Urinary prolactin is correlated with mothering and allo-mothering in squirrel monkeys

Physiology & Behavior 84 (2005) 295 – 301

Urinary prolactin is correlated with mothering and allo-mothering in squirrel monkeys Joseph Soltisa, Frederick H. Wegnerb, John D. Newmana,* a

Laboratory of Comparative Ethology, National Institute of Child Health and Human Development, National Institutes of Health, Department of Health and Human Services P.O. Box 529, Poolesville, MD 20837, United States b The Wisconsin National Primate Research Center, University of Wisconsin, Madison, WI 53715, United States Received 25 June 2004; received in revised form 2 November 2004; accepted 15 December 2004

Abstract The hormone prolactin is implicated in infant care-giving by parents and allo-parents in a variety of species. Adult female squirrel monkeys (Saimiri spp.) engage in allo-mothering behavior, which includes carrying and nursing infants, but communal care of offspring has not been investigated from an endocrine standpoint in this taxon. We attempted to fill this gap by examining prolactin levels in squirrel monkeys (Saimiri sciureus) as a function of parental responsiveness. Subjects were housed at the National Institutes of Health Animal Center and assays were performed at the Wisconsin National Primate Research Center. To test for the presence of prolactin in squirrel monkey, saliva, blood and saliva were simultaneously collected from anesthetized subjects during routine health examinations. Prolactin was detectable in serum but not in saliva samples. In the core investigation, behavioral data were collected by focal animal sampling on three 1male multi-female groups, and individually identified urine was collected non-invasively from foil containers underneath group cages on a daily basis throughout the behavioral study. Changes in urinary prolactin over time reflected changes in the reproductive state of a female who was pregnant, gave birth and lactated during the study. Mean urinary prolactin levels in non-lactating females and a male housed with infants in one group were higher than in adults from 2 groups without infants. In the group with infants, mean urinary prolactin levels in adults increased with the amount of infant contact and care-giving. The squirrel monkey may represent a new primate model for investigating the endocrinology of infant care-giving. Published by Elsevier Inc. Keywords: Communal breeding; Allomothering; Parenting; Salivary prolactin; Primates

1. Introduction The actions of the pituitary hormone prolactin are strongly implicated in the expression of maternal behavior in mammals [9,23]. For example, prolactin injections into the hypothalamus of ovariectomized, prolactin-depleted female rats primed with pregnancy hormones stimulated maternal behavior toward foster pups [3]. Prolactin has also been implicated in the care-giving behavior of non-mothers in both birds and mammals [34,35,38]. Among primates, prolactin is associated with increased paternal responsiveness in new and expectant human fathers [28], and with * Corresponding author. Tel.: +1 301 496 0835; fax: +1 301 496 0066. E-mail address: [email protected] (J.D. Newman). 0031-9384/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.physbeh.2004.12.006

allo-parenting in two species of New World monkey, the common marmoset (e.g. [5,21]) and the cotton-top tamarin (e.g. [37]). The squirrel monkey (Saimiri spp.) is another primate taxon in which extensive allo-parenting occurs, but in this taxon only females are involved (captivity: [11,24,33]; freeranging: [1,4]; wild: [18]). Squirrel monkeys live in mixedsex social groups, and in most species females are philopatric while males disperse at sexual maturity [2,4,17]. Female squirrel monkeys are known to form affiliative relationships, engage in collective defense, form coalitions, and preferentially exchange vocalizations with one another [1,4,12,32,26,27]. As noted above, these affiliative relationships among females include communal care of offspring. Adult males do not participate in infant

296

J. Soltis et al. / Physiology & Behavior 84 (2005) 295–301

care, although subadult males have been observed occasionally to interact affilliatively with infants, including carrying or protecting them [1,11]. Allo-mothering by female squirrel monkeys consists of active care-giving, including protecting, retrieving, carrying and nursing infants, and passive forms of affiliation, such as maintaining proximity and huddling with infants. Allo-mothering in squirrel monkeys has not been investigated from a physiological standpoint. We explored the feasibility of using the squirrel monkey as an additional non-human primate model for examining the neuro-endocrinology of infant care-giving. In this report, we validate prolactin assays for squirrel monkey serum and urine, and examine whether prolactin excreted in the urine is associated with exposure to infants and allo-mothering.

