Simulation of tree and stand development under different environmental conditions with a physiologically based model

A simulation approach is used to describe annual tree growth and tree mortality from the output of a physiologically based model (FORSANA). Height and diameter growth are calculated directly from the amount of carbon allocated to sapwood by considering an optimum height/diameter ratio, which depends on stand density. Tree mortality is defined by means of a relation between net primary production and carbon loss due to compartment senescence. Thus, all responses to environmental conditions considered in the physiological part of the model are implicitly considered in the stand development description. The dynamic simulation of stand properties, on the other hand, is required to apply the physiological based process description to long-term assessments. The model is used to describe height and diameter development of three Scots pine (Pinus sylvestris L.) stands in eastern Germany which are exposed to different levels of nitrogen deposition and SO2 air pollution. Results are compared with tree ring analysis covering a period of 27 years. For further evaluation, the model is initialised with forest inventory data of 288 pine stands and is run over 23 years using daily weather and deposition data as well as fertilisation information as input. The results ate compared to data from a second inventory of the same stands. This comparison is conducted separately for regions exposed to high and low deposition. The model represents annual height and diameter development at two of the three selected sites. With respect to the third site, considerable disturbances in the early years of stand development are assumed to be responsible for the unusual growth trend. The regional evaluation of the model yields correlation coefficients with forest inventory data between 0.57 and 0.86, with a generally better fit on diameter and stemwood volume than height. The approach demonstrates the uncertainty of estimations which are based on investigations at only few sites, and is discussed as a possible method for regional assessment of forest development under environmental change. (C) 1999 Elsevier Science B.V. All rights reserved.