oN GNNIiA NR LSOéIAL 5:: NIX 1:4 2:: :5... X1 eunaaww.» ; :31. ; . .. O 3.4 . . . .r . fade. . DC? cm 0 NI mvEAN com PEI AP. 5_( J.- J. a? .1: 5 +Amx¢ £1.04». NTAEG... ,nx_._...5:,.,q.1.s..:..1 . .... 1...... .13.. .51.. Tariff: HE. lllllll lllllllllllljlllfiljll l/ “BRAR y 3 1293 00679 Will 1 ’ Michigan State . University This is to certify that the thesis entitled Social Organization in Japanese Quail (Coturnix coturnix jagonica): Appetitive and Consummatory Components presented by Robert Edward Otis has been accepted towards fulfillment of the requirements for Ph.D. degfiein Psychology @M’o‘f zm Major professors Date 2/24/72 M f 64:15: l 0-7639 l .4? ’. 993 ABSTRACT SOCIAL ORGANIZATION IN THE JAPANESE QUAIL (COTURNIX COTURNIX JAPONICA): APPETITIVE AND CONSUMMATORY COMPONENTS BY Robert Edward Otis Studies of avian social organization have utilized the social dominance theory for 50 years despite the con— fusing array of data it has generated. Objections to this theory are twofold: First, linear hierarchies are infre— quently found in avian societies and second, various defi— nitions of dominance are frequently found to be unrelated. Some of this confusion may be the result of meas- uring aggression without regard to its appetitive and consummatory aspects. Nine flocks of Japanese quail (Coturnix coturnix japonica) were used to test the follow— ing questions: 1. Can dominance behavior in quail be viewed as appetitive behavior for incentive-acquisition (food, water)? That is, are there differences in dominance relationships and aggressive behavior at different distances from incen— tives or between incentives? Robert Edward Otis 2. What are the relationships between the appeti- tive and consummatory aspects of food- and water-getting sequences of behavior? More specifically, are measures of dominance at varying distances from the incentive good predictors of consummatory activity, such as latency and total amounts consummed? 3. Can a dominance hierarchy, when based on ag— gression at a food source, be modified by depriving indi— viduals of food? Similarly, can a dominance hierarchy based on aggression at a water source be modified by depriving individuals of water? The results of this experiment indicated the incentive—related (or appetitive) character of aggressive behavior. Japanese quail were found to have different aggressive-response profiles at different distances from food and water incentives. No differences in response profiles were observed when incentives were absent. Fre- quency of aggression increased as quail moved closer to the incentives while no differences in amounts of aggres- sion were found across the cage when incentives were absent. A trend towards unilateral aggressive interaction was found as incentive-distance increased, reflecting a change from a peck—dominance to a peck—order form of social organization. When incentives were absent social organi— zation approximated a peck order. Robert Edward Otis Dominance hierarchies varied between different incentive-distances and between incentives. Dominance behavior in the appetitive phase could not be used to accurately predict measures of consummatory behavior (latency, total time and average time of incentive-access). However, when incentives were present, there were positive correlations between dominance hierarchies in the three incentive-distance sectors, indicating that the appetitive phase was a single component. Effects of incentive deprivation on dominance behavior could not be interpreted because of instability in dominance relation- ships among control groups. The traditional view of aggression as a consummatory event was re-examined, particularly as it related to the measure of dominance in studies of avian social organiza— tion. SOCIAL ORGANIZATION IN THE JAPANESE QUAIL (COTURNIX COTURNIX JAPONICA): APPETITIVE AND CONSUMMATORY COMPONENTS BY Robert Edward Otis A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology l972 To my Wife Barbara ii ACKNOWLEDGMENTS The author wishes to express his gratitude to Dr. Stanley C. Ratner for serving as chairman of the thesis committee and for his support and encouragement. Special thanks are also given to Dr. Ralph Levine for co-chairing the thesis committee in Dr. Ratner's absence. Dr. M. Ray Denny and Dr. Robert K. Ringer served on the thesis com- mittee and provided constructive criticisms and helpful suggestions for which the author is grateful. Special thanks are also given to the staff of the Poultry Science Department, to Sulo J. Hulkonen and Dennis Dodson in particular, for providing the quail as well as equipment and information essential for the care and maintenance of these birds. And to Gary F. Connor of the Psychology Department for providing many hours of technical assistance, a special thanks. iii TABLE OF CONTENTS Page DEDICATION. . . . . . . . . . . . . . . ii ACKNOWLEDGMENTS . . . . . . . . . . . . . iii LIST OF TABLES . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . viii LIST OF APPENDICES . . . . . . . . . . . . ix INTRODUCTION . . . . . . . . . . . . . . l Appetitive and Consummatory Aspects of Dominance . . . . . . . . . . . . 5 Specific Goals of this Experiment . . . . . 9 METHOD . . . . . . . . . . . . . . . . ll Subject Population . . . . . . . . . . ll Identification of Birds . . . . . . . . l6 Observation Cages and Room Conditions. . . . 17 Measures of Incentive—Distance . . . . . . 20 Measures of Dominance . . . . . . . . . 2l Recording Techniques . . . . . . . . . 22 Procedure . . . . . . . . . . . . . 23 Procedure for Study 1 . . . . . . . . 23 Procedure for Study 2 . . . . . . . . 26 Procedure for Study 3 . . . . . . . . 27 RESULTS AND DISCUSSION. . . . . . . . . . . 29 Dominance Behavior as Appetitive Behavior . . 29 Aggressive Behavior at Different Incentive- Distances . . . . . . . . . . . 29 Unilateral Aggressive Interactions Measured at Different Incentive—Distances . . . 35 Dominance Status Related to Incentive- Distance . . . . . . . . . . . 39 Comparisons Between Morning and Evening SessionS‘ . . . . . . . . . . . 44 iv Page Between- -Incentive Comparisons of Dominance Hierarchies . . . . . . 44 Dominance Behavior Related to Consummatory Activity . . . . . . . . . . . 44 Discussion of Dominance Behavior as Appetitive Behavior . . . . . . . . . . . . 51 Alternate Explanations. . . . . 52 Dominance Behavior and the Consummatory Theory . . . . . . . . . . . . 55 Modification of Dominance Behavior. . . . . 60 Discussion of Modification of Dominance Behavior . . . . . . . . . . . . 61 GENERAL DISCUSSION . . . . . . . . . . . . 64 The Stressful Nature of Dominance Studies . . 64 The Appetitive Nature of Aggression . . . . 66 Measures of Social Dominance. . . . . . . 68 SUMMARY. . . . . . . . . . . . . . . . 70 REFERENCES. . . . . . . . . . . . . . . 74 APPENDICES. . . . . . . . . . . . . . . 81 Table 10. LIST OF TABLES Page The number of birds in each group and the amount of time required to achieve group stability, the time of year each Study was begun and the number of Ss dying during this experiment . . . . . . . . . . . 15 Mean frequency of response occurrence for every possible dyadic interaction in Study 3 . 27 Results of analysis of variance for total amounts of aggression between Sectors in the Evening and Morning sessions of Study 3. . . 34 Mean prOportion of sessions per dyad in which unilateral aggressive interactions were observed . . . . . . . . . . . . . 37 Results of analyses of differences between amounts of unilateral and bilateral aggres- sion occurring in different Sectors for each incentive. . . . . . . . . . . . . 38 Correlations between dominance hierarchies measured in different Sectors during each phase of the Food session in Studies 2 and 3 . 4l Correlations between dominance hierarchies measured in different Sectors during each phase of the Water sessions in Studies 2 and 3 . . . . . . . . . . . . . . 43 Correlations between dominance hierarchies between Morning and Evening Sessions. . . . 45 Correlations between dominance hierarchies in similar Sectors during Food and Water sessions . . . . . . . . . . . . . 46 Correlations between Dominance Index and latency for initial eating and drinking behavior . . . . . . . . . . . . . 47 vi Table 11. 12. 13. Correlations between Dominance Index and total time of eating and drinking behavior. . . . Correlations between Dominance Index and average time spent at the food and water cups per visit. . . . . . . . . . Average correlation coefficients (rSav) for dominance hierarchies measured daily over each phase in the time-control group of Study 3. vii Page 49 50 61 LIST OF FIGURES Figure Page 1. Top and side views of observation cages. . . l8 viii LIST OF APPENDICES Appendix . Page A. Body Weights of Subjects . . . . . . . 82 B. General Characteristics of Aggressive Behavior in Japanese Quail . . . . . . 84 C. Dominance Correlated Variables . . . . . 97 D. Intragroup Response Matrices for Study 3 . . 106 ix INTRODUCTION Social dominance has been variously defined in terms of aggressive superiority of one individual over another and dominance relationships are known to exist in a stereotyped, predictable manner within social groups of many avian species. Social organization is frequently measured by allocating each individual to a rank order according to the number of flockmates dominated. This hierarchy reflects relative predictability of approach/ retreat patterns of behavior when individuals meet in a competitive situation. A dominant bird is usually charac- terized by pecking or threatening behavior (approach) . while the subordinate shows escape or avoidance behavior (retreat). Dominance hierarchies were first described by the Norwegian Schjelderup—Ebbe (1922) in a flock of barnyard fowl. Nearly 10 years went past before the first extensive investigations of dominance hierarchies were begun in this country by W. C. Allee (1931). Allee's associates, Guhl and Collias, were preeminent in carrying this interest into the 1960's and today Allee's social dominance theory stands as the most widely used social-structuring mechanism for avian groups. Although the dominance theory was primarily developed from observations of captive domestic fowl researchers have felt free to utilize these theoreti- cal constructs across a wide range of species, both in the field and in the laboratory (see reviews in Armstrong, 1965, and Noble, 1939). The social dominance theory as developed by Allee and his associates is founded on two basic assumptions. First, a continuum of rank-order criteria is assumed to exist in all social groups of birds and second, dominance is assumed to be a cluster of interrelated behaviors which ramify into other social relationships. When ir- regularities have occurred in observations (ex. nonlinear hierarchies) researchers have tended to modify the theory 3d hog rather than critically examine the theory's basic postulates (Gartlan, 1968). Recently, however, evidence has been gathered which renders both postulates invalid for primate societies (Gartlan, 1968; Bernstein and Sharpe, 1966; Rowell, 1966). The validity of these assumptions for avian societies has never