Markov Models and Succession: A Test from a Heathland in the Netherlands

Abstract

(1) In the search for models of vegetation succession, Markov models have received much attention. This work has been largely theoretical because of a lack of experimental data. (2) A Markov chain model is examined for an Empetrum nigrum heath in The Netherlands, based upon observations over 19 years. (3) The assumption that the transition probabilities are stationary does not hold for these data. The transition matrix is affected by: (i) identifiable influences due to climate and insect plagues; (ii) directional changes of the transition probabilities with time; and (iii) spatial influences. (4) Nevertheless, in certain circumstances, transition matrices can be useful descriptive devices even when the system is not Markovian.