Behavioral Ecology doi:10.1093/beheco/arq125 Advance Access publication 2 September 2010

Chacma baboon mating markets: competitor suppression mediates the potential for intersexual exchange P.M.R. Clarke,a,b J.E.B. Halliday,b,c L. Barrett,b,d and S.P. Henzib,d Department of Anthropology, University of California, One Shields Avenue, Davis, CA 95616, USA, b Behavioural Ecology and Research Group, School of Psychology, University of KwaZulu-Natal, Durban 40412, South Africa, cThe Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush Veterinary Centre, Roslin, EH25 9RG, UK, and dBehavior and Evolution Research Group, Department of Psychology, University of Lethbridge, Lethbridge, Alberta T1K 3M4, Canada

a

T

he metaphor of the market and the economic principles it invokes has a long and fruitful history in evolutionary biology (Bowles and Hammerstein 2003; Hammerstein and Hagen 2005). One that has culminated in a formal biological market theory of cooperation where cooperative acts are viewed as commodities whose value vary with the supply and demand of trading parties (Noe¨ and Hammerstein 1994, 1995). The canonical example is the so-called ‘‘mating market’’ in which males and females represent distinct trading classes, each seeking to acquire the other’s gamete (Bowles and Hammerstein 2003). Due to a male-biased intersexual asymmetry in a propensity to mate, however, the supply of sperm typically exceeds demand and, as a consequence, females are expected to require supplementary commodities in exchange for their gametes. Accordingly, males from a broad array of species are regularly observed exchanging resources and/or services for mating opportunities with females and doing so in a manner sensitive to fluctuations in the operational sex ratio (OSR) (e.g., Gumert 2007b; Metz et al. 2007; Norscia et al. 2009). Address correspondence to P.M.R. Clarke. E-mail: pclarke@ucdavis .edu. Received 1 December 2009; revised 26 June 2010; accepted 2 July 2010. ! The Author 2010. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: [email protected]

This widespread occurrence of intersexual cooperation in the context of mating is arguably unexpected. On the one hand, one of the central predictions of biological market theory is that cooperative exchange will break down whenever members of the demanding or competing class are able to control/dominate members of the supplying or choosing class and are therefore in a position to appropriate commodities by force (Noe¨ and Hammerstein 1994, 1995). On the other hand, we know that sexual coercion/control of females by males is a common feature of mating systems, often an effective means of increasing mating success, and, along with intermale competition and female choice, can represent a fundamental agent of sexual selection (Smuts BB and Smuts R 1993; Clutton-Brock 1995; Arnqvist and Rowe 2005). Given this, we may reasonably expect cooperative intersexual exchange in the context of mating to be the exception rather than the rule. The fact that it is not implies, therefore, that the ability to control trading partners does not inevitably preclude the need for cooperation. Work on the relationship between cleaner fish and their clients suggests that this may be, in part, because exploited parties often have the option of terminating an interaction, thereby preventing the opportunity for not only further exploitation but also further profitable cooperation (Bshary and Wu¨rth 2001; Bshary and Noe¨ 2003). This idea of an ‘‘exit threat’’ (sensu Johnstone and Bshary 2008) seems germane to the

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Contrary to the expectations of biological market theory, in species where sexual coercion is effective males often exchange resources or services with females for the opportunity to mate. This suggests that an ability to control mating partners does not preclude the need for their cooperation. We argue that this is because, in many systems, female resistance to coercion can precipitate intermale competition and, as such, a male’s mating strategy may often be better served by securing female compliance through affiliative rather than agonistic means. Based on this reasoning, we predicted that the need for males to exchange resources/services for mating access with females will be absent only when intermale power differentials are such that dominant individuals can secure uncontested access to receptive females. Accordingly, data from our long-term study site revealed no support for a biological markets model of intersexual mating exchange in chacma baboons, where the mating monopolies of alpha males are near absolute. Specifically, we found that males groomed females substantially less than they were groomed by them, their propensity to groom was poorly described by measures of female fertility, male rank, and the operational sex ratio, and their mating success was not associated with their grooming effort. We further predicted that an additional consequence of the degree of competitor suppression seen in chacma would be a reversal in the direction of intersexual trade of services for mating. We found that female grooming was positively associated with the probability that they would successfully initiate a copulation. Our study strongly suggests that it is variance in competitor suppression, not partner control, that mediates the potential for and direction of intersexual cooperative exchange. Key words: biological markets, coercion, competitor suppression, intersexual cooperation. [Behav Ecol 21:1211–1220 (2010)]

1212

Takefushi 1993) and possibly stress alleviation (Keverne et al. 1989). As such, it can be considered a service that can be exchanged for itself or other services/resources (Henzi and Barrett 1999). Indeed, numerous studies, many on chacma, have shown that individuals trade grooming reciprocally (e.g., Barrett et al. 1999; Stopka and Graciasova´ 2001), as well as for other commodities, such as the opportunity to handle infants (e.g., Henzi and Barrett 2002; Gumert 2007a) or tolerance at feeding sites (e.g., Barrett et al. 2002; Port et al. 2009). Furthermore, work on long-tailed macaques and mice has shown that males may exchange grooming for the opportunity to mate (Gumert 2007b) or for information on female reproductive status (Stopka and Macdonald 1999). A standard interpretation of a mating-market model of intersexual grooming predicts that male grooming effort will be a function of male rank, the OSR, and the probability that copulation will lead to fertilization (Gumert 2007b). We therefore tested whether models containing these variables could explain patterns of intersexual grooming by males. We placed them in competition with models based on reciprocal trade alone, the simplest variant of which suggests that grooming effort by males is solely a function of the grooming behavior of their partners. If competitor suppression eliminates the potential for market-based effects, then models based on reciprocal trade alone should provide the better fit. Using reciprocal trade as the counter model was appropriate because the reciprocal exchange of grooming between females is well established (Barrett et al. 1999, 2000), for many males mate guarding represents a rare opportunity to be consistently groomed and, conversely, it limits a female’s opportunity to exchange grooming with other females (Rasmussen 1980; Collins 1981). The central prediction in a mating-market is that male grooming effort will be positively associated with mating success. Therefore, in order to identify whether male grooming had any effects on mating success, we also modeled the probability of mating success for both sexes. However, if the incitement of intermale competition is the means by which females circumvent the constraints of coercion, then male grooming effort may reflect attempts to secure female compliance and maximize the longevity of guarding bouts rather than increase the chance of individual mating. Indeed, the fact that all mating occurs within the context of mate guarding suggests that if there is any intersexual exchange in chacma, guarding duration should be the most salient variable (Clarke, Henzi, and Barrett 2009). Therefore, we also examined the relative contribution of each sex in maintaining the integrity of guarding bouts, using models containing the variables invoked by the mating market framework. If the nature of intermale relations is such that, in the context of mating, female compliance is of no tradable value, then contrary to the expectations of a classic mating market, we may expect the balance of trade to fall in favor of males. Specifically, we may predict that it will be females who exchange grooming for any services/resources males may have to offer. To investigate this, we also examined a number of models of female grooming and copulation. We have shown that in response to the threat of infanticide by recently immigrated males, female chacma have been selected to predispose multiple males to protect their infants from attack (Clarke, Henzi, and Barrett 2009). Consequently, they favor very high mating rates throughout their receptive period that, importantly, are far higher than those at which males initiate mating (Clarke, Henzi, and Barrett 2009). If selection tailors male mating rates and sperm usage to the probability of fertilization (Wedell et al. 2002), as seems to be the case in chacma (Clarke, Henzi, and Barrett 2009), then we may reasonably expect females to groom males as a means to reduce their

