and the Prairie Studies of the —a secretive, mouselike —have revealed that may be responsible for monogamous behavior

by C. Sue Carter and Lowell L. Getz

bservation of the mating and females of monogamous species tend from studies of prairie was the pup-rearing habits of nonde- to be about the same in size and ap- observation that social cues regulate Oscript, brown that live pearance. Mated pairs will defend the the reproductive physiology of this spe- under weeds and grasses might not nest and territory from intruders, and cies. Even to enter estrus (sexual heat), seem an obvious way to improve knowl- both parents care for the young. Monog- a female prairie vole must sniÝ a male. edge of monogamy. After all, most hu- amous may form complex Indeed, Milo E. Richmond, now at Cor- mans can attest to the complexity of social groups that include an extended nell University, found that female prai- male-female relationships. Yet studies family and oÝspring of various ages. rie voles do not have the ovarian cycles of the prairie vole ( ochrogas- Incest is avoided within these families; that are typical of nonmonogamous ter), a common pest throughout the adult young usually do not reproduce mammals. In monogamous voles, a fe- midwestern U.S., have led us on a fasci- as long as they live with related family male must have a male partner to in- nating scientiÞc journey from our start- members. Finally, we should point out duce estrus. ing point in ecology to the exploration that although common in birds, monog- Furthermore, not just any male can of the neuroendocrinology of social amy is rare in mammals. In an exhaus- bring a female into heat. Fathers and bonds. Unlike most rodents, prairie tive survey, Devra G. Kleiman of the Na- brothers do not seem capable of elicit- voles form long-lasting pair bonds, and tional Zoological Park in , ing sniÛng. This may be an adaptive both parents share in raising their D.C., found that only about 3 percent mechanism designed to prevent incest. young. Our studies have provided a new of mammals are monogamous. In fact, both males and females will es- understanding of the importance of two Sexual exclusivity, however, is not a sentially remain prepubescent as long hormones, and , feature of monogamy. Studies of the as they stay with their families. which are well known for their respec- prairie vole as well as those of other tive roles in reproduction and body wa- mammals and birds have indicated that y sniÛng an appropriate male, ter regulation. Work with voles now sug- absolute sexual monogamy is not nec- the female picks up a chemical gests that these hormones are involved essarily associated with social monog- B signal called a pheromone. Phero- in the development of monogamy. amy. In fact, DNA Þngerprinting tests mones in turn trigger the hormonal The chief criterion that deÞnes mo- have shown that oÝspring of female events needed to activate the ovaries nogamy is a lifelong association be- prairie voles are not always fathered by and to induce heat. A small chemical tween a male and a female. Within this the cohabiting males. In some cases, a sense organ, known as the vomerona- broad deÞnition lie several characteris- litter may have mixed paternity. sal organ, helps to mediate the eÝects tics that are easily observed. Males and Because prairie voles incorporate the of pheromones. John J. Lepri, now at the deÞning features of monogamy, they University of North Carolina at Greens- make excellent subjects for the explo- boro, and Charles J. Wysocki of the Mo- ration of the biological foundations of nell Chemical Senses Center in Philadel- C. SUE CARTER and LOWELL L. GETZ monogamy, at least as it exists among phia found that removal of the vomero- have collaborated on their respective studies of behavioral endocrinology and nonhumans. Prairie voles are also small, nasal organ in the female prevented the Þeldwork to investigate the biology of weighing only a few ounces, and are eas- start of heat. A similar eÝect occurs monogamy. Both are fellows of the Amer- ily reared in the laboratory. But of par- when the olfactory bulb is detached. By ican Association for the Advancement of ticular importance for understanding removing the bulb, Jessie R. Williams, Science. Carter received her Ph.D. from the biology of monogamy is the fact Brian Kirkpatrick and Burton Slotnick, the University of Arkansas at Fayette- that not all voles are monogamous. The working in our University of Maryland ville. Before taking on her current posi- (M. pennsylvanicus) and laboratory, disrupted the sexual and tion as professor of zoology at the Uni- the (M. montanus) show social behaviors of the prairie voles. versity of Maryland, she was professor of psychology and ecology, ethology and no indications of monogamy. Voles of In our laboratory, Dean E. Dluzen evolution at the University of Illinois. these species are rarely retrapped with found that once a female is exposed Getz heads the department of ecology, the same partner and do not establish ethology and evolution at the University stable families, and males of these spe- of Illinois. He received his Ph.D. from cies do not usually care for their young. the University of Michigan. The work de- PRAIRIE VOLES (Microtus ochrogaster) Therefore, comparisons of prairie voles engage in prolonged periods of mating, scribed here reßects a 15-year collabora- with their nonmonogamous relatives tion and arose from observations made often past the time that is needed to en- during approximately 35 years of Þeld- can yield insights into the causes of sure pregnancy. These extended bouts work by Getz. monogamy. may help to facilitate the formation of One of the Þrst surprises that came monogamous social bonds.

