|Home | About | Journals | Submit | Contact Us | Français|
Behavioral and endocrine suppression of reproduction in subordinate females produces the high reproductive skew that characterizes callitrichid primate mating systems. Snowdon et al. reported that the eldest daughters in tamarin families exhibit further endocrinological suppression immediately following the birth of siblings, and suggested that dominant females exert greater control over subordinate endocrinology during this energetically challenging phase of reproduction. We monitored the endocrine status of five Wied’s black tufted-ear marmoset daughters before and after their mother delivered infants by measuring concentrations of urinary estradiol (E2), pregnanediol glucuronide (PdG), testosterone (T), and cortisol (CORT). Samples were collected from marmoset daughters 4 weeks prior to and 9 weeks following three consecutive sibling-litter births when the daughters were prepubertal (M=6.1 months of age), peripubertal (M=11.9 months), and postpubertal (M=17.6 months). The birth of infants was associated with reduced ovarian steroid excretion only in the prepubertal daughters. In contrast, ovarian steroid levels tended to increase in the postpubertal daughters. Urinary E2 and T levels in the postpubertal daughters were 73.8% and 37.6% higher, respectively, in the 3 weeks following the birth of infants, relative to prepartum levels. In addition, peak urinary PdG concentrations in peri- and postpubertal daughters were equivalent to luteal phase concentrations in nonpregnant, breeding adult females, and all of the peri- and postpubertal daughters showed clear ovulatory cycles. Cortisol excretion did not change in response to the reproductive status of the mother, nor did the concentrations change across age. Our data suggest that marmoset daughters of potential breeding age are not hormonally suppressed during the mother’s peripartum period or her return to fertility. These findings provide an additional example of species diversity in the social regulation of reproduction in callitrichid primates.
In callitrichid social groups, breeding activity is typically limited to a single adult male and female [French, 1997]. The mechanism underlying this high reproductive skew in males appears to be primarily behavioral in nature, since testosterone levels do not differ dramatically between breeders and nonbreeders, and sons removed from natal family groups readily mate with unrelated female partners [Baker et al., 1999; Ginther et al., 2001]. However, reproductive suppression in subordinate females is mediated by a complex combination of behavioral and endocrine mechanisms that differ by species and social context. Although captive golden lion tamarin (Leontopithecus rosalia) daughters show normal ovarian cycles, high levels of aggression from breeding females can prevent reproduction [French et al., 2002]. In contrast, daughters in captive cotton-top tamarin (Saguinus oedipus) families do not exhibit ovarian cyclicity while they are in their natal family group [French et al., 1984; Ziegler et al., 1987b]. In marmosets, there is clear evidence of physiological suppression of subordinate females in groups comprised of unrelated males and females [e.g., Abbott, 1993]; however, in family groups, one or more daughters may escape suppression and commence ovulatory cycles [Carlson et al., 1997; Saltzman et al., 1997a, b; Smith et al., 1997; Ziegler & Sousa, 2002]. Regardless of its expression, a prominent functional interpretation of reproductive suppression in family-living callitrichids involves the minimization of intrasexual reproductive competition within the group [Abbott, 1993; French, 1997].
In marmosets and tamarins, breeding females return to fertility and conceive within 2–3 weeks after parturition [French et al., 1996; Ziegler et al., 1987a]. The early return to fertility places the breeding female in direct competition with other potential breeding females, including daughters, for the resources necessary to sustain pregnancy, and for the postpartum resources necessary for care of the subsequent litter, including the alloparental care provided by older offspring [Digby, 1995]. To the extent that breeding females can regulate reproductive function in daughters via olfactory and behavioral cues [French, 1997], mothers might be expected to exert greater control over reproduction in daughters during the postpartum period. The eldest daughters in cotton-top tamarin families do, in fact, show further reproductive suppression following the birth of siblings, with decreased levels of urinary estrone glucuronide (E1C) and luteinizing hormone (LH) during the first 3 weeks postpartum [Snowdon et al., 1993]. Because daughters in this species are already anovulatory, however, they pose little threat to the reproductive hegemony of the breeding female. For callitrichid species in which daughters can and do exhibit ovulatory cycles in their natal groups (e.g., marmosets), suppression of nonbreeder reproduction during the postpartum period should be particularly pronounced if the phenomenon reflects avoidance of reproductive competition.
