Integrative and Comparative Biology Integrative and Comparative Biology, pp. 1–14 doi:10.1093/icb/icx047 Society for Integrative and Comparative Biology

SYMPOSIUM

Paternal Care in Biparental : Intra- and Inter-individual Variation Wendy Saltzman,*,†,‡,1 Breanna N. Harris,§ Trynke R. De Jong,¶ Juan P. Perea-Rodriguez,k Nathan D. Horrell,*,† Meng Zhao*,‡ and Jacob R. Andrew*,‡ *Department of Biology, University of California, Riverside, California, USA; †Neuroscience Graduate Program, University of California, Riverside, California, USA; ‡Evolution, Ecology and Organismal Biology Graduate Program, University of California, Riverside, California, USA; §Department of Biological Sciences, Texas Tech University, Lubbock, Texas, USA; ¶Department of Behavioural and Molecular Neurobiology, University of Regensburg, Germany; kDepartment of Anthropology, Yale University, New Haven, CT, USA From the symposium “The Development and Mechanisms Underlying Inter-individual Variation in Pro-social Behavior” presented at the annual meeting of the Society for Integrative and Comparative Biology, January 4–8, 2017 at New Orleans, Louisiana.

1E-mail: [email protected]

Synopsis Parental care by fathers, although rare among mmmals, can be essential for the survival and normal devel- opment of offspring in biparental species. A growing body of research on biparental rodents has identified several developmental and experiential influences on paternal responsiveness. Some of these factors, such as pubertal matura- tion, interactions with pups, and cues from a pregnant mate, contribute to pronounced changes in paternal respon- siveness across the course of the lifetime in individual males. Others, particularly intrauterine position during gestation and parental care received during postnatal development, can have long-term effects on paternal behavior and contribute to stable differences among individuals within a species. Focusing on five well-studied, biparental species, we review the developmental and experiential factors that have been shown to influence paternal responsiveness, and consider their roles in generating both intra- and inter-individual variation. We also review hormones and neuropeptides that have been shown to modulate paternal care and discuss their potential contributions to behavioral differences within and between males. Finally, we discuss the possibility that vasopressinergic and possibly oxytocinergic signaling within the brain, modulated by gonadal steroid hormones, may represent the “final common pathway” mediating effects of developmental and experiential variables on intra- and inter-individual variation in paternal care.

Introduction behavior, neural development, endocrine function, In all , mothers invest heavily in producing emotionality, and cognition that persist into adult- and rearing their offspring. Fathers, in contrast, pro- hood (reviewed in Braun and Champagne 2014; vide care for their young in only an estimated 5–10% Bales and Saltzman 2016). of mammalian taxa (Kleiman and Malcolm 1981). Not surprisingly, our understanding of the proxi- Paternal care is often associated with social monog- mate basis of paternal care has lagged far behind that amy and can involve such behaviors as warming, of maternal care. In recent years, however, consider- feeding, protecting, retrieving, and grooming infants, able progress has been made in elucidating develop- depending on the species (Kleiman and Malcolm mental, social, hormonal, and neural determinants of 1981). In these biparental mammals, fathers can paternal behavior, primarily in biparental rodents. make important and lasting contributions to the sur- Much of this work has focused on determinants of vival and development of their offspring, including within- changes in males’ responses to pups effects on social, reproductive and aggressive across the lifespan: males of some species undergo