100 SCIENTIFIC AMERICAN June 1993 Copyright 1993 Scientific American, Inc. to male odors, levels of norepinephrine, vomeronasal system. The stimulation of luteinizing begins a cascade of a neurotransmitter, and luteinizing hor- the olfactory system and the secretion of chemical and neural events that stimu- moneÐreleasing hormone (LHRH) change LHRH cause the pituitary gland to re- lates the ovary to secrete gonadal ste- within minutes. These biochemical lease a surge of luteinizing hormone into roids. Two of the most important ste- events occurred within the area of the ol- the bloodstream. In conjunction with roids secreted are estradiol, a potent factory bulb that receives input from the other endocrine changes, the release of kind of estrogen, and progesterone.

Copyright 1993 Scientific American, Inc. SCIENTIFIC AMERICAN June 1993 101 PRAIRIE VOLE (MONOGAMOUS)

FEMALE

MALE

MEADOW VOLE (NONMONOGAMOUS)

FEMALE

MALE

0 10 20 30 40 50 60 70 80 90 100 TIME IN NEST WITH PUPS (PERCENT)

PARENTAL CARE demonstrated by prairie voles far exceeds that shown by nonmonogamous meadow voles. The diÝerence is most apparent with male prairie voles, which are with the pups four times as often as male meadow voles are.

Might estrogen and progesterone also gins. This rise in progesterone probably and Syrian hamsters become nonrecep- be involved in monogamous behav- helps to regulate the duration of sexual tive after about 45 minutes of mating. ior? In collaboration with Janice M. activity by bringing these rodents into It is possible that the lengthy sexual Bahr of the University of Illinois, we and out of heat. In contrast, we found interactions of prairie voles help the searched for patterns of gonadal steroid that in prairie voles progesterone lev- sperm enter the uterus and reach the production that varied between estrus els in the blood did not increase until egg. Studies of rats by Norman T. Adler and nonestrus female prairie voles and many hours after coitus began. of the University of Pennsylvania have compared the results with data from The delayed secretion of progester- shown that complex patterns of sexu- nonmonogamous species. Estradiol, a one explains an observation made in al behavior can inßuence the release of hormone known to be essential in in- previous studies: that female prairie hormones and alter the ability of sperm ducing estrus in rodents, was elevat- voles in their Þrst estrus mate for pro- to enter the femaleÕs reproductive tract ed only in female prairie voles in heat. longed periods. In our laboratory, Di- and fertilize an egg. It declined after mating. This pattern ane M. Witt observed that when the fe- Yet improving the chances of fertil- is similar to that displayed by polyga- male was in natural estrus, males and ization is probably not the sole reason mous rodents. Analysis of the patterns females continued to engage in bouts for these extended bouts of mating. of progesterone levels, however, pre- of mating for about 30 to 40 hours. Once mating begins, females ovulate sented an unexpected Þnding. In the This extended mating period contrasts within about 12 hours, and successful nonmonogamous rats and montane sharply to that seen in nonmonoga- pregnancy can occur shortly thereafter. voles, progesterone is released in the mous species. Mating in meadow and Thus, prairie voles in their Þrst heat bloodstream shortly after mating be- montane voles persists for a few hours, continue to copulate for hours after they have met the requirements for 18 pregnancy. We suspect that, like and 16 some other primates, prairie voles may copulate to facilitate the formation of 14 monogamous social bonds. Protracted 12 mating would be particularly crucial for prairie voles that are interacting for the 10 Þrst time, because they need to estab- PRAIRIE VOLES lish their lifelong monogamous bond. 8 Indeed, some evidence for this idea comes from observations of females 6 that have previously mated and become 4 MEADOW VOLES

