Allometric biomass equations for young broadleaved trees in plantations in Romania

The possibility of estimating young trees biomass is rather limited because forest yield tables are constructed starting from higher thresholds of proxy, such as diameter or height, and lack of availability of allometric equations. The aim of this study is to provide species-specific and general biomass equations for young plants often used for plantations on marginal lands in southern and eastern Romania. Power functions based on log-transformed data were applied to seven tree species (Robinia pseudoacacia (L), Quercus sp., Populus alba (L.), Gleditsia triacanthos (L.), Elaeagnus angustifolia (L.). Salix alba (L.) and Fraxinus excelsior (L.)), one shrub (Rosa canina L.) and to the overall dataset with all the species pooled together (406 plants), using the diameter at collar height (D(ch)), diameter at breast height (D(bh)) or height (H) as single predictor. D(ch) resulted as being the best predictor for each compartment for very young trees, but H also proved to be a promising individual predictor both for species-specific or general equations. D(bh) could satisfactorily predict the aggregated aboveground biomass, but generally could not adequately estimate all biomass components (i.e., belowground biomass or foliage). The goodness of regression was lowest for the foliage and highest for the stem and aggregated biomass compartments. The scaling coefficient (a) and exponent (b) of power functions were influenced both by species-specific factors and by the growth stage of the trees. Parameters to be used for a general equation were also provided for each biomass compartment and predictor. Using D(ch) as independent variable, we observed that the value of the general scaling exponent estimated to predict total aboveground biomass was the same as the value (2.66) predicted by the WBE functional model. Using D(bh) as predictor for the general allometric equation, the resulting value of b (2.36) coincided with the values empirically estimated by previous studies. Equations were finally compared against three independent datasets. Parameters provided by the general equation highlighted permanent overestimation for aggregated biomass compartments and underestimation for branches or roots, but always fell into the range provided by the upper and lower values estimated for a and b. This suggests that, at least for young trees, our equation could be applied without regard for local fertility conditions or plantation management. (C) 2011 Elsevier B.V. All rights reserved.