Aggression is a fascinating and distressing topic for the pet rat owner. We are intrigued or amused by aggressive behaviors like chasing and sidling, but we hate it when our pets hurt each other. Males who played harmlessly together for months may start to harass or bite each other around six months of age. Peaceful rats may attack a newcomer, making introductions between old and new rats difficult or impossible. Tame female rats may become aggressive while their litters are young.
The field of rat aggression is enormous, with thousands of publications and a specialized journal (Aggressive Behavior). I cannot hope to summarize all the literature here. My intent is to present the information on rat aggression that is most relevant to (a) the owners of pet rats, and (b) people interested in rat behavior under natural or semi-natural conditions.
Agonistic behavior refers to the complex of aggression, appeasement and avoidance behavior that occurs between members of the same species. Agonistic behavior is a much broader term than "aggression," which refers to behavior patterns which serve to intimidate or damage another (for more, see McFarland, 1982)
Agonistic behavior behavior involves several actions, or motor patterns, including chasing, sidling, boxing, biting, and kicking, as well as audible and ultrasonic vocalizations. Agonistic behavior can occur between rats in a colony, and between resident rats and intruders.
Aggressive behavior: sequences and strategies
Agonistic behaviors are strung together into sequences that vary
in their intensity and duration. The lowest intensity encouters are
chases. As the intensity increases, you may see stand-offs and
physical contact like boxing and sidling. These physical encounters
may escalate in rare instances into a fight.
Aggressive neck grooming is a low-intensity form of agonistic behavior. Grooming consists of rapid little nibbles in which the groomer seizes folds of neck skin between his teeth (Miczek and Boer 2004). The groomed rat remains immoble and may peep or squeak softly. Any sudden movement by the groomed rat may trigger a bite and kick from the groomer.
A low intensity agonistic interaction is a chase. Many agonistic interactions consist of one rat chasing another for a few seconds. The chased rat flees and either outruns or hides from the pursuer, or sometimes the pursuer desists. If the pursuer catches up, he may nip or bite the rump of the fleeing rat and may attempt to engage him in an encounter. Sometimes, the pursuer may mount the fleeing rat.
If the chased rat holds his ground, chaser and chasee may have an encounter. A pursued rat may turn to face the pursuer, initiating a nose-off. The defending rat may show an open-mouth tooth display, with long squeaks and sometimes hisses. Sometimes the tails of both rats writhe on the ground, and both rats' fur may be piloerect. Frequently the encounter goes no further, and one of the rats (usually the pursued one) flees.
If neither rat flees, the encounter may escalate with physical contact. The rats may box and/or sidle, and if the sidling rat gets very close he may push or kick out with a hind foot. The subordinate rat, especially if he is young or the rats are confined in a cage, may roll over into a belly-up roll. I've found among my rats that most of the few encounters that get this far end here, with the flight of the pursued rat. After the flight another sequence may occur, and on a few occasions combinations of chases, sidling, and flight have lasted for tens of minutes or even an hour or more.
A variation on the sidle sequence is seen on ledges such as shelves and hammocks. On ledges, the sidling rat crowds the other toward the edge. This may be a very slow process, with the sidling rat maneuvering a fraction of an inch at a time, with long pauses when nobody moves. If the second rat does not jump off the ledge himself, he will be pushed off. Sometimes the pursuer will follow the second rat around the cage, pushing him off ledge after ledge. If the second rat resists, the encounter may escalate.
Rarely, physical contact may escalate into a fight. On a handful of occasions the encounter escalates beyond the sidle-and-kick stage, and the two rats close into a fighting ball, in which the attacking rat may bite the other's flank or rump. These fights are usually brief (lasting just a few seconds) and end with one rat fleeing, hiding and staying staying quiet and subdued for a while -- an hour or more.
Offensive and defensive tactics
The immediate goal of serious fighting is the attack and defense of the rump and lower flanks. The goal of the aggressor is to inflict a bite on the rump, and the goal of the defender is to prevent such a bite (Blanchard et al. 1975, 1977, Blanchard & Blanchard 1977, Pellis & Pellis 1987, Takahashi and Blanchard 1982).
In groups of three rats, subordinates tend to employ one of two strategies in their relationship with the dominant rat. Subordinates either avoid the dominant rat or they stay close to him. "Avoiders" are called omega subordinates, and the "close-in" rats are called beta subordinates (Barnett 1975 p. 125).
Omega avoiders are more severely pummeled by the dominant rat when encountered (Blanchard & Blanchard 1990), and if free to do so omegas are more likely to emigrate from the colony (Barnett 1975 p. 125). Omegas who cannot escape may be so severely attacked and harassed that they die (Adams and Boice 1983). However, in some cases, if the dominant rat is removed, an omega subordinate may be more likely than a beta to rise to become the new dominant rat.
In contrast to omegas, betas become fully submissive toward the dominant rat, who in turn tolerates their presence more (Pellis et al. 1993).
So, the two strategies have different costs and benefits. Beta subordinates live in relative peace with the dominant rat and tend to gain more access to food and females. Omegas, on the other hand, tend to be the target of more aggression from the dominant rat. The omega and the alpha may maintain an uneasy truce. But the omega may be in a better position to take over the dominant's role if it becomes available (Pellis et al. 1993).
What causes a rat to adopt one strategy or the other? Rats vary in temperamental factors that influence aggressiveness and boldness, which may have a genetic component. In colonies of rats that differ in boldness, the boldest rat may become dominant, intermediates may adopt the omega strategy, and the least bold may become betas (Pellis et al 1993). In any case, dominance is usually not correlated with body size (e.g. Stewart and Palfai 1967).
