Psychopharmacology (1989) 97:149-151 Psychopharmacology 9 Springer-Verlag 1989 Temporal and sequential patterns of agonistic behavior: effects of alcohol, anxiolytics and psychomotor stimulants K.A. Miczek, M. Haney, J. Tidey, T. Vatne, E. Weerts, and J.F. DeBold Department of Psychology Tufts University, Medford, MA 02155, USA Abstract. Social and agonistic interactions are composed of a range of species-typical acts, postures, displays and other communicative signals that follow characteristic pat- terns. Descriptive and analytic methods permit an assess- ment of the temporal and sequential features of highly probable patterns of agonistic interactions. Analysis of the intervals that separate consecutive attacks by a resident mouse or rat toward an intruder identifies bursts or epochs of attacks. Amphetamine (1.25, 2.5 mg/kg), but not diaze- pam or alcohol, alters the burst pattern of attack behavior. Higher doses of alcohol, but not diazepam, in either resi- dent male rats or in lactating rats confronting an intruder, reduce the sequences of aggressive acts and postures with high transition probabilities as identified by lag sequential analysis. These results suggest that temporal and sequential patterning mechanisms may be differentially altered by am- phetamine- and alcohol-type substances. These neural mechanisms for behavioral patterning appear to be relevant for many types of behavior. Key words: Amphetamines Alcohol - Benzodiazepines - Dopamine receptor agonists - Agonistic behavior - Ag- gression - Anxiolytics - Statistical analysis Agonistic and social interactions involve behavior that is intricately patterned both in time and in sequence. We illus- trate how simple quantitative ethological methods detect changes in the patterning of agonistic behavior that are produced by alcohol, anxiolytics and psychomotor stimu- lants, in addition to or even in the absence of marked chan- ges in the total incidence of this behavior. For example, in rats the behavioral elements of pursuit, threat, attack bite and aggressive posture can be identified to occur (1) as part of a sequence with one element following the next with high probability and (2) as part of an epoch or burst of aggressive behavior that alternates with periods of rela- tive behavioral quiescence. Temporal patterns of agonistic behavior The initiation, execution, and termination of many behav- iors such as sleeping, feeding, drinking, reproducing, mater- nal care, aggressive and social interactions occur in predict- able temporal patterns. Ethologists have characterized these Offprint requests to: K.A. Miczek, Research Building, Tufts Uni- versity, 490 Boston Avenue, Medford, MA 02155, USA temporal patterns quantitatively using various tools (e.g., Fagen and Young 1978; van Hooff 1982). One such typical pattern is the alternation between epochs of intense and frequent behavior with behaviorally quiet periods; agonistic behavior of many species follows this course. Ever since detailed ethological assessments of drug effects on agonistic behavior became available (e.g., Miczek and Krsiak 1979; Miczek 1987), many students of behavior have noted marked changes in the pattern of behavior, not only in the total frequency or duration of agonistic behavioral ele- ments during experimental sessions. During confrontations between a resident male mouse and an intruder, the resident engages in pursuit, sideways threat, attack bites, and tail rattles, in addition to several non-agonistic activities such as grooming, rearing and walk- ing (Miczek and O'Donnell 1978). Each occurrence of these and other behavioral elements as well as those exhibited by the intruder mouse are measured in terms of start and end time from video records of the agonistic interactions, using a computerized data acquisition system (Miczek 1983). When the resident is administered d-amphetamine sul- fate at different dose levels 30 min before a weekly 5-min confrontation with a male intruder mouse, significant dose- dependent changes in agonistic and non-agonistic behaviors are seen (e.g., Miczek 1981). d-Amphetamine decreases the frequency of pursuits, sideways threats, attack bites, and tail rattles by the resident as well as increases the frequency and duration of non-agonistic walking. The changes in at- tack behavior reach statistical significance only at higher amphetamine doses (5.0 mg/kg). Substantial changes in the temporal pattern of agonistic behavior are detected in amphetamine-treated resident mice at doses that are lower than those necessary to decrease the total incidence (Fig. 1 A). An exponential curve fits ade- quately the distribution of the intervals between consecutive attack bites exhibited by all members of the treatment group. Two component curves describe two segments of the interval distribution, a steep one accounting for 70-80% of all cases and fitting the short intervals and a shallow one fitting the remaining long intervals. At the 2.5 mg/kg amphetamine dose, the distribution of the short intervals characterizing the bursts or epochs of attack behavior is markedly wider, including 4-5 times longer intervals than under saline control conditions. At the 5.0 mg/kg dose, the short intervals do not form a separate segment of the curve any more, indicating a random distribution of intervals be- tween consecutive attacks; the characteristic burst pattern