Biological Journal of the Linnean Society (1990) 41: 257-269 Variation in aggressiveness in house mouse populations PAUL F. BRAIN Biological Sciences, University College of Swansea, SA2 8PP, Wales AND S T E F A N 0 PARMIGIANI Istituto di Zoologia, Universitri degli Studi di Parma, 43100 Parma, Italy Clearly the ability of ‘house mice’ to vary their social structures is an important feature contributing to their success in a wide range of habitats. Social structure is strongly influenced by aggressiveness and other behaviours in male and female mice. Material is presented illustrating how genotype, intrauterine location and social experiences influence dyadic encounters in this ‘species’. KEY WORDS:-House structure. mouse - aggressiveness - genes - intrauterine location - experience CONTENTS Introduction . . . . . . . . . . . . . Genetic influences on ‘social aggression’ in male mice . . . Material and methods . . . . . . . . . . . Husbandry . . . . . . . . . . . . Behavioural tests . . . . . . . . . . . Results . . . . . . . . . . . . . . Individually housed male ‘residents’ us. ‘TO’ strain intruders Individually housed male ‘residents’ us. like-strain intruders. Responses of intruders to residents . . . . . . . . . . . . . ‘Female aggression’ and social structure Intrauterine location phenomena . . . . . . . . Effects of prior social experiences on individual aggressiveness . Concluding comments . . . . . . . . . . . Acknowledgements . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ social 257 258 258 258 258 259 259 260 26 1 262 263 264 267 267 267 INTRODUCTION Competition for space, mates, social status and resources is a n almost ubiquitous phenomenon in the Animal Kingdom. Intraspecific aggression crucially shapes social structure and spatial distribution of conspecifics via territorial and/or of migratory responses (Van Oortmerssen & Busser, 1989). 0024-4066/90/090257 + 13 S03.00jO 257 01990 The Linnean Society of London 258 P. F. BRAIN AND S. PARMIGIANI Understanding relationships between aggressiveness and distribution may o n b be currently possible in the house mouse. The ‘house mouse’ is clearly a very divergent ‘species’, essentially consisting of a group of seven recognized species, numerous subspecies and several chromosomal races. This highly opportunistic animal flourishes in environments as diverse as fields, cold stores, warehouses, hayricks, Pacific atolls and islands close to Antartica. Individuals may be largely surface-dwelling or inhabitants of complex burrow systems. Different populations of mice, at different locations and a t varied times within the same location, exhibit different social organizations (e.g. male territoriality and hierarchial groups). Such organizations (which often seem suited to prevailing local conditions of food distribution, available cover and animal density) generally reflect variations in female and male aggressiveness. Genetic endowment, intrauterine location and experience give rise to individual differences in murine aggressiveness which could underpin variations in the probability of assuming particular social structures. Different behavioural phenotypes may be suited to different phases of the population ‘cycle’ seen in such animals. GENETIC INFLUENCES ON ‘SOCIAL AGGRESSION’ IN MALE MICE Attack and threat may serve different functions. Fighting and threat may be employed in offence or defence (both personal or of a nest site or parental investment) in such animals. These activities may also be used to assess fitness of potential mates. Over 200 strains of inbred laboratory mice offer unique opportunities to assess whether laboratory tests of ‘aggression’ actually measure the same attribute. This has been doubted (see Brain, 1989); indeed preliminary accounts suggest that such tests assess mixtures of offensive, defensive and (in some cases) predatory motivations (Mainardi et al., 1986; Jones & Brain, 1987). More detailed measures of behaviour in murine ‘social aggression’ encounters in which the genotype of the resident and of the intruder are varied are reported here. Using an ‘ethoexperimental’ approach (Brain, 1989) one can distinguish variations of aggressiveness from changes in attack mediated by differences in sociability or fearfulness and/or altered in social communication. MATERIAL AND METHODS Husbandry Outbred albino, Tuck Ordinary (‘TO’) and Swiss-Webster and inbred NZW/Ola, BALB/c, C57BL/10, DBA/2, CBA/Ca and C3H/He stocks were used in this study. All animals were bred and housed in the Animal Facility of the University College of Swansea under standardized housing and rearing conditions (see Jones, 1983). Behavioural tests Intermale fighting induced by individual housing (otherwise known as ‘social aggression’) is the most investigated form of ‘aggression’ in laboratory rodents, is AGGRESSIVENESS I N THE HOUSE MOUSE 259 reliably generated in many strains and is comparable to behaviour seen in natural populations (Corti et al., 1989). One week after weaning, males of each line were individually housed for 28 days. Each ‘resident’ had a single 10-minute home cage encounter with an anosomic (largely lacking a sense of smell) grouphoused male ‘intruder’. Such opponents elicit attack but rarely initiate such behaviour or retaliate when attacked (Brain & Al-Maliki, 1978). Male