Intermale and maternal aggression in adult rats tested at different ages

Physiology & Behavior, Vol. 29, pp. 1013-1018.PergamonPress, 1982. Printedin the U.S.A.

Intermale and Maternal Aggression in Adult Rats Tested at Different Ages L O R E Y K. T A K A H A S H I 1 A N D R I C H A R D K. L O R E z

D e p a r t m e n t o f Psychology, Douglass College, Rutgers University, N e w Brunswick, N J 08903 R e c e i v e d 19 J u l y 1982 TAKAHASHI, L. K. AND R. K. LORE. lntermale and maternal aggression in adult rats tested at different ages. PHYSIOL. BEHAV. 29(6) 1013--1018, 1982.--In Experiment 1, rat colonies consisting of two males and one female were established when the animals were either 100, 200, or 300 days old. All colony members were reared from weaning in small, stable isosexual groups prior to colony formation. Males that were 200 days old engaged in significantly more fighting at colony formation, and, three weeks later, were more likely to wound an intruder during a 24 hr aggression test than either younger or older animals. These findings demonstrate that colony aggression in male rats varies markedly as a function of age when reared under housing conditions that are employed in most laboratories. In Experiment 2, colony females were individually housed after their second pregnancy and their response to a male intruder was evaluated for a 24 hr period after nine days of lactation. Maternal aggression occurred at all ages, indicating that age may be of lesser importance in nest defense of females than intermale conflict. However, maternal aggression was ineffective in preventing the destruction of a high proportion of the litters by intruders in the three maternal age groups. Intermale aggression

Maternal aggression

Age

SOME investigators who work with adult laboratory rats use animals that are less than 125 days old whereas others use much older animals. The implicit assumption here is that once an animal reaches sexual maturity, its behavior remains fairly constant thereafter [7]. In intraspecies aggression research, this assumption is particularly suspect since there is no information available on the relationship between age and fighting in laboratory rats during the period from early sexual maturity (at about 60 days of age) through late adulthood. In contrast to the lack of information on the relationship between age and aggression in laboratory rats, studies on wild Norway rats indicate that age is a crucial factor influencing the readiness of an animal to attack a conspecific. For example, Calhoun's [6] extensive study of a captive population of wild rats showed that animals below 85 days of age rarely engaged in serious fighting, and a consistently high level of fighting, as indicated by the prevalence of wounds, did not occur until the animals were more than 250 days old. Similarly, Robitaille and Bovet [ 13] observed in wild populations of rats that fighting was rare in animals estimated to weight less than 150 g. Agonistic encounters appeared most frequently among larger animals that were estimated to weight 150 to 350 g. However, these 150 to 350 g animals avoided confrontations with still larger individuals and always lost in fights with these heavier animals. Since the size of rats increases with age, these findings indicate a strong and positive relationship between age and aggression and that older animals are more aggressive than younger but sexually mature adults. Domestic rats living in established colonies sometimes

Rats

Intruder

Social experience

Infanticide

ignore unfamiliar intruders, or at the opposite extreme, attacks resulting in death to the intruder may occur. These differences in colony aggression (reviewed by Lore et al. Aggress. Behav., in press) are partially explained by subtle differences in the previous social/aggressive experience of the animals, as well as the nature of the physical environment and duration of the aggression test. However, differences in the age of the colony residents may also account for some of the inconsistencies noted between different laboratories. Often age differences are confounded with other variables. For example, rat colonies consisting of three males and three females were given weekly exposures to conspecific intruders [5]. At the end of nine weeks, these colonies ("early-introduction colonies") showed higher levels of alpha male attack than colonies not exposed to intruders ("late-introduction colonies"). Thus, previous exposure to intruders produced higher levels of aggression relative to animals that were comparable in age but given no intruder experiences. In the same study, the emergence of an alpha male in the late-introduction colonies occurred with fewer weekly intruder exposures. Although it was concluded that extensive intracolony social experiences potentiated intruder-directed aggression, the animals in the lateintroduction colonies were also older than those in the earlyintroduction colonies when weekly intruder tests were initiated. Hence, the more rapid emergence of alpha male attack may have been, in part, a consequence of their greater age. In this study, small colonies of rats consisting of two

1Present address: Department of Biology, Princeton University, Princeton, NJ 08544. ~Present address: Department of Psychology, Busch Campus, Rutgers University, New Brunswick, NJ 08903.

