Plant invasion community model with pos/neg ringmodel

This is the R code used to run a stochastic grid-based simulation model of plant community dynamics, used to explore how different plant–plant interaction networks affect community invasibility. The model simulates plant community dynamics with an annual time step on a square grid whose size is determined by the parameter worldsize (100 = 100 × 100 cells). Each grid cell can be empty or occupied by at most one individual plant. All cells are set to be empty at the start of a simulation, and the soil in all cells is set to be unconditioned. At each time step, the following processes occur: immigration, recruitment, growth and mortality. Immigration allows empty cells to be colonised from an external species pool at a low probability. Recruitment fills remaining empty cells either via local dispersal from neighbouring plants or via global dispersal from the entire community. Plant growth is determined by plant–soil feedbacks defined by species interaction matrices, and mortality occurs with a fixed probability, after which soil is conditioned by the deceased plant. The simulation begins by manually setting model parameters at the top of the script, including world size, dispersal type and interaction effect type. These parameters can be modified to reproduce the main analysis or conduct sensitivity analyses for the baseline model and the positive-negative ring model. Simulation outputs are written to a folder named “outputs”, which must be created in the working directory prior to running the model.