Forests play a crucial role by providing a wide range of goods and services, including industrial wood, fuel wood, and non-wood forest products such as animal fodder, essential oils, and food Planted forests contribute to carbon sequestration, recreation opportunities like hunting and hiking, erosion control, and the rehabilitation of degraded lands For countries with low forest cover, establishing new forests through planting is essential to harness these multiple benefits.
From 1990 to 2015, the global area of planted forests expanded from 167.5 million hectares to 277.9 million hectares, with significant regional variations (Payn et al 2015) In Vietnam, the area of planted forests grew from 0.58 million hectares in 1985 to 4.13 million hectares by 2016 (Ministry of Agriculture and Rural Development of Vietnam 2017) These plantations not only provide materials for the pulp and paper industry but also contribute to environmental protection by reducing greenhouse gas emissions and alleviating rural poverty (Kim 2009) As wood resources diminish and processing costs rise, there is an increasing focus on timber production from plantations However, current wood resources are often under-utilized and poorly managed, highlighting the need for effective and sustainable practices to enhance timber quality and prevent further resource decline Research into wood properties is essential for the sustainable utilization of these forest resources.
Wood's biological origin leads to significant variability, which can be categorized into radial variation (from pith to bark) and axial variation (along the stem) for a given species This inherent variability poses challenges in accurately predicting wood performance, complicating its processing and utilization However, it also presents opportunities for genetic enhancement and diverse applications Understanding the variability within trees is crucial for improving wood quality and optimizing processing methods.
Table 1.1 The area of natural and planted forest in Vietnam from 1985 to 2016
A re a (m il li on ha )
Melia azedarach L is a deciduous tree belonging to the family of Meliaceae It is native to the
The M azedarach species, native to the Himalaya region of Asia, thrives in warm climates, poor soils, and seasonally dry conditions This tree typically reaches heights of 7 to 12 meters but can grow up to 45 meters under exceptional conditions Its rounded crown is complemented by long-petioled, alternate leaves that can measure up to 50 cm, featuring dark green leaflets on top and lighter green underneath, with serrate margins The tree produces small, fragrant flowers with five pale purple or lilac petals that grow in clusters The fruit, a marble-sized drupe, matures to a light yellow color and remains on the tree throughout winter, eventually becoming wrinkled and almost white.
M azedarach has significant potential in combating global warming due to its exceptional carbon-storing capabilities (Osei et al 2018) This species, along with other fast-growing trees, is utilized as a pulp material because of its high productivity (Ministry of Agriculture and Rural Development of Vietnam 2014) Additionally, the use of M azedarach wood in construction, such as for posts and beams, is anticipated to enhance its value (Hasegawa et al 2015) Various studies have explored the physical and mechanical properties of M azedarach wood, highlighting its versatility and benefits.
Research conducted in 2006 highlighted the diverse wood properties of M azedarach in Japan, indicating its potential as a new timber source Further studies by Venson et al in 2008 explored the physical, mechanical, and biological characteristics of M azedarach in Mexico, confirming its suitability as structural lumber when the right genotypes and clones are utilized.
In northern Vietnam, M azedarach is widely cultivated, primarily in short rotations of 5-6 years to provide raw materials for the pulp and particleboard industries Due to a decline in available wood resources and rising wood processing costs, there is a growing interest in utilizing wood from plantations.
M azedarach is known for its durability, resistance to termites and insects, and ease of workability (Nghia 2007) However, there has been a lack of research on the within-tree variations in wood properties, both radially and axially, of this important species planted in Vietnam, despite its versatile uses.
The rapid and nondestructive evaluation of wood properties is crucial for tree breeders and optimal timber utilization, as it enhances quality control by minimizing property variation in raw materials and by-products (Oliveira et al 2005) However, the inherent variability of wood, a natural material, poses challenges in accurately predicting its performance, complicating the application of nondestructive testing methods for wood quality assessment Research indicates that nondestructive techniques can effectively evaluate wood properties in small specimens, lumber, and even assess the quality of logs and standing trees (Wang et al 2001, Carter et al 2005, Ishiguri et al 2006, Ishiguri et al 2008) Among these, stress wave techniques are widely recognized for their accuracy, relying on the relationship between wave propagation characteristics—such as velocity and attenuation—and the mechanical and physical properties of the wood, as well as any irregularities present (Ross and Pellerin 1994, Montero et al 2015) Consequently, this study aims to explore the effectiveness of stress wave techniques in evaluating the wood properties of M azedarach cultivated in northern Vietnam.
