Unidirectional response to bidirectional selection on body size II. Quantitative genetics.

Area of Science:

• Evolutionary biology
• Quantitative genetics
• Animal breeding

Background:

• Accurate estimation of trait evolvabilities is crucial for predicting evolutionary changes under selection.
• Multivariate quantitative genetics models are used to predict trait evolution, but their precision for fitness-related traits is empirically uncertain.

Purpose of the Study:

• To assess the predictive accuracy of quantitative genetics models for bivariate artificial selection on body size and maturity in medaka fish.
• To investigate the extent to which observed evolutionary responses align with theoretical predictions.

Main Methods:

• Bivariate artificial selection was applied to body size and maturity in three lines of medaka fish (Large, Small, Control).
• A random effect ('animal') model was used to remove nonheritable trends and noise.
• Observed selection responses were compared to theoretical predictions.

Main Results:

• The Small line did not evolve smaller body length; its response was smaller than predicted.
• The Large and Control lines showed expected responses for body size.
• The response for maturity was opposite to predictions in both selected lines, with the Large line evolving late maturity and the Small line evolving early maturity.

Conclusions:

• Observed evolutionary responses in medaka fish substantially mismatched theoretical predictions from quantitative genetics models.
• The discrepancies could not be explained by common sources of uncertainty like sampling effects or genetic drift.
• This highlights limitations in the predictive power of current models for complex trait evolution.