How common is it?
Agonistic behavior is actually not very common between rats in an established colony. Blanchard et al. (1988) studied aggression in six mixed-sex colonies of Norway rats over their entire lifespans (the colonies consisted of 3 males and 3 females, all unrelated, and placed together at age 3 months). The rats in his study spent less than 1% of their time in agonistic encounters. The most aggressive animal in the Blanchard study spent less than 2 minutes per hour in aggressive interactions.
Alberts and Galef (1973) established colonies made of pairs of wild-stock brothers. If one of the males was removed for 24 hours, then replaced in the home cage, the remaining brother was not aggressive toward the returnee. Instead, both rats simply began feeding.
Males are much more likely to engage in agonistic behavior (both offensive and defensive) than females. In Blanchard et al.'s study (1988), males accounted for 93.2% of the time spent performing agonistic behavior. Males tend to develop dominance hierarchies, which are less common in females.
In mixed-sex colonies, resident males are far more likely to attack intruders than resident females. However, this low level of female aggression toward intruders may be due to the presence of males. Females living in all-female groups do attack intruders (Blanchard et al. 1984), and as with males, one female tends to account for most of the attacks.
Biting: Males rarely bit females in the Blanchard et al. (1988) study. Only 5 out of 18 females were bitten by males (1-3 bites each), and these bites occured only during the first month after the colonies were formed. Female-female biting was nearly as rare (<1 bite per female), and also occured only during the month after the rats were first introduced. In contrast, male-male biting was much more frequent (0.6-3.2 bites per male) and continued throughout the lifespan of the colony.
Age and aggression within a colony
In general, young male rats play fight together from age 5 weeks to around age 5-6 months. Play fighting isn't serious, nobody consistently wins or loses, and nobody gets injured. As the young males approach social maturity at around 5-6 months of age, however, their fighting becomes serious, with chasing, sidling, and biting, and one rat emerges as a consistent winner. In this way a stable, long-term dominance hierarchy emerges between the males. The dominant male also tends to be the most active and to copulate most frequently with the females. Females, for their part, tend to behave less aggressively toward each other, but a female dominance hierarchy may emerge while they are pregnant or lactating.
Here are two studies of rats in a semi-natural setting that examine dominance in mixed-sex colonies of rats over the rats' entire lifespans:
Study 1: Adams and Boice (1983) established a mixed-sex colony of 4 male and 4 female 45 day old albino rats in a large, outdoor semi-natural enclosure and recorded the dominance relationships among the rats and their offspring every day for 15 months.
Young rats play fight, but their play fighting doesn't predict adult dominance: They observed that young male rats engaged in lots of play fighting: they pinned each other, wrestled, and boxed. There was no sidling, chasing, or biting. None of the rats were injured. Winners and losers were frequently reversed, so that no rat was consistently dominant. These play-wins and play-losses did not predict the adult male hierarchy at all. Females didn't interact with each other as much, but when they did they also play fought.
At social maturity, fighting turns serious, and a stable male dominance hierarchy emerges: When the rats approached social maturity at around age 5-6 months, adult aggressive behavior made its appearance: sidling, chasing, and biting. A stable social hierarchy emerged, with one male rat consistently winning contests over the others. This male became the dominant rat, the other three became subordinates. Two were socially active "betas," while one was a socially isolated, viciously attacked "omega." The omega lost weight, acquired scars, and eventually died at a little over 6 months of age.
Interestingly, five weeks after the hierarchy was established, the dominant male became ill and stayed in his burrow for a week. During his absence, two of the subordinates started confronting each other and one became dominant over the other. When the original alpha male got better and emerged from his burrow, the new alpha male severely attacked him, leading to the original alpha's death the following day. This new alpha remained dominant in a stable hierarchy for the rest of the study. This male copulated with females twice as often as any other male.
Response to newcomers: When new males were introduced, the resident males, especially the resident dominant male, attacked them. The newcomers' activity was greatly suppressed for several months and they spent most of their time underground. When new females were introduced, one resident female attacked them persistently, but these attacks waned after a few days and the new females were eventually accepted.
Female dominance: Females dominance wasn't as salient as that of males. A female dominance hierarchy only appeared when females were pregnant or lactating. Females circled each other, showed piloerection, tail swishing, boxing, and tooth chattering. Generally, these conflicts were resolved when one female dropped to a sidling-biting posture. One female consistently won most of these contests. She was also the most socially active female, the most active against male or female intruders. This dominant female was the most attractive to males, she was the first to raise a litter to weaning age, and two of her sons became the dominant males in the next generation.
The next generation: The appearance of the juveniles at age 22-30 days of age didn't change the social hierarchy of the adult males at all. The juveniles tended to dart from one burrow to the next, or to follow adults to food sources. The juvenile males started play fighting at around 40 days of age: they pinned and pounced, wrestled and boxed. As with the previous generation, winners and losers were reversed frequently and they didn't develop any stable dominance hierarchy. Two young males, however, were particularly socially active: they investigated females and copulated with them a great deal. These two young males grew up to be dominant, even though they weren't the biggest play fighters.
When the second generation approached social maturity at around 5 - 6 months of age they established a new dominance hierarchy within a few days. Two of the young males started sidling, chasing, biting, and winning. They fought with any young males that crossed their paths. Only two adult rats of the previous generation were still alive -- the old alpha male and the old alpha female -- but the new young dominant males didn't interact with the old alpha male at all. Over time, one of the young dominant males established himself as dominant over the other.
