Rectangular and hexagonal grids used for observation, experiment and simulation in ecology

Abstract Regular grids or lattices are frequently used to study ecosystems, for observations, experiments and simulations. The regular rectangular or square grid is used more often than the hexagonal grid, but their relative merits have been little discussed. Here we compare rectangular and hexagonal grids for ecological applications. We focus on the reasons some researchers have preferred hexagonal grids and methods to facilitate the use of hexagonal grids. We consider modelling and other applications, including the role of nearest neighbourhood in experimental design, the representation of connectivity in maps, and a new method for performing field surveys using hexagonal grids, which was demonstrated on montane heath vegetation. The rectangular grid is generally preferred because of its symmetrical, orthogonal co-ordinate system and the frequent use of rasters from Geographic Information Systems. Cells in a rectangular grid can also easily be combined to produce new grids with lower resolutions. However, efficient co-ordinate systems and multi-resolution partitions using the hexagonal grid are available. The nearest neighbourhood in a hexagonal grid is simpler and less ambiguous than in a rectangular grid. When nearest neighbourhood, movement paths or connectivity are important, the rectangular grid may not be suitable. We also investigate important differences between visualizations using hexagonal and rectangular grids. A survey of recent uses of grids in Ecological Modelling suggested that hexagonal grids are rarely used, even in applications for which they are more suitable than rectangular grids, e.g. connectivity and movement paths. Researchers should consider their choice of grid at an early stage in project development, and authors should explain the reasons for their choices.