C o p y r i g h t © 1982 P e r g a m o n Press---0031-9384/82/121013-06503.00/0

1014

T A K A H A S H I A N D LORE

males and one female were established when the animals were either 100, 200, or 300 days of age. All animals were reared in small, isosexual groups to duplicate the housing conditions of laboratory rats commonly used in aggression research. In Experiment 1, fighting at colony formation and intruder-directed attack was assessed for the resident males. In Experiment 2, maternal aggression directed toward an unfamiliar male was examined following delivery of the second litter. EXPERIMENT 1 METHOD

Colony Animals Male and female Long-Evans rats obtained from Perfection Breeders, Douglassville, Pennsylvania served as colony residents. All animals were weaned at 21 days of age and reared in isosexual groups of four animals. Two weeks prior to the start of colony formation, the animals were separated and placed in standard laboratory cages. Purina rat chow and water were freely available throughout the study.

TABLE 1 INTERMALE AGGRESSIONDURING THE INITIAL 20 MIN OF COLONY FORMATIONAT EITHER 100, 200, OR 300 DAYS OF AGE Age in Days

Proportion of colonies exhibiting aggression Mean duration in seconds of lateral and on-top postures (SE)* Mean number of bites (SE)*

100

200

300

2/9

8/9

2/9

19.0 (5.0)

29.1 (12.4)

98.8 (55.2)

0.5 (0.5)

1.6 (1.3)

1.5 (0.5)

*Scores are based only upon those colonies exhibiting aggression.

ony males and the intruder were weighed, anesthetized with ether, and examined for body wounds. The two resident males and the previously separated female were then returned to their respective colony.

Colony Formation Three groups of nine colonies, each consisting of two males and one female, were formed when the animals were either I00, 200, or 300 days old (_+2 days). The males averaged 295.0, 484.3, and 513.3 g in body weight whereas females averaged 208.5, 294.5, and 319.7 g at 100, 200, and 300 days of age, respectively. One male in each colony was earpunched and marked with dye to facilitate rapid identification. At colony formation, two nonlittermate males were placed simultaneously into a stainless steel cage (63.5 x 41.5 x 25.5 cm) with Plexiglas fronts and pine chips scattered on the floor. The aggressive interactions between the two males were observed for the first 20 min of colony formation. The specific behaviors recorded for each male were: (1) combined duration of on-top and lateral postures and (2) number of bites. These behaviors have been described previously [3]. At the conclusion of the 20 min observation session, the female was placed into the colony.

Intruder Tests After 21 days of cohabitation, the female was removed and a naive male Long-Evans intruder was placed into each colony for a 24 hr period. Prior to this aggression test, all intruders were reared in isosexual groups of 4 to 6 animals from weaning at 21 days of age and were 100 to 114 days old with an average body weight of 346.5 g. The 24 hr intruder test consisted of two successive phases: (1) the aggressive behaviors of the colony males toward the intruder were recorded for the first 30 min and (2) immediately after this observation period, the intruder was removed and placed into a Plexiglas escape chamber (28.4x 12.7x 11.4 cm) and returned to the colony. At one end of the chamber, a 5. I cm diameter hole was cut to provide passage. The use of the escape chamber minimizes damaging attack, but still provides a sensitive long-term assessment of the colony's aggressive reaction to the intruder [18]. That is, intruders that are unable to leave the escape chamber because of the readiness of residents to initiate attack will incur greater body weight losses after 24 hr than intruders that are attacked less frequently. After the aggression test, the col-

RESULTS

Colony Formation Fights Aggression occurred in nearly 90% of the 200 day old group at the time the males were initially placed into the colony (Table 1). In marked contrast, fighting was observed in only 22.2% of the 100 and 300 day old males. A Fisher exact probability test was used to compare the number of 200 day old colonies exhibiting aggression to the combined total of the 100 and 300 day old groups. Significantly more fighting occurred at 200 days of age than in the other two age groups (p =0.004 two-tailed). Although the aggressive behaviors were recorded separately for each male in the colony, the behavioral scores were combined because of the low number of 100 and 300 day old animals that engaged in fighting. The combined aggression scores for those colonies that exhibited fighting are presented in Table 1. Interestingly, there was an increase in lateral and on-top posturing as the animals became older as well as a slight increased in the number of bites. Closer examination of each colony which engaged in aggression revealed that 87.1% (100 days) 94.6% (200 days), and 100% (300 days) of the lateral and on-top behaviors were made by one male. Furthermore, this aggressive male accounted for all the bites that were observed during fighting.