Study 2: Blanchard et al. 1988. In a similar experiment, Blanchard et al. (1988) studied aggression in mixed-sex colonies of Norway rats over their entire lifespans. The authors formed six colonies, consisting of 3 males and 3 females each, aged 95-115 days old.
During the first month after the colonies were created, aggression levels were relatively high in four of the colonies. However, the offense-defense relationships during that month didn't predict the social hierarchy that developed later. After a month, male-female and female-female aggression decreased, but male-male aggression continued.
These four colonies showed stable levels of offense and defense over the colony lifespan. They developed clear, stable dominance hierarchies between the males, which lasted more than a year. The fifth colony did not develop a clear dominance hierarchy right away: it took many months of moderate levels of aggression before one male to emerged as dominant.
In the sixth colony, the three males lived peacefully together and exhibited little aggression toward each other at all ages and developed no clear dominance relationships. When a newcomer was introduced, however, all three males attacked it intensely.
In contrast, female-female and male-female aggression declined after the first month and remained relatively infrequent, while male-male aggression remained high in comparison.
Established rat colonies tend to attack intruders, and can do so with great violence. However, aggression toward intruders is not automatic, and the intensity of an attack is variable. The expression of aggression is quite responsive to differences in development, experience, and context. Aggression depends on the sex, age, experience and hormonal status of the resident and the intruder, and the presence of females and juveniles (for a review, see Lore et al. 1984).
Who attacks intruders?
In mixed-sex colonies, male residents -- especially the socially dominant male -- attack intruders more than females and subordinates do (Blanchard et al. 1977b, 1988). For example, one study found that 90% of the bites inflicted on an intruder were made by the resident alpha male (Takahashi and Blanchard 1982).
There are exceptions, however: Blanchard et al. (1988) observed a colony in which the dominant and a subordinate male shared the attack on an intruder equally, and another colony in which a group of three males who got along well and appeared to have no dominance hierarchy between them intensely attacked an intruder all together.
Resident rats tend to attack intruders of the same sex: males attack intruder males, while females primarily attack females (DeBold and Miczek 1984). A male rat confronted with a strange, intact female tends to investigate her, sniff her, and mount her. If the female is spayed before introduction, the male's behavior is similar, but he investigates, sniffs, and mounts her somewhat less than intact females. Aggressive behavior toward any of these females is brief. However, if a male is confronted with an intact male, he investigates and sniffs the newcomer only briefly, attacks within a few minutes, spends a lot of time behaving aggressively, and rarely mounts him (Flannelly and Lore, 1977).
Resident females in mixed-sex colonies tend not to attack intruders. However, females housed in all-female colonies may be as aggressive toward intruders as males in all-male colonies (McDonald 1971). Therefore, it appears that in mixed-sex colonies, male aggression is enhanced while female aggression is reduced. In the absence of males, females may be more aggressive toward strange rats.
Description of resident aggression toward an intruder
Duration of the encounter
Aggression toward intruders changes over time. For example, resident colonies may "warm up" gradually, working themselves up into an attack on an intruder. In this case, fighting may not occur during the first few minutes but may break out over the following hours. In other cases, aggression may be intense at first but may decline over time.
• Short exposure (6-10 minutes)
Blanchard et al. (1975) created six mixed-sex colonies of rats (3 males and 4 females aged 85-90 days). After three months, the authors tested the colony's response to an intruder by removing one subordinate male and replacing him with a 133 day old male intruder. Both dominant and subordinate resident males approached and sniffed the intruder, became piloerect, sidled toward the intruder, boxed with him, bit him, and chased him around the cage. However, the dominant rat spent much more time attacking the intruder than the subordinate did. After 10 minutes in the residents' cage the intruders had received an average of 15 bites each, demonstrating that these attacks were serious. Most of the bites were to the back.
In a similar experiment, Takahashi and Blanchard (1982) established seven colonies of domestic and seven colonies of wild-trapped Norway rats consisting of 3 adult males and 3 adult females each. Pups born were moved at age 30 days. After 15 weeks, these colonies were tested with an intruder for 6 minutes. The dominant male attacked the intruder about one minute after introduction, and chased, sidled, bit, and stood on top of the intruder. The dominant male inflicted 90% of the bites on intruders, and 80% of these bites were to the rump. Almost all the remaining bites were to the limbs or tail. Bites to the head or belly were extremely rare (less than 1 in 40).
Anesthetized intruder: If an intruder was anesthetized before being placed, sleeping, in the cage, the residents still attacked and bit him, though less than when the intruder was awake. Sleeping intruders received about 5 bites each (Blanchard et al. 1975). Residents did not show sidling, boxing, or lunging toward a sleeping intruder. They sometimes sniffed and licked the intruder's mouth and head before and after biting. Bites were made directly without any preliminaries. Most bites were delivered by the alpha male, but females and subordinate males delivered a few bites as well. Males tended to bite the rump, while females tended to bite the head. Some rats pulled and combed the intruder's fur wtih their teeth. In a few cases, the residents briskly buried the sleeping intruders under sawdust (Takahashi and Blanchard 1982).
• Medium exposure (1 day):
Luciano and Lore (1977) placed intruders in a colony consisting of two males and three females and left the intruders there for 21 hours. This procedure was repeated every two days. Resident aggression toward the intruders was intense: each intruder received an average of 50.9 small, 12.6 medium, and 4.6 large body wounds. One intruder was killed within 2 hours. Aggression toward subsequent intruders increased over time as well. Upon removal, the intruders were found to have gastrointestinal pathology: ulcers, gastric lesions, thinning of the stomach lining.