ß The Author 2017. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved. For permissions please email: [email protected]. 2 W. Saltzman et al. predictable changes in their behavioral responses to that may be broadly applicable across taxa, while pups, transitioning, in some cases, between aggression, highlighting observed differences among species. indifference, and nurturance at different life stages. Fourth, although these five species are classified as More recently, attention has begun to focus on biparental based on observations in both free-living between-individual variation in pup responsiveness, and captive , mechanistic studies of paternal addressing both its causes and consequences. While behavior have been performed almost exclusively un- it is becoming apparent that behavioral differences der highly controlled laboratory conditions. among fathers can translate into long-term differences Moreover, although some investigators characterize among their offspring, the environmental and organ- fathers’ behavior toward their own pups in the con- ismal sources of these differences have, for the most text of the entire family, the most common experi- part, not yet been addressed in an integrated manner. mental paradigm is to quantify behavior of a juvenile In this article, we discuss proximate determinants or adult male toward a single, experimentally pre- of males’ behavioral responses to pups in biparental sented, unrelated pup in the absence of other indi- rodents within the context of both intra- and inter- viduals or stimuli with which to interact. This individual variation in paternal and allopaternal care. paradigm is extremely useful for controlling and We begin by briefly reviewing findings on endocrine standardizing the physical and social environment and neuroendocrine influences on paternal behavior. during testing, but clearly is highly artificial and sim- Next, we describe developmental and experiential in- plistic. Thus, these studies almost certainly underes- fluences, focusing first on determinants of longitudi- timate the complexity and diversity of environmental nal changes in males’ responses to pups across the influences on paternal behavior. lifespan and second on sources of variation among Fifth, although nurturant behavior by any male individuals within a particular species. Finally, we toward a non-descendant, immature conspecific examine the possibility that the brain’s vasopressi- should, technically, be referred to as “allopaternal nergic systems, and possibly the oxytocinergic sys- behavior” or “allopaternal care,” we use these tem, in association with gonadal steroid hormones, phrases specifically to refer to nurturant or affiliative may represent a “final common pathway” mediating behavior performed toward conspecific pups by im- both within- and between-animal variation. mature males, since alloparents in these species are Several important points should be noted. First, likely to be juveniles. We use “paternal behavior” or we focus exclusively on naturally biparental rodents, “paternal care” to refer to such behavior by adult in which fathers routinely provide care for their off- males, whether or not the male and the pup are spring; we do not include studies of uniparental ro- related, to facilitate comparisons among animals in dents (e.g., Rattus, Mus) or other taxa, although different reproductive conditions. Similarly, we use findings from these groups might complement or the terms “allopaternal responsiveness” and “paternal extend those from biparental rodents. Second, al- responsiveness” to refer to the propensity of an im- though approximately 6% of rodent genera are esti- mature or adult male, respectively, to perform nur- mated to perform biparental care (Kleiman and turant behavior towards pups. Malcolm 1981), paternal care has been studied in Finally, it should be noted that different compo- only a handful of these species, most notably the nents of paternal or allopaternal behavior, such as prairie ( ochrogaster), mandarin vole huddling, licking/grooming, and retrieving a pup, (M. mandarinus), California mouse (Peromyscus cal- may be influenced by somewhat different neural, en- ifornicus), Campbell’s dwarf hamster (Phodopus docrine, and experiential factors. Furthermore, al- campbelli), and Mongolian gerbil (Meriones unguicu- though we discuss several studies that focus on latus). These species come from only two families infanticide, aggression toward pups should not be (: , mandarin vole, California assumed to represent, at a mechanistic level, the in- mouse, Campbell’s dwarf hamster; Muridae: verse of parental care (Dulac et al. 2014). Mongolian gerbil) and only one of the major rodent lineages (Fabre et al. 2012), and therefore may not be representative of biparental rodents in general. Neuroendocrine influences on paternal Third, in spite of the relatively close phylogenetic behavior relationships among these species, it should not be Males in biparental mammalian species undergo sys- assumed that influences on and mechanisms underly- tematic changes in hormonal and neuropeptide sig- ing the expression of paternal behavior are consistent naling during the transition to fatherhood, in across taxa. Thus, we attempt to develop a general association with pair formation, mating, cohabita- model of the proximate basis of paternal behavior tion with a pregnant female, and/or exposure to Paternal care in biparental rodents 3 infants. Some of these changes differ across species, extended amygdala (reviewed by Bales and and their functional significance, including potential Saltzman 2016). Moreover, central injection of AVP effects on paternal behavior, is generally unknown. promotes paternal behavior whereas central infusion Endocrine and neuroendocrine changes in fathers, as of AVP receptor antagonists has the opposite effect well as their possible functions, have been reviewed in both the obligately biparental prairie vole (Wang in detail recently (Saltzman and Ziegler 2014; Bales et al. 1994; but see Bales et al. 2004) and the facul- and Saltzman 2016). Therefore, we discuss them only tatively biparental meadow vole (M. pennsylvanicus; briefly here, focusing primarily on experimental Parker and Lee 2001). On the other hand, castration rather than correlational findings. of male prairie virtually eliminates AVP- The anterior pituitary hormone prolactin has been immunoreactivity in the LS and lateral habenular referred to as “the hormone of paternity” (Schradin nucleus (LHN) but does not alter paternal behavior, and Anzenberger 1999), as circulating or excreted levels indicating that AVP signaling in these areas is not are elevated in fathers of numerous biparental species essential for expression of paternal care (Lonstein and often correlate with males’ expression of allopater- and De Vries 1999). nalorpaternalbehavior(Saltzman and Ziegler 2014; Oxytocin, the “sister” neuropeptide of AVP, facili- Hashemian et al. 2016). To our knowledge, however, tates both maternal and allomaternal behavior in ro- prolactin has not been shown to have a causal influence dents (Bridges 2015; Kenkel et al. 2017). In contrast, on allopaternal or paternal care in any rodent species. little is known about effects of oxytocin on paternal Circulating testosterone concentrations, on the and allopaternal care. Paternal behavior in adult virgin other hand, are typically reduced in fathers and male prairie voles was inhibited by combined intra- have been shown convincingly to influence the ex- cerebroventricular treatment with an AVP receptor an- pression of paternal behavior; however, effects may tagonist and an oxytocin receptor antagonist, but not differ both within and among species. In California by either antagonist alone (Bales et al. 2004); however, mice, castration reduces and testosterone or estrogen the oxytocin receptor antagonist that was used was replacement restores parental behavior (Trainor and later found to more potently antagonize the AVP re- Marler 2001, 2002). Similar results were found in vir- ceptor than the oxytocin receptor (Kenkel et al., 2017), gin male Mongolian gerbils housed in same-sex groups leaving the results of this study difficult to interpret. (Martınez et al. 2015); however, virgin male gerbils More recently, treatment with a different oxytocin re- housed with a lactating female showed the opposite ceptor antagonist acutely inhibited parental behavior pattern (Clark and Galef 1999). Studies of prairie voles in adult male prairie voles in a dose-dependent man- have likewise yielded mixed results: castration either ner (described in Kenkel et al. 2017). reduced (Wang and De Vries 1993)ordidnotalter Very few studies have addressed possible organi- (Lonstein and De Vries 1999) responses to pups in zational effects of early-life hormone exposure on virgin males. Finally, castration did not alter paternal paternal behavior. In male prairie voles, however, behavior in a study of Campbell’s dwarf hamster fa- exposure to androgens, estrogens and oxytocin dur- thers (Hume and Wynne-Edwards 2005). ing the postnatal period appears to be vital for ex- The stimulatory effect of testosterone in California pression of paternal behavior in adulthood (reviewed mice is mediated by aromatization of testosterone to by Bales and Saltzman 2016). estrogen in the brain (Trainor and Marler 2002). Some or all of the hormones and neuropeptides Few other studies have investigated effects of estro- thought to influence paternal behavior might also gen on allopaternal or paternal behavior in juvenile play key roles in determining within- and between- or adult males. In prairie voles, however, Cushing individual differences in paternal and allopaternal et al. (2008) found that experimentally increasing responsiveness. Therefore, in describing sources of the expression of estrogen receptor a (ERa) in the variation in paternal and allopaternal care, below, medial amygdala (MeA) via a viral vector inhibited we point out, where possible, hormonal and neuro- parental behavior in adult males, while increasing peptide differences that correlate with—and might ERa expression in the bed nucleus of the stria ter- underlie—this variation (Fig. 1). minalis (BNST) had no effect (Lei et al. 2010). The neuropeptide vasopressin (AVP) has been Within-animal changes in behavioral strongly implicated in the expression of paternal responses to pups behavior. Both within and among rodent species, paternal behavior correlates with patterns of AVP- Age immunoreactivity and AVP binding, particularly in Allopaternal or paternal responsiveness in biparental the lateral septum (LS) and other parts of the rodents is modulated by males’ age and/or 4 W. Saltzman et al.

Intrauterine Parental care Exposure Cues from Cues from Paternal position received to pups mate mate & pups experience Stress/ T,E AVP,OT, OT AVP,T AVP,T anxiety ? Cort

AVP, Cort

Prenatal Early Juvenile/ Mated New Experienced postnatal young adult father father