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

maintenance of cooperative exchange in the context of mating, given the known interplay between female choice and intermale competition and the phenomenon of indirect mate choice (Wiley and Poston 1996; Wong and Candolin 2005; Hunt et al. 2009). Data from numerous species indicate that while females may be unable to overcome the constraints of coercion directly, they are often able to gain access to alternative males by inciting intermale competition through signaling, using, for example, calls or actual physical resistance (e.g., Cox and LeBoeuf 1977; Koprowski 1993; Pizzari 2001). Invoking indirect mate choice as the mechanism by which intersexual cooperation is maintained in the face of coercion implies that exchange will be a negative function of the extent to which a single male is able to suppress/limit the mating behavior of his rivals. At one extreme, therefore, we expect intersexual exchange, despite sexual coercion, in species where the power differentials between males are relatively small. This is because intermale competition will be intense and frequently lead to the usurpation of guarding males and female resistance will, therefore, be effective in promoting challenges by rival males (e.g., Moore AJ and Moore PJ 1999; Moore et al. 2001). Support for this line of reasoning comes, in part, from the nature of intersexual exchange seen in species like chimpanzees, where males commonly provide services and/or foodstuffs to females with whom they mate and where the frequency with which they mate is correlated with the degree of provision (Hemelrijk et al. 1992; Hockings et al. 2007; Gomes and Boesch 2009). At the other, where power differentials between males are large so that rank is the sole determinant of mating success/ access, inter-sexual exchange related to mating is expected to be largely absent. This is because female resistance/choice will have little bearing on interactions between males and mating outcomes (e.g., Reichard et al. 2005) and, concomitantly, the need for males to secure female compliance will be absent (see also Johnstone and Bshary 2002, 2008). A complete proof of the hypothesized relationship between intermale competition and intersexual exchange requires, therefore, evidence of an absence of exchange in species where power differentials between males are substantial. With this in mind, we undertake an examination of intersexual exchange in wild chacma baboons (Papio hamadryas ursinus). Chacma are ideally suited to such an analysis for 3 reasons. First, they exhibit one of the strongest correlations between male rank and reproductive success seen in the primates and are the most despotic of all the savannah baboon subspecies (Bulger 1993; Weingrill et al. 2000, 2003). This is largely due to the absence of any alternative male strategies, such as the coalitions formed by subordinates in the East African subspecies (e.g., Alberts et al. 2003), so that mating access is determined, in the main, by physical condition and attendant fighting prowess. Second, males are nearly twice the size of females and, consequently, male control of females is assumed to be profound (Kitchen et al. 2009). This is well evidenced by the fact that all mating occurs within the context of mate guarding, bouts of which can last up to 6–7 days (Clarke, Henzi, and Barrett 2009). Finally, after a given period of residency, male chacma, like in other subspecies (Gesquiere et al. 2007), are able to distinguish between conceptive and nonconceptive cycles (Weingrill et al. 2003). As conceptive cycles rarely overlap, this means that dominant individuals are able to monopolize access to females whose gametes are actually of value with near total impunity (Weingrill et al. 2003; Henzi et al. 2010). We use an analysis of intersexual grooming to investigate the propensity of male chacma to exchange commodities for mating opportunities/access. Grooming provides direct benefits to recipients in the form of ectoparasite removal (Tanaka and

Behavioral Ecology

Clarke et al.

•

Chacma baboon mating markets

1213

reluctance to mate. Alternatively, it may simply be the case that females exhibit a preference for dominant males alone because they are best placed to provide protection from other potentially infanticidal males (van Schaik et al. 2004). As with the analysis of male grooming, we place these models in competition with one predicting that intersexual grooming reflects reciprocal exchange. METHODS Chacma study subjects and data collection

Response variables Grooming Traditionally, market-based studies of grooming have considered only immediately reciprocated grooming and, in doing so, have discarded all unreciprocated bouts (e.g., Barrett et al. 1999; Chancellor and Isbell 2009). While this may be suitable for investigating the issue of reciprocal exchange (i.e., like for like), it may not give a clear representation if different commodities (e.g., mating vs. grooming) are being traded. With this in mind, we modeled patterns of grooming using 3 different measures: Binary occurrence of grooming. Grooming by either sex was only seen in 49.51% of all samples, meaning that over 50% of the total data set contained zeros. This excess of zeros, known as zero inflation, meant that the variable could not be normalized. Therefore, we binary coded all samples, for each sex, as to whether they had or had not been seen grooming. This enabled us to benefit from the analytical power associated with the complete data set and, empirically, allowed covariates of when an individual did and did not groom to be identified. Mean time spent grooming (seconds) per focal sample. To model grooming duration, we restricted the data set to only those focal samples where at least one member of a given dyad had been observed grooming. For these samples, we then took the mean amount of time an individual was observed grooming (log(x 1 1)) and, in doing so, included bouts that commenced prior to the sample and those that finished after it; although, we only considered grooming that occurred within the confines of the sample. Mean contribution (seconds) per reciprocated bout. Following others, we also assessed the amount each sex contributed during immediately reciprocated grooming bouts. We defined immediate reciprocation as that occurring within 30 s of the end of a previously received episode. We considered 30 s to be the

Table 1 Demographic breakdown (! x (min-max)) of the 2 study troops over the study period (04/2003 to 10/2004)

Male number Female number Estrus female number Sex ratio OSR

Troop 1

Troop 2

6.31 10.32 3.87 1.68 0.63

5.71 11.61 4.46 2.06 0.79

(3–9) (9–11) (1–8) (1.11–3.67) (0.14–1.67)

(5–7) (10–13) (2–8) (1.57–2.60) (0.29–1.33)

