charisma: An R package to perform reproducible colour characterization of digital images for biological studies

Abstract

Advances in digital imaging and software tools have provided increasingly accessible datasets and methods for analysing colour evolution. Despite the variety of computational packages available, most rely on colour classification before running analyses. Previous methods to characterise colour limit the ability to analyse large-scale image databases and are not always representative of biologically relevant colour classes, which decrease the accuracy of downstream analyses. Here, we present charisma, an R package designed to characterize the distribution of distinct colour classes in images suitable for large-scale studies of biological organisms. We demonstrate the utility of our package through an analysis of colour evolution in a sample of diverse and charismatic birds, tanagers, in the subfamily Thraupinae. We show that charisma can quickly and accurately classify every pixel in an image and validate these results using pre-identified, canonical colour swatches. We find that charisma colour classifications are consistent with those made by colour pattern experts in the field. Applying charisma to tanager colour evolution, we find that charisma outputs seamlessly integrate with downstream evolutionary analyses. Our results demonstrate that using charisma to manually curate and characterize colours in images provides a standardized, reliable and reproducible framework for high-throughput colour classification.