The overall objective of this study was to investigate within-tree variations in fulfilment wood properties including information on wood structure as well as physical and mechanical properties of
M azedarach planted at two different sites in northern Vietnam The specific objectives were:
• To determine and estimate the within-tree variations in intrinsic wood properties such as growth ring width, wood density, fiber length, and microfibril angle
• To determine and estimate the within-tree variations in radial and tangential shrinkages
• To investigate the usefulness of a stress wave technique for prediction wood dimensional stability
• To determine and estimate the within-tree variations in mechanical properties
• To investigate the usefulness of a stress wave technique for prediction wood strength and stiffness
This article aims to inform wood industry experts about the sustainable processing of M azedarach logs for timber, highlighting the species' potential applications It emphasizes the effectiveness of rapid, nondestructive methods in predicting the dimensional stability and mechanical properties of M azedarach wood Additionally, the findings serve as a foundation for implementing machine grading of M azedarach timber in Vietnam, promoting better utilization of this resource.
Introduction
Wood serves numerous purposes, each requiring specific quality criteria, such as knot size, wood density, cell wall thickness, fiber length, and fiber content, which must be assessable and linked to the wood resource (Walker et al 1993) Due to its biological origin, wood exhibits significant variability, with properties differing between stands, trees, and even within small sections like growth rings (Sharma et al 2014) This inherent variability complicates the accurate prediction of wood performance, making efficient processing and utilization challenging The primary sources of variation in wood properties significantly impact its application across various industries.
• Between-tree variations on similar sites;
• Between-site variations of the same genotype growing in different geographic regions
Radial variation is the most recognized and extensively researched aspect of within-tree wood variability, typically observed as a pattern of changes in wood characteristics between core and outer wood, as well as juvenile and mature wood (Anoop et al 2014).
Within-tree variations in wood properties
Growth rings are layers of wood that form during a single growth period in plants, characterized by a contrast in color between light-colored earlywood and dark-colored latewood Earlywood, produced at the beginning of the growing season, consists of larger cells with thinner walls, while latewood, formed later in the season, features smaller cells with thicker walls Understanding these distinctions is essential for studying tree growth and wood properties.
The growth rate minimally impacts the wood properties of diffuse-porous species, as these species maintain a consistent vessel proportion throughout their annual rings regardless of growth rate In contrast, the density of ring-porous hardwoods is significantly affected by growth rate, with faster growth typically resulting in denser wood Moreover, the volume of vessel tissue in ring-porous hardwoods remains constant each year, leading to a smaller proportion of vessel tissue in wider growth rings (Walker et al 1993).
Ring width variation with age significantly influences wood properties, typically showing a trend of decreasing width as trees mature due to competition within the stand However, this pattern can be affected by soil and climatic conditions, as noted by several studies (Zobel and Sprague 1998, Matsumura et al 2006, Adamopoulos et al 2010, Kiaei et al 2016).
In a study conducted in 2006, it was found that the growth ring width of M azedarach in Japan decreases from the pith to the periphery This phenomenon is particularly evident in ring-porous hardwoods, where the growth rate, indicated by ring width, shows a positive correlation with wood density This correlation arises because the earlywood zone remains relatively constant annually, resulting in wider rings that consist of denser latewood with fewer vessels.
De Zeeuw 1980, Adamopoulos et al 2010) Nevertheless, other works failed to find any relationship between ring width and wood density (Van Eck and Woessner 1964, Taylor and Wooten 1973, Taylor
1977) A possible reason could be that the earlywood-latewood proportion, the percentage of cell wall material and the tissue proportions differ between the annual rings (Zang and Zhong 1992)
Ring width in trees varies significantly due to factors like environmental changes and competition for nutrients and sunlight Research has shown a strong correlation between ring width and growth conditions, indicating that variations in ring width serve as a detailed record of tree growth and health in response to climate changes At the edges of their geographical distribution, tree growth is primarily limited by temperature, with tree-ring data acting as a proxy for local temperature variations Additionally, studies have found that trees planted with wider spacing exhibit larger growth rings, as they experience less competition for vital resources such as nutrients, water, and sunlight, leading to accelerated growth compared to crowded trees.
