The six observations reported here show that the transfer of food between strangers is not unique to humans or other great apes. In each case, the food items that were transferred could have been obtained individually, without any requirement for cooperation or specialised skills. This is particularly true for observations 1–3 as food was available in several areas other than the place where the food transfer occurred. In these three cases, individuals could have easily taken food directly from the pots, and even sometimes from individuals of their own group. In observations 4–6, however, there was only one source of food present, leading to congregation around the device. This lack of multiple food sources in observations 4–6 could have meant that it was easier for an individual to obtain food from another individual, even if that individual was from another group, than obtain food from the device, despite the agonistic interaction that could have arisen as a consequence of this. However, these six food transfer events were very similar to those that occur between individuals of the same group.
These instances of transfer of food between members of different groups of golden lion tamarin may show, for the first time, a level of tolerance in this species similar to that seen in bonobos (Fruth and Hohmann 2018). However, the social structures of golden lion tamarins and bonobos are very different: while bonobos show moderate aggression towards other individuals, including individuals of other communities, and live in fission–fusion groups within a community, golden lion tamarins are aggressive towards potential immigrants, and live in a family-structured group where offspring emigrate, on average, when they are around 2.5 years old (Baker and Dietz 1996; Fruth and Hohmann 2018; Romano et al. 2019). Moreover, unlike in bonobos, the food resources that were transferred between the golden lion tamarins could have been easily monopolised by just one individual. Both the social and physical structure of golden lion tamarins’ environment are very different to those of other primate species, such as bonobos, in which intergroup food transfers have been observed. It is therefore unlikely that tolerance between different groups or resource defence help explain intergroup food transfers in golden lion tamarins, and likely that this type of sharing between groups evolved independently in different primate linages. In primates, there is evidence that food transfers may be used by females to test a male’s tolerance (Goffe and Fischer 2016; van Noordwijk and van Schaik 2009; Yamamoto 2015). Yamamoto (2015) proposed a begging-for-social-bond hypothesis, in which individuals beg to strengthen social bonding as well as to gain access to food. Although insufficient, if supported by further experiments, the observations described here would indicate that this hypothesis is applicable to species other than those of great apes, and also that social bonds may extend beyond a group. Given that the observations of golden lion tamarins reported here violate previous expectations of which social and environmental conditions drive food transfers between groups, additional work on this species could further our understanding of the function of intergroup food transfers.
Although food transfers between different groups of golden lion tamarins could be indicative of a new level of social tolerance in this species, it is important to note that they are highly territorial, with intergroup interactions usually being aggressive (French and Inglett 1989; Peres 1989; Ruiz-Miranda et al. 2002). During dispersal events, resident golden lion tamarins are also aggressive toward immigrants (Baker and Dietz 1996), so the six observations of food transfer between individuals of different groups reported here are particularly interesting, and also inconsistent with previously suggested functions of food transfer in this species. In golden lion tamarins, food transfers have mainly been studied in the context of providing nutrition or information to juveniles, or nutrition to pregnant females. Here I suggest that food transfers may be used to create and/or strengthen social bonds with non-group members.
Subordinate golden lion tamarins have two main reproductive options: to wait for a breeding opportunity in their natal group, while caring for the young of the breeding pair, or to emigrate to explore their own breeding opportunities (Romano et al. 2019). Both male and female golden lion tamarins disperse from their natal group and settle in the first available breeding position or unoccupied area that they encounter; however, males tend to disperse more frequently than females, and are more successful when dispersing (Baker and Dietz 1996; Dietz and Baker 1993; Moraes et al. 2018; Romano et al. 2019). Furthermore, males and females use different strategies to emigrate: males are more likely to immigrate into established groups, whereas females are more likely to form new groups (Romano et al. 2019), and are also more likely to inherit their natal territory than males (Baker and Dietz 1996). Leaving the natal group for reproduction is very risky for tamarins, but encounters with neighbouring groups provide opportunities to identify potential sexual partners or new group members (Nascimento et al. 2014). Food transfers between individuals of different groups could therefore be used to create a social bond prior to immigration, which could either facilitate acceptance and reduce aggression when immigrating to a new group, or enable individuals to find social partners to form a new group with. Work on captive callitrichids suggests that immigration might be limited by aggression from resident individuals, with female residents being particularly intolerant of other females (e.g., French and Inglett 1989; French and Snowdon 1981; Harrison and Tardif 1989). Food transfers with members of a different group could therefore be a potential mechanism to recruit new members into a group by modulating the level of tolerance towards unfamiliar conspecifics (French and Inglett 1989). Food transfers with individuals from a different group, particularly with individuals of a different sex, might enable future immigrants to assess their likelihood of being integrated into a new group, or of finding a mate. Miller et al. (2003) found evidence that scent markings are not used for territorial defence in golden lion tamarins, but might be used as a way to communicate information for mate selection, extra-group copulation, and/or to attract immigrant partners. Food transfers could similarly be used to decrease aggression in order to communicate information beyond the group. If food transfers are indeed used as a means of creating a social bond with non-group individuals prior to dispersal, one would expect individuals involved in between-group food transfers to be of an age to disperse. According to Romano et al. (2019) natal emigration occurs around 2.5 years of age (range: 12–70 months). All the receivers of the between-group transfers in the six reported observations were within this range, except for one tamarin which was 11 months old, and all the receivers were younger than 2.5 years old, except for one which was 3 years 4 months old. The age of the donors was more variable, i.e. from 1 year 4 months to 7 years 4 months, but all the donors were within the age range given by Romano et al. (2019) except for one individual. However, we would not necessarily expect the age of a donor to fall within the natal emigration age range, as the receiver might be immigrating into the donor’s group, whereas the donor might not necessarily emigrate.