2. Methods 2.1. Blood and saliva collection and assays In an initial experiment prior to the core study in observation rooms described below, we investigated the possibility of assaying prolactin in saliva by simultaneously collecting saliva and blood from subjects who were anesthetized during a routine health examination. Blood was collected from 26 adult female squirrel monkeys (14 Gothic Arch, Saimiri sciureus, and 12 Roman Arch, S. boliviensis, [10,15], and saliva was collected from 16 of the 26 females (8 Gothic and 8 Roman Arch). Blood (2 cc) was drawn with a 27-gauge needle on a 3-ml syringe by femoral venipuncture and kept in a vacutainer on ice. Blood was centrifuged at 3500 rpm for 20 min at 4 8C, and the supernatant was recovered and stored at 75 8C until the time of the assay. Saliva (200–1000 ul) was collected with disposable pipettes, transferred into plastic 1.5-ml tubes and placed in ice. Saliva samples were centrifuged at 3500 rpm for 1 s (flash centrifuge) at 4 8C, and stored at 75 8C until the time of the assay. Serum and saliva immunoradiometric assays (IRMA) for prolactin were performed at the Assay Services Unit of the Wisconsin National Primate Research Center, with a human prolactin (hPRL) IRMA kit (Medicorp, Montreal, Quebec, CA) which uses mouse antibody in a solid phase two-site immunoradiometric method. The prolactin reacts simultaneously with solid phase hPRL antibody and radiolabelled (I125) hPRL antibody. Accuracy was determined by measuring a pooled serum sample and, in the same assay, adding that same amount of sample to doses of a second standard curve. Expected values are the known standard curve values plus the obtained value of the serum sample when measured alone, and the observed values were those obtained from the second standard curve with the added serum sample. Accuracy was 89.87% (F1.53 S.E.M.). The minimum detectable level was 1 ng/mL serum, and a single assay was performed with an intra-assay coefficient of variation of

2.7%. No prolactin was detectable in any squirrel monkey saliva sample. 2.2. Study subjects and housing Core study subjects were 14 adult, 1 juvenile and 2 infant Gothic Arch (see [15]) squirrel monkeys (S. sciureus; [10]) residing in 3 social groups. Each group consisted of 1 adult male and several adult females, but differed with respect to the presence of juveniles and infants. Group A contained no offspring, Group B contained a single juvenile but no infants, and Group C contained 2 infants (Table 1). In the group with infants, 1 infant aged 6–12 months across the study, and 1 infant was born and aged 0–2 months during the study. All groups were housed at the National Institutes of Health Animal Center in Poolesville, Maryland, in 2–3 attached 1.50.70.8 m (H, L, W) cages. Groups A and C were housed in the same indoor room within visual contact, while group B was housed in a separate room with a nonfocal Gothic group with no infants. Animals were fed daily with monkey chow and supplemental nuts and/or fruit on a constant schedule. Water was available at all times. A 12-h light/dark cycle (lights on 700–1900 h) was maintained throughout the study. 2.3. Urine collection and assays Urine samples were recovered from sheets of ReynoldsR aluminum foil placed in waste pans at the bottom of subject cages. Urine was collected from 3 groups of squirrel monkeys (Table 1) during daily focal observations described below (always between 1000 and 1400 h). Urine collections were conducted throughout the behavioral study. When an identified animal urinated and there was no cross-contamination from the urine or feces of another animal, the tray was pulled from the cage floor and the urine was collected with a disposable plastic pipette and transferred to a 1.5-ml tube and placed in ice. Foil was turned over or replaced after each sample was collected. Animals quickly habituated to the urine collection procedure so that ongoing behavior (e.g. relaxed huddling) was not normally disrupted during urine collection. At the end of each daily collection, tubes were centrifuged at 3500 rpm for 5 min at 4 8C. Centrifuged urine was transferred into a new tube, and one drop of Table 1 Group compositions a

