Inference on diversity from forest inventories: a review

摘要: A number of international agreements and commitments emphasize the importance appropriate monitoring protocols assessments as prerequisites for sound conservation management world’s forest ecosystems. Mandated periodic surveys, like inventories, provide a unique opportunity to identify properly satisfy natural resource information needs. Distinctively, there is an increasing need detecting diversity by means unambiguous measures. Because all measures are functions tree species abundances, estimation indices profiles inevitably performed estimating abundances then abundance estimates. This strategy can be readily implemented in framework current inventory approaches, where routinely estimated plots placed onto surveyed area accordance with probabilistic schemes. The purpose this paper assess effectiveness reviewing theoretical results from published case studies. Under uniform random sampling (URS), that when uniformly independently located on study region, consistency asymptotic normality index estimators follow standard limit theorems effort increases. In addition, variance bias reduction achieved using jackknife method. Despite its simplicity, URS may lead uneven coverage region. order avoid unbalanced sampling, use tessellation stratified (TSS) suggested. TSS involves covering region polygonal grid randomly selecting plot each polygon. TSS, consistent, asymptotically normal more precise than those URS. Variance possible no reduce bias.