residents encountered a ‘TO’ strain intruder or an intruder of their own strain (detailed results are only given for the former). Tests took place during the ‘dark’ phase of the light cycle under dim, red lighting (c. 9 lux) using transparent cage-lids to facilitate videotaping. There were 8- 12 animals per observation category. Tapes were transcribed as durations of some 45 behavioural components (defined in Brain, McAllister & Warmsley, 1989). Elements were assigned (on the basis of sequence analysis) to broad categories of activity (see Brain et al., 1989) and times allocated by experimental subjects to categories summed. Encounters between individually housed males and ‘TO’ or same-strain opponents were examined because:(i) some attack behaviour was evident in most strains; (ii) this was the only form of aggression examined in all strains using both ‘TO’ and same-strain opponents; (iii) videotape recordings of encounters involving individually-housed males could be analysed en masse increasing observer reliability. RESULTS Durations of broad categories of activity are tabulated as medians, with ranges in parentheses. The data refer to all animals tested, irrespective of whether they showed overt attack. Individually housed male ‘residents’ vs. ‘TO’ strain intruders Table 1 lists times allocated to behaviours in individually housed males of the strains in response to TO standard opponents. Interestingly, durations of ‘aggression’ (comprising e.g. ‘chase’, ‘tail rattle’, ‘biting attack’ and threat postures) do not show a consistent relationship to the more conservative measure of accumulative attacking time (i.e. number of seconds allocated to biting attack on the ‘standard opponent) across strains. Perhaps proportions of overt attack to other components of aggressive behaviour (e.g. threat postures) vary from strain to strain? Overt attack was not associated with particular distributions of other broad activities. For example, DBA/2 and C3H/He males showed very different levels of attack but similar durations of social behaviour. Other strains showing levels of attack greater than those of DBA/2 males (i.e. T O , Swiss and C57BL/10) have lower incidences of social behaviour. The ‘non-aggressive’ CBA/Ca and C3H/He strains, although showing comparable levels of social and non-social behaviour, have very different durations of defence and immobility. Higher levels of social behaviour were, however, generally associated with reduced defence. Such associations indicate ‘tolerant’ residents, since defence is often evoked by the intruder’s behaviour (e.g. reciprocated social investigation) P. F. BRAIN AND S. PARMIGIANI 260 TABLE 1. Durations of activities shown by individually-housed male residents in encounters with ‘TO’ strain intruders Strain of resident Category of activity Social sexual TO SWISS NZWjO 1a BALB/c C57BL/IO DBA/2 CBA/Ca C3H/He 179.7 (74.0-460.4) 188.5 (123.6-388.4) 291.1 (114.2-413.6) 149.2 (46.5-223.3) 177.6 ( 1 17.4-33 1.O) 321.9 (231.5-422.0) 241.9 (168.3-47 1.4) 360.8 (227.6-462.4) Active non-social Immobile Aggression Defence/flight 239.9 45.1 (0.4-72.4) 35.4 (8.1-99.6) 19.5 (3.6-53.6) 101.5 (51.3-242.5) 24.6 (5.6-41.9) 30.2 (2.5-188.6) 53.3 (15.4-139.1) 37.3 ( 7.3-88.4) 61.0 (2.1-105.0) 49.9 (5.6-158.0) 7.6 (4.8-14.0) 10.7 (3.6-36.5) 7.3 (1.8-14.5) 13.6 (5.1-33.2) 15.9 (1 .l-70.2) 1.1 (0.0-32.6) 35.8 (4.9-59.2) 4.8 (2.0-1 1.8) 11.4 (5.3-2 1.3) ( 108.3-424.3) 308.6 (162.6-335.0) 219.0 (150.2-360.6) 251.1 ( 148.7-343.9) 306.1 (230.6-424.4) 218.8 (146.4-304.9) 280.0 (8 1.6-357.2) 210.9 ( 1 18.0-333.0) 20.2 (5.9-160.2) 11.2 (3.0-15.3) 15.7 (6.1-38.1) 16.6 (1 1.9-34.4) 9.4 (6.0-29.1) during periods of close physical contact (recorded as ‘social’ activity). Using this line of reasoning, C3H/He and BALB/c residents are respectively the most and least ‘tolerant’ to TO opponents. Significant strain differences were evident in durations of social and defence but not non-social behaviour. Differences in ‘aggression’ durations largely reflected strain differences in the incidence and intensity of attack fJones & Brain, 1987). The most obvious strain difference, however, concerns the extremely high levels of immobility shown by BALB/c males. Blanchard & Blanchard (1989a, b) interpret immobility as indicative of ‘fear’; others suggest it represents passive avoidance. If one applies the latter interpretation, BALB/c males ‘passively avoid’ their opponent when not engaged in attack. Van Oortmerssen (1971) suggested that extremes of avoidance and defence (also of long duration in this line) might typify a strain descended from surface-living (c.f. burrow-living) ancestors, which could not easily hide from an aggressor and would be forced to flee. Since ‘crouching’ and ‘freezing’ (components of immobility) have been interpreted (Grant, 1963) as responses of animals prevented from fleeing by the cage’s physical restrictions, the immobility of BALB/c mice seems consistent with thwarted flight and hence with a surface-living ancestry. Individually housed male ‘residents’ vs. like-strain intruders Most of the conclusions reached above concerning strain differences in behavioural characteristics were confirmed when using like-strain opponents (consequently, these data are not given in detail). Some differences in the resident’s behaviour may be, at