A tentative dynamic site index model for Scots pine (Pinus sylvestris) plantations in Bulgaria

The site index tables for Scots pine (Pinus sylvestris L.) plantations applied in the forest inventory and management in Bulgaria nowadays consist of relative site index classes based on the mean height. They have limited applicability because of their narrow age-height range and poor fit to the actual growth trajectories. The main objective of this investigation is to derive a tentative dynamic, base-age invariant site index model for Scots pine plantations in Bulgaria based on the inventory of permanent sample plots, which would allow simple site index estimation and height growth prediction. Five of the most often used growth functions for modeling site index curves (Bertalanffy-Richards, Lundqvist-Korf, Hossfeld, Schumacher and Gompertz) have been tested in all their algebraic difference forms (ADA models), in several published generalized algebraic difference forms (GADA models) and a generalized algebraic difference form of Schumacher's 2-parameter model was proposed and examined in the present study as well. The candidate regression models, expanded with a first-order autoregressive error structure, were fitted using the base-age invariant dummy variable method. Lundqvist-Korf's growth function estimated in its anamorphic algebraic difference form was derived to characterize the site quality indices of Scots pine plantations in Bulgaria. The available data from plantation-grown Scots pine in Bulgaria had some limitations and model re-fitting on a data set expanded by additional measurements and stem analysis data is recommended. The study confirmed the reliability of the difference equation approach which can be widely employed for dynamic site index modeling of the forest stands in Bulgaria.