been critically examined. The first postulate assumes the dominance relation— ship to be transitive, allowing all individuals in a group to be ordered along on ordinal scale according to the number of flockmates dominated. Among avian groups, however, "triangular" (and "square") relationships are frequently reported where, for example, one bird is dominant to a second and the second is dominant to a third which, however, is dominant to the first (cf. Masure and Allee, 1934a). Researchers have generally eXplained intransitivity in any one of three ways: (1) as being caused by situational influences on aggression, or (2) because insufficient time was allowed for flock stabilization, or (3) because observational techniques were inadequate (cf. Guhl and Fischer, 1969; Masure and Allee, 1934a). Holabird (1955, p. 253). for example, suggested subordinate chickens to be "belligerent" when they pecked birds normally dominant to them and Armstrong (1965, p. 269) called such interactions "accidental." Two forms of linear hierarchies were identified by Allee and his associates (Masure and Allee, 1934a). In the peck-right or peck order society aggression between dyadic members was unilateral with one bird always dominating the other. In the peck—dominance society aggression was bilateral, of the "now-one-wins—now-the— other" variety and dominance was assigned to the more frequent winner. Lack of a precise operational definition of these two hierarchy forms has led to arbitrary decisions as to which organization is being observed. Domestic fowl are always considered as fitting a peck order criteria yet, upon close examination, aggression between dyad members in flocks of fowl is frequently bilateral rather (than unilateral (cf. Guhl, 1968; Masure and Allee, 1934a; Williams and McGibbon, 1956). Forcing data from a peck—dominance organization into a linear hierarchy model is obviously convenient for statistical analyses (e.g. rank order correlations) but it avoids a critical examina- tion of the underlying theoretical assumptions. The second postulate of the social dominance theory concerns the definition of dominance as a cluster of inter— related responses. Relationships between dominance hierarchies and other social interactions have produced a confusing array of correlational figures. The social dominance theory would predict highly significant cor- relations between dominance and other social behaviors which are essential for survival and reproduction. How— ever, little or no relationship has been found between dominance and roosting priority or clumping organization (Andrew, 1957a; Guhl, 1953; L111, 1968), approaching and consuming food (Guhl, 1953; Masure and Allee, 1934a), frequency of mating, including courting, crouching and treading (Guhl, 1953; Guhl and Warren, 1946; Guhl, Collias and Allee, 1945), number of eggs laid (Guhl, 1953), dusting and social preening (Lill, 1968), crowing (Allee, Collias and Lutherman, 1939), scratching and random walking (Banks, 1956), and approaching strange objects (Andrew, 1957; Ratner and Denburg, 1959). Dominance hierarchies based on fighting often show little reliability across various localities (Masure and Allee, 1934b), between paired and flock situations (Guhl, 1953), and between seasons (Bennett, 1939; Coutlee, 1967; Masure and Allee, 1934a,b; Shoemaker, 1939). That is, a bird may be dominant in one part of the cage but not in another, in a paired contest but not in a flock situation, or he may be dominant to females during the winter but not during the breeding season. Furthermore, dominance hierarchies do not represent an accurate gradient of aggressiveness (Guhl, 1968; Lill, 1968). In fact, various components of fighting have been found not to correlate with each other (Williams and McGibbon, 1956). Considering the foregoing evidence it is of interest that only a few researchers have spoken out against the unitary dimension model of social dominance and that the theory still remains essentially unchanged from its original conception. Scott (1956) long ago indicated that a description of any society solely in terms of dominance-subordination relationships was not enough. Appetitive and Consummatory Aspects of Dominance Some of the confusion that has arisen with regards to social dominance may be due to the wide variety of responses used in its measurement. Dominance has usually been defined in terms of aggressive responses observed during competitive situations where access to some incen- tive has been restricted. Aggression in these cases is appetitive or instrumental in incentive-getting. Since appetitive behaviors are specific to their consummatory stimuli, measures of aggression in various appetitive situations might not yield the same dominance orders. Even when the appetitive motor responses are similar in the different cases, they are directed toward different goals and involve responsiveness to different stimuli (Hinde, 1966). A bird will show one type of appetitive behavior when seeking food, another when selecting a perch or roost, and still another when attempting copula— tion. In group situations, aggressive responses are appetitive components of most of these behavior classes. In effect, dominance hierarchies can be conceptualized around different consummatory classes which have aggression— appetitive components. It is therefore not surprizing that researchers measuring dominance in group situations over long periods of time and without regard to consummatory classes often report fluctuating heirarchies (cf. Bennett, 1939). Of 63 studies sampled where dominance was measured in birds (57% of which were domestic fowl), 22 recorded aggressive behavior at a food source, two measured aggres— sion over food plus other incentives, one was in reference to perch attainment while 38 studies measured aggression during paired contests or in flock situations where the incentive conditions were difficult to define. Dominance hierarchies measured on the basis of these response classes would be expected to interrelate only to the extent that response components and motivational state (i.e. satiation or deprivation of consummatory stimulus) were similar. A food satiated bird, for example, might rank low in a hierarchy based on aggressive deference at a food source, but rank high, perhaps, in competition for a mate or for a perch or roost. Fighting over territory or in paired-contests is consummatory behavior since the response sequence terminates after the other member avoids or escapes. Aggression in these cases may not correlate highly with aggression which occurs as appetitive elements of other behavioral sequences. Species in which males become dominant to females (or vice versa) during the breeding season may only re- flect the fact that dominance hierarchies vary according to the consummatory class being measured. During the breeding season fighting may be appetitive to sexual behavior whereas fighting during other times of the year may be related to other goals. Likewise, territoriality in peck—dominance societies (e.g. pigeons) may often reflect differences in aggression around various incen- tives (e.g. territory, food, perch). The relationship between dominance rank and consum— matory behavior is not well understood. This relationship has often been found to be very unpredictable. For example, dominant birds, as measured by aggression around a food source, may or may not eat sooner or eat more food than subordinates (Marler, 1955a). When other consummatory classes are related to this dominance hierarchy, the cor— relations are frequently low and unreliable as was mentioned earlier. Since consummatory behaviors are influenced by conditions of deprivation and satiation (Denny and Ratner, 1970) and since appetitive responses are functionally related to the consummatory act, modification of an indi— vidual's dominance status (e.g. upward mobility) would be a predictable consequence of food deprivation, given that the dominance hierarchy was based on aggression at the food source. This same individual would predictably rise _in other dominance hierarchies to the extent that the consummatory classes were interrelated (e.g. thirst and hunger, Hinde, 1966, p. 150). This hypothesis has not been tested and literature support is scattered. Collias (1950) noticed that sub— ordinate chickens, which had been starved for two days, approached and ate from the food source despite the presence of dominant peers. No records of aggression were reported, but it was apparent that appetitive responses for food—attainment increased in strength following food deprivation. Many other studies have reported increased aggression at the food source when the flock as a whole was placed on food deprivation (Ellis, 1966; Marler, 1957) and in some cases dominance reversals have been observed (Guhl and Allee, 1944). Specific Goals of this Experiment The present experiment was designed to test the hypothesis that measures of dominance are, in effect, measures of appetitive behavior for incentive acquisition and that dominance status measured at varying distances from the incentive or between incentives might vary as a function of differing motivational conditions. This hypothesis will be referred to as the consummatory theory of social dominance. Japanese quail (Coturnix coturnix japonica) were used as subjects. Cages were constructed which allowed the measurement of aggression in quail at different distances from food and water cups. Incentive— distance was presumed to reflect different components of a consummatory sequence with consummatory and highly specific appetitive components occurring nearest the incentive while aggression which was less specific to or unrelated to incentive-acquisition was assumed to occur at the furthest distances from the incentives. Utilizing this operational definition of the consummatory sequence the present eXperiment sought to answer the following questions: 1. Can dominance behavior in quail be viewed as appetitive behavior for incentive-acquisition (food, water)? That is, are there differences 10 in dominance relationships and aggressive behavior at different distances from incentives or between incentives? What are the relationships between the appetitive 'and consummatory aspects of food- and water- getting sequences of behavior? Are measures of dominance at varying distances from the incentive good predictors of consummatory activity, such as latency and total amounts consummed? Can a dominance hierarchy, when based on ag- gression at a food source, be modified by depriving individuals of food? Similarly, can a dominance hierarchy based on aggression at a water source be modified by depriving individuals of water? METHOD Subject Population The subjects used in this experiment were Japanese quail. In order to understand the ramifications of using the quail in a study of social organization, it is necessary to review its background information. The Japanese quail is a small galliform which has been placed in the family PHASIANIDAE. It is apparently one of a number of subspecies of the European Coturnix coturnix although Vaurie (1965, p. 293) has listed it as a separate species Coturnix japonica. It is native to the Orient, particularly to the islands of Japan. Several attempts have been made since 1870 to introduce this bird to the United States as a game animal. These attempts have all failed, perhaps due to what Labiskey (1961) calls a genetic loss of wildness and vigor due to domestication. As a result of these conservation efforts, researchers at Auburn University and the University of California recog- nized the quail's value as a biomedical research animal (National Academy of Sciences, 1969), and today its popularity as a subject for behavioral research appears to be spreading (cf. Farris, 1967; Reese and Reese, 1962; Selinger and Bermant, 1967). 