Key message

Models for quantifying tree biometric properties, imperative for forest management decision-making, including height, diameter, bark thickness and volume were developed, and wood basic density was documented for dry Afromontane forests of south-central Ethiopia.

Abstract

Tree biometric properties such as height (ht), diameter at breast height (dbh), bark thickness (bt), volume and wood basic density (wbd) are imperative for forest management decision-making. For dry Afromontane forests in south-central Ethiopia, models for quantifying such tree properties are totally lacking. This study, therefore, aimed at developing models for ht based on dbh, for dbh based on stump height diameter (dsh), for bt based on dbh, for volume based on dbh, ht and crown width (crw), as well as documenting wbd data. Comprehensive and representative datasets were collected from Degaga–Gambo and Wondo Genet forests. The ht, dbh and bt modelling were based on 1345 sampled trees during forest inventories, while the volume modelling and wbd documentation were based on 63 destructively sampled trees from 30 species covering 87% of the total basal area in the study sites. Weighted least squares regression was applied for modelling and leave one out cross-validation was used for evaluation. The ht–dbh and dbh–dsh models performed well (pseudo-R² = 0.72 and 0.98), while bt–dbh performed poorer (pseudo-R² = 0.42). Models for the total tree, merchantable stem and branches volume were developed with different options for independent variables, where pseudo-R² varied from 0.74 to 0.98, with smallest values for the branches models The models may be applied to forests outside the present study sites provided that the growing conditions are carefully evaluated. The species-wise wbd was ranging from 0.426 to 0.979 g cm⁻³, with the overall mean of 0.588 g cm⁻³. The wbd data will be useful for building up a national wbd database and may also be included in the Global Wood Density database. The study represents a significant step towards sustainable forest management including REDD + MRV practices in the dry Afromontane forests of south-central Ethiopia.

中文翻译：

---

建模和量化埃塞俄比亚中南部干旱紫檀林的树木生物特征

关键信息

开发了用于量化树木生物特征的模型，这对于森林管理决策是必不可少的，包括高度，直径，树皮厚度和体积，并记录了埃塞俄比亚中南部干旱山毛榉森林的木材基本密度。

抽象

树木的生物特征，例如身高（ht），胸径（dbh），树皮厚度（bt），体积和木材基本密度（wbd），对于森林经营决策至关重要。对于埃塞俄比亚中南部的干燥Afromontane森林，完全缺乏用于量化此类树木特性的模型。因此，本研究旨在开发基于dbh的ht，基于树桩高度直径（dsh）的dbh ，基于dbh的bt的模型，基于dbh，ht和冠宽（crw）的体积，以及记录wbd数据。从Degaga–Gambo和Wondo Genet森林收集了全面而有代表性的数据集。ht，dbh和bt建模基于森林清查中的1345棵采样树，而体积建模和wbd文档基于30个物种的63种破坏性采样树，覆盖了研究地点总基础面积的87％。加权最小二乘回归用于建模，而遗漏的交叉验证用于评估。ht–dbh和dbh–dsh模型表现良好（伪R ²  = 0.72和0.98），而bt–dbh模型表现较差（伪R ² = 0.42）。针对总树，可交易茎和枝条体积的模型开发了不同的独立变量选项，其中伪R ²从0.74到0.98不等，分支模型的值最小。该模型可能适用于当前研究地点以外的森林只要仔细评估生长条件。物种方面的wbd范围为0.426至0.979 g cm ^-3，总体平均值为0.588 g cm ^-3。wbd数据对于建立国家wbd数据库非常有用，也可以包含在Global Wood Density数据库中。这项研究是迈向可持续森林管理的重要一步，包括在埃塞俄比亚中南部干旱的Afromontane森林中实施REDD + MRV实践。