Aggression of Colony Males Toward an Intruder As Table 2 indicates, 33.3% of 121 day and 55.6% of 221 day old resident males showed aggression toward the intruder during the first 30 min. In contrast, fighting was not observed in the 321 day old group. This difference in the number of 121 and 221 day old colonies exhibiting aggression toward an intruder in comparison to the older 321 day old animals was significant (Fisher exact probability test, p =0.04, two-tailed). It should be noted that in both these groups, the aggression directed toward the intruder was made by only one male. Furthermore, the duration of lateral and on-top posturing and the number of bites were considerably higher in the 221 day old colonies. There were no observed instances of intruders attacking the resident males.

AGE AND AGGRESSION

1015

TABLE 2 BEHAVIORAL AND PHYSICAL MEASURES ASSOCIATED WITH INTRUDER PRESENCE DURING THE INITIAL 30 MIN OBSERVATION PERIOD AND AFTER 24 HR Age in Days

Initial 30 min period Proportion of colonies exhibiting aggression Mean duration in seconds of lateral and on-top postures (SE)* Mean number of bites (SE)* After 24 hr Mean number of intruder wounds (SE) Mean percent body weight change of intruder (SE) Mean percent body weight change of colony males (SE)

121

221

321

3/9

5/9

0/9

72.8(20.4)

184.8(104.8)

0.0(0.0)

0.7 (0.7)

4.0 (2.1)

0.0(0.0)

0.3 (0.2)

6.9 (2.9)

0.0(0.0)

-0.5 (0.6)

-2.9

(0.8)

-0.9(0.7)

0.I (0.6)

-0.8

(0.7)

-0.4(0.3)

*Scores are based only upon those colonies exhibiting aggression.

Physical Measures After 24 hr Intruders placed into the 221 day old colonies lost somewhat more body weight after 24 hr than those placed into either the 121 or 321 day old groups (Table 2). However, a Kruskall-Wallis one-way analysis of variance conducted separately on the percent body weight change of intruders and colony males indicated no reliable differences as a function of colony age. In contrast to the nonsignificant loss in body weight among intruders, there was a high number of wounds on intruders placed into the 221 day old colonies (KruskallWallis H = I 4 . 1 0 , p<0.01). Subsequent tests indicated that the number of intruder wounds was reliably higher after fighting in the 221 day old colonies than in the other two groups (p<0.05 in both cases, Mann-Whitney test, twotailed). There were no reliable differences in the number of intruder wounds following exposure to 121 and 321 day old colonies. No wounds were found on any of the colony residents. EXPERIMENT 2

(day of birth = Day 0), when maternal aggression is at its peak [9], a naive male Long-Evans intruder was placed into the colony of each female for a 24 hr period. The rearing conditions of these intruders were identical to that described previously. The intruders were 100 to 125 days old with a mean body weight of 370. I g. The 24 hr aggression test was conducted in the same manner as indicated in the previous experiment: A 30 min observation period followed by the use o f the escape chamber. Female body weights were taken approximately 15 to 30 min prior to the intruder test. The mean body weights of the three groups of females on Day 9 of lactation were, in terms of increasing ages, 331.5 g, 353.1 g, and 342.2 g, respectively. In addition, the number of pups in each litter was counted at this time. Pups were counted in their nest to minimize nest disturbance. Differences in behavior among groups were tested for significance by Kruskall-Wallis one-way analysis of variance and Fischer exact probability tests.

RESULTS

Initial 30 min Observation Period METHOD When the females in the three age groups were visibly pregnant with their second litter the colony males were removed. One female in the 100 day old group died of unknown cause prior to her second pregnancy and one 300 day old female lost her pups prior to testing. Furthermore, four females (one 200 day old, three 300 day old) failed to successfully reproduce. Thus, eight females in each of the 100 and 200 day old group and five females in the 300 day old condition were tested for maternal aggression. At the time of their second delivery, the females averaged 173.9 days (range=154--192), 268.1 days (range=247-302), and 359.4 days (range=342-379), in the young, middle, and older age conditions, respectively. On postpartum Day 9

