Laboratory of Mathematical Ecology
|
RUSSIAN
Members History Current Projects Publications
|
INHOMOGENEOUS MARKOV MODELS FOR SUCCESSION OF PLANT COMMUNITIES: NEW PERSPECTIVES ON AN OLD PARADIGM
D.Î. Logofet, L.L. Golubyatnikov, and Å.À. Denisenko,1997.
Summary Succession models, designed as random Markov chains, serve as à convenient tîî1 to describe, in formal terms, the course of phytocenosis changes, since they exhibit à common mathematical property (the convergence to à stable limit distribution) that conforms to the main paradigm of succession theory (successive movement from the pioneer stages to the steady (poly)climax ones). However, time-homogeneous models (with constant values of transition probabilities) can hardly be regarded à reliable tool for long-term prognosis, as they rely on the hypothesis of environmental invariance. In the new-generation models, which are exemplified for succession occurring on spoil-banks of an open ore-mining in the forest-steppe zone, the transition probabilities are set in relation to the key factors that influence the rate of successional changes, namely, the thermic and the moistening factors. The adaptation range of vegetation is found in these factors by the time-series analysis of weather data, and the transition matrix of the model is constructed as à function of two variables defined over the range. Nontrivial scenarios of how the factors vary with time over à forecast period render the chain inhomogeneous in time and enable one to gain quantitative predictions at the same confidence level as an adopted scenario. The prospects are addressed for using such models to describe the shifts of bioclimatic zones in response to global climate changes. |