11 12 In its natural habitat the Japanese quail is highly solitary and territorial in its social behavior (Wetherbee, 1961). Stanford (1957) reports that quail roost separately in both cold and hot weather. There is no mention of the formation of winter covies. In the spring this bird migrates in large flocks but only at night (Meinertzhagen, 1954; Taka-Tsukasa, 1935). During daylight the birds separate. Groups of quail have also been noted around dusting areas, but, except for mating pairs, the quail does not form any permanent social bonds. This suggests the "unnaturalness" of housing quail in flocks in a laboratory setting. Flocks of penned Japanese quail may form dominance hierarchies, but the exact nature of this organization and the factors that influence it are unknown. Only two researchers have investigated this behavioral process and their results, both unpublished, are conflicting. Preston (1961) studied the development of social behavior in Japanese quail up to the time that a dominance organi— zation was recognizable (between 9 and 10 weeks). He found that ”all of the quail in the flocks were controlled by a single dominant bird with little evidence occurring of a descending order among the subordinates" (p. 36). However, Preston reports little in the way of quantitative information regarding the nature of this social arrangement. 13 As a consequence his conclusions appear to be based more on anectodal observations than on carefully derived data. This oligarchy organization was not seen by Eynon (1968) but rather a highly linear peck order. Eynon made an excellent ethological analysis of the quail's aggressive and sexual behavior and, apparently as a side interest, described the peck order in a small group (5 females, 1 male) of birds over a three-week period. Dominance was observed at the feed and watering troughs. The observed hierarchy did not remain stable over this period, and the factors controlling this fluctuation were not identified. The reason why Eynon and Preston found different forms of social organization is not clear. Since neither research paper gives a detailed account of dominance rela— tionships within the flock, it is difficult to obtain a precise understanding of the quail's social organization. Three studies (1,2,3) were conducted with three groups of quail in each study. The mean body weights of all subjects (SS) before and after each study are reported in Appendix A. Each group was composed of four males and four females. This was an arbitrary decision to use half males. Woodard and Abplanalp (1967) found highest fertility measures in flocks containing one male for every two females. However, in nature quail appear to be monogamous unless there are an excess of females (Wetherbee, 1961). Since 14 little is known about sexual composition of quail flocks (e.g. during migration) and since intersex encounters are maximized when an equal number of each sex are present, it was decided to use a 1:1 sex ratio in the treatment groups. The use of eight birds to a group was based on some observations made earlier by this researcher. The size of the apparatus apparently favors groups of ten or smaller. Flocks of 15 and 13 were placed in the cage on separate occasions and within one month their numbers had decreased to ten in both cases. No further reduction in floCk size was ever observed. Since ten appeared to be the upper limit of flock size for the experimental cages, groups of eight were expected to further minimize any stress caused by pOpulation size. The genetic origin of these birds is unknown. The original stock has been maintained by the Poultry Science Department here at Michigan State University since the mid 1950's. The birds in this experiment represent at least the 25th generation of interbreeding within this closed flock (Dennis Dodson, personal communication). All SS were transferred to the observation cages between 11 and 16 days prior to the beginning of a study in order that flock aggression might stabilize. As shown in Table 1 many birds had to be replaced to achieve this stability (ex. 14 in Group 1 of Study 3). Highly 15 o . oH N m 0 SH 0 m 0 mm Loud: mH VH H m H NH OH m o NH m m m o mumsunmm NH 6 H m H HH 0 m 0 HH v m mm m HmQEoomo HH H H H hpgpm mcHHsm :mmmm Qmmmm mpspm mnowwm meMHmmm mpuHm pmHo nonfisz hpspm mama @060 CH “whopv mEHB mo quESZ . msouw mUSHm .pcmEHquxm mH u msHHSU mcH%© mm mo Hmnfidc may paw Gamma mmB Mpsvm £06m Ham» mo mSHp map .ngHHQMHm msoum m>mH£om ou UmHHsva mEHp mo pssoem map can msoum £06m CH mpuHQ mo Hmflasc m£B||.H MHm¢B l6 aggressive individuals frequently caused fatal wounds on their flockmates, resulting in both the aggressor and the wounded being replaced. All birds were randomly assigned to treatment groups. gs were 10, 14 and 11 weeks old when testing began for Studies 1, 2 and 3 respectively. Preston (1961) reported the appearance of a social organization in 10 week old quail, so these SS were of sufficient age. Dead birds were not replaced. Sickness, injury and death were recognized as a predictable occurrence in flocks of quail. Every effort was made to maintain healthy birds. Once each week gs were weighed and examined for ectoparasites and for other evidence of deterioration in general health. Identification of Birds When studying the social organization of birds it is necessary to be able to recognize each individual in the flock under observation. Quail in this experiment were identified according to (1) their sex, and (2) according to'a coded arrangement of metal bands on each leg. Since there were four of each sex in a cage, one of each sex had two leg bands on its left leg (referred to as MLL and FLL, where M means Male and F means Female); one had two leg bands on the right leg (MRR and FRR); one had one band on the left leg and one on the right leg (MLR and FLR); and 17 the final two gs had no bands at all (M and F). In addi— tion, the male's leg bands were painted light blue during Studies 2 and 3 so as to facilitate identification. M had spots of blue paint on both legs. Since quail have a tendency to remain in crouched positions, particularly the females and subordinates, their fluffed feathers and drooping primaries retard observation of leg bands. The blue paint greatly facilitated identification in these situations, especially when the sex of a bird was not obvious due to its posture. The female is relatively larger than the male and has a "spreckled" breast caused by a dark and light feather arrangement. In most cases sexual differences were readily apparent. Observation Cages and Room Conditions Three observation cages were constructed so as to allow measurement of aggressive interactions at varying distances from a food and water cup. Each cage measured 17 inches high, 18 inches wide (or deep) and 63 inches long (see Figure 1). The front and back walls were built of 5/8 inch hardware cloth through which observations could be made. The floor was similarly constructed to allow fecal material to drop through to the table some 6 inches below the cage floor. The top or ceiling consisted ofa sheet of 1/4 inch plywood. It could be removed to allow egg gathering and the giving of food. 18 .AmBOHHmV unmEQUMHm HHwoouonm muH auHs mso sumo Hmw>mn on pm>oEmu Gown mm: mmso “mums tam poow map pcsoum HHMB mco Any 30H> mch QH .mm3m Hmnuusm mechmmHocH mama u tam m muouomm new mac Hmumz may ummummc mchn é Houomm suHB .Umucmmmum mmz umum3 cmsz mmmo m£p mo paw Hmauo may on wmmum>mu GOHuMOHuHucmpH Houomm .Emwnmmouo HOOHM amHHm mHnu pcommn wmmo wnu mo ummu map mo mumfimcoo O Houomm .Emmnmmouo Hoon umHHm may :Hnqu pan < uouomm mUkuso mwum mnu mmmHumEoo m Houomm .mso poem may wasoum aHmHMHUwEEH wHHms may cHngz mmum map an pochmo mH 4 Houomm .pmucmm Iona msHmn mmB 000m on» MH mm czonm mum O cam m .d muouomm .AuanHv mmso HmuMB paw HummHv poem mo cowumooH m3osw Amy 3mH> mos .mwmmo GOHum>umeo mo m3mH> wUHm paw QOBII.H musmHm 3mu> «mum lac Hmumz 2/ O Coon 3mH> 909 Amy Hmumz HHS boom O O m «ll/,mEmemmOHO\& Hoon 19 Each cage had a water cup in the middle of one endewall and a similarly placed food cup at the other end. The water cup consisted of a modified Hart cup (H. W. Hart Manufacturing Co., Glendale, California) with the operant lever removed and a plastic tubing coupled to a water container located outside the cage. The food cup was made from a sheet metal eaves trough. Four-inch high walls built up along the sides of both the water and food cups allowed a two—inch wide access to each cup. Only one quail could eat or drink at any one time. Photocell systems were situated at the top rim of each cup, allowing measure- ment of the duration of time any individual remained in this area. Interruption of this light beam did not consti- tute eating behavior but its relationship to eating was obviously very high. The ceiling of each cage had two 25-watt light bulbs attached to it. Lights were on a fixed photoperiod of 10 hours dark, 14 hours light (8:00 AM to 10:00 PM). Humidity and temperature were subject to variation in the experimental room. A 24—hour Bendix hygrothermograph was used to record this fluctuation. Temperature averaged 74° F during this eXperiment but slowly rose approximately 2° F (72° to 74°) with the presence of the experimentors in the room. A range of 23 to 35% relative humidity was recorded during the four months of this experiment. 20 Measures of Incentive-Distance In order to answer the first two goals of this experiment, dominance behavior was measured at three distances from each incentive, defined as Sectors A, B and C. Cross-beams dividing the floor of the cage into three equal sectors were used to identify these three Sectors (see Figurelj. Dominance was always measured as to where the target bird was standing. Sector A was the area within the walls immediately surrounding the incentive cup, Sector B was the remaining area around the incentive walls defined by the nearest cross-beam in the floor, and Sector C was the rest of the cage beyond this cross-beam. Figure 1 shows these Sectors around the food cup only. When water was presented Sector identifi- cation reversed to the other end of the cage with Sector A being nearest to the water cup and Sectors B and C being increasingly further away. During Study 3 Sector B for each incentive was made smaller in size by constructing an imaginary boundary line (in parallel to a cross-beam) 10 inches from each incentive cup. These lines were defined by placing tape on the side walls of the cage, making them visible to the observors but not to the quail inside the cage. Sector C was consequently made larger in this process for both incentives. 