Markov models of forest-type secondary succession in coastal British Columbia

Publisher Website
Google Scholar
Abstract
Sample plots of five biogeoclimatic subzones of the British Columbia coastal forest were classified into forest types at 5-year intervals during 40 years. Matrices of transition probabilities from one forest type to another were calculated and stationary Markov models were built for each subzone. Simulation runs were tested against observations. The time and composition of steady state were predicted to evaluate the likelihood of the models. Succession and behaviour of forest types were discussed within and between subzones. It was concluded that forest succession was not Markovian for reasons that depended on the data set and on properties of stationary Markov models. The future of a forest type cannot be determined solely on the basis of its present state, and its prediction necessitates a sound knowledge of how the forest got there in the first place.

This publication has 3 references indexed in Scilit: