سلسله مراتب فضیلت: همزیستی، نوع دوستی و سیگنالینگ در شکار مشارکتی زنان مارتو

Cooperative hunting is often assumed to be mutualistic, maintained through returns to scale, where, by working together, foragers can gain higher per capita return rates or harvest sizes than they can by hunting alone. We test this hypothesis among Martu hunters and find that cooperation only provides increased returns to poorer hunters while disadvantaging better hunters. Even so, better hunters still cooperate as frequently as poorer hunters. We ask whether better hunters are advantaged in secondary sharing distributions or whether they bias their partner choice to kin or household members. We find that better hunters are not more likely to pair up with kin and they do not gain consumption benefits from acquiring more. They share a greater proportion of their harvest than poorer hunters: no matter how much one produces — better hunter, worse hunter, cooperator, solitary hunter — all eat the same amount in the end. Such a result suggests the hypothesis that cooperation might be a costly signal of commitment to the public interest on the part of better hunters, which generates trust among camp members and facilitates strong social networks, particularly among women, who cooperate more than men. While some foragers may benefit through cooperation from returns to scale or risk reduction, others may benefit more through signaling commitment and generating trust.

When hunters have the option of jointly pursuing prey, they are expected to do so primarily when there are direct, mutualistic benefits in the form of increasing returns to scale. Previous work has shown that cooperative hunting is likely to be maintained when coordinated action increases individual hunting success, prey encounter rates or harvest size obtained, or reduces the costs of search and pursuit leading to increasing per capita foraging return rates (Alvard, 2001, Alvard and Nolin, 2002, Packer and Ruttan, 1988, Smith, 1981 and Smith, 1991). The benefits achieved by cooperative hunting are gained through distribution of the subsequent harvest among the members of the hunting party, such that those who cooperate gain increased individual consumption return rates (measured as kilocalories consumed per unit of time spent foraging) relative to what they could achieve through hunting alone. Despite the synergistic benefits possible with cooperative hunting, conflicts of interest, failures of coordination, extensive free riding and heterogeneity among individual foragers in ability and access to group production can produce differential benefits to cooperation, raising questions about how collective action in group hunting and other forms of production is sustained (Hawkes, 1992 and Ostrom, 1991). If group foraging involves striking a balance between the benefits of cooperating and the costs of interference competition, groups might become larger than optimal for all members if solitary foragers still do better to join them rather than hunt alone (Smith, 1981, Smith, 1985 and Smith, 1991). When group members find it costly to exclude joiners, the benefits of cooperative hunting will be lower for all than if group sizes are kept close to optimal. Cooperation may also fail to provide higher per capita benefits if the distribution of prey following the hunt is biased or non-cooperators are allowed to gain access to the group's production. This can lead to a reduction in cooperation; Sosis, Feldstein and Hill (1998) found that the distributions of cooperative fishing groups on Ifaluk were biased toward canoe owners and large landholders, and, as expected, such individuals fished cooperatively more often than small-holders and men who did not own canoes. Cooperation is also sensitive to partner choice if hunters differ in their hunting abilities or effort, especially when solitary hunting offers returns that are just a bit lower than those from a cooperative hunt. If better hunters do not pair up with other good hunters, or exclude poor hunters from the party, they might not see any synergistic effects of cooperation relative to what they could achieve by hunting alone (Packer and Ruttan, 1988 and Winterhalder, 1996). If it is too costly for hunters to assert control over the composition of hunting groups, better hunters are expected to hunt alone more often than poorer hunters. Finally, cooperative hunting can only provide synergistic benefits if the benefits of cooperation are outweighed by the costs of harvest losses to non-hunters through demand sharing. Among Lamaleran whale hunters, crews of at least eight gain higher returns from whale hunting than from solitary fishing, but the payoffs to this strategy depend upon strict rules for division that specify certain portions to those who play certain roles on the hunt, thus minimizing losses to free riders (Alvard & Nolin, 2002). In this article, we ask whether or not returns to scale structure the benefits of cooperative hunting among Martu, Aboriginal foragers of the Western Desert in Australia. Martu are the indigenous owners of the estates that surround Lake Disappointment and the Percival Lakes in the northwest section of Australia's Western Desert (see Tonkinson, 1974, Tonkinson, 1978, Tonkinson, 1988, Tonkinson, 