21 Measures of Dominance It was pointed out earlier (Introduction) that definitions of dominance behavior have varied across species. In the present experiment dominance behavior was defined as the occurrence of supplanting or avoidance behavior. The specific agonistic responses involved are summarized below. 1. Peck--rapid thrusting of the beak, making contact with another bird usually on the head or shoulder region. Included in this category were pecks which resulted in grabs on the head, neck or back. Some grabs on the back or side region were oriented towards pulling out feathers and were not counted as aggressive pecks. The head or neck grab was also a component of the sexual response. Pecks to the beak, feet and leg region were frequently made in a slower, more deliberate manner, particularly when wounds were obvious in these regions. These pecks were not counted as aggressive pecks. Push—-causing another bird to be supplanted by a rapid thrust and contact with the breast or shoulder region. Threat--rapid thrusting with beak, stopping short of contact with the target bird or the approach of one bird causing another to sup- plant. Situations where one quail suddenly moved while others around it remained stationary were not counted as avoidance responses since the stimulus was not apparent. Kick—-use of the feet to supplant another. Usually observed as a quick shuffling motion of legs and feet while the aggressor was on the back of his subordinate. Similar motions of feet seen when quail are dusting or scratching the floor. Fight—~aggressive interaction involving two individuals facing each other using any of the above aggressive responses. 22 Supplants involving birds engaged in the following behavior were not counted as aggressive supplants: resting, dusting, preening, defecating, pecking walls, pecking floor and eating feces. It was felt that these behaviors repre- sented other consummatory sequences which, if counted, would constitute a source of variance in unilateral dominance relationships measured around food and water incentives. Components of the male's courtship behavior fre— quently caused supplanting or avoidance responses in other males and females. These responses included Neck-and- Body-Tonus, Toe Walk and Churring (Otis, 1968). These responses were recorded but analyzed only in a sexual context. Recording Techniques Two observors were involved; one recorded agonistic and other social interactions, the other reCorded sequential order of eating and drinking. Two channels of a lO-pen Esterline Angus event recorder were wired into the photocell circuits at the food and water cups in each cage. Since observations could be made on only one cage at a time, these two channels were switched to the cage of interest, allowing the other eight pens to be free for recording each indi— vidual's behavior. These eight pens were, in turn, wired 23 to a key board on one observor's chair, enabling her to record the sequential order of eating and drinking by each individual in the flock. Agonistic and other social behaviors were recorded by speaking into a tape recorder. Procedure Three studies were conducted (labeled 1, 2 and 3) with only slight differences in each. Each study involved three groups of quail (two experimental and one control) being observed over three phases of treatment. Each phase lasted for six days with a day of no observations interceding between phases. Ss were weighed and feces were removed from beneath the cages on this day between phases. Because of their overall similarity the procedure for Study 1 alone will be portrayed in detail followed by the changes which were made in the designs of Studies 2 and 3. Procedure for Study 1 During each phase (labeled I, II and III) observa- tions were made three times daily; during a nonincentive Morning session and during food and water sessions in the Evening. Morning sessions.-—Morning observations were con- cerned with social interactions which occurred following simultaneous access to both food and water. Both incentives were given between 9:00 and 9:45 a.m. (randomly staggered 24 at 15 minute intervals to accommodate each of the three groups) and 15-minute observations of each group were made between 10:30 a.m. and 11:45 a.m. following removal of the food and water. Records were kept of aggressive inter- actions occurring throughout the cage. Evening sessions.--Evening sessions began between 7:00 p.m. and 8:00 p.m. (randomly staggered at 30 minute intervals to accommodate each of the three groups) and involved the presentation of food and water during succes- sive periods. Food always preceded water. While food was being offered, records were made of (l) the order and frequency with which individuals ate, (2) the amount of time each individual spent near the rim of the food cup, and (3) aggressive interactions occurring within the three Sectors of the cage. Food was presented for 15 minutes. Water presentation followed the same format as food and involved similar measures being recorded. Water was also presented for 15 minutes. In Phases I and III dominance was measured during identical conditions. Phase II was designed to answer the third goal of this experiment, namely what are the effects of food and water deprivation on dominance behavior. During Phase II one experimental group was examined for food deprivation effects while the second experimental 25 group was examined for water deprivation effects. For the time—control group treatment conditions remained the same over all three phases, thus providing a measure of dominance stability over this period of time. In Phase II, two mid-ranking Ss (one male, one female) in the food-deprived experimental group were selected as test Ss following the end of Phase I. Ninety minutes prior to the regular evening session, between 5:30 p.m. and 6:00 p.m. (staggered at thirty minute intervals to accommodate each of the experimental groups) these two test Ss were removed from the observation cage and placed singly in small wooden cages where water but not food was available for 30 minutes. During this time the remaining 6 Ss in the observation cage were allowed access to both food and water. Observations revealed that these 6 birds were always satiated on food and water (i.e. stopped eating and drinking) before 30 minutes had passed. After 30 minutes the two test Ss were returned to the observation cage where all food and water had been removed. One hour later the regular evening food session was begun, lasting for 15 minutes, followed by the regular 15 minute water session. The two test Ss were 7% to 8% hours food- deprived at the start of the evening food session while the others in the group were one hour deprived. Similar procedures were followed for two midranking Ss (one male, one female) in the water—deprived group 26 except that water, rather than food, was deprived from these test Ss during Phase II while their flockmates back in the observation cage were given access to both food and water. Procedure for Study 2 Study 1 was essentially an exploratory attempt to define food and water deprivation parameters and to refine measures of dominance. As Table 1 shows four quail died during Study 1. In Study 2 quail were allowed longer access times to food and water. During the Morning sessions food and water were given between 9:00 a.m. and 9:45 a.m. They were removed 180 minutes later, between 12:00 noon and 12:45 p.m. During the evening sessions the regular food session lasted for 30 minutes rather than 15 minutes. All other experimental parameters were the same as in Study 1. 1 Only one quail died in Study 2 (see Table 1). In both Studies 1 and 2, however, the occurrence of dead quail raised questions about uncontrolled health-related variables Operating (e.g. pathology). Since dead quail were not replaced and since space allotment per bird is different in flocks of seven or less than it is in flocks of eight, there arose questions as to what effects these two variables had on dominance expression. As a result of these questions statistical comparisons in Studies 1 and 2 were difficult to interpret. 27 Procedure for Study 3 In Study 3 an additional 15 minutes of simultaneous food and water was given to all groups following the evening sessions. lNone of the quail died during Study 3 (Table 1), apparently as a result of this increased food and water supply. During Studies 1 and 2 it was felt that a small area around the incentive cups, extending approximately 8 to 10 inches into Sector B, yielded different forms of aggressive behavior from that seen in Sectors A and C alone. This effect was apparently being masked because Sector B, being as large as it was, encompassed both this suspected dominance form and dominance forms associated with areas further away. In an effort to delineate this apparent effect in Study 3 Sector B was reduced in size (see Measures 9: Incentive—Distance) for both incentives by moving the outer boundary toward the incentive to a point ten inches away from the incentive cups. Other than these changes in food and water access times and reduction in size of Sector B, the procedure for Study 3 was the same as for Study 2. RESULTS AND DISCUSSION Dominance Behavior as Appetitive Behavior Aggressive Behavior at Different Incentive-l~ Distances Data from Study 3 alone provide a representative picture of aggressive interactions in quail in the present experiment. The measures reported in Table 2 are mean frequency of aggressive supplants (i.e. dominance) per every possible dyadic interaction. In a flock of eight birds there are 56 possible interactions with each bird being capable of dominating every other flockmate in an intransitive manner. The means reported in Table 2 were obtained by first combining all food and water sessions as well as all phases in each of the three groups in Study 3. The frequencies of aggressive responses used in all dominance displays were then summed for all 168 interactions (i.e. 56 in each of the three groups) for each response type and these totals were divided by 168 for each Sector during the Evening sessions and for all sectors combined in the Morning session. The averages reported in Table 2 provide a standard basis for comparison between response types within each incentive-distance as well as between the various incentive-distances and the 28 29 mm.¢H mH.mH mo.mm ov.NOH HMHOB mo. vo. mo. «m.H MOHM Hm.n mm.OH w¢.