The estimation of the site productivity of forest stands is a major facet of the study of growth and yield in the forest sciences. Initial approaches were formulated by ASSMANN (1961) and expanded upon by FRANZ (1968, 1971). To date, reliable site-based yield modelling is possible only in relatively simply structured forests such as pure, single-layered stands and in short rotation coppice stands (Fig. 1, Fig. 2). Tree species with the capacity to coppice and with rapid early growth such us poplar, willow and black locust are suitable foremost for cultivation in short rotation plantations. Depending on site conditions, the mean annual biomass increment ranges from 4 to more than 20 tonnes dry bulk ha(-1) year(-1) for poplar. Optimal growth is achieved at high stocking densities on soils with good water availability (Fig. 3). A two-tiered model was applied for site-based estimations of poplar production in short rotation coppice. First the dominant height was estimated in a multiple linear regression analysis in relation to stand age and the following four site factors (Fig. 4): available water storage capacity (AWSC), soil quality index (SQI), mean temperature between April and July (T4-7) and total precipitation between May and June (P5-6). Subsequently the forest biomass was determined in a non-linear regression analysis of the dominant stand height. Simulation calculations showed that, for a given stand density and a rotation of 9 years, a mean annual biomass increment of more than 8 tonnes dry bulk ha(-1). year(-1) can be achieved in the first rotation on all agricultural land in the German state Saxony (Fig. 5). The yield estimator for short rotation poplar plantations was developed based on the modelling of coefficients from allometric biomass functions (Fig. 6) relative to mean stand height [m] and stocking density (shoot number ha(-1)) (Fig. 7, Fig. 8, Fig. 9). Use of the yield estimator, i.e. to calculate the biomass of a short rotation coppice plantation, requires only data pertaining to the diameter, height and the shoot number per hectare. Preliminary studies to develop a yield estimator for short rotation willow plantations have shown the approach adopted for poplar to be unsuitable and that, instead of coefficient modelling, alternatives such as the extrapolation of diameter distributions using the Weibull function linked with biomass functions should be investigated.