2.2.2 Wood density and specific gravity
Specific gravity is the most crucial physical property of wood, closely linked to its mechanical and physical characteristics While specific gravity and wood density are often used interchangeably, they have distinct definitions that refer to the same attribute To accurately assess wood density, one must consider its moisture content, as this significantly impacts both mass and volume In physics, density is defined as mass per unit volume, but with wood, fluctuations in moisture complicate this relationship.
Wood density is a crucial property that significantly influences mechanical resistance and wear, making it a key factor in determining the technological quality of wood It is one of the first characteristics assessed when evaluating the potential value of a timber species This vital property contributes to the strength and stiffness of solid lumber and impacts the physical yields of fiber used in composite products, as well as in pulp and paper production.
Higher wood density is increasingly valued for its role in wood technology and its significance in estimating forest biomass and carbon storage The carbon stored in trees is influenced by both biomass and the carbon content of wood and other tissues Consequently, wood density and stem volume are crucial factors that can determine carbon storage at the tree level.
Wood density variation within hardwoods is complex, exhibiting a range of patterns throughout the stem Research by Knapic et al highlights the diverse radial variations in wood density among hardwood species.
A study in 2011 revealed a trend of decreasing wood density with cambial age in Quercus faginea, a pattern that is also observed in other Quercus species, including Q garryana and Q suber.
Wood density in certain Meliaceae species, such as Melia azedarach and Toona ciliate, shows a linear increase from the pith to the bark.
2009), and Swietenia macrophylla (Lin et al 2012) On the other hand, Wahyudi et al (2016) reported a nearly constant basic density of Azadirachta excelsa from pith to bark Besides, Ofori and Brentuo
Research from 2005 indicated that Cedrela odorata exhibits low wood density at the pith, which then increases sharply outward to a peak before gradually declining towards the outer edge In contrast, the axial wood density of Acacia melanoxylon, when planted in New Zealand, showed minimal variation with height.
Zealand (Nicholas and Brown 2002), while a significant increase of wood density with height was found in Acacia melanoxylon planted in Portugal (Machado et al 2014)
Corewood typically exhibits lower density compared to outerwood, primarily consisting of earlywood, which contributes to its reduced density The thin S2 layer in corewood results in a higher lignin content and a lower cellulose content than found in outerwood The transition from corewood to outerwood occurs between growth rings from the pith, varying by species and the specific property being analyzed.
The average basic density of trees in a stand is significantly influenced by environmental factors, while genetic control accounts for variations between individual trees regardless of their location When a wood property is influenced by environmental control, it experiences considerable changes with shifts in the environment Conversely, properties under strong genetic control may remain consistent despite varying growth conditions or may vary independently of environmental influences.
Both environmental factors and genetic control play crucial roles in influencing tree populations While the environment significantly impacts the average basic density of a population, genetic factors account for the variations among individual trees within that environment Consequently, introducing a species to a new region can lead to unpredictable outcomes, making it essential to cultivate the species on a smaller scale before committing to large-scale planting Many unforeseen results have occurred in the past, leading to several disappointments in such endeavors.
Nondestructive wood evaluation
Nondestructive evaluation (NDE) plays a crucial role in the forest products industry by enhancing structural product grading programs, leading to engineered materials with defined performance standards There is a growing interest in innovative and cost-effective NDE technologies for assessing wood quality Among these, stress wave propagation techniques have gained significant attention over the past few decades Research has explored the use of vibration modes in wood materials to evaluate quality for various forms, including standing trees, logs, and small-dimension wood specimens.
Numerous studies have demonstrated that the modulus of elasticity of lumber can be effectively predicted using the stress wave velocity of trees, logs, or small specimens (Nanami et al 1993; Ross et al 1997; Ikeda and Arima 2000; Wang et al 2001; Carter et al 2005; Ishiguri et al 2006; Ishiguri et al 2008) For instance, Wang et al (2001) identified a strong correlation between the dynamic modulus of elasticity of wood in trees, measured via the stress wave method, and the modulus of elasticity of small, clear specimens obtained from these trees using destructive testing methods, specifically in Tsuga heterophylla and Picea sitchensis Additionally, Ishiguri et al (2008) reported a significant positive correlation between the stress wave velocity of trees and the modulus of elasticity in static bending of lumber in Larix kaemferi.