NUMBER OF MALE-FEMALE PAIRS 2 CAPTURED PER TRAPPING SESSION MALE-FEMALE PAIRS of prairie voles are caught far more frequently than are such 0 pairs of meadow voles. Furthermore, the 0 5 10 15 20 25 30 35 same pairs are often captured repeat- POPULATION DENSITY edly. Such studies provided the Þrst clue (NUMBER OF POTENTIAL PAIRS PER HECTARE) of monogamy in the prairie vole.

102 SCIENTIFIC AMERICAN June 1993 Copyright 1993 Scientific American, Inc. STRANGER

AGGRESSION

MATING

PARTNER

AGGRESSION

MATING

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 DURATION OF BEHAVIOR (MINUTES)

AGGRESSION by female prairie voles is revealed in 10-minute tests comparing hostility with mating preference. Females spent more time attacking strangers rather than mating with them. In contrast, they show little aggression toward their partners. pregnant. Witt found that these experi- this behavior translates into territorial- had lived with a male, females showed a enced females engaged in brief copula- ity or mate defense. In the laboratory tendency to engage in nonsexual phys- tions, sometimes limited to a few min- we have used this model to examine ical contact with the familiar male, not utes. Having established a social bond, the physiological processes responsible the stranger. These results illustrate the experienced males and females may for pair bonding. importance of social contact as an in- not need to mate for long periods. SpeciÞcally, we hypothesized that dex of partner choice. They also con- Social interaction that follows mating hormonal events induced by copula- Þrm the DNA tests revealing that in na- may be one of the mechanisms that re- tion might account for the dramatic be- ture female voles do not show absolute inforces monogamy in a species. Such havioral changes that occurred after sexual monogamy. interplay in nonmonogamous species mating. Working in our laboratory, Ker- More recently Williams examined fe- often is restricted to a brief interval ry OÕBanion took a Þrst step toward male preferences in tests that lasted when the female is in heat. For example, examining this idea. OÕBanion studied for at least three hours. She placed fe- Michael H. Ferkin, now at Cornell, ob- how females choose male partners be- male prairie voles in a relatively large served that male and female meadow fore and after mating. In his experi- maze that contained three chambers. voles did not remain in physical con- ments, familiar and unfamiliar males The females could elect to spend time tact after mating. In the Syrian ham- were tethered at opposite ends of an alone or with males tethered in the two ster, which is an especially solitary ani- arena. OÕBanion gave a female 10 min- other chambers. The were mon- mal, one of us (Carter) found that a fe- utes to choose. For the most part, fe- itored on videotape for their social and male that has mated becomes extremely males chose to mate with familiar and sexual preferences. After exploring both aggressive toward the male. In fact, the unfamiliar males equally. But if they the stranger and the partner for about female may try to kill her sexual part- ner if he does not leave after coitus. In contrast, mated monogamous PRAIRIE VOLE mammals remain highly social toward CONTACT their mates, even during nonreproduc- WITH tive periods. Leah Gavish, in our labora- PARTNER tory, demonstrated that prairie voles often touch and remain near their sex- CONTACT WITH ual partner. But this friendliness does STRANGER not extend to strangers. After mating, both males and females became excep- tionally aggressive toward unfamiliar MONTANE VOLE members of their own sex. In , CONTACT WITH PARTNER