1991, Tonkinson, 1990, Tonkinson, 2007 and Walsh, 1990). As highly mobile hunter–gatherers in one of the most remote and arid regions of Australia, many Martu managed to maintain their lifeway relatively autonomous of colonial influence well into the second half of the 20th century. We focus particularly on Martu women, who frequently hunt small animals, and supply half of the bush meat that people consume (Bliege Bird & Bird, 2008). Historic band composition and contemporary residential patterns among Martu may have been critically shaped by the extent to which women engage in cooperative task specialization (Scelza & Bliege Bird, 2008). Sharing among Martu can also be extremely equitable for some resources, despite variance in productive effort (Bird and Bliege Bird, 2010 and Bliege Bird and Bird, 2008). Some of this variability stems from gender- and age-related specialization on different productive activities: some individuals spend more time hunting, others more time collecting; some produce more small game, others produce more large game (Bliege Bird and Bird, 2008 and Bird et al., 2009). Such variability provides an ideal case study of the factors influencing cooperative production. We first take a broad perspective on cooperative foraging among Martu, asking whether cooperation provides increasing returns to scale across foraging activities and how gender affects cooperation. We then narrow our focus to the hunting activity that provides the most meat by weight and comprises the majority of all hunting bouts, sand monitor hunting, asking whether cooperation on monitor hunts provides increasing returns to scale over solitary hunting, and whether there is individual heterogeneity in the benefits of cooperation. We ask whether better hunters compensate for losses incurred in cooperating with poorer hunters by cooperating less often, keeping more for themselves after sharing, or choosing to cooperate more often with kin. 1.1. Cooperation among Martu Extensive descriptions and analysis concerning the nature of different types of Martu foraging activities are provided elsewhere (Bird and Bliege Bird, 2005, Bird et al., 2009, Bliege Bird and Bird, 2005, Bliege Bird and Bird, 2008, Codding et al., 2010 and Bliege Bird et al., 2009). Below we focus on the most frequent hunting activity (sand monitor hunting) and provide some brief comments on the cooperation and patterns of distribution associated with other foraging activities used in our analyses. 1.1.1. Sand monitor hunting Sand monitor (Varanus gouldii) hunting is the most common foraging activity and mostly conducted by women. In this foraging activity, there are several different ways individuals might cooperate. In the winter season, where successful hunting involves targeting patches of old-growth spinifex grass for burning ( Bird et al., 2005 and Bliege Bird et al., 2008), two or more individuals often burn the same patch and share search costs for the prey revealed within. Each hunter will separate by about 100 m, calling the other over to assist in pursuit and capture if they find fresh tracks or a likely burrow. The hunters then coordinate in probing around the burrow in wide concentric circles with the point of their digging sticks to locate the den. One might dig up the entrance hole to determine the direction of the tunnel, the other probing for the terminal chamber. Depending on the depth of the den, hunters may then take turns in its excavation. In the summer season, when sand monitor are active on the surface, two or more hunters might cooperate to track and chase a single prey item, attempting to capture it before it retreats to its deep summer den. During a cooperative hunt, most partners pool their harvest: one hunter, usually the older or more skilled hunter, will transport all prey in her own bag. Typically, she will then take responsibility for cooking their pooled returns at the dinner-time camp (hereafter, DTC). As the individual monitor lizards are removed from the fire, she initially divides the harvest evenly between the partners she cooperated with while hunting (primary distribution). Each individual hunter then distributes her own portion to others in the DTC: her family, her children, her spouse, her brother or others with whom she did not hunt (secondary distribution). This is the only hunting activity that is usually characterized by producer control over distributions. While there is a strong negative relationship between the number of people in the DTC and the amount a producer keeps, some producers keep more than others, and it is in these secondary distributions where reputations for stinginess or generosity can be built and maintained. On cooperative hunts there are thus two ways that a hunter might bias the amount she keeps for herself: she might give less or more to her hunting partner(s) in primary distribution, and she might keep a portion for herself that is smaller or larger than the individual portions she gives away to others in secondary distribution. 