Proximity and compliance Female compliance during mate guarding was quantified using the frequency with which females approached and withdrew from their guarding partner. Approaches and withdrawals were defined as any movement into or out of a 3 m radius of the focal animal. We used the proportion of all movements constituting withdrawals as the measure of compliance. Mating success We considered 2 measures of mating success: 1) successful copulations initiated by males and 2) successful copulations initiated by females. A successful copulation was defined as one that resulted in a female withdrawal and/or a copulation call. A copulation attempt by a male was characterized by mounting and attempted intromission, whereas for females, it entailed the clear and deliberate orientating of the anogenital region toward a male. Both were easily and unambiguously scored (see also Clarke, Henzi, and Barrett 2009). When considering the entire data set, all measures of mating success exhibited pronounced zero inflation. Consequently, we modeled both as binary variables, with samples being scored as either one where a successful copulation was observed or one where it was not. Explanatory variables We used a suite of explanatory variables aimed at capturing the effects of variance in female fertility and intrasexual competition for mates. Female fertility. Two measures of the immediate fertilizability of females were used: ovulation probability (OP) and cycle conception status (CC). The probability of ovulation was quantified in terms of the number of days prior to the onset of swelling detumescence (D-0) (see also Hausfater 1975; Bulger 1993; Weingrill et al. 2000, 2003). Following Clarke et al. (Clarke, Barrett, and Henzi 2009; Clarke, Henzi, and Barrett 2009), it was treated as a 3-level categorical variable: Phase 1: D-1 to D-4; Phase 2: D-5 to D-9; and Phase 3: D-10 to D-14. Cycle conception status was treated as a 2-level categorical variable: 1) conceptive (ultimate cycle before pregnancy); 2) nonconceptive (all remaining cycles). Intrasexual competition. We considered 2 measures: male rank (MR) and the OSR. Male hierarchies were quantified using the frequency of approaches and supplants and were linear with minimal transitivity. To minimize the effects of localized reversals, particularly among the lower ranks, adjacent ranks were collapsed into each other to give a 3-level factor: high rank, mid rank, and low rank (see also Clarke, Barrett, and Henzi 2009; Clarke, Henzi, and Barrett 2009). The OSR was treated as a continuous variable and defined as the ratio of adult nonnatal males to sexually mature (i.e., non-nulliparous) swollen females in the D-14 to D-1(see above) phase of swelling. This allowed us to assess the effects of both intermale and interfemale competition for mates. Candidate model set Rather than using stepwise multiple regression and P value based null hypothesis testing, we compared different models using likelihood-based information criteria (Burnham and Anderson 1998; Johnson and Omland 2004). We formulated a set of candidate models a priori based on the expectations

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Data come from 2 troops of wild chacma baboon found at the De Hoop Nature Reserve, South Africa (for details, see Barrett et al. 2004) and were collected over an 18-month period (04/2003 to 10/2004) using 20 min continuous focal animal sampling (Altmann 1974). Seventeen nonnatal adult males and 21 adult (non-nulliparous) females (see Table 1) were sampled over the course of 73 separate menstrual cycles.

appropriate cutoff because a frequency plot of times between episodes revealed this to be a natural point of inflection.

Behavioral Ecology

1214

of the hypotheses outlined above and an understanding of the chacma baboon mating system. We define and justify these models below, using x, u, h, and g to denote models of male grooming, female grooming, female compliance, and mating success, respectively, and the acronyms specified above to denote the explanatory variables. When defining models, 3 indicates interactions between terms, as well as the associated main effects, whereas 1 indicates main effects alone.

Models of female grooming Females generally favor a higher mating rate than do males, whose mating effort is tailored both to their rank and the probability of ovulation (Clarke, Henzi, and Barrett 2009). If, therefore, females use grooming as a means to encourage males to mate, then we may expect their grooming effort to be sensitive to these variables. A possibility denoted by the models 1) uMR 1 OP and 2) uMR 3 OP. To accommodate the possibility that females exhibit a preference based solely on male rank, we included 3) uMR. Variance in the OSR during the study occasionally meant that there were more sexually receptive females than males present. It is conceivable, therefore, that females had to compete for male attention. We considered this with 4) uOSR, 5) uMR 1 OP 1 OSR, 6) uOSR 3 MR 1 OSR 3 OP, and 7) uOSR 3 MR 3 OP. As with the modeling of male grooming, the simplest model of reciprocal trade was 8) uMG (MG, male grooming). To accommodate the possibility that females may exchange grooming both for itself and sexual access, we also included 9) uMG 1 MR 1 OP, 10) uMG 1 MR 3 OP, 11) uMG 1 MR, 12) uMG 1 OSR, 13) uMG 1 MR 1 OP 1 OSR, 14) uMG 1 OSR 3 MR 1 OSR 3 OP, and 15) uMG 1 OSR 3 MR 3 OP. Models of female compliance If males groom females in an effort to increase their compliance during guarding, then a biological markets perspective predicts that the propensity of females to maintain proximity to their guarding partners will be a function of male grooming effort. The simplest model, therefore, was 1) hMG. Given the emphasis placed on male rank, ovulation probability, and

Models of mating success To increase the probability that model selection reflected the genuine contribution of grooming effort, as opposed to an association with unconsidered effects, we assessed the contribution of grooming relative to previously identified best-fit models. Previously (Clarke, Henzi, and Barrett 2009), we have demonstrated that the probability of successful copulation initiation by males, g#, is well characterized by a model including the main effects of ovulation probability and male rank alone, 1) g#MR 1 OP. Therefore, we examined the relationship between intersexual grooming and successful copulation initiation by males by comparing this model with 2) g#MR 1 OP 1 MG and 3) g#MG. The probability of a successful copulation attempt by females, g$, is well described by a model containing only the main effects of ovulation probability and cycle conception status, 1) g$OP 1 CC (Clarke, Henzi, and Barrett 2009). Therefore, we examined the effects of grooming on female mating success with 2) g$OP 1 CC 1 FG and 3) g$FG. Data analysis Model selection Model selection was based on the Akaike information criterion (Akaike 1974), with small sample size correction (AICc). AICc provides a measure of a model’s fit scaled to its complexity, with smaller values indicating better fit and allows models to be ranked according to their ability to account for the data. In addition, we calculated Akaike weights, wi, which provide a measure of the probability that a model is the best fit, within the context of the candidate model set (Burnham and Anderson 1998). We provide details of how AICc and Akaike weights are derived in the Supplementary Material (see also Burnham and Anderson 1998). We assessed effect size and estimate precision on the basis of estimates taken from all models, using model averaging (Burnham and Anderson 1998). Following Burnham and Anderson (1998), we present only models that collectively accounted for 95% of the available model weight in the main text, but provide full model comparisons in the Supplementary Material. Variance components We assessed, using model comparison, the variance attributable to male identity, female identity, troop identity, and sampling year, by including them in models as random intercepts. We also assessed the effects of time of day when the sample was collected as baboons tend to groom disproportionately more during early morning and late evening. Although neither troop identity nor sampling year accounted for any variance in any of the responses, both male and female identity did for all but one. Time of day also accounted for substantial variance in the binary occurrence of male grooming but not female grooming.