least in part, consequences of the opponent’s behaviour. Responses of opponents to a given ‘behavioural stimulus’ of a resident may be determined by many factors, e.g. including the genotype and previous experience of that opponent. For example, C57BL mice (Van Oortmerssen, 1971) and juvenile TO males (Brain et al., 1981) tend to flee rather than assume AGGRESSIVENESS IN THE HOUSE MOUSE 26 1 defensive postures in response to being bitten. Since attacking mice frequently chase fleeing opponents (Childs, 1979), ‘strain differences’ in chasing by residents might simply reflect variability in fleeing by opponents. To detect behaviour which is ‘typical’ of a resident strain therefore, one should seek features which are consistent across different types of opponent. As BALB/c mice showed immobility in both situations, this may be ‘characteristic’ of this strain. Similarly, high levels of social behaviour seem typical of isolated C3H/He males. All resident strains (except C3H/He) showed more defence against TO than like-strain opponents; significantly so in five out of eight strains. Perhaps TO opponents have a strong tendency to reciprocate social investigation during periods of physical contact with the resident and hence stimulate more defence? Alternatively, residents may be more generally ‘tolerant’ of opponents of their own strain. Comparing social behaviour of TO and same-strain intruders does not resolve this issue, since the recorded durations are similar. Perhaps the social behaviour of TO intruders differs in composition from that of like-strain intruders e.g. involving more physical contact, such as grooming? Same-strain intruders, conversely, might engage in activities such as ‘attend’, which involve no direct contact with residents and are unlikely to stimulate defence. General assumptions that similar durations of activity reflect similar responses to particular experimental situations are unjustified where the composition of individual components is unknown. This is a common omission in investigations employing this type of measure, however such studies are still more informative than those which simply utilize measures of attack. Responses of intruders to residents Table 2 provides data for responses of TO intruders to different residents. Comparisons of opponent behaviour in encounters with particular strains of resident may suggest which behaviours are ‘typical’ of the resident strain. For example, immobility is particularly evident in the opponents of NZW/Ol a residents. Grant (1963) noted that ‘crouch’ (the major contributor to immobility) is a frequent response to being groomed. One might, therefore, reasonably predict high levels of grooming of their opponents by NZW/Ol a residents (as is the case). ‘Crouch’ is shown by animals when their movement is restricted by highly ‘aggressive’ residents. Opponents exposed to other residents as aggressive as NZW/Ola males (e.g. Swiss) do not, however, show this degree of immobility. The durations of defence shown by opponents are roughly proportional to the amount of overt attack received. Durations of defence are, however, disproportionately high in both the TO and same-strain intruders encountering Swiss and NZW/Ol a residents. I n particular, the ratio of passive (rigidly-held defence postures) to active (e.g. flight and kicking) defence is high in these opponents. Perhaps more damage is inflicted by bites delivered by these residents stimulating more extreme defence? The ratio of passive : active defence is low in the remaining strains of opponents and especially in the case of C57BL/10 intruders, where it indicates a greater tendency for flight (see Van Oortmerssen, 1971). Irrespective of their strain or the strain of the resident encountered, opponents showed elevated non-social behaviour (largely cageexploration) and reduced social behaviour and aggression (certainly involving no overt attack). P. F. BRAIN AND S. PARMIGIANI 262 TABLE 2. Durations of activities shown by ‘TO’ strain intruders in encounters with individuallyhoused male residents Strain of resident Category of activity Social sexual 14.1 (7.4-27.3) 23.1 SWISS (9.8-28.1) 12.7 NZW/OlOa (8.8-42.1) 23.8 BALB/C (13.6-116.1) 16.0 C57BL/10 ( 12.9-25.0) 12.5 DBA/P (9.9-16.5) 13.2 CBA/Ca (9.6-17.0) 14.0 C3H/He (6.0-79.7) TO Active non-social Immobile Aggression Active defencelflight Passive defence 458.4 (161.7-587.0) 383.1 (156.6-562.7) 332.9 (79.9-542.2) 514.9 (283.1-583.2) 37.7 (2.6-167.0) 79.0 (0.4-1 11.1) 126.5 (39.2-223.9) 12.2 (0.0-165.0) 0 (0.0-1.0) 0 (0.0-111.1) 16.6 (0.2-245.0) 66.3 (13.5-192.2) 477.2 (183.9-585.1) 486.5 (323.2-586.3) 581.5 (395.7-589.3) 533.4 (430.5-585.5) 13.8 ( 1.2-63.7) 0 (0.0-1 16.5) 33.5 (0.0-181.4) 60.2 (0.0-217.9) 0 (0.0-96.7) 0 (0.0-124.8) 6.4 (0.0-37.9) 0 59.6 (2.0-163.5) 2.9 (0.9-192.4) 23.9 ( 1.3-155.1) 0 (0.0-0.6) 0 (0.0-7.6) 0 (0.0-0.3) 0.1 (0.0-2.8) 0 (0.0-0.7) 0 (0.0-0.4) 52.2 (7.1-209.9) 57.1 (0.0-13 1.4) 58.8 (0.0-2 13.9) 31.7 (0.3-51.0) 1.1 (0.3-7.3) 2.1 (0.0-3.4) (0) 0 (0) ‘FEMALE AGGRESSION’ AND SOCIAL STRUCTURE Female mice were thought to attack conspecifics only when performing parental care (i.e. maternal aggression), a behaviour protecting the young (Ostermeyer, 1983). T h e mother’s attack is uninhibited, with bites to the most vulnerable regions (i.e. head, ventral and inguinal