Plant cover is a measure of the two-dimensional projection of plant species onto the surface and is the most commonly used measure of plant abundance. There are numerous historic and current plant ecological investigations and monitoring programs where plant cover has been measured by visual estimation in small plots or using the pin-point method.
The collected plant cover data has often been analyzed with standard statistical techniques without taking the typically observed spatial aggregation of plant species into account. This practice has been continued within the plant ecological community although a more suited statistical method using the beta distribution has been suggested in the literature for more than sixty years.
In our essay review we advocate the use of the beta distribution for analyzing plant cover data and suggest relevant R packages for analyzing different types of plant cover data.
Among other features it is demonstrated how:
- one can best model visually estimated ordinal plant cover data without using the midpoint values of the ordinal intervals.
- it becomes possible to study spatial aggregation as an ecological phenomenon by analyzing plot based plant cover data by estimating a parameter that represents the degree of spatial aggregation.
- different methods of measuring plant cover may be analyzed in a common beta distribution framework in order to synthesize information across studies or time-series.
Christian Damgaard (Aarhus University, Denmark) and Kathryn Irvine (U.S. Geological Survey, USA)
Read the full paper on the journal website: Using the beta distribution to analyse plant cover data.