Abstract
Wood permeability has a major effect on industrial wood processing and utilization. Wood anatomy and resin influence the permeability of wood. Understanding and manipulating these influences is important to optimize the manufacture and use of forest products. This study investigated the relationships between wood anatomical traits, radial permeability, and resin content of samples collected from 19-year-old hybrid pines (Pinus elliottii var. elliottii × Pinus caribaea var. hondurensis) from Queensland, Australia. The earlywood tracheid tangential lumen diameter and axial resin canal diameter were statistically positively correlated with radial permeability. The heartwood proportion and the frequency of axial resin canals were statistically negatively correlated with radial permeability and positively correlated with resin content. The axial resin canal diameter, sapwood proportion, latewood content, ray frequency, and earlywood tracheid lumen diameter increased from pith to bark, whereas the axial resin canal frequency decreased. Resin was found distributed throughout the wood microstructure, from pith to bark in many samples, and in both heartwood and sapwood. However, there was a much greater quantity of resin in heartwood and wood from the middle (inner radius) of the tree, with widespread occurrence of resin impregnation in the axial tracheids.
Download PDF
Full Article
Influence of Wood’s Anatomical and Resin Traits on the Radial Permeability of the Hybrid Pine (Pinus elliottii × Pinus caribaea) Wood in Australia
William Leggate,a,b,* Maryam Shirmohammadi,b Robert L. McGavin,b Kerri Chandra,c Mark Knackstedt,a Lydia Knuefing,a and Michael Turner a
Wood permeability has a major effect on industrial wood processing and utilization. Wood anatomy and resin influence the permeability of wood. Understanding and manipulating these influences is important to optimize the manufacture and use of forest products. This study investigated the relationships between wood anatomical traits, radial permeability, and resin content of samples collected from 19-year-old hybrid pines (Pinus elliottii var. elliottii × Pinus caribaea var. hondurensis) from Queensland, Australia. The earlywood tracheid tangential lumen diameter and axial resin canal diameter were statistically positively correlated with radial permeability. The heartwood proportion and the frequency of axial resin canals were statistically negatively correlated with radial permeability and positively correlated with resin content. The axial resin canal diameter, sapwood proportion, latewood content, ray frequency, and earlywood tracheid lumen diameter increased from pith to bark, whereas the axial resin canal frequency decreased. Resin was found distributed throughout the wood microstructure, from pith to bark in many samples, and in both heartwood and sapwood. However, there was a much greater quantity of resin in heartwood and wood from the middle (inner radius) of the tree, with widespread occurrence of resin impregnation in the axial tracheids.
Keywords: Wood anatomy; Wood permeability; Resin content; Pinus elliottii; Pinus caribaea
Contact information: a: Research School of Physics and Engineering, The Australian National University, Canberra, ACT 0200, Australia; b: Queensland Department of Agriculture and Fisheries, Horticulture and Forestry Science, Salisbury Research Facility, 50 Evans Rd, Salisbury, Qld 4107, Australia; c: Queensland Department of Agriculture and Fisheries, Eco Sciences Precinct, 41 Boggo Rd, Dutton Park, Qld 4102, Australia; *Corresponding author: william.leggate@daf.qld.gov.au
INTRODUCTION
Permeability is a measure of the ease with which liquids and gases flow through a porous substance under the influence of a pressure gradient (Comstock 1968; Tesoro 1973; Milota et al. 1994). The permeability of wood influences many of its important processing and utilization properties, such as drying, preserving, wood modification systems, pulping, gluing, finishing, and even durability (Fogg 1968; Tesoro 1973; Hansmann et al. 2002; Zimmer et al. 2014). Generally, less permeable wood types are more difficult to dry and to treat with chemical preservatives (Milota et al. 1994). Wood permeability is also one of the main controlling factors influencing the depth of adhesive penetration (Kumar and Pizzi 2019). Lower wood permeability has also been linked to higher durability in some cases (Nicholas et al. 2005; Sandberg and Salin 2012).
