Grooming relationships between breeding ... - Wiley Online Library

American Journal of Primatology 69:1159–1172 (2007)

RESEARCH ARTICLE Grooming Relationships Between Breeding Females and Adult Group Members in Cooperatively Breeding Moustached Tamarins (Saguinus mystax) ¨ TTKER1, MAREN HUCK1, DIETMAR P. ZINNER2, PETRA LO 1 AND ECKHARD W. HEYMANN 1 Department of Behavioral Ecology and Sociobiology, German Primate Centre, Go¨ttingen, Germany 2 Department of Cognitive Ethology, German Primate Centre, Go¨ttingen, Germany

Grooming is the most common form of affiliative behavior in primates that apart from hygienic and hedonistic benefits offers important social benefits for the performing individuals. This study examined grooming behavior in a cooperatively breeding primate species, characterized by single female breeding per group, polyandrous matings, dizygotic twinning, delayed offspring dispersal, and intensive helping behavior. In this system, breeding females profit from the presence of helpers but also helpers profit from staying in a group and assisting in infant care due to the accumulation of direct and indirect fitness benefits. We examined grooming relationships of breeding females with three classes of partners (breeding males, potentially breeding males, (sub)adult non-breeding offspring) during three reproductive phases (post-partum ovarian inactivity, ovarian activity, pregnancy) in two groups of wild moustached tamarins (Saguinus mystax). We investigated whether grooming can be used to regulate group size by either ‘‘pay-for-help’’ or ‘‘pay-to-stay’’ mechanisms. Grooming of breeding females with breeding males and non-breeding offspring was more intense and more balanced than with potentially breeding males, and most grooming occurred during the breeding females’ pregnancies. Grooming was skewed toward more investment by the breeding females with breeding males during the phases of ovarian activity, and with potentially breeding males during pregnancies. Our results suggest that grooming might be a mechanism used by female moustached tamarins to induce mate association with the breeding male, and to induce certain individuals to stay in the group and help with infant care. Am. J. Primatol. 69:1159–1172, 2007. c 2007 Wiley-Liss, Inc. Key words: grooming; Saguinus mystax; tamarins; pay-to-stay; pay-for-help Correspondence to: Petra Lo ¨ttker, Department of Research and Documentation, Bavarian Forest National Park, Freyunger Str. 2, Grafenau, Germany. E-mail: [email protected]

Received 12 February 2006; revised 21 December 2006; revision accepted 21 December 2006 DOI 10.1002/ajp.20411 Published online 1 March 2007 in Wiley InterScience (www.interscience.wiley.com).

r 2007 Wiley-Liss, Inc.

¨ ttker et al. 1160 / Lo

INTRODUCTION Allogrooming is the most common affiliative behavior in primates and constitutes a major part of the total time spent in social interaction [Goosen, 1987]. It brings hygienic and hedonistic benefits to the recipients through removal of dirt and ectoparasites, increased production of X-endorphins or decreased heart rate [Boccia et al., 1989; Hart et al., 1992; Keverne et al., 1989], and also social benefits for the performing individuals [Dunbar, 1991]. Allogrooming can be used to establish and maintain mate association, for social integration, calming effects, and reconciliation (reviewed in Chadwick-Jones [1998]), and also as long-term investment, e.g., to obtain coalitionary support [Seyfarth, 1977]. Hence, allogrooming can indirectly affect an animal’s fitness by influencing a partner’s future behavior. This influence could be particularly important in social systems that depend on a high degree of cooperation, like in callitrichids. In these small Neotropical primates adult and subadult group members cooperate in the care for twin offspring born to the usually single breeding female [see Heymann, 2003a, for review]. Dispersal is often delayed, and breeding females seem to be able to regulate group size by adapting their degree of aggressiveness, which could lead to immigrations and evictions [McGrew & McLuckie, 1986; Schaffner & French, 1997]. Non-breeding individuals, on the other hand, are valuable assistants in vigilance and territorial defence, and are crucial for successful infant rearing [e.g., Goldizen, 1987]. Therefore, breeding females might use grooming (besides aggressiveness) as an additional mechanism to regulate group size by using it as an incentive for non-breeding individuals (‘‘helpers’’) to stay in the group and help with infant care (‘‘pay-for-help’’; Lazaro-Perea et al. [2004]). However, in callitrichids helpers can profit as well from staying in a group and assisting with offspring care [Huck et al., 2004b]. Direct benefits include enhanced survival, and the acquisition of care-giving experience which translates into higher reproductive success when helpers become breeders [cf. Emlen et al., 1991]. As genetic relatedness in callitrichid groups was found to be high [Huck et al., 2005; Nievergelt et al., 2000], helpers can gain indirect fitness benefits through assisting in rearing the infants of close kin. Thus, non-breeding individuals may pay breeding females for their tolerance, thereby increasing the time they can stay in their natal group ńchez et al. [2002]; food for (‘‘pay-to-stay’’; infant-carrying for tolerance: Sa tolerance: Rapaport [2001]). Besides offering infant-carrying and provisioning, grooming might be a mechanism used by non-breeding individuals to maintain group membership. Whether ‘‘pay-for-help’’ or ‘‘pay-to-stay’’ actually work in a callitrichid group should depend on the importance of helping and the risk of eviction at a given point of time and hence should be influenced by group size and composition (presence of dependent offspring, number of helpers), and the reproductive phase of breeding females. Despite the advantages of larger groups in terms of infant care and predator avoidance, group size in callitrichids seems to be limited as the number of members usually does not exceed nine in tamarins and 16 in ¨ttker et al., 2004a]. Not all group members are marmosets [Heymann, 2003a; Lo equally threatened by eviction, and eviction risk is not equally distributed over time. Older (female) offspring seem to face an especially high risk of eviction due to aggression received from breeding females [McGrew & McLuckie, 1986; ńchez et al., 2002], and the probability of eviction seems to be higher around Sa breeding females’ conceptive periods [Lo¨ttker et al., 2004a].