Luciano and Lore (1975) examined the role of the intruder's social experience on its reception on a colony by placing isolation-reared or socially-reared intruders in the residents' cage for 21 hours. During the first hour, all intruders were attacked with similar degrees of aggression. After 21 hours, however, isolation-reared intruders had received many more bites than socially-reared intruders. This demonstrates that while aggression was similar in the first hour, aggression increased in severity during the following 21 hours toward some types of intruders. Such attacks took more than an hour to develop.
In a similar experiment, Adams and Boice (1983) established a mixed-sex colony of rats in a semi-natural outdoor enclosure. They introduced male intruders to the colony one at a time, for 24 hours over many weeks, and found comparable results: all resident males were aggressive toward the intruder, but the socially dominant male was by far the most aggressive. The residents (usually the socially dominant male) approached, sniffed and aggressively groomed the intruder, then chased and bit him. Aggression was severe: three out of 14 intruders died before 24 hours, and the remaining intruders lost weight and received an average of 4.8 large wounds each.
• Long exposure (1 week)
Blanchard et al. (1975) placed one intruder in each of his six mixed sex colonies and left it there for 7 days (colonies were formed of 3 males and 4 females aged 85-90 days and were tested 3 months later). The attacks proved to be lethal. Two of the intruders were dead after 28 hours, a third died after 32 hours, a fourth after 36 hours. The remaining two rats were alive after 7 days but were badly wounded. Each of the six intruders had received an average of 48 bites (range: 37-63 bites), most of which were to the back. During the same period of time, none of the resident rats were wounded. In a control procedure, one of the subordinate males was removed and replaced after 3 hours. He received no wounds (Blanchard et al. 1975).
• Permanent exposure
Adams and Boice (1983) introduced two new males to an established mixed-sex colony of rats in a semi-natural outdoor enclosure. Both newcomers were attacked by the residents, especially by the dominant male. Over the ensuing six weeks, the activity of the newcomers was severely repressed: they spent 90% of their time hiding underground (the dominant male spent only 30% of his time underground). The newcomers had extremely low levels of social contact, copulations with females, or wins against the residents. From six weeks to four months after the introduction, the newcomers appeared to be more accepted into the group: they spent more time above ground, achieved a few copulations and engaged in some social contact, though they engaged in these activities less than the other long-term male resident, and far less than the dominant male.
Attacking rats prefer to bite intruders on the rump (Blanchard et al. 1977, Takahashi and Blanchard 1982). If the intruder is anesthetized residents prefer to bite the head, but if thwarted they prefer to bite on the back. In fact, if a sleeping intruder is placed on its back with its head protected, the resident rats tend not to bite the belly but direct their attacks at the hindquarters, or burrow under the animal to bite the rump (Blanchard et al. 1977).
Very little has been written describing the aggression of female residents toward intruders. DeBold and Miczek (1984), and Takahashi and Blanchard (1982) note that females tend to attack with a nipping or striking bite toward the face. Because bites to the face are characteristic of defensive rats, females may may be behaving defensively rather than offensively toward intruder rats.
Inter-resident hostility in the presence of an intruder
Sometimes, the presence of an intruder leads to a breakdown in friendly interactions between residents. Residents may sniff each other more, and while chasing the intruder may bump into each other and assume a boxing stance or lateral postures toward each other. These confrontations are usually brief and end amicably after mutual sniffing (Alberts and Galef 1973).
Do wild rats attack intruders?
Wild rats attack introduced intruders. However, these attacks are not sustained and the intruder is rarely seriously injured because he runs away. Once he runs far enough away, pursuit stops, and he is ignored (Telle 1966). Aggression toward intruders therefore seems to function to exclude intruders from resources like food, shelter, and sexual access to females available to residents, and to chase away surplus animals into adjacent areas. In captivity, intruders cannot run away, so conflicts tend to escalate into serious and sometimes lethal fights (more on wild rat aggression).
What do the intruders do?
In general, intruders do not attack resident rats. They spend most of their time defending themselves by boxing, laying on the back, and fleeing. They spend very little time being aggressive toward residents (Blanchard et al. 1975, Christie and Barfield 1979, Takahashi and Blanchard 1982). Intact male intruders tend to show a little more aggression when they meet a castrated resident (Christie and Barfield 1979).
Rats defend their heads well, so intruders typically focus on defending the next major target: their backs. Laying on the back and boxing appear to be attempts to hide the back from attack. Both behaviors interpose the rat's body between the attacker and the defender's rump. In boxing, the defending rat's whiskers and face get in the way of the attacking rat. As long as the defender can maintain the whisker-to-whisker contact seen during boxing, he is never bitten (Takahashi and Blanchard 1982). If the defender's whiskers are anesthetized the attacker inflicts bites to the face. The lateral display, or sidle, is used by the attacking rat to get around the intruder's boxing and deliver a bite to the rump (Blanchard et al. 1977, Pellis and Pellis 1987).
When the defending rat lies on his back his belly is in the way of a rump-bite. The dominant rat presses the attack by laying on top of the defender at right angles to his body, thus avoiding the defender's whiskers, and trying to bite the defender's sides (Blanchard et al. 1977). The attacker was frequently successful: the on-the-back posture did not prevent bites. This indicates that the intruder's on-the-back posture is not an attack-inhibiting "submission" signal, but is rather a form of self-defense in which the defender tries to keep his back from being bitten.
