Simulating vegetation dynamics and land use in a mire landscape using a Markov model

The paper consists of a dynamical approach to mire-related vegetation and land use development. Investigations were carried out on two different study levels. First, successions of mire plant communities (community or C-level), based on stratigraphical core sample data, and, second, dynamics of plant cover and land use (PC&LU) types of mire landscapes (landscape or L-level), based on sequential aerial photographs (AP), were studied. The first reveals mainly autogenic successions of plant communities over several thousand years, the second, natural and anthropogenic successions of PC&LU types over a 30 year time span. The source data for the quantitative analysis on the basis of Markov model (MM) were provided by 17 peat types as the representatives of corresponding plant communities, taken from cores through the genetic centres of mires (C-level). Nineteen PC&LU types as the representatives of hybrid systems, were deduced from AP of three time periods (L-level). The first order MM, mainly used in ecological modelling, was also applied in addition to the C-level treatment, to a more complicated investigation, i.e. L-level. As an extension of the first order MM, second order effects were modelled. For ease of analysis of three-dimensional matrices, three states of succession (after reduction 17- and 19-element classification schemes to 3 element) only were used. The treatment represented has been used to reconstruct historical development of mire plant communities and PC&LU types in mire landscapes and analyse the future development of the corresponding types. The analysis of the first and second order models and predictions made on this basis are given.