Am. J. Primatol. DOI 10.1002/ajp

Grooming in Wild Tamarins / 1161

The aim of our study was to investigate grooming behavior of breeding females with three different classes of helpers (breeding males, potentially breeding males, non-breeding offspring) in different reproductive phases (postpartum ovarian inactivity, ovarian activity, pregnancy) in two groups (E and W) of wild moustached tamarins (Saguinus mystax) with known genetic relation¨ttker et al., 2004a]. If grooming serves to regulate ships [Huck et al., 2005; Lo group size (‘‘pay-for-help’’ or ‘‘pay-to-stay’’), it is predicted that (1) during times of infant dependency (or shortly before parturition), breeding females should increase their investment in grooming relationships with helpers to give them an incentive to stay and help; (2) during fertile periods of the breeding female, grooming between breeding females and breeding males should be intensified both as a form of courtship behavior and to establish or maintain mate association (i.e., to secure the breeding position); (3) during pregnancy, i.e., when a new breeding cycle starts and previous infants are available as new helpers, (older) helpers should increase their investment in grooming relationships with breeding females to be allowed to remain in the group; (4) in groups with less helpers (group W) breeding females should groom helpers more than in groups with more helpers (group E).

METHODS Study Site The study was carried out at the Estacioń Biolo´gica Quebrada Blanco (EBQB), in northeastern Peruvian Amazonia (41210 S 731090 W; for details of the ¨ttker et al. [2004a]). Total annual rainfall was 2,781 mm, with the study site see Lo dry season (monthly precipitation o200 mm) lasting from June to September. Study Animals We studied two groups (W, E) of well-habituated and individually known moustached tamarins with known genetic relationships [Table I; Huck et al., 2005; Lo¨ttker et al., 2004a]. The breeding females’ potential grooming partners were assigned into three classes: (1) breeding males (BM 5 fathers of offspring, additionally characterized by mate-guarding behavior: Huck et al. [2004a]), (2) potentially breeding males (PBM 5 adult males unrelated to and copulating with the breeding female), and (3) non-breeding offspring (NBO 5 (sub)adult offspring of the breeding pair).

Reproductive Phases of the Breeding Females Hormonal profiles of breeding females as determined by measuring progestogen- and estrogen metabolites in fecal samples revealed three different reproductive phases: (1) post-partum ovarian inactivity (PP 5 consistently low baseline hormone levels following parturition), (2) ovarian activity (oA 5 regularly fluctuating hormone levels), and (3) pregnancy (P 5 150 days preceding parturition, with elevated hormone levels during mid/late gestation) [for details ¨ttker et al., 2004b]. Phases of post-partum ovarian inactivity lasted for 54 see Lo (WF1) and 64–82 days (EF1). In both groups, PP was synchronous with infant dependency (lactation and obligatory infant-carrying: M. Huck & P. Lo¨ttker, unpublished data). Phases of oA lasted for 88 (WF1) and 124–142 days (EF1).



Breeding female Breeding male Potentially breeding male Non-breeding offspring Non-breeding offspring E: 16 Sep 2001 Non-breeding offspring E: 16 Sep 2001 Non-breeding offspring B: 12 May 2000

F M M M F F M

B: 22 Feb 2000 E: 05 Dec 2001 B: 22 Feb 2000

Breeding female Breeding male Potentially breeding male Potentially breeding male Non-breeding offspringd Non-breeding offspringd

