Abstract
AbstractEstimating underlying cooccurrence relationships between pairs of species has long been a challenging task in ecology as the extent to which species actually cooccur is partially dependent on their prevalences. While recent work has taken large steps towards solving this problem, the next question is how to assess the factors that influence cooccurrence. Here I show that a recently proposed cooccurrence metric can be improved upon by assigning Bayesian priors to the latent cooccurrence relationships being estimated. In the context of analysing the factors that affect cooccurrence relationships, I demonstrate the need for a generalised linear model (GLM) that takes cooccurrences and species prevalences (not cooccurrence metrics) as its data. Finally, I show the form that such a GLM should take in order to perform Bayesian inference while accounting for non-independence of dyadic matrix data (e.g. distance and cooccurrence matrices).
Publisher
Cold Spring Harbor Laboratory
Reference20 articles.
1. “Using Network Analysis to Explore Co-Occurrence Patterns in Soil Microbial Communities;The ISME Journal,2012
2. Julia: A Fresh Approach to Numerical Computing
3. “Confidence Limits for Regression Relationships Between Distance Matrices: Estimating Gene Flow with Distance;Journal of Agricultural, Biological, and Environmental Statistics,2002
4. Makie.jl: Flexible high-performance data visualization for Julia
5. Ge, Hong , Kai Xu , and Zoubin Ghahramani . 2018. “Turing: A Language for Flexible Probabilistic Inference.” In International Conference on Artificial Intelligence and Statistics, AISTATS 2018, 9-11 April 2018, Playa Blanca, Lanzarote, Canary Islands, Spain, 1682–90. http://proceedings.mlr.press/v84/ge18b.html.