Introduction of male intruders into colonies on postpartum Day 9 elicited maternal aggression in all three groups of females (Table 3). Although more young and middle-age females attacked (87.5%), relative to older females (40%) there were no reliable differences in aggression between these groups (Fisher exact probability test, p=0.13, two-tailed). Furthermore, the middle-age females engaged in more aggressive posturing and made more bites than the other two groups, but neither difference was statistically significant. In addition to the occurrence of maternal attacks, intruder aggression was observed in approximately 50% of the colonies (Table 3). There were no signficant differences in the combined duration of lateral and on-top posturing or the fre-

1016

T A K A H A S H I A N D LORE TABLE 3 BEHAVIORAL AND PHYSICAL MEASURES ASSOCIATEDWITH MATERNAL AGGRESSIONDURING THE INITIAL 30 MIN OBSERVATION PERIOD AND AFTER 24 HR Age in Days (range)

Initial 30 min Proportion of colonies exhibiting maternal aggression Mean duration in seconds of lateral and on-top postures by females (SE)* Mean number of bites by females (SE)* Proportion of colonies exhibiting male intruder aggression Mean duration in seconds of lateral and on-top postures by male intruders (SE)t Mean number of bites by male intruders (SE)¢ After 24 hr Mean percent body weight change of intruders (SE) Mean percent body weight change of females (SE)

154-192

247-302

342-379

7/8

7/8

2/5

81.3(28.8)

162.3(61.2)

46.0(40.0)

2.6 (1.1)

5.0 (2.2)

0.5 (0.5)

4/8 40.0(20.6)

3/8 80.0(61.8)

3/5 134.7(71.9)

1.0 (1.0)

2.3 (0.3)

2.7 (0.9)

-0.3 (0.5)

1.6 (0.7)

1.3 (1.8)

3.1 (2.8)

5.5 (2.2)

-1.3 (1.3)

*Scores are based only upon those colonies exhibiting maternal aggression. i'Scores are based only upon those colonies exhibiting male intruder aggression.

quency of bites made by intruders. In the majority of cases (60%), intruders attacked following female-initiated fights, but in four colonies intruders initiated the fights. The most dramatic finding during the 30 min observation session was the high occurrence of male infanticide in all three groups of females (Table 4). Destruction of entire litters was observed in nearly 60% of the 21 colonies tested during the first 30 min. Our informal observations indicated that bites on the young occurred rapidly and appeared to be made indiscriminately on all areas of the body. Moreover, once infanticide began, most pups were killed in rapid succession. There was not a single observed instance in which a female killed a pup.

After 24 hr No significant changes in body weight were found for either females or intruders after the aggression tests (Table 3). Moreover, wounds were confined to three intruders (mean n u m b e r = l . 3 ) in the middle-age group and no body wounds were found on any female residents. As Table 4 indicates, 85.7% of the litters were completely destroyed after 24 hr. Although it is uncertain whether the male intruder was solely responsible for the death of all pups, the occurrence of killing observed when the intruder was first placed into the colony strongly suggests that the male was responsible for these acts. GENERAL DISCUSSION The results of Experiment 1 indicated that males that were 200 days old at the time of colony formation engaged in

significantly more fighting and, three weeks later, were more likely to injure unfamiliar intruders during a 24 hr aggression test than both younger (100 day old) and older (300 day old) animals. It should be emphasized that the peak in intruderdirected aggression is not attributable to a greater difference in age between residents and intruders at 221 days than at 121 days of age because (1) 321 day old resident males which were considerably much older than intruders failed to exhibit aggression and (2) the same relationship between age and aggression was demonstrated at the time of colony formation, which consisted of pairing males that were similar in age. Taken together, the increase and subsequent decline in aggressiveness as a function of age provide strong evidence that the propensity to exhibit intermale aggression will vary significantly long after the animals attain sexual maturity. These findings agree, in part, with Calhoun's [6] work on wild rats: Adult males did not engage in serious fighting until they were 200 or more days old. However, the almost total absence of aggression in our 300 day old males differs strikingly from Calhoun's findings that animals at this age and even much older (450 days) continued to fight readily and dominated younger males. This contradiction is very likely a consequence o f experiential differences between laboratory and wild rats. In the field, the older rat that survives is probably an experienced fighter, whereas our 300 day old rats were reared for prolonged periods in small, stable groups where aggression occurs infrequently [1,8]. Thus, in the absence of varied aggressive experiences for an extensive period of time, a reduction in aggression will occur. Certainly, laboratory rats given varied experiences with intruders---hence partially duplicating the wild rat's