D: 03 Dec 2001 E: 14 Dec 2001

Demographic notesb

F M M M M M

Classa

Z31 Z31 Z31

13 13

PP

Z34 Z34 Z34 13–14

14–16 14–16

oA

Z39 Z39 Z39 15–19

17–21 17–21

P

Age (months) of non-breeding offspring in the reproductive phasesc

c

b

The classes were determined using genetic data on parentage (see Huck et al. [2005]) and observations on copulatory behavior (see text) ¨ttker et al. [2004a]. D, died; E, emigrated; B, born; for details on the demography of the study groups see Lo PP, post-partum phase of ovarian inactivity; oA, ovarian activity; P, pregnancy; EM1, EF2, and EF3 were already adult , i.e. Z19 months old, when habituating the group in January 2000. d WM4 and WM5 were not offspring of female WF1 but of her probable sister WF2, the former reproductive female in the group, and male WM1. After WF2 left the group, WF1 suckled WM4 and WM5 [Smith et al., 2001], and took over the breeding position together with WM1 [Huck et al., 2005; Lo¨ttker et al., 2004a]. e EM4 was still juvenile in PP so that he was included in data analysis only from oA onward (compare Methods).

a

Group W WF1 WM1 WM2 WM3 WM4 WM5 Group E EF1 EM3 EM2 EM1 EF2 EF3 EM4e

Individual Sex

TABLE I. Summary Characteristics of the Study Animals



Collection of Behavioral Data Between January and December 2001, both groups were followed daily from the time they left a sleeping site in the morning till retirement to a sleeping site in the afternoon (average daily observation time 5 9.2 h), for a total of 3,004 contact hours on 330 days (W), and 3,257 contact hours on 351 days (E). Behavioral data were collected by the first two authors and two assistants (Marcos Oversluijs V. and Jenni Pe´rez J.) using a rotatory schedule so that each observer worked with each group in each reproductive phase. Given potential inconsistencies in collecting behavioral data using multiple data collectors, we examined interobserver reliability. Mean (7SD) inter-observer reliability was 91.570.13% [Martin & Bateson, 1993], and we feel this is an indication of overall consistency and reliability. Grooming data were collected during 10-min focal-animal samples with continuous recording, and by behavioral sampling [Martin & Bateson, 1993] between focal samples. The majority of data stemmed from behavioral sampling (93%), whereas 7% of data derived from focal samples. Focal samples were conducted on all adult and subadult group members. Focal animals were selected following a rotational scheme, in which the order was randomly determined before the start of observations. Additionally, we conducted opportunistic samples on individuals that were actually visible, whenever a selected animal could not be detected. We further conducted hourly scan samples in which we recorded the activity of each identified individual that became visible within 2 min. These data were used to estimate the visibility of individual animals. As ‘‘grooming’’ we defined any interaction in which one animal picked through the fur of another with its hands or mouth. For each grooming interaction we noted groomer and groomee, and determined its duration. The interaction was defined to end when a shift in activity occurred for more than 10 s. If an animal was involved in grooming with more than one partner simultaneously, we recorded grooming time separately for each dyad involved.

Data Analysis With breeding females as the focus of our study, we only analyzed dyads involving these females. We restricted data to adult and subadult individuals ¨ttker et al. [2004a]). According to their ages, males (age Z13 months; Lo WM4 and WM5 were included in the analyses from 22 March, EM4 from 12 June onward (Table I). In both groups, we started data analyses with the parturition date, and continued throughout the breeding females’ three reproductive phases. For each reproductive phase we summed up the duration of grooming that breeding females directed to and received from each group member. As the majority of grooming data stemmed from behavioral sampling records (see above), individuals were not equally visible and the three reproductive phases differed in duration, values were corrected (D) by including visibility data of each animal, using the results of our scan samplings (see Table II). The total duration of grooming an animal A directed to an animal B in each phase (tDA phase n) was divided by the sum of scans for animal A and animal B in that phase (SA phase n and SB phase n, respectively, note that the sum of scans for each individual refers to the number of scans it was actually visible; see above). These corrected figures were then multiplied by an integer which represented the number of scans collected for



PP oA P

EF1

WM2

WM3

WM4

WM5

170/770 (103/55) 1,420/380 (138/88) 5,610/5,380 (251/173)

EM3 0/600 (103/73) 70/190 (138/146) 860/170 (251/201)

EM2

EF2d

0/0 (103/73) 230/1,130 (103/83) 580/860 (138/140) 1,160/1,880 (138/119) 16,680/18,670 (251/151) 340/220 (251/11)

EM1

250/470 (103/93) 510/100 (138/215) 840/950 (251/35)

EF3d

1,230/980 (148/104) 750/0 (148/64) 340/220 (148/77) 80/10 (148/44) 190/430 (148/43) 8,710/1,400 (237/199) 0/0 (237/82) 1,490/430 (237/149) 360/50 (237/172) 1,380/1,310 (237/139) 13,170/9,200 (324/279) 3,160/290 (324/136) 10,250/5,910 (324/206) 12,240/5,230 (324/213) 18,700/9,540 (324/220)

WM1

/ (103/) 0/0 (138/47) 2,530/1,470 (251/154)

EM4e

b

Refers to the number of scans an individual was actually visible. Bold letters indicate breeding individuals, normal letters indicate potentially breeding males, italic letters indicate non-breeding offspring; ‘‘F’’ stands for female, ‘‘M’’ stands for male individual. c PP, post-partum phase of ovarian inactivity; oA, ovarian activity; P, pregnancy. d EF2 and EF3 emigrated during P so that they are represented in fewer scans than the other animals. e EM4 was still juvenile in PP so that he was included in data analysis only from oA onward (compare Methods).