Recent studies indicate that nondestructive methods can effectively evaluate the dimensional stability of structural lumber Research by Yamashita et al (2009a, 2009b) revealed a strong correlation between longitudinal-transversal shrinkage and the modulus of elasticity in logs of Cryptomeria japonica, assessed through tapping while in a green state.
Acoustic analysis has emerged as a valuable presorting criterion for identifying warp-prone Pinus ponderosa boards before kiln-drying, demonstrating a significant correlation between the boards' acoustic properties and grade loss due to warp Additionally, research by Dundar et al (2013, 2016) highlights the effectiveness of ultrasonic measurements in green conditions for accurately predicting transverse shrinkages in both softwood and hardwood species.
Conclusion of literature review
Hardwoods possess a more intricate structure than softwoods due to their diverse cell types and unique arrangements, resulting in a varied appearance and grain The differences in wood properties within the same species arise from genetic variations and ecological factors such as altitude, precipitation, temperature, soil, water, and nutrients These elements influence tree growth and development, with genetic structure being the primary determinant of wood properties, while ecological conditions directly or indirectly impact tree development, fertility, form, and height.
Within a species, wood variability can be categorized into radial variation, which occurs from pith to bark, and axial variation along the stem Radial variation is the most extensively studied aspect of wood characteristics, showcasing distinct patterns between inner and outer wood This significant variability complicates the prediction of wood performance and its processing efficiency However, it also presents opportunities for genetic enhancement and diverse applications Given the differences among species, it is essential to examine wood property variations in the context of specific species.
Abstract
The study examined the variation in intrinsic wood properties of 17–19-year-old Melia azedarach L trees in northern Vietnam, focusing on growth ring width (GRW), specific gravity (SG), fiber length (FL), and microfibril angle (MFA) Discs were collected from five heights above ground, revealing an average GRW of 7.44 mm, SG of 0.548, FL of 1.07 mm, and MFA of 14.65 degrees Significant differences (P < 0.05) were found in SG, FL, and MFA among trees and between the two sites Additionally, the longitudinal position of the samples had a notable impact on these wood properties.
Radial position significantly influences wood properties, with a high impact on growth ring width (GRW) and specific gravity (SG), contributing 52.58% and 58.49% to total variations, respectively Fiber length (FL) and SG increase from the pith to the bark, while GRW and microfibril angle (MFA) decrease in the same direction Notably, fiber length increment stabilizes between the 7th and 10th rings, which is crucial for log processing These findings provide essential insights for developing effective management strategies for structural timber production from M azedarach plantation trees in northern Vietnam.
Keywords: Melia azedarach, Growth ring width, Specific gravity, Fiber length, Microfibril angle
Introduction
Melia azedarach L is a deciduous tree belonging to the family of Meliaceae It is native to the
The M azedarach species, native to the Himalaya region of Asia, thrives in warm climates, poor soils, and seasonally dry conditions This tree typically reaches a height of 7–12 meters, but can grow up to 45 meters under exceptional circumstances Its leaves, measuring up to 50 cm long, are alternate, long-petioled, and odd-pinnate, featuring dark green upper surfaces and lighter green undersides with serrate margins The small, fragrant flowers have five pale purple or lilac petals and grow in clusters The fruit is a marble-sized drupe, light yellow when mature, that remains on the tree throughout winter, gradually becoming wrinkled and almost white.
M azedarach is primarily valued for its high-quality timber, which is easy to season as the planks dry without cracking or warping This wood is naturally resistant to fungal infections, making it ideal for manufacturing a variety of products, including agricultural implements, furniture, plywood, boxes, poles, and tool handles.
M azedarach, a versatile tree species, is utilized in cabinet making and construction, as noted by Nghia (2007) Its leaves serve as both green manure and insecticides, while in the Middle East and Assam, India, it is cultivated for fuel supply, particularly on tea estates (EL-Juhany 2011) In Vietnam, M azedarach is typically planted in short rotations of 5 to 6 years to provide raw materials for the pulp and particleboard industries.
