Enhancing a permanent sample plot system in natural forests. The Statistician.pdf

Postprint of: The Statistician (1992) 41, pp. 525-538 Enhancing a permanent sample plot system in natural forests TREVOR BEETSON,1 MARKS NESTER1 and JEROME VANCLAY2 1 Queensland Forest Service, M.S. 483, Gympie QLD 4570, Australia 2 Faculty of Forestry, Royal Veterinary and Agricultural University, Thorvaldsensvej 57, DK-1871 Frederiksberg C, Denmark Abstract. Permanent sample plots (PSPs) provide the basis for growth models and thus determine the quality of inferences made from decision support systems for forest management. Existing plots may have been adequate in the past, but changing needs and possibilities may have made data suboptimal for present demands. Funds for the establishment of additional plots are generally limited, so supplementary sampling should be based on an optimal sampling strategy employing prior information. This case study in south-east Queensland used topoclimatic strata and existing inventory data to identify where additional PSPs should be established, so as to increase the geographical and silvicultural range of the growth data collected. Some existing PSPs in areas considered over- sampled may be abandoned. Introduction Decision making by natural resource managers has been and continues to be impaired by deficiencies in the quality and extent of relevant information. This paucity of information on natural resources restricts the ability of management agencies to assess the impacts of alternative management strategies. One important objective for forest managers is sustainable timber production which requires reliable yield estimates and the ability to predict the future nature of the forest resource under alternative management strategies. Essential components for such estimates include the area information, resource inventory, a growth model and volume equations (Vanclay, 1989, Vanclay and Preston, 1989). The integrity of the growth model is dependent largely upon the quality of the data from permanent sample plot