Fig. 1 Experiential effects on allopaternal and paternal responsiveness acting at different life stages in biparental rodents, as well as hormones and neuropeptides that have been implicated correlationally and/or experimentally. Bold text and thick arrows indicate cues or experiences routinely encountered during reproduction. Italicized text and dashed arrows indicate early-life influences that may differ across animals and contribute to stable inter-individual differences in (allo)paternal responsiveness. Non-italicized text and thin arrows indicate factors that might contribute to either intra- or inter-individual differences over a variable time span. T: testosterone; E: estrogen; AVP: vasopressin; OT: oxytocin; and Cort: corticosterone. developmental stage (e.g., juvenile, adult) as well as pups declines across the juvenile period and remains reproductive status (e.g., sexually naı¨ve, mated, fa- low into adulthood (Gubernick and Laskin 1994); ther), with males typically showing (1) high levels however, testosterone levels have not been character- of allopaternal care and little or no infanticide to- ized in young males across the pubertal period. In ward younger siblings or unfamiliar pups, as juve- contrast to Mongolian gerbils, prairie voles, and niles, (2) a decrease in nurturant responses to pups California mice, sexually naı¨ve male Campbell’s and, in some cases, increased rates of infanticide, as dwarf hamsters do not show significant changes in sexually naı¨ve adults, and (3) pronounced paternal affiliative behavior toward pups from the early post- responsiveness and little or no infanticide toward pubertal period to mid-adulthood (Gregg and either descendant or unrelated pups during cohabi- Wynne-Edwards 2005; Vella et al. 2005). Again, tation with a pregnant or lactating female. In male however, hormonal changes across this period have Mongolian gerbils, for example, (allo)paternal re- not been characterized. sponsiveness, as measured by time spent with pups as well as preference for pups over an empty nest, Cues from the mate declines from the age of weaning until young adult- Males’ responses to pups can be influenced dramat- hood, and thereafter begins to increase again (Clark ically by cues from or interactions with a female and Galef 2001), whereas infanticide toward unre- mate. Bamshad et al. (1994) compared behavioral lated pups shows the opposite pattern (Elwood responses to unfamiliar pups among male prairie 1980). Interestingly, testosterone levels follow voles 0, 2, 13 or 21 days after they were paired roughly the same pattern across development as in- with an adult female, or 6 days following the birth fanticide (Probst 1987), such that the peak in testos- of their first litter (approximately 30 days after pair- terone levels (75 days of age) coincides with the ing). Males’ duration of time spent in paternal be- peak in infanticide and the nadir in nurturant be- havior toward an unfamiliar pup was positively havior toward pups (Clark and Galef 2001). correlated with length of cohabitation with a female, In prairie voles, males and females of all ages tend and males whose mates were in the late stages of to behave affiliatively toward pups. Nonetheless, pregnancy or early stages of lactation were markedly they, like Mongolian gerbils, exhibit a shift in behav- more likely than other males to behave paternally. ioral responses to pups around the age of puberty: Circulating testosterone concentrations did not differ for virgin males and females combined, individuals among the groups of males, but AVP- older than 40 days both retrieve and attack unfamil- immunoreactivity in the LS and LHN changed with iar pups more than do individuals 40 days and duration of cohabitation: AVP-immunoreactivity in younger; testosterone levels in males rise at around both regions declined significantly within 3 days af- 45 days of age (Roberts et al. 1999). In virgin male ter pair formation and increased gradually across the California mice, affiliative behavior toward unrelated mate’s pregnancy. In a separate study of prairie Paternal care in biparental rodents 5 voles, 3 days of cohabitation with a female elevated hand, fathers that are permanently separated from males’ plasma testosterone levels and increased ex- both their mate and pups on the day of parturition pression of AVP mRNA in the BNST, a likely source are much less likely to behave paternally toward an of AVP detected within fibers in the LS and LHN unrelated pup 3 days later, and more likely to com- (Wang and De Vries 1993). Thus, the authors spec- mit infanticide, compared to males that remain with ulated that cohabitation with a female increases syn- their mate (Gubernick and Alberts 1989). Thus, con- thesis of AVP in the BNST and release of AVP in the tinuous exposure to the mate, but not to the pups, is LS and LHN, potentially leading to enhanced pater- important for the maintenance of paternal respon- nal responsiveness. siveness in new fathers. Finally, paternal responsive- In Mongolian gerbils, similar to prairie voles, pa- ness is maintained in California mouse fathers that ternally inexperienced males show a sharp reduction are housed without their mate and pups but are ex- in infanticide and an increase in paternal behavior posed to urine from their mate, indicating that che- toward an unrelated pup across the course of their mosensory cues from the mate, but not necessarily mate’s first pregnancy (Elwood 1977). The same pat- from the pups, are important for the maintenance of tern is seen in virgin male gerbils housed with a paternal responsiveness in fathers during the early female during roughly the second half of her preg- postpartum period (Gubernick 1990). nancy (Clark et al. 2001). In expectant first-time fa- thers, this inhibition of infanticide appears to be mediated largely by physical and visual contact Previous exposure to pups with the mate: males that were prevented from en- Interactions with younger siblings or unrelated pups, gaging in physical contact with their mate, especially either during the juvenile period or in adulthood, those that were also prevented from seeing the mate, may contribute to both intra- and inter-individual during the last few days of pregnancy were signifi- differences in (allo)paternal responsiveness in some cantly more likely to kill an experimentally pre- biparental rodents. Virgin male prairie voles that sented, unrelated pup than expectant fathers with have lived with younger siblings are significantly full physical and sensory access to their mates more likely to behave paternally to an unfamiliar (Elwood 1980). Olfactory cues from the pregnant pup than those that have no experience with younger mate, on the other hand, did not appear to play a siblings, although most males in both conditions be- role in the inhibition of infanticide in expectant fa- have paternally (Roberts et al. 1999). Similarly, in thers (Elwood and Ostermeyer 1984). Interestingly, California mice, virgin males that have lived with experienced fathers, in contrast to first-time fathers, their parents and younger siblings show high levels show pronounced paternal behavior toward unre- of paternal care toward an unrelated pup, compared lated pups whether or not the mate has been re- to virgin males that have lived with only their par- moved, indicating that experience with pups has a ents and a littermate but no younger siblings, or long-term facilitatory effect on the maintenance of with only a littermate (Gubernick and Laskin paternal responsiveness (Elwood and Ostermeyer 1994). This effect is age-limited, however: nurturant 1984). California mouse fathers whose mates have behavior toward experimentally presented pups de- recently given birth to their first litter show more creases from the young juvenile period to adulthood, paternal behavior toward an unrelated pup than do and previous cohabitation with younger siblings does age-matched virgin adult males housed with another not affect pup-directed behavior in older juveniles or male, whereas responses to pups do not differ mark- adults (Gubernick and Laskin 1994). In Mongolian edly between new fathers and males whose mates are gerbils, unlike both prairie voles and California mice, pregnant with their first litter (De Jong et al. 2009). cohabitation with younger siblings does not appear Finally, in adult male mandarin voles, both mating/ to alter behavioral responses to pups in adult virgin cohabitation with a female and fatherhood Increase males, in terms of frequency of infanticide (Elwood expression of some components of paternal behavior 1980; Saltzman et al. 2009); however, effects on pa- (Song et al. 2010). rental behavior per se have not been evaluated. Cues from a mate can facilitate not only the onset Species also differ in whether repeated exposure to but also the maintenance of paternal behavior. pups during adulthood facilitates the onset of pater- California mouse fathers housed with their mates nal care. In a recent study of California mice, Horrell tend to be highly paternal toward an unrelated pup et al. (2017) found that adult virgin males with no 3 days after the birth of their litter, regardless of previous exposure to pups engaged in less paternal whether their own litter is removed on the day of behavior than new fathers, as observed in other stud- birth or remains with the parents. On the other ies, but that virgins’ paternal responsiveness was 6 W. Saltzman et al. increased by repeated, brief (20-min) exposure to supraphysiological levels, were nominally slower to pups: after 2-3 exposures, virgin males’ behavioral contact their own pups when tested in the absence responses to pups did not differ from those of new of the mate, but no effects of corticosterone were fathers. In adult, virgin male mandarin voles, even a seen on the frequency or duration of fathers’ inter- single, 10-min exposure to an unrelated pup in- actions with their offspring when tested either with creased paternal responsiveness to an unrelated pup or without the mate present (Harris et al. 2011). 