Relationships between landcover proportion and indices of landscape spatial pattern

Recent studies have related percolation theory and critical phenomena to the spatial pattern of landscapes. We generated simulated landscapes of forest and non-forest landcover to investigate the relationship between the proportion of forest (P_i) and indices of patch spatial pattern. One set of landscapes was generated by randomly assigning each pixel independently of other pixels, and a second set was generated by randomly assigning rectilinear clumps of pixels. Indices of spatial pattern were calculated and plotted against P_i. The random-clump landscapes were also compared with real agricultural landscapes. The results support the use of percolation models as neutral models in landscape ecology, and the performance of the indices studied with these neutral models can be used to help interpret those indices calculated for real landscapes.