Measuring Natural Selection in the Wild, Max Gotts, UG '24 (9D2FCEE5)

Fitness landcapes are incredibly powerful tools for visualizing the action of natural selection on rapid timescales. However, collecting the relevant empirical data to generate them is costly and time-consuming. This article creates a simple model for natural selection according to a fitness landscape, and uses it to demonstrate that each fitness landscape produces a unique distribution of traits. An algorithm is designed to reverse this process that is robust to noise. This algorithm is tested on a dataset of Iberian bluethroat morphology to show that visualizing fitness is valuable and necessary to understand evolution. This method has the potential to open a new way of investigating evolution, and supplements current methods in evolutionary ecology to help biologists interpret and understand phenotypic adaptation from readily-accessible field data. This project is currently unpublished, so please do not take or share any photos of the equations or results. (This project is currently unpublished, so please do not take or share any photos of the equations or results.)