Why May Repeatability of Highly Flexible Traits Say Little about their Evolutionary Potential?~!2009-10-27~!2010-02-24~!2010-03-22~!

A huge advance in our understanding on the relevance of phenotypic plasticity in evolutionary processes has taken place during the last decades. However, there are several specific topics in which the phenotypic plasticity framework has not been deeply considered. One of these topics is our current interpretation of trait repeatability, which is defined as the consistency of a trait over time (typically measured by the intraclass correlation coefficient). Common sense, as well as consideration based on quantitative genetics, suggests that the demonstration of significant repeatability of a trait necessarily precedes any attempt to demonstrate its selective significance. However, this statement is based on the vision that trait variation over time is in some point in between no change (i.e., fixed traits) and completely random variation. Here, I show, through a hypothetical (but realistic) example, that flexible traits able to respond to environmental factors randomly distributed among individuals can evolve through natural selection without showing a significant repeatability value.