Despite all the defending rat's efforts, however, attacks by dominant colony rats often succeed and are frequently lethal (Blanchard et al. 1975). The intruder's defensive behavior may be an attempt to delay or thwart the biting attack until flight becomes possible. In a cage, however, turning and running were extremely dangerous: most bites were inflicted when the defender turned and fled (65% of bites; Blanchard et al. 1977). In a more open setting the defending rat could flee. Encounters in the open might not escalate as far as they do in a confined space. Indeed, wild rats fights are brief, on-the-back postures are never seen, and most fights end in the flight of the loser (Robitaille and Bovet 1976). (More on wild rat behavior)
Being an intruder is stressful: attacked intruders develop ulcers within hours of introduction
Luciano and Lore (1977) placed intruders every other day into a colony consisting of two males and three females and left the intruders there for 21 hours. Resident aggression toward the intruders was intense: each intruder received an average of 50.9 small, 12.6 medium, and 4.6 large body wounds. One intruder was killed within 2 hours.
After 21 hours, each intruder was removed and its stomach and intestines were examined and were found to have a wide variety of stress-induced gastrointestinal pathologies. All of the intruders had developed between 1 and 4 stomach ulcers each. Intruders also displayed hypersensitive, reddened, inflammed stomach linings (diffuse erythema and gastritis), and prominent folds (ruggae) in the stomach. The intruders had transparent, thin intestinal walls consisting of a series of distended pockets instead of normal opaque, thick walls and a smooth, regular surface. The intruders had no recently ingested food or fecal matter, mucosal lining, or villi, indicating that they did not feed in the residents' cage. All of the intruders were in poor condition when removed. In comparison, 10 control animals had stomachs filled with food, with normal stomach and intestinal linings.
In conclusion, intruder rats develop stress-induced gastrointestinal pathology soon after being attacked by other rats.
The role of odor cues
How do residents tell familiar and unfamiliar rats apart? Odor cues play an important role. Alberts and Galef (1973) and Flannelly and Thor (1976) reduced resident rats' sense of smell by applying a solution of zinc sulfate inside the nose (Alberts and Galef 1971, Thor et al. 1976, but see Slotnick et al. 2000). Resident rats with an impaired sense of smell tended not to attack intruders in their cages. Anosmic residents also approached and investigated intruders far less than intact ones. The decline in resident aggression may be due to the elimination of odor cues, but it could also be due to a general decrease in activity following the loss of a vital sensory system, or to systemic poisoning from the zinc sulfate (Sieck and Baumbach 1974).
However, disabling the intruder's sense of smell has no effect on the resident's aggression toward the intruder: residents attack intact and anosomic intruders similarly (Flannelly and Thor 1976).
Do colonies have a "colony odor"?
One hypothesis about how residents differentiate familiar from unfamiliar rats is that a colony has a colony-specific odor, acquired from the outside through mutual contact with the urine and feces of colony members. If this is true, then any rat that carries the colony odor would be tolerated, and any rat that does not carry that odor, or carries a strange odor, would be attacked. Alberts and Galef (1973) contucted a pair of experiments to test these two predictions.
Odor swapping tests
Alberts and Galef (1973) scrubbed strange rats with soap and rinsed them with alcohol and water to reduce their own odor, then placed each rat for a week in a box containing the urine and feces-covered shavings from another rat colony. The day before testing, each intruder was doused in fresh urine collected from the colony. The intruder was then placed into the colony in whose shavings they had been kept. In a second but similar study, the authors took a subordinate male rat from a colony and followed the same procedure, using the urine and feces of a strange colony, and re-introduced the familiar rat to his cagemates.
All this odor swapping had no effect on how the newcomers were treated by the residents. Regardless of the odor treatment, strange rats were treated aggressively and familiar rats were treated amicably. Specifically, housing a strange rat with familiar-smelling urine and feces, then dousing him in familiar urine, didn't fool the residents. They still attacked him. Similarly, housing a familiar rat with the urine and feces of unfamiliar rats, then dousing him in unfamiliar urine, didn't fool the residents either. They recognized their cagemate anyway and responded to him amicably.
Conclusion: rat colonies do not have a "colony odor"
Therefore, rat colonies do not have a shared urine/feces-based colony-specific odor. Rats do not gain acceptance in a colony by picking up external olfactory cues acquired by contact with the feces or urine of colony members. Instead, rats appear to react to other rats based on their individual odor.
The physical environment
• The effect of the home cage
Rats tend to be more aggressive toward intruders in their home cage than two strange rats are toward each other in a neutral area. When an intruder is introduced to a resident's home cage, the resident almost always emerges dominant. When two rats are introduced in a neutral area, however, dominance relationships are unpredictable (Barfield et al. 1972). Thor (1976) notes that changing the location of the home cage, or placing the resident rats in a new cage just prior to the introduction, may also reduce a colony's attack on a new rat, but these conditions have not yet been empirically tested.
• Presence of an escape chamber
The provision of an escape chamber within the resident colony's cage -- a small box with a narrow entrance -- reduces fighting between residents and an intruder. If such an escape chamber is present, the intruder tends to run to the chamber and defend himself from inside. Intruders who were given an escape chamber fought less, received fewer wounds, and were less likely to be killed than intruders without such a refuge. An escape chamber therefore provides the intruder with the option of avoiding serious injury, especially during prolonged exposure to resident rats (Takahashi et al. 1980).
Design of the escape chamber: Takahashi et al (1980) provided two types of escape chamber. The Vertical Escape Chamber was a 10 cm wide x 10 cm deep x 20 cm tall hardware cloth enclosure. The 10 x 10 top was open; access could only be gained by climbing the 20 cm walls and entering through the top. The Horizontal Escape Chamber was a Plexiglas box measuring 28.4 cm x 12.7 cm x 11.4 cm. At one end was a single 5.2 cm diameter hole.
