Abstract
Integration of conservation into sustainable forest management plans requires the accurate assessment of pre-harvesting biodiversity. An open question is how to efficiently and accurately sample biodiversity in production forests. Here we investigate one possible answer: the use of existing pre-felling inventories designed for the estimation of timber stocking. Using data drawn from two large permanent forest sample plots in Malaysia, we tested the ability of the Malaysian SMS pre-felling inventory (Pre-F) to accurately assess alpha (α)- and beta (β)-tree diversity. We found that the Pre-F did not accurately estimate variation in α-diversity at small spatial-scales, but could be used to accurately extrapolate α-diversity on larger spatial scales. The Pre-F also adequately assessed β-diversity in heterogeneous environments, but did not perform as well in homogeneous environments. The use of tree species codes as opposed to Latin nomenclature affected estimates of β- but not α-diversity.