Minimum grid-cell size based on home-range connectivity

Grid-based home-range estimates depend on the spatial resolution of the grid. When cells are too small relative to the spatial distribution of animal locations, the resulting home range may consist of several disconnected components. min_cell() evaluates this issue by calculating the topological connectivity of the occupied grid cells across a sequence of cell areas.

For each tested cell size, animal locations are overlaid on a regular grid and cells containing at least one location are retained as the grid-based home range. Connectivity is then evaluated according to shared cell edges: cells that touch only at a corner are not considered connected. Connectivity is expressed as the proportion of occupied cells belonging to the largest connected component.

The function identifies the smallest tested cell area at which all occupied cells form a single connected component. This provides a way to evaluate whether a chosen grid resolution gives a spatially coherent representation of the observed home range while retaining as much spatial resolution as possible.

library(gridHR)

min_cell(
  gibbons,
  min = 1,
  max = 1000,
  interval = 100,
  cell_shape = "hex",
  show_plot = TRUE
)

The connectivity profile shows that all occupied cells form a single connected component at a hexagonal cell area of 250 square meters. Thus, even at this relatively fine spatial resolution, the observed locations produce a fully connected representation of the home range.