Tree biomass models for the entire production cycle of Quercus suber

The demand for accurate biomass estimates is more pressing than ever. Cork oak (Quercus suber) presents particular challenges for biomass estimation due to variability among trees, even within the same stand. One issue is that small cork oak trees in the understory are often overlooked, leading to an underestimation of stand biomass. Characterising stand biomass is a significant concern for process-based modelling, as inaccuracies in observed values can result in further output uncertainty. Moreover, the production and management of cork oak are primarily focused on cork extraction, which alters several tree biomass components, particularly after the first debarking. This study provides a set of models for estimating cork oak biomass (total and per-tree component) across all stand layers. The models were developed using seemingly unrelated regression, ensuring additivity between biomass compartments. The separation of the cork oak life cycle reflects tree development and the production cycle of interest for managers, resulting in three sub-groups (juvenile trees, adult virgin trees, and debarked trees), along with four compartments (leaves, branches, stem, and aboveground). The juvenile model requires only total tree height, while the virgin model requires diameter at breast height and total tree height. No other tree variable was tested in the juvenile model, and only the number of first-order main branches was tested in the virgin model, but the associated parameter was not significantly different from zero. Two alternative models were developed for the debarked trees: a simple model that offers an alternative when only diameter under bark is available (predictive modelling efficiency, EF = 0.35 for branches and EF = 0.89 for total aboveground biomass); and a complex model that incorporates crown and stand variables, demonstrating better performance for branches and stem when parameters depend on the crown length and stand basal area, respectively (EF = 0.63 for branches and EF = 0.90 for total aboveground biomass). Our results underscore the need to consider crown and stand variables in the later stages of development.