OOS 18-2 - How many individuals do you need to calculate ITV?

One of the main objectives of ecological studies is to assess how species relate to their biotic and abiotic environment. Recently, the role of biotic and abiotic factors in community assembly has been explored using intraspecific trait variation (ITV). Although ITV has been used to answer questions in population and community ecology across a vast range of taxa and systems, key questions remain in terms of robust methodology for calculating ITV. First and foremost is, how many individuals are needed to accurately quantify ITV? In this study, we use simulations of population-level ITV along an environmental variable to determine the chances of committing Type I and Type II errors. We simulated a series of scenarios where multiple populations differed in population level ITV and ITV showed a range of relationships with the environmental variable. Across scenarios, we also altered the number of individuals sampled from each population and the number of populations. Overall, Type I error increased when there was a high correlation between the number of individuals sampled and the environmental gradient. Type II error increased significantly when ITV showed lower correlations with the environmental gradient, and less than 20 individuals were sampled per population. Our results indicate that the number of individuals necessary to accurately characterize population-level ITV depends on ITV-gradient relationships. In order to minimize the chance of Type I and Type II errors, population level ITV, the number of individuals sampled, the environmental gradient assessed, and the number of populations must be taken into account. These parameters will likely differ across taxa and study systems. However, as researchers detail and report these parameters for their ITV studies, a general consensus for particular taxa and systems may emerge to help guide ITV sampling design.