Adult males Adult females Juveniles (1–3 years) Infants (b1 years) a

A

B

C

1 4 0 0

1 3 1 0

1 4 0 2b

Adults ranged in age from 8 to 22 years. One infant aged 6–12 months across the study, and one was born and aged 2 months during the study. b

J. Soltis et al. / Physiology & Behavior 84 (2005) 295–301

glycerol solution (0.52 mol/L) was added per 0.75 ml of urine to prevent prolactin breakdown [14]. Tubes were rotated by hand several times to facilitate admixture and stored at 75 8C until the time of the assay. Prolactin assays from squirrel monkey urine required 3.5 ml of urine, but squirrel monkeys urinate in much smaller amounts (generally b1 ml) and they do not produce large early morning voids. Therefore, multiple samples collected across different consecutive observation days were pooled into a combined sample (of z3.5 ml) from which a single prolactin measure could be obtained. As a result, urinary prolactin measures presented here represent integrated measures taken over a period of several days to several weeks. Subjects provided 3–15 (median=6) such combined samples. The male in Group B rarely urinated during observations and was excluded from the analysis. Urinary creatinine enzyme immunoassays (EIA) and prolactin radioimmunoassays (RIA) were performed at the Assay Services Unit at the Wisconsin National Primate Research Center. Creatinine concentration was measured for each combined urine sample as described previously [36]. Mean intra-assay coefficient of variation was 1.81%, and the inter-assay coefficient of variation was 13.42%. Urine samples were then concentrated. For each assay tube, 1500 AL of urine was concentrated to approximately 50 AL using CentriconR YM-10 centrifugal filtering device from Millipore. After concentration, 250 AL assay buffer was added to the 50 AL concentrate, and centrifuged to transfer the concentrated urine to the bottom of the assay tube. The prolactin RIAs were performed with a human prolactin (hPRL) RIA kit (UCB-Bioproducts, distributed by Accurate Chemical and Scientific Corp., Westbury, NY), using anti-hPRL-I125 as the tracer [37]. The prolactin assay was validated for squirrel monkey urine by determining parallelism and accuracy. Serially diluted squirrel monkey urine (N=9) paralleled the standard curve (N=6) with no difference in slope (t=0.265, df=41, PN0.05). Accuracy, as described above for serum assays, was 102.45%F3.43 (S.E.M.). The minimum detectable level was 0.025 ng/mL urine. Mean intra-assay coefficient of variation was 13.74%, and the inter-assay coefficient of variation was 19.43%. Creatinine concentration (mg/mL) was divided into prolactin levels to control for dilution variability [36]. 2.4. Behavioral observations Behavioral observations conducted by J.S. consisted of 10-min focal follows on randomly ordered adults within a group, always between 1000 and 1400 h. Social and sexual behaviors with all group members were recorded. Table 2 reports those social behaviors used in the analyses reported here. Animals were identified by uniquely colored beads around the neck and by distinguishing physical characteristics. Groups A, B and C were observed for 2O, 2 and 6 months, respectively, and every adult in each group was

297

Table 2 Behavioral measures Term

Definition a

Huddle Carry Nurse Approachb Leave In proximitya

Animal Animal Animal Animal Animal Animal

sits in physical contact with another transports another on its back breast-feeds another moves within 10 cm of another moves further than 10 cm from another within 10 cm of another

a Huddling and 10 cm proximity are robust measures of affiliation in squirrel monkeys [26]. Incidental proximity or touching due only to the tail was not scored as proximity or huddling, respectively. b Approaches that did not result in at least 3 s of proximity were not scored in order to avoid scoring proximity when an animal walked by another. bApproachesQ due only to the tail were not scored as an approach.

observed for 410, 310 and 880 min, respectively. Observations on Groups A and B were not as extensive because they remained demographically stable, whereas a female in group C gave birth during the study. 2.5. Statistical analyses and data transformation Group means were compared using Mann–Whitney U tests, and relationships between continuous variables were examined with Spearman’s correlations (SPSS ver. 11.0) with two-tailed alpha set at 0.05. Prolactin values were log10-tranformed when graphically presenting values for the reproductive female because she showed extremely high prolactin levels.

