paleobuddy: An R package for flexible simulations of diversification and fossil sampling

Simulations are powerful tools for investigating evolutionary questions. Statistical methods for the analysis of diversification and fossil sampling rates require extensive testing through simulation. Simulations also aid in the investigation of evolutionary dynamics that lack analytical solutions, besides model adequacy tests. Flexible simulation of fossil records and phylogenies is a necessary tool to fully investigate diversity through time. paleobuddy is an R package that aims to fill several gaps between current simulators by providing a more general framework, with flexible, robust and independent simulations of birth-death processes, fossil records and phylogenetic trees, while allowing for straightforward implementation of novel scenarios. Here we describe the main algorithms involved in paleobuddy simulations, give a summary of the capabilities of the package in terms of evolutionary scenarios and visualization and illustrate a paleobuddy workflow. paleobuddy expands the available parameter space to investigate complex inference models. Given that it allows for simulations of phylogenies (complete, reconstructed or ultrametric) and fossil records from the same underlying birth-death simulations, it can increase our understanding about the differences in palaeontological and neontological approaches, and consequently expand our capacity of analysing a combination of evidence from extinct and extant species. Finally, the focus on birth-death simulations with budding speciation events allows for more generality in the diversification processes considered, expanding the flexibility and capabilities of simulation in the field.

Automated summary

Simulations are important tools for investigating evolutionary questions, particularly in the analysis of diversification and fossil sampling rates. paleobuddy is a flexible R package that allows for simulations of birth-death processes, fossil records, and phylogenetic trees, providing a more general framework for investigating complex inference models. It expands the available parameter space and increases our understanding of the differences between palaeontological and neontological approaches.