On the dual nature of maternal aggression in rats

Presentation of a natural predator, a cat, was used to differentiate elements of maternal attack by female rats on a male intruder. Following exposure (without direct physical contact) of post-partum females to a cat or to a toy stuffed cat (control group), the females were replaced in their home cages and presented with a male intruder rat. Cat exposure reliably decreased lateral attack to the intruder, as well as locomotion, but had no effect on either jump attack or an upright defensive posture (boxing). Since predator exposure produces a somewhat durable increase in defense, along with inhibition of nondefensive behavior, these results suggest that maternal aggression represents a mixture of offensive, usually related to competition, and defensive (protective) behaviors. The results indicate that maternal aggression, as a parental care behavior, appears to be at least partially resistant to fear. © 1996 Wiley-Liss, Inc.     

Does fear modulate defensive and offensive types of maternal attack in mice?

Swiss CD‐1 lactating mice show a different pattern of attack toward intruders of differing sex, displaying defensive attack against the male (bites on the head and ventrum associated with fear) and offensive attack against the female (bites on the back and flanks with no elicitation of fear). This dichotomy may reflect diverse functions of maternal aggression: the attack toward males (the more infanticidal gender in laboratory strains) has been interpreted as a counterstrategy to infanticide, whereas the attack toward females may serve to establish a social hierarchy or to space rivals of the same sex. In terms of proximal mechanisms, fear may be a key factor involved in the modulation of the different patterns of attack. In Experiment 1 we compared the pattern of attack of lactating females in Swiss CD‐1 and Wild mice toward male and female intruders in relation to fear components of behavior of the attacking dams. Results showed that in Swiss mice, male intruders were attacked with a defensive type of attack accompanied by high levels of fear, whereas female intruders did not elicit fear in the attacking animal but were attacked with an offensive pattern. In Wild mice, both types of intruders were attacked with a defensive pattern; notwithstanding, fear was evident only toward male intruders. This suggests that fear is not totally responsible for the expression of the defensive type of attack. To test the hypothesis that defensive attack toward male and female intruders may be related to the infanticidal potential of the intruder, Experiment 2 examined levels of infanticide in both male and female Swiss CD‐1 and Wild mice. Swiss female mice showed virtually no infanticidal behavior, whereas Swiss males and both sexes of Wild mice showed similarly high levels of infanticide (55%–75%). From a game theory perspective, the defensive pattern of maternal attack toward female intruders in Wild mice is discussed as “extreme” defense of a high value resource and thus, functionally, a competitive form of aggression. Aggr. Behav. 26:193–203, 2000. © 2000 Wiley‐Liss, Inc.     

A pinch or a lesion: a reconceptualization of biting consequences in mice

Intruder and resident male colony mice exhibit an array of distinct defensive and offensive behaviors. Intruders typically show more boxing, flight, defensive sideways position, on the back position and general locomotion, while residents exhibit higher levels of attack, olfactory investigation, aggressive grooming, and biting, with a preference for dorsal bite locations. Here, analysis of bite locations on the body of the intruder mice showed that the majority of bites produced few lesions (i.e. actual puncturing of the skin) when compared to scrapes or no markings. Most bites were directed to the back of the opponent animal with very few bites directed towards the opponents' vulnerable ventrum. In particular, bites directed at the relatively hairless ventrum produced no lesions. These findings, along with previous work on mice and rats, suggest that intraspecific offense with preferred target sites for biting, facilitates an effective but largely nonwounding interaction between resident and intruder mice. Furthermore, bruise and wound analyses suggest an association between bite targets and tissue damage. The preference for specific bite locations may be complimented by a differential intensity of attack, with the back attracting the strongest bites and the ventrum the weakest. This apparent nonwounding fighting pattern contradicts the current evaluation of rodent wounding severity in this paradigm and can therefore refine the usage of this model and of the protocols associated with it.     