3. Results 3.1. Prolactin in squirrel monkey serum, saliva and urine Prolactin was detectable in squirrel monkey serum and urine, but not in saliva. All serum samples from the 26 females collected during the health examination had detectable amounts of prolactin (Roman range: 5.49– 102.54; Gothic range: 12.01–141.30 ng/mL). On the other hand, detectable levels of prolactin were not found in any of the 16 saliva samples, even though serum taken simultaneously from the same individuals all showed detectable levels of prolactin. Finally, nearly all urine samples yielded detectable levels of prolactin (range: 0–288.21 ng/mg creatinine/mL). Levels of prolactin were undetectable in 9% (8/89) of the urine samples. 3.2. Urinary prolactin and infant contact and care-giving In the group containing infants, the mean urinary prolactin level across non-lactating adult members was significantly higher than the mean across adults in the two groups lacking infants (Mann–Whitney U test, Z= 2.256, N 1=3 adult means, N 2=8 adult means, P=0.024; Fig. 1). This difference is stronger when the data for the two

J. Soltis et al. / Physiology & Behavior 84 (2005) 295–301

.7

3.3. Urinary prolactin in a pregnant female

.6

Fig. 3 shows urinary prolactin levels across the study for the female who gave birth. Prolactin levels were low during early pregnancy when interactions with her 6month-old infant were rare. Prolactin rose substantially during late pregnancy and returned to intermediate and decreasing levels during lactation, a pattern consistent with marmosets who nurse singletons [16] and humans [19,20].

.5 .4 .3 .2 .1 5

3

A

B

3

C

Fig. 1. Adult urinary prolactin (PRL) levels in three groups of squirrel monkeys. Urine was collected throughout the study. The grand mean (FS.E.M.) of individual adult means is shown for each group. Number of adults (N) in each group is indicated. Adults in groups without infants (A,B) show lower mean levels of urinary PRL compared to non-lactating adults in the group with infants (C). Note that groups A and C were housed in the same room but in different group cages, and that group B was housed in another room containing no infants. In group C, data for the mother (who gave birth during the study) and a lactating allo-mother were excluded. Mean PRL with data from the mother and lactating allo-mother included is 22.28 (ng/mg creatinine/mL).

lactating females, including the mother, are included (Z= 2.789, N 1 =5 adult means, N 2 =8 adult means, P=0.005). Fig. 2a shows urinary prolactin levels for the five adult members of the group with infants (group C). Fig. 2b shows the amount and type of affiliative contact with infants by adult members of group C. Both infants in the group were born to the same female, so each infant had the same three female kin in the group, the mother, grandmother and aunt. The father and an unrelated female also resided in the group. The father spent only 1% of focal time in proximity with infants and never huddled with them, while the unrelated female spent 14% of her time in close proximity or huddling with infants. The grandmother spent 22% of her time in close proximity or huddling. The aunt spent 23% of her time in affliative contact with the infants, but also engaged in active care-giving such as carrying and allo-nursing. Finally, the mother spent 38% of her time with infants, primarily carrying and nursing. The three allo-mothers directed their behaviors primarily toward the older infant. The care-giving behavior of the mother, on the other hand, was primarily directed toward her newborn infant. Comparing Fig. 2a and b shows that the mean level of urinary prolactin increases across adults in a stepwise fashion with the amount and quality of infant care. The data set is too small for statistical analysis when including only 3 non-lactating adults (although they do exhibit the expected pattern), but across all five adults, prolactin is significantly associated with infant contact and care (all proximity and physical contact: Spearman’s correlation, r S=1.00, N=5, Pb0.01; physical contact only: r S=1.00, N=5, Pb0.01).