AGE AND AGGRESSION

1o17

TABLE 4 LITTER SIZE AND OCCURRENCE OF INFANTICIDEBY INTRUDERS AS A FUNCTION OF MATERNALAGE Age in Days (range)

Mean litter size (SE) Proportion of litters completely killed after 30 min Proportion of litters completely killed after 24 hr

154-192

247-302

342-379

13.5(0.9) 3/8

9.0(1.1) 5/8

5.8(1.5) 4/5

7/8*

7/8i

4/5*

*One pup survived in a litter of eight. ?Five pups survived in a litter of five. ~:Six pups survived in a litter of six.

e x p e r i e n c e - - a r e capable of intense aggression toward strangers at 415 days o f age [4]. Other investigators have found that older laboratory rats reared under conditions comparable to ours were equally nonaggressive. F o r example, Barnett and Hocking [2] reported that laboratory rats reared in mated pairs or in groups of three of like sex showed little evidence of aggression when colonies were formed between 322 to 350 days of age. Although their study purported to demonstrate that laboratory rats were considerably less aggressive than wild rats, their findings, instead, inadvertently provide further evidence that the lack of varied aggressive experiences can attenuate the motivation for rats to exhibit fighting at an old age. An alternative, but not entirely incompatible explanation for the reduction in aggression among 300 day old rats is that social housing inhibited fighting. This suggestion that prolonged periods of social rearing may reduce aggression in animals is not a new one. Long ago, Scott [16] found that mice reared together became completely nonaggressive toward each other even after attaining adulthood. Scott suggested that these animals developed a habit of notfighting which was termed passive inhibition. Eibl-Eibesfeldt [8] reported a similar finding in rats. However, both investigators indicated that unfamiliar conspecifics introduced into the group were often attacked and wounded. Hence, our findings of no intruder-directed aggression in 321 day old males may be the result of extensive and stable social rearing conditions which decreased the propensity of males to aggress even toward unfamiliar conspecifics. If social rearing rather than the lack of varied conspecific experiences is responsible for the attenuation of aggression in old males, then animals raised under conditions with little or no opportunity to interact with other conspecifics should exhibit higher levels of fighting when tested at 300 or more days o f age relative to socially reared animals.

The results of Experiment 2 showed that socially reared females were capable of exhibiting maternal aggression irrespective o f their age and litter size at the time of testing. Whether female rats at different ages show comparable levels of aggression throughout the entire lactation period is unknown, but at least when tested on postpartum Day 9, the duration of aggressive posturing and the number of bites were not reliably different. Svare and Gandelman [17] reported a similar finding in a longitudinal study conducted in mice. They found that the majority of females that were able to reproduce displayed aggression during six successive lactation periods despite a gradual reduction in litter size. However, they did note a decline in the intensity of maternal attacks across successive lactation periods that indicates a possible age effect. Although we found no significant differences in aggression between females tested at various ages, our findings should be interpreted cautiously due to the small sample size of the oldest group of animals. Experiment 2 further revealed that unlike the majority of intruders that rarely attack resident males [3], intruder aggression directed toward the lactating female occurred in nearly 50% of the 21 colonies that were tested during the initial 30 rain observation period. In addition, an unexpectedly high number of intruders committed infanticide. In some intruder tests, once pup-killing began, the entire litter was destroyed within the first 30 min. Previously, Erskine, Denenberg, and Goldman [10] found that 38% of 110 day old male rats attacked and killed the litters of lactating females. Although the percentage of males that killed in their study is somewhat lower than the percentage we found after 24 hr (85%), their animals were tested for only 45 min. Indeed, several o f the males in our study required a period of greater than 30 min to initiate the killing response. Other investigators [14,15] have also reported that 55-80% of male rats will kill pups that are introduced into their cage for a 24 hr period. The high number of males that engaged in pup-killing using a test session that is considerably longer than commonly used to study maternal aggression is especially noteworthy because it offers an interesting laboratory model to study the reproductive strategies of males. To date, several studies have examined the propensity for males to exhibit pup-killing in collared lemmings [12], meadow voles [20], and mice [11,19] and have interpreted their findings in terms of a reproductive tactic for males to increase their fitness relative to other males. That is, males that kill appear to benefit by impregnating the female that has lost her litter, thereby siring young that are his after a shorter time period than males that do not kill. Whether pup-killing by laboratory male rats will enhance their reproductive output is a question that awaits further investigation. ACKNOWLEDGEMENTS This research was supported by grants from the Rutgers Research Council and the Busch Fund, Rutgers University. We are especially grateful to Kevin Flannelly for technical assistance in Experiment 1.

