Main Article Content
Miombo woodlands cover large areas in Tanzania but very little reliable data on forest dynamics for the woodlands exist. The main objective of this study was to develop a model system describing such dynamics based on easily measurable tree variables. Individual tree diameter growth and mortality models, and area-based recruitment models were developed. The modelling was based on data from 117 permanent sample plots established between 1997 and 2009 on seven sites in the Iringa and Manyara regions. The data set comprised 2 314 trees at establishment and 4 758 observations of individual tree growth, and covered a large number of species (141). Based on diameter growth and by using a model-based clustering procedure, three tree species groups were developed. Several model forms with different independent variables were tested. Simple and robust models with few independent variables, which agreed with expected biological behaviour, were finally selected. Although biological behaviour was the principal model selection criterion, the models also performed well statistically. The developed model system, which is the first of its kind for miombo woodlands in Africa, may be used to develop a decision support tool for practical forest management planning in the country.
Keywords: empirical models, forest dynamics, permanent sample plots, tree species grouping