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Models of male grooming A complete biological markets model of intersexual grooming predicts that, in the context of mate guarding, a male’s grooming effort will be a function of his rank, the ratio of receptive females to competing males, and the probability that mating will lead to fertilization (Gumert 2007b). We considered this possibility with the main-effects model 1) xMR 1 OSR 1 OP, as well as the more complex 2) xMR 3 OSR 1 MR 3 OP 1 OSR 3 OP. It is, however, known that male chacma are able to discriminate between conceptive and nonconceptive cycles (Weingrill et al. 2003). Therefore, if male grooming effort is sensitive to the immediate fertility of females it may also be expected to vary with cycle conception status, a possibility denoted by the models 3) xMR 1 OSR 1 OP 1 CC and 4) xMR 3 OSR 1 MR 3 OP 1 OSR 3 OP 1 MR 3 CC 1 OSR 3 CC 1 OP 3 CC. The simplest model of reciprocal trade we examined was 5) xFG (FG, female grooming), which suggests that grooming effort by males is solely a function of the grooming behavior of their partners. It seems feasible to suggest, however, that grooming by males may be simultaneously sensitive to the potential for the reciprocal exchange of grooming and the interchange exchange of grooming for sexual access and/or compliance. Therefore, we also considered all previously defined models with the addition of female grooming as a main effect: 6) xFG 1 MR 1 OSR 1 OP, 7) xFG 1 MR 3 OSR 1 MR 3 OP 1 OSR 3 OP, 8) xMR 1 OSR 1 OP 1 CC, and 9) xMR 3 OSR 1 MR 3 OP 1 OSR 3 OP 1 MR 3 CC 1 OSR 1 OP 3 CC.

OSR by the market model, we also included 2) hMG 1 MR 1 OSR 1 OP and 3) hMG 3 MR 1 MG 3 OSR 1 MG 3 OP. Again, in light of the ability of males to discriminate between conceptive and nonconceptive cycles (Weingrill et al. 2003), we also considered 4) hMG 1 MR 1 OP 1 CC 1 OSR and 5) hMG 3 MR 1 MG 3 OSR 1 MG 3 OP 1 MG 3 CC. We also accommodated the possibility that the female strategy is shaped by a preference for high rank and/or the potential for interfemale competition for mates with the models 7) hMR, 8) hOSR, and 9) hMR+OSR. We placed these models in competition with one that predicts that female proximity maintenance was simply the reciprocal of their guarding partner’s tendency to maintain proximity, denoted by 6) hMC (MC, male compliance).

Clarke et al.

•

Chacma baboon mating markets

1215

RESULTS Intermale relations and mate guarding The mean distance between a guarded female and a nonguarding male was 21.599 (60.966) m, whereas the distance between a guarded female and their guarding partner was 2.556 (60.133) m. A guarding pair was approached by nonguarding male at a rate of 0.107 (60.072) times per hour, equivalent to one approach approximately every 9 h. No approaches resulted in the usurpation of the incumbent guarding male and in only 3 cases did they culminate in aggression between the guarding male and the approaching male. As a consequence, a guarding male was subject to intermale aggression at a rate of 0.029 (60.017) times per hour, equivalent to one contest every 34 h. Mate guarding and intersexual grooming

Female grooming when mate guarded

The amount of time males spent grooming females increased when mate guarding compared with when not. So too, however, did the amount of time they spent being groomed by females. Consequently, they received nearly twice as much grooming from their partners as they gave (see Figure 1). Per hour, males requested grooming less frequently than females and received more unrequested grooming from females than they gave in return (Male grooming. Requested: 0.346 6 0.04; unrequested: 0.370 6 0.44. female grooming. Requested: 0.404 6 0.046; unrequested: 0.259 6 0.037).

Table 4 shows the top models for the 3 measures of female grooming. Model selection revealed that a model containing only the binary occurrence of male grooming, with over 50% of the model weight, provided the best account of the binary occurrence of female grooming. Model averaged parameter estimates (Table 5) indicate that the occurrence of male grooming had a strong positive effect on the occurrence of female grooming. In addition to the occurrence of male grooming, the second- and third-ranked models also contained the main effects of male rank and OSR, respectively (Table 4). However, comparing the likelihood of these models with that of the top model, indicates that, beyond the provision of additional parameters, these variables provided little improvement in fit. Parameter estimates echo this and indicate that they had negligible effect on the occurrence of female grooming (Table 5). Modeling the duration of female grooming using all bouts revealed that a model containing only male rank was ranked top (Table 4). A model containing only the mean was ranked second, whereas one containing the duration of male grooming and male rank was ranked third. The relatively small difference in weights between these models indicates that none of them can be definitively considered the best fit. Model averaged parameter estimates suggest that females may have altered their grooming effort in relation to male rank; although confidence intervals for the variable caution against putting too much weight in such a conclusion

Male grooming when mate guarding Table 2 shows the top models of the 3 selected measures of male grooming. For the binary occurrence of male grooming, a model containing only the binary occurrence of grooming by females was the best fit, with 98% of the model weight. The extremely high weight of this model indicates that it can be considered the best fit, within the context of the candidate model set, with reasonable certainty. Model averaged parameter estimates indicate that males were more likely to groom females when females had groomed them (Table 3). In contrast, male rank, ovulation probability, OSR, and cycle conception status had little or no effect on the occurrence of male grooming.

Table 2 Parameter number, K, log-likelihood, log L, AICc and Akaike weights, wi, for models of the 3 measures of grooming by males Response Variable

Model

K

log L

AICc

wi

Binary occurrencea Durationb (all bouts)

xFG xb xFG xb xFG

5 4 5 2 3

2160.018 2354.947 2356.915 2105.620 2104.754

330.233 722.216 724.319 215.335 215.699

0.981 0.726 0.254 0.539 0.450

Durationc (reciprocated) a

b

Figure 1 Time, in seconds, each sex spent grooming the other outside (dark gray) and during mate guarding (light gray).

c

Link function: logit. Random effect structure: male ID 1 female ID 1 time of day. Link function: identity. Random effect structure: male ID 1 female ID. Link function: identity. Fixed effects only.

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

When modeling, the duration of all grooming bouts performed by males, model selection revealed that the mean, xb, provided the best fit (Table 2). Appreciable support was, however, also found for a model containing only the duration of grooming received from females, xFG. Although the secondranked model received fairly substantial support, model averaged parameter estimates indicate that, as with the variables invoked by the market model, the duration of female grooming had little to no effect on the duration of grooming performed by males (Table 3). Models were ranked in the same order when considering only immediately reciprocated bouts (Table 2). Greater weight was, however, assigned to the second-ranked model, xFG, at the expense of the best fit, xb. This was to be expected, given that the appearance, statistically speaking, of reciprocity may be inflated by considering only bouts where both sexes groomed. Nevertheless, as with male rank, ovulation probability, OSR, and cycle conception status, female grooming effort had only a very weak effect (Table 3).

All analyses were carried out using the ‘lme4’ package (Bates et al. 2008) in R (R Development Core Team 2009).