3.4. Effects of seasonality, age and contamination Seasonality and age are possible confounds that may affect prolactin levels [7,37]. Samples were collected across

a 2.0 1.5 1.0 .5 0.0 -.5 -1.0 N=

10

8

M

F

9

8

15

GM Aunt Mother

b Proportion of Time with Infants

0.0 N=

Mean ( ± SE) log Urinary PRL (ng/mg creatinine/mL)

(ng/mg creatinine/mL)

Mean (±SEM) Urinary PRL

298

.5 .4 .3

Nursing Carrying Huddling Proximity only

.2 .1 0.0

M

F

GM Aunt Mother

Fig. 2. Mean (FS.E.M.) adult urinary prolactin (PRL) levels for adults in group C in order of increasing PRL level (a), and proportion of time spent in affiliative interaction with infants in order of increasing affiliation (b). Urine was collected throughout the behavioral study. Urinary PRL increases in the same order as increasing infant affiliation. M=father of both infants, F=unrelated female, GM=grandmother of both infants, Aunt=aunt of both infants, and Mother=mother of both infants. For panel (a), number of combined samples (N) for each adult is indicated (see Methods). Adult means derived from log10-transformed individual sample values. Nontransformed PRL means are as follows (ng PRL/mg creatinine/mL urine): M=0.28, F=0.42, GM=0.78, Aunt=1.00, and Mother=108.91. For panel (b), one infant was born before the study began (age 6–12 months across observation period) and one infant was born during the study (age 2 months at end of observation period).

J. Soltis et al. / Physiology & Behavior 84 (2005) 295–301

Log Urinary PRL (ng/mg creatinine/mL)

3.0

299

4. Discussion

.1

2.5 2.0

4.1. Measuring prolactin in squirrel monkeys

1.5 1.0

-.7

M F GM A

.5 0.0 -.5 -1.0 -140

-100

-60

-20

20

60

Day from Parturition Fig. 3. Main graph: Log10 prolactin (PRL) levels in a pregnant female who gave birth and lactated during the study. Average gestation length in Saimiri sciureus is 170 days [8], so this female may have been 1–2 months pregnant at the beginning of the study. Non-transformed PRL values range from 0.17 to 288.21 (ng PRL/mg creatinine/mL urine). Urine collections ceased on day 82 and began again on day 55 when the pregnancy was discovered. Inset: Mean log10 prolactin values (ng PRL/mg creatinine/mL urine) for other members of the group were lower before the birth of the infant (filled bars), compared to after the birth (open bars). Relationship to infant: M=father, F=unrelated female, GM=grandmother, A=aunt.

different seasons. The group with infants was studied for 6 months, from Summer to Winter, while the remaining 2 groups were studied one after the other for shorter periods (2 to 2O months) during the 6 months of observation on the group with infants. In mammals, seasonal changes in prolactin are mediated by changes in daylight length [7], but these subjects all experienced a 12-h light/dark cycle throughout the study in indoor, temperature-controlled rooms, so the effects of seasonality should be minimal. Also, we could not find evidence for age effects. Age was not significantly correlated with urinary prolactin levels among non-lactating adult subjects (Spearman’s correlation, r S=0.331, N=11, P=0.320), and the mean age of nonlactating adults in the group with infants (18.0F3.0 S.E.M.) was similar to the mean age of adults in groups without infants (15.0F1.5; Mann–Whitney U test, Z= 0.840, N 1=3, N 2=8, P=0.497). The pregnant female exhibited extremely high levels of urinary prolactin, and cross-sample contamination could account for the overall high prolactin levels in that group. We were careful not to collect samples that were potentially cross-contaminated (see Methods), but we nevertheless tested for this possibility. If such contamination occurred with any regularity then the prolactin levels of other animals in the group should track the distinctive pattern found in the pregnant female, in which there was a peak during late pregnancy (see Fig. 3). In fact, all other members of the group show an increase in prolactin after the birth of the infant. Although insufficient power does not allow for statistical significance (Wilcoxon signed ranks test: Z= 1.826, N=4, P=0.068), the temporal pattern of prolactin values does not track that of the mother (see Fig. 3 and inset).

