Lukášek , Zeidler , Barcík : Shrinkage of Grand Fir Wood and its Variability within the Stem

This study presents the results of the experiment focused on assessment of the shrinkage of Grand fi r (Abies grandis /Douglas/ Lindl.) wood and its variability in relation to the position in the stem. The examined sample trees came from the Černokostelecka Area in the Czech Republic. Tangential shrinkage was 7.1 %, radial 3.3 % and volumetric shrinkage 10.5 %. The results show evident infl uence of the horizontal position on the evaluated features. The lowest shrinkage values are attained close to the pith; shrinkage toward the bark shows a rising trend. In terms of vertical position, the distribution of shrinkage in the stem did not show any apparent trend. Although the dependence of shrinkage on density proved to be statistically signifi cant, density explains only a small portion of the shrinkage variability..


INTRODUCTION 1. UVOD
Central European forest management is characterized by dominant usage of autochthonic forest woody plants.Silver fi r (Abies alba /Mill.)has for centuries ranked among the most signifi cant European trees.Due to its favorable growth and technological properties for the forest and wood industry, it was hardly replaceable.In comparison with the initial natural presence in the forest vegetation of the Czech Republic of 19.8 %, its current portion is about 1 % (MZE 2010).Beran (2006) states the clear absence of Silver fi r in the Czech forests in age groups 40 -60 years.In connection with massive recession of Silver fi r from the Central European localities, extensive research was started already in the last century for the purpose of fi nding a tree, which could substitute the favorable characteristics of Silver fi r.
Grand fi r (Abies grandis /Douglas/ Lindl.) was selected based on positive references as one of the possible trees because of small site requirements, resistance and above-average volume production (Beran and Šindelář, 1996;Šindelář et al, 2006;Podrázský and Remeš, 2008;Podrázský and Remeš, 2009;Mitze, 2010).The natural habitat of Grand fi r is the Northwest coast of North America.The climatic conditions in its natural habitat range from mild maritime to downright continental climate (Foiles et al, 1991;Alden, 1997;Klinka, 2007).Grand fi r is considered in the Central European region to be a prospective introduced tree species.The quick and high volume yield coupled with the lower density predetermines this species for efficient industrial application (Hapla, 2006;Vos and Kharazipour, 2010;Hof et al, 2008;Mitze, 2010).In terms of growth and production, Grand fi r is being monitored in many experiments and provenance tests.Current research supports its possibilities and growth potential mainly in juvenile years (König, 1995;Podrázský and Remeš 2008;Rau et al. 2008;Hof et al, 2008).
One of the most important physical properties of wood is shrinkage.This is a process in which the water in the cell walls evaporates from the wood resulting in dimensional and volume changes.Shrinkage describes the intensity of these changes in the wood.Wood shrinkage has a substantially anisotropic character (Bosshard, 1974;Požgaj et al, 1997;Niemz, 1993).Shrinkage is most intensive in the transverse directions, mainly in the tangential direction.Knowledge of the shrinkage values is fundamental for quality drying of the wood, hydrothermal treatment, impregnation, assurance of shape consistency and quality processing and working of the wood including surface fi nishes.Shrinkage of the wood continues to be a wood research subject (Kärki, 2001;Pelz et al, 2003;Riebel, 2007;Kord et al, 2010).
Wood as an organic material of plant origin shows substantial variability of properties (Josza and Middleton, 1994;Hapla and Wellhausen, 2003).The different shrinkage values in relation to the position in the trunk are confi rmed by many authors (Ying et al, 1994;Dumail and Castera, 1997;Kärki, 2001;Perstorper et al, 2001;Wang et al, 2008;Yamashita et al, 2009;Kord et al, 2010;Leonardon et al, 2010).Wood density can be considered the key indicator of wood quality.From the density value, it is possible to derive other signifi cant mechanical and physical properties of wood that are also important for practice (Bosshard, 1974;Niemz, 1993;Požgaj et al, 1997;Niemz and Sonderegger, 2003).This dependence can also be found between density and shrinkage.This is one of the results of the extensive study focused on the assessment of the quality of Grand fi r wood in the Czech Republic.The main idea of the experiments was to evaluate the position of Grand fi r among the domestic fi rs, and acquire data for the application of Grand fi r wood in the wood processing industry.Last but not least, the assessment was to be made of the possibilities for replacing the domestic fi r wood.The objective of this work was to determine the shrinkage of grand fi r wood from the Czech Republic and to establish its trunk variability in relation to the vertical and horizontal position.The dependence of shrinkage on density was also evaluated.