Offensive and defensive aggression in humans: A longitudinal perspective

Offensive and defensive aggression were distinguished on the basis of associated events. Behavior delivering noxious stimuli was defined as defensive when it was a response a threatening situation and as offensive when it was an unprovoked act. The existence of a general aggression factor was implied by the finding that initiators of aggression also readily defend themselves if attacked. The use of a peer nomination and teacher rating variable for offensive aggression, “Attacks without reason, teases others, naughty things”; and another for defensive aggression, “Defends him/herself if teased, but does not tease or attack others without reason,” yielded differential results. Offensive aggression at age 14 was significantly predicted by aggressive behavior age 8, while defensive aggression was not. Intercorrelations of rating variables at 14 showed that offensive aggression was related to weak self-control and defensive aggression to strong self-control. Only offensive aggression at age 14 predicted criminality at age 20. The results indicate a diversity of motivations for aggressive behaviors, as suggested by observations on animal behavior.     

Relationship between mousekilling and conspecific aggression in the male rat

In order to investigate the relationships between mousekilling and conspecific aggression, behavioral variables of killer and nonkiller rats were compared in a “resident-intruder” paradigm, in resident as well as in intruder animals. The occurrence of offensive items (offensive sideways, attack) was significantly higher in killer rats when they were residents; their corresponding opponents displayed more defensive behaviors. No significant difference in aggressive behaviors was noted when the comparison was done in the intruders. These results and those of previous studies suggest that there is a correlation between mousekilling and intraspecific offensive behaviors. Some similarities in the situations where both behaviors are elicited–eg, introduction of an unfamiliar intruder into a familiar environment–may contribute to the existence of such a correlation and the possibility of common mechanisms underlying both behaviors is discussed.     

Offensive and defensive bite-target topographies in attacks by lactating rats

The attacks by resident lactating Wistar rats on sexually naive conspecifics of both sexes were examined. Male and female intruders were equally attacked in terms of frequency and number of bites, but the topographies of biting seen in these encounters were different. Similarly to male-male agonistic interactions, females were attacked in a fashion which avoided bites to the head and snout (“offensive” attack), whereas males were frequently bitten on such vulnerable regions (“defensive” attack). This dichotomy in bite pattern suggests that different motivations and functions underlay maternal aggression in these situations. The defensive attack on males may be a deterrent to infanticide since only male intruders counterattack lactating females and kill their pups. The attack on females may be concerned with resource competition.     

Social dominance and individual aggressiveness

Male rats exhibiting high, moderate, or low levels of offensive aggressive behavior in interactions with intruders in their home cage were grouped in mixed-sex colonies with 1 male of each aggression-level group in each colony. Agonistic interactions measured 1, 2, 3, 7, 14, 21, and 22 days after colony formation indicated that highly aggressive males on pretests continued to be more aggressive, becoming the dominant colony male in five of seven colonies and attacking intruders more often than less aggressive males. In the two remaining colonies the moderately aggressive male became dominant. This relationship, which was consistent over a number of indices, including offensive and defensive behaviors, and wound counts and wound sites, was seen even when a substantial weight differential favored the less aggressive animal. Dominance relationships were rapidly established and within-group fighting declined significantly over the 21-day test period. Pretest offensive levels also influenced the behavior of subordinates, with high or moderately aggressive subordinates showing more defense in interactions with dominants and receiving more wounds than did low-aggression subordinates. Dominant males also showed more defense in interacting with those subordinates which had been more aggressive during pretests. This pattern of results suggests that aggression level of the subordinate as well as the dominant may be an important factor determining the intensity of agonistic interactions in male rats.     

The effects of negative reinforcement for irritable aggression on resident-intruder behavior

This experiment demonstrated that rats trained to display elevated levels of shock-induced aggression in a negative reinforcement paradigm displayed more boxing behavior than yoked control groups in a later test in which intruder rats were placed in the home cage of resident rats. Resident or intruder status did not affect the influence of the negative reinforcement procedure on the observed resident-intruder behavior of trained animals; however, naive intruders paired with trained residents displayed increased defensive behavior, suggesting that negative reinforcement for shock-induced aggression affected the behavior of these residents.     

Habituation of aggressive behavior in mice: A parametric study

The decline of aggression as a result of repeated episodes of fighting may represent a habituation phenomenon. The frequency of biting and sideways threat by male mice toward conspecific intruders declined over ten 5-min confrontations. Aggression returned to about 50% of original levels when a new intruder was introduced. Variations in length or frequency of confrontation indicated that the rate of decline of attack bites and sideways threats depended on these confrontation parameters. The condition that provided the fewest opportunities to fight provided the least decline in the frequency of attack and threat. The frequency of attack bites and sideways threats elicited by the presentation of a new intruder after repeated confrontations was greater than the frequency of attack bite and sideways threat in the last confrontation with the original intruder, when the resident and intruder were relatively inexperienced. Spontaneous recovery of attack and threat occurred but rate of recovery did not depend on confrontation parameters. The results indicate that the decrement of aggression in mice in repeated confrontations may be the result of habituation but fatigue is an important cofactor.     

Male-female differences and the influence of neonatal and adult testosterone on intraspecies aggression in rats.

Male and female albino rats were tested for intraspecies aggression without the use of shock. In the first experiment, male pairs showed more biting attacks, offensive sideways movements, and self-grooming than did female pairs; male pairs also showed more stereotyped defensive/submissive behaviors and were wounded more frequently. The second experiment examined the effects of neonatal castration and testosterone propionate (TP) administration on fighting. Males castrated at birth attacked other males less frequently than did controls when tested with TP treatment as adults. The TP given at birth to neonatally castrated males restored attacks to control levels. Females given TP as neonates did not differ from either male or female controls. Other aggressive/defensive behaviors, however, did not show this pattern. The results suggest that while the presence of testosterone during a brief postnatal period and during adulthood is necessary for attack behavior to occur, other related behaviors may not be affected in a similar manner.     

The offensive nature of maternal aggression in mice: Effects of fluprazine hydrochloride

The phenylpiperazine fluprazine hydrochloride was used to test the hypothesis that maternal aggression in laboratory mice reflects offensive motivation. Lactating females (n = 8 per condition) tested for aggression following drug treatment (1.0 mg/kg and 5.0 mg/kg of fluprasine) were significantly less aggressive than saline controls according to all measures used. The findings support the hypothesis that maternal aggression is offensive in nature, and provide no evidence that maternal attack is a mixture of offensive and defensive tendencies.     

The effects of urine on aggressive responses by male golden hamsters

Twenty adult male golden hamsters were isolated into individual cages for a period of six weeks, at the end of which time they had introduced into their home cages, on three occasions, a castrated male intruder. On each occasion the castrated intruder was daubed on the anogenital region with urine from one of three sources: (1) intact females, (2) other castrated males, and (3) intact males. Urine from a different source was applied to the castrated intruder on each of the three tests. Resident males consistently showed more aggression, sniffing, and following and less defensive behavior than the intruders. However, aggression by the resident males showed a significant variance over the three urine treatments given to the intruder. It is concluded that like male mouse urine, male hamster urine contains attack-provoking cues, but that unlike that of the mouse, female urine does not appear to be attack-inhibiting in this species.     

The ventromedial hypothalamus and aggressive behaviour in rats

Male rats were given bilateral lesions in either the anterior or posterior ventromedial hypothalamus (VMH). The intermale aggressive behaviour of these animals within their own territory was observed before and after the surgical procedure and compared with the behaviour of sham-operated animals. The effects of anterior VMH lesions include an increased tendency to respond with frontal threatening upon approach of a conspecific male. This behaviour closely resembles the aggressive responses described in “shock-induced aggression” tests. Posterior VMH lesions facilitate territorial aggressive behaviour characterized by approaching the opponent followed by lateral threatening and fighting. It is suggested that 2 distinct neural substrates exist, which serve to inhibit defensive (anterior-VMH) and offensive (posterior-VMH) intermale aggression, respectively.     

