Knowledge of the potential productivity of forest sites is fundamental for making strategic decisions in forest management. Site productivity is usually evaluated using the site index, and therefore the development of site index models is one of the crucial tasks in forest research and forest management. This research aims to develop an effective method for building top-growth and site index models using data from temporary sample plots (TSP). Exploiting the advantages of the generalised algebraic difference approach (GADA), the proposed method overcomes the limitations of the guide curve method that has been to date used in site index modelling using TSPs data and allows to obtain only a set of anamorphic site index curves. The proposed approach enables the construction of dynamic site index models with polymorphism and variable asymptotes. Such models better reflect local, site-specific height growth trajectories and therefore allow more appropriate site index estimation. We tested the proposed method using data collected from 5105 temporary sample plots in Poland. Our results indicate that growth trend estimates using height–age measurements of TSPs may be valuable data for modelling top height growth. For these reasons, the proposed method can be very useful in forest management.

The site index (SI) is the most commonly used and representative measure of the phytocentric approach; it evaluates the site productivity based on the stand height and age. In the case of mixed stands with complex structures, phytocentric methods are very limited, while in non-forest areas, they are not ap­plicable. In situations where the applicability of phytocentric methods is limited, the site productivity is de­termined by geocentric methods. Geocentric methods allow direct modelling of site productivity, expressed by SI predicted from various environmental variables. The aim of this study was to develop a geocentric model for oak. Site productivity expressed by SI was described by the environmental variables and stand characteristics. To develop the SI model, we used the data from 2490 NFI plots with dominant oak species (Quercus sessilis and Quercus robur). A generalized additive model was used in modelling site productivity. We documented a significant relationship between SI and the environmental variables, age of stands and stand density. Furthermore, the site productivity for oak is shaped by climate factors, soil type, geology, and altitude. The model developed based on the geocentric method, explained 55.1% of the variation of SI

No abstract