In this study, we investigated patterns of urinary hormone excretion in pre-, peri-, and postpubertal female marmosets (Callithrix kuhlii) 4 weeks prior to and 9 weeks following the birth of three successive sibling-litters. We measured ovarian steroid activity using patterns of urinary estradiol (E2) and pregnanediol glucuronide (PdG). Urinary cortisol (CORT) excretion was measured to monitor hypothalamic-pituitary-adrenal axis (HPA) activity. Finally, we measured urinary testosterone (T) because there is accumulating evidence that androgens may be important modulators of sexual and aggressive behavior in females [Christiansen, 2001], and evidence that excreted androgens in female marmosets reflect variations in circulating T [Armstrong et al., 2003]. E2, PdG, and T concentrations were expected to increase significantly as the daughters developed through the pubertal stages. Concentrations of cortisol were also expected to increase in correspondence with increases in ovarian activity [Saltzman et al., 1998]. To the extent that breeding females exert control over reproductive potential in their daughters, we also expected to see significant decreases in sex steroid levels in the weeks immediately following the birth of infants, particularly in postpubertal daughters. In contrast, significant increases in cortisol excretion were expected after the birth of a sibling, to the degree that the arrival of new infants and return to fertility in the mother would be potent social stressors for the daughters [Smith&French, 1997b; Ziegler et al., 1995].
The subjects of this study were five Wied’s black tufted-ear marmoset (C. kuhlii) daughters that were housed with their natal family groups at the University of Nebraska at Omaha’s Callitrichid Research Center (Table I). (For a description of the housing and husbandry practices used, see French et al. .) The females were sampled for three successive sibling-litters: one in which they were prepubertal (M=6.1 months old), one in which they were peripubertal (M=11.9 months old), and one in which they were postpubertal (M=17.6 months old; categorization based on Smith et al. ).
Urine samples were collected between 0600 and 0800 hr from all subjects one to five times per week, with an average (±SD) of 1.86 (±1.18) samples each week, by means of a previously described noninvasive technique [French et al., 1996]. The samples were stored at −20°C until they were assayed. Enzyme immunoassays were utilized to monitor hormone concentrations after enzyme hydrolysis and ether extraction (E2 [Fite&French, 2000] and T [Nunes et al., 2000]) or via direct assay (PdG [French et al., 1996] and CORT [Smith&French, 1997a]). Recovery of 3H-labeled steroid after extraction was 57.7% for E2, and 63.1% for T. The intra- and interassay coefficients of variation for all steroid assays were <7.6% and 15.7%, respectively. Creatinine was measured by a modified Jaffé end-point assay [French et al., 1996], and all hormone concentrations were divided by the creatinine concentrations to control for variable fluid intake and output.
Hormone values for each daughter were averaged for the 4 weeks prior to the birth of siblings to establish a prepartum baseline, and in three 3-week blocks following the birth of siblings, encompassing a majority of the maternal lactation period in Callithrix [Missler et al., 1992; Tardif et al., 2001]) and maternal infant-care effort [Nunes et al., 2001]. To assess maturational effects across the three sibling-litter births, we conducted a completely within-subjects, two-way analysis of variance (ANOVA; litter (n=3) × weeks (n=4)) on the urinary excretion profiles for E2, T, PdG, and CORT. Planned within-subjects, one-way ANOVAs were used to assess changes in steroid excretion across the pre- and postpartum phases for each developmental stage. To assess the ovulatory capacity of daughters, we compared the peak PdG concentrations in daughters at each age with luteal-phase peak PdG concentrations in breeding adult females (n=6 females) known to be cycling but not pregnant. We conducted post-hoc analyses using the least-significant difference (LSD) test.