Behavioral Ecology

1216

Table 3 Model averaged parameter estimates (695% unconditional confidence intervals) from all models of the 3 measures of male grooming Response Binary

Intercept Female grooming Male rank High versus mid High versus low Mid versus low Ovulation probability High versus mid High versus low Mid versus low OSR Cycle conception status Conceptive versus nonconceptive

22.000 (60.940) 2.765 (60.664)

3.189 (60.883) 0.004 (60.064)

3.965 (61.173) 0.069 (60.223)

0.001 (60.237) 0.013 (60.311) 0.012 (60.236)

0.009 (60.265) 0.026 (60.405) 0.016 (60.262)

0.001 (60.221) 0.011 (60.212) 0.009 (61.955)

20.005 20.006 0.000 0.006

Duration (all bouts)

(60.124) (60.186) (60.164) (60.312)

20.001 (60.071)

20.006 0.000 0.006 0.014

(60.127) (60.180) (60.199) (60.352)

Duration (reciprocated)

20.001 0.000 0.008 20.016

20.005 (60.132)

(60.050) (60.061) (60.094) (60.457)

0.003 (60.082)

Only parameter estimates for main effects are given as all models containing interaction terms had model weights ,0.0001.

(Table 5). All other variables appeared to have had only a negligible effect on the duration of female grooming. Modeling the duration of only immediately reciprocated female grooming bouts largely mirrored the modeling of all female grooming bouts, except that the model containing only the duration of male grooming carried more weight (Table 4). Model averaged parameter estimates indicate, however, the effect of male grooming was limited, as was the case with all selected explanatory variables, including male rank (Table 5). Female compliance in the market place Guarded females withdrew from their guarding partners at a rate of 6.139 (60.333) times per hour and approached them at a rate of 3.178 (60.261) times per hour. On average, therefore, withdrawal represented 69.16% of all female movements. Variance in female compliance had no clear effect on the Table 4 Parameter number, K, log-likelihood, log L, AICc and Akaike weights, wi, for models of the 3 measures of grooming by females

a b c

Response variable

Model

Binary occurrencea

uMG 3 uMG1MR 5 uMG1OSR 4

2173.290 352.658 0.508 2172.173 354.346 0.198 2173.239 354.609 0.191

Durationb (all bouts)

uMR ub uMG1MR uMG uOSR uMR1OP uMG1OSR

6 4 7 5 5 8 6

2388.794 2391.601 2388.381 2391.385 2391.586 2388.643 2391.379

790.107 791.446 791.457 793.139 793.539 794.185 795.276

0.360 0.184 0.183 0.079 0.065 0.045 0.027

Durationc (reciprocated) ub uMR uMG uMG1MR uOSR uMG1OSR uMR1OP

4 6 5 7 5 6 8

2106.237 2104.221 2105.397 2103.484 2106.139 2105.334 2104.157

220.719 220.960 221.161 221.664 222.645 223.186 225.214

0.235 0.208 0.188 0.146 0.090 0.068 0.025

K

log L

AICc

wi

Link function: logit. Random effect structure: male ID 1 female ID. Link function: identity. Random effect structure: male ID 1 female ID. Link function: identity. Random effect structure: male ID 1 female ID.

frequency with which nonguarding males approached guarding pairs (Figure 2a) nor did it appear to influence proximity between guarded females and nonguarding males (Figure 2b). During mate guarding, rates (per hour) of female withdrawal were lower in the 5 min proceeding the occurrence of intermale aggression than during other periods (Outside of intermale aggression: 6.154 6 0.337; Prior to intermale aggression: 5.00 6 1.225), suggesting that variance in female compliance had little influence on patterns of intermale competition. Modeling of female compliance revealed a model containing only the relative movements of male-guarding partners to be the best fit, carrying ’ 100% of the model weight (hMC: logL ¼ 2296.889, AICc ¼ 604.0, wi ’ 1). This model predicts that variance in the propensity of females to maintain proximity to their guarding partners was simply a function of the extent to which their partners move toward or away from them (Male Approach: b!ˆ 6 95%: 21.631 6 0.018. Male Withdrawal: bˆ! 6 95%: 1.033 6 0.015). Figure 3 shows the relationship between male and female compliance and indicates that the more a guarding male tended to move away from their partner the more their partner tended to move toward them and vice versa. This, together with the fact that models containing male rank, ovulation probability, male grooming, and OSR performed so poorly (see Supplementary Material), suggests that female movements were aimed simply at maintaining a constant distance between themselves and their guarding partner. Intersexual grooming and mating success Table 6 shows AICc values and weights for models of the binary occurrence of male and female mating success. When considering successful copulations initiated by males, the previously identified best fit, g#MR 1 OP, was ranked top. Male grooming appeared in the second-ranked model, carrying nearly 25% of the model weight. However, when comparing the likelihood of this model to that of the top model it can be seen that the addition of male grooming provided only a marginal improvement in fit. This, together with the fact that a model containing only male grooming fared worse than a model containing only the mean, indicates that male grooming had little appreciable effect on the probability that a male would successfully initiate a copulation. This was further confirmed by model averaged parameter estimates for b ! : 0:01760:294Þ: the effects of male grooming (bˆ!6se

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Explanatory

Clarke et al.

•

Chacma baboon mating markets

1217

Table 5 Model averaged parameter estimates (695% unconditional confidence intervals) from models of the 3 measures of female grooming Response Binary

Duration (all bouts)

Intercept Male grooming Male rank High versus mid High versus low Mid versus low Ovulation probability High versus mid High versus low Mid versus low OSR

21.597 (60.684) 2.343 (60.582)

3.905 (60.959) 20.026 (60.122)

4.893 (61.095) 0.066 (60.219)

0.038 (60.341) 20.001 (60.246) 20.036 (60.369)

0.718 (61.300) 0.069 (60.870) 20.649 (61.289)

20.274 (60.812) 20.065 (60.480) 20.072 (60.197)

0.002 20.162 20.023 0.086

0.006 0.018 20.081 0.019

0.003 20.004 0.025 20.050

(60.194) (60.808) (60.373) (61.057)

(60.284) (60.732) (60.994) (60.839)

Duration (reciprocated)

(60.149) (60.253) (60.292) (60.816)

Only parameter estimates for main effects are given as all models containing interaction terms had model weights ,0.01

When considering successful copulations initiated by females, a model containing female grooming was ranked top, with nearly 80% of the model weight. Indeed, collectively, models containing female grooming carried over 95% of the available weight, suggesting that it played an important role in the realization of female mating success. The model averaged parameter estimate confirms this, indicating that the occurrence of female grooming had a strong and consistently positive effect on the likelihood that a female would successfully b ! 0.785 6 0.343). initiate copulation (bˆ! 6 se:

