This repository is used to give an example of how to perform analysis of the elements of metacommunity structure, followed by constrained correspondence analysis. I used this method to infer how metacommunities form along environmental gradients. Main R packages showcased here include metacom, and various vegan tools.
The R scripts were created for tutoring colleagues. It is currently not reproducible as the datasets are not included in this repository.
The 2 datasets used were provided by Tiago Teixeira, which consists of frequencies of bat species across 13 sites in the Atlantic Forest in the form of a presence/absence matrix, and data on landscape properties of the 13 sites for the subsequent CCA analysis.
This tutorial has 2 parts: "EMS_Basics_OKR.R", and "CCA_Basics_OKR.R" intended to be viewed in that order.
The R script "EMS_Basics_OKR.R" includes a series of commands to:
- Prepare the environment
- Load and prepare the data set "Bats_and_Sites.csv" (not included)
- Perform the EMS calculation with the R function
metacom::Metacommunity - Generate a visualisation of the ordinated matrix using
metacom::Imagine(available in plot folder)
You may then follow-up with R script "CCA_Basics_OKR.R" to:
- Fetch the ordinated matrix
- Select environmental variables from "Landscape_var.csv" (not included), create CCA model
- Interpret CCA: Inertia, eigenvalues
- Interpret CCA: biplot (available in plot folder)
No longer updating this, thanks for visiting. -OKR