areas, see Fig. 1) of the intruders (Brain, 1981). As a consequence, aggression in lactating mice is termed ‘defensive’ to distinguish it from the ‘offensive’ behaviour of males in which the topography of biting attack rarely involves vulnerable body regions. Nevertheless, recent studies using Swiss albinos suggest that maternal aggression is a heterogeneous phenomenon ranging from offensive to defensive attack. These different patterns of attack are generated when lactating females encounter females in the same condition, sexually naive males or virgin female intruders (Parmigiani et al., 1988a). Sexually naive males are particularly subjected to intense defensive attack whereas virgin and lactating female intruders are attacked offensively. ‘Fear’ is also more characteristic of lactating females responding to strange males (Parmigiani et al., 1988a). Studies with the opiate antagonist naloxone (Parmigiani et al., 198813) and the antiaggressive drug Fluprazine (Parmigiani et al., 1989) also suggest that maternal attack on female and male intruders tap different neural substrates. Indeed, the female’s attack on males may serve a sexual selection function (Parmigiani, 1989). Comparing maternal aggression among the different lines and strains shows that this form of attack usually covaries with the level of intrasexual aggression and infanticide by males (Parmigiani, 1989), suggesting that aggressive behaviour in the two sexes is controlled by the same autosomal genes (St. John & Corning, 1973). Female aggression is not, however, restricted to the lactation period and pup defence. I n fact, females become aggressive towards conspecifics AGGRESSIVENESS IN THE HOUSE MOUSE 263 Figure 1 . Lateral view of a mouse in a defence posture showing the location of the bite targets. The unshaded areas (basically the head and the ventral surface) are the regions animals are inhibited from biting in offensive attack. In defensive attack, animals show a rather random distribution of bites. (especially adult females) when associated with a territorial male. Virgin Swiss albino females individually housed for 24 hours in cages previously inhabited by a male, exhibit increased levels of attack and mounting-like behaviour towards same sex intruders cf. counterparts isolated in cages containing clean sawdust. Consequently, male urinary odour may stimulate female intrasexual aggression. This observation suggests that inter-female attack may be a form of territorial aggression and may also be important in population dynamics. INTRAUTERINE LOCATION PHENOMENA The intrauterine location phenomenon is a further source of variance in aggressiveness. It is a naturally occurring consequence of foetuses being positioned randomly next to others of the same or opposite sex, in mammals that produce multiple offspring, e.g. mice, rats and hamsters. Fifty percent of the offspring of such animals are I M (being located between male and female foetuses), 25% 2M (being located between two males) and 25% O M (being located between two females) (Vom Saal, 1981). Mice at the ends of a row are simply classified with respect to their nearest partner so can only be OM or l M , but it is recognized that such animals occupy an unusual position. Male and female mouse foetuses secrete differing titres of steroid hormones during the prenatal period. Steroids secreted by one foetus seem to be transmitted to contiguous foetuses by unknown mechanisms where they modify development in a predictable manner. Numerous comparisons of male and female mice from different intrauterine positions have revealed radical differences in morphology, physiology and behaviour (reviewed in Vom Saal, 1989). Many comparisons have used only OM and 2M individuals, since, in all studies in which OM individuals have been investigated, they are intermediate between OM and 2M individuals (Vom Saal, 1981). For example, comparison of female mice differing in their in utero proximities to males, reveals varied degrees of masculinization (Meisel & Ward, 1981; Vom Saal & Bronson, 1978). Most studies have concerned the ontogeny of copulatory behaviour. Fewer studies have addressed the role of sex steroid exposure during perinatal life on other activities, such as infanticide, intersexual aggression, postpartum aggression and social interaction (Vom Saal, 1989). 264 P. F. BRAIN AND S. PARMIGIANI 2M female mice are more aggressive towards female intruders than OM counterparts regardless of the hormonal state a t the time of testing. Activation by a specific hormone appears unnecessary to observe such aggression (Vom Saal, 1983a). OM females, in contrast, display more active avoidance in adulthood than 2M counterparts (Hauser & Gandelman, 1983). Following castration at birth, 2M CF1 male mice, exhibited inter-male aggression after 16 days of testosterone treatment, while OM males showed no aggression after this duration of treatment (Vom Saal et al., 1983), indicating that adult 2M males are the more sensitive to the ‘activational effects’ of testosterone on the neural substrate. Studies of gonadally-intact male mice with newborn young showed that most 2M males exhibited parental behaviour, whereas most OM males exhibited infanticide (Vom Saal, 1983b). Obviously, the intrauterine position phenomenon is capable of producing phenotypic variation