1. University of Natural Resources and Applied Life Sciences, Vienna Department of Applied Plant Sciences and Plant Biotechnology Universität für Bodenkultur Wien Department für Angewandte Pflanzen- wissenschaften und Pflanzenbiotechnologie Yield estimation at poplar in short rotation coppice in Austria (preliminary results) C. Zeitlhofer, E. Hochbichler, J. Schweinberger & P. Liebhard Introduction and Aims: In Austria, energy wood production in short rotation coppice systems (SRC) becomes increasingly important to meet the demands of the growing bioenergy sector. In order to successfully develop the SRC market, the achievement of high and constant yields in SRC management is just as important as a reliable harvesting technology, which facilitates the production of high quality wood chips. Yield models are needed to estimate biomass increment with respect to clones, site factors and management alternatives. Single shoot surveys and biomass functions in combination with stand inventories form the general basis for estimating yield and productivity. They also help to optimize yield and rotation length, while taking the maximum harvestable tree diameter, which is determined by harvesting technique, into account. At present, the use of common harvesting technologies is limited by the tree diameter at cutting height (diameter at tree/stool height of 10cm - D 10 ). Trees/stools with D 10 of more than 14 cm cannot be harvested with available forage harvesters (Fig. 1). For optimizing the yield estimation of SRC stands, we developed preliminary clone specific yield functions for poplar clones, which are based on common yield estimation functions, but also take the characteristics of the newly used clones (e.g. faster growth, lower wood density) into account. Material and Methods: The experimental site is situated in Austria in the east of Vienna. In spring 2007, eight poplar clones were planted out in three replications, resembling a fully randomized block design. The size of the entire experimental plot is 1,07 ha. The trees were planted in single rows with a distance of three meters between rows and 0,58 meters between trees within the rows (Fig.2). Each plot consists of three single rows with a length of 20 meters. Standard yield measurements were carried out in December 2007 and November 2008. The experimental plot was harvested in November 2008 with a „Claas Jaguar Forage Harvester“ in a single-phase process. Before mechanized harvesting, 12 trees of each clone, representing the range of D 10 were used for a destructive sample inventory. The allometric equation of the biomass function: ln dm (kg) = [a + b * ln D 10 (cm)] * C [a,b: regression coeff.; corr. element: C = e^(SEE²/2) after SPRUGEL 1983] was used for predicting dry matter (dm) of tree biomass (dried to constant weight at 105° C). The above ground biomass of each plot (kg per hectare) was then calculated by extrapolation with the number of trees of each plot (number of trees per hectare), and related to the D 10. Summary: With adequate yield models and inventory techniques specific to clones, site factors, field areas and management alternatives (planting design, rotation period) we can support decisions for optimizing SRC-management. The adapted biomass functions should take the specific weight of different poplar clones into account. SRC yield models are very helpful to predict e.g. the entire harvestable biomass, to optimize harvesting technology, date of harvest and to plan the required storage and transportation capacities. Results and Discussion: The common biomass function, which is available for poplar planted in the 1980s in Austria, does not describe the biomass increment of most newly used poplar clones sufficiently. Because of their fast growth, their wood density is substantially lower than that of common varieties. Therefore, the yield function needed to be adjusted to the new clones. Thus, it was adapted to eight different poplar clones from the experiment in Austria. With the preliminary adjusted clone specific yield functions, estimated yields only differ about plus minus 6,4 percent from observed yields. Figure 3 shows the goodness of fit of both, the common and the clone specific yield function. Figure 4 shows the difference between the common yield function and the clone specific ones. Again, we emphasize, that the proposed clone specific yield functions are preliminary and require further improvement. Fig. 1: Claas Jaguar forage harvester Fig. 3: Preliminary biomass functions for various „new“ clones and the average of all „new“ clones (BF allg. 2008) in comparison with the common biomass function (BF allg. 2007) Fig. 2: Typical single row planting design used for poplar SRC Fig. 4: Comparison of biomass estimation functions and actually harvested biomass LIEBHARD_BENWOOD_Beijing 2010