a

PP oA P

WF1

Femaleb Phasec

Group memberb

TABLE II. Total Duration [s] of Grooming Given/Received by Breeding Females Toward/From Different Group Members in Each Reproductive Phase (No of Scansa per Female/Grooming Partner is given in Parenthesis)



the most visible dyad (XY) of each study group in each phase [Arnold & Whiten, 2003]: tDA phase n SXY phase n: D¼ SA phase n þ SB phase n For each dyad, we also determined indices of reciprocity in time (RIs) per reproductive phase following Nishida [1988], with TA being the time animal A grooms B, and TB the time animal B grooms A: RI ¼

ðTA TB Þ : ðTA þ TB Þ

Indices of reciprocity can range from 1 to 1. Values from 0.2 to 0.2 indicate balanced grooming, values from 0.2 to 1.0 indicate that animal A grooms animal B longer than vice versa, and values from 0.2 to 1.0 indicate that animal B grooms animal A longer than vice versa [Nishida, 1988]. In our case, A refers to the breeding female and B to the different grooming partners. We calculated means for each partner class in each phase in each group from corrected durations and RIs and applied them to two-way repeated measures ANOVAs using partner and group as independent variables and the three phases as within-subject factor, to evaluate differences in grooming behavior of breeding females with the three classes of grooming partners in the three reproductive phases. We included only main effects in the analyses as our ANOVA design did not allow testing interactions. All statistics were two-tailed and were performed with STATISTICA 6.0. Significance levels were set at Pr0.05, whereas values between 0.05oPr0.1 indicated a trend toward a significant result. RESULTS Overall Time Spent Grooming by Individual Tamarins Over the whole study period, animals from both study groups spent on average 3.2% of their total activity time in grooming (Table III). With 3.2 and 5.4% for WF1 and EF1, respectively, the percentage of time breeding females spent in social grooming was equal to or slightly higher than the mean. Most grooming occurred during pregnancies, where animals spent on average 6.8% of time in grooming interaction. Grooming Given by Breeding Females The amount of grooming breeding females provided varied between the two groups and the reproductive phases, and tended to vary between the classes of partners (two-way repeated measures ANOVA: group: F1,2 5 41.63, P 5 0.023; partner: F2,2 5 12.57, P 5 0.075; reproductive phase: F2,4 5 29.18, P 5 0.004; Fig. 1a). The breeding female from group W groomed her partners longer than the breeding female from group E. In both groups, there was a tendency toward BM and NBO obtaining the highest amounts of grooming. Both females groomed their partners longer during pregnancy than during PP and oA. Apart from high amounts of grooming directed toward all classes of partners during their pregnancies, breeding females directed conspicuously high amounts of grooming toward the respective breeding male during oA. Most of the variance was explained by the variable ‘‘reproductive phase’’ (49.2%). The variables ‘‘group’’ and ‘‘partner’’ explained 15.3 and 9.0% of the variance, respectively.


¨ ttker et al. 1166 / Lo TABLE III. Percentage of Time Spent Grooming in Focal Samples by Each Individual in Each Reproductive Phase (No. of Focal Minutesa is Given in Parenthesis) Reproductive phasec Individualb Group W WF1 WM1 WM2 WM3 WM4 WM5 Group E EF1 EM3 EM2 EM1 EF2d EF3d EM4e Mean

PP

oA

5.6 (249) 0 (274) 0 (359) 0 (214) 0 (51) 0 (52)

1.1 0.5 0.3 3.0 0 0.3

(452) (389) (377) (364) (167) (363)

1.2 (322) 0.9 (222) 4.1 (271) 1.3 (306) 3.6 (253) 3.3 (273) (0) 1.7 (2,864)

3.3 (364) 1.0 (391) 1.3 (468) 0 (406) 0 (431) 0.7 (539) 0 (59) 0.9 (4,770)

P 3.8 4.6 7.8 8.3 5.2 5.8

(581) (630) (708) (654) (230) (226)

9.3 (551) 2.8 (466) 4.7 (513) 6.2 (500) 8.3 (72) 15.0 (60) 6.4 (216) 6.8 (5,348)

Total 3.2 2.4 3.9 5.3 2.7 2.2

(1,282) (1,293) (1,444) (1,232) (448) (641)

5.4 1.8 3.3 2.9 2.0 2.5 4.6 3.2

(1,237) (1,079) (1,252) (1,212) (756) (872) (216) (12,964)

a

Refers to the time animals were actually visible and not ‘‘out of sight’’. Bold letters indicate breeding individuals, normal letters indicate potentially breeding males, italic letters indicate non-breeding offspring; ‘‘F’’ stands for female, ‘‘M’’ stands for male individual. c PP, post-partum phase of ovarian inactivity; oA, ovarian activity; P, pregnancy. d EF2 and EF3 emigrated during P so that they are represented in fewer focal samples than the other animals. e EM4 was still juvenile in PP so that he was included in data analysis only from oA onward (compare Methods). b