Although the sample size was limited, we noted significant increases in levels of excreted sex steroids across consecutive sibling-litter births (Fig. 1) that coincided with the developmental stages of the daughters (E2: F(2,8)=15.74, P=0.002; T: F(2,8)=10.01, P=0.007; PdG: F(2,8)=16.14, P=0.002). All three profiles were characterized by extremely low mean levels in prepubertal daughters, and a slight rise in hormone levels in peripubertal daughters (LSD, E2: NS; T: P=0.017; PdG: P=0.021). The mean levels continued to rise as the daughters became postpubertal (LSD, E2: P=0.014; T: NS; PdG: P=0.017). All increases in sex steroid concentrations between pre- and postpubertal stage daughters were significant (LSD, E2: P=0.010; T: P=0.024; PdG: P=0.015). There was also clear evidence that peri- and postpubertal daughters displayed one or more ovulatory cycles during the peripartum period of their mother, marked by luteal phase PdG concentrations of ≥10 μg/mg Cr (see Fig. 2). ANOVA and subsequent post-hoc tests revealed no significant differences in peak luteal PdG concentrations among peripubertal daughters, postpubertal daughters, and cycling adult females (13.4 ± 2.0 μg/mg Cr, 14.1 ±1.6 μg/mg Cr, and 12.1 ± 1.1 μg/mg Cr, respectively). However, prepubertal daughters exhibited significantly lower maximum PdG concentrations compared to all other females (2.1 ± 0.7 μg/mg Cr; F(3,17)=15.5, P<0.001). In contrast to sex steroids, concentrations of excreted cortisol did not significantly change across the developmental stages of the daughters (F(2,8)=2.973, NS).
The birth of siblings had little influence on hormone excretion across the developmental stages of the daughters. Prepubertal daughters exhibited a nonsignificant pre- to postpartum decrease in E2, T, and CORT levels across all three 3-week blocks (F(3,12)<1.20; NS). However, prepubertal daughters exhibited significant changes in pre- to postpartum levels of PdG (F(3,12)=4.37, P=0.027). A decrease from prepartum levels of PdG was evident for all three postpartum blocks, although only the 55.8% decrease for weeks 1–3 was significant (LSD, P<0.05). Peripubertal daughters did not exhibit significant changes in PdG, E2, and CORT levels (E2: F(3,12)=1.19, (NS); PdG: F(3,12)=2.65, (NS)). In contrast, peripubertal daughters showed significant changes from pre- to postpartum concentrations of urinary T (F(3,12)=7.10, P=0.005). T levels decreased nonsignificantly from prepartum values during weeks 1–3 following a birth (LSD, P>0.05), but significantly increased 50.3% over weeks 4–6 and 45.3% over weeks 7–9 (LSD, P<0.05). Postpubertal daughters exhibited increases of 73.8% E2, 37.6% T, 5.3% PdG, and 48.3% CORT from prepartum levels during weeks 1–3. However, a statistical analysis revealed that these changes, and those of the subsequent weeks, were not significant (F(3,12)<1.71, NS).
Because reproductive suppression in family-living callitrichids involves the minimization of intrasexual reproductive competition within the group, it has been suggested that the breeding female may exert control over her daughters’ reproduction during the time of postpartum ovulation, when the breeding position is highly vulnerable [Lazaro-Perea et al., 2000]. Endocrine profiles obtained in a previous study of subordinate tamarins [Snowdon et al., 1993] were consistent with this prediction, but our results from marmosets were not. For the marmoset daughters in our study, there were no systematic reductions in ovarian hormone concentrations across the immediate postpartum period in the mother. As predicted, maturational effects on ovarian activity were observed in marmoset daughters as they passed through the developmental stages of puberty [Smith et al., 1997]; however, we failed to detect the expected corresponding increase in hypothalamic-pituitary-adrenal axis activity [Saltzman et al., 1998].
The only finding consistent with the predictions of mother–daughter reproductive competition was the reduction in excreted PdG in the first 3 weeks postpartum in prepubertal daughters. However, the concentrations of PdG are exceedingly low in prepubertal daughters, and even maximum PdG concentrations in these daughters (2.1 ± 0.7 μg/mg Cr) are characteristic of anovulatory adult females [Smith et al., 1997]. It is likely, therefore, that the significant reduction in PdG exhibited by prepubertal daughters does not reflect a functional change in reproductive potential. In contrast to our predictions, we identified significant increases in T in daughters exposed to a second birth of siblings. Although this study did not address the origin of elevated excreted T, previous work on a variety of female mammals (e.g., dogs, baboons, and human females) reveals that ovarian (especially luteal) tissue is a prominent source of androgen [e.g., Castracane et al., 1998]. The functional significance of elevated T may be that it facilitates aggressive and sexual behavior in developing females. Other steroid levels tended to increase (73.8% in E2, 37.6% in T, and 5.3% in PdG) in postpubertal daughters after the birth of siblings; however, these changes in steroid excretion were not significant. Interestingly, both peri- and postpubertal daughters exhibited peak PdG concentrations during the peripartum period of their mother that were similar to those of nonpregnant, breeding adult females, providing quantitative evidence that the mother’s reproductive state has little impact on reproductive function in her older daughters.
