Individual-based models allow accurate prediction of introduced large herbivore populations in rewilded landscapes - agent-based model and associated scripts

Code and ABM for the paper "Individual-based models allow accurate prediction of introduced large herbivore populations in rewilded landscapes". This is split into three main sections as below: Data/current_deer_pops includes datasets and spatial information concerning the current estimated deer populations in Corsica (both min and max). The most important files here are 'Tableau_estimations populations cerf de Corse.xlsx' and 'aires repartitions cerf 2025.qgz', which contain both minimum and maximum estimated population sizes and their distribution on Corsica. Modelling/NetLogo contains both the NetLogo individual-based model and all required raster layers, the 50 min and 50 max starting estimates for population sizes, the release points for GPS tagged deer, and both population-level outputs (population sizes, number of visited patches etc) and maps for both evaluation and predictive scenarios. R_scripts contains all R scripts required for validation and evaluation of the outputs of the ABM. It includes: 1) 'sample_means' - for constructing figure S3 2) 'validation' - for performing home range and mother-offspring centroid distance validations 3) 'Initialising 2025 deer location for NetLogo' - for generating the 100 starting distributions of deer for both min and max population size estimated 4) 'Prediction mapping' - for producing the visit maps in figure 2 5) 'Prediction pop dynamics' - for calculating estimated population sizes, number of visited patches, and estimated population growth rates into the future 6) 'Splitting range expansion into north, central, and south' - for estimating the range expansion rates split by the north, central, and southern deer populations