in the behaviour of male and female rats and mice, as are changes in gene frequency. These changes would not require microevolution. There has been considerable speculation about the impact of different behavioural phenotypes especially in mice with their in naturally-occurring mammal populations cycles of ‘plagues’ and ‘crashes’. ~ EFFECTS OF PRIOR SOCIAL EXPERIENCES ON INDIVIDUAL AGGRESSIVENESS Prior social experiences can also affect the natural propensity for individuals to show aggressive responses. In spite of this, the impact of winning or losing encounters on subsequent performance in laboratory mice has not been systematically studied. Andrade, Kamal & Brain (1989) showed that Swiss mice that had repeatedly defeated conspecifics in daily tests, attack and threaten intruders in subsequent encounters to a greater extent than counterparts without positive fighting experience. The most marked changes in aggressive behaviour occurred between the first and second daily test, there being little further change between the second the fifth test. Latency to first attack significantly declined, whereas numbers of attacks, as well as times allocated to attacking and threatening significantly increased over consecutive tests. The detailed ethological analysis confirmed that other behavioural consequences of repeated experience were evident. Social investigatory behaviour and timid/defensive behaviour decreased as mice spent more time attacking. Although aggression was the focus of the study, these additional alterations may be important for interpreting the effects of experience. Brazin & Al-Maliki (1978) also showed that repeated exposure to male conspecifics is sufficient to elevate offensive behaviour in mice. Such experience is superior to individual housing in producing sustained attack, although isolates attack as readily as males which have had prior experience with intruders in fights. Strange or unfamiliar objects (Barnett, 1975) elicit fear in rats. Avoidance of unfamiliar objects is most pronounced when the object is encountered in a familiar place, such as the home-cage (Cowan & Barnett, 1975). Eventually, however, rodents attend to such items for extended periods, a response obtained even if the object has generated pain on previous occasions. Subjects also frequently sniff and contact such items (Lore, Nikoletseas & Takahashi, 1984). As an intruder is, in a sense, a strange object, one might expect its presence to AGGRESSIVENESS IN THE HOUSE MOUSE 265 elicit ‘fear’ in the resident. Blanchard & Blanchard (1981) suggested that experience with intruders enhances aggressive behaviour by decreasing the resident’s fear of strangers. Blanchard, Shelton & Blanchard ( 1970) confirmed that even brief periods of familiarization with intruders substantially alter future reactions. First experiences of naive residents with intruders are especially important. These initial experiences are characterized by extensive investigation, reminiscent of the interest shown towards inanimate objects. This behaviour increases resident familiarity with ‘strangers’ and presumably decreases associated ‘fear’. Even if the encounter is too brief to lead to attack in this first meeting, residents may, over repeated encounters, become familiar with intruders as a class of object. This use of ‘familiarity’, does not imply recognition of the intruder as an individual but as a Qpe of object. This is a n important distinction because individual recognition has the opposite effect on attack. Familiar conspecifics (animals known to one another) are generally less likely to fight than are strangers (see Kamal & Brain, 1985). Kamal (1986) also observed that, after the initial increase, aggressive behaviour declines over an extended series of consecutive daily tests. The most marked decrement evident was in the tenth test. Winslow & Miczek (1984) also reported that the frequency of biting and lateral threat by male mice toward conspecific intruders declined over a series of ten confrontations. Decrements in murine aggression in repeated confrontations may be a consequence of habituation but fatigue is also important. Declines may also be discussed in terms of the waning of aggressive motivation with repeated experience. Similar phenomena occur in stable groups. Krsiak & Janku (1969) noted that groups of male mice initially housed together show a period of intense fighting, followed by a gradual decline in such behaviour. Three possible causes for such declines in aggression have been advanced. 1. Increasing familiarity of subordinates makes them less efectiue stimuli for eliciting aggression. Unfamiliar opponents are generally attacked more vigorously than familiar counterparts. Kamal (1986) confronted Swiss males with the same intruder each day but there was no decline in the level of aggression from the initial confrontations with these intruders. If intruders became familiar, the aggressive behaviour should have declined and Kamal (1986) suggested that familiarity with an opponent lasts less than 24 hours in this strain of mouse. 