Distribution model and spatial variation of cover in grassland vegetation

Abstract

Abstract In addition to biomass and frequency of occurrence, cover is often used as a measure in vegetation surveys. However, the theoretical and statistical bases underlying cover measures are not fully understood. Here, based on actual survey data, we showed that the frequency distributions of cover in grassland vegetation measured using the point‐grid plate method were well described by the beta distribution. The beta distribution expresses various types of cover frequency distributions, such as the ‘J’, ‘L’, ‘one‐peak’, ‘U’, and ‘rectangular’ types. We defined an index that expresses the spatial heterogeneity of cover based on the beta distribution. Further, we derived a simple method to estimate cover and the spatial heterogeneity index based on the beta distribution using a point‐grid plate. In this simple method, we classified the plant cover observed in each quadrat into the following three categories: cover < c 1 ; cover between c 1 and c 2 ; and cover > c 2 (0 < c 1 < c 2 < 1). We counted the number of grid points only up to c 2 ; when the cover exceeded c 2 , counts were halted and the plant cover was assigned to cover > c 2 . This method saves time and labor. We showed that the parameter values (cover and spatial heterogeneity index) estimated from the data obtained for plant species in a grassland using the simple method agreed well with those obtained by counting all grid points covered by the plant species.