Annuaire IPHC

DEPE | Séminaires BEEPSS / BEEPSS Seminars » Individual-based Models of Reciprocation and Interchange of Social (...)

Individual-based Models of Reciprocation and Interchange of Social Behaviors

Last update: : Tuesday 3 October 2017, by Nicolas Busser

Presented by: Ivan Puga-Gonzales, IPHC-DEPE
Date: Thursday 28th Sept 2017, 13:30-15:00
Place: IPHC, Amphithéâtre Grünewald, bâtiment 25

The proximate mechanisms giving rise to reciprocation and interchange of social behaviours in primates have been extensively debated. Three mechanisms have been proposed: ‘calculated reciprocity’, ‘emotional bookkeeping’ and ‘symmetry-based reciprocity’. Based on experimental and observational data, researchers appear to favour emotional bookkeeping. A different approach to study potential mechanisms underlying reciprocation and interchange is by means of individual-based computer models (IBMs). Several different IBMs have been developed in this regard. In this talk, I will focus on three IBMs, identical regarding the rules for individuals’ social interactions but different regarding the mechanisms given rise to patterns of reciprocation and interchange of social services: ‘GrooFiWorld’ suggests proximity-based interactions; ‘FriendsWorld’ suggests proximity-based interactions and preferential interactions; and the ‘Reaper model’, suggests preferential interactions alone as the basis of these patterns. Because the three IBMs reproduce patterns of reciprocation and interchange, it is difficult to tell which one, if any, represents more accurately empirical data. To clarify this, we used a social network approach. We collected data from 14 groups belonging to eight macaque species and did a social network analysis. Then, we simulated each group in each IBM, analyzed the emergent social networks, and quantitatively compared empirical and models’ networks. Results show that the three IBMs capture to some degree the features observed in social networks of macaques; although, some global network metrics (e.g. modularity) were poorly fit by all IBMs; suggesting then, that in the IBMs social interactions may be simpler than in reality. Further, although no IBM seemed better than the others at fitting the empirical data, the analysis indicates that spatial-structure is an important factor influencing distribution of social grooming since the two spatial-explicit IBMs performed best.