Understanding wood properties is crucial for enhancing the quality of wood products Research indicates significant variability in wood properties, which can differ by stand, tree, circumference, radius, height, and even within growth rings While these variations offer opportunities for sustainable wood utilization, there is a lack of information regarding the wood properties of M azedarach in Vietnam This study aims to explore the variations in wood properties—such as growth ring width, specific gravity, fiber length, and microfibril angle—within individual trees, among different trees, and across various sites in northern Vietnam The findings will inform management strategies for the sustainable use of M azedarach wood in the region.
Materials and methods
Samples were collected from two M azedarach state-owned plantations in Vietnam The location and detailed information of the two sites are given in Table 3.1 The trees were around 17–
At 19 years old, the trees, planted at a density of 830 per hectare with a spacing of 4 m × 3 m, are sourced from seeds of natural forests nearby The species is expected to have a rotational age of 15-20 years Thinning was conducted at 3 and 6 years, removing 50% of the standing trees each time, with thinned trees repurposed as poles and branches for firewood In August 2016, six trees were harvested based on criteria such as straightness and absence of disease, felled at 15 cm above ground level, with diameter at breast height and total stem height measured prior to felling.
Cross-sectional discs, each 3 cm thick, were extracted from various heights (0.3, 1.3, 3.3, 5.3, and 7.3 m above ground) to analyze growth ring width (GRW) and specific gravity (SG) Additionally, a disc was taken from each tree at 1.3 m height to measure fiber length (FL) and the microfibril angle of the S2 layer of the cell wall (MFA).
Pith-to-bark strips measuring 30 mm tangentially and 15 mm longitudinally were extracted from the south side of the discs and air-dried These strips were then conditioned in a controlled environment at 20°C and 60% relative humidity until they reached a constant weight Subsequently, images of the conditioned strips were captured using a Canon MP-650 scanner connected to a computer, and GRW measurements were obtained through image analysis.
J software version 1.50i (Image-J) GRW of each tree was expressed as a mean value of all rings in that tree
Table 3.1 General characteristics of the study sites
Province Tuyen Quang Son La
Table 3.2 Age, diameter at breast height, and total stem height of sampled Melia azedarach trees
DBH diameter at breast height (at 1.3 m above the ground), H tree height a Measured by ring counting at the 15 cm above the ground
After measuring the growth ring width (GRW) of the wood samples, distinct growth rings were cut into individual rings for air-dry specific gravity (SG) measurement In instances where the rings were too small, two or more rings were combined for measurement Specific gravity, defined as the ratio of the density of wood to that of water at 4 ºC (Zobel and Van Buijtenen 1989), was determined using an electronic densimeter MD-300S from Alfamirage Co Ltd, Japan, with each sample taking approximately 10 seconds to measure.
3.3.5 Fiber length and microfibril angle
Pith-to-bark strips measuring 20 mm in thickness and 10 mm in length were extracted from discs at a height of 1.3 meters to assess fiber length (FL) and microfibril angle (MFA) The outer latewood layers from specific ring numbers (1, 2, 3, 5, 8, 10, 13, 15, and 17) were macerated in a 1:1 solution of 65% nitric acid and distilled water, combined with potassium chlorate, for a duration of five days Following this process, the samples were rinsed three times with distilled water, stained with safranin, and prepared for examination on glass slides.
30 fibers was measured by using a profile projector (V-12, Nikon) at a 50-fold magnification
Small blocks measuring 10 (R) × 10 (T) × 10 (L) mm were prepared from strips at ring numbers 1, 2, 5, 10, and 15 Radial sections of 8 µm thickness were cut using a microtome, macerated for 40 minutes, and cleaned with distilled water The sections underwent dehydration in a series of ethanol solutions, starting with 10% ethanol and progressing through 30%, 60%, 80%, and finally 100% ethanol, with each step lasting 5 minutes After dehydration, the sections were placed on a glass slide and immersed in a 3% iodine-potassium solution for 2-5 seconds A few drops of 60% HNO3 were added, and a coverslip was applied over the specimen The mean fiber area (MFA) of 25 fibers per small block was measured using a light microscope (Olympus DP70, Nikon) and Image J software version 1.50i.
Fig 3.1 Tree ring in cross section obtained from Melia azedarach (at 3.3 m height, Tree
3.3.6 Determination of fiber length increment (FLI )
Wood fiber length variations were modeled using a logarithmic relationship based on the distance from the pith to the annual ring The Fiber Length Index (FLI) was calculated according to the methodology established by Honjo et al (2005), with annual FLI values determined from ring to ring using a specific formula.
