Tree allometric equations for estimating biomass and volume of Ethiopian forests and establishing a database: Review

Abstract

Biomass measurements are central to understanding and monitoring carbon stocks in the forests. They also assists in assessing the growing stock of a species and biomass energy. Efforts in allometric equation development have grown. This compiled information is an important tool for assessing biomass and associated carbon stocks, the growing stock of a species and biomass energy. Despite this, except for a few attempts, the compilation of available equations has received little attention, particularly in Ethiopia. The uncertainty involved in biomass estimation is also associated with the choice and errors of the equations. The aim of the present paper is thus to review and compile available biomass and volume allometric equations and to critically synthesise quality control procedures for the equations. A total of 767 equations for both volume and biomass were collected from 47 articles, theses and dissertations, representing 47 sites in Ethiopia. Of these, 49% of the equations were developed for the Dry Afromontane biome, 22% for the Acacia-Commiphora biome, 15% for the Combretum-Terminalia biome and 14% for the Moist Afromontane biome. Among the compiled equations, biomass equations comprised 86.7% while the remainder 13.3% were for volume. Although a wide range of natural variations resulted in diverse species in Ethiopia, the already available equations were constrained by the size of individual samples, tree diameter range, tree species and spatial coverage. Moreover, variability in equation presentation and the missing of some relevant statistical parameters or inconsistent statistical analysis were among the problems observed. A quality control procedure revealed that 45 allometric equations had verification problems and were inappropriate for application. Invalid numeric values that may over- or underestimated values were predicted in 42% of the equations. Moreover, seven equations predicted “unrealistic” values, which were low values (4 equations) or both negative and low estimates (3 equations) within their interval of calibration. Altogether, 174 equations (22.7%) were statistically credible equations. These missing information affected the use of the equations and led to the preference of the generalised equations to assess forest biomass stocks. This highlighted important research gaps. The compiled database will also be useful to minimise duplication of efforts and enable efficient utilisation of limited resources, particularly in developing countries such as those in Eastern Africa and beyond in the tropics.