rats. Journal of Pharmacology and Experimental Therapeutics, 194, 343 350 . Sanger, D. J., & Blackman, D. E. (1976). Rate depend ent effects of drugs: a review of the literature. Pharmacology, Biochemistry, and Behavior, 4 , 73 83.
114 Schama, K. F., & Branch, M. N. (1989). Tolerance to effects of cocaine on schedule controlled behavior: effects of fixed interval schedule parameter. Phar macology, Biochemistry, and Behavior, 32 , 267 274. repeated administration. Journal of the Experimental Analysis of Behavior, 62 , 45 56. Schuster, C. R. (1978). The oretical basis of behavioral tolerance: Implications of the phenomenon for problems of drug abuse. In Krasnegor, N. A. (Ed.), Behavioral Tolerance: Research and Treatment Implications, National Institute Drug Abuse Research Monograph (pp. 4 17), Washington , DC: U.S. Government Printing Office. Schuster, C. R., Dockens, W. S., & Woods, J. H. (1966). Behavioral variables affecting the development of amphetamine tolerance. Psychopharmacologia, 9, 170 182. Simon, N. W., Gilbert, R. J., Mayse, J. D., Bizon, J. L ., & Setlow, B. (2009). Balancing risk and reward: A rat model of risky decision making. Neuropsychopharmacology, 34 , 2208 2217. Smith, C. B. (1964). Effects of d amphetamine upon operant behavior of pigeons: enhancement by reserpine. Journal of Pharmaco logy and Experimental Therapeutics, 146, 167 174. Smith, J. B. (1986). Effects of chronically administered d amphetamine on spaced responding maintained under multiple and single component schedules. Psychopharmacologia, 88, 296 300. Spealman, R. D., Katz, J. L., & Witkin, J. M. (1978). Drug effects on responding maintained by stimulus reinforcer and response reinforcer contingencies. Journal of the Experimental Analysis of Behavior , 30 , 187 196. Stafford , D., & Branch, M. N. ( 1996 ). Relations between dose magnitude, subject sensitivity, and the development of tolerance to cocaine induced behavioral disruptions in pigeons. Behavioural Pharmacology, 7 , 324 333. Stitzer , M., & McKearney, J. W. ( 1977 ). Drug effects on fixed interval responding with pause requ irements for food presentation. Journal of the Experimental Analysis of Behavior , 25 , 51 59. Ta, W. M., Pitts, R. C., Hughes, C. E., McLean, A. P., & Grace, R. C. (2008). Rapid acquisition of preference in concurrent chains: effects of d amphetamine on s ensitivity to reinforcement delay. Journal of the Experimental Analysis of Behavior, 89 , 71 91. Tanno, T., Maguire, D. R., Henson, C., & France, C. P. (2014). Effects of amphetamine and methylphenidate on delay discounting in rats: interactions with ord er of delay presentation. Psychopharmacologia, 231 , 85 95. Thurlby, P. L., & Samanin, R. (1981). Effects of anorectic drugs and prior feeding on food rewarded runway behavior. Pharmacology, Biochemistry, and Behavior, 14 , 799 814.
115 Tilson, H. A., & Sparb er, S. B. (1973) . The effects of d and l amphetamine on fixed interval and fixed ratio behavior in tolerant and nontolerant rats. Journal of Pharmacology and Experimental Therapeutics, 187 , 372 379. Tsunematsu, S. (2000). Effort and time cost effects o n demand curves for food by pigeons under short session closed economies. Behavioural Processes, 53 , 47 56. Van Haaren, F., & Anderson, K. G. (1994). Behavioral effects of acute and chronic cocaine administration in male and female rates: effects of fixed ratio schedule parameters. Behavioural Pharmacology, 5 , 607 614. Wade, T. R., de Wit, H., & Richards, J. B. (2000). Effects of dopaminergic drugs on delayed reward as a measure of impulsive behavior in rats. Psychopharmacologia, 150 , 90 101. Weaver, M. T., & Branch, M. N. (2008). Tolerance to the effects of cocaine on behavior under a response initiated fixed interval schedule. Journal of the Experimental Analysis of Behavior, 90 , 207 218. Weaver, M. T., Dallery, J., & Branch, M. N. (2010). Response to pography in behavioral tolerance to cocaine with rats. Behavio u ral Pharmacology, 21 , 660 667. Wellman, P. J., Davis, K. W., Clifford, P. S., Rothman, R. B., & Blough, B. E. (2009). Changes in feeding and locomotion induced by amphetamine analogs in rats. Drug and Alcohol Dependence, 100 , 234 239. Wentink, E. A. ( 1938 ). The effects of certain drugs and hormones upon conditioning. Journal of Experimental Psychology, 22 , 150 163. Will, B., & Checchinato, D. (1973). Effects of d amphetamine on operant beh aviour in the rat. Psychopharmacologia, 29 , 141 149. Wolgin, D. L. (1989). The role of instrumental learning in behavioral tolerance to drugs. In A. J. Gouldie & M. W. Emmett Oglesby (Eds.), Psychoactive drugs: Tolerance and sensitization (pp. 17 114). Cli fton, NJ: Humana Press. Wolgin, D. L. (2000). Contingent tolerance to amphetamine hypophagia: New insights into the role of environmental context in the expression of stereotypy. Neuroscience & Behavioral Reviews, 24, 279 294. Wolgin, D.L., & Wade, J.V. (1 995). Learned suppression of stereotypy in amphetamine treated rats: Implications for understanding tolerance to amphetamine 'anorexia.' Behavio u ral Pharmacology, 6 , 254 262. Yanovski, S. Z., & Yanovski, J. A. (2014). Long term drug treatment for obesity: a systematic and clinical review. Journal of the American Medical Association, 311 , 74 86.
116 Yoon, J. H., & Branch, M. N. (2004). Interactions among unit price, fixed ratio value, and dosing regimen in determining effects of repeated cocaine administration . Behavioural Processes, 67 , 363 381.
117 BIOGRAPHICAL SKETCH Vanessa Minervini grew up in Pawleys Island, SC . S he attended the College of Charleston, where s he became fascinated with the experimental analysis of behavior while shaping a lever press respons e in a rat as part of her laboratory course requirements . She c ompleted an honors thesis, the focus of which was food demand elasticity as a function of session duration in rats , and graduated from the College of Charleston in 2010 . Vanessa then attended t he University of Florida to complete her graduate training in the laboratory of Dr. Marc Branch . Throughout her graduate training, she has been passionate about elucidating behavioral principles in animal models and understanding the way in which environm ental manipulations affect responding. In addition to maintaining an active research interest in behavioral economics, the focus of her masters and doctoral research was identifying environmental variables that modulate the development of tolerance to the effects of central nervous system stimulants , such as cocaine and d amphetamine. Upon graduating from the University of Florida in 2015, Vanessa plans to complete a post doctoral fellowship that will allow her to pursue her interest in mechanisms of drug action, both behavioral and neurobiological, and the ways in which these mechanisms are involved in and contribute to drug dependence as well as therapeutic drug effects. Ultimately, she aims to obtain a position in an academic setting where she can produce meaningful research as well as train graduate students.