Wood microstructure and wood extractives have been shown to exert a major influence on permeability (and the treatability of wood), contributing to the problems that arise when treating refractory heartwood and sapwood. The wood anatomical characteristics and phenomena that have been shown to affect wood permeability in softwoods include heartwood content, pit aspiration, latewood proportion, cellular dimensions, tylosoids (outgrowths in resin canals), the presence of interstitial spaces, and the frequency of longitudinal tracheids, longitudinal parenchyma, ray parenchyma, ray tracheids, and resin canals (Stamm 1931; Fogg 1968; Tesoro 1973; Hansmann et al. 2002; Martín et al. 2010; Leggate et al. 2019). Permeability is also influenced by wood chemistry, in particular extractives, including resins that can block liquid flow (Ellwood and Ecklund 1961; Fogg 1968; Olsson et al. 2001; Baraúna et al. 2014). The migratory characteristics of resins (e.g. during high temperature drying) also influence the penetration of preservative liquids including penetration depth and uniformity (Ahmed et al. 2012). Resins are produced as a natural tree defense mechanism against pests, diseases, and mechanical damage. A key function of resin, apart from some fungicidal and insecticidal effects, is to seal wood tissue to prevent the invasion of microbes (such as fungal hyphae) and to fill air spaces (Back and Allen 2000). The resin seal is water resistant, therefore creating impermeable barriers to liquid movement (Back and Allen 2000).
The hybrid pine, (Pinus elliottii var. elliottii [PEE] × Pinus caribaea var. hondurensis [PCH]) now dominates exotic softwood plantations and log supply in Queensland, Australia, with approximately 94,100 ha planted and an annual log production soon to exceed 1 million cubic meters per year (Leggate et al. 2019). This important resource supports a diverse processing sector that includes sawn timber, engineered wood products, reconstituted panels, landscaping, and lower-grade end uses (Lee 2015).
Leggate et al. (2019) reported a study on the radial gas and liquid permeability and resin content of 19-year-old PEE × PCH plantation-grown hybrid pine trees from Queensland, Australia, which were represented by various genotypes and tree stocking rates (stems per hectare). The key findings from this study were that there was no important effect of genotype or stocking rate on radial permeability; both gas and liquid radial permeability increased significantly from pith to bark positions within the tree. Conversely, resin content decreased from pith to bark. Gas and liquid permeability were significantly correlated and a highly significant negative relationship was found between permeability (gas and liquid) and resin content. However, the Leggate et al. (2019) study did not investigate the relationships between wood anatomical traits and these factors. Understanding and manipulating (e.g. genetically, silviculturally, or via wood product manufacturing techniques) these relationships is important for optimizing the manufacture and use of forest products. This paper reports on a study that investigated the influence of wood anatomical and resin traits on radial permeability.
EXPERIMENTAL
Wood Samples
The wood samples used for this study were derived from a previous study reported by Leggate et al. (2019) on the radial permeability and resin content of 19-year-old PEE × PCH trees. These trees were represented by three commercially relevant genotypes: the F1 seedling, which was selected as a routine plantation seedlot (a single cross family); Clone 887, which was selected for good growth and flat, fine branches; and Clone 625, which was selected for heavy branches to contrast with clone 887. Five plantation tree-stocking rates (200, 333, 500, 666, and 1000 stems per hectare [spha]) represented a low to high range of stocking rates. A set of 30 trees was harvested from Experiment 622NC, an F1 (first filial hybrid of PEE × PCH) taxon (clones and F1 family) spacing trial planted during March of 1997 within compartment 202 Donnybrook LA (26° 58’ 30” S; 152° 59’ 00” E), SF 611 Beerburrum, north of Brisbane, Queensland, Australia. Experiment 622NC was established for wood production research purposes and not for resin production – therefore no trees in Experiment 622NC were tapped for resin collection. The trees were 19 years old at the time of harvest in October 2016. As per Table 1, two trees from each stocking rate and genotype combination were harvested to provide samples for wood property analysis.
Table 1. Tree Sampling Criteria (Number of Trees Sampled for Each Genotype and Stocking Rate Combination) (Leggate et al. 2019)
Wood samples were obtained from one transverse disc (35-mm thickness) that was taken from each tree at 2.34 m height. From each disc, samples for radial permeability and resin content determination were taken at three radial locations from pith to bark as per Fig. 1 (from Leggate et al. 2019). Further description of the methodology used for tree sampling, permeability (gas and liquid [water]), and resin content measurement is provided in Leggate et al. (2019).
Fig. 1. Permeability and resin content samples cut from discs at three radial locations from pith to bark (Leggate et al. 2019)
A subset of permeability samples was selected from the study undertaken by Leggate et al. (2019) to facilitate anatomical analysis that formed the basis for the present study (Table 2).
Table 2. Sub-sampling for Anatomical Studies (number of wood samples)
These permeability samples (24 mm in diameter and 8 mm in thickness) were cut into three sections as per Fig. 2.