Although there were numerous similarities in reproductive parameters between marmosets and tamarins, including suppression of daughters, twin litter births, and a fertile postpartum ovulation, this study also identified important interspecific differences in the timing and intensity of suppression following the birth of siblings. Reproduction may be more costly for tamarins than for marmosets, since tamarins have a longer periods of gestation [French&Fite, 1999] and infant dependence [Snowdon, 1996] than marmosets. Further, observations of free-ranging populations suggest that marmosets have greater mean group sizes than tamarins [Mittermeier et al., 1988], potentially providing breeding female marmosets with more alloparental assistance than may be available to tamarins. These differences may place a greater premium on dominant control of subordinate reproduction in tamarins, and may account for the divergent endocrine responses to maternal reproductive state in daughters. Moreover, this divergence in daughters’ endocrine responses may help explain the more frequent observations of plural breeding females in marmosets than in tamarins [French, 1997]. However, further research is necessary to determine the ultimate causes and underlying mechanisms of the observed variation in the timing and intensity of suppression in callitrichid daughters.
There are two possible explanations for the absence of further suppression and a trend toward increased steroid activity following the birth of siblings in marmoset daughters. First, under conditions of abundant resources, the suppression of daughters may be relaxed. Because marmosets are able to give birth to twin litters approximately every 6 months, family groups with sexually mature daughters (12–18 mo) may be saturated with alloparental help. Therefore, a lack of further hormonal suppression at this time may facilitate voluntary emigration of daughters from the natal group, or increase the inclusive fitness of the breeder by allowing a daughter to produce offspring. Second, daughters may increase their resistance to suppression following the birth of infants in preparation for emigration or to take over the breeding position in the natal group. In wild populations of common marmosets, infant birth and conception appear to be socially transitional times, which are marked by the emigration of subordinates and frequent loss of breeders [Lazaro-Perea et al., 2001]. Among captive marmosets, some mature daughters escape from suppression while in their natal groups [Smith et al., 1997; Saltzman et al., 1997a], but most fail to reproduce [Saltzman et al., 1997a]. However, changes in the social environment rapidly induce ovarian activity and sexual behaviors in daughters [Saltzman et al., 1997b; Ziegler&Sousa, 2002]. This swift adaptation to alterations in the daughter’s social environment could be extremely beneficial during times of forced or voluntary emigration, or the replacement of resident breeders. Thus, the absence of further suppression and a trend toward increasing steroid levels immediately following the birth of siblings in postpubertal daughters may reflect the priming that allows for this rapid response to social change and successful transition into a breeding role.
Our data present evidence that postpubertal marmoset daughters are not more hormonally suppressed following the birth of siblings, and exhibit a trend of increased steroid production, allowing for greater adaptability to the social change that often follows parturition and the mother’s return to fertility. These data provide an example of species diversity in the social regulation of reproduction in callitrichid primates, and add support to the growing consensus that marmoset mothers have little suppressive influence on reproductive function in their daughters [Saltzman et al., 1997a; Ziegler&Sousa, 2002].
We thank animal-care technicians MaLinda Henry, Denise Hightower, Heather Jensen, and Danny Revers, and veterinarian Tom Curro for their dedication to the care of the callitrichid colony at UNOmaha. We also thank the numerous people who assisted with urine collection over the years encompassing this study. We especially thank Toni Ziegler for providing useful comments on an earlier version of this paper. The research presented was approved by the UNMC/UNO Institutional Animal Care and Use Committee (95-103-07).
Contract grant sponsor: NSF; Contract grant number: IBN 00-91030; Contract grant sponsor: NIH; Contract grant number: HD 42882.