Data from: The evolution of cooperation: interacting phenotypes among social partners

Models of cooperation among nonkin suggest that social assortment is
important for the evolution of cooperation. Theory predicts that
interacting phenotypes, whereby an individual’s behavior depends on the
behavior of its social partners, can drive such social assortment. We
measured repeated indirect genetic effects (IGEs) during cooperative
predator inspection in eight populations of Trinidadian guppies (Poecilia
reticulata) that vary in their evolutionary history of predation. Four
broad patterns emerged that were dependent on river, predation history,
and sex: (i) current partner behavior had the largest effect on focal
behavior, with fish from low-predation habitats responding more to their
social partners than fish from high-predation habitats; (ii) different
focal/partner behavior combinations can generate cooperation; (iii) some
high-predation fish exhibited carryover effects across social partners;
and (iv) high-predation fish were more risk averse. These results provide
the first large-scale comparison of interacting phenotypes during
cooperation across wild animal populations, highlighting the potential
importance of IGEs in maintaining cooperation. Intriguingly, while focal
fish responded strongly to current social partners, carryover effects
between social partners suggest generalized reciprocity (in which one
helps anyone if helped by someone) may contribute to the evolution of
cooperation in some, but not all, populations of guppies.