REFERENCES 1. Barnett, S. A. The Rat: A Study in Behavior, revised edition. Chicago: University of Chicago Press, 1975. 2. Barnett, S. A. and W. E. Hocking. Further experiments on the social interactions of domestic "Norway" rats. Aggress. Behay. 7: 259-263, 1981.

3. Blanchard, R. J. and D. C. Blanchard. The organization and modeling of animal aggression. In: The Biology of Aggression, edited by P. F. Brain and D. Benton. Leiden: Sijthoffand NoordholT, 1981, pp. 529-561.

1018 4. Blanchard, R. J., D. C. Blanchard, T. Takahashi and M. J. Kelley. Attack and defensive behaviour in the albino rat. Anim. Behav. 25: 622-634, 1977. 5. Blanchard, R. J., L. K. Takahashi and D. C. Blanchard. The development of intruder attack in colonies of laboratory rats. Anim. Learn. Behav. 5: 365-369, 1977. 6. Calhoun, J. B. The Ecology and Sociology o f the Norway Rat. USPHS Publication No. 1008. Washington, DC: U. S. Government Printing Office, 1962. 7. Denenberg, V. H. Paradigms and paradoxes in the study of behavioral development. In: Origins o f the Infant's Social Responsiveness, edited by E. B. Thoman. Hillsdale, N J: Lawrence Erlbaum Associates, 1979, pp. 251-289. 8. Eibl-Eibesfeldt, I. The fighting behavior of animals. Scient. Am. 205: 112-122, 1961. 9. Erskine, M. S., R. J. Barfield and B. D. Goldman. Intraspecific fighting during late pregnancy and lactation in rats and effects of litter removal. Behav. Biol. 23: 206-218, 1978. 10. Erskine, M. S., V. H. Denenberg and B. D. Goldman. Aggression in the lactating rat: Effects of intruder age and test arena. Behav. Biol. 23: 52-66, 1978. 1 i. Labov, J. B. Factors influencing infanticidal behavior in wild male house mice (Mus musculus). Behav. Ecol. Sociobiol. 6: 297-303, 1980. 12. Mallory, F. F. and R. J. Brooks. Infanticide and other reproductive strategies in the collared lemming, (Dicrostonyx groenlandicus. Nature 273: 144-146, 1978.

TAKAHASHI

AND LORE

13. Robitaille, J. A. and J. Bovet. Field observations on the social behaviour of the Norway rat, Rattus norvegicus (Berkenhout). Biol. Behav. 1: 289-308, 1976. 14. Rosenberg, K. M. Effects of pre- and postpubertal castration and testosterone on pup-killing behavior in the male rat. Physiol. Behav. 13: 159-161, 1974. 15. Rosenberg, K. M., V. H. Denenberg, M. X. Zarrow and B. L. Frank. Effects of neonatal castration and testosterone on the rats' pup-killing behavior and activity. Physiol. Behav. 7: 363-368, 1971. 16. Scott, J. P. Animal Behavior. Chicago: University of Chicago Press, 1958. 17. Svare, B. and R. Gandelman. A longitudinal analysis of maternal aggression in Rockland-Swiss albino mice. Devl Psychobiol. 9: 437-446, 1976. 18. Takahashi, L. K., S. Grossfeld and R. K. Lore. Attack and escape in the laboratory rat: A modification of the colonyintruder procedure. Behav. Neural Biol. 29: 512-517, 1980. 19. vom Saal, F. S. and L. S. Howard. The regulation of infanticide and parental behavior: Implications for reproductive success in male mice. Science 215: 1270-1272, 1982. 20. Webster, A. B., R. G. Gartshore and R. J. Brooks. Infanticide in the meadow vole, Microtus pennsylvanicus: Significance in relation to social system and population cycling. Behav Neural Biol. 31: 342-347, 1981.