Nikoletseas and Lore (1981) took 30 day old rats and housed them individually in either burrow-digging boxes or cages of similar size. At age 92-95 days of age, an intruder of the same sex and age was placed in the resident's cage and left there for 24 hours. The researchers observed the rats for 20 minutes, then after 24 hours removed both animals and weighed them, counted wounds, and looked at whether their stomachs were empty or full. During the 20 minute observation period the authors found no difference in duration of social interaction, latency to aggression, and duration of aggression. However, burrow-housed rats exhibited significantly less aggression than those housed in cages during those 20 minutes, possibly because intruders in the burrow-digging environment were able to hide.
After 24 hours, however, the picture is quite different between the two environments: intruders lost more body weight and more of them had empty stomachs when exposed to burrow-digging residents. Some intruders were able to use the more complex environment of the burrow-system to escape attack: they squeezed themselves into the water compartment, sought refuge in the burrows and defended the tunnel entrance or plugged the entrance with dirt.
This study indicates that burrow-housed rats tend to be much more aggressive toward intruders over 24 hours than cage-housed rats, despite the fact that the burrow environment is more complex and provides more opportunities for intruders to hide from the resident.
Blanchard et al. (1988) found that the aggression of resident males depended on their age. When resident males of mixed-sex groups were under six or seven months of age, they showed relatively little aggression toward the intruders. Resident male aggression toward intruders increased from age 6.7 to 20 months. The resident males attacked intruders more and more quickly. Female aggression toward intruders did not increase like this.
Takahashi and Lore (1983) also found high levels of aggression in male rats that were 6.7 months old. The authors formed colonies of rats consisting of 2 males and 1 female, when the animals were either 100 (3.3 months), 200 (6.7 months), or 300 days old (10 months). Prior to colony formation, all the animals had been reared in small, same-sex groups. The authors found that the 200 day old males were significantly more aggressive than the younger or older males: they fought the most with their new colony mates, and three weeks later were more aggressive toward unfamiliar intruders than either the 100 or 300 day old males.
Blanchard et al. (1988) also founcd that the aggression of resident males toward intruders declined somewhat with age. However, older males who were still inclined to fight with intruders did so readily, with the same intensity and duration as when they were young, though with less coordination. The few males who survived to 26 months continued to attack intruders.
Effect of social experience on aggression
• Social experience of residents and intruders
Luciano and Lore (1975) compared the response of socially-experienced and isolation-reared colonies of rats to intruders. The authors created these two types of colonies by combining 2 male and 2 female rats of differing social experience. The "isolation reared" colonies were composed of rats that had been housed in isolation after weaning. The "socially experienced" colonies were made of rats that had been housed with other rats after weaning. At colony formation, all socially-experienced rats were strangers to each other. These colonies were presented with unfamiliar socially-experienced or isolation-reared intruders.
The authors found that aggressive behavior (such as aggressive grooming), serious physical injury, stomach ulcers, and intruder weight loss occured only when an isolation-reared intruder was placed in a socially-reared colony. Socially reared rats are violently intolerant of a rat that lacks similar social experience (see also Flannelly and Thor 1978)
None of the other combinations -- socially-experienced colony and socially-experienced intruder, isolation-reared colony with socially- or isolation-reared intruder -- produced such severe aggression. Socially-experienced colonies displayed almot no aggression toward socially-experienced intruders, while isolation-reared colonies exhibited low levels of aggression toward both types of intruders.
Isolation-reared intruders may have failed to respond appropriately to the vocal and postural signals of socially-experienced residents, or may have failed to produce the appropriate defensive or appeasement signals to deflect attack.
• Prior experience of intruders with residents, and of residents with intruders
Intruders also learn from experience: intruders that experience a defeat by one resident are less likely to elicit aggression during a second encounter with a different resident (Lore et al. 1976).
Colonies with prior experience with intruders are more aggressive toward intruders than intruder-naive colonies of equivalently aged rats. Blanchard et al. (1977b) formed 16 colonies of rats (3 males, 3 females) when the rats were an average of 131 days old. Half of the colonies were presented with a different intruder for 10 minutes every week for 10 weeks. The other half were not presented with any intruders. After 10 weeks (when the rats in the colonies were 7 months old), all colonies were presented with an intruder. Blanchard et al. found that males from the intruder-experienced colonies become piloerect faster, boxed more with the new intruder, and bit the intruder more than the males from the intruder-naive colonies (Blanchard et al. 1977b).
Effect of other rats on male aggression
• Presence of females
Aggression between male rats is enhanced in the presence of female rats (Barnett et al. 1968, Barnett and Stoddard 1969).
Flannelly and Lore (1977) housed males for one week with intact females, or spayed females, or small intact males. The authors measured the male rat's aggression toward a strange male before and after living with these different cagemates. The authors found that the male's aggression was enhanced after living with intact females, but not after living with other types of rats. Males housed with females fought intruders more frequently, with greater intensity and severity than males housed with spayed females or intact males (Flannelly and Lore 1977).
The odor of a female in heat appears to be an important factor in intensifying male-male aggression. Taylor (1975) found that male rats attacked others at much higher frequencies after just smelling a female rat in heat.
• Presence of nursing young
The presence of nursing young may intensify the aggression of male colony members. Dominant males from colonies containing at least one lactating female and litter attacked male intruders severely and at great length (Lore and Luciano 1977).
• Presence of juveniles
Males are less aggressive toward each other when they can smell a juvenile nearby. Males who were briefly exposed to a 17-27 day old juvenile -- especially a juvenile female -- were subsequently less aggressive toward a strange adult male (Taylor and Weiss 1982).