1 week later (Song et al. 2010). In contrast, repeated Harris et al. (2013) evaluated effects of a 7-day 10-min exposure to a pup did not reliably alter pa- chronic variable stress paradigm on paternal behav- ternal behavior in adult virgin Campbell’s dwarf ior of California mouse fathers beginning 1-3 days hamsters, even after four exposures (Vella et al. after the birth of their first litter. Despite the clear 2005). Similar to dwarf hamsters, adult, virgin male physiological effects of the stress paradigm (De Jong prairie voles showed no change in parental behavior et al. 2013), effects on paternal behavior were mod- after 3 consecutive 20-min exposures to pups over 6 est. Most notably, upon being reunited with their days (Kenkel et al. 2013). mate and pups following acute exposure to a re- straint stressor, fathers spent less time engaging in paternal behavior and grooming their female mate, Stress and anxiety compared to fathers that were removed from and Age, cues from a mate, and experience with pups are returned to their families without being exposed to predictable developmental factors that influence the an additional stressor. Aside from these behavioral expression of affiliative behavior toward pups over differences immediately following stressor expo- the life of an individual animal in several biparental sure, only minor behavioral differences were ob- rodent species. In contrast, affective factors, espe- served between fathers in the chronic variable cially stress and anxiety, may exert more subtle and stress and control conditions. These results suggest possibly more transient effects on males’ responses to that males increase pup- and mate-directed behav- pups, and are likely to differ among individuals as a iors following a brief separation from their family result of environmental and genetic influences. Thus, (consistent with Bredyetal.2004), but that stress stress and anxiety could potentially contribute to can block this effect. Interestingly, chronically both within- and between-animal differences in allo- stressed fathers in the same study had increased paternal and paternal care. levels of AVP mRNA in the paraventricular nucleus Maternal care by rodent mothers can be impaired of the hypothalamus (PVN) compared to non- by exposure to acute or chronic stressors or to ex- stressed fathers and also showed higher levels of perimental increases in corticosterone or autogrooming following stress. Both PVN expres- corticotropin-releasing factor (e.g., Nephew and sion of AVP and autogrooming are linked to anx- Bridges 2011; Klampfl et al. 2014; Pereira et al. iety in rodents (Ferre´ et al. 1995), suggesting that 2015). Relatively little is known, however, about ef- anxiety, as well as stress, might inhibit paternal fects of stress or stress-related hormones or neuro- behavior. peptides on allopaternal or paternal behavior, and in In addition to experimental studies, several corre- the few studies addressing this topic, results have lational studies, mostly in California mice, provide mostly been fairly subtle. evidence for links between paternal responsiveness Two studies have found that acute stress or acute and stress or anxiety. Latency of adult male increases in stress hormones can alter males’ inter- California mice to approach a novel object (a mea- actions with pups. Virgin male prairie voles tested sure of neophobia) was positively and significantly with an unfamiliar pup for 10 min were more likely correlated with latency to perform paternal behavior to behave paternally and spent more time engaged in toward an unfamiliar pup (Chauke et al. 2012). In paternal behavior if they underwent a swim stressor the same species, Lambert et al. (2011) found a neg- 45 min before exposure to the pup than if they did ative correlation between the total time that virgin not. Moreover, males’ circulating corticosterone con- males spent performing paternal behavior toward an centrations immediately after interactions with the unrelated pup and the expression of the immediate- pup correlated negatively with the amount of time early gene Fos in the PVN, a region strongly in- spent licking and grooming the pup (Bales et al. volved in the stress response, following a 10-min 2006). In contrast, California mouse fathers exposure to the pup. Finally, De Jong et al. (2012) that underwent acute corticosterone treatment, found that latencies of adult virgin males to sniff an which raised their circulating corticosterone to unfamiliar pup correlated positively with expression Paternal care in biparental rodents 7 of AVP mRNA in the PVN and negatively with urine Between-animal differences in marking in the center of a novel environment; both behavioral responses to pups of these measures are considered indices of anxiety. In the biparental rodents that have been studied, Taken together, these findings suggest that paternal behavioral responses to pups may differ markedly responsiveness may be inversely related to anxiety among individual sexually naı¨ve males as well as and stress-reactivity, at least in the short term. among individual fathers within a species. Intriguingly, two studies have provided evidence Although these differences are likely to arise in part that consumption of placenta by adult male from genetic influences, early-life experience can also California mice can reduce males’ anxiety and in- contribute to long-term differences among individ- crease their willingness to interact with pups. First, uals. To date, three important sources of inter- Perea-Rodriguez (2016) treated adult virgin males individual variation in paternal responsiveness have with either near-term placenta homogenized in oil been identified: intrauterine position during gesta- or oil alone via oral gavage, and subsequently char- tion, parental care received during the pre-weaning acterized behavioral and neural responses to a pup period, and early-life handling. or novel object. Seven hours after treatment, males that had been administered placenta had lower laten- Intrauterine position cies to approach both a novel object and a pup, as well as reduced Fos expression in the dorsal BNST Working with Mongolian gerbils, Clark et al. (1998) following exposure to either stimulus. In a separate demonstrated that intrauterine position can influ- study, Perea-Rodriguez (2016) found that both vir- ence males’ behavioral responses to pups in adult- gin males and new fathers treated with placenta trav- hood, as well as potential hormonal mediators of eled longer distances in an open field 4 h after paternal behavior. During the first 20 days following treatment, compared to control males treated with the birth of their first litter, fathers that had gestated oil vehicle; distance traveled in an open field is con- between two sisters (2-F males) had significantly sidered an inverse measure of anxiety in rodents more contact with pups than fathers that had ges- (Sestakova et al. 2013). As fathers from several bipa- tated between two brothers (2-M males). In addi- rental rodent species [Campbell’s dwarf hamster tion, 2-M male gerbils had higher circulating (Jones and Wynne-Edwards 2000), California mouse testosterone levels in adulthood than 2-F males (Lee and Brown 2002; Perea-Rodriguez and Saltzman (Clark et al. 1992). Studies in other rodents indicate 2014), prairie vole (K. L. Bales, personal that differences in males’ intrauterine position are communication.)] have been observed eating their associated with differences in exposure to androgens mates’ placentas, this behavior might reduce anxiety and estrogens during gestation (Vom Saal et al. 1983; in new fathers and increase their willingness to in- Pei et al. 2006), as well as with differences in expres- teract with their newborn pups. sion of androgen receptors and at least one steroido- In sum, several findings from biparental rodents genic enzyme, 5a-reductase, in peripheral suggest that males—especially virgin males—with reproductive organs (Nonneman et al. 1992; re- high levels of anxiety-related behavior or neuroendo- viewed in Ryan and Vandenbergh 2002). Therefore, crine correlates of anxiety may engage in less nur- intrauterine position likely affects males’ parental be- turant behavior toward unrelated pups, and that this havior in adulthood by modulating exposure to ste- behavior may be inhibited by acute or chronic stress roid hormones during both early development and as well as by the stress-related glucocorticoid hor- adulthood. Effects of intrauterine position on pater- mones. These effects of stress and glucocorticoids nal behavior have not, to our knowledge, been in- have mainly been fairly subtle, however; to our vestigated in other species. knowledge, no studies have found marked effects of stress on either paternal care or infanticide. Parental care received Thus, stress and anxiety might play a limited role A second early-life influence on the expression of in determining intra- and inter-individual differences paternal care in biparental rodents is the quality in behavioral responses to pups in biparental ro- and/or quantity of care that fathers received from dents. Importantly, however, almost all of the studies their own parents (reviewed by Braun and on associations between paternal care and stress or Champagne 2014; Bales and Saltzman 2016). anxiety have been performed in adult California Broadly, males that were reared uniparentally (i.e., mice; thus, the extent to which these findings can without their fathers present) subsequently perform be generalized to other species or age classes is not less paternal care toward their own offspring than do known. males reared biparentally (i.e., by both parents). 8 W. Saltzman et al.

In prairie voles, males raised by only their mothers California mouse fathers and, subsequently, by their spend significantly less time licking and grooming sons. When tested in their home cage with one of their own pups during the first 6 postpartum days their pups, castrated fathers took significantly longer than fathers that were raised by both parents (Ahern than intact fathers to approach and begin caring for et al. 2011). On the other hand, virgin male prairie their pups, and spent significantly less time huddling voles raised uniparentally show no differences in be- and grooming their pups. These differences were re- havioral responses to pups, compared to virgins peated in the subsequent generation: gonadally intact reared biparentally (Ahern and Young 2009). sons of castrated fathers spent significantly less time Importantly, prairie vole mothers engage in compa- huddling and grooming their pups, and performed rable levels of infant care whether rearing pups with significantly more retrievals of pups, than sons of or without a mate; consequently, pups reared by sin- intact fathers. Although neural and endocrine mea- gle mothers receive less total parental care than those sures were not characterized in the offspring in this raised by both parents (Ahern and Young 2009). study, another study in the same species found that Thus, effects of father absence on behavioral devel- sons of castrated males had lower AVP- opment in pups do not appear to be mediated by immunoreactivity in the dorsal region of the BNST changes in maternal behavior. As in prairie voles, compared to sons of intact males, as well as higher mandarin vole (Jia et al. 2011; Yu et al. 2015) and AVP-immunoreactivity in the PVN (Frazier et al. Mongolian gerbil (Gromov 2009, cited by Braun and 2006). Together, these studies suggest that individual Champagne 2014) fathers raised by only their moth- differences in paternal behavior may be transmitted ers perform less paternal behavior toward their own across generations, potentially mediated by changes offspring than do fathers that were raised by both in AVP signaling within the brain. parents. Perkeybile and colleagues (2013) compared allo- Even among males raised by both parents, the paternal behavior in prairie voles that had received early rearing environment can influence paternal different patterns of care from their own parents. care performed towards pups in adulthood. In a se- Offspring of “high-contact” parents experienced ries of studies on the California mouse, Marler and high total levels of contact with their parents but colleagues have demonstrated that experimental ma- relatively low levels of contact with their fathers nipulation of parental care received by young males specifically, compared to offspring of “low-contact” can lead to long-term changes in these males’ behav- parents. When tested with an unfamiliar pup shortly ior toward their own offspring. Bester-Meredith and after weaning, sons of high-contact pairs engaged in Marler (2003) compared paternal behavior of male more non-huddling contact with the pup than sons California mice that were raised either by conspecific of low-contact pairs; no other behaviors differed foster parents or by foster parents of a congeneric reliably between the two groups. Cross-fostering species, the white-footed mouse (Peromyscus leuco- studies demonstrated that this effect was mediated pus), in which fathers provide little care of offspring. primarily by experiential, rather than genomic, Male California mice reared by white-footed mouse transmission of behavior, as juvenile males’ behav- pairs received fewer retrievals by their foster fathers ioral responses to a pup correlated with several and subsequently, as adults, spent less time retrieving components of parental behavior that they had re- their own pups, compared to males reared by ceived from their foster parents (Perkeybile et al. P. californicus pairs. Across all groups of male off- 2015). In addition, binding of AVP and oxytocin spring, AVP-immunoreactivity in the BNST correlated in the BNST of juvenile males correlated signifi- positively with paternal care performed toward their cantly or marginally, respectively, with several as- own pups; however, cross-fostering across species did pects of parental care received, as well as with not alter expression of AVP in any of the brain re- AVP and oxytocin binding in their biological gions examined (BNST, MeA, PVN, supraoptic nu- parents. cleus of the hypothalamus [SON]). In another cross- Taken together, these studies demonstrate that the fostering study, male meadow voles raised by prairie quality and/or quantity of parental care that males vole foster parents received higher levels of parental receive during early development influences their be- care during pre-weaning development and subse- havioral responses to pups—both their own off- quently performed more paternal behavior toward spring and unrelated pups—in adulthood and the their own offspring, compared to male meadow voles juvenile period, and that these differences in parent- raised by conspecific foster parents (McGuire 1988). ing style can be transmitted to the next generation. Gleason and Marler (2013) characterized paternal They also indicate that these developmental effects behavior performed by castrated and sham-castrated on paternal behavior are associated with, and Paternal care in biparental rodents 9 perhaps mediated by, changes in oxytocin and AVP well as in perpetuating individual differences in pa- signaling within the brain. ternal behavior across generations (see Fig. 1). Numerous correlational and experimental studies in Early-life handling the prairie vole, meadow vole, mandarin vole, and Alloparental responsiveness in male prairie voles can California mouse (reviewed by Frazier et al. 2006; be modulated by experimental handling during the Bales and Saltzman 2016; Perkeybile and Bales 2017) early postnatal period. Bales et al. (2007) found that have implicated AVP, especially vasopressinergic pro- males who, along with their families, were manually jections from the MeA and BNST to the LS and LHN, removed from their cage during routine cage-changing in regulating paternal care. The experiments reviewed in the first two postnatal days were more likely to above further indicate that AVP signaling is altered by perform alloparental behavior, and engaged in higher several experiential factors that simultaneously influ- levels of alloparenting, when tested with an unrelated ence paternal behavior. In prairie voles, for example, pup during the juvenile period, compared to males interactions with a pregnant female both increase pa- that, with their families, were removed from the cage ternal behavior and alter AVP-immunoreactivity in in a plastic cup rather than by hand. Notably, pups adult males, such that high paternal responsiveness were not touched directly in either handling paradigm: appears to be associated with increased synthesis of pups in the “manual” condition dangled from their AVP in the MeA and BNST as well as increased re- mothers’ nipples while the mother was lifted and lease of AVP in the LS and LHN (Bamshad et al. moved to a new cage, and those in the “cup” condi- 1994). In the same species, castration reduces and tion were scooped into the cup before being trans- testosterone replacement increases paternal behavior ferred to a new cage. In a separate study, juvenile as well as AVP expression in the MeA, BNST, LS males that, with their families, had undergone the and LHN (Wang and De Vries 1993; De Vries and “manual” handling paradigm before postnatal Day 7 Miller 1999). In California mice, low levels of paternal showed more alloparental behavior than males that behavior in adult sons of castrated males are associ- had not undergone handling manipulation during ated with low expression of AVP in the dorsal BNST this period (Bales et al. 2011). (Frazier et al. 2006; Gleason and Marler 2013). Early handling in the latter study also altered AVP expression in the brain has not, to our knowl- binding of oxytocin in the BNST and density of edge, been characterized in male rodents from known oxytocin-containing cell bodies in the SON; how- intrauterine positions; however, given that 2-M male ever, effects of handling on oxytocin signaling did gerbils, compared to 2-F males, are exposed to high not correspond closely to effects on alloparental be- prenatal and adult androgen levels and have low pa- havior. In contrast, early handling did not signifi- ternal responsiveness in adulthood, and given the pro- cantly alter AVP expression or AVP binding to V1a nounced effects of gonadal steroids on the BNST- receptors in any of the five brain regions studied. MeA-LS-LHN AVP pathway (De Vries and Miller Importantly, both mothers and fathers increased 1999), this system seems likely to be affected by in- their pup-directed behaviors immediately after the trauterine position. Increases in testosterone coincide handling manipulation on postnatal Day 1, suggest- with decreases in (allo)paternal behavior in virgin ing that effects of handling on pups’ behavioral and male Mongolian gerbils (Elwood 1980; Probst 1987; neuroendocrine development might have been medi- Clark and Galef 2001) and prairie voles (Roberts et al. ated by effects on the parental care that they received 1999) suggest that maturational effects, too, could po- (Bales et al. 2007, 2011). tentially be mediated by changes in AVP signaling. Finally, although prior experience with pups promotes Correlated variation in males’ (allo)paternal responsiveness in prairie voles (Roberts et al. 1999) and California mice (Gubernick and behavioral responses to pups and Laskin 1994), possible effects on testosterone and neuropeptide signaling AVP signaling have not been examined, to our As described above, the mechanisms underlying de- knowledge. velopmental and environmental influences on In addition to associations between paternal behav- (allo)paternal responsiveness include multiple, inter- ior and expression of AVP in the gonadal-steroid- acting endocrine and neuroendocrine systems and dependent BNST-MeA-LS-LHN pathway, several neural pathways. Nonetheless, AVP signaling within studies have found negative associations between pa- the brain, possibly modulated by gonadal steroids, ternal behavior and expression of AVP or AVP mRNA appears to play a key role in mediating effects of in the PVN (prairie vole: Perkeybile et al. 2013, 2015; the social environment on paternal behavior, as mandarin vole: Wang et al. 2014; California mouse: 10 W. Saltzman et al.

