Multi-model inference of non-random mating from an information theoretic approach

Area of Science:

• Evolutionary biology
• Quantitative genetics
• Behavioral ecology

Background:

• Non-random mating significantly influences evolutionary trajectories.
• Understanding the interplay between mating causes (mate choice, competition) and consequences (sexual selection, assortative mating) is crucial.
• Existing methodologies often struggle to disentangle these complex mating dynamics for discrete traits.

Purpose of the Study:

• To develop a standardized modeling framework for discrete traits to infer mating parameters.
• To connect causal mechanisms of non-random mating with observable mating patterns.
• To provide a tool for distinguishing between alternative processes driving observed mating patterns.

Main Methods:

• Developed a modeling framework for discrete traits with multiple phenotypes.
• Derived maximum likelihood estimates for various mating models.
• Employed information criteria for multi-model inference and selection.
• Applied the methodology to empirical data from Littorina saxatilis.

Main Results:

• The framework demonstrated good performance in model selection and parameter estimation via simulations.
• Analysis of Littorina saxatilis ecotypes revealed mating patterns best explained by models incorporating both mate choice and competition.
• These factors jointly generated positive assortative mating and female sexual selection.

Conclusions:

• This work presents the first standardized methodology for model selection and multi-model inference of mating parameters for discrete traits.
• The InfoMating software facilitates the analysis of complex mating systems.
• The framework offers a powerful tool to empirically investigate the causal mechanisms underlying sexual selection and assortative mating patterns.