NESTED REGRESSION FOR ESTABLISHING TREE BIOMASS EQUATIONS
嵌套式回归建立树木生物量模型

Aims The objective was to introduce a simplified method called nested regression for quick establishment of allometric models for tree biomass. Methods Main branches at all orders were used as basic sample units. The biomass of a branch was obtained by summing that of its sub-branches or sub-main branches, either observed or estimated. Thus the span of branch diameter for the equation was extended. The branch was dissected into main braches, and a main branch is the axis of a branch at any order. Biomass equations were established for main branches, branches and individual trees, progressively. By summing the values of main braches, the biomass of a branch in a higher order was obtained. Important findings Because of data nesting, i.e., small branches were encompassed in their upper-order branches, the measured sample was reduced to a minimum, making the method time- and cost-effective. Satisfactory accuracy of predicted values was shown by testing with observed data.