Romano et al. (2019) found that conspecific attraction, where individuals leave their natal group because they are attracted to potential extra-group mates and/or emigrating group mates, characterises emigration for both male and female golden lion tamarins. Food transfers might be a way of assessing potential extra-group mate quality or acceptance, and Hankerson and Dietz (2014) suggest that males in particular might prospect neighbouring groups for breeding opportunities. Hence food transfers might be particularly useful for males deciding where and when to immigrate to reduce the probability of eviction. Romano et al. (2019) also found evidence for parallel dispersal (emigration with peers or close kin) in golden lion tamarins. Since females are more likely to start new groups than males, they might evaluate potential mates or social partners for this through intergroup food transfers. Long-term data are required to determine whether individuals involved in food transfers with members of other groups then immigrate preferentially into these groups, or start new groups with members who have shown tolerance towards them.
Four out of the six observations reported here were food transfers between individuals of different sex: in three observations (1, 2, 3) the food went from a male to a female, while in one observation (4) it went from a female to a male. These four observations might be examples of individuals assessing the quality of a potential mate through food transfer prior to dispersing. If so, these observations would support the sexual selection hypothesis, which postulates the occurrence of competition in the choice of a mate (West-Eberhard 1983). However, two of the observations were of food transfer between individuals of the same sex: in observation 5 the food went from a female to a female, and in observation 6 the food went from a male to a male. It is possible that instead of being a means of helping a tamarin to choose between potential mates, these food transfer events also help individuals to select a future social partner prior to dispersing. It is also interesting to note that the two food transfers between individuals of the same sex took place rapidly, and were more akin to what is described in the literature as “food stealing”. Overall, I suggest that food transfer to a member of a different group may be a means of creating a social bond with that individual, especially prior to dispersal. Further data are required to assess which of these outcomes is more likely to primarily drive food transfers between adults of different groups.
One limitation of the observations reported here is that they were made during an experiment where food items were provided to golden lion tamarin groups. Thus it is possible that the increase in food availability induced an atypical level of tolerance towards non-group members, resulting in the observed food transfers. However, I think it is unlikely that the increase in food availability created this high level of tolerance because half of the observations (3, 5, and 6) were made during the fruiting season, when so much food was available that it was sometimes difficult to interest the groups in the fruit provided in the experiments, as they were more interested in the fruit in the trees. Furthermore, some experiments undertaken on tamarins in captivity have indicated that food transfers are less likely when food is abundant (e.g. Price and Feistner 1993) because it is easier for the tamarins to acquire food personally instead of getting it from another individual. Moreover, although interactions between different groups are usually quite aggressive initially, they can last several hours, and lose intensity over time (Peres 1989). During face-to-face encounters, when no experiments are taking place, individual golden lion tamarins still forage, although less than at other times (Peres 1989). When a group’s territory overlaps with that of another group and when encounters de-escalate, individuals from different groups have been observed feeding on the same tree (C. R. Ruiz-Miranda, personal communication). Although not widely reported in the literature, this is relatively common, particularly in large ficus trees (C. R. Ruiz-Miranda, personal communication). Juveniles from different groups have also been observed playing together, without much interference from adults (C. R. Ruiz-Miranda, personal communication), suggesting a certain level of tolerance, at least towards some individuals. Hence, given those observations, it seems probable that food transfers between individuals of different groups occur outside of the context of human provisioning.
Conclusion
Most previous work on food transfers in golden lion tamarins has focussed on transfers from adults to young to determine whether the function of the transfer is to provide nutrition, information or both (Price and Feistner 1993; Rapaport 1999; Troisi et al. 2020). Up until now, adult-adult food transfers have only been reported in captive or reintroduced individuals (Ruiz-Miranda et al. 1999). Here there is not only evidence of adult-adult food transfers in the wild, but also of food transfers between individuals of different groups, which is inconsistent with the previously suggested functions of food transfers in this species. I suggest an additional function of food transfers in wild golden lion tamarins: that they create and/or strengthen social bonds with individuals outside of the family group, which could be particularly useful for tamarins prior to immigrating to a new group or founding a new group with individuals from other groups. Although the function of intergroup encounters in lion tamarins is not fully understood, the six observations reported here contribute to the growing body of literature showing the flexibility of social behaviour in callitrichids. Taken together, the different functions of food transfers in wild populations of primates offer us insights into their social behaviour (Goffe and Fischer 2016).