No abstract

Summary1. Fires are nowadays small, yet frequent, in temperate Central European conifer forests, but little is known about the fire history in this region. This is likely due to the lack of intact forests that contain old trees and dead wood from which fire history may be reconstructed. An exception is the Białowie_ za Primeval Forest (BPF) in Poland for which we were able to reconstruct the fire history in detail. 2. From 886 tree ring samples collected in a 13-ha conifer-dominated area, we reconstructed fire events and tree establishment back to the mid 1600s. From 1653 to the late 1700s fires were very frequent, with mean point (single tree) fire intervals of 18 years and mean stand scale fire intervals of 6 years. After 1781, the intervals between the fires increased dramatically, and since 1874 no major fire was recorded. 3. Tree establishment underwent substantial changes, closely tracking shifts in fire frequency. When fires were frequent, Pinus sylvestris establishment occurred only sporadically. Later, less frequent fires promoted massive establishment of both P. sylvestris and Picea abies. At present, only P. abies and a few deciduous trees regenerate. 4. Synthesis. We present the first high-resolution fire history in the Central European temperate lowland forest area. The discovery of old P. sylvestris trees and stumps with fire scars in many conifer-dominated parts of BPF show that fire was a major component in the past dynamics of this forest. We also show that historically, fires were recurring at very close intervals, supporting an open, Pinus-dominated forest. These result contrasts with the written history of BPF, which focus on a few, large fires from the past. Human influence on the fire regime was probably substantial, although the disentangling of climatic and human impacts needs further studies. We propose that fire should be increasingly taken into consideration in models of disturbance, vegetation development and forest openness in the whole Central European lowland forest region.

Question: What are the main driving factors in 70 years of natural dynamics in tree recruitment in the Biazowiez 'a National Park? Location: Biazowiez 'a National Park, Poland, is one of the least disturbed temperate, lowland forest systems in Europe. Methods: We tested whether fluctuations in large herbivore populations, changes in climate and openness of the forest explained compositional dynamics. Tree recruitment (to size class DBH ! 5 cm) was measured on permanent transects (in total, 14.9 ha) six times between 1936-2002. These data were related to existing data on ungulate density, climatic parameters and estimates of forest openness collected during the same period. Results: Total recruitment of all tree species combined was negatively correlated with total ungulate density and red deer density. The variation in response between species was related to the preferences of herbivores; the more preferred forage species (especially Carpinus betulus) were positively and the less preferred species negatively related to herbivore density. Total tree recruitment rates were not related to climatic parameters and openness of the forest. Only Alnus glutinosa recruitment was significantly related to climatic parameters, and Ulmus glabra related to forest openness, but there were no predictable patterns in recruitment among species in relation to these factors. Conclusion: The present study indicated that changes in large herbivore density have played an important role in driving patterns in tree recruitment and species composition during the last 70 years in Biazowiez 'a National Park. In contrast to other studies, increasing herbivore numbers were associated with higher recruitment of preferred and browsing-tolerant species. Periodical crashes in ungulate numbers, whether human-induced or caused by natural factors, may offer windows of opportunity for regeneration of a range of tree species and facilitate more diverse and dynamic forest development.

Using empirical data (ca. 7500 phytosociological releves), a simple, probabilistic ‘vegetation‐site’ model was developed, to simulate geographical distribution of 71 forest community types, representing the potential natural vegetation (PNV) of Switzerland. The model was interfaced to a geographic information system (GIS) and used to generate a numerical vegetation map, on the basis of digital maps of 12 environmental variables including climatic conditions (temperature and precipitation), topography (elevation, slope, aspect), and soil parameters (soil pH and physical soil parameters). The predicted distribution of forest communities was compared with several vegetation maps, prepared for some subregions of Switzerland by means of traditional field methods. Similarity ranged from 50 to 80 %, depending on the community type, level of vegetational hierarchy and the geographical region. The current resolution and accuracy of the simulated vegetation map allows us to study the vegetational patterns on the level of the entire country or its major geographical and climatic regions. The simulated vegetation map is potentially an important tool in ecological risk assessment studies concerning the possible impacts of climate change on the ecological potential of forest sites and biological diversity of forest communities.

Questions: Does the size structure characterizing particular tree species in natural forest stands converge towards a steady state? What is the main reason for departures of observed tree size distributions from equilibrium models? What is the consequence of long-term strict protection for the demography of tree species and the conservation of associated biota?Location: Stands of natural forest in Białowie_ za National Park (NE Poland) that have been under strict protection for about 90 yr.Methods: Data on recruitment, growth and mortality of approximately 10 000 trees representing 11 species (aspen, birch, alder, pine, oak, maple, ash, elm, spruce, lime and hornbeam), from seven replicated surveys of five permanent monitoring plots (total size = 15.44 ha), established in 1936, were used to derive theoretical, species-specific equilibrium distributions as a basis for evaluating the demographic changes of the species involved.Results: Over a monitoring period of 76 yr, there were strong downward trends in the population densities of several tree species of different successional status and light requirements. A vast majority of the conspecific equilibrium curves, modelled on the basis of growth and mortality data, revealed a 'rotated sigmoid' shape when plotted in semi-log plots. In contrast to these equilibrium curves, the observed tree size distributions of most species were unimodal, at both the beginning and the end of the study period. Departures between theoretical and observed distributions increased over time, particularly for shadeintolerant and intermediate species.Conclusions: The populations of several tree species occurring in the Białowie_ za National Park are demographically unstable -a fact reflected in large and steadily increasing differences between theoretical and actual tree size distributions. The main reason for this are the insufficient recruitment rates typical of most species. Some tree species may very soon disappear almost entirely, and this will certainly lead to far-reaching changes in the functioning of local forest communities. Among other consequences, the current demographic trends, as observed for Białowie_ za tree species, are very likely to have a strong effect on the long-term survival of numerous forest organisms (such as insects, fungi, lichens, etc.), which depend, directly or indirectly, on the presence and balanced demography of particular tree species.