The juvenile exemption from attack
Adult rats almost never attack juvenile rats -- adults are far more likely to attack another adult or an infant (see infanticide) than a juvenile. Rats aren't the only rodents with a juvenile exemption: adult deermice don't attack juveniles during the nonbreeding season (Healey 1967), and Chinese hamster juveniles are partially protected from adult attack as well (Dasser 1981).
Adult wild rats living naturally in a garbage dump never attacked juveniles under 50 grams (Robitaille and Bovet 1976). Similarly, adult wild rats in a semi-natural enclosure did not attack young rats under 50 days of age. As the juveniles grew older, they were attacked more, but even then they enjoyed partial immunity: until 86 days these attacks almost never included bites (Calhoun 1963).
In the laboratory, Thor and Flannelly (1976) housed domestic adult male rats individually and presented them with juvenile rats of different ages. The youngest rats were immune from attack, but older juveniles were not. The older the juvenile, the more severely it was attacked. Specifically, adult males rarely attacked 21-40 day old juvenile rats, but followed the young intruders around the cage without harming them. Adult males attacked almost all 41-60 day old juveniles after considerable sniffing and posturing, and killed 25% of them. Adult males attacked 61-80 day old juveniles almost immediately, bit them many times, and killed 42% of them. Relative juvenile immunity holds in mice, too: Taylor (1982) found that adult male mice rarely attacked 15-25 day old juveniles but readily attacked adults.
What protects juveniles from adult attack?
Juveniles appear to be immune from attack because of their low levels of testosterone, their small size, and possibly their juvenile behavior. As males grow older they lose this immunity: they grow in size and their urine reflects the hormonal changes associated with puberty, and these features trigger attacks from other adult males. Females, however, retain some of their immunity from attack by adult males, a male-aggression-inhibiting immunity that is reflected in the adult female's urine chemistry (Haag et al. 1974). Adult female rats are not, however, immune from aggression by other females (Debold and Miczek 1984).
Testosterone profiles of young rats: Testosterone is low in young rats until age 40 days (1.6 ng/2 ml), increases from age 40 to 60 days and reaches its highest peak around 60 days (11.4 ng/2 ml). Testosterone subsequently declines and stabilizes at an intermediate level at around age 90 days (3.67 ng/2 ml) (Grota 1971; see also Knorr et al. 1970, Resko et al. 1968 for similar results). This increasing level of testosterone may therefore be indirectly responsible for increased adult aggression associated with age. If male rats are castrated before puberty, they remain immune from attack as adults (Flannelly and Thor 1978).
Higher testosterone triggers more attacks, low testosterone inhibits attacks: In mice, adult males attack juveniles dabbed with adult male urine. Conversely, adult males attack reduce their attacks on adult males dabbed with juvenile urine (Taylor 1982, Dixon and Mackintosh 1976). Juvenile urine therefore provides some protection from attack, while adult male urine reduces the immunity enjoyed by juveniles. However, urine isn't the whole story: dabbing urine on an adult or juvenile doesn't completely reverse the resident adult male's normal response to it (Taylor 1982).
Similarly, adult male mice attack females who have been injected with testosterone (Mugford and Nowell 1970). The cue appears to be olfactory. Testosterone injections into female mice cause chemical changes in their urine and the secretions of their preputial glands that elicit aggression from adult males (Mugford and Nowell 1971a and b).
In rats, castrated adult male intruders elicit less aggression than intact ones (DeBold and Miczek 1984, Flannelly and Thor 1978).
Low testosterone not only inhibits attacks on the juvenile, it also reduces aggression between adult males. The presence of a juvenile reduces aggression between adult males, but juvenile females and castrated juvenile males are more effective in inhibiting adult male aggression than intact juvenile males (Taylor and Weiss 1982).
Testosterone profiles vary with experience: Experience and events in the environment influence testosterone levels. For example, male rats who are stressed in infancy (e.g human handling, or electric shock) develop sooner than unhandled males: their testosterone peaks between 50 and 60 days, and declines by 70 days. Therefore, rats who are stressed in infancy may have an earlier rise in testosterone, which may decrease age of onset of attack by resident adult males. As adults, males who were stressed as infants had the same levels of testosterone as unstressed males (Grota 1971).
Small body size also helps inhibit aggression: In mice, dwarf adult mice are used to study whether small size inhibits aggression. Dwarf mice have a form of pituitary dwarfism (recessive pituitary dwarfism: dw/dw) which results in normally proportioned adults that weigh 6-12 grams and are about the size of a normal 20 day old juvenile. Adult dwarf mice have normal gonads, normal levels of testosterone, and normal urine. Small body size appears to provide some protection: dwarf males are attacked less readily, and less frequently, than normal sized males. However, small size isn't enough: dwarf adult males are still attacked more than normal juvenile males (Taylor 1982).
Other factors: Low testosterone and small size may not be the only features of juveniles that inhibit adult attack. For example, juvenile behavior, like fearless exploration, may also play a role.
The neutering of male animals is an ancient technique used to make domestic male animals docile (think of oxen, geldings, and wethers as compared to bulls, stallions, and rams). Castration has a similar effect on male rats.
Castration of all male rats in a group reduces intermale aggression. If all rats in a group are castrated, they generally behave less aggressively toward each other, and their social hierarchy is disrupted (Stewart and Palfai 1967).