Grooming Received by Breeding Females The amount of grooming breeding females received varied between the classes of partners and the reproductive phases and tended to vary between groups (two-way repeated measures ANOVA: group: F1,2 5 16.93, P 5 0.054; partner: F2,2 5 35.84, P 5 0.027; reproductive phase: F2,4 5 447.66, P 5 0.00002; Fig. 1b). The breeding female from group W received more grooming from her partners than the female from group E. In both groups, breeding females received more grooming from BM and NBO than from PBM. Breeding females received more grooming while being pregnant than during oA and PP. Again, most of the variance was explained by the variable ‘‘reproductive phase’’ (63.1%). The variable ‘‘partner’’ explained 12.5% of the variance. Almost no variance was explained by the variable ‘‘group’’ (2.9%). Indices of Reciprocity Indices of reciprocity tended to vary between groups and reproductive phases, but partners had no effect (two-way repeated measures ANOVA: group: F1,2 5 12.69, P 5 0.071; partner: F2,2 5 0.05, P 5 0.95; reproductive phase ANOVA: F2,4 5 6.71, P 5 0.053; Fig. 2). Indices of reciprocity tended to be lower in group E than in group W, i.e., grooming in group E was either more balanced or skewed in favor of the breeding female’s partners. In both groups, the difference between the three partner classes was not significant. In both groups, the lowest values ( 5 skew in favor of the females’ partners) were found during PP. Indices


Grooming in Wild Tamarins / 1167 300 (a) Grooming given 250 200 150

Corrected duration [min]

100 50 0 300 (b) Grooming received 250 200 150 100 50 0 Phase: Partner:

PP

oA

P

breeding male

PP

oA

P

potentially breeding male

PP oA

P

non-breeding offspring

Fig. 1. Mean duration (7SE) of grooming breeding females (a) gave to and (b) received from the three partner classes during post-partum ovarian inactivity (PP), ovarian activity (oA), and pregnancy (P). , group W; J, group E. Values for ‘‘breeding males’’ in both groups and for ‘‘potentially breeding males’’ in group E represent individual values of one dyad, whereas all other values are means of 2–4 dyads (compare Table I). For original values see Table II.

of reciprocity were skewed in favor of the breeding females with BM during oA and with PBM during their pregnancies. Breeding females groomed most balanced with BM and NBO during their pregnancies. Most of the variance was explained by the variables ‘‘group’’ and ‘‘reproductive phase’’ (30.9 and 25.4%, respectively). Almost no variance was explained by the variable ‘‘partner’’ (0.3%). DISCUSSION Grooming was a prominent behavior in the daily activity of the moustached tamarins. Time spent in grooming interaction (3.2%) fell well within the range reported previously for wild tamarin and marmoset species (1.4–2.4% for S. mystax: Heymann [1996]; 9.1% for S. fuscicollis: Goldizen [1989]; 14.0–14.7% for C. jacchus: Digby [1995]; Lazaro-Perea et al. [2004]). Over all reproductive phases, breeding females spent more than average time in giving and receiving



grooming, and maintained intensive grooming relationships with all group members. A number of studies confirmed the central role of breeding females in grooming interactions [e.g., Digby, 1995; Lazaro-Perea et al., 2004; Vogt, 1978]. The expected and observed directions of grooming between breeding females and the different partner classes according to the four predictions we made are summarized in Table IV. Our first prediction was that during times of infant dependency, breeding females should increase their investment in grooming relationships with helpers to give them an incentive to stay and help. Opposing this prediction we found that grooming was either balanced or that partners groomed the breeding females longer than vice versa during this phase. Time or energy constraints during infant dependency could preclude immediate reciprocation of grooming and services provided, and helpers could be paid at another time (see below; LazaroPerea et al., 2004). In callitrichids, all group members are involved in rearing the

1.0 0.8 breeding female grooms longer

Index of reciprocity

0.6 0.4 0.2

balanced grooming

0 -0.2 -0.4

partner grooms longer

-0.6 -0.8 -1.0 Phase:

PP

Partner:

breeding male

ooA

P

PP

oA

P

potential breeding male

PP

oA

P

non-breeding offspring

Fig. 2. Mean indices of reciprocity (7SE) for the breeding females with the three partner classes during post-partum ovarian inactivity (PP), ovarian activity (oA), and pregnancy (P). , group W; J, group E. Values for ‘‘breeding males’’ in both groups and for ‘‘potentially breeding males’’ in group E as well as in group W during oA represent individual values of one dyad, whereas all other values are means of 2–4 dyads (compare Table I).

