It is generally agreed that there are at least several different types of aggressive behavior in mammals, and, in recent years, interest has focused especially upon three common types: offense, defense, and predation [7, 19]. Offense is characterized by a bite-and-kick attack and offensive sideways posture, and it may be elicited by confronting a mature male cat, rat, or mouse with an unfamiliar male opponent of the same species. Offense may also be elicited in either males or females by giving one piece of food to two hungry mice [11] or rats [21]. Defense is characterized by a lunge-and-bite attack or upright posture with boxing movements of the forelimbs in the mouse or rat [3] and by hissing, arching of the back and striking with the forelimbs in the cat [14]. Both offense and defense involve extensive autonomic arousal. Predation is exemplified by the mouse-killing of the rat and the killing of mice or rats by the cat, a behavior that involves minimal autonomic arousal and has therefore been called "quiet attack" [10].
Reviewing the literature until 1979, I proposed that offense and defense could be distinguished in terms of their neural organization and that probably predation could be distinguished as well, although it was less well understood [2]. A more specific prediction was that it should be possible to abolish offense by making a specific lesion in the brainstem, since it had already been shown that defense could be abolished by lesions of the midbrain central gray [8]. The reason that brain stem rather than forebrain regions were predicted was because studies had shown that the forebrain is not necessary for any of these behaviors. Predation is retained after surgical isolation of the hypothalamus [9]; offense is retained after cross-sectional destruction of the hypothalamus [2]; and defense is also retained after cross-sectional destruction of the hypothalamus [1].
The experimental clue that guided the direction of the present study was provided by a previous report [13] that median raphe lesions abolish offense in mice. It is the only study in the literature that claims to abolish offense with brain stem lesions [2]. Although we failed in pilot studies to replicate those findings in the rat, we did find that larger lesions that destroyed part of the tegmentum on either side of the median raphe were sufficient to abolish offense. The present study replicates our pilot results and extends the testing procedure to include defense and predation as well as offense.
The experimental subjects were adult male rats of a hybrid strain (DA by S2). Preoperatively, they were tested for predation by introducing an unrestrained adult male mouse into their home cage, and they were tested for offense by introducing an adult male rat of a different strain into their home cage. Ten rats, approximately 10% of all the rats given behavioral testing, fulfilled the testing criteria which were three consecutive kills of a mouse within 20 minutes, each test being on a different day, and three consecutive tests with a bite-and-kick attack against a rat intruder within 20 minutes, again each test being on a different day. The percentage of animals showing predation and offense was consistent with findings in other rat strains [4, 12].
The scoring of offense and predation were as follows. Predation was scored in terms of the latency to kill the mouse, or, in a few tests when no kill occurred, the latency to a non-lethal biting attack directed at the mouse. Offense was given an "offense score" consisting of one point for the presence of offensive sideways posture, which is a low level of offense, and additional points for each occurrence of bite-and-kick attack, which is a higher level of offense [15]. Offense tests were 20 minutes in duration and predation tests were 20 minutes in duration or until the killing of the mouse, whichever came first.
 |
 |
 |