Do social partners affect same-sex sexual behaviour in male water striders?

Same-sex sexual behaviour (SSB) represents an emergent trait of two interacting same-sex individuals. Although empirical studies have investigated how social environments can influence SSB, little is known about the effect of the interacting partner and its associated phenotype on SSB. In species where females are larger than males or males express male-specific behaviour, a male's morphology or behaviour can serve as a way for other males to recognize its sex and express SSBs. Here we used both a trait-based and variance-partitioning approach to test for the effect of the interacting male's identity and his multivariate phenotype on SSB, using water striders, Gerris lacustris, as a model. We repeatedly subjected males to dyadic interactions with the same and different partners and measured their SSBs. We used the variance-partitioning approach to estimate the partner's identity effect, and the trait-based approach to assess which trait of the partner explains changes in SSB. We found that the partner's SSB reduced the tendency of males to show SSB. SSB was affected by their interacting partner's identity, but the partner effect was not due to the repeatable components ('personality') of the partner's SSB. Males also did not differ in their responses to variation in the partner's SSB at different levels (between-partner, or partner 'personality', and within-partner levels, or partner 'plasticity'). Taken together, these findings provide the first empirical evidence that SSB can be plastically expressed in response to traits in social partners. We also highlight the usefulness of combining the trait-based and variance-partitioning approach to test whether partners represent a component of the social environment affecting the expression of labile traits.