¢m om.H HmmHCB mo. oH. mm.oH mm.mm Cmsm vw.m mH.p mn.mm Hm.mo mxoom COHmmmm U Cowomm m Houomm C Howowm mCHCHoE COHmmwm @CHCm>m .Conmmw mCHCHOZ mCH mCHHsp mHOHomm HHm m>mn mm meHQEoo Comb m>mn mm>HpCmoCH pCm mmwmsm .mmsoum HH< .m mpsvm CH CoHHOMHmHCH UHUmmp mHQHmmom mHm>m How mOCmHHCooo mmCommmH Ho SOCqumHH Cmmzll.m mHmde 30 nonincentive Morning session. For example, in Sector A for Evening sessions, quail used the Push response an average of 33.55 times whereas Threat responses were used an average of 1.30 times. Morning sectors were combined due to the small amounts of aggression occurring in each _sector alone. Comparisons within each incentive condition revealed differing response profiles at different distances from the incentive. In Sector A quail utilized the Peck and Push responses more than 97% of the time to supplant others from near the incentive cup. A Friedman xi test (Siegel, 1956) revealed significant differences (p<.001) between frequencies of different aggressive responses in this sector. Comparisons between specific response categories (Wilcoxin signed-ranks test) indicated Pecks to be signifi- cantly greater in frequency than each of the other response l . . .Nonparametric analyses were necess1tated because of heterogeneity of variance among several response classes. Although 10,881 aggressive responses were observed during Study 3, relatively few Kicks and Pushes were counted during this time. Unless otherwise indicated Kruskal- Wallis one-way analyses of variance were used to test for overall differences between independent samples. Where differences were found to occur beyond the .05 level of significance a Mann—Whitney U test for independent samples was used to detect differences between pairs of samples. Likewise the Friedman two-way analysis of variance and Wilcoxin signed-ranks tests were used to analyse differences among related samples (Siegel, 1956). 31 categories (p<.01) while the Push was used more frequently than the Threat and Kick (p<.01). Threats and Kicks were .observed at the same frequency in Sector A. In Sector B significant differences were again detected between aggressive responses (Friedman x: test, p<.001). Here, however, Pecks, Pushes and Threats accounted for 99% of the aggression. Subsequent comparisons of specific response categories (Wilcoxin signed-ranks test) revealed a different profile of behavior than that observed in Sector A. Threat responses were used as frequently as the Peck response in Sector B but whereas the Peck was significantly greater in frequency than the Push (p<.001), Threat responses were found not to differ in frequency from the Push. The Kick response was utilized less than each of the other aggressive responses in Sector B (p<.001). Quail again showed a different pattern of aggressive behavior in Sector C, the furthest distance from the incentive cups. Overall differences in response categories were again highly significant (Friedman Xi test, p<.001). Unlike Sectors A and B, the Peck and Threat responses in Sector C accounted for more than 98% of the aggression. Threats were used more often than Pecks (p<.003), Pushes and Kicks (p<.001), while Pecks in turn were observed more frequently than Pushes and Kicks (p<.001). Frequencies of Push and Kick behavior were found not to be different. 32 The profile of aggression observed throughout the cage during Morning sessions was most similar in appearance to the pattern observed in Sector C, with Pecks and Threats accounting for more than 99% of all aggression. Overall differences between response categories in the Morning 2 session were found to be highly significant (Friedman Xr test, p<.001). Threats were observed more frequently than each of the other response types (p<.001) while Pecks were used more frequently than Pushes and Kicks (p<.001). Pushes and Kicks were used equally often. Patterns of aggressive activity in Study 3 were also analyzed according to the sex combination involved (Male—supplant—Male, labeled as Male—Male; Male—Female, Female—Male, and Female-Female) in each Sector of the Evening sessions and in all Sectors combined for the Morning session. The results of these analyses and a discussion of these findings in relation to what other researchers have reported for the quail are presented in Appendix B. In general these data reflect the overall data presented in Table 2. Both male and female quail used different ag- gressive responses depending upon how far from the food and water cups they were standing. So great was this effect that a description of male-female interactions was meaning— less unless incentive—distance was taken into consideration. These data suggested that aggressive behavior was under the control of different stimuli in each Sector during 33 the Evening session and in the Morning session when neither food nor water was present. The Totals reported in Table 2 for each Sector and the Morning session indicate a trend of increasing amounts of aggressive behavior as incentive—distance decreased. This effect was tested by analysis of variance for each group separately and for each incentive separately. The results of these analyses are shown in Table 3. As Table 3 indicates, this trend was statistically significant for Food sessions but not for Water sessions. Although Sector A with Sector B (A,B) and Sector B with Sector C (B,C) comparisons revealed significant differences (p<.01, Newman-Keuls test, Winer, 1962) for only two of the three Groups during Food presentation (A,B, Groups 1 and 3; B,C, Groups 2 and 3) the probability of this occurring this many times by chance alone was .001 (Sakoda, Cohen and Beall, 1954). It was therefore concluded that frequencies in Sector A were greater than in Sector B and frequencies in Sector B were greater than in Sector C for Food sessions. Water sessions revealed a different distribution of aggression. There were as many aggressive interactions in Sector A as there were in Sector B, while Sector C proved to have the fewest. Morning sessions were analyzed somewhat differently. Since there were no incentives presented during the Morning session a Sector differentiation as used during the Evening 34 TABLE 3.——Results of analysis of variance for total amounts of aggression between Sectors in the Evening and Morning sessions of Study 3. Group Incentive F(df = 2,54) Sector Comparisons A,B B,C A,C 1 Food - 43.20** ** ** Water 7.09** * ** None 1.53 (Morning) 2 Food 15.64** ** ** Water 18.20** ** ** None 2.51 (Morning) 3 Food 21.07** ** ** ** Water 5.26** * ** None 1.30 (Morning) 7': p<.05. * * p<.01. 35 sessions would be inappropriate. Consequently a Left and Right sector was defined by the area corresponding to Sector B at both ends of the cage as if both food and water were presented at the same time. The remaining middle section was defined as the Middle sector. During Study 3, the Middle sector was made larger with the reduction in size of Sector B. Aggressive interactions rarely occurred within the areas corresponding to Sector A during incentive pre— sentation and they were not analyzed for the Morning session. Unless otherwise indicated all further analyses involving the Morning session will utilize this Left, Middle and Right classification. As Table 3 indicates no differences could be detected between these three Morning sectors for amount of aggressive behavior. Intragroup response matrices for Study 3 are pre— sented in Appendix D. These data reveal individual inter— actions among all Ss in each group during Study 3 and serve to indicate absolute levels of activity in each incentive condition during each phase. Unilateral Aggressive Interactions Measured at Different Incentive—Distances Members of dyads may either show unilateral or bilateral aggression (or, of course, not respond at all) during any particular session. Avian societies charac— terized by unilateral aggressive interactions have been referred to as peck orders while societies characterized 36 by bilateral aggression have been referred to as peck-dominance organizations. There was evidence in the present experiment suggesting that the form of social organization expressed by a flock of quail is a function of how close Ss are to the incentive when aggreSSive behavior is measured. Table 4 presents the proportion of unilateral inter— actions within each Sector of the Evening session and within the Left, Middle and Right sectors of the Morning session. In most cases interactions become more bilateral as the quail approach the incentive. This is shown in Table 4 by the smaller proportion of unilateral interactions in Sector A.) The Friedman Xi test (Siegel, 1956) was used to evaluate overall differences in these proportions across the three Sectors. This nonparametric analysis was chosen because a predonderance of "zero" interactions occurred in Sector C, causing heterogeneity of variance and hence violating a basic assumption of the parametric analysis of variance model. I Results of analyses presented in Table 5 can best be understood by recalling the changes in Sector size that were made in Study 3. It was felt that by making Sector B smaller in size this would serve to delineate dominance behavior occurring in a small area around the incentive cUp. It was suggested that dominance interactions in Sector B were different from those in Sectors A and B. Analyses of Study 3 subsequently revealed a different oo.H mm. oo.H m oo.H oo.H oo.H N mm. mm. mm. H m oo.H oo.H mm. m oo.H oo.H oo.H m oo.H oo.H no. H m vm. oo.H vm. m mm. mm. mm. m mm. oo.H mm. H H nanm chsz puma macho swsum mCOHmmmm mCHCHoz mm. on. Hm. mm. Hm. mv. m mm. Hm. mm. wa. mm. ww. m 7 mm. mm. mm. mm. om. mv. H m 3 mm. mm. mm. mm. mm. me. m mm. mm. mm. mm. vm. we. m mm. mm. mm. mm. mm. mm. H m mm. mm. mm. mm. om. mm. m v0. Hm. mm. mm. vm. mm. m mm. mm. mm. om. mm. mm. H H mConmom mCHC®>m o uomm m uomm 4 Doom 0 Doom m uomm 4 pomm QCOHo hospm COHmmmm Hmpmz Conmmm poom .pm>Hmeo mums mCoHpomnmpCH m>HmmmHmmm HMHmHMHHCC CUHCB CH pmwo Hog mCOHmmmm Ho COHHHOQOHQ mez||.v mHm m> m> m> m> m> Hmme Slm SIS SIS SIS SIS SIS mflHHMZIHMMmDHM COHHHUCOU “mCOmHHmmfioo m>HquoCH HHMHm>o umme D meCHSBICsz .mCOmHmeEoo CmHHmm .AmHmEmm H mm smfimz UmCHQEoo mConmmm CmHMB CCm COOM.How CoHHHCCoo m>HHCmoCH Comm CHCHHB mCoHHMCHQEoo xmm CmmemQ Amm mHnme mmHoCvamuw mmCommmH m>memHmmm mo mmmmHMCm mo mHHsmmmlu.vm mHmSB CH CBOan 90 different parts of the cage (see Table 2). These effects are apparent in Table B3 as well. For Male—Male inter— actions a Friedman Xi test revealed significant differences (p < .001) existing between response frequencies in every incentive condition. Subsequent paired-comparisons (Wilcoxin signed—ranks test) for all sex combinations in all incentive conditions produced the results shown in Table B5. In Sector A males tended to supplant other males more fre- quently with a Peck response than with any other type of aggressive response (p < .01). In Sector B Pecks and Threats were used equally often and both were more frequently given (p < .01) than were Pushes. In the outer Sector C Threats were more fre- quent than all other responses (P < .01), with Pecks next in order (p < .01). The Morning session was most similar to Sector B in that frequencies of Pecks and Threats could not be differentiated though both were used more often than all other aggressive responses (p < .01). Male-Female Behavior Male-supplant-Female behavior ranked second to Male-Male behavior in total frequency of occurrence (see Tables B3 and B4), indicating the extensiveness of male- initiated aggressive activity in the Japanese quail. Male- Female aggression proved to occur more frequently than 91 “GMOHMHGmHm pom n ma Ho. v SSS mo. v QS S ms S SS m: SS mHmEmmlmHmEmm ms