Frazier et al. 2006; De Jong et al. 2012). PVN AVP can In summary, experimental and correlational stud- facilitate activation of the hypothalamic-pituitary- ies have implicated both AVP and oxytocin in the adrenal axis in response to stress and promotes regulation of paternal responsiveness in rodents. In anxiety-related and depression-like behavior particular, fairly strong evidence indicates that AVP (Caldwell et al. 2008). Correspondingly, in three of may be an important determinant of both intra- and these studies, inter-individual differences in paternal inter-individual differences in (allo)paternal care. behavior and PVN AVP expression were associated Evidence for a role of oxytocin is less consistent; how- with differences in circulating corticosterone concen- ever, this difference might arise, in part, from the trations and/or anxiety-like behavior (Frazier et al. traditional focus on AVP as influencing prosocial be- 2006; De Jong et al. 2012; Perkeybile et al. 2013, 2015). havior in males and oxytocin playing a comparable Oxytocin, in addition to AVP, has been implicated role in females (e.g., Lim and Young 2006). More in the activation of allopaternal and paternal behavior, studies are needed in order to further elucidate the especially in voles. To our knowledge, only two stud- specific effects of both neuropeptides on pup-directed ies, both in prairie voles, have directly manipulated behavior in male rodents, as well as their contribu- oxytocin signaling and evaluated effects on behavioral tions to intra- and inter-individual variation. responses to pups. As described above, these studies found that an oxytocin receptor antagonist, either alone (Kenkel et al. 2017) or in conjunction with an Conclusions and future directions AVP receptor antagonist (Bales et al. 2004), reduced The studies reviewed above indicate that allopaternal the expression of paternal behavior in adult males. and paternal responsiveness in male rodents can be Additionally, several correlational studies have modulated by numerous developmental and envi- found that differences among males in (allo)paternal ronmental factors. Several of these, including age, responsiveness were associated with differences in experience with pups, and cues from a reproductive oxytocin signaling. Mandarin vole fathers that were female, are relatively consistent across most individ- treated with cocaine for 4 days, followed by 24 h of uals and can result in predictable changes in males’ withdrawal, spent less time in contact with and lick- behavioral responses to pups across the life span, as ing/grooming their own pups, and had fewer oxyto- males transition from juveniles to sexually naı¨ve cin- (and AVP)- immunoreactive neurons in the adults to mated males to fathers. Others, including PVN, compared to control males treated with saline stress and anxiety, are likely to be more variable over (Wang et al. 2014). In another study of mandarin time and among individuals, and may have relatively voles, adult males that either had been previously short-term effects on behavioral responses to pups; exposed to an unrelated pup, had been mated with longer-term effects are likely to occur as well but a female, or were new fathers showed more nurtur- have not yet been investigated. Finally, environmen- ant behavior toward an unrelated pup and had more tal differences among males during early develop- oxytocin-immunoreactive neurons in the SON and/ ment, particularly intrauterine position and or PVN, compared to virgin males (Song et al. patterns of parental care received, can have persistent 2010). Among virgin male mandarin voles, individ- effects on paternal responsiveness, potentially leading uals that exhibited high levels of paternal behavior to stable inter-individual differences in adults. toward unrelated pups also had high expression of Both within- and between-individual differences oxytocin in the PVN and SON, compared to males in (allo)paternal responsiveness have been linked to that engaged in low levels of paternal behavior (Li differences in several hormonal and neuropeptide et al. 2015). Similarly, in the facultatively biparental signaling systems, including androgens, estrogens, meadow vole, new fathers that behaved paternally glucocorticoids, oxytocin, and AVP. Several lines of when tested with one of their own pups had higher evidence, however, point to AVP as playing a partic- levels of oxytocin binding in the BNST, LS, lateral ularly prominent role. AVP signaling within the sex- amygdala, and accessory olfactory nucleus, compared ually dimorphic BNST-MeA-LS-LHN pathway, to sexually and parentally inexperienced males tested which is positively influenced by testosterone, com- with an unfamiliar pup (Parker et al. 2001). On the monly shows a positive association with paternal re- other hand, effects of cross-fostering on alloparental sponsiveness, whereas AVP activity within the PVN, behavior in juvenile male prairie voles were not ac- associated with stress and anxiety, typically shows a companied by significant changes in expression of negative correlation with paternal care. The similar- oxytocin receptors in any of the five brain regions ity of these patterns across several rodent species examined (Perkeybile et al. 2015). suggests that at least some components of the Paternal care in biparental rodents 11 proximate regulation of paternal care have evolved Funding by a common mechanism in biparental rodents, at This work was supported by the U.S. National least within the Muridae and Cricetidae. Institutes of Health [R21MH087806, R21HD075021] The mechanisms by which early-life experience and the U.S. National Science Foundation [IOS exerts long-term effects on (allo)paternal responsive- 1256572, IOS 1407370]. The symposium from which ness and neuroendocrine activity are not yet known. this paper originated was supported by the National In rats, patterns of maternal care affect behavioral Science Foundation [IOS-1634027], Society of and neuroendocrine function in offspring through Integrative and Comparative Biology Division of epigenetic regulation of gene expression Animal Behavior, Division of Comparative (Stolzenberg and Champagne 2016), and a similar Endocrinology, Division of Ecology and Evolution, mechanism is likely to subserve early-life effects of and Division of Neurobiology, Neuroethology, and paternal behavior as well. Elucidation of the specific Sensory Biology. genes that may be epigenetically regulated by pater- nal care, as well of interactions between these persis- tent epigenetic effects and shorter-term experiential References influences, will yield important insights into the Ahern TH, Young LJ. 2009. The impact of early life family transgenerational transmission of paternal behavior. structure on adult social attachment, alloparental behavior, Finally, the functional significance of inter- and the neuropeptide systems regulating affiliative behav- individual variation in paternal care is not clear. iors in the monogamous prairie vole (Microtus ochrogaster). Infant-directed behavior expressed by a father can Front Behav Neurosci 3 published online (doi: 10.3389/ influence the morphological, behavioral, endocrine, neuro.08.017.2009). Ahern TH, Hammock EA, Young LJ. 2011. Parental division neuroendocrine, cognitive, and affective development of labor, coordination, and the effects of family structure of his young; however, whether this ultimately affects on parenting in monogamous prairie voles (Microtus reproductive success for the father and/or his off- ochrogaster). Dev Psychobiol 53:118–31. spring is, for the most part, unknown. Under natural Bales KL, Boone E, Epperson P, Hoffman G, Carter CS. 2011. conditions, the expression and consequences of dif- Are behavioral effects of early experience mediated by oxy- ferent patterns of allopaternal and paternal care may tocin? Front Psychiatr 2:24. depend heavily on the environment. For example, Bales KL, Kim AJ, Lewis-Reese AD, Carter CS. 2004. Both oxytocin and vasopressin may influence alloparental behav- high levels of nest attendance and contact with off- ior in male prairie voles. Horm Behav 45:354–61. spring might be beneficial in cold climates, in which Bales KL, Kramer KM, Lewis-Reese AD, Carter CS. 2006. these behaviors may be crucial for maintaining pups’ Effects of stress on parental care are sexually dimorphic body temperature, whereas lower levels of nest atten- in prairie voles. Physiol Behav 87:424–9. dance might facilitate foraging and food intake by the Bales KL, Lewis-Reese AD, Pfeifer LA, Kramer KM, Carter CS. father or alloparent, especially under conditions of 2007. Early experience affects the traits of monogamy in a limited or highly dispersed food resources. sexually dimorphic manner. Dev Psychobiol 49:335–42. Bales KL, Saltzman W. 2016. Fathering in rodents: Integration of mechanistic laboratory experiments Neurobiological substrates and consequences for offspring. with studies of rodents living under naturalistic con- Horm Behav 77:249–59. ditions will greatly advance our understanding of the Bamshad M, Novak MA, De Vries GJ. 1994. Cohabitation al- sources and consequences of variation in paternal care. ters vasopressin innervation and paternal behavior in prairie voles (Microtus ochrogaster). Physiol Behav 56:751–8. Bester-Meredith JK, Marler CA. 2003. Vasopressin and the transmission of paternal behavior across generations in Acknowledgments mated, cross-fostered Peromyscus mice. Behav Neurosci 117:455–63. We thank Dr. Miyetani Chauke, Dr. Theodore Braun K, Champagne FA. 2014. Paternal influence son off- Garland, Jr., and the many undergraduates and collab- spring development: behavioural and epigenetic pathways. oratorswhocontributedtoourresearch,aswellas J Neuroendocrinol 26:697–706. three anonymous reviewers who provided helpful Bredy TW, Lee AW, Meaney MJ, Brown RE. 2004. Effect of comments on a previous draft of the manuscript. neonatal handling and paternal care on offspring cognitive We also thank Dr. Ben Dantzer and Dr. Dustin development in the monogamous California mouse Rubenstein for inviting WS to participate in the sym- (Peromyscus californicus). Horm Behav 46:30–8. Bridges RS. 2015. Neuroendocrine regulation of maternal be- posium “The Development and Mechanisms havior. Front Neuroendocrinol 36:178–96. Underlying Inter-individual Variation in Pro-social Caldwell HK, Lee H-J, Macbeth AH, Young WS. III. 2008. Behavior” at the 2017 meeting of the Society for Vasopressin: behavioral roles of an “original” neuropeptide. Integrative and Comparative Biology. Prog Neurobiol 84:1–24. 12 W. Saltzman et al.