We found detectable levels of the hormone prolactin in squirrel monkey serum and urine samples. It has already been shown that prolactin is measurable in squirrel monkey serum [25], but this report demonstrates that prolactin is measurable in squirrel monkey urine, as it is in other New World primates (marmosets: [31]; tamarins: [37]). Squirrel monkeys do not usually urinate in large amounts or at specific times (e.g. an early morning void), however, so in most cases analyses of urine will not yield information on acute changes in prolactin (e.g. daily or diurnal changes). Assaying urine in squirrel monkeys is a useful non-invasive means of acquiring integrated prolactin measures over relatively long periods of time, however. There is controversy as to whether prolactin is excreted into and measurable in saliva [6,29]. We did not obtain detectable levels of prolactin in any of 16 squirrel monkey saliva samples although all subjects showed detectable levels of prolactin in simultaneously obtained serum samples. Similarly, prolactin was undetectable in human saliva using four different assays [30], but prolactin has been detected in rhesus monkey saliva samples [6,13]. 4.2. Factors associated with elevated prolactin in squirrel monkeys These results show for the first time in squirrel monkeys that prolactin is associated with mothering and allo-mothering as it is in many other species (sources in Introduction). Adults in a group with infants, including non-mothers, showed higher mean levels of urinary prolactin compared to adults in groups without infants. Furthermore, urinary prolactin increased with the amount and quality of infant care, and prolactin tracked changes in the reproductive state in one female who gave birth. Nevertheless, these results should be replicated with larger samples. While no causal model associating prolactin and parenting behavior can be developed from these correlational data, physical contact was one factor associated with increased PRL. Animals in visual, vocal and perhaps olfactory contact with infants in an adjacent cage (30 cm apart) showed no increase in prolactin (Fig. 1). In the group with infants, however, prolactin values were increased among adults, and prolactin increased in accordance with the level of physical contact with infants (Fig. 2). These results are consistent with studies of the common marmoset, another small-bodied new world primate. These studies showed that physical contact is necessary for an increase in plasma prolactin levels in non-mothers [5,22], and the relative increases in plasma PRL were similar in degree to the relative increases in urinary PRL shown here. Kinship was also associated with increases in PRL. Unrelated animals in an adjacent cage with visual, auditory