ms mg mg ma ms mamzlmamfimm SS mc SS SS m: SS meEmmImHmS SS ms SS SS ma SS mamzlmamz mafiauoz ¥¥ mg «LS ¥¥ mg 91% GHQEQMIGHQEQHH SS SS SS SS SS SS mHmSImHmEmm SS mc SS SS ms SS wamfimMImHmz SS ma SS SS S SS meSlmfimS U HOpowm SS SS mm SS ms S mamfimmlmamfimm SS SS SS SS SS SS mHmSImHmE®m SS SS SS SS ma SS GHMEmMIwHwS SS SS SS SS ma S mHmSIGHmS m HO#O®m SS SS SS SS SS ms mamEmmImHmEmm mu SS SS SS SS ma mHmSImHmEmm SS SS SS SS SS SS mamfimmImeS mm SS SS SS SS SS mHmSIQHmS d HO¢U®m MOHM MUHM #mmnnB MOHM #mmnflfi szm . . . COH @QH O C .H H m> m> mxy m> w> m> .u. X095 U QMWHVCWMMU pmmuse gmsm swam xomm Somm Somm m .u H cflnufls Amm mfinme cfl caogmc .coflpmcfinaoo xmm 30mm mmfloswsqmum mmcommmu mo mommamcm mo mpHSmmmll.mm mqmde 92 all Female-initiated aggression in all three Evening Sectors as well as during the Morning (p < .02). Response profiles for Male—Female behavior reflected those found for Male—Male interactions. Near the incentive cup (Sector A), the Peck was the most observed response (p < .01). As the male quail moved further from this area, to Sectors B and C, he gave both the Peck and Threat response to females with equal fre- quency to the exclusion of other behavior (p < .01). Morning sessions yielded this same heavy reliance on Pecks and Threats. Female—Male Behavior In Sector A females tended to supplant males more frequently on the average than they did other females (p < .002). This tendency reversed in the other two Sectors (B and C, p < .02), while in the morning no dif- ferences between male and female targets could be detected. Response profiles shown by females on males were of considerable interest as they reflected marked dif— ferences from Male-initiated aggression. Whereas males supplanted females more frequently with Pecks in Sector A, females used Pecks and Pushes equally often to force males out of the incentive area. In Sector B the Push response was most frequently used (p < .01). No significant dif— ferences were found during the Morning session. -93- Female-Female Behavior In Sector A Female-supplant—Female behavior was the least observed of the four sex combinations (p <.002). Again Female-Female response profiles are of particular interest for they reflect changes in stimulus control. Females used Pecks and Pushes equally often near the incentive cup (Sector A) followed in order of frequency by Kicks (p <.Ol), then Threats (p <.Ol). In Sector B, Peeks, Pushes and Threats were used with equal frequency while in the outer Sector C and in the Morning session Peeks and Threats were indistinguishable in frequency. Discussion of General Characteristics of Aggressive Behavior The aggressive behavior of Japanese Quail has been characterized by Farris (1964), Selinger and Bermant (1967), and Eynon (1968). Eynon's report gives the clearest picture of aggression in quail. He recorded responses emitted by pairs of quail placed together in neutral cages. The dominant bird was characterized by such behaviors as "attack, biting, bristle display, 'churring,‘ feeding, 'keking,‘ ' whereas the subordinate more pecking, preening and strutting,’ frequently showed such responses as ”avoidance run, backward shuffle, crawling under the other, crouch, crowing, distress call and peering" (p. 67). 94 Aggression recorded in the flock situation of this experiment was considerably different from that described by Eynon for paired-contests. This is not particularly surprizing since different stimulus conditions were Operating in the two conditions. A bird being attacked in a flock situation must respond not only to the "attacker" but also so as to avoid other potential attackers in the flock. Thus a full expression of some response categories might be prevented in the flock situation. This, in fact, was found to be the case. The dominant quail in the present experiment (i.e., the aggressor in a particular dyadic interaction) corres- ponded to Eynon's in terms of "attack, biting (or head, shoulder grabs) and pecking" behavior. "Bristle displays, churring, keking and strutting" have also been observed by this researcher in paired contests but only rarely in the larger flock situation. If "bristle displays" occurred in the flock condition they were too rapid for precise identification. Eynon does not mention whether "feeding and preening" were performed before or after aggressive contact. "Churrs" consisted of low—frequency component tones which were difficult to localize in the flock condi- tion. "Keks" had a number of behavioral accompaniments as well as higher frequency tones which made it more easily localized. A "keking" quail was usually in either an aggressive crouch ready to attack or actually 95 attacking other quail in the cage. There were also move— ments of the throat which accompanied the "kek." Except .in sexual encounters, vocalizations are not reported in this paper because of the high probability of error in identifying the vocalizer. Nevertheless, "churrs" were rare during this experiment and their few occurrences were as components of the Threat sequence. "Keking" vocalization often characterized quail which appeared to go "crazy" during a session, running around the cage Pecking and Threatening essentially every flockmate. The stimulus for this sudden Pecking and Threatening behavior was never understood. Its effect was to cause many of the attacked birds to "pop" over and over again into the ceiling of the cage, resulting in numerous wounds. This frantic Pecking and Threatening behavior often ended in exhaustion. Quail aggressive behavior in this experiment was primarily a male-initiated response. This effect was accentuated when the quail were further from the incentive, with males providing 81% and 78% of the aggression in Sectors B and C, respectively. It was interesting to find that males in Sector B primarily used the Peck and Threat to supplant other females while females supplanted males primarily with the Push response. Some of the most ferocious aggression occurred here in Sector B and this tendency on the part 96 of females not to Peck other males was considered to be an avoidance of aggressive interaction. The relative effects of each type of aggressive response was not measured but it appeared that a Peck signified greater expression of dominance than a Push, causing greater degrees of avoidance behavior. 9 Again in Sector C males used both Pecks and Threats to supplant females while females either were unable to utilize the Threat or found it unrewarding to threaten other males. The effectiveness of a Threat lies most likely in the individual's conditioned aversive properties acquired during successful dominance behavior. Since it will later be shown that females ranked lower, on the average, than males (see Appendix C) it follows that their success as threatening" stimuli would be lower. APPENDIX C DOMINANCE CORRELATED VARIABLES 97 I DOMINANCE CORRELATED VARIABLES Relationship Between DOminance Rank and Sex Table Cl reports mean social ranks for all males combined and all females combined in Study 3. Results of the Mann-Whitney U test (Siegel, 1956) for differences between sexes showed that males were significantly higher in rank (lower numerically) in every case but one (Sector B during Water sessions). Similar analyses were not made on Study 2 because of unequal group sizes. TABLE Cl.--Mean dominance rank for males and females in Study 3. Means were computed from data collated over all phases and, for the morning session, overall sectors as well. Results of Mann-Whitney U test for differences between sexes in each incentive condition are also shown. Food Sectors Water Sectors Sect A Sect B Sect C Sect A Sect B Sect C Morning Males 3.84* 3.46* 3.46* 3.86** 3.60 3.18 3.42** Females 5.16 5.54 5.54 5.14 5.40 5.82 5.58 * p < .029 ** p <1 .014 98 99 The percentage of sessions in which males assumed the alpha position in Study 3 is shown in Table C2. Sexual differences were detected by using the binomial test (Siegel, 1956). Males clearly held the number one rank more frequently than did females. However, no sexual differences were ever found in Sector A during water presentation or in any Water Sector for Group 2. TABLE C2.--Percentage of sessions over all phases combined in Study 3 in which males were alpha. Results of binomial test of significance are also shown. Food Sectors Water Sectors Group A B C A B C Morning l .37 .88*** .77* .55 .75* .80** .88*** 2 .83** .94*** .72* .65 .67 .64 .73 3 .83** .88***l.00*** .68 .8l** .94*** .78** * p < .05 ** p < .01 *** p < .OOl Relationship Between Dominance Rank and Weight Table C3 reports correlations between dominance index and weight for Studies 2 and 3. Although a tendency towards negative relationships was apparent, as would be expected since males were higher in rank and weighed less, 100 TABLE C3.--Correlations between Dominance Index and weight in each incentive condition for both Study 2 and Study 3. Food sessions Water sessions Sect' Sect Sect Sect Sect Sect Group Phase _ A B C A B C Morning Study 2 1 I .221 .787* .082 .514 .659 .602 .770* II .814* .123 -.350 -.351 —.374 - -.393 II .668 —.069 -.533 -.293 -.494 —.448 - 2 I .133 -.174 - -.251 -.304 - -.303 II -.309 —.O93 - -.l66 -.O43 -.142 -.103 III .577 - — .511 -.O63 .198 -.000 3 I .671 .238 - .054 .063 - .640 II .483 .145 .030 .210 .094 .045 .132 III .116 -.l89 .071 .354 .090 -.278 .086 Column Mean .426 .130 -.154 .075 -.035 .005 .126 Study 3 l I -.325 -.709*—.276 —.357 -.544 -.472 —.563 II .474 -.357 -.l71 —.263 -.247 -.556 -.086 III .397 -.O85 .038 .350 -.269 -.303 .029 2 I .226 .223 —.151 —.609 -.111 .217 .017 II, -.O48 .222 .000 -.502 .271 — .117 III .622 .485 -.416 .306 .343 -.356 .210 3 I —.118 -.015 -.227 .040 -.O62 - —.163 II .152 -.270 -.353 -.667 -.498 —.424 — III .377 -.367 - -.475 -.263 -.603 -.441 Column Mean .213 é.115 -.l99 -.27l -.l65 —.374 _.123 *p_.05. .707 and .834 needed for the first correlation to be significant at the .05 and .01 levels, respectively. lOl essentially all correlations were found to be nonsignificant. Overall mean correlations ranged from -.357 to .375 indi— cating weight to be a poor predictor of dominance status in all Sectors of all incentive conditions. 'Relationship Between Dominance Rank and Sexual Behavior There were 103 instances of heterosexual behavior in Study 3, only one instance in Study 2 and none in Study 1. 'The relationship between sexual behavior and dominance status was analyzed by first ranking males and females within their own sex groups (n=4) according to dominance status and then correlating ranks of each male with the rank of the female partner involved in the sexual relationship. Correlations of -.OO7, .344 and .106 were obtained for Groups 1, 2 and 3, respectively, during all sessions combined of Study 3. The relationship between social rank and frequency of sexual behavior was determined by first ranking males and females within their own sex group (from one to four) according to their mean DI over all phases combined in Study 3. These dominance rank orders were then correlated with rank orderings of males and females based upon fre- quency of sexual behavior. Correlations obtained for males and females were -.986 and -.788, respectively. Only the males' correlation was significant (p < .02). 