MATERIJALI I METODE
The locality for the collection of Grand fi r sample trees was situated close to Kostelec nad Černými lesy village.The territory lies at a distance of 25 -50 km south-east of Prague.The height above sea level of these localities is in the range of 325 -430 m.A closer description of the localities is given by Podrázský and Remeš (2009) and Tauchman et al. (2010).A total of six representative Grand fi r specimens were taken from  63 (2) 121-128 (2012) 123 tory with a controlled climate.The test samples used had the shape of squared timber of dimensions 20 mm x 20 mm x 25 ± 5 mm in tangential, radial and axial direction, respectively.
For each sample, the density was set at 12 % moisture content (ČSN 49 0108, 1993) and subsequently shrinkage in a tangential direction, radial direction and volume shrinkage (ČSN 49 0128, 1989).Total shrinkage, from green state to owen dry condition, was always evaluated in the respective direction.

REZULTATI
The average shrinkage value in the tangential direction was 7.1 % and in the radial direction 3.3 %.Volume shrinkage was 10.5 %.The rest of the descriptive statistics values for shrinkage in the individual anatomical directions and volume shrinkage are given in Table 1.
The progress of tangential shrinkage in relation to the trunk horizontal position shows a rising trend regardless of the vertical position and thus reaches the several different stands in this area.The age of the felled trees was in the range of 29 -38 years.The height of the sample trees was in the range of 24 -30 m and the breast-height diameter in the range of 27 -36 cm.
Three sections were always taken from the trunks of the felled trees -bottom, middle and top part (Vytisková, 1973;Langum et al, 2009).The sections represent different vertical positions in the trunk.When taking the specimens, the initial orientation in the plants was considered in terms of horizontal position (Sonderegger et al, 2008).
The centre planks were taken from individual sections, and test samples were subsequently prepared in a manner that enabled clear identifi cation of their trunk horizontal position (Gryc and Horáček, 2007;Sonderegger et al, 2008;Langum et al, 2009).Numbers 1, 2, 3, 4, 5 show relative position to the pith, where number 1 means the closest position to the centre of the tree.A closer defi nition of the method for collecting sections and positions of the test samples in the section is given in Figure 1.Production of the test samples and subsequent wood testing were done in accordance with the national standard (ČSN 49 0101, 1980) in a labora- ...Lukášek, Zeidler, Barcík: Shrinkage of Grand Fir Wood and its Variability within the Stem highest value in the zone furthest from the trunk pith.A similar trend is also observed in radial shrinkage and volumetric shrinkage (Figure 2).Density shows a similar trend to shrinkage, whereby it rises from the pith to the cambium (Figure 2).The most intensive rise in density appears in the bottom section of the trunk, while the mildest is in the top section.Vertical variability does not have a clear trend in the case of radial shrinkage.The shrinkage values initially rise with height, and then drop towards the top.The lowest radial shrinkage value in the vertical direction is in the bottom section of the trunk.The highest shrinkage value is in the middle section of the trunk.ANOVA (Tukey test, α = 0.05) confi rmed the statistically signifi cant difference in the radial shrinkage values between the middle and the bottom sections.Tangential shrinkage is also highest in the middle section of the trunk, the lowest shrinkage values are, however, reached in the top section.The top section shows statistically the lowest tangential shrinkage values.Volumetric shrinkage copies the trend of tangential shrinkage.The lowest shrinkage values are in the top section and the lowest in the middle section of the trunk.The top section shows statistically slower volumetric shrinkage values in comparison with the bottom or middle section.Radial, tangential and volumetric shrinkage has a comparable trend in the vertical direction.Shrinkage initially rises towards the middle section of the trunk and then slightly drops towards the top (Figure 3).
Although the relationship between shrinkage and density has been confi rmed as statistically signifi cant, the infl uence of density on the shrinkage of Grand fi r wood is relatively low.The infl uence of density is most signifi cant on radial shrinkage.However, the determination coeffi cient is only 0.28.In the remaining two instances of shrinkage evaluation, the values are even lower, 0.23 for volumetric shrinkage and only 0.13 for tangential shrinkage (Figure 4).

