Data
The American black bear Ursus americanus was arbitrarily chosen
to illustrate the method because it was expected to have a large,
well-sampled range. Data were drawn from the Global Biodiversity
Information Facility (GBIF) using the occ_data function in R
library rgbif (Chamberlain et al. 2020) on 2 September 2020.
After processing the data with the st_polygon and st_sfcfunctions in the R library sf (Pebesma 2018), convex hull areas
were computed using st_area .
Monte Carlo simulations were used to compare the four methods. Data were
placed randomly within ranges having simple geometrical shapes. Four
sets of simulations were implemented. (1) Comparisons of all four
methods using five points randomly drawn from within a circle in each
trial. (2) Comparisons with 20 points drawn. (3) Comparisons with 20
points drawn from a circle and 80% of the data points in the right-hand
side of the shape randomly removed. (4) Trials using six different
shapes, with 10 points being drawn during each trial and only the mosaic
area method being applied in most cases.