1.1.2. Perentie and cat hunting Men and women are equally represented on feral cat (Felis catus) and perentie (Varanus giganteus and V. panoptes) hunts ( Bird et al., 2009). For these medium-sized mobile prey, cooperation generally entails two or more individuals searching separately and then coordinating in tracking the same animal when fresh tracks are found, pursing it to exhaustion. Unlike with sand monitors, but as with larger animals like kangaroo and bustard, the hunters typically do not cook and distribute their own harvests. If an older individual is present, she will generally take responsibility for cooking and dividing prey into between four and six standardized portions. Like the primary distribution of kangaroo, if the DTC is large, shares from medium-sized animals like these are sometimes distributed to “hearth groups” (individuals sitting at separate fires) within a DTC, rather than to individuals, and secondary distributions occur among members of the hearth group. This contrasts with the distribution of sand monitor, which is always from the acquirer to individual consumers. 1.1.3. Vehicle hunting Long-distance search using a vehicle is most common for hunts targeting the Australian bustard (Ardeotis australis), although if other high-ranked prey are encountered (perentie, kangaroo, monitor lizards, skink or patches of Solanum fruit) they will usually be pursued. Vehicle hunts are nearly always cooperative to some extent: large-mixed sex groups are common, and even if there is only a single rifle in the party, others will act as spotters and trackers, tracking prey and following it while in the vehicle. Mothers and grandmaternal caretakers often leave young children with vehicle hunters while they hunt on foot, in order to avoid having to carry them long distances. At the dinner camp, distributions of cooked bustard proceed in the manner described for other medium- to large-sized prey. 1.1.4. Kangaroo hunting While men most frequently hunt hill kangaroo (Macropus robustus) alone, they do sometimes cooperate. Cooperation often involves one hunter moving ahead on the hilltop, another in the acacia brush at the base of the hill to drive prey out of hiding and toward the rifle. If more than one hunter has a rifle, hunts became less coordinated, with each hunter separating on opposite sides of the rocky range so that prey missed by one might be shot at by the other. Sharing of kangaroo is highly formal, with the hunter playing no role in preparation or distribution but for exceptional circumstances. The formality allows for the cook/distributor to make the parcels in primary distribution equal for each recipient or hearth group present at the dinner camp (see Bird and Bliege Bird, 2010, Bird et al., 2009, Bliege Bird and Bird, 2008 and Codding et al., 2010, for detailed description and analysis of kangaroo hunting and sharing). 1.1.5. Collecting Collecting activities include picking Solanum fruit (S. centrale and S. diversifolium), digging roots or corms (Vigna lanceolata or Cyperus bulbosis), chopping tree-boring grubs from trunks or roots (Endoxyla spp.), harvesting the nectar of Hakea and Grevillia flowers, or collecting feral European bee (Apis) honey. For most of these non-hunting activities, cooperation is minimal, usually involving acquiring resources in the same patch or in close proximity, and pooling prior to returning to the dinner camp. An older woman with a large harvest might dump a portion of her surplus into the container of a younger person with less prior to leaving the patch, “so they could take home more to share”. This type of cooperation is most common for fruit, nectar, root or bulb digging or in acquiring grubs. Honey is the only collecting activity in which pursuits were necessarily collective: several individuals often took turns with the axe to extract honey from a single hive, some staying below the tree to tend smoky fires to keep the bees calm. 1.2. Hypotheses The first step in our analysis is to investigate whether or not cooperative foraging in general seems to be maintained through the benefits of increasing returns-to-scale; that is: H1:. Cooperation increases per capita return rates in all foraging activities The pattern of variability in the frequency of cooperation across foraging activities by the gender of the forager could tell us something about the benefits that influence cooperation. If cooperation is patterned by the benefits of returns-to-scale, then it should be most frequent in activities that show strong positive effects of cooperation on returns, less frequent in activities that show weak effects on returns and most infrequent in activities that show negative effects on returns. As an alternative, cooperation could be a means to overcome risk by increasing the chances of a successful harvest. If so, we might expect to see a correlation between the proportion of bouts in any activity that are cooperative and the expected probability of failure in that foraging activity. As we have suggested in previous publications (Bliege et al., 2008), Martu women may be more risk sensitive than men and, accordingly, we would expect this relationship to be much stronger for women than for men. Specifically we suggest that: H2:. Men and/or women both cooperate more frequently in activities that are associated with higher chances of bout failure We then focus our analysis on sand monitor hunting, which offers the most variability in cooperation and thus the greatest opportunity to link cooperative