SOCIAL PREFERENCE for sexual partners CONTACT is demonstrated by female prairie voles WITH in three-hour tests. They prefer contact STRANGER with their partners over contact with male strangers. Female montane voles 0 10 20 30 40 50 60 70 80 actually spend more time alone. CONTACT WITH MALES (MINUTES)

Copyright 1993 Scientific American, Inc. SCIENTIFIC AMERICAN June 1993 103 30 minutes, females usually chose the Based on these studies, we hypothe- prairie vole. These receptors are scat- familiar male. sized that in prairie voles stimulation tered throughout the mammalian cen- In similar studies, Williams discovered experienced during mating, or perhaps tral nervous system. Witt found that that a female in her Þrst heat developed more slowly by touch and cohabitation, the distribution of oxytocin receptors a preference for a male if she was al- might release oxytocin. Oxytocin would, in prairie voles diÝered from the pat- lowed to live with him for at least 24 in turn, hasten the formation of social tern in rats. The diÝerences were espe- hours. If the pair copulated, however, bonds between males and females. cially striking within the limbic system, cohabitation produced clear social pref- Several recent Þndings support this the area of the brain involved in sexual erences in as few as six hours. These supposition. Witt injected oxytocin into and social behavior. Insel and his NIMH studies demonstrate that some aspect the central nervous system of females. colleague Larry E. Shapiro subsequent- of the sexual interaction hastens the on- As a result, the females became more ly showed that the distribution of oxy- set of a partner choice. We believe that sociable and less likely to Þght with tocin receptors in prairie voles and in hormones or neurochemicals released males, as compared with females that pine voles, another monogamous spe- during mating or cohabitation may ex- did not receive the oxytocin or females cies, diÝers from that in the polyga- plain the experimental results. that received the hormone administered mous montane and meadow voles. That into the peripheral circulation. The pos- the patterns of oxytocin receptors cor- ne clue to the identity of the itive social eÝects of oxytocin in the relate with monogamy further substan- hormones came from work by brain have now been documented in tiates the idea that oxytocin has an es- O Peter H. Klopfer of Duke Univer- other species. Witt found improved so- sential role in social organization. sity. He recognized that social bonds be- cial behavior in rats, and James Winslow tween mothers and their oÝspring were and Thomas R. Insel of the National In- he pair bonding in monogamy associated with the release of oxytocin stitute of Mental Health (NIMH) report- also leads mated pairs to guard and hypothesized that the compound ed similar results in squirrel monkeys. T one another or the shared terri- might be the hormone of Òmother love.Ó Williams examined the role of oxyto- tory. Reproductively naive prairie voles Niles Newton of Northwestern Universi- cin more directly. She repeated her pref- rarely Þght, but mated prairie voles can ty extended these observations to spec- erence tests on prairie voles whose cere- be extremely vicious toward strangers. ulate that maternal and sexual bonds bral ventricles were infused with oxy- Because studies have implicated testos- could be inßuenced by the secretion tocin. She found that females formed terone, a major reproductive hormone of the hormone. Oxytocin is produced rapid preferences for males if they were manufactured by the testes, in aggres- primarily as a result of breast or gen- exposed to oxytocin over a six-hour pe- sion in other animals, we initially hy- ital stimulation, such as that which riod. But when combined with a drug pothesized that testosterone might also occurs during mating, birth and lacta- that blocks the oxytocin receptors, oxy- be responsible for the postmating hos- tion. More recently E. Barry Keverne and tocin no longer exerted the social eÝect. tility in prairie voles. But in our labora- Keith M. Kendrick of the University of These results suggest that oxytocinÕs ac- tory Nicholas Hastings found that nei- Cambridge have shown that in sheep tion within the brain may be one of the ther castration nor testosterone injec- either vaginal stimulation or oxytocin physiological events that lead to the tions had an eÝect on male aggression treatments can speed the formation of formation of monogamous pairs. after mating. mother-infant bonds. Kerstin UvnŠs- Because the receptors for a hormone If testosterone does not regulate ag- Moberg of the Karolinska Institute in can regulate the behavioral eÝects of gression or mate guarding, then what Stockholm has demonstrated that even that hormone, we also looked at the does? Many pieces of evidence suggest- simple touch can release oxytocin. patterns of oxytocin receptors in the ed that vasopressin, a hormone best known for its role in regulating the hu- man bodyÕs water content, might play a role in mate guarding. First, Craig F. MALE Ferris of the University of Massachu- PARTNER setts Medical Center in Worcester and Elliott H. Albers of Georgia State Uni- versity had implicated vasopressin in territoriality and aggression in ham- sters. Second, vasopressin shares a mo- lecular structure similar to that of oxy- tocin; the molecules diÝer from one MALE FEMALE STRANGER another in only two of their nine amino FINAL acids. In addition, both hormones may POSITION be released under similar circumstanc- es, such as during sexual behavior and other social actions. The cellular and be- havioral functions of vasopressin and FEMALE oxytocin, however, tend to be antagonis- STARTING tic. Therefore, we reasoned that if oxy- POSITION tocin encourages social contact, perhaps vasopressin causes the antisocial or mate-guarding behavior shown by male prairie voles after they have mated. CHOICE TEST given to female prairie voles in the laboratory reveals a social pref- Winslow, Hastings and Insel tested erence for the mated males. Initially, females enter the cages of both strangers and this hypothesis in a collaborative study. their partners (represented by brown lines) and will mate with both. Within about In one experiment, males were injected 30 minutes, however, females tend to remain near the familiar male. before mating with a drug that blocks

