Decomposing social interactions: a statistical method for estimating social impact and social responsiveness

Area of Science:

• Behavioral Ecology
• Quantitative Genetics
• Evolutionary Biology

Background:

• Social interactions significantly influence how traits are expressed, impacting fitness.
• Phenotypic variation in social traits comprises direct effects, social responsiveness, and social impact.
• Existing models often fail to distinguish between responsiveness and impact, limiting understanding of social dynamics.

Purpose of the Study:

• To develop and evaluate a novel model for dissecting variation in phenotypic expression due to social interactions.
• To assess the performance of this model across different experimental designs using simulations.
• To investigate the impact of individual variation in social responsiveness on selection responses.

Main Methods:

• Developed a new statistical model to decompose social effects into direct, responsiveness, and impact components.
• Employed extensive simulations to test model performance under various conditions.
• Analyzed the influence of sample size, number of individuals, and social partners on estimation accuracy.

Main Results:

• Accurate estimation of variance components is achievable with a total sample size of at least 3200 individuals.
• Covariance estimation is most improved by increasing the number of unique individuals, followed by unique social partners.
• Ignoring individual responsiveness and measurement error leads to biased and imprecise trait-based analyses.

Conclusions:

• The proposed model effectively disentangles social responsiveness from social impact, offering a more nuanced understanding of social behavior.
• Adequate sample sizes are crucial for reliable estimation, with unique individuals being key for covariance accuracy.
• Failing to account for individual variation in responsiveness hinders accurate estimation of indirect genetic and social effects, impacting evolutionary studies.