Last update: : Wednesday 8 March 2017, by
Par : Cédric Sueur, IPHC-DEPE
Date : jeudi 9 mars 2017 de 15h à 16h
Lieu : CNRS, Amphithéâtre Marguerite Perey, bâtiment 01
Network optimality has been described in genes, proteins and human communicative networks. In the latter, optimality leads to the efficient transmission of information with a minimum number of connections. While studies have shown that differences in node centrality exist in animal affiliative networks, e.g. central individuals having higher fitness, network efficiency has not much been studied in animal groups. Living in groups has many advantages but it also involves certain disadvantages such as increased disease transmission and the need to make collective decisions (time costing). In theory, the social network properties optimizing decision accuracy and the spread of information should also increase disease transmission rates, creating a trade-off between decision-making efficiency and infection risk. We aim to explore this trade-off by examining social network properties and investigating how they might interact to maximize decision accuracy and minimize infection risk. Concerning information transmission during collective decision-making, we show in several species that individual centrality in the affiliative network elicit stronger follower behaviour and are crucial to the achievement of consensus decisions. This functional embedded leadership improves the efficiency of the decision-making process, enabling faster decisions. Our data also suggest that a behavioural rule-of-thumb ‘follow my close affiliate’ can result in groups, in order to benefit from the knowledge of elder, dominant, or natal individuals (who are often central in social networks) while simultaneously maintaining bonds with highly social individuals that may provide indirect fitness benefits. However, when doing a meta-analysis on primates, we found that centralisation and modularity had opposing effects on network efficiency, showing that tolerant species had more efficient networks to spread information. Moreover, neocortex ratio was positively correlated with network efficiency. Such network properties affecting individual fitness could be shaped by natural selection affecting both information and disease transmission. Modularity and efficiency also to be higher in primate networks compared to artificial networks. These results highlight that the transmission of information versus disease in animal groups may be favoured by an equilibrium between centralized and decentralized networks.