Compositional dynamics of natural forests in the Bialowieza National Park, northeastern Poland

. This paper presents results of a long-term study on natural forest dynamics in the Bialowieza National Park (BNP), northeastern Poland. Five permanent sample areas were used, each consisting of a transect of varying width (40 - 60 m) and length (200 - 1380 m). The total sample area is 14.9 ha. The study covers the period 1936–1992. During this period measurements were made on five occasions at approximately 10-yr intervals. On each measurement date all trees with DBH > 5 cm were identified and their spatial location, diameter, crown condition and position in the canopy determined. During the study period the stands underwent noticeable changes, mainly in terms of tree species composition. The major change was a quantitative increase of the late-successional species: Tilia cordata and Carpinus betulus, also to a lesser degree Fraxinus excelsior and, in the last period, of the early successional Alnus glutinosa. Declining species included both early- and late-succession species. Among the latter group, Picea abies ranked first. This species lost much of its importance during the last few decades. P. abies was followed by Pinus sylvestris which is an important component of the climax vegetation under the conditions prevailing in Bialowieza, at least on more oligotrophic sites. Still, this species has not been able to regenerate during the whole study period. Some other late-succession species, Acer platanoides and Quercus robur, were also amongst the declining species. Although the basal area of Q. robur increased, its population was getting older and the process of natural regeneration was markedly impeded. All typical pioneer, short-lived species: Betula pendula and B. pubescens, Salix caprea and Populus tremula also decreased, which was probably caused by a lack of major disturbances during the study period. In general, the results obtained for the semi-natural conditions of Bialowieza during the 56-yr observation period suggest a rather high compositional instability of the forest stands there. A more precise identification of the role of particular factors in the observed stand dynamics is difficult because of the paucity of appropriate historical and environmental data which refer directly to the study plots; moreover, the data are generally incompatible and of different resolution.