The ecospat R Package: A Collection of Pre-, Core-, and Post-Modeling Tools to Investigate Species Niches and Distributions

The R package ecospat is a collection of R functions and datasets designed to support spatial ecology analyses, with a focus on pre-, core-, and post-modeling analyses of species distributions (SDMs). Pre-modeling analyses include data extractions, designing sampling strategies, sub-sampling data, assessing spatial autocorrelation, reducing collinearity among predictors, measuring environmental similarity and analog/non analog situations, assessing niche stability within a species or niche difference among species, and measuring species cooccurrences to investigate community assembly. Core-modeling analyses cover modeling fitting, spatial predictions and projections, and model evaluation.
It includes in particular the fitting of SDMs with the “ensemble of small models” (ESMs) approach, which applies especially to species with small sample sizes (e.g., rare and endangered species). The post-modelling analyses comprise variance partitioning, determination of occupied patches and range size using the IUCN criteria, SDM prediction map binarization, and predictions of community properties from stacking individual species, including also functional and phylogenetic diversity. In this book section, we first illustrate two study cases where the ecospat package can be used: (a) modeling rare species distribution and quantifying species range size and (b) anticipating biological invasions via niche comparison. We then discuss other applications of the functions in the ecospat package and provide some perspectives about the future development of the package.