TABLE IV. Expected and Observed Directions/Amounts of Grooming According to the Predictions Prediction 1 2 3 4 a

Phasea Expected direction of groomingb PP oA P All

BF4BM; NBO; PBM BF4BM BFoNBO WF14EF1

Observed direction of groomingb BFrBM; NBO; PBM BF4BM BF 5 NBO WF14EF1

PP, post-partum phase of ovarian inactivity, oA, ovarian activity, P, pregnancy. BF, breeding female; BM, breeding male; NBO, non-breeding offspring; PBM, potentially breeding male.

b



twin infants and carrying and feeding them are time and energy demanding tasks [Huck et al., 2004b]. Therefore, there might not only be less time for grooming due to the handling of offspring but also due to the need to restock energy, e.g., by increased feeding activity. In fact, time spent grooming was generally lower during infant dependency (Table III). The skew in favor of helpers could likewise result from increased attraction and as a means to obtain access to infants [Goosen, 1987; Henzi & Barrett, 2002; O’Brien, 1993]. Infant callitrichids indeed are attractive for all group members, as quarrels over infant-carrying, i.e., a noncarrier attacking a current carrier and trying to pull down the infant from his ¨ttker & M. Huck [personal observation]; back, were frequently observed (P. Lo see also Price [1991]). Thus, it is possible that non-breeding individuals pay for the privilege to provide care because it gives them rearing experience [Emlen et al., 1991; Henzi & Barrett, 2002] or puts them in good social relation with the breeding female already in this phase of the reproductive cycle (‘‘pay-to-stay’’; see below). Our second prediction was that during fertile periods of the breeding female, grooming between breeding females and breeding males should be intensified both as a form of courtship behavior and to establish or maintain mate association. This is supported by our result that reciprocity indices between breeding females and breeding males were clearly skewed in favor of females in their phases of ovarian activity. Similarly estrous female capuchin monkeys give more grooming to adult males than vice versa, which is consistent with active female solicitations of copulations in this species [DiBitetti, 1997]. Female-skewed grooming during the receptive phase in our study is in accordance with the general need for female tamarins to constantly advertise her quality in a system with strong female breeding competition and expected male choosiness [Heymann, 2003b]. Supporting this interpretation, a young adult female moustached tamarin increased her grooming contribution to about 30% and groomed the adult males more often after the death of the breeding female [Heymann, 1990]. At the same time, her vulva increased from medium to full adult size, which might be interpreted as the onset of breeding activity. When the breeding female died, the group comprised (in addition to the young adult female) two adult males, one subadult male, two juvenile males, and a subadult female and had been stable for at least 6 months (E.W. Heymann, personal communication). Strengthening the social relation with adult males may be important even when no other adult female is present in a group as they face the risk of adult female immigration and breeding male emigration [Lo¨ttker et al., 2004a]. As in our study, grooming relations between breeding female and male common marmosets and pygmy marmosets were more intense around conception, but the direction of grooming was not specified [Albuquerque et al., 2001; Soini, 1987]. Studies in other wild callitrichids where males groomed breeding females longer than vice versa [Goldizen, 1989; Heymann, 1996; Lazaro-Perea et al., 2004] did not control for female reproductive phase and thus are not directly comparable. Our third prediction was that during pregnancy, i.e., when a new breeding cycle starts and previous infants are available as new helpers, (older) helpers should increase their investment in grooming relationships with breeding females to be allowed to remain in the group. This prediction is not supported by our result that grooming between breeding females and non-breeding offspring was highly balanced. However, apparently balanced grooming relationships between these classes might result from the lack of distinction between ages and sexes within the non-breeding offspring. In larger groups of common marmosets (Z6 non-breeding females) the oldest ones groomed the breeding female more than



vice versa (‘‘pay-to-stay’’), whereas younger females received more grooming from the breeding female than they gave (‘‘pay-for-help’’) [Lazaro-Perea et al., 2004]. Indeed, in our study, too, some offspring appeared to groom their mothers more than vice versa whereas others were groomed by their mothers more, a difference which disappears when building the mean. In group E, it might well be that age differences have played a role, as the youngest offspring (EM4) received more grooming than he gave and two of the older offspring (EM1, EF3) gave more than they received (Tables I and II). However, unfortunately it is not known which of the already adult offspring was indeed the oldest. Notably, during pregnancy grooming between breeding females and potentially breeding males was highly skewed toward females providing more than they received. Additionally, the breeding male from group W (WM1) directed comparatively high amounts of grooming toward one of the two potentially breeding males (WM2), while receiving little (data not shown), an observation that was also reported by Heymann [1996]. In both groups it was one of the potentially breeding males that did most of the infant-carrying, whereas both mother and father carried less than all any other adult group member [Huck et al., 2004b]. Hence, the high value of potentially breeding males for infant care might have led to the exchange of grooming by breeders to pay them for their help (see above). Our forth prediction was that in groups with less helpers (group W) breeding females should groom helpers more than in groups with more helpers (group E). This prediction is supported by our results that across all phases the breeding female from group W groomed her partners longer than the breeding female from group E, and that indices of reciprocity were either more balanced or partners groomed the female longer than vice versa in group E. Group E had four nonbreeding offspring and only one potentially breeding male, whereas in group W two non-breeding offspring and two potentially breeding males were present. This would represent a situation under which biological market mechanisms operate, so that according to the law of supply and demand helpers would be more valuable in group W than in group E and thus would be able to demand a higher price for their help [cf. Barrett et al., 1999; Noe¨ & Hammerstein, 1994, 1995]. Additionally, both non-breeding offspring in group W were not the sons of the breeding female (Table I). Thus, group members in group W were less closely related to recent infants in the group, and might have needed a higher incentive to help with infant care. In contrast, in group E they might have had to pay for staying, especially since with nine group members the group was already quite large. A negative correlation between group size and asymmetry in grooming between breeding and non-breeding female common marmosets [Lazaro-Perea et al., 2004] supports our assumption. Whether differences in group composition can really lead to differences in grooming investment as discussed here, or whether the differences found between the two groups were just caused by female idiosyncracies [Borries et al., 1994; Izawa, 1980] remains to be shown. In sum, the fact that grooming patterns changed over the annual reproductive cycle and that breeding females preferred or were preferred by different partners at different stages of the reproductive cycle suggests that grooming may be used by moustached tamarins as an additional mechanism to regulate group size either by ‘‘pay-for-help’’ or by ‘‘pay-to-stay’’. More specifically, breeding females might use grooming to induce mate association with the breeding male, as well as to induce certain individuals to stay in the group and help with infant care. As our results are based on a small sample size our