Where: FLI is the fiber length increment; ΔFL is the change in fiber length; and ΔRN is the change in ring number The FLI was then expressed as a percentage
An analysis of variance (ANOVA) was conducted on wood properties such as growth rate (GRW), specific gravity (SG), fiber length (FL), and microfibril angle (MFA) to evaluate the significance of site, tree, height level, and radial position effects Trees were treated as random effects, while other sources of variation were considered fixed effects Additionally, variance components for these sources of variation were estimated The statistical analysis utilized R software version 3.2.3.
Table 3.3 Model used in the analysis of variance
11 Residuals a Source of variations excluded in fiber length and microfibril angle analysis, since wood specimens were collected at 1.3 m stem height only
Results and discussion
The average growth ring width (GRW) of M azedarach was 7.44 mm, with a range of 6.53 to 8.64 mm among trees Variations in GRW were minimally influenced by site and tree-to-tree differences, accounting for only 0.28% and 1.58% of the total variation, respectively The radial variation of GRW was highly significant (P < 0.001) and contributed the most to the total variation at 52.58% GRW was largest near the pith and decreased rapidly with cambial age up to 5-6 years, stabilizing thereafter, though some trees exhibited fluctuations Additionally, the longitudinal variation of GRW was significant (P < 0.05) but contributed only 2.67% to the total variation, with mean GRW decreasing with height from 8.70 to 6.34 mm.
The current study's findings align with existing literature on M azedarach, particularly the work of Matsumura et al (2006), which examined wood properties in 17-year-old plantation trees in Japan They noted that growth ring width (GRW) was substantial near the pith up to 3 meters in height and stabilized beyond the fourth ring, irrespective of stem height GRW variability is influenced by several factors, including environmental changes (Zobel and Van Buijtenen, 1989) and plant spacing, with more widely spaced trees exhibiting accelerated growth due to reduced competition for resources such as nutrients, water, and sunlight (Zhu et al., 2000) In this study, consistent plant spacing across two sites resulted in no significant difference (P > 0.05) in mean GRW between them.
Table 3.4 Mean values per site and tree for selected wood properties of Melia azedarach
Category Growth ring width (mm)
2 7.15 ± 0.32 a 0.559 ± 0.004 a 1.12 ± 0.01 a 13.17 ± 0.14 b Mean 7.44 ± 0.24 0.548 ± 0.003 1.07 ± 0.01 14.65 ± 0.12 Mean values are followed by standard errors a,b,c,d Means with different superscript within a column significantly differ (P < 0.05)
Table 3.5 Variance components for growth ring width and specific gravity of Melia azedarach
Growth ring width Specific gravity
Residuals 250 11.34 9.33 df degrees of freedom, Var variance (%)
Fig 3.2 Variation of growth ring width in the radial and vertical directions of Melia azedarach in two sites
G row th ri ng w idt h (m m )
G row th ri ng w idt h (m m )
The study on the specific gravity (SG) of M azedarach revealed significant influences from site, tree-to-tree variations, stem height position, and radial position, with radial position contributing the most to the total variation at 58.49% The SG values ranged from 0.523 to 0.572, aligning with previous literature that reported values between 0.5 and 0.65, although some studies noted lower values Variations in SG may be due to factors such as tree age and geographic conditions, including latitude and climate Differences in altitude, annual rainfall, and soil types between study sites also likely impacted SG variations Further research is necessary to explore the genetic effects on SG variation in M azedarach in northern Vietnam.
The study found that the wood density of Melia azedarach increases consistently from the pith to the bark across all stem heights, aligning with previous research (Matsumura et al 2006, Nock et al 2009, El-Juhany 2011) This trend is also observed in other Meliaceae species, including Toona ciliata and Swietenia macrophylla In contrast, Wahyudi et al (2016) noted that Azadirachta excelsa exhibits a nearly constant basic density from pith to bark, while Ofori and Brentuo (2005) reported that Cedrela odorata shows a low density at the pith, which peaks before declining towards the bark.
