Determining the minimal background area for species distribution models: MinBAR package

One of the crucial choices when modelling species distributions using pseudo-absences and background approaches is delineation area to fit model. We hypothesise that there a minimum around geographical centre distribution characterises well enough range environmental conditions needed by survive. Thus, fitting model within this should be optimal solution in terms both quality execution time. MinBAR an R package calculates means sequentially several concentric models (SDMs) until satisfactory included metrics reached. The version 1.1.2 implemented for MaxEnt (using either maxnet or original java program) uses Boyce Index as metric assess performance. Three case studies are presented test hypothesis package's functionality. show how partial trained with part often perform equal better than those fitted on entire extent. versatile tool helps modellers objectively define solution.