Fig. 2. Sectioning of the permeability samples for wood anatomical studies
Light Microscopy
Section B shown in Fig. 2 was used for light microscopy studies. The top and bottom transverse surfaces of section B were prepared using a Reichert sliding microtome (Model Number 338188; Reichert, Vienna, Austria) with steel blades to provide a clean, smooth surface. Images from both surfaces were captured using a Nikon Eclipse LV100ND microscope (Tokyo, Japan) at 50x magnification. Image J (IJ 1.46r) (U.S. National Institutes of Health, Maryland, USA) was used to measure:
- The cell lumen diameter in both the tangential and radial directions of the earlywood [first-formed wood within an annual growth ring that has a high proportion of relatively large diameter thin-walled cells (Zobel and Sprague 1998)] and latewood [later formed wood within an annual growth ring that has thicker cell walls (Zobel and Sprague 1998)] tracheids;
- The diameter and area of individual axial resin canals (a tubular intercellular duct lined by an epithelium, conducting secondary plant products secreted by the epithelial cells and that is surrounded by subsidiary cells [Richter et al. 2004]); resin canal diameter was determined as the average of tangential and radial diameters, and resin canal area was measured as the surface area occupied by the resin canal on the transverse surface. Resin canal diameter and area measurements included all components of the resin canal complex (includes epithelial, parenchyma and other subsidiary cells).
Scanning
The same middle section (Section B) used for light microscopy studies was also scanned on the transverse surface using an Epson Perfection Dual Lens System V850 Scanner (EPSON, Suwa, Japan) at high resolution (9600 dpi). Image J (IJ 1.46r) was used to process the scanned images to measure the following features:
- The earlywood and latewood proportions (based on surface area);
- The total number of axial resin canals per mm2;
- The percentage of axial resin canals in both earlywood and latewood; and
- The percentage of resin-filled/closed axial resin canals in both earlywood and latewood.
Section A shown in Fig. 2 was used to determine the heartwood and sapwood proportions on the transverse surface. Heartwood is defined as the inner layers of the wood in a growing tree that no longer contain living cells, and in which the reserve materials (e.g., starch) have been removed or converted into heartwood substances (International Association of Wood Anatomists 1957). It is usually notably darker than the surrounding sapwood. Sapwood is defined as the outer layers of wood in a living tree containing living cells and reserve materials (e.g., starch) (International Association of Wood Anatomists 1957). Sapwood is usually markedly paler in color than the surrounding heartwood.
A chemical spot test method for differentiating heartwood and sapwood in Pinus species (QDPI 2003) was applied. After chemical application, the samples were scanned using an Epson Perfection Dual Lens System V850 Scanner at high resolution (9600 dpi). After scanning, the heartwood and sapwood proportions were measured using Image J (IJ 1.46r)
Image Capture Using a Micro-capture Camera
Section B (shown in Fig. 2) used for light microscopy studies and scanning as discussed above, was photographed using a USB Micro Capture Plus camera at a magnification of 10x (Model Number QC 3199; DIGITECH Industries, Hong Kong, SAR, China). The camera used was equipped with eight LED lights with adjustable brightness and had an image resolution of 2592 × 1944 pixels. The images were then assessed using Image J (IJ 1.46r) to determine the number of rays per mm2 on the transverse surface. A ray is defined as a ribbon-like aggregate of cells extending radially in the xylem and phloem (International Association of Wood Anatomists 1957).
Micro-computer Tomography
To examine the micro-distribution of resin within the wood structure, eighteen permeability samples with dimensions (24 mm diameter and 8 mm thickness) representing three radial positions (from pith to bark) and from low to high permeability were scanned at the National Laboratory for X-ray Micro Computed Tomography (CTLab) based at the Australian National University (ANU, Canberra, Australia). A HeliScan MicroCT system with an optimized space-filling trajectory (ANU, Canberra, Australia) (Kingston et al. 2018) was used to yield sharp images (Latham et al. 2008; Myers et al. 2011) at a resolution of 10.9 μm. A subset of these samples (seven) was also scanned at a smaller dimension (of 8 mm or 6 mm diameter × 8 mm thickness) to achieve a higher resolution image of 4.6 μm. These images were then assessed using the open-source scientific visualization software DRISHTI (Version v2.6.6; ANU, Canberra, Australia) (Limaye 2012).
Scanning Electron Microscopy
For detailed observations of the wood’s anatomical features, six samples with dimensions of 8 mm (radial) × 12 mm (tangential) × 8 mm (longitudinal) representing different radial positions and permeability values were used for scanning electron microscopy (SEM). Prior to imaging, samples were kept in an oven at 50 °C for 7 days to reduce the sample’s moisture content. To increase the contrast of samples, transverse and tangential surfaces of oven-dried samples were platinum coated using a Leica EM-SCD005 sputter coater (Leica, Wetzlar, Germany). A Zeiss Sigma VP field emission SEM (Zeiss, Oberkochen, Germany) was used to capture images of transverse and tangential surfaces. Images were taken in different magnifications (200x, 500x, and 1.5kx) to visualize the cell size and types in each plane.