102 Sexual behavior in the male quail is characterized by a sequence of measurable components (Otis, 1968). The colorful courtship phase consists of Neckéand-Body—Tonus, Toe-Walk, and Churring (a vocalization) activity followed in order by Body-Orient, Neck-Grab, Mount and Tread. There may be chase components in the sexual sequence but these were not counted here because of their possible role in other behavior sequences (e.g., aggression). The question arose as to whether more dominant males were exhibiting more sexual-response components than others. In order to answer this question the sexual-response components were given values from 1 to 7 according to their sequential position in the chain, with Neck-and-Body—Tonus given the value of one. Dominance rank was then cor- related with the observed sexual components during a particular sexual relationship. Correlations of .328, -.353 and .307 were obtained for Groups 1, 2 and 3 of Study 3. It was of secondary interest that the 13 instances of courtship behavior never went beyond the Body Orient stage. That is, Mounting and Tread components of the sexual sequence were never preceded by these "courtship" responses. The role of the Neck-and-Body—Tonus and Toe Walk responses as sexual components should perhaps be questioned. 103 Discussion of Dominance—Correlated Variables Male quail were clearly the more dominant sex - during this experiment (Table C1). Male dominance is a well documented phenomenon in other gallinaceous species, particularly in the domestic fowl (Domm, 1939;Guhl, 1950; Masure and Allee, 1934a,b). Some avian researchers have felt this male dominance effect to be so phylogenetically widespread as to warrant the concept of "sexual dominance” apart from social dominance (Armstrong, 1965; Guhl, 1950; Noble, 1939; wynne-Edwards, 1962). Since many species seem to require a dominating male for successful copula— tion with females, this widespread male dominance phenomenon is viewed as a highly adaptive feature (Armstrong, 1965). Among captive Japanese quail the female is capable of assuming the alpha position (Table C2), though never to the same extent as the male quail. There was no evidence in this experiment that male and female quail had separate social orders as has been found in turkeys (Hale, Schleidt and Schein, 1969) and chickens (Guhl, 1953; Guhl and Fisher, 1969). Since male quail held the more dominant ranks in the flock it is not surprising that body weight was negatively correlated with dominance rank albeit non- significantly so. 'Male quail ranged in average weight 104 from 88 to 108 grams while females ranged from about 100 to 117 grams (Table A1, Appendix A). ' The relationship between weight and social status is not at all clear in the domestic fowl. Both a lack of correlation (Collias, 1943; Potter, 1949) and positive correlations (Craig, Ortman and Guhl, 1965) have been reported. Fennell found dominant game cocks to be the largest birds while Shoemaker (1939) found no relationship between dominance and weight in canaries. There was no evidence that a dominating male quail was an essential factor for copulation. Female quail showed no preference for particularly ranked males. High ranking males showed the most sexual behavior while an apparent trend towards this relationship proved insignifi- cant in females. Among Gambel's quail it is known that the female will form an attachment to the more aggressive males (Gorsuch, 1934). Similar findings have been reported in other avian species (Heinroth, 1911, for ruddy shield- ducks; Murchison, l935a,b for domestic fowl) although it is also known that a highly aggressive male may deter the female (see review in Armstrong, 1965). There was some suggestion of this latter finding in Study 1 where extent of aggression appeared to be higher throughout the day. There were no sexual responses observed during observation session of Study 1. 105 In domestic fowl it is known that dominant males will fight subordinates who are attempting to mate with hens. Guhl, Collias and Allee (1945) found the resulting suppression of sexual behavior to persist even when the dominants were removed from the group. They labeled this phenomenon "psychological castration." On many occasions subordinate quail were attacked by superiors while mating but these low ranking males were rarely, if ever, supplanted from their sexual activity. Though records of this event were never kept, it was apparent that these low ranking males (as based on aggression during food- or water—getting behavior) were "unsupplantable" while per— forming the consummatory sequence of sexual behavior. It should be recalled that measures of sexual behavior during Study 3 were relatively few (103) due, perhaps, to the imposed food and water deprivation schedules. It was most interesting to find that courtship responses were never observed in sequence with copulatory behavior. This may reflect an artefact of the Study itself. An investigation of food deprivation effects on somponents of the male quail's sexual behavior would be of considerable interest. APPENDIX D INTRAGROUP RESPONSE MATRICES FOR STUDY 3 106 INTRAGROUP RESPONSE MATRICES FOR STUDY 3 Tables D1 through D21 show frequency of dyadic interactions in all groups during Study 3. Subjects are identified by sex (M means male, F means female) and place- ment of leg bands (MLL means two bands on left foot of male; MRR means two bands on right foot of male; M means a male with no leg bands; MLR means one band on each leg of a male). Subjects listed in the left column of each table represent aggressors and dominance interactions (i.e., where supplants occur by the aggressor) with each of the other seven flock members are found by reading across the row for each subject. All types of aggressive responses have been combined in these tables. The mean DI for each subject is shown in the column to the right of each table. This mean DI was computed as an average of the DIs measured over the six days of each phase. It was ngt_computed directly from the accumulated response frequencies shown in Tables Dl through D21. 107 108 TABLE Dl.--Intragroup response matrices for Group 1 showing aggressive supplants (by Ss in left column) during Evening Food sessions in Sector A—for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) . denote those gs Water-deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M* - 23 8 9 3 11 3 7 .31 MLL 23 — 24 92 4 35 15 12 .41 MRR 25 68 - 61 3 35 11 14 .62 MLR 14 83 34 - 3 31 7 12 .48 F 6 10 4 12 - 3 4 .28 FLL 10 39 7 26 — 2 3 .30 FRR* 1 9 6 3 1 5 — 1 .27 FLR 3 19 17 10 2 6 3 - .40 (b) Phase II: M* - 2 2 1 3 .10 MLL 2 - 1 19 14 15 3 20 .40 MRR 1 - 6 2 5 2 4 .20 MLR 6 11 13 - 8 6 1 50 .34 F 5 5 7 - 7 7 .25 FLL 3 15 4 15 12 - 3 9 .34 FRR* 1 1 2 1 — 2 .13 FLR 11 14 9 30 13 11 1 — .47 (c) Phase III: M* - 12 6 11 1 4 5 .27 MLL 12 — 14 28 34 16 5 28 .43 MRR 5 7 - 10 3 5 4 .27 MLR 19 33 23 - 7 8 32 .42 F 10 31 3 9 — 4 1 7 .30 FLL 12 17 7 12 12 - 3 15 .52 FRR* 2 3 1 - 2 .13 FLR 11 26 10 17 9 6 4 - 42 109 TABLE D2.--Intragroup response matrices for Group 1 showing aggressive supplants (by §$ in left column) during Evening Food sessions in Sector B for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those §s Water-deprived during Phase II. M MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M* - 3 1 1 1 .10 MLL 4 — 9 1 1 4 3 .24 MRR 17 13 — 25 3 5 2 11 .61 MLR 2 7 5 - 4 2 10 .39 F 1 1 1 1 - 1 3 .15 FLL 1 6 3 3 - 2 .27 ' FRR* 1 3 - .07 FLR 2 1 1 — .05 (b) Phase II: M* — 1 2 .07 MLL — 1 1 3 2 .15 MRR' 2 — 1 4 3 10 13 .34 MLR 4 4 85 — 5 2 3 3 .50 F 1 1 1 - 1 1 .09 FLL 2 5 1 2 — 1 .17 FRR* - .02 FLR 2 3 2 4 3 - .20 (c) Phase III: M* - 2 1 2 1 .11 MLL 2 — 1 6 3 2 1 1 .27 MRR 1 - 2 2 4 1 .20 MLR 2 53 - 6 2 3 .47 F 1 - 1 .05 FLL 2 2 1 — 1 3 .15 FRR* - .00 FLR 1 2 l l 6 2 2 - .22 110 TABLE D3.--Intragroup response matrices for Group 1 showing aggressive supplants (by Ss in left column) during Evening Food sessions in Sector C—for PhaSes I(a), II(b) and III (c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those SS Water-deprived during Phase II. M ' MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M* — 1 1 6 .07 MLL - .00 MRR 7 1 - 5 1 1 6 .28 MLR — 1 1 .05 F — 2 .05 FLL - 1 .02 FRR* 1 - .02 FLR 1 1 - .05 (b) Phase II: M* — 2 1 2 .08 MLL 3 — 2 .08 MRR 2 11 - 4 2 11 .28 MLR 3 1 43 — 1 .22 F 1 - 1 1 .07 FLL 9 1 — .11 FRR* 1 — .01 FLR 1 2 3 6 — .21 (c) Phase III: M* - 2 1 .07 MLL — 3 4 1 1 2 .15 MRR 2 - 1 2 1 .14 MLR 1 5 11 - 1 1 .24 F 2 4 2 1 - 6 1 .16 FLL 5 - 2 .06 FRR* l 2 - .04 FLR 1 '2 1 1 1 2 - 16 111 TABLE D4.--Intragroup response matrices for Group 1 showing aggressive supplants (by SS in left column) during Evening Water sessions in Sector A for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily 015 measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those gs Water-deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I:' M* - 1 2 3 4 _ 3 2 .23 MLL 1 - 2 2 2 .15 MRR 5 - 5 10 6 10 .30 MLR 5 11 - 5 5 7 .27 F — 1 2 .07 FLL 8 1 6 1 3 - 6 4 .27 FRR* 2 7 4 - 4 .25 FLR 4 11 8 3 1 - .23 (b) Phase II: M* - 1 6 1 3 8 3 .21 MLL — 1 . 1 .05 MRR 4 - 1 4 2 6 7 .32 MLR 2 3 - 1 2 4 9 .31 F 3 1 1 - 1 5 7 .26 FLL 5 2 4 — 4 2 .21 FRR* 5 4 1 1 1 — 5 .23 FLR '6 ' 5 2 4 - .16 (c) Phase III: M* — 4 4 1 2 4 .19 MLL - 1 .02 MRR 2 1 - 2 1 2 2 7 .18 MLR 3 5 — 2 1 20 .25 F 3 3 4 — 3 4 3 .30 FLL 1 1 - 1 7' .17 FRR* 1 1 2 - .07 2 8 2 2 2 — .26 FLR 3 l 112 TABLE D5.--Intragroup response matrices for Group 1 showing aggressive supplants (by SS in left column) during Evening - Water sessions in Sector B for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those SS Water-deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR Mgin (a) Phase I: M* - 2 1 3 2 .16 MLL — 1 1 2 .10 MRR 16 0 25 6 19 7 .63 MLR 1 2 — 2 2 .18 F 2 1 - 2 .07 FLL 2 1 4 - 2 3 29 FRR* 2 2 2 — .06 FLR 2 2 1 2 - .11 (b) Phase II: M* - 3 1 2 1 .13 MLL .. - 1 .02 MRR 4 1 - 1 11 10 1 9 .37 MLR 2 19 - 1 1 .23 F 4 1 4 - 2 1 2 .20 FLL 2 1 1 - 1 2 .14 FRR* 3 2 3 1 - 2 .18 FLR 2 1 4 3 2 2 — .18 (c) Phase III: M* - 2 2 3 1 2 .14 MLL 1 - 2 .07 MRR 1 1 - 3 6 1 3 .22 MLR 2 1 28 - 2 .25 F 3 3 3 1 — 3 1 .18 FLL 2 - 3 .10 FRR* 1 1 — 1 .06 FLR 1 5 1 2 — .14 113 TABLE D6.--Intragroup response matrices for Group 1 showing aggressive supplants (by SS in left column) during Evening Water sessions in Sector C for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those SS Water-deprived during Phase II. M ‘MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M* .. l l 1 .07 MLL - .00 MRR 17 6 - 3 3 1 2 .38 MLR 2 - 1 .07 F 1 - .02 FLL 1 1 - .05 FRR* 1 - .02 FLR — .00 (b) Phase II: M* — 1 1 1 .07 MLL - .00 MRR 1' 5 - 3 18 .27 .MLR 1 22 — 1 1 .24 F - 1 .02 FLL 1 — 1 .05 FRR* 1 - .02 FLR 1 3 1 — .06 (c) Phase III: M* — 3 2 .07 .MLL 2 - 5 1 1 2 .14 MRR 1 1 - 1 2 1 .14 MLR 6 10 - 1 .16 F 1 1 1 - 1 .10 FLL 1 4 - .07 FRR* - .00 FLR - 2 1 2 - .08 114 TABLE D7.