Capuchin monkeys (Sapajus[Cebus]apella) play Nash equilibria in dynamic games, but their decisions are likely not influenced by oxytocin

Abstract

Comparative approaches to experimental economics have shed light on the evolution of social decision-making across a range of primate species, including humans. Here we replicate our previous work looking at six pairs of capuchin monkeys' (Sapajus [Cebus] apella) responses to scenarios requiring both coordination (Assurance Game) and anticoordination (Hawk-Dove Game). This then provides a foundation for assessing their responses to two additional games, one with a scenario of beneficial cooperation with a temptation to defect (Prisoner's Dilemma) and one with an environment requiring changing strategies within short temporal proximity (Alternating Economic Game). We additionally explored the effects of exogenous oxytocin on decision-making. Oxytocin did not affect decisions in any of our games. Results from the first two games largely replicated our previous findings. Responses to the Prisoner's Dilemma were more varied than was seen in previous games, with pairs respectively cooperating, defecting, and failing to establish stable strategies. Such variability indicates that this game may be a good assay for individual differences in social decision-making. Finally, capuchins were able to flexibly switch between their previously established strategies within each of the different games, even when the games were presented within the same session, requiring strategy adjustments within short temporal proximity. These results build on earlier findings showing that capuchins can alter decision-making strategies as the context demands, which is likely essential for decision-making in naturally occurring contexts.

Shifts between cooperation and antagonism driven by individual variation: A systematic synthesis review

“…Conversely, another study on the Southern hairy-nosed wombat (Lasiorhinus latifrons) did not find a relationship between hormone profiles and decision-making (Walker, Taylor, & Sunnucks, 2007); instead the cooperativecompetitive outcome was determined by a state-dependent response to their physical environment. A similar pattern was found in capuchins (Cebus apella), where decisions to cooperate were context-dependent, but not influenced by exogenous oxytocin (Smith et al, 2019), suggesting that the relationship between individual physiology, behavioural strategy and environmental factors are likely to be species-specific. Interaction outcomes can also influence the hormonal state of individuals, as in the African striped mouse (Rhabdomys pumilio).…”

What behaviour in economic games tells us about the evolution of non-human species' economic decision-making behaviour

“…This game is particularly interesting because the best choice depends on whether it is a one-shot interaction, in which case the best outcome is to defect, or repeated, in which case cooperation dominates. In our results, capuchins' responses differ across pairs, with some tending to the one-shot NE of mutual defection and others to the mutually cooperative outcome, which maximizes benefit in the long term within stable, mutually cooperative relationships, which these highly social monkeys exhibit 3 [33]. Moreover, subjects seem to be sensitive to previous outcomes.…”

“…In anti-coordination games, such as the Hawk Dove game (figure 1b; also the Chicken or Snowdrift games), chimpanzees [30] and squirrel monkeys [28] rarely settle on a consistent pattern of play. Capuchin and rhesus monkeys are able to find the NE, but only when they can see their partners' choices [25,33]. One important royalsocietypublishing.org/journal/rstb Phil.…”

Anything for a cheerio: Brown capuchins (Sapajus [Cebus] apella) consistently coordinate in an Assurance Game for unequal payoffs

“…On the other hand, even the partner getting the lower value reward does absolutely better playing Stag , suggesting that they may still prefer to play it even if it results in a relatively less good payoff. Supporting this, in other economic games in which outcomes are not equal, such as the Hawk‐Dove game, capuchins continue to play a Nash equilibrium despite receiving less than their partner (Smith et al, 2019).…”

“…Finally, a few capuchins have coordinated in manual tasks, but capuchins have generally all succeeded in coordinating in computerized tasks, a difference that may be due to the increased number of trials per session and the shorter latency between choice and reward in computerized testing, both of which may support learning. However, capuchins appear to play the Stag / Stag Nash equilibrium by matching their partner's play, only succeeding when they can see what their partner has already played (Brosnan, 2011; Brosnan et al, 2012; Smith et al, 2019).…”