pycoalescence and rcoalescence: Packages for simulating spatially explicit neutral models of biodiversity

Abstract

A priority for community ecology is to understand the mechanisms determining species composition (Agrawal et al., 2007; Vellend, 2010) and unravel the interplay between ecology, biogeography and evolution. One theoretical approach is to generate biodiversity patterns using theoretical models and compare the patterns to empirical data. The models incorporate processes deemed important from real-world observations or prior theoretical studies. This interplay of theoretical and empirical work helps reveal the elusive links between ecological patterns and their mechanisms. Theoretical models can be studied mathematically or via computer simulations. In this paper, we focus on the latter. Neutral theory represents one notable branch of ecological theory (Hubbell, 2001; Rosindell, Hubbell, & Etienne, 2011). It captures three key mechanisms (dispersal, drift and speciation; Vellend, 2010) in a species-agnostic manner, meaning that species identities do not influence the dynamics. Numerous neutral models have been used

Early tetrapod diversification under neutral theory

“…Unfortunately, almost all non-neutral models cannot benefit from the use of coalescence and associated abilities to account for sampling in huge spatially explicit systems. We use the pycoalescence package available for Python and R (Thompson et al , 2020), which uses coalescence methods implemented in C++ for high performance spatially explicit neutral simulations. All simulations were performed on high-throughput computing systems at Imperial College London.…”

“…Under coalescence methods, higher speciation rates can be applied post-simulation for generating communities (Rosindell et al . 2008; Thompson et al . 2020).…”

“…We explored 5 density and 5 dispersal parameters giving 25 combinations using Latin hypercube sampling (McKay et al 1979) to evenly sample from arithmetic parameter space. Under coalescence methods, higher speciation rates can be applied post-simulation for generating communities (Rosindell et al 2008;Thompson et al 2020). We performed simulations using a minimum speciation rate of and applied all other speciation rates afterwards to generate additional communities.…”

Adding stage‐structure to a spatial neutral model: implications for explaining local and regional patterns of biodiversity

“…We coded two versions of the zero‐sum simulation in R using different algorithms: one that runs forwards in time, and one that runs backwards in time using coalescence methods (Rosindell et al 2008, Thompson et al 2020). The backwards‐in‐time algorithm has two major advantages: it runs much faster than the forwards‐in‐time version; and it can run on effectively infinite landscapes (limited only by numerical precision) consistent with our original model definition above.…”

Biogeographic Drivers of Evolutionary Radiations

“…Any 'clock' is initialized with τ generations (here 100), and will then count down over time but increase by G whenever a dispersal event of a conspecific occurs (gene flow). Thompson et al, 2020). This is because a model with multiple habitat patches and protracted speciation with gene flow can only be approximated with coalescence (Rosindell and Phillimore, 2011;Gascuel et al, 2016).…”