Iterated Prisoner Dilemma boost the emergence of high reciprocal altruism-based cooperative behaviors in rats

DELMAS GUILLERMO; LEW S. E; B. S. ZANUTTO

Washington

Congreso; Neuroscience 2017; 2017

SFN

The underlying mechanisms that explain cooperative behaviors among unrelated remain unclear in evolution theory. However, it was shown that the use of appropriate 2x2 pay-off matrices in the iterated Prisoner Dilemma (iPD), boost the emergence of reciprocal altruism-based cooperative behaviors, as a way to learn to maximize reward. When subjects face an opponent with cooperative strategy (Tit for Tat) and iPD matrix, the biggest long-term reward is available when both actor and opponent subjects cooperate mutually. It happens when mutual cooperation is the most probable state out of the four possible states in the game. Although, iPD is successfully learnt in humans it has not been found in animals. Here, we show for the first time that using a specific pay-off matrices and punishment (timeout), long-Evans rats were able to learn reciprocal altruism behaviors with cooperation rates over 80%. Two groups of unrelated male Long-Evans rats (two months old) were trained using dual opposite operant chambers, visual contact between animals was allowed through a small window. The experimental and opponent groups were food restricted in order to maintain their weights at 90-95% and 80%-85% of their ad libitum weights, respectively. When modeled as Markov chain transition probabilities, rats behaviors show a strong preference to stay in the mutual cooperation state or to return to it as soon as a wrong response is executed.