2. Behauioural changes in subordinates may result f r o m their learning to avoid or paclfr dominants. Grant & Mackintosh (1963) reported that subordinates responded to attacks by fleeing or assuming defensive upright postures. Burg & Slotnik (1963) suggested that declines in colony aggression occurring soon after groups are established, are a consequence of the improved ability of subordinates to avoid dominant mice. Indeed, Connor & Lynds (1977) reported that subordinate animals quickly learn to avoid and run when approached by dominants. Grimm (1980) also claims that defeated mice learn to assume submissive postures more quickly on successive encounters with fighters, thus reducing the attacks they receive. Strasser & Dixon (1988) suggest that decreased movement is an additional factor reducing probability of attack in rats and mice. Previous agonistic experience of intruders is critical. Lore et al. (1976) placed the same male intruder rat, at weekly intervals, into the cages of two different resident males. O n the first encounter, the resident fought and defeated the intruder but when the now-experienced intruder was placed into the second 266 P. F. BRAIN AND S. PARMIGIANI resident’s cage no serious fighting occurred. They suggested that since the same intruders (but not the same residents) were used in both tests, the experienced intruders modified their subsequent behaviour when confronting a resident animal in a way that reduced attack. Brain et al. (1981) have also reported differences between previously defeated and naive house mice in the defensive strategies they show in response to conspecific attack. Male laboratory mice that have been subjected to defeat, show greater use of the forepaws to push away the partner and receive fewer bites overall. They are also subject to fewer bites on their back (the preferred target site of attackers) as they keep that target away from their attacker. I n Kamal’s (1986) studies, intruders used in consecutive daily trials increased their timidity and defensive behaviour over days. These animals were still vigorously attacked by residents which had experienced winning, but less experienced residents showed lessened attack in response to the pronounced defensive behaviour of these defeated intruders. When animals became highly aggressive, intense attacks may be directed even against intruders which are reintroduced each day. Lagerspetz (1964) reported that induction of an ‘anger’ state in mice, results in attacks being directed even towards motionless opponents or inanimate objects. 3. Temporal decline in aggressiveness may result from motivational changes in the dominant. Kamal’s ( 1986) repeated tests with docile ‘standard opponents’, produced declines in fighting and threat on the tenth day, favouring the explanation that the progressive decline in aggressiveness resulted from motivational changes in these mice with positive fighting experience. Fatigue may also be important for this reduction in aggressiveness but testing at 24-hour intervals seems likely to minimize its relevance. I n cases where animals confronted the same intruder in every trial, residents persisted in attack. The decline in the attack on intruders could not be attributed to increasing familiarity with the opponents or to a change in intruder behaviour with time. Thus, the resident’s changed motivation seems the most likely explanation of Kamal’s results and the phenomenon of ‘social inertia’ in stable groups of mice (one should note, however, that the odour characteristics of subordinates are changed due to suppressed androgen secretion; see Brain, 1983). Andrade et al. (1989) have shown that the effects of exposure to defeat are reversible. Mice repeatedly defeated over several days showed little sign of aggression during the first two days of confrontation with docile anosmic intruders displaying, instead, high levels of defence/timid behaviour. O n the third daily confrontation, however, defeated mice started to threaten and attack their anosmic apponents. Consequently, favourable daily encounters enable defeated mice to recover from their submissive condition within three days of the last defeat. A subsequent experiment used defeated mice with previous experience of victory (unlike defeated mice in the above experiment). After the last of an actually more intense series of defeats, these subjects confronted anosmic intruders. Although they showed no fighting or threat immediately after or up to 6 hours after the last defeat, the majority attacked their anosmic opponents 12 hours after this experience. The recovery from substantial defeat in experienced ‘winners’ thus occurs between 6 and 12 hours from the last negative fighting experience. Positive experience before the defeat thus profoundly changes the dynamics of the recovery from defeat. This characteristic may be of great importance to animals such as the house mouse which may assume a AGGRESSIVENESS IN THE HOUSE MOUSE 267 territorial organization. Subjects must be able to assume the characteristics of a territory holder relatively quickly if an area becomes available. Also, territory holders may be subject to defeat if they stray outside their defended areas. Obviously such animals must be able to tolerate such occasional negative experiences (these are actually likely to be much milder and shorter than the exposures used here) without impairing their ability to hold their territory. The point to emphasize is that it is not solely the outcome of immediately preceding social experiences that determine ‘aggressiveness’ even in the ‘lowly’ mouse. Accumulated positive and negative fighting experiences essentially determined the behavioural characteristics of the individual animal. CONCLUDING COMMENTS There are obviously many potential sources of variation in aggressiveness which underpin the wide range of ‘social’ organizations assumed by the highly opportunistic ‘house mouse’, facilitating its successful exploitation of a wide range of environments. The behavioural characteristics of male and female individuals are influenced by the genotype of the individuals. Rather than this being a simple variation in ‘aggressiveness’, genotypes differ on a range of attributes. I t is probable that any elicited responses are consequences of the resident and the intruder’s behaviours. One should also note that social structure of mouse populations is much influenced by female individuals - that their contribution is not essentially passive. Another source of phenotypic variance is the intrauterine location phenomenon. This can throw up three different ‘types’ of male and females with varied physiological, morphological and behavioural attributes. Finally, one should also record that prior social experiences have considerable impact on the organism’s subsequent aggressive (and other) behaviours which could also affect the range of social structures available to the ‘house mouse’. These data suggest that any relationship between ‘aggressiveness’ and success in feral mouse populations will be a complex one but they may explain why such animals are able to readily vary their social organizations. ACKNOWLEDGEMENTS Data generated by Sandra Jones, Khalid B. H. Kamal and Yousif Y. Yousif is cited in this account. Support by the SERC and the Governments of Italy, Iraq and Saudi Arabia is acknowledged. REFERENCES ANDRADE, M. L., KAMAL, K. B. H. & BRAIN, P. F., 1989. Effects of positive and negative fighting experience on behaviour. In P. F. Brain, D. Mainardi & S. Parmigiani (Eds), House M o u s e Aggression: 223-232. Chur: Harwood Academic Publishers. BARNETT, S. A,, 1975. The Rat, A Study in Behauior. Chicago: University o f Chicago Press. BLANCHARD, R. J. & BLANCHARD, D. C., 1969a. Crouching as an index of fear. Journal ofComparative and Physiological Psycholou, 67: 370-375. BLANCHARD, R. J. & BLANCHARD, D. C., 1969b. Passive and active reactions to fear-eliciting stimuli. J ~ ~ n t oaf l Comparative and Physiological Psychologv, 68: 129-135. BLANCHARD, R. J. & BLANCHARD, D. C., 1981. The organization and modelling of animal aggression. In P. F. Brain & D. Benton (Eds), The Biology of Aggression: 529-561. Alphen aan den Rijn: Sijthoff and Noordhoff b.v. 268 P. F. BRAIN AND S. PARMIGIANI BLANCHARD R. J., SHELTON, V. F. & BLANCHARD, D. C., 1970. Historical effects of stimulus exposure: readiness to eat and object exploration. Learning and Motivation, I : 432-444. BRAIN, P. F., 1981. Differentiating types of attack and defense in rodents. In P. F. Brain & D. Benton (Eds), Multidisciplinary Approaches to Aggression Research: 53-78. Amsterdam: Elsevier/North-Holland Biomedical Press. BRAIN, P. F., 1983. Pituitary-gonadal influences on social aggression. In B. B. Svare (Ed), Hormones and Aggressive Behavior: 3-25. New York: Plenum Press. BRAIN, P. F., 1989. The adaptiveness of house mouse aggression. In P. F. Brain, D. Mainardi & S. Parmigiani (Eds), House Mouse Aggression: 1-22. Chur: Harwood Academic Publishers. BRAIN, P. F. & AL-MALIKI, S., 1978. A comparison of ‘intermale fighting’ in ‘standard opponent’ tests and attack directed towards locusts by ‘TO’ strain mice: Effect of simple experimental manipulations. Animal Behauiour, 26: 723-737. BRAIN, P. F., BENTON, D., CHILDS, G. & PARMIGIANI, S., 1981. The effect of the type of opponent in tests of murine aggression. Behauioural Process, 6: 3 19-327. BRAIN, P. F., McALLISTER, K. J. & WALMSLEY, S. V., 1989. Drug effects on social behaviour: methods in ethopharmacology. In A. A. Boulton, G. B. Baker & A. J. Greenshaw (Eds), Neuromethods, Volume 13: Psychopharmacology: 687-7 139. Clifton, New Jersey: The Humana Press Inc. BURG, R. D. & SLOTNIK, B. M., 1983. Response of colony mice to intruders with different fighting experience. Aggressive Behauior, 9: 49-58. CHILDS, G., 1979. Videotape analysis of social behauiour in laboratov mice. Unpublished M S c . Thesis, University of Wales. CONNOR, J. L. & LYNDS, P. G., 1977. Mouse aggression and the intruder familiarity efkct; evidence for multiple factor determination. Journal qf Comparative and Physiological Pgchology, 91: 270-280. CORTI, M., PARMIGIANI, S., MAINARDI, D., CAPANNA, E. & BRAIN, P. F., 1989. The role of intermate aggression in speciation processes in chromosomal races of house mice. In P. F. Brain, D. Mainardi & S. Parmigiani (Eds), House Mouse Aggression: 49-67. Chur: Harwood Academic Publishers. COWAN, P. E. & BARNETT, S. A,, 1975. The new object and new place reactions of Rattus L. <oological Journal of the Linnean Society, 56: 2 19-234. GRANT, E. C., 1963. An analysis of the social behaviour of the male laboratory rat. Behauiour, 21: 260-281. GRANT, E. C. & MACKINTOSH, J. H., 1963. A comparison of the social postures of some common laboratory rodents. Behauiour, 21: 246-259. GRIMM, V. E., 1980. The role of submissiveness in isolation-induced intermale fighting in