--Intragroup response matrices for Group 1 showing 'aggressive supplants (by Ss in left column) during Morning sessions (all sectors combined) for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those gs Water-deprived during Phase II in the ‘ EVening sessions. M MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M* - 1 2 1 1 .08 MLL — 1 1 1 .07 MRR 33 7 — 5 14 6 6 9 .72 MLR 4 19 3 — 3 5 9 .55 F - 1 1 .05 FLL — 2 .07 FRR* - 1 1 1 — .07 FLR — .00 (b) Phase II: M* - 2 1 .05 MLL 2 . - 2 2 .08 MRR 7 9 - 2 11 6 17 .07 MLR 13 7 141 — 1 1 3 16 ~.53 F 2 — 2 .09 FLL 2 - .07 FRR* 1 2 — 1 .04 FLR 1 1 2 2 2 4 - 20 (c) Phase III: M* - 1 7 .05 MLL 4 - 1 4 6 2 1 .24 MRR 1 3 — 1 2 .22 MLR 12 14 22 - 1 4 .44 F 1 - .01 FLL 2 2 1 - .12 FRR* 1 1 - .05 FLR 2 4 3 3 1 1 - .29 115 TABLE D8.——Intragroup response matrices for Group 2 showing aggressive supplants (by §S in left column) during Evening Food sessions in Sector A for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phvse in this incentive condition, is shown as well. Asterisks (*) denote those SS Food-deprived during Phase II. M ‘MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M — 24 11 18 25 12 31 27 .57 MLL* 17 - 2 1 1 6 .18 MRR 33 5 — 3 5 2 15 7 .41 MLR 6 4 4 — 4 .21 F 10 2 4 — 4 5 .23 FLL* 16 2 2 — .18 FRR 37 2 14 5 3 — 7 .41 FLR 10 3 4 5 1 5 - .21 (b) Phase II: M’ — - 31 41 9 17 24 22 24 .55 MLL* 22 — 12 7 6 3 2 .35 MRR 40 19 — 6 7 8 13 19 .54 MLR 8 4 8 - 1 12 3 5 .37 F 7 1 4 1 - 3 1 .16 FLL* 19 . 4 14 8 2 — 2 4 .35 FRR 16 6 7 2 7 — 6 .36 FLR 13 3 4 7 2 2 5 — .33 (c) Phase III: M — 10 42 7 1 5 14 14 .38 MLL* 16 - 9 3 7 1 6 2 .26 "MRR 45 20 - 8 7 6 7 28 .49 MLR 3 5 11 - 1 2 2 2 .31 F 3 2 3 — 3 .11 FLL* 8 1 6 1 — 1 3 .27 FRR 5 5 7 5 2 2 - 3 .26 FLR 3 1 10 3 2 2 — .16 116 TABLE D9.--Intragroup response matrices for Group 2 showing aggressive supplants (by Ss in left column) during Evening Food sessions in Sector B_for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, c0mputed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those SS Food-deprived during Phase II. M ‘MLL MRR MLR F FLL FRR FLR Mgin (a) Phase I: M - 38 49 52 32 28 22 28 .86 MLL* - 1 2 2 1 .10 MRR 3 11 — 9 4 6 3 .42 MLR 35 9 15 — 16 12 4 3o .38 F 2 3 3 - 13 2 17 .22 FLL* - 1 .02 FRR 1 2 3 - 1 .09 FLR 2 2 8 1 - .13 (b) Phase II: M - 41 29 3 5 13 2 .48 MLL* 1 - 2 .03 MRR 1 22 — 2 1 4 3 .29 MLR 32 11 3 — 3 2 7 .44 F 1 — 1 1 .07 FLL* 1 — , .02 FRR I 1 1 4 2 - .12 FLR 1 1 1 4 2 — .12. (c) Phase III: M - 8 13 1 3 5 4 .31 MLL* - 1 1 1 1 .08 MRR 14 41 - 10 8 15 5 14 47 MLR 5 3 — 1 .12 F 1 1 1 - 3 2 10 .12 FLL* 1 — .02 FRR 1 1 — .05 FLR 2 4 6 1 21 25 4 — 36 117 TABLE DlO.--Intragroup response matrices for Group 2 showing aggressive supplants (by SS in left column) during Evening Food sessions in Sector C for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the Six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those gs Food-deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 2 5 2 2 1 3 .27 MLL* — .00 MRR 3 - 3 1 l .09 MLR 2 2 4 - 4 7 2 6 .33 F _ l .02 FLL* - .00 FRR - .OO FLR 1 1 5 3 - .12 (b) Phase II: M - l 9 1 .12 MLL* - 1 . 02 MRR f 2 — 3 1 .12 MLR 4 3 — 2 l .17 F — .00 FLL* - . 00 FRR - .00 FLR 6 l 2 - .10 .(c) Phase III: M - l 1 1 .07 MLL* l - 1 .05 MRR 1 - .02 MLR - 1 .02 F - 4 .05 FLL* l - .02 FRR 1 2 - .07 FLR 1 1 26 4 7 - .24 118 TABLE Dll.--Intragroup response matrices for Group 2 showing aggressive supplants (by SS in left column) during Even Water sessions in Sector A for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those §s Food-deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 5 2 1 1 1 8 .22 MLL* 10 - 6 4 1 4 3 6 .35 MRR 3 5 — 4 2 8 .23 MLR 1 4 3 — 1 1 8 .20 F 3 2 1 — 1 2 2 .19 FLL* 2 4 2 - 3 2 .21 FRR 3 1 2 2 - 3 .17 FLR 5 1 5 9 4 2 2 — .29 (b) Phase II: M - 1 3 5 2 .12 MLL* 3 - 4 1 4 2 2 .27 MRR 6 1 — 1 3 1 8 .22 MLR 1 1 - 2 3 2 .18 F 1 1 - 1 .06 FLL* 6 2 7 3 1 - 2 .24 FRR 2 3 1 — .11 FLR 1 8 1 2 — .16 (c) Phase III: M - 2 3 2 1 2 .11 MLL* 1 — 2 2 .06 MRR 9 3 - 4 1 1 5 25 MLR 3 1 - 2 3 .12 F 1 - .07 FLL* 2 2 3 1 1 — .19 FRR 3 1 1 — .11 FLR 7 1 2 1 .13 119 TABLE D12.--Intragroup response matrices for Group 2 showing aggressive supplants (by SS in left column) during Evening Water sessions in Sector B for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those gs Food—deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 3 4 1 1 1 2 .20 MLL* 5 - I 1 2 .09 MRR 5 6 - 6 4 1 5 .26 MLR 4 10 3 - 5 7 .34 F - .05 FLL* 1 2 - .05 FRR 1 1 - .04 FLR 1 1 5 1 1 — .14 (b) Phase II: M — 7 5 5 3 1 2 .32 MLL* _ 3 - 3 2 1 3 .15 MRR .5 13 - 2 1 3 5 .26 MLR 8 7 3 — 2 6 .34 F 3 2 1 1 - 3 .16 FLL* 2 2 10 1 — 1 2 .26 FRR 1 3 1 - .10 FLR 4 1 3 3 3 1 1 - 20 (c) Phase III: M - 2 10 5 2 1 5 7 .32 MLL* 4 — 2 1 1 1 .11 MRR 15 14 - 4 2 3 2 .32 MLR 3 2 5 - 1 3 1 2 .22 F 2 1 - 2 1 5 .18 FLL* 2 3 1 - 3 .16 FRR ' 3 2 2 - .12 FLR . 7 2 3 9 1 2 5 - .41 120 TABLE Dl3.--Intragroup response matrices for Group 2 showing aggressive supplants (by Se in left column) during Evening Water sessions in Sector C for Phases I(a), II(b) and II(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. Asterisks (*) denote those SS Food-deprived during Phase II. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M — 1 5 1 1 2 .15 MLL* 1 - 2 .07 MRR 1 9 — 2 5 1 1 3 .25 MLR 1 1 '— .04 F - 1 .02 FLL* — .00 FRR 1 - .02 FLR 2 — .02 (b) Phase II: M - 2 .05 MLL* — 2 .02 MRR 1 5 - 2 1 1 .12 MLR — .00 F 1 1 — 2 .09 FLL* — .00 FRR - .00 FLR — .00 (c) Phase III: M - 1 .02 MLL* 7 — 1 3 .07 MRR 4 1 — .09 'MLR 1 - 1 .07 F - 1 .02 FLL* - .00 FRR - .00 FLR 1 5 1 1 11 12 5 - .29 121 TABLE Dl4.--Intragroup response matrices for Group 2 showing aggressive supplants (by §$ in left Morning sessions (all sectors combined) for II(b) and III(c) of Study 3. Each S's mean as an average of the six daily DIs measured phase in this incentive condition, is shown column) during Phases I(a), DI, computed during each as well. Asterisks (*) denote those gs Food-deprived during ‘ Phase II during the Evening sessions. ‘7 M MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 4 1 2 .10 MLL* 1 — 2 .05 MRR 1 1 - 1 1 .05 MLR 4 3 - 8 6 2 .35 F - 1 .05 FLL* - l .02 FRR ' 1 - .02 FLR - .00 (b) Phase II: M - 1 .02 MLL* , — 1 .02 MRR 2 — 2 .10 MLR - 1 .02 F — .00 FLL* — .00 FRR 1 - .02 FLR 1 - .02 (c) Phase III: M — 1 1 .05 MLL* 1 — .02 MRR 3 - 1 .07 MLR - .00 F — 1 .02 FLL* - .00 FRR 2 - .02 FLR 2 4 3 - .08 122 TABLE D15.——Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by SS in left column) during Evening Food sessions in Sector A for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 7 9 2 14 6 13 9 .36 MLL 11 — 7 2 12 11 8 24 .54 MRR 9 3 — 15 13 16 41 29 .44 MLR 2 5 — 1 1 1 .11 F 13 6 11 1 - 2 4 15 .34 FLL 2 1 7 - 2 .11 FRR 9 9 16 2 15 2 — 21 36 FLR 13 16 12 1 23 5 15 — 44 (b) Phase II: M - 3 11 23 17 9 17 12 .46 MLL 11' - 6 12 12 8 11 8 .62 MRR 7 1 - 14 6 5 8 7 .40 MLR 7 5 8 — 1 5 7 1 .22 F 13 7 7 9 — 16 7 7 .45 FLL 3 6 — 2 .24 FRR 12 13 10 18 7 7 - 7 .47 FLR 6 5 6 4 7 8 11 - .39 (c) Phase III: M - 6 13 18 17 11 14 27 .46 MLL 12 — 2 13 11 9 16 11 .50 MRR 4 6 — 8 11 12 12 4 .42 MLR 6 - 3 6 .08 F 12 7 8 - 13 2 15 22 FLL 6 3 4 - 2 7 28 FRR 8 7 13 12 — 18 .43 FLR 18 5 8 6 17 16 15 — 45 123 TABLE Dl6.--Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by §s in left column) during Evening Food sessions in Sector B for Phases I(a), II(b) and III(c) of Study 3. Each §fs mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 1 1 .03 MLL 3 - 2 5 1 1 1 2 .22 MRR 60 . 34 - 75 34 50 14 9 .78 MLR 1 — 2 .07 F 3 1 - 1 1 11 FLL 1 1 — 04 FRR 1 3 1 - 1 .15 FLR 8 5 3 3 16 6 1 — .26 (b) Phase II: M - 1 .02 MLL 1 - 3 2 1 2 1 .21 MRR 3 — 1 09 MLR 59 23 89 - 32 38 2 6 .88 F 2 - 1 .07 ELL 1 1 - 1 03 FRR 1 4 2 3 - .22 FLR — 00 (c) Phase III: M - 4 25 19 5 10 3 3 .34 MLL 2 - 1 1 2 4 4 .25 MRR 1 - 1 2 .10 MLR 12 14 26 — 2 13 5 .51 F 1 - 1 .05 FLL 2 — .05 FRR 4 1 1 3 - 4 .16 FLR 1 1 1 5 2 - .15 124 TABLE Dl7.--Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by SS in left column) during Evening Food sessions in Sector C for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. M 'MLL MRR MLR F FLL FRR FLR Mgin (a) Phase I: M - l .02 MLL 2 - l l 3 l .14 MRR 5 12 - 10 1 5 l .33 MLR - l l .04 F — .00 FLL 1 — .02 FRR - .00 FLR 1 - .02 (b) Phase II: M - l .02 MLL 2 - 3 5 1 2 .19 MRR - l .01 MLR 7 ' 3 30 — 3 2 .43 F ' — .00 FLL l 2 l 1 - .09 FRR l l — 2 .07 FLR - .00 (c) Phase III: M - 7 8 l .17 MLL - 3 l 3 l .14 MRR - .00 MLR 3 25 - 2' l 2 l .34 F — .OO FLL .00 FRR - .00 FLR 1 2 - .05 125 TABLE D18.--Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by SS in left column) during Evening Water sessions in Sector A for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shoWn as well. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M -. 4 1. 6 8 1 .23 MLL 6 - 6 1 5 10 4 3 .40 MRR 5 3 — 1 2 8 5 3 .26 MLR 1 2 - 1 1 1 .11 F 1 2 3 1 - 3 6 2 .28 FLL 1 3 6 — 5 1 .17 FRR 6 8 10 1 9 8 — 1 .37 FLR 1 1 — .03 (b) Phase II: M — 6 2 4 5 1 .26 MLL 2 — 17 1 6 14 7 .32 MRR 2 8 - 2 7 1 1 .22 MLR 2 . 2 - 3 2 2 .24 F 2 1 4 - 2 2 .17 FLL 1 4 4 — 1 .12 FRR 10 1 1 1 - .15 FLR 1 1 - 2 — .10 (c) Phase III: M — 5 4 1 2 15 1 .36 MLL 2 — 2 2 4 2 .20 MRR 2 - 2 12 4 1 .25 MLR 1 2 - .07 F 1 - 3 2 .12 FLL 1 3 2 - 1 1 .13 FRR 3 1 6 - .13 FLR 2 . - .05 126 TABLE D19.-—Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by SS in left column) during Evening Water sessions in Sector B for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition,is shown as well. . Mean M MLL MRR MLR F FLL FRR FLR DI (a) Phase I: M - l l 1 2 l .12 MLL l - 2 2 2 l l .11 MRR 8 ll - 15 3 9 6 .43 MLR l 3 2 - — .11 F l — .01 FLL l 3 l — 2 15 FRR 2 1 5 - .13 FLR 2 l 1 1 .09 (b) Phase II: M - 2 4 2 2 2 .18 MLL 3 — 4 2 2 2 l l .21 MRR 2 — 1 4 2 .17 MLR 15 15 31 — 8 5 1 .58 F l 1 - 1 l 1 .ll FLL 2 l — .06 FRR 2 1 - 16 FLR 1 - 02 (c) Phase III: M - 6 4 8 2 l 22 MLL 7 — 2 1 l l .10 MRR 2 3 - 1 l 2 3 .20 MLR 4 l 12 - .10 F l l l l - .08 FLL 2 5 2 2 - 2 .22 FRR 2 2 l l - .15 FLR 1 - .02 127 TABLE D20.-—Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by SS in left column) during Evening Water sessions in Sector C for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 1 1 1 1 .09 MLL 3 - 3 1 2 7 .24 MRR 4 9 - 10 1 .19 MLR - .00 F 1 - .02 FLL — .00 FRR 1 — .00 FLR - .00 (b) Phase II: M — 1 02 MLL 4 - 2 1 2 1 .14 MRR — 1 .02 MLR 7 2 40 - 1 1 1 1 .36 F - 1 .02 FLL 1 - .02 FRR 1 ‘ - .02 FLR - .00 (c) Phase III: M - 4 3 12 1 1 1 .19 MLL 1 - 1 3 2 2 .14 MRR 1 - 1 1 1 .10 MLR 1 1 9 — 1 .19 F - .00 FLL - .00 FRR 1 — .02 FLR . - .00 128 TABLE D21.--Intragroup response matrices for Group 3 (time controls) showing aggressive supplants (by Ss in left column) during Morning sessions (all Sectorg combined) for Phases I(a), II(b) and III(c) of Study 3. Each S's mean DI, computed as an average of the six daily DIs measured during each phase in this incentive condition, is shown as well. M 'MLL MRR MLR F FLL FRR FLR MS?“ (a) Phase I: M - 2 . .02 MLL 1 - l 2 1 l .14 MRR 9 - 22 8 6 2 4 .24 MLR l — l 1 .12 F - .00 FLL - 2 .02 FRR 1 - .02 FLR 2 5 l 7 . 2 4 - .17 (b) Phase II: M — l l .05 MLL - 3 3 .14 MRR - .00 MLR 2 4 12 — l l l .31 F l — l _ . .05 FLL - .00 FRR - .OO' FLR - .00 (c) Phase III: M - 6 l3 2 2 4 4 .24 MLL - 1 l l 1 .10 MRR l — .02 MLR 2 1 28 — 3 3 5 3 .38 F - .00 FLL - .OO FRR l — .Ol FLR 2 1 2 - .10 MICHIGRN STATE UNIV. LIBRRRIES 1|HIWIHIIIIHIIIIWIIIIIIHIIHIIIHIIIIHIIHHIIIHIHI 31293006793404