The planning and implementation of the majority of professional measures and activities in forest production management requires the knowledge of the basic quantitative elements (tree diameter, height, volume) and value elements (value of wood volume). In that sense, the aim of this research was to define the relationship between the value of timber and tree dimensions in beech forest thinning in Serbia (regression models, average growth rates, etc.). This will make it possible to predict in advance some value parameters, to raise the quality level of thinning operation planning, as well as to achieve better management of financial flows in the forestry sector (e.g. assessment of efficiency and profitability). The sample consisted of 7,815 beech trees harvested by thinning (5,973 in high forests and 1,842 in coppice forests), and the measurements included total volume and the volume and value of timber assortments.

The main scientific method applied in the study was the modelling method, and the analysis of causality between the study elements was performed using mathematical-statistical methods, such as descriptive statistics and regression and correlation analysis.

Based on the study results, the conclusions were as follows: (a) there is a high variability in timber value depending on forest origin and assortment category, (b) the effect of the diameter at breast height on timber value is stronger (the coefficient of linear regression R=0.52-0.83) than the effect of height (R=0.47-0.68), (c) the obtained regression models are a very good description of the relation between diameter at breast height and total timber value (R=0.58-0.97 in high forests and R=0.64-0.97 in coppice forests). An especially high quality model is the model which refers to the effect of diameter at breast height on the value per tree (R=0.97 in high forests and coppice forests), (d) all models show unequal distribution of the error along regression line (heteroskedasticity), but as this occurrence is more related to large trees, which are relatively few in thinning operations, this does not affect the accuracy of timber value assessment in thinning operations. The study conclusions point out that the timber value in beech forest thinning can be assessed with relatively high reliability.

The obtained models can be used when it is necessary to know the total value of thinned material (aiming at the determination of financial assessment of thinning efficiency), or when it is necessary to create the plans for selling the felled timber (estimation of the potential revenue).

Key words: diameter, height, timber value, beech, thinning, Serbia

Decreased growth of pedunculate oak (Quercus robur L.) during the 20th century has been documented in Europe and is considered as one of the symptoms of oak decline. However, there is insufficient information on wood formation of oak in Latvia (Baltic countries). Tree-ring width is a commonly used proxy for evaluation of growth, but other wood proxies, such as earlywood vessel size, can provide additional information on mechanisms affecting growth. In this study, we investigated changes in tree-ring width and cross-section area of earlywood vessels for the periods of 1899–2009. Since the 1980s, abrupt changes in wood formation were observed, particularly in the more continental eastern Latvia. Tree-ring width showed significant decrease, while lumen cross-section area of earlywood vessels significantly increased in the eastern part of Latvia. These changes might have been triggered by an extremely cold December in 1978.

Key words: earlywood vessel, tree-ring width, growth changes, climatic extreme, Quercus robur

An approach combining the information generated from different stochastic differential equations was developed to improve the predictive qualities of stem taper and volume. The stochastic differential equations and the stem taper and volume models were fitted to data from Scots pine and Norway spruce trees that were collected from across the entire Lithuanian territory. New models deduced from the Gompertz and Ornstein-Uhlenbeck shape stochastic differential equations were tested against the classical Kozak's stem taper model, q-exponential segmented stem taper model, classical Schumacher-Hall’s volume model, and q-exponential volume model based on allometric and geometric concepts. Comparison of the predicted stem taper and stem volume values with those obtained using regression fixed-effects models demonstrated the predictive power of the stochastic differential equations models.

Key words: stochastic differential equation, stem taper model, volume model

Variations in Brinell hardness of Scots pine wood from mature mineral soil stands in five regions in Finland and Sweden were studied by means of linear mixed models. Hardness was studied from planed tangential surfaces of the specimens, i.e., the force acted in radial direction. One model consisted of only readily available background variables, whereas, in the other model wood density was included as an independent variable. Geographical origin, as well as the longitudinal and radial location of the specimens within a tree was significantly affecting hardness. In addition, if the indentation was placed on latewood section the hardness was clearly higher compared to earlywood and combined earlywood and latewood. With the best fit model approx. 50% of the hardness variation could be described. Wood density was the most important variable affecting hardness and simple linear relationship was found between hardness and wood density. Hardness values based on the diameters of the residual indentations and the initial depths of the indentations were compared, and the differences between these two, as well as the possible sources of error were discussed.

Key words: Brinell hardness, linear mixed model, Pinus sylvestris L., Scots pine, wood quality