Castration of just the dominant male decreases his aggression and status. Castrated dominant rats show about 15% of the aggression displayed by intact rats toward intruders. Castration may also cause a dominant rat to lose his social position. An intact subordinate male may become dominant in his place (Albert et al. 1986). Testosterone implants brings male aggression back to pre-castration levels (Albert et al. 1986, Bean and Conner 1978, Christie and Barfield 1979b), though it may take 2-3 weeks (DeBold and Miczek 1981).
This decline in aggression after castration may take several weeks. DeBold and Miczek (1984) castrated the dominant rats of several colonies, then introduced a new, intact male once per week. They found that aggression toward the intruder had greatly declined by 5 weeks post-castration. The castrated residents attacked and threatened intruders much less.
Intact male residents attack castrated male intruders about half as much as they attack intact intruders (Debold and Miczek 1984, Christie and Barfield 1979, see also Flannely and Thor 1976, Barfield et al. 1972). In general, the greatest amount of aggression occurs between intact residents and intact intruders. Encounters between castrated residents and intact intruders are only about half as aggressive. Encounters between castrated or intact residents and castrated intruders are about one third as aggressive as the intact-intact pairing. In all, castrate-castrate pairs show minimal aggression toward each other (Christie and Barfield 1979).
However, castration does not eliminate aggression entirely. Castrated rats can be aggressive in their home cage. Castrates rarely initiate an aggressive encounter with an intruder, but will display aggression if provoked (Christie and Barfield 1979) and will fight effectively (Barfield et al. 1972). Castrated residents may still dominate over intact intruders: Albert et al. (1986) found that castrated residents were more aggressive than an intact intruder 50% of the time, while Christie and Barfield (1979) found that castrates dominated intact intruders 1/3 of the time.
Therefore, castration of resident male rats results in a powerful decline in aggression toward intruders, but not necessarily a complete elimination of aggression. Castration results in a loss of social position, as demonstrated by the subordinate cagemates becoming more aggressive toward intruders after the alpha rat's castration. Castrated rats also elicit less aggression from other rats.
So, castration of a male tends to reduce his aggressive behavior, and castration also makes him less likely to be attacked by other males. Does spaying have the same effect on females? In a nutshell, no. Spaying has much less effect on female aggressive behavior than castration does on male aggressive behavior.
Spayed resident females' response to female intruders: Spaying (ovariectomy) does not reduce a resident female's aggression. A spayed female continues to show pre-spay levels of aggression toward intruder females even 7 weeks after the spay surgery. In contrast, a castrated male's aggression toward intruders declines markedly during the 7 weeks after his castration (Debold and Miczek 1981, 1984).
Resident females' response to spayed/unspayed female intruders: Spaying female intruders does not reduce the amount of aggression they receive from resident females: resident females attack spayed females just as much as they attack intact females. However, resident females do tend to refrain from attacking lactating intruders (DeBold and Miczek 1984).
Resident males' response to spayed/unspayed female intruders: Spaying also has little effect on how an intruder female is treated by resident males. Resident males inspect intruder females but rarely attack them, whether they are intact or spayed. Mounting of an intruder female may depend on whether resident females are present: if resident females are present, males may show only low levels of sexual behavior toward female intruders, but if there are no resident females, the males may be more sexually active (DeBold and Miczek 1984).
Flannelly and Lore (1977) paired a single intact male with two intact or two spayed females in a neutral area. The male rats tended to investigate, sniff, and mount them. The male investigated, sniffed, and mounted spayed females somewhat less than intact females, though the differences were not significant. Aggressive behavior toward any of these females was brief.
Note, however, that Alberts and Galef (1973) found that many resident males initially attacked spayed and unspayed female intruders, but also attempted to copulate with them.
In maternal aggression a pregnant or lactating mother rat attacks intruders (other rats, humans, or other animals).
Maternal aggression can include a lunging attack, typically directed at the neck or back of an intruding rat. The mother rat may also bite, sidle, and kick (thrust of one hind leg toward intruder during a sidle). Intruders usually respond with nose-offs or belly-up postures (Erskine et al. 1978).
Maternal aggression is present but occurs relatively infrequently at the end of pregnancy, and starts to increase after the birth of the litter. Maternal aggression increases after birth of the young and peaks during the 9th day of lactation. On the ninth day, the highest percentage of females attacks and bites intruders, with the lowest hesitation before doing so. After the ninth day, maternal aggression declines (Erskine et al. 1978).
The presence of the litter is an important in triggering maternal aggression: if the litter is removed the mother rat's aggression subsequently decreases. Four hours after the litter's removal, the mother rat's aggression is down to nearly normal levels (Erskine et al. 1978). Pup age appears to be an important factor in maternal aggression as well. Regardless of time since partuition, mothers fostering young pups are more aggressive than mothers fostering older pups (Giovenardi et al. 2000).
Maternal aggression may discourage other animals from approaching and possibly harming her offpsring (e.g. predators, unfamiliar rats who could commit infanticide).
Maternal aggression is therefore a normal behavior associated with parenting. It probably helps protect the vulnerable young from animals who might harm them. Maternal aggression may be distressing to the pet rat owner, but it is entirely adaptive. Maternal aggression tends to wane as the youngsters grow older, and the mother rat eventually returns to normal levels of aggression.
Male aggression toward other rats is a poor predictor of aggression toward humans.
Handling and socialization of young rats by humans greatly reduce a rat's aggression toward humans. But such experience with humans has no impact on how rats behave toward other rats. For example, Galef (1970) handled wild rat pups from infancy, and these rats grew up to be docile toward humans but very aggressive toward intruder rats. Therefore, a male rat may be socially dominant and aggressive toward intruders, but quite docile and tame in human hands.