300

J. Soltis et al. / Physiology & Behavior 84 (2005) 295–301

and olfactory contact did not experience prolactin increases, and closely related females in the group with infants experienced the highest prolactin levels. Finally, the one female who allo-nursed an infant had lost her own (stillborn) infant 6 months before the study, and she may have been hormonally primed to lactate. In previous work on squirrel monkeys, kinship and reproductive priming have been implicated in allo-mothering [4,18,33], but this is the first showing a role for prolactin. Further work will be required to develop causal models for alloparental behavior in squirrel monkeys. Acknowledgements We thank the Priority One animal care staff at the NIH Animal Center, and Deborah Bernhards and Hope Donkin for many contributions including help with blood and saliva sample collection. Toni Ziegler facilitated conducting hormone assays at the Assay Services Unit at the Wisconsin National Primate Research Center. R. Lucille Roberts and three reviewers provided helpful comments on earlier drafts. J.S. was supported by a National Institutes of Health Intramural Training Award. References [1] Baldwin JD. The behavior of squirrel monkeys (Saimiri) in natural environments. In: Rosenblum LA, Coe CL, editors. Handbook of squirrel monkey research. New York7 Plenum Press, 1985. p. 35 – 53. [2] Boinski S. The social organization of squirrel monkeys: implications for ecological models of social evolution. Evol Anthropol 1999;8: 101 – 12. [3] Bridges RS, Numan M, Ronsheim PM, Mann PE, Lupini CE. Central prolactin infusions stimulate maternal behavior in steroid-treated, nulliparous female rats. Proc Natl Acad Sci 1990;87:8003 – 7. [4] DuMond FV. The squirrel monkey in a semi-natural environment. In: Rosenblum LA, Cooper RW, editors. The squirrel monkey. New York7 Academic Press, 1968. p. 87 – 145. [5] Dixson AF, George L. Prolactin and parental behaviour in a male new world primate. Nature 1982;299:551 – 3. [6] Dvoskin R, Lindell SG, Higley JD, Suomi SJ, Shoaf S. Reply [to 29]. Biol Psychiatry 2001;50:9 – 10. [7] Freeman ME, Kanyicska B, Lerant A, Nagy G. Prolactin: structure, function, and regulation of secretion. Physiol Rev 2000;80:1523 – 631. [8] Goss CM, Popejoy LT, Fusiler JL, Smith TM. Observations on the relationship between embryological development, time of conception, and gestation. In: Rosenblum LA, Cooper RW, editors. The squirrel monkey. New York7 Academic Press, 1968. p. 171 – 91. [9] Grattan DR. Behavioural significance of prolactin signalling in the central nervous system during pregnancy and lactation. Reproduction 2002;123:497 – 506. [10] Hershkovitz P. Taxonomy of squirrel monkeys genus Saimiri (Cebidae, Platyrrhhini): a preliminary report with description of a hitherto unnamed form. Am J Primatol 1984;6:257 – 312. [11] Hunt SM, Gamache KM, Lockard JS. Babysitting behavior by age/sex classification in squirrel monkeys (Saimiri sciureus). Primates 1978;19:179 – 86. [12] Leger DW, Mason WA, Fragaszy DM. Sexual segregation, cliques, and social power in squirrel monkey (Saimiri) groups. Behaviour 1981;76:163 – 81.

