EcoCoMix (Compositional Mixed Model in Ecology) addresses compositional autocorrelation in community analyses.
The logo was created using DeepAI.
Installation
You can install the development version of EcoCoMix from GitHub with:
# install.packages("devtools")
devtools::install_github("tpaknok/EcoCoMix")Main function
The main function of this package is EcoCoMix, which address compoisitional autocorrelation in regression. Briefly, the function automatically constructs the compositional (after considering phylogenetic covariance between species if provided) matrix between communities based on their composition and the evolutionary history of the species.
Background
In many ecological analyses, we obtained different metrics at community levels and analyzed them using LM/GLM. Nevertheless, this assumes statistical independence between all communities. Species however is not independent, given their shared evolutionary history. Indeed, in many comparative analyses (e.g., correlating different traits), the shared evolutionary history between species has to be controlled using Generalized Least Square Regression or Generalized Linear Mixed Model.
The functions in EcoCoMix help users to control phylogenetic non-independence between communities without intensive coding. Users only need 1) species composition in each community and 2) a phylogenetic tree including all species in the community data.