effort and its benefits. Sand monitor hunting parties are cooperative 40% of the time, women cooperate more often than men (65% of women's vs. 43% of men's bouts are cooperative) and these hunts account for nearly half of all foraging bouts on dinner camps (713 of 1532), giving us a large sample of both cooperative and noncooperative hunts. Furthermore, the number of hunts per individual is quite high, allowing us to examine intra-individual variability in hunting returns. The mutualism hypothesis suggests that: H3:. Cooperation within a single foraging activity (sand monitor hunting) should (a) increase the per capita return rate, (b) increase total harvest weight (in kilocalories of edible flesh), (c) reduce the time needed for foraging in sand monitor hunting or (d) increase the chances of a successful hunt Because our statistical methods are correlative, we also attempt to test for directional causality: does cooperation cause changes in return rates, or do external changes in return rates affect whether or not people cooperate? For example, a correlation between low returns and cooperation could mean either that cooperation causes low returns or that people are more likely to cooperate when returns are likely to be low. If so, we would expect a greater percentage of individuals to hunt cooperatively on dinner camps where solitary hunting offers lower returns than cooperative hunting. We would also expect solitary and cooperative hunting returns to be correlated across time and space: camps that see lower solitary returns should also see lower cooperative hunting returns. Because sand monitor harvests are always shared (on average an individual keeps 55±3% of any catch), return rates for cooperative vs. solitary hunting should be based on what a hunter (and his or her dependents) actually consumes after all sharing has taken place. Cooperative hunting may not provide higher acquisition returns, but cooperative hunters may be able to collectively defend a greater percentage of their catch and thus keep more for themselves and dependents (lose less to sharing) than will solitary hunters (producer priority). Or, alternatively, cooperative hunters may be able to collectively exert stronger claims to the catches of other hunters, receiving more than solitary hunters. The end result should be that cooperative hunters have higher consumption returns (grams of meat consumed by self and dependents per hour of foraging) than solitary hunters: H4:. Cooperators have producer priority: higher consumption returns (grams of meat consumed by self and dependents per hour of foraging) than solitary hunters Behavioral ecologists have long recognized that if individuals consistently differ in their pre-distribution foraging return rates due to skill or effort, and do not assort according to those skill levels, better or more productive hunters will be disadvantaged by cooperative hunting. If in the sharing of the harvest, better hunters do not receive a greater proportion of the catch than poorer hunters, they will not gain from cooperation and so should cooperate less frequently than poorer hunters: H5:. If there is heterogeneity in forager skill or production, better hunters should cooperate less frequently than poorer hunters If we find that better hunters continue to cooperate at the same frequency as poorer hunters, we can ask what direct or indirect benefits they might be receiving. We first focus on explanations for why there are consistent differences in hunter production — why do better hunters not simply quit earlier? One plausible explanation for overproduction is that it does benefit hunters to produce more in the short-term because those who demonstrate that they are better hunters may eat more in the end, either because others might be more likely to give to them (reciprocity) or they might be able to keep more of their own production for themselves (producer priority). We then turn toward explaining why better hunters cooperate so much. One plausible explanation is that better hunters have the social capital to be able to discriminate against their partners more readily, and while they may not be choosing to assort based on hunting skill, they may be more likely to pair up with kin or co-resident household members than are poorer hunters. Thus, while good hunters take on a cost to cooperate, they mediate that cost by directing the benefits of cooperation to close kin or other with whom they interact reciprocally on a daily basis. If better hunters choose more often to cooperate with kin or members of their own residential camps, they ensure that the subsequent pooling in primary distribution provides benefits to kin or residential camp members who might reciprocate in other ways. If so, we would expect a strong interaction between kinship and/or residency and the difference in hunt rank between partners. As the difference in rank between two cooperating partners increases, the effects of kinship and residential camp membership on the percentage of time partners affiliate should become stronger, thus: H6:. Better hunters should (a) benefit directly through having higher consumption returns or (b) benefit indirectly through biasing cooperative partner choice to kin or co-resident household members