Human and non-human animals live in "social networks" of interactions and relationships. For example, one animal might interact frequently with its groupmates, while another rarely does so. An animal might be aggressive to another, but interact peacefully with yet others. The researchers will investigate how these social networks change after conflict to either make conflict worse in the future, or to prevent conflict from re-emerging. The researchers hypothesize that social networks that we observe in nature will tend to change in ways that limit further conflict for two reasons. First, we might not have much of a chance to observe networks that do not do so, as otherwise conflict becomes so severe that the whole group quickly breaks apart. Second, individual animals may benefit from avoiding conflict, and so they modify their behavior in ways that reduce the risks of further conflict. To test this hypothesis, the researchers propose to explore how social networks change after conflict in the cichlid fish Neolamprologus pulcher, which live in stable, long-lasting groups using dynamical network models coupled with experiments in nature. If social networks change in ways that limit conflict, this would help understand how groups of individuals with competing interests are able to stay together and prevent day-to-day quarrels from escalating. This project involves international collaboration in Zambia and training of students in both mathematical and biological research. This project is co-funded by the Animal Behavior program in the Division of Integrative Organismal Systems, the Mathematical Biology program in the Division of Mathematical Sciences, the BIOMAPS program for proposals at the interface of Biology, Math and the Physical Sciences, and from the Office of International Science and Engineering.
The researchers propose to perturb groups of N pulcher in the laboratory and the field, for example by removing individuals to create opportunities for others to change social status. They will conduct intensive behavioral observations on all group members before and after perturbation. They propose to use stochastic actor oriented models (SAOM) to analyze social network change. Stochastic actor-oriented models are a type of time ordered dynamic network model in which individual behaviors and network structures coevolve. They propose to use mathematical and computational models to test whether the patterns of connections found from the SAOM analyses are associated with low levels of conflict and high group stability. Finally, the researchers propose to measure stress hormone (cortisol) levels and reproductive success to investigate whether individuals benefit from being in groups that are better able to limit conflict. Findings from this research will provide insights into how individuals' social environment emerges through the effects of their interactions with one another, the resilience of the social environment to perturbation, and how the emergent social environment feeds back on individual performance and reproductive success.