DISCUSSION

Our data indicate that patterns of grooming between the sexes in chacma baboons do not conform to the expectations of a classic mating market model. We show that 1) males groomed females for approximately half as much time as they were groomed by them, 2) they received more unrequested grooming from females than they provided, 3) male grooming had no effect on mating success nor female proximity maintenance, whereas 4) female grooming appeared to promote mating and, finally, 5) models including measures of male rank, the immediate fertility of females and the ratio of males to sexual receptive females all provided a poor account of both the occurrence and duration of intersexual grooming

by males. Concurrent with these findings, our data also reveal that levels of intermale competition over mating access were minimal. Specifically, we found that guarding males were rarely approached by rival males and were even less frequently involved in aggression with them. Furthermore, the behavior of females when guarded had no bearing on these measures of intermale interaction and, indeed, seemed aimed more at maintaining the integrity of guarding bouts than disrupting them. These findings contrast with those from species, such as chimpanzees and macaques, where the strictures of rank are not as severe as in chacma, but the potential for coercion is comparable. In these species, males are the primary groomers, the degree to which they groom is a function of both their rank and the OSR and their grooming is positively associated with mating success (e.g., Gumert 2007b; Gomes and Boesch 2009; Norscia et al. 2009). Taken together with these findings, our data support the prediction that it is competitor suppression, not partner control that mediates the potential for cooperative exchange. Further support for this conclusion comes from the patterns of intersexual grooming in anubis and yellow baboons. As in chacma, males in these subspecies are demonstrably capable of physically controlling females (Hausfater 1975; Rasmussen 1980; Collins 1981; Kitchen et al. 2009). In contrast to chacma, however, the formation

Figure 2 Relationship between female compliance and (a) the frequency, per hour, with which nonguarding males approached a guarding pair and (b) the distance, in meters, between a guarded female and nearest nonguarding male.

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Explanatory

Behavioral Ecology

1218

of rank leveling coalitions by subordinates mean that alpha males are frequently subject to challenges from rivals when mate guarding and, consequently, often unable to secure uncontested mating monopolies (Alberts et al. 2006). Accordingly, data (Table 7) from these subspecies indicate not only that the amount of grooming given to females by males increases sharply during consortship but also that consorting males are responsible for most of the observed grooming. Furthermore, the few studies that reported the effects of ovulation probability found a positive association with male grooming effort. Although much remains unknown, these findings point toward a pattern consistent with the expectations of a biological market model and, in turn, our hypothesis that it is competitor suppression not partner control that ultimately mediates the potential for exchange. The persistence of trade between the sexes where effective coercion is possible, but its absence where competitors can be suppressed, suggests that the latter more effectively undermines the market. This is expected because the commodity of interest—the opportunity to fertilize eggs—is finite and nonrenewable and, as such, members of the bidding class (i.e., males) are involved in a zero-sum game, where an increase in one male’s mating access inevitably equates to a decrease in another’s. Given this, it can be seen that the use of sexual coercion by one male not only negatively affects females but also rival males because it increases mating access for the coercer at the expense of the rest of the male cohort. This is why cooperative intersexual exchange is seen in species where males are coercive, but power differentials between males are small because it is the readiness and ability of males Table 6 AICc and Akaike weights, wi, for models of mating success Response variable Male initiated

Model g#MR 1 g#MR 1 g#b g#MG

Female initiated g$OP g$FG g$OP g$b

K log L OP OP 1 MG

1 CC 1 FG 1 CC

AICc

wi 0.518 0.251 0.165 0.066

7 8 3 4

2156.162 2155.829 2161.452 2161.340

326.695 328.139 328.982 330.812

7 4 6 3

2159.759 2164.555 2163.990 2173.229

333.890 0.814 337.242 0.152 340.257 0.034 352.536 ,0.001

All models run using logit link function and random effect structure: male ID 1 female ID.

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Figure 3 Relationship between male and female compliance.

to usurp each other and, in turn, undermine the coercive behavior of their rivals that gives currency to female compliance. The extent to which the exchange of services by males for female compliance will be sensitive to market effects is difficult to say. However, if the rate of intermale competition over mating access is a function of the OSR, then so should the ability of females to incite intermale competition. Therefore, the value of female compliance as a tradable commodity and the attendant need for cooperative exchange may vary with market pressures. This need not be a linear function of OSR because beyond a certain intensity of intermale competition coercion will become a more effective tactic, in the short term at least (e.g., Hausfater 1975; Hogg 1984). Ultimately, the reason that competitor suppression eliminates the need for trade by males is that it renders female resistance to monopolization pointless—resistance is only worthwhile if it elevates the probability that a rival male will challenge the incumbent. In light of this, competitor suppression may have consequences that go beyond diminishing cooperative mating–orientated exchange by males. Most notably, if resistance is costly for females and has no bearing on the stability and attendant longevity of guarding bouts, then selection should favor females who simply acquiesce to monopolization (Ha¨rdling et al. 1999); just as promiscuity may be observed, despite a female preference for monandry, simply because it is more convenient for females to acquiesce rather than resist persistent male advances (‘‘convenience polyandry’’: Thornhill and Alcock 1983). This would, then, account for 1) our finding that, despite tangible conflict between the reproductive strategies of males and females (Pradhan and van Schaik 2008; Clarke, Henzi, and Barrett 2009), guarded females are generally compliant regardless of male rank, ovulation probability, and OSR and 2) the fact that males rarely, if ever, direct aggression toward their partners in the subspecies (Seyfarth 1978; Clarke 2006; Kitchen et al. 2009). The fact that females were the primary groomers and that their grooming effort was positively associated with the probability that they would successfully initiate a copulation, confirms our prediction that an additional consequence of the nature of intermale competition in chacma is that the direction of intersexual trade of grooming for mating is reversed. Given how common promiscuity and male prudence in sperm allocation are across the animal kingdom (Jennions and Petrie 2000; Wedell et al. 2002; Clarke, Pradhan, and van Schaik 2009), this corollary of competitor suppression is unlikely to be unique to chacma. Having said this, although female promiscuity is clearly a necessary condition for a reversal in intersexual trade, conservatism in sperm allocation may not be. It could simply be the case that in situations where females have no influence over intermale competition and, in turn, the identity of their guarding partner, males may refrain from mating so as to force females to exchange. Here, then, demand is created not by asymmetries in the number of competing and choosing parties but by virtue of the degree of control males can exercise over females. Therefore, in addition to circumventing the need for cooperative exchange, high levels of competitor suppression could permit males to create market effects that provide them with services that could not have been extracted by force. In sum, by showing that the nature of mating exchange can only be understood by considering the interdependency between partner control and competitor suppression, our study lends additional weight to the growing realization that the individual components of sexual selection—intrasexual competition, intersexual choice, and intersexual coercion— cannot be considered in isolation (Wong and Candolin 2005; Hunt et al. 2009). Rather, the dynamic interplay

Clarke et al.

•

Chacma baboon mating markets

1219

Table 7 Patterns of intersexual grooming by male East African baboons in relation to the biological market predictions that 1) males should be the primary groomers when mate guarding, 2) male grooming should be negatively associated with male rank, and 3) positively associated with the probability of ovulation

b c

Site

Papio hamadryas anubis P. h. anubis P. h. anubis Papio hamadryas cynocephalus P. h. cynocephalus P. h. cynocephalus P. h. cynocephalus

Captive Gilgil Gombe Mongiro Amboseli Mikumi Ruaha

Male primary groomera U(34) U

Ub

U U U(35) U

Male rank

Ovulation probability

OSR

References

? ? ? ? ? ? ?