Pain-induced aggression: An ethological perspective

Pain-induced aggression has been studied mainly in animals, but generalizations are made to human aggression, noteably in Berkowitz’s “aversive-stimulation” view. This review reevaluates the motivational basis of pain-induced attack by considering it from an ethological perspective. Studies supporting the view that shock-induced attack is fear-motivated are briefly reviewed before consideration of evidence that does not support this view, from naturally-ocurring examples of pain-induced attack, and from examples of sex and hormonal effects. A reconcilliation between the apparently conflicting views is suggested by distinguishing between the functional (competitive-protective) dimension and the causal (offensive-defensive) dimension. It is proposed that pain-induced attack is protective in function but can be offensive or defensive (or both) in its form; the offensive form is only rarely observed under laboratory conditions. The implications of this view are discussed (1) for the use of shock-induced fighting as an animal model of aggression; and (2) for studies of human pain-induced aggression. Ethical issues surrounding shock-induced aggression research are considered.     

Pain-induced aggression: An ethological perspective

Pain-induced aggression has been studied mainly in animals, but generalizations are made to human aggression, noteably in Berkowitz’s “aversive-stimulation” view. This review reevaluates the motivational basis of pain-induced attack by considering it from an ethological perspective. Studies supporting the view that shock-induced attack is fear-motivated are briefly reviewed before consideration of evidence that does not support this view, from naturally-ocurring examples of pain-induced attack, and from examples of sex and hormonal effects. A reconcilliation between the apparently conflicting views is suggested by distinguishing between the functional (competitive-protective) dimension and the causal (offensive-defensive) dimension. It is proposed that pain-induced attack is protective in function but can be offensive or defensive (or both) in its form; the offensive form is only rarely observed under laboratory conditions. The implications of this view are discussed (1) for the use of shock-induced fighting as an animal model of aggression; and (2) for studies of human pain-induced aggression. Ethical issues surrounding shock-induced aggression research are considered.     

Rank order in pairs of communally nursing female mice (Mus musculus domesticus) and Maternal Aggression Towards Conspecific Intruders of Differing Sex

This study examined social behavior between pairs of unfamiliar lactating females, with litters of the same age, at different periods after parturition (3, 7, and 17 days). Tests were generally followed by the formation of communal rearing nests, and subsequent maternal attack on intruders of differing sex was assessed. In all three intervals lactating females showed ritualized attack with formation of clear dominance-subordination relationships before combining litters into communal nests. The dominant females in 90% of cases started to retrieve alien pups into their nests. Agonistic behavior and communal nest formation were most rapid when pups were around 3 days old. Maternal attack on conspecific intruders was mainly displayed by the dominant lactating females. Male and female intruders were equally attacked (in terms of frequency and intensity of attack), but there was less such aggression when pups were around 17 days of age. Nevertheless the topography of biting attack employed against female and male conspecific intruders was different. Females were attacked using a strategy avoiding bites to the head and ventral surface (indicative of “offensive” behavior) whereas males were severely bitten on vulnerable body regions (indicative of “defensive” behavior).     

Reflexive fighting in the albino rat: Aggressive or defensive behavior?

Specific agonistic responses of albino rats were compared for dominant colony rats and intruders, and for rats in a “reflexive fighting” task. The “reflexive fighters” showed high levels of defensive responses such as boxing and freezing, and very low levels of aggressive behaviors such as piloerection, biting, and the lateral display. This pattern clearly suggests that the behaviors measured in the reflexive fighting task reflect conspecific defensive reactions, rather than “shockelicited aggression.” Moreover, striking responses seen in the reflexive fighting task also occur at a high rate when footshock is given to a solitary rat held in a boxing posture. Thus it appears that “reflexive fighting” primarily involves defensive rather than aggressive responses.     

Attack and defensive behaviors in the albino mouse

Attack by dominant male colony mice on intruders included chasing and lateral attack behaviors, while the corresponding intruder behaviors were flight, boxing, and checking. Both of these are similar to the attack and defensive behaviors of colony rats and intruders. However, mice did not show a significant constraint on bites to ventral areas, and the rat defensive behavior of lying on the back, which is effective because of this constraint, was rare; the corresponding “on-top” behavior of attackers was almost absent in mice. These findings strongly support the view that intraspecific attack and defensive behaviors, and target sites for bites, are interrelated factors facilitating effective but nonlethal agonistic interactions in muroid rodents.