Chauke M, de Jong TR, Garland T Jr, Saltzman W. 2012. aggression and vasopressin expression in male California Paternal responsiveness is associated with, but not medi- mouse offspring. Horm Behav 50:699–707. ated by reduced neophobia in male California mice Gleason ED, Marler CA. 2013. Non-genomic transmission of (Peromyscus californicus). Physiol Behav 107:65–75. paternal behaviour between fathers and sons in the monog- Clark MM, Galef BG Jr. 1999. A testosterone-mediated trade- amous and biparental California mouse. Proc Biol Sci 280 off between parental and sexual effort in male Mongolian published online, (doi: 10.1098/rspb.2013.0824). gerbils (Meriones unguiculatus). J Comp Psychol 113:388–95. Gregg JK, Wynne-Edwards KE. 2005. Placentophagia in naı¨ve Clark MM, Galef BG Jr. 2001. Age-related changes in paternal adults, new fathers, and new mothers in the biparental responses of gerbils parallel changes in their testosterone dwarf hamster, Phodopus campbelli. Dev Psychobiol concentrations. Dev Psychobiol 39:179–87. 47:179–88. Clark MM, Liu C, Galef BG Jr. 2001. Effects of consanguinity, Gromov VS. 2009. Interactions of partners in family pairs, exposure to pregnant females, and stimulation from young on care of the offspring, and the role of tactile stimulation male gerbils’ responses to pups. Dev Psychobiol 39:257–64. in formation of parental behavior of the Mongolian gerbil Clark MM, Vom Saal FS, Galef BJ Jr. 1992. Intrauterine po- (Meriones unguiculatus) under laboratory conditions. Izv sitions and testosterone levels of adult male gerbils are cor- Akad Nauk Ser Biol 5:569–79. related. Physiol Behav 51:957–60. Gubernick DJ. 1990. A maternal chemosignal maintains pa- Clark MM, Vonk JM, Galef BG. Jr. 1998. Intrauterine posi- ternal behaviour in the biparental California mouse, tion, parenting, and nest-site attachment in male Peromyscus californicus. Anim Behav 39:936–42. Mongolian gerbils. Dev Psychobiol 32:177–81. Gubernick DJ, Alberts JR. 1989. Postpartum maintenance of Cushing BS, Perry A, Musatov S, Ogawa S, Papademetriou E. paternal behaviour in the biparental California mouse, 2008. Estrogen receptors in the medial amygdala inhibit the Peromyscus californicus. Anim Behav 37:656–64. expression of male prosocial behavior. J Neurosci Gubernick DJ, Laskin B. 1994. Mechanisms influencing sib- 28:10399–403. ling care in the monogamous biparental California mouse, De Jong TR, Chauke M, Harris BN, Saltzman W. 2009. From Peromyscus californicus. Anim Behav 48:1235–7. here to paternity: neural correlates of the onset of paternal Harris BN, Perea-Rodriguez JP, Saltzman W. 2011. Acute ef- behavior in California mice (Peromyscus californicus). fects of corticosterone injection on paternal behavior in Horm Behav 56:220–31. California mouse (Peromyscus californicus) fathers. Horm De Jong TR, Harris BN, Perea-Rodriguez JP, Saltzman W. Behav 60:666–75. 2013. Physiological and neuroendocrine responses to Harris BN, de Jong TR, Yang V, Saltzman W. 2013. Chronic chronic variable stress in male California mice variable stress in fathers alters paternal and social behavior (Peromyscus californicus): influence of social environment but not pup development in the biparental California and paternal state. Psychoneuroendocrinology 38:2023–33. mouse (Peromyscus californicus). Horm Behav 64:799–811. De Jong TR, Korosi A, Harris BN, Perea-Rodriguez JP, Hashemian F, Shafigh F, Roohi E. 2016. Regulatory role of Saltzman W. 2012. Individual variation in paternal re- prolactin in paternal behavior in male parents: a narrative sponses of virgin male California mice (Peromyscus califor- review. J Postgrad Med 62:182–7. nicus). Physiol Biochem Zool 85:740–51. Horrell ND, Perea-Rodriguez JP, Harris BN, Saltzman W. De Vries GJ, Miller MA. 1999. Anatomy and function of 2017. Effects of repeated pup exposure on behavioral, neural, extrahypothalamic vasopressin systems in the brain. Prog and adrenocortical responses to pups in male California Brain Res 199:3–20. mice (Peromyscus californicus). Horm Behav 90:56–63. Dulac C, O’Connell LA, Wu Z. 2014. Neural control of ma- Hume JM, Wynne-Edwards KE. 2005. Castration reduces ternal and paternal behaviors. Science 345:765–70. male testosterone, estradiol, and territorial aggression, but Elwood RW. 1977. Changes in the responses of male and not paternal behavior in biparental dwarf hamsters female gerbils (Meriones unguiculatus) toward test pups (Phodopus campbelli). Horm Behav 48:303–10. during the pregnancy of the female. Anim Behav 25:46–51. Jia R, Tai F, An S, Zhang X. 2011. Neonatal paternal dep- Elwood RW. 1980. The development, inhibition and disinhi- rivation or early deprivation reduces adult parental behav- bition of pup-cannibalism in the Mongolian gerbil. Anim ior and central estrogen receptor a expression in Behav 28:1188–94. mandarin voles (Microtus mandarinus). Behav Brain Res Elwood RW, Ostermeyer MC. 1984. Infanticide by male and 224:279–89. female Mongolian gerbils: ontogeny, causation, and func- Jones JS, Wynne-Edwards KE. 2000. Paternal hamsters me- tion. In: Hausfater G, Hrdy SB, editors. Infanticide: compar- chanically assist the delivery, consume amniotic fluid and ative and evolutionary perspectives. NY: Aldine. p. 367–86. placenta, remove fetal membranes, and provide parental Fabre P-H, Hautier L, Dimitrov D, Douzery EJP. 2012. A care during the birth process. Horm Behav 37:116–25. glimpse on the pattern of rodent diversification: a phylo- Kenkel WM, Perkeybile AM, Carter CS. 2017. The neurobi- genetic approach. BMC Evol Biol 12:88. ological causes and effects of alloparenting. Dev Neurobiol Ferre´ P, Fernandez-Teruel A, Escorihuela RM, Driscoll P, 77:214–32. Corda MG, Giorgi O, Tobena~ A. 1995. Behavior of the Kenkel WM, Paredes J, Lewis GF, Yee JR, Pournajafi- Roman/Verh high- and low-avoidance rat lines in anxiety Nazarloo H, Grippo AJ, Porges SW, Carter CS. 2013. tests: relationship with defecation and self-grooming. Autonomic substrates of the response to pups in male prai- Physiol Behav 58:1209–13. rie voles. PLoS One 8:e69965. Frazier CR, Trainor BC, Cravens CJ, Whitney TK, Marler CA. Klampfl SM, Brunton PJ, Bayerl DS, Bosch OJ. 2014. 2006. Paternal behavior influences development of Hypoactivation of CRF receptors, predominantly type 2, Paternal care in biparental rodents 13