RASPRAVA
The ascertained shrinkage values for Grand fi r from the locality in the Czech Republic correspond to the values attained in the natural habitat.Principally, identical values were obtained for both anatomical directions and volumetric shrinkage (Table 2).
Table 3 shows the comparison of the ascertained values with the quality of grand fi r wood from the plan-tations in the rest of the European countries.It is clear that the shrinkage values of Grand fi r in the Czech Republic are similar to those obtained by the authors who evaluated the shrinkage of this introduced wood tree grown in the neighbouring states.
The comparison with our native and commercial softwoods that Grand fi r could suitably supplement or even replace is contained in Table 4.In terms of shrinkage values in the individual anatomical directions and volumetric shrinkage, Grand fi r is closest to our domestic Silver fi r.Riebel ( 2007) also arrived at the same conclusion.From just the shrinkage aspect, Grand fi r is a suitable replacement.
In terms of shrinkage progress in grand fi r trunk in the horizontal direction, the lowest shrinkage value is obtained near the pith.This distribution of shrinkage in the trunk is more or less similar for tangential, and radial as well as volumetric shrinkage.With the exception of the bottom section in the case of radial shrinkage, the vertical position has no infl uence and shrinkage shows a rising trend from the pith to the bark regardless of the section.The radial distribution of the density copies the shrinkage progress, whereby the lowest  shrinkage value is near the pith, regardless of the vertical position, and has a rising trend toward the bark.The rising shrinkage values from the pith toward the bark are confi rmed by Herritsch (2007), Wang et al. (2008).This phenomenon is attributed mainly to the presence of juvenile wood in the central zone of the trunk, and to the different microfi bril angle of the cell wall of the juvenile wood in comparison with the angle of mature wood (Panshin and De Zeeuw, 1980;Larson et al, 2001;Perstoper et al, 2001;Yamashita et al, 2009;Leonardon et al, 2010).
Concerning the infl uence of the vertical position in the trunk on shrinkage distribution, a marked trend is not discernible in the case of Grand fi r.Mild rise in the shrinkage from the bottom to the tree-top is in the middle section of the trunk replaced by a mild drop.With the exception of radial shrinkage, statistically the lowest value is obtained in the upper section of the trunk.The ambiguous trend in the vertical distribution of the dimensional changes of wood in the trunk in relation to the change of moisture content is confi rmed by Gryc et al. (2007) and Wang et al. (2008).Just like in the case of radial variability, the vertical progress of the features is infl uenced by the portion of juvenile wood in individual trunk sections and its structural difference from mature wood (Zobel and Van Buitenen, 1989).A relatively surprising factor is the vertical progress of density, whereby its value rises with rising height and is the highest at the top part.Higher values of density in the upper section of the trunk are given by Dumail and Castera (1997).A different statistical density value is only in the middle section of the trunk.The differences in the density values in the bottom section and the top section cannot, however, be statistically proven (ANOVA, Tukey test, α = 0.05) and it is not possible to talk about any signifi cant trend in dependence of density on vertical position.The small infl uence of tree height on density is confi rmed by Herritsch (2007) for Pinus radiata.
Shrinkage of both juvenile and mature wood is closely correlated to the wood density (Panshin and De Zeeuw 1980;Larson et al. 2001).Positive correlation of density and shrinkage was found by Bengtsson (2001), Perstoper et al. (2001) and Yamashita et al. (2009).The infl uence of density on the shrinkage of Grand fi r wood is relatively low and is the most signifi cant in radial shrinkage (Figure 4).The low value of the determination coeffi cient for the dependence of shrinkage on density is given by Kärki (2001).Niemz and Sonderegger (2003) state the determination coefficient for tangential swelling in relation to the density of wood at an average value of 0.39 for selected industrially processed woods.Also Kord et al. (2010) state a determination coeffi cient of 0.39 for the dependence of tangential shrinkage on wood density.

ZAKLJUČAK
The conclusions of this study targeted at the assessment of the shrinkage of Grand fi r wood and its variability within the stem from the selected area in the Czech Republic are as follows: The shrinkage values for Grand fi r in the climatic conditions of the Czech Republic correspond to the values obtained in the natural habitat.
The shrinkage values are identical to those obtained in other European countries.
Based on the shrinkage, Grand fi r wood from do-mestic trees is the closest to the Silver fi r wood.
The infl uence of radial position on shrinkage varia-bility was clearly proved, whereby the lowest values of all the assessed shrinkages are obtained in the centre of the trunk and rise towards its outer perimeter.
The vertical position does not clearly infl uence the size of the shrinkage, whereby it was not possible to monitor the signifi cant trend in the distribution of the shrinkage over the height of the trunk.
The positive dependence of all the evaluated types of shrinkage on density was proven.The usage of the density value as an indicator of the degree of shrinkage is, however, disputable because the ascertained determination coeffi cients are relatively low.
In terms of the overall concept of research focused on the evaluation of Grand fi r wood quality, consequences for its processing and possibilities for its application, it can be concluded that Grand fi r is a potential substitute for the indigenous Silver fi r.However, this is a partial conclusion only based on comparison of shrinkage.Due to the substantially lower density in comparison with the domestic fi r, additional tests shall be done, mainly of mechanical characteristics, and a complex comparison of these species will be performed.

Table 3
Comparison of shrinkage values with the values of the rest of the European plantations Tablica 3. Usporedba dobivenih vrijednosti utezanja s vrijednostima utezanja za drvo s europskih plantaža