38 previous reports, radial variation of wood SG depends on species For M azedarach planted in northern Vietnam, this study showed that wood SG increases gradually from pith toward outside
Significant differences (P < 0.05) in specific gravity (SG) were observed at various height levels, showing a general trend of decreasing SG from 0.3 to 3.3 meters, followed by a slight increase at the top This pattern aligns with findings from Kim et al (2008), who noted that SG was highest at the stump and decreased before increasing toward the top in Acacia mangium and Acacia auriculiformis in northern Vietnam Additionally, as illustrated in Fig 3.4, Matsumura et al (2006) found that the stem contains both low and high SG zones, with higher SG levels located in the outer stem area and lower levels in the inner region.
Fig 3.3 Variation of specific gravity in the radial and vertical directions of Melia azedarach in two sites
Spe ci fi c gr avi ty
Spe ci fi c gr avi ty
Tree stem maps illustrate the differences in growth ring width and specific gravity across various sites Each graph corresponds to an individual tree, with light colors indicating low specific gravity and dark colors representing high specific gravity.
The analysis of the microfibril angle (MFA) in the S2 layer of the cell wall of M azedarach wood fibers reveals significant differences (P < 0.001) across various sites, among individual trees, and along the radial direction Notably, radial position emerged as the primary source of variation, accounting for 26.20% of the total variation observed A declining trend in mean MFA was noted from the pith to the bark, aligning with previous findings reported by Matsumura et al (2006) and other related species.
(Ishiguri et al 2012, Todoroki et al 2015) High MFA in rings close to the pith ensure flexibility and protect the young shoots from wind damage (Walker and Butterfield 1995)
The findings on the fruit length (FL) of M azedarach indicate a mean FL of 1.07 mm, with variations ranging from 0.98 to 1.15 mm across different trees Significant sources of variation in FL include site, tree-to-tree differences within the site, and radial position, with radial position accounting for the majority of the variation at 77.83% Additionally, FL at breast height demonstrates an increase from the pith to the bark, attributed to the lengthening associated with cambial age (Matsumura et al 2006).
The length of fibers in the present study is in agreement to those in literature for M azedarach Abdul (2007) reported a 0.78–1.3 mm length for M azedarach fibers, while Richter and Dallwitz
(2000) reported an average FL of 0.8–1.65 mm Contrary, El-Juhany (2011) reported a lower average
FL of 0.742–0.792 mm for eight-year-old M azedarach The significant variation between trees in
Research indicates that while some studies show consistent fiber length (FL) among trees (Leal et al 2006), others report slight variations (Gartner et al 1997) This tree-to-tree variability in FL may stem from the unique capabilities of individual trees to generate fibers of differing lengths compared to their neighbors (El-Juhany 2011).
Table 3.6 Variance components for fiber length and microfibril angle of Melia azedarach
Residuals 1566 13.12 720 45.86 df degrees of freedom, Var variance (%)
Fig 3.5 Radial variation of MFA for two different sites of Melia azedarach
M ic rof ibr il a ngl e ( o )
Fig 3.6 Radial variation of fiber length for two different sites of Melia azedarach y 1 = 0.2196ln(x 1 ) + 0.6309
3.4.5 Stabilizing point of fiber length increment
The logarithmic regression analysis of fiber length (FL) for site 1 and site 2 revealed distinct radial patterns, represented by the equations y1 = 0.2196ln(x1) + 0.6309 and y2 = 0.1784ln(x2) + 0.8039, where y denotes fiber length and x indicates ring number Additionally, the fiber length index (FLI) was calculated and illustrated, indicating that FLI began to stabilize between the 7th and 10th rings for both sites, suggesting that wood beyond the 7th ring contains comparatively longer fibers.
3.4.6 Implications for wood utilization of M azedarach in northern Vietnam
The length of fibers in wood is crucial for optimizing timber quality and value, with mature wood exhibiting long fiber length (FL), high specific gravity (SG), and low microfibril angle (MFA) being ideal for structural applications In this study, FL was found to increase from the pith to the bark, with wood beyond ring number 7 displaying favorable properties for structural use Additionally, M azedarach trees from site 2 (Son La provenance) showed superior SG, longer FL, and lower MFA compared to those from site 1 (Tuyen Quang provenance), suggesting that site 2 or similar environments are preferable for M azedarach plantations in northern Vietnam Further research is needed to assess the impact of seed source and mechanical properties for sustainable wood utilization of M azedarach in the region.