conclusions will have to remain tentative and will have to be confirmed in future studies. ACKNOWLEDGMENTS This study is part of the PhD project of the first author at the Department of Behavioral Biology, Institute for Neuro- and Behavioral Biology, University of Muenster, Germany. Fieldwork was carried out under a letter of understanding between the Universidad Nacional de la Amazonıá Peruana (Iquitos, Peru) and the Deutsches Primatenzentrum and complied with Peruvian laws. Financial support was provided by the DFG (HE 1870/10-1,2). We heartily thank our field assistants Camilo Flores A. and Ney Shahuano T. for their invaluable work in the field. We thank Louise Barrett, Ronald Noe¨, Markus Port, Roland Hilgartner, Paul Garber, and two anonymous referees for helpful and valuable comments on earlier drafts of this manuscript, and Traute Huck for improving the English. REFERENCES Albuquerque ACSR, Sousa MBC, Santos HM, Ziegler TE. 2001. Behavioral and hormonal analysis of social relationships between oldest females in a wild monogamous group of common marmosets (Callitrix jacchus). Int J Primatol 22:631–645. Arnold K, Whiten A. 2003. Grooming interactions among the chimpanzees of the Budongo forest, Uganda: tests of five explanatory models. Behaviour 140:519–552. Barrett L, Henzi SP, Weingrill T, Lycett JE, Hill RA. 1999. Market forces predict grooming reciprocity in female baboons. Proc R Soc Lond B 266:665–670. Boccia ML, Reite M, Laudenslager M. 1989. On the physiology of grooming in a pigtail macaque. Physiol Behav 45:667–670. Borries C, Sommer V, Srivastava A. 1994. Weaving a tight social net: allogrooming in free-ranging female langurs (Presbytis entellus). Int J Primatol 15:421–443. Chadwick-Jones J. 1998. Social exchange and grooming partnerships. In: Chadwick-Jones J, editor. Developing a social psychology of monkeys and apes. East Sussex: Psychology Press. p 101–109. DiBitetti MS. 1997. Evidence for an important social role of allogrooming in a platyrrhine primate. Anim Behav 54:199–211. Digby LJ. 1995. Social organization in a wild population of Callithrix jacchus: II. intragroup social behavior. Primates 36:361–375. Dunbar RIM. 1991. Functional significance of social grooming in primates. Folia Primatol 57:121–131. Emlen ST, Reeve HK, Sherman PW, Wrege PH, Ratnieks FLW, Shellman-Reeve J. 1991. Adaptive versus nonadaptive explanations of behavior: the case of alloparental helping. Am Nat 138:259–270. Goldizen AW. 1987. Facultative polyandry and the role of infant-carrying in wild saddle-

backed tamarins. Behav Ecol Sociobiol 20: 99–109. Goldizen AW. 1989. Social relationships in a cooperatively polyandrous group of tamarins (Saguinus fuscicollis). Behav Ecol Sociobiol 24:79–89. Goosen C. 1987. Social grooming in primates. In: Mitchell G, Erwin J, editors. Comparative primate biology Vol. 2 Part B: behavior, cognition, and motivation. New York: Alan R. Liss. p 107–131. Hart BL, Hart LA, Mooring MS, Olubayo R. 1992. Biological basis of grooming behaviour in antelope: the body-size, vigilance and habitat principles. Anim Behav 44: 615–631. Henzi SP, Barrett L. 2002. Infants as a commodity in a baboon market. Anim Behav 63:915–921. Heymann EW. 1990. Social behaviour and infant-carrying in a group of moustached tamarins, S. mystax (Primates: Platyrrhini: Callitrichidae), on Padre Isla, Peruvian Amazonia. Primates 31:183–196. Heymann EW. 1996. Social behavior of wild moustached tamarins, Saguinus mystax, at the Estacioń Biolo´gica Quebrada Blanco, Peruvian Amazonia. Am J Primatol 38: 101–113. Heymann EW. 2003a. New world monkeys II: marmosets, tamarins, and Goeldi’s monkeys (Callitrichidae). In: Kleiman DG, Geist V, Hutchins M, McDade MC, editors. Grzimek’s animal life encyclopedia, Vol. 14: mammals III. Farmington Hills: Gale Group. p 115–133. Heymann EW. 2003b. Scent marking, paternal care, and sexual selection in callitrichines. In: Jones CB, editor. Sexual selection and reproductive competition in primates: new perspectives and directions (Special Topics in Primatology, Vol. 3).


¨ ttker et al. 1172 / Lo Oklahoma: American Society of Primatologists, Norman. p 305–326. ¨ttker P, Heymann EW. 2004a. Huck M, Lo Proximate mechanisms of reproductive monopolization in male moustached tamarins (Saguinus mystax). Am J Primatol 64: 39–56. Huck M, Lo¨ttker P, Heymann EW. 2004b. The many faces of helping: Possible costs and benefits of infant-carrying and food transfer in moustached tamarins (Saguinus mystax). Behaviour 141:915–934. ¨ttker P, Bo¨hle U-R, Heymann EW. Huck M, Lo 2005. Paternity and kinship patterns in polyandrous moustached tamarins (Saguinus mystax). Am J Phys Anthropol 127: 449–464. Izawa K. 1980. Social behavior of the wild black-capped capuchin (Cebus apella). Primates 21:443–467. Keverne EB, Martensz ND, Tuite B. 1989. Beta-endorphin concentrations in cerebrospinal fluid of monkeys are influenced by grooming relationships. Psychoneuroendocrinology 14:155–161. ´tima Arruda M, SnowLazaro-Perea C, de Fa don CT. 2004. Grooming as a reward? Social function of grooming between females in cooperatively breeding marmosets. Anim Behav 67:627–636. Lo¨ttker P, Huck M, Heymann EW. 2004a. Demographic parameters and events in wild moustached tamarins (Saguinus mystax). Am J Primatol 64:425–449. Lo¨ttker P, Huck M, Heymann EW, Heistermann M. 2004b. Endocrine correlates of reproductive status in breeding and nonbreeding wild female moustached tamarins. Int J Primatol 25:919–937. Martin P, Bateson P. 1993. Measuring behaviour—an introductory guide. Cambridge: Cambridge University Press. McGrew WC, McLuckie EC. 1986. Philopatry and dispersion in the cotton-top tamarin, Saguinus (o.) oedipus: an attempted laboratory simulation. Int J Primatol 7:401–422. Nievergelt CM, Digby LJ, Ramakrishnan U, Woodruff DS. 2000. Genetic analysis of


group composition and breeding system in a wild common marmoset (Callithrix jacchus) population. Int J Primatol 21:1–20. Nishida T. 1988. Development of social grooming between mother and offspring in wild chimpanzees. Folia Primatol 50:109–123. Noe¨ R, Hammerstein P. 1994. Biological markets: supply and demand determine the effect of partner choice in cooperation, mutualism and mating. Behav Ecol Sociobiol 35:1–11. Noe¨ R, Hammerstein P. 1995. Biological markets. Trends Ecol Evol 10:336–339. O’Brien TG. 1993. Allogrooming behaviour among adult female wedge-capped capuchin monkeys. Anim Behav 46:499–510. Price EC. 1991. Competition to carry infants in captive families of cotton-top tamarins (Saguinus oedipus). Behaviour 118:66–88. Rapaport LG. 2001. Food transfer among adult lion tamarins: mutualism, reciprocity or one-sided relationships? Int J Primatol 22:611–629. ńchez S, Pela éz F, Gil-Bu Sa ¨ rmann C. 2002. Why do cotton-top tamarin female helpers carry infants? A preliminary study. Am J Primatol 57:43–49. Schaffner CM, French JA. 1997. Group size and aggression: ‘recruitment incentives’ in a cooperatively breeding primate. Anim Behav 54:171–180. Seyfarth RM. 1977. A model of social grooming among adult female monkeys. J Theoret Biol 65:671–698. Soini P. 1987. Sociosexual behavior of a freeranging Cebuella pygmaea (Callitrichidae, Platyrrhini) troop during postpartum estrus of its reproductive female. Am J Primatol 13:223–230. Smith AC, Tirado Herrera ER, BuchananSmith HM, Heymann EW. 2001. Multiple breeding females and allo-nursing in a wild group of moustached tamarins (Saguinus mystax). Neotrop Primates 9:67–69. Vogt JL. 1978. The social behavior of a marmoset (Saguinus fuscicollis) group II: behavior patterns and social interaction. Primates 19:287–300.