Comparison of four inventory designs for estimating yields from the thinning of pine plantations

Abstract

Four inventory designs were used to predict the yields of commercial products from a second thinning of a 21-year-old Pinus radiata plantation: (I) standard, single stage inventory with visual assessment of products and volumes derived from taper tables; (2) model based inventory with visual assessment of products and volumes measured by Telerelaskop; (3) model based inventory with visual assessment of products and volumes measured by Spiegel Relaskop; (4) single stage inventory with computer modelling of both thinning and utilisation and volumes derived from taper tables. Both fixed area and variable radius (point sample) plots were tested with each design.Most accurate results were achieved with designs I and 2 but the model based inventory was quicker and cheaper to implement because assessment of products was confined to a small sub-sample of trees. The Spiegel Relaskop underestimated volumes in the small diameter classes, whilst the modelling inventory gave the least accurate estimates due mainly to errors in the thinning model. Inventories using fixed-area plots were more accurate than those based on variable-radius plots but equally precise (i.e. had the same sampling errors). As more trees were measured on the fixed-area plots, they were less efficient than variable-radius plots.Model-based sampling may be the most suitable method for estimating the yields of products from thinning of pine plantations.Simulation of thinning operations using computer modelling systems such as Pars may be preferred where different management options are to be evaluated but the accuracy of the models must be monitored against field practice.