The ontogeny of repeatability: measurement choices shape individuality estimates in a clonal fish

Abstract

Consistent among-individual differences in behavior across time and contexts ("individuality", "animal personality") are increasingly recognized as playing an important role in both the ecology and evolution of species. Such individuality is not static: individual differences emerge, shift, and crystallize across ontogeny, shaped by developmental plasticity, stochastic processes, and gene-environment interactions. Despite growing awareness of this developmental dimension, the field has relied predominantly on snapshot designs with few repeated measures, raising the question of how much reported repeatability estimates reflect biology versus methodological choices. Here, we leverage high-resolution tracking data from 45 clonal Amazon mollies (Poecilia formosa) recorded at 5 frames per second across the first 28 days of life to systematically examine how temporal aggregation, sampling rate, frame rate, and the number and placement of measurement days affect repeatability estimates and their underlying variance components. We find that repeatability increases with higher temporal aggregation, decreases with lower sampling rates, and varies substantially depending on the developmental stage sampled. These results demonstrate that commonly used measurement practices can substantially distort repeatability estimates, and that conclusions about behavioral consistency may be as much a product of study design as of biology. We close with practical recommendations for researchers studying individuality across development.