Fig 3.7 Fiber length increment with cambial age of Melia azedarach in Vietnam (Bars mean standard error)
Fi be r l engt h inc re m ent (% )
Conclusions
The study revealed significant variations in wood properties within M azedarach trees, with longitudinal position affecting growth ring width (GRW) and wood specific gravity (SG) Specifically, mean GRW decreased with higher stem positions, while wood SG showed a decline from the stump to intermediate stem before slightly increasing towards the top Radial position significantly impacted all wood properties, contributing the most to total variation; fiber length (FL) and wood SG increased from the pith to the bark, whereas GRW and microfibril angle (MFA) decreased in the same direction Notably, fiber length index (FLI) stabilized beyond the seventh ring from the pith, highlighting important considerations for processing logs in northern Vietnam.
Abstract
This study quantified variations within tree stems in tangential shrinkage (αT), radial shrinkage
(αR), and tangential/radial shrinkage ratio (αT/αR) of Melia azedarach grown in two different sites in northern Vietnam The overall values of αT, αR, and αT/αR were 7.05%, 4.38%, and 1.64, respectively
The variation in αT and αR increased from pith to bark, showing a consistent trend across both sites Radially, the αT/αR ratio significantly decreased from 10% to 50% of the radial length from the pith before stabilizing outward Transverse shrinkage variation with height was minimal and statistically insignificant Strong positive correlations were observed between transverse shrinkage and basic density (BD), indicating that selecting for higher wood density may increase transverse shrinkage Additionally, αT and αR exhibited significant positive linear relationships with both acoustic wave velocity (VL) and dynamic modulus of elasticity of log (DMOElog), suggesting that lumber with greater transverse shrinkage could potentially be sorted using the stress wave method for M azedarach in northern Vietnam.
Keywords: Melia azedarach, Transverse shrinkage, Nondestructive evaluation, Radial position
Introduction
Wood's biological origin contributes to its significant variability, which can be categorized into radial variation (from pith to bark) and axial variation (along the stem) This inherent variability complicates the prediction of wood performance, posing challenges for efficient processing and utilization However, it also presents opportunities for genetic enhancement and diverse applications Thus, gaining a deeper understanding of wood variability within a tree is crucial for improving wood quality and optimizing processing methods.
Dimensional stability and warping are critical issues in wood drying and processing, primarily due to the anisotropic shrinkage gradients in both radial and tangential directions (Wang et al 2008) The shrinkage of wood, a significant physical property, impacts its usability in various products, with certain wood species benefiting from minimal shrinkage Volume change is not uniform; hardwoods can experience tangential shrinkage of 6 to 12% when transitioning from green to oven-dry conditions, while radial shrinkage is approximately half of the tangential shrinkage for the same specimen This differential shrinkage is a key factor contributing to shape distortion during lumber seasoning and its subsequent use (Dundar et al 2016).
Wood density is the primary factor influencing shrinkage, as the amount of shrinkage correlates directly with the moisture lost from the cell wall (Skaar 1988) Understanding the relationship between transverse shrinkage and wood density is crucial for effective lumber management.
Research indicates that sorting lumber by shrinkage properties can enhance drying yields while reducing costs and energy consumption Additionally, studies propose that acoustic wave techniques could effectively evaluate the dimensional stability of structural lumber.
Research by Yamashita et al (2009a, 2009b) established significant correlations between longitudinal and transversal shrinkage and the dynamic modulus of elasticity in green Sugi (Cryptomeria japonica) logs Wang and Simpson (2006) demonstrated the effectiveness of acoustic analysis as a presorting method to identify warp-prone Ponderosa pine (Pinus ponderosa) boards prior to kiln-drying, revealing a statistically significant relationship between the acoustic properties and grade loss due to warping Furthermore, Dundar et al (2013, 2016) highlighted the potential of ultrasonic measurements in the green condition for accurately predicting transverse shrinkages in both softwood and hardwood species.
Melia azedarach, a deciduous tree belonging to the family of Meliaceae, is native to the
The M azedarach species, native to the Himalaya region of Asia, thrives in warm climates and poor soils, and has become naturalized in many subtropical and tropical areas worldwide In Vietnam, it is widely planted across northern provinces The growing demand for plantation wood, driven by dwindling natural wood resources and rising processing costs, has increased interest in M azedarach wood in the furniture industry This wood is noted for its fine grain, durability, termite and insect resistance, and ease of workability However, there