? ? Uc ? U U U

? ? ? ? ? ? ?

Rowell (1968) Smuts (1985) Ransom (1981); Domb (2000) Rowell (1966, 1974) Hausfater (1975) Rasmussen (1980) Collins (1981)

U Prediction satisfied; ?, Prediction not tested. Figures in parentheses indicate, where known, the extent to which males groomed more than females. Fully swollen females never observed grooming consort partner (Ransom 1981). Male grooming effort positively associated with swelling size/length (Domb 2000).

between the 3 agents must be incorporated into accounts of sexually selected traits if a complete understanding is to be obtained. By applying this approach to the study of intersexual exchange, we also highlight how the operation and integrity of a biological market is a function not only of partner preference, partner control, and bidding competition but also the extent to which these feedback on each other. This will be true of more than just mating markets and should apply to all situations where the commodity of interest may reasonably be considered finite. SUPPLEMENTARY MATERIAL Supplementary material can be found at http://www.beheco .oxfordjournals.org/. FUNDING University of Bolton bursary, Wingate Foundation Scholarship (UK) and a post-doctoral fellowship from the department of anthropology, University of California, Davis to P.M.R.C; Leverhulme Research Fellowship (UK) to L.B; National Research Foundation (South Africa) to S.P.H. We thank Cape Nature for permission to work at De Hoop. We are also grateful to Richard McElreath and Mark Grote for advice and helpful discussions on mixed modeling and model comparison.

REFERENCES Akaike H. 1974. A new look at the statistical model identification. IEEE Trans Automat Contr. 19:716–723. Alberts SC, Buchan JC, Altmann J. 2006. Sexual selection in wild baboons: from mating opportunities to paternity success. Anim Behav. 72:1177–1196. Alberts SC, Watts HE, Altmann J. 2003. Queuing and queue-jumping: long-term patterns of reproductive skew in male savannah baboons, Papio cynocephalus. Anim Behav. 65:821–840. Altmann J. 1974. Observational study of behaviour; sampling methods. Behaviour. 49:227–267. Arnqvist G, Rowe L. 2005. Sexual conflict. Princeton (NJ): Princeton University Press. Barrett L, Gaynor D, Henzi SP. 2002. A dynamic interaction between aggression and grooming reciprocity among female chacma baboons. Anim Behav. 63:1047–1053. Barrett L, Gaynor D, Rendall D, Mitchell D, Henzi SP. 2004. Habitual cave use and thermoregulation in chacma baboons (Papio hamadryas ursinus). J Hum Evol. 46:215–222.

Barrett L, Henzi S, Weingrill T, Lycett J, Hill R. 1999. Market forces predict grooming reciprocity in female baboons. Proc R Soc Lond Biol Sci. B266:665–670. Barrett L, Henzi S, Weingrill T, Lycett J, Hill R. 2000. Female baboons do not raise the stakes but they give as good as they get. Anim Behav. 59:763–770. Bates D, Maechler M, Dai B. 2008. lme4: linear mixed-effects models using S4 classes. R package version 0 999375-28. Bowles S, Hammerstein P. 2003. Does market theory apply to biology? In: Hammerstein P, editor. Genetic and cultural evolution of cooperation. Cambridge (MA): MIT Press. p. 153–165. Bshary R, Noe¨ R. 2003. Biological markets: the ubiquitous influence of partner choice on cooperation. In: Hammerstein P, editor. Genetic and cultural evolution of cooperation. Cambridge (MA): MIT Press. p. 167–184. Bshary R, Wu¨rth M. 2001. Cleaner fish labroides dimidiatus manipulate client reef fish by providing tactile stimulation. Proc R Soc Biol Sci. B268:1495–1501. Bulger J. 1993. Dominance rank and access to estrous females in male savanna baboons. Behaviour. 127:67–103. Burnham KP, Anderson DR. 1998. Model selection and inference: a practical information-theoretic approach. New York: Springer. Chancellor RL, Isbell LA. 2009. Female grooming markets in a population of gray-cheeked mangabeys (Lophocebus albigena). Behav Ecol. 20:79–86. Clarke PMR. 2006. Inter-sexual mating conflict in the chacma baboon (Papio hamadryas ursinus) [PhD thesis]. University of Bolton. Clarke PMR, Barrett L, Henzi SP. 2009. What role do olfactory cues play in chacma baboon mating? Am J Primatol. 71:493–502. Clarke PMR, Henzi SP, Barrett L. 2009. Sexual conflict in chacma baboons, Papio hamadryas ursinus: absent males select for proactive females. Anim Behav. 77:1217–1225. Clarke PMR, Pradhan G, van Schaik CP. 2009. Intersexual conflict in primates: infanticide, paternity allocation, and the role of coercion. In: Muller MN, Wrangham RW, editors. Sexual coercion in primates and humans: an evolutionary perspective on male aggression females. Cambridge (MA): Harvard University Press. p. 42–77. Clutton-Brock GA, Parker TH. 1995. Sexual coercion in animal societies. Anim Behav. 49:1345–1365. Collins DA. 1981. Social behaviour and patterns of mating among adult yellow baboons (Papio c. cynocephalus. l. 1766) [PhD thesis]. University of Edinburgh. Cox CJ, LeBoeuf BJ. 1977. Female incitation of male competition: a mechanism of mate selection. Am Nat. 111:317–355. Domb LJG. 2000. Sexual swellings in wild baboons (Papio cynocephalus anubis) at Gombe National Park, Tanzania [PhD thesis]. Harvard University. Gesquiere LR, Wango EO, Alberts SC, Altmann J. 2007. Mechanisms of sexual selection: sexual swellings and estrogen concentrations as fertility indicators and cues for male consort decisions in wild baboons. Horm Behav. 51:114–125.