[13] Lindell SG, Suomi SJ, Shoaf S, Linnoila M, Higley JD. Salivary prolactin as a marker for central serotonin turnover. Biol Psychiatry 1999;46:568 – 72. [14] Livesey JH, Roud HK, Metcalf MG, Donald RA. Glycerol prevents loss of immunoreactive follicle-stimulating hormone and luteinizing hormone from frozen urine. J Endocrinol 1983;98:381 – 4. [15] MacLean PD. Mirror display in the squirrel monkey (Saimiri sciureus). Science 1964;146:950 – 2. [16] McNeilly AS, Abbott DH, Lunn SF, Chambers PC, Hearn JP. Plasma prolactin concentrations during the ovarian cycle and lactation and their relationship to return of fertility post partum in the common marmoset (Callithrix jacchus). J Reprod Fertil 1981; 62:353 – 60. [17] Mitchell CL, Boinski S, van Schaik CP. Competitive regimes and female bonding in two species of squirrel monkeys (Saimiri oerstedi and S. sciureus). Behav Ecol Sociobiol 1991;28:55 – 60. [18] L.S. Morton, Maternal and non-maternal contributions to infant caregiving in wild and captive Peruvian squirrel monkeys (Saimiri spp.). Unpublished Doctoral dissertation (in Animal Behavior). University of California, Davis; 2000. [19] Neville MC, Morton J, Umemura S. Lactogenesis: the transition from pregnancy to lactation. Pediatr Clin N Am 2001;48:35 – 52. [20] Ostrom KM. A review of the hormone prolactin during lactation. Prog Food Nutr Sci 1990;14:1 – 43. [21] Roberts RL, Jenkins KT, Lawler T, Wegner FH, Newman JD. Bromocriptine administration lowers serum prolactin and disrupts parental responsiveness in common marmosets (Callithrix j. jacchus). Horm Behav 2001;39:106 – 12. [22] Roberts RL, Jenkins KT, Lawler T, Wegner FH, Norcross JL, Bernhards DE, et al. Prolactin levels are elevated after infant carrying in parentally inexperienced common marmosets. Physiol Behav 2001;72:713 – 20. [23] Rosenblatt JS. Psychobiology of maternal behavior: contribution to the clinical understanding of maternal behavior among humans. Acta Paediatr 1994;397:3 – 8. [24] Rosenblum LA. Mother–infant relations and early behavioral development in the squirrel monkey. In: Rosenblum LA, Cooper RW, editors. The squirrel monkey. New York7 Academic Press, 1968. p. 207 – 33. [25] Scammell JG, von Haven R, Friesen HG, Wear LB, Thompson ME, Brady AG, et al. Characterization of prolactin and growth hormone immuno- and bioactivities in the pituitary gland and serum of the squirrel monkey (Saimiri boliviensis boliviensis). Am J Primatol 1992;26:35 – 46. [26] Smith HJ, Newman JD, Symmes D. Vocal concomitants of affiliative behavior in squirrel monkeys. In: Snowdon CT, Brown CH, Petersen MR, editors. Primate communication. Cambridge7 Cambridge University Press, 1982. p. 30 – 49. [27] Soltis J, Bernhards D, Donkin H, Newman JD. The squirrel monkey chuck call: vocal response to playback chucks based on acoustic structure and affiliative relationship with the caller. Am J Primatol 2002;57:119 – 30. [28] Storey AE, Walsh CJ, Quinton RL, Wynne-Edwards KE. Hormonal correlates of paternal responsiveness in new and expectant fathers. Evol Hum Behav 2000;21:79 – 95. [29] Thijssen JHH, Matute LMP, van Goozen SHM, van Engeland H, Blankenstein MA. No detectable prolactin in human saliva. Biol Psychiatry 2001;50:9. [30] Thijssen JHH, van Goozen SHM, van Engeland H, Matute LMP, Blankenstein MA. None of four commercially available assays detects prolactin in human saliva. Clin Chem 2000;46:1409 – 10. [31] Torii R, Moro M, Abbott DH, Nigi H. Urine collection in the common marmoset (Callithrix jacchus) and its applicability to endocrinological studies. Primates 1998;39:407 – 17. [32] Vaitl EA, Mason WA, Taub DM, Anderson CO. Contrasting effects of living in heterosexual pairs and mixed groups on the structure of social attraction in squirrel monkeys (Saimiri). Anim Behav 1978;26:358 – 67.

J. Soltis et al. / Physiology & Behavior 84 (2005) 295–301 [33] Williams L, Gibson S, McDaniel M, Bazzel J, Barnes S, Abee C. Allomaternal interactions in the bolivian squirrel monkey (Saimiri boliviensis boliviensis). Am J Primatol 1994;34:145 – 56. [34] Wynne-Edwards KE. Hormonal changes in mammalian fathers. Horm Behav 2001;40:139 – 45. [35] Wynne-Edwards KE, Reburn CJ. Behavioral endocrinology of mammalian fatherhood. Trends Ecol Evol 2000;15:464 – 8. [36] Ziegler TE, Scheffler G, Snowdon CT. The relationship of cortisol levels to social environment and reproductive function in female

301

cotton-top tamarins, Saguinus oedipus. Horm Behav 1995;29: 407 – 24. [37] Ziegler TE, Wegner FH, Snowdon CT. Hormonal responses to parental and nonparental conditions in male cotton-top tamarins, Saguinus oepidus, a new world primate. Horm Behav 1996; 30:287 – 97. [38] Ziegler TE. Hormones associated with non-maternal infant care: a review of mammalian and avian studies. Folia Primatol 2000; 71:6 – 21.