Statistical Analysis
Gas and liquid permeability, resin content, and other wood anatomical features and their relationships were initially explored using a graphical compilation of scatterplots, histograms and Pearson correlation values (Fig. 3) from a data matrix using the pysch package (v1.9.12; Northwestern University, Evanston, IL, USA) (Revelle 2019) in R (R Foundation for Statistical computing, Vienna, Austria) (R Core Team 2019). A data matrix of the most significant variables, as chosen in the statistical analysis detailed below, was created to compile the graphical output showing the individual relationships and their correlation.
Statistical analyses were performed in GenStat v19 (VSN International 2018, Hemel Hempstead, UK). The wood anatomical features were each analyzed with a general linear regression. However, the heartwood proportion did not have a wide enough range to analyze via linear regression and was thus converted into proportional categories of All (samples with100% Heartwood), Intermediate (samples with 1 to 99% Heartwood), and None (samples with 0% Heartwood). The heartwood was subsequently analyzed using a multinomial regression. Stocking rate, genotype, position, and their interactions were fitted as explanatory variables/factors and any interactions that were found to be non-significant were dropped from the model via backwards selection.
Gas and liquid permeability as well as resin content were analyzed using general linear regressions. Rather than using the design factors of stocking rate, genotype, and position as explanatory variables (as per Leggate et al. 2019), the wood anatomical features were considered as potential explanatory variables to determine the underlying biological relationship brought on by the experimental design. To determine the most significant variables, multiple all-subset regressions were performed on groups of variables. The most influential variables were selected, and interactions between them were also tested. The final model was chosen as the model that included all significant interactions and their individual effects, as well as any other significant variables. For all models, the means and standard errors were calculated and, where relevant, pairwise comparisons were performed using Fishers Protected Least Significant Difference at a 5% significance level.
RESULTS AND DISCUSSION
Summary of Statistical Relationships between Wood Anatomical Traits, Permeability, and Resin Content
Using only the wood anatomical features (heartwood proportion, earlywood proportion, ray frequency, resin canal characteristics and dimensions, and tracheid dimensions) the linear regressions were able to account for 78.6%, 74%, and 63.2% of the variation in gas permeability, liquid permeability, and resin content, respectively. For each of these models, the individual contributions from the wood anatomical features are discussed below. Figure 3 shows a graphical matrix for the direct relationships between key variables independent of interaction with other variables.
Fig. 3. Graphical representation showing direct relationships between key variables (where significant Pearson correlation values are shown as: * p < 0.05, ** p < 0.01, and *** p < 0.001). In1Gasperm = gas permeability; In1Liquidperm = liquid permeability; In1ResinContent = resin content; HW = heartwood; ResinCanals2 = axial resin canal frequency; DiamEW.T – tangential diameter of earlywood tracheids; RCDiam = resin canal diameter; RCArea = resin canal area.
Heartwood Proportions
The statistical analyses of heartwood proportion split into categories of All (samples with 100% Heartwood), Intermediate (samples with 1 to 99% Heartwood), and None (samples with 0% Heartwood), showed a significant difference for radial positions (p < 0.001) with the pith zone having significantly more heartwood compared with the mid-radius and bark positions (Table 3). The bark and mid-radius positions were not significantly different in any heartwood category, and had none or almost none in the All and Intermediate categories. Neither genotype nor stocking rate had a significant relationship with the heartwood proportion categories.
The heartwood proportion category and resin canal area (for individual resin canals) had a significant interaction (p < 0.001) for gas permeability; permeability did not significantly change with increasing resin canal area for those with no heartwood (None category). However, for those samples with some/all heartwood (Intermediate and All categories), as the resin canal area increased, the gas permeability decreased. The decrease appeared more pronounced in samples consisting of only heartwood. The heartwood proportion category had a significant effect (p < 0.001) on liquid permeability in those with no heartwood, which had a significantly higher liquid permeability than those with some/all heartwood. The analysis of resin content showed that samples with no heartwood had significantly lower (p < 0.001) resin content than those with some/all heartwood.