mice. International Journal of Neuroscience, 11: 115-120. HAUSER, H. & GANDELMAN, R., 1983. In utero contiguity to males affects level of avoidance responding in adult female mice. Science ( N e w York), 220: 437-438. JONES, S. E., 1983. Videotape analysis of dzfferent forms of “aggression” in various strains o f the laboratory m o u e (Mus musculus 15.).Unpublished Ph.D. Dissertation, University of Wales. JONES, S. E. & BRAIN, P. F., 1987. Performances of inbred and outbred laboratory mice in putative tests of aggression. Behauioral Genetics, 17: 87-96. KAMAL, K. B. H., 1986. Effects of experience andfamiliarity on social interactions of laboratory albino mice (Mus musculus L.). Unpublished Ph.D. Dissertation, University of Wales. KAMAL, K. B. H. & BRAIN, P. F., 1985. Effects of separation quality on subsequent social encounters in male mice. In F. Le Moli (Ed.), Multidisciplinary Approaches to Conflict and Appeasement in Animals and Man: 160. Parma: University of Parma Press. KRSIAK, M. & JANKU, I., 1969. The development of aggressive behaviour in mice by isolation. In S. Garattini & E. B. Sigg (Eds), Aggressive Behauiour: 101-105. Amsterdam: Excerpta Medica Foundation. LAGERSPETZ, K. M. J., 1964. Studies on the aggressive behaviour of mice. Annals of the Finnish Academy of Science, Series B, 131: 1-131. LORE, R., FLANNELLY, K . & FARINA, P., 1976. Ultrasounds produced by rats accompany decreases in intraspecific fighting. Aggressive Behavior, 2: 175-181. LORE, R., NIKOLETSEAS, M. & TAKAHASHI, L., 1984. Aggression in laboratory rats: a review and some recommendations. Aggressive Behavior, 10: 59-7 1. MAINARDI, D., PARMIGIANI, s.,JONES, s. E., BRAIN, P. F., CAPANNA, E. & CORTI, M., 1986. Social conflict and chromosomal races of feral male house mice: An assessment combining laboratory and field investigations. Accademia Narionale dei Lzncei, 259: 11 1-139. MEISEL, R. L. & WARD, I., 1981. Female rats are masculinized by male littermates located caudally in the uterus. Science (New York), 213: 239-242. OSTERMEYER, M. C., 1983. Maternal aggression. In R. W. Elwood (Ed.), Parental Behauiour in Rodents. London: J. Wiley and Sons. PARMIGIANI, S., 1989. Maternal aggression and infanticide in the house mouse: consequences on the social dynamics. In P. F. Brain, D. Mainardi & S. Parmigiani (Eds), House Mouse Aggression: 161-178. Chur: Harwood Academic Publishers. PARMIGIANI, S., BRAIN, P. F., MAINARDI, D. & BRUNONI, V., 1988a. Different patterns of biting attack employed by lactating female mice ( M u domesticus) in encounters with male and female conspecific intruders. Journal of Comparative Psychology, 102: 287-293. AGGRESSIVENESS IN T H E HOUSE MOUSE 269 PARMIGIANI, S., RODGERS, R. J , , PALANZA, P. & MAINARDI, M., 1988b. Naloxone differentially alters parental aggression by female mice towards conspecific intruders of differing sex. Aggressive Behavior, 14: 213-224. PARMIGIANI, S., RODGERS, R. J., PALANZA, P., MAINARDI, M. & BRAIN, P. F., 1989. The inhibitory effects of Fluprazine on parental aggression in female mice are dependent upon intruder sex. Physiology and Behavior, 46: 455-459. ST. JOHN, R. D. & CORNING, P. A., 1973. Maternal aggression in mice. Behavioral Biology, 9: 635-639. STRASSER, S. & DIXON, K., 1987. The role of senses in house mouse social aggression. Paper presented at Conference on House Mouse Aggression: A Model for Understanding the Evolution of Social Behauiour. Erice, Italy (10th-15th September). VAN OORTMERSSEN, G. A,, 197 1. Biological significance, genetics and evolutionary origin of variability in behaviour within and between inbred strains of mice ( M u s musculus). Behaviour, 36; 1-92. VAN OORTMERSSEN, G. A. & BUSSER, J., 1989. Studies in wild house mice. Disruptive selection of aggression as a possible force in evolution. In P. F. Brain, M. Mainardi & S. Parmigiani (Eds),House Mouse Aggression: 87-1 17. Chur: Hanvood Academic Publishers. VOM SAAL, F. S., 1981. Variation in phenotype due to random intrauterine positioning of male and female fetuses in rodents. Journal of Reproduction and FertiliQ, 62: 633-650. VOM SAAL, F. S., 1983a. Models of early hormonal effects of intrasex aggression in mice. In B. B. Svare (Ed.), Hormones and Aggressive Behavior: 197-222, New York: Plenum Press. VOM SAAL, F. S., 1983b. The interaction of circulating oestrogens and androgens in regulating mammalian sexual differentiation. In J. Balthazart, E. Prove. & R. Giles (Eds), Hormones and Behaviour in Higher Vertebrates: 159-1 7 7 . Berlin: Springer-Verlag. VOM SAAL, F.S., 1989. Perinatal testosterone exposure has opposite effects on adult intermale aggression and infanticide in mice. In P. F. Brain, D. Mainardi & S. Parmigiani (Eds), House Mouse Aggression: 179-204. Chur: Harwood Academic Publishers. VOM SAAL, F. S. & BRONSON, F. H., 1978. In ulero proximity of female mouse fetuses to males: Effect of reproductive performance during later life. Biology of Reproduclion, 19: 842-853. VOM SAAL, F. S., GRANT, W., McMULLEN, C. & LAVES, K., 1983. High fetal estrogen titers correlate with enhanced sexual performance and decreased aggression in male mice. Science ( N e w York), 220: 1306-1 309. WINSLOW, J. T. & MICZEK, K. A,, 1984. Habituation of aggressive behavior in mice: a parametric study. Aggressive Behavior, 10: 103-1 13.
© Copyright 2026 Paperzz