Abstract
The use of size frequency distributions of lichen populations has, in the past, been recommended as a lichenometric technique. Theoretically, saxicolous lichen populations should be density-dependent with the result that the size frequency distributions of older populations should be best described by a Poisson model. Evidence from Scotland and Norway indicates that the size frequency distributions are extremely variable, ranging from truncated log-normal to Poisson in form. Consequently, the use of the 1-in-1000 thallus, derived from a log-normal model, is not universally applicable as a descriptive index, particularly in the case of older lichen populations.