Resumen
Quantifying the carbon balance in forests is one of the main challenges in forest management. Forest carbon stocks are usually estimated indirectly through biomass equations applied to forest inventories, frequently considering different tree biomass components. The aim of this study is to develop systems of equations for predicting tree biomass components for the main forest softwood species in Spain: Abies alba Mill., A. pinsapo Boiss., Juniperus thurifera L., Pinus canariensis Sweet ex Spreng., P. halepensis Mill., P. nigra Arn., P. pinaster Ait., P. pinea L., P. sylvestris L., P. uncinata Mill. For each species, a system of additive biomass models was fitted using seemingly unrelated regression. Diameter at the breast height and total height were used as independent variables. Diameter appears in all component models, while tree height was included in the stem component model of all species and in some branch component equations. Total height was included in order to improve biomass estimations at different sites. These biomass models were compared to previously available equations in order to test their accuracy and it was found that they yielded better fitting statistics in all cases. Moreover, the models fulfil the additivity property. We also developed root:shoot ratios in order to determine the partitioning into aboveground and belowground biomass. A number of differences were found between species, with a minimum of 0.183 for A. alba and a maximum of 0.385 for P. uncinata. The mean value for the softwood species studied was 0.265. Since the Spanish National Forest Inventory (NFI) records species, tree diameter and height of sample trees, these biomass models and ratios can be used to accurately estimate carbon stocks from NFI data.