in the medial-posterior BNST is vital for adequate maternal Pereira AS, Giusti-Paiva A, Vilela FC. 2015. Central cortico- behavior in lactating rats. J Neurosci 34:9665–76. sterone disrupts behavioral and neuroendocrine responses Kleiman DG, Malcolm JR. 1981. The evolution of male pa- during lactation. Neurosci Lett 606:88–93. rental investment in mammals. In: Gubernick DJ, Klopfer Perkeybile AM, Bales KL. 2017. Intergenerational transmis- PH, editors. Parental care in mammals. New York (NY): sion of sociality: the role of parents in shaping social be- Plenum. p. 347–87. havior in monogamous and non-monogamous species. Lambert KG, Franssen CL, Bardi M, Hampton JE, Hainley L, J Exp Biol 220:114–23. Karsner S, Tu EB, Hyer MM, Crockett A, Baranova A, et al. Perkeybile AM, Griffin LL, Bales KL. 2013. Natural variation 2011. Characteristic neurobiological patterns differentiate in early parental care correlates with social behaviors in paternal responsiveness in two Peromyscus species. Brain adolescent prairie voles (Microtus ochrogaster). Front Behav Evol 77:159–75. Behav Neurosci 7 published online (doi: 10.3389/ Lee AW, Brown RE. 2002. The presence of the male facilitates fnbeh.2013.00021). parturition in California mice (Peromyscus californicus). Perkeybile AM, Delaney-Busch N, Hartman S, Grimm KJ, Can J Zool 80:926–33. Bales KL. 2015. Intergenerational transmission of allopar- Lei K, Cushing BS, Musatov S, Ogawa S, Kramer KM. 2010. ental behavior and oxytocin and vasoapressin receptor dis- Estrogen receptor-a in the bed nucleus of the stria termi- tribution in the praiarie vole. Front Behav Neurosci 9 nalis regulates social affiliation in male prairie voles published online, (doi: 10.3389/fnbeh.2015.00191). (Microtus ochrogaster). PLos One 5:e8931. Probst B. 1987. Developmental changes in the pituitary- Li Y, Lian Z, Wang B, Tai F, Wu R, Hao P, Qiao X. 2015. gonadal axis in male Mongolian gerbils from birth to Natural variation in paternal behavior is associated with adulthood. Exp Clin Endocrinol 90:157–66. central estrogen receptor alpha and oxytocin levels. Roberts RL, Miller AK, Taymans SE, Carater CS. 1999. Role J Comp Physiol A 201:285–93. of social and endocrine factors in alloparental behavior of Lim MM, Young LJ. 2006. Neuropeptidergic regulation of prairie voles (Microtus ochrogaster). Can J Zool 76:1862–8. affiliative behavior and social bonding in animals. Horm Ryan BC, Vandenbergh JG. 2002. Intrauterine position effects. Behav 50:506–17. Neurosci Biobehav Rev 26:665. Lonstein JS, De Vries GJ. 1999. Sex differences in the parental Saltzman W, Ziegler TE. 2014. Functional significance of hor- behaviour of adult virgin prairie voles: independence from go- monal changes in mammalian fathers. J Neuroendocrinol nadal hormones and vasopressin. J Neuroendocrinol 11:441–9. 26:685–96. [PMC][10.1111/jne.12176] [25039657] Martınez A, Ramos G, Martınez-Torres M, Nicolas L, Saltzman W, Thinda S, Higgins AL, Matsumoto WR, Ahmed S, Carmona A, Cardenas M, Luis J. 2015. Paternal behavior McGeehanL,KolbEM.2009.Effectsofsiblingsonrepro- in the Mongolian gerbil (Meriones unguiculatus): estrogenic ductive maturation and infanticidal behavior in cooperatively and androgenic regulation. Horm Behav 71:91–5. breeding Mongolian gerbils. Dev Psychobiol 51:60–72. McGuire B. 1988. Effects of cross-fostering on parental be- Schradin C, Anzenberger G. 1999. Prolactin, the hormone of havior of meadow voles (Microtus pennsylvanicus). paternity. New Physiol Sci 14:223–31. J 69:332–41. Sestakova Puzserova A, Kluknavsky N, Bernatove I. 2013. Nephew BC, Bridges RS. 2011. Effects of chronic social stress Determination of motor activity and anxiety-related behav- during lactation on maternal behavior and growth in rats. iour in rodents: methodological aspects and role of nitric Stress 14:677–84. oxide. Interdiscip Toxicol 6:126–35. Nonneman DJ, Ganjam VK, Welshons WV, Vom Saal FS. Song Z, Tai F, Yu C, Wu R, Zhang X, Broders H, He F, Guo 1992. Intrauterine position effects on steroid metabolism R. 2010. Sexual or paternal experiences alter alloparental and steroid receptors of reproductive organs in male behavior and the central expression of ERa and OT in male mice. Biol Reprod 47:723–9. mandarin voles (Microtus mandarinus). Behav Brain Res Parker KJ, Lee TM. 2001. Central vasopressin administration reg- 214:290–300. ulates the onset of facultative paternal behavior in Microtus Stolzenberg DS, Champagne FA. 2016. Hormonal and non- pennsylvanicus (meadow voles). Horm Behav 39:285–94. hormonalbasesofmaternalbehavior:theroleofexpe- Parker KJ, Kinney LF, Phillips KM, Lee TM. 2001. Paternal rience and epigenetic mechanisms. Horm Behav behavior is associated with central neurohormone receptor 77:204–10. binding patters in meadow voles (Microtus pennsylvanicus). Trainor BC, Marler CA. 2001. Testosterone, paternal behav- Behav Neurosci 115:1341–8. ior, and aggression in the monogamous California mouse Pei M, Matsuda K-I, Sakamoto H, Kawata M. 2006. (Peromyscus californicus). Horm Behav 40:32–42. Intrauterine proximity to male fetuses affects the morphol- Trainor BC, Marler CA. 2002. Testosterone promotes paternal ogy of the sexually dimorphic nucleus of the preoptic area behaviour in a monogamous mammal via conversion to in the adult rat brain. Eur J Neurosci 23:1234–40. oestrogen. Proc R Soc Lond B 269:823–9. Perea-Rodriguez JP, Saltzman W. 2014. Differences in placen- Vella ET, Evans CCD, Ng MWS, Wynne-Edwards KE. 2005. tophagia in relation to male and female reproductive status Ontogeny of the transition from killer to caregiver imn in in the California mouse (Peromyscus californicus). Dev dwarf hamsters (Phodopus campbelli) with biparental care. Psychobiol 56:812–20. Dev Psychobiol 46:75–85. Perea-Rodriguez JP. 2016. Placentophagia in the California Vom Saal FS, Grant WM, McMullen CW, Laves KS. 1983. mouse (Peromyscus californicus)anditspotentialrolein High fetal estrogen concentrations: correlation with in- the onset of male parental care [dissertation]. [Riverside creased adult sexual activity and decreased aggression in (CA)]: University of California, Riverside. male mice. Science 220:1306–9. 14 W. Saltzman et al.

Wang Z, De Vries GJ. 1993. Testosterone effects on paternal vasopressin, oxytocin and tyrosine hydroxylase in mandarin behavior and vasopressin immunoreactive projections in voles. Neuropeptides 48:29–35. prairie voles (Microtus ochrogaster). Brain Res 631:156–60. Yu P, Zhang H, Li X, He F, Tai F. 2015. Early bi-parental Wang Z, Ferris CF, De Vries GJ. 1994. Role of septal vaso- separation or neonatal paternal deprivation in mandarin pressin innervation in paternal behavior in prairie voles voles reduces adult offspring paternal behavior and alters (Microtus ochrogaster). Proc Nat Acad Sci U S A 91:400–4. serum corticosterone levels and neurochemistry. Horm Wang J, Taia F, Lai X. 2014. Cocaine withdrawal influences Behav 73:8–14. paternal behavior and associated central expression of