The aim of this study was to evaluate the prognosis of the diameter structure of an experimentally managed forest, utilizing a system of differential equations of the first order, adjusted for each treatment. The basic data were obtained from the Bom Manejo Projet (Embrapa Eastern Amazon/CIFOR/ITTO), carried out in the Monte Dourado Forest Company (Jarí) area, Vitória do Jari, Amapá, Brazil. The statistical delineation was designed as random blocks with three replications. The treatments applied were combinations of logging intensities (15%, 25% and 35% of the total volume of the trees larger than 60 cm of DBH) and thinning intensities (0%, 30%, 50% and 70% of the original basal area). The non-logged forest was used for the control plots. Stand development was monitored through 40 permanent sample plots of 1 ha each (36 logged and 4 unlogged). The stand was measured in 1984, logged in 1985 and remeasured in 1986, 1988, 1990, 1994, 1996 and 2004. All the trees with DBH ≥ 20 cm were measured. While keeping the same simplicity of transition matrix based models, the proposed model allows for determining the state of the system for any given point in time, not being confined just for multiplying time intervals as was the former. This method was verified for different evaluation periods and the results showed that not even the longer periods of projections (break-even point) will be enough to restore the initial conditions of the stand.
Forest management; diametric distribution; growth models; forest dynamics; Amazon