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

a

Subspecies

1220

Pizzari T. 2001. Indirect partner choice through manipulation of male behaviour by female fowl, gallus gallus domesticus. Proc R Soc Biol Sci. B268:181–186. Port M, Clough D, Kappeler PM. 2009. Market effects offset the reciprocation of grooming in free-ranging redfronted lemurs, Eulemur fulvus rufus. Anim Behav. 77:29–36. Pradhan GR, van Schaik C. 2008. Infanticide-driven intersexual conflict over matings in primates and its effects on social organization. Behaviour. 145:251–275. R Development Core Team. 2009. R: a language and environment for statistical computing. Vienna (Austria): R Foundation for Statistical Computing. Ransom T. 1981. Beach troop of the Gombe. Lewisburg (PA): Bucknell University Press. Rasmussen KLR. 1980. Consort behaviour and mate selection in yellow baboons (Papio cynocephalus) [PhD thesis]. University of Cambridge. Reichard M, Bryja J, Ondrackova´ M, Da´vidova´ M, Kaniewska P, Smith C. 2005. Sexual selection for male dominance reduces opportunities for female mate choice in the European bitterling (Rhodeus sericeus). Mol Ecol. 14:1533–1542. Rowell T. 1966. Forest living baboons in Uganda. J Zool. 149:344–365. Rowell TE. 1968. Grooming by adult baboons in relation to reproductive cycles. Anim Behav. 16:585–588. Rowell TE. 1974. Contrasting adult male roles in different species of nonhuman primates. Arch Sex Behav. 3:143–149. Seyfarth R. 1978. Social relationships among adult male and female baboons. I. Behaviour during sexual consortship. Behavior. 64: 204–226. Smuts BB. 1985. Sex and friendship in baboons. New York: Aldine. Smuts BB, Smuts R. 1993. Male aggression and sexual coercion of females in nonhuman primates and other mammals: evidence and theoretical implications. Adv Study Behav. 22:1–63. Stopka P, Graciasova´ R. 2001. Conditional allogrooming in the herbfield mouse. Behav Ecol. 12:584–589. Stopka P, Macdonald DW. 1999. The market effect in the wood mouse, apodemus sylvaticus: selling information on reproductive status. Ethology. 105:969–982. Tanaka I, Takefushi H. 1993. Elimination of external parasites (lice) is the primary function of grooming in free-ranging Japanese macaques. Anthropol Sci. 101:187–193. Thornhill R, Alcock J. 1983. The evolution of insect mating systems. Cambridge (MA): Harvard University Press. van Schaik CP, Pradhan GR, van Noordwijk MA. 2004. Mating conflict in primates: infanticide, sexual harassment and female sexuality. In: Kappeler PM, van Schaik CP, editors. Sexual selection in primates: new and comparative perspectives. Cambridge: Cambridge University Press. p. 131–150. Wedell N, Gage MJG, Parker GA. 2002. Sperm competition, male prudence and sperm-limited females. Trends Ecol Evol. 17: 313–320. Weingrill T, Lycett J, Henzi S. 2000. Consortship and mating success in chacma baboons (Papio cynocephalus ursinus). Ethology. 106: 1033–1044. Weingrill T, Lycett JE, Barrett L, Hill RA, Henzi SP. 2003. Male consortship behaviour in chacma baboons: the role of demographic factors and female conceptive probabilities. Behaviour. 140:405–427. Whittingham MJ, Stephens PA, Bradbury RB, Freckleton RP. 2006. Why do we still use stepwise modelling in ecology and behaviour? J Anim Ecol. 75:1182–1189. Wiley RH, Poston J. 1996. Indirect mate choice, competition for mates, and coevolution of the sexes. Evolution. 50: 1371–1381. Wong BBM, Candolin U. 2005. How is female mate choice affected by male competition? Biol Rev Camb Philos Soc. 80:559–571.

D ownloaded from be he co.oxfordjourna ls.org at U niversity of C alifornia, D avis on Ja nua ry 23, 2011

Gomes CM, Boesch C. 2009. Wild chimpanzees exchange meat for sex on a long-term basis. PLoS One. 4:e5116. Gumert MD. 2007a. Grooming and infant handling interchange in Macaca fascicularis: the relationship between infant supply and grooming payment. Int J Primatol. 28:1059–1074. Gumert MD. 2007b. Payment for sex in a macaque mating market. Anim Behav. 74:1655–1667. Hammerstein P, Hagen EH. 2005. The second wave of evolutionary economics in biology. Trends Ecol Evol. 20:604–609. Ha¨rdling R, Jormalainen V, Tuomi J. 1999. Fighting costs stabilise aggressive behaviour in intersexual conflicts. Evol Ecol. 13:245–265. Hausfater G. 1975. Dominance and reproduction in baboons (Papio cynocephalus). Contrib Primatol. 7:1–150. Hemelrijk CK, van Laere GJ, van Hoof JARAM. 1992. Sexual exchange relationships in captive chimpanzees? Behav Ecol Sociobiol. 30: 269–275. Henzi SP, Barrett L. 1999. The value of grooming to female primates. Primates. 40:47–59. Henzi SP, Barrett L. 2002. Infants as a commodity in a baboon market. Anim Behav. 63:915–921. Henzi SP, Clarke PMR, van Schaik CP, Pradhan GR, Barrett L. 2010. Infanticide and reproductive restraint in a polygynous social mammal. Proc Natl Acad Sci U S A. 107:2130–2135. Hockings KJ, Humle T, Anderson JR, Biro D, Sousa C, Ohashi G, Matsuzawa T. 2007. Chimpanzees share forbidden fruit. PLoS One. 2:e886. Hogg J. 1984. Mating in bighorn sheep: multiple creative male strategies. Science. 225:526–529. Hunt J, Breuker CJ, Sadowski JA, Moore AJ. 2009. Male–male competition, female mate choice and their interaction: determining total sexual selection. J Evol Biol. 22:13–26. Jennions MD, Petrie M. 2000. Why do females mate multiply? A review of the genetic benefits. Biol Rev Camb Philos Soc. 75:21–64. Johnson JB, Omland KS. 2004. Model selection in ecology and evolution. Trends Ecol Evol. 19:101–108. Johnstone RA, Bshary R. 2002. From parasitism to mutualism: partner control in asymmetric interactions. Ecol Lett. 5:634–639. Johnstone RA, Bshary R. 2008. Mutualism, market effects and partner control. J Evol Biol. 21:879–888. Keverne EB, Martinez N, Tuite B. 1989. Beta-endorphin concentrations in cerebrospinal fluid of monkeys are influenced by grooming relationships. Psychoneuroendocrinology. 14:155–161. Kitchen DM, Beehner JC, Bergman TJ, Cheney DL, Crockford C, Engh AL, Fischer J, Seyfarth RM, Wittig RW. 2009. The causes and consequences of male aggression directed at female chacma baboons. In: Muller MN, Wrangham RW, editors. Sexual coercion in primates and humans. Cambridge (MA): Harvard University Press. p. 128–156. Koprowski JL. 1993. Alternative reproductive tactics in male eastern gray squirrels: ‘‘making the best of a bad job’’. Behav Ecol. 4:165–171. Metz M, Klump GM, Friedl TW. 2007. Temporal changes in demand for and supply of nests in red bishops (Euplectus orix): dynamics of a biological market. Behav Ecol Sociobiol. 61:1369–1381. Moore AJ, Gowaty PA, Wallin WG, Moore PJ. 2001. Sexual conflict and the evolution of female mate choice and male social dominance. Proc R Soc Biol Sci. B268:517–523. Moore AJ, Moore PJ. 1999. Balancing sexual selection through opposing mate choice and mate competition. Proc R Soc Biol Sci. B266:711–716. 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. Norscia I, Antonacci D, Palagi E. 2009. Mating first, mating more: biological market fluctuation in a wild prosimian. PLoS One. 4:e4679.

Behavioral Ecology