Heartwood is much less permeable compared with sapwood, which is one major factor explaining the lower permeability of the wood from the middle of the tree (on the horizontal axis of the tree) (Leggate et al. 2019). The heartwood of the Pinus genus is less permeable compared with sapwood due to factors, such as higher extractives (resin) content, greater pit aspiration, pit encrustation, reduced size and frequency of pits, and the presence of tylosoids (Fig. 4). Heartwood formation also entails a shift in the mean pore size toward smaller radii (Schneider and Wagner 1974).
Fig. 4. An SEM image of a tylosoid in an axial resin canal from a heartwood sample at magnification of 500x
Koch (1972) defines the southern pines as those species whose major natural range is in the United States south of the Mason-Dixon line (39 °43´ N.) and east of the Great Plains. The southern pine group of species includes PEE. Pinus caribaea is not strictly a southern pine species according to Koch’s definition, but it is viewed as closely related, and has been crossed with other southern pine species to yield hybrids. Koch (1972) reports considerably higher extractives content in the heartwood of southern pines compared with sapwood. Other studies on similar species within the southern pine group have shown that the permeability of sapwood can be up to 12,000 times more permeable than heartwood (Stamm 1931; Fogg 1968; Tesoro 1973). However, even in the absence of heartwood, permeability has been shown to decrease in the sapwood from the outer to the inner in all three anatomical directions (Hansmann et al. 2002), a tendency also observed in the present study. Juvenile wood has also been shown to have lower permeability than mature wood (Milota et al. 1994).
The extent of heartwood formation in the Pinus species varies by age. Koch (1972) commented that until 50 years of age southern pines have minimal heartwood, with most of the wood volume comprising light-colored sapwood. Benson (1930) reported an examination of stumps and cross-sections of several hundred southern pine trees, ranging in age from 10 to 300 years and in size from saplings to old forest veterans. His observations indicated that heartwood formation in southern pines usually began when the trees were between 20 and 30 years of age. In trees under 20 years of age, regardless of size, little or no heartwood was found (Benson 1930). However, this study by Benson (1930) examined heartwood proportions in southern pines in the United States, and not in trees grown in Australian conditions. The wood samples in the present study were sourced from 19-year-old plantation-grown trees in Queensland, Australia, and were already showing heartwood development at the time of sampling. The age that heartwood formation begins has been shown to be influenced by various climatic conditions—for example, in Scots pine (Pinus sylvestris), heartwood is formed after approximately 20 years in Alsace, France, compared with 70 years in northern Sweden (Back and Allen 2000).
Earlywood and Latewood Proportions
Overall, this study revealed a much greater proportion of earlywood (the average across all samples was 86%) compared with latewood (the average across all samples was 14%) (Table 3). The earlywood proportion decreased and the latewood proportion significantly increased from pith to bark (p < 0.001), with the average earlywood proportion for both regions 100% and 71%, respectively (Table 3). This pattern has been a typical trend in southern pines, where the percentage of latewood and the wood density increases from the pith toward the bark until the maximum is reached (Hakkila 1989). There was no statistically significant relationship between the proportion of earlywood or latewood and tree stocking rates or genotype.
When considered independently of other variables, the earlywood/latewood proportion accounted for approximately 25 to 30% of the variability in both gas and liquid permeability and resin content. However, once other more statistically significant wood anatomical variables were accounted for, the earlywood/latewood proportion did not significantly add to the model. In dried softwood, latewood has been shown to have greater permeability compared with earlywood (Hansmann et al. 2002). As such, the decreased proportion of latewood in samples taken from near the radial center of the log may be an additional factor that explains the lower permeability of wood from this zone. Latewood (in dried wood) is usually more permeable than earlywood because of its higher resistance to pit aspiration (Phillips 1933; Liese and Bauch 1967; Siau 1984, 1995; Richter and Sell 1992; Hansmann et al. 2002). In latewood, the higher resistance to pit aspiration is usually attributed to the greater rigidity of pit membranes and thicker cell walls (Comstock and Côté 1968). However, in some cases, such as with wood at higher moisture content, the earlywood has been shown to be more permeable than latewood (Erickson 1970).
Ray Frequency
Ray frequency (Fig. 5, Table 3), as number per mm2, was found to significantly decrease as tree stocking rate increased (p = 0.035). Genotype was also found to be significant (p = 0.034), with the F1 seedling having a significantly lower ray frequency than C887. C625 had a ray frequency in the middle that was not significantly different from either of the other genotypes. There did appear to be an increasing trend in ray frequency toward the outer wood, but it was not significant (p = 0.06). While rays have been shown to act as important flow paths for liquids (Côté 1963; Liese and Bauch 1967; McQuire 1970; Olsson et al. 2001), in this study there was no significant relationship between ray frequency and radial permeability.