104 SCIENTIFIC AMERICAN June 1993 Copyright 1993 Scientific American, Inc. BEHAVIORAL ESTRUS UTERUS GROWS TO PREPARE FOR PREGNANCY

HYPOTHALAMUS

ESTROGEN

OLFACTORY BULB ESTROGEN OVARY NE MALE LHRH UTERUS PHEROMONE ANTERIOR LUTEINIZING PITUITARY HORMONE

VOMERONASAL VAGINA ORGAN

HORMONAL CASCADE that triggers estrus in the female prai- (LHRH) are secreted and start the production of luteinizing rie vole begins when she sniÝs a male. The vomeronasal organ hormone. Luteinizing hormone reaches the ovaries via the picks up pheromones, stimulating the olfactory bulb. Norepi- bloodstream and stimulates them to produce estrogen. Estrogen nephrine (NE) and luteinizing hormoneÐreleasing hormone is then carried to the hypothalamus, where it induces estrus.

vasopressin receptors. The injections interactions between oxytocin and va- roids in rats. Exposure to stress during eliminated the increase in attacks di- sopressin are particularly apparent. the perinatal periodÑthe period of sex- rected toward strangers that usually fol- Although these studies have provid- ual diÝerentiation in mammalsÑinßu- lows mating. The eÝect was not a gen- ed strong clues to some of the neuro- ences subsequent reproductive devel- eral inhibition of aggression. The anti- endocrine mechanisms underlying mo- opment. For example, male rats that vasopressin drug did not stem attack nogamous behavior, a major puzzle per- have been stressed early in life tend behavior when given to males that had sists. Besides prairie voles, monogamy to show a more feminine pattern of de- completed their mating. In a separate ex- occurs in such diverse species as wild velopment as adults. Even the genital periment, Winslow and Insel infused va- dogs, tamarins and marmosets. Why, in anatomy is somewhat demasculinized. sopressin into a male while a female was a physiological sense, should such taxo- Apparently, high levels of stress dur- present. Such males then displayed in- nomically diÝerent mammals show the ing the perinatal period inhibit the nor- creased hostility toward male intruders. unique features of monogamy? mal secretion or the action of mascu- Vasopressin may also play a role in linizing hormones called androgens. male parental care. Recent Þndings by ne solution to this mystery Craig H. Kinsley, now at the University Maryam Bamshad, Melinda A. Novak may be found in the adrenal of Richmond, and Robert S. Bridges and Geert J. De Vries of the University O system and its eÝects on the of Tufts University demonstrated that of Massachusetts show the behavior in developing embryo. The adrenal sys- perinatal disturbance also increases the male prairie voles correlates with char- tem produces steroids called glucocor- probability that male rats will show, as acteristic changes in vasopressin levels. ticoids. Individuals release these hor- adults, parental care. Thus, in rats stress The experiments on the eÝects of oxy- mones, particularly corticosterone and appears to alter reproductive functions tocin and vasopressin on prairie vole cortisol, in response to stress. Yet the in a direction that is considered normal behavior suggest that these two com- endocrine systems of adult prairie voles in monogamous mammals. pounds have a much broader behav- and marmosets secrete unusually copi- We believe that adrenal activity in ioral signiÞcance than was previously ous amounts of glucocorticoids, even prairie voles might account in part for thought. Rather than just being a ho- when the animals are not under stress. their monogamy. Shapiro and Insel meostatic compound, vasopressin may Our work with prairie voles leads us to found that shortly after birth prairie have a more general role as a neuropep- hypothesize that the interactions be- voles have an unusually reactive adre- tide involved in eliciting parental care tween the adrenal and gonadal hor- nal system. Simply removing the moth- and defensive behavior with respect mones during early life might account er for a few minutes elevates the levels to self and family. Oxytocin, which has for some of the monogamous patterns of glucocorticoids in pups. In contrast, a well-established role in reproduction, that emerge later. the nonmonogamous montane voles might block the more primitive, anti- This supposition is based in part on and rats require greater disturbance social actions induced by vasopres- research pioneered by Ingeborg L. Ward before adrenal activity increases. sin, thus permitting social behaviors to of Villanova University. Ward document- In prairie voles the reactivity of the emerge. Finally, monogamy may be a ed the developmental eÝects of interac- adrenal glands, during late pregnancy reÞned expression of sociality in which tions between adrenal and gonadal ste- or the early postnatal period, might con-

Copyright 1993 Scientific American, Inc. SCIENTIFIC AMERICAN June 1993 105 cial behavior and bonds, which are at MONOGAMOUS CHARACTERISTIC HORMONE least as meaningful to humans as to prairie voles. Disruption of these bonds OXYTOCIN (RELEASED BY in humans, such as that which occurs in FEMALE BONDING TO MALE MATING OR CONTACT) the loss of a child, parent or spouse, can have disastrous consequences for men- tal and physical health. Yet we under- stand little about the behavioral physi- AGGRESSION BETWEEN VASOPRESSIN (HIGH) ology of the formation of such relation- MALES AFTER MATING ships in humans. In fact, the notion that social bonds have a ÒbiologyÓ has not been generally understood. SIMILAR SIZE AND APPEAR- CORTICOSTERONE (HIGH) It is tempting to speculate that oxy- ANCE OF MALE AND FEMALE TESTOSTERONE (LOW) tocin and vasopressin could also aÝect behavior. But the role of these hormones, even in animals, is diÛcult CORTICOSTERONE (HIGH) to document, and many obvious ques- MALE AND FEMALE tions remain unexplored. Most human PARENTAL CARE VASOPRESSIN (HIGH) research is limited to correlations be- tween changes in hormonal blood lev- els and behavior. SOCIAL REGULATION CORTICOSTERONE (HIGH) The paucity of human data raises con- OF REPRODUCTION cerns about the medical eÝect of these hormones, which are often adminis- tered as treatments. For example, oxy- FEATURES OF MONOGAMY in the prairie vole correlate with speciÞc hormones tocin is used to induce uterine contrac- and their amount in the body. The high and low levels of hormone are relative to tions in childbirth, and vasopressin is those found in nonmonogamous but closely related species. prescribed to treat bed-wetting in chil- dren. Actions that indirectly aÝect hor- monal levels, such as delivering a child tribute to the appearance in the adult of commitment from the male is a beneÞt by cesarean section or bottle-feeding, some of the deÞnitive characteristics of of monogamy because males can in- may also aÝect the amount of oxytocin monogamy, including reduced sexual crease their own reproductive success received by either a mother or her in- dimorphism and increased male paren- by caring for their oÝspring. Perhaps fant. Because oxytocin and vasopressin tal care. Preliminary experiments con- this strategy represents a probabilistic were at Þrst believed to act outside the ducted in our laboratory by Luci Rob- function. Monogamous males are in- brain, on tissues such as the uterus, erts oÝer some evidence. In those anal- creasing their Þtness in general while breast and kidney, little attention has yses, postpartum exposure to higher accepting the burden of rearing some been given to the behavioral conse- than normal levels of androgen reduced pups that are not their own. quences of these treatments or actions. the tendency of adult male prairie voles The theoretical implications of this We now know that oxytocin and vaso- to care for pups. Research in progress Þnding is uncertain. The prairie voles pressin are active in the central nervous is examining the importance of interac- we have studied in Illinois live in an system. Findings from animal research, tions between the adrenal and gonadal environment that provides abundant such as those described here for prairie systems in the development of other supplies of food, water and other es- voles, should encourage the biomedical components of monogamy. sential resources. We believe that mo- community to look more carefully at nogamy in prairie voles evolved when the potential eÝects of these powerful lthough we can now identify food was not plentiful. Under such con- hormones on behavior. some of the physiological under- ditions, monogamy might oÝer addi- A pinnings of monogamy in the tional beneÞts not evident in the habi- prairie vole, the ultimate (or evolution- tats we have examined. We are current- ary) cause of this behavior and its adap- ly comparing Illinois prairie voles with FURTHER READING tive signiÞcance remain unclear. It is those that live in a much harsher envi- THE COMPARATIVE PSYCHOLOGY OF MO- widely assumed that, from the perspec- ronment in Kansas. NOGAMY. D. A. Dewsbury in Nebraska tive of the oÝspring, having two parents Our studies also frequently elicit Symposium on Motivation, 1987, Vol. is better than having one. Trapping questions concerning the applicability 35, edited by D. W. Leger, pages 1Ð50; studies in prairie voles, however, have of our Þndings to human behavior. Mo- 1988. SOCIAL ORGANIZATIONS AND MATING not provided support for this assump- nogamy in Old World primates and in SYSTEM OF THE PRAIRIE VOLE, MICROTUS tion. Based on Þeld data from more humans probably takes on a diÝerent OCHROGASTER. L. L. Getz, B. McGuire, than 700 prairie vole families, single form from that described here for voles. J. Hofmann, T. Pizzuto and B. Frase in mothers are as successful as mother-fa- Yet there are some parallels. Clearly, hu- Social Systems and Population Cycles in ther pairs in rearing litters to maturity. man monogamous partners do not al- Voles. Edited by R. H. Tamarin, R. S. Ost- That sexual exclusivity is not a domi- ways stay sexually exclusive. Animals, feld, S. R. Pugh and G. Bujalska. Birk- nant feature of monogamy in prairie including humans, may be more partic- hauser Boston, 1990. voles also raises an evolutionary ques- ular in the selection of a social com- OXYTOCIN AND SOCIAL BONDING. C. S. Carter, J. R. Williams, D. M. Witt and tion. Parental care in mammals, and es- panion than in their choice of a sexu- T. R. Insel in Annals of the New York pecially in prairie voles, represents a al partner. Academy of Sciences, Vol. 652, pages signiÞcant investment of time and en- In addition, our research highlights 204Ð211; 1992. ergy. It is usually assumed that such the general signiÞcance of positive so-

106 SCIENTIFIC AMERICAN June 1993 Copyright 1993 Scientific American, Inc.