| Resumo: | The assessment of cork quality and the estimation of cork value are very important to forest landowners, for management purposes and for cork commercialisation. The Forest Producers Associations have been using a sampling scheme with the objective of estimating cork value (price per unit of weight, usually kg) before extraction, based on the sampling of individual trees along a zigzag transect that covers the entire stand. The sampling error is usually too high, but, from a practical standpoint, it is difficult to increase the sampling intensity if it would imply an increase in sampling costs. The aim of this work was to propose, from data collected in six stands representative of the cork oak stands in Portugal, an alternative sampling methodology with a more efficient precision/ cost ratio. Precision and costs of alternative sampling designs based on clusters of different sizes, complemented with analysis of the intracluster correlation coefficient, were studied in order to propose the most adequate sampling strategy. Single-stage cluster sampling with clusters of 5–7 trees guarantees a reasonable sampling error (10–15%) and can be conducted without a large increase in cost
|
|---|