Allometric equations for estimating aboveground biomass of Coffea arabica L. grown in the Rift Valley escarpment of Ethiopia

Coffee, Coffea arabica L., which is native to Ethiopia, is the world’s most widely traded tropical agricultural commodity. While much is known about the productivity and management of coffee for coffee beans little attention has been given to the plants overall biomass production and carbon sequestration. The objective of this study was to develop and evaluate allometric equations for estimating the aboveground biomass of C. arabica plants growing in indigenous agroforestry system in the Rift Valley escarpment of south-eastern Ethiopia. Coffee plays an important role in providing income and in sustaining these productive systems. Biomass harvesting of 31 plants with 54 stems was carried out in a 40 km² area varying in elevation from 1,500 to 1,900 m. The stem accounted for most (56 %) of plant biomass, followed by branches (39 %) and twigs plus foliage (5 %). Plant mean biomass was 22.9 ± 15.8 kg. Power equations using stem diameter measured at either 40 cm (d ₄₀) or at breast height (d, 1.3 m) with and without stem height (h) were evaluated. The square power equation, $ Y = b_{ 1} d_{ 40}^{ 2} $ , was found to be the best (highest ranked using goodness-of-fit statistics) for predicting total and component biomass. The reliability of the prediction decreased in the order: stem > branches > twigs plus foliage. A cross-validation procedure showed that equation parameterization was stable and coefficients reliable. Our parameterized square power equation for total aboveground biomass was also found to be better than the equations parameterized by Hairiah et al. (Carbon stocks of tropical land use systems as part of the global C balance: effects of forest conversion and options for clean development activities, International Centre for Research in Agroforestry, Bogor, 2001) and Segura et al. (Agroforest Syst 68:143–150, 2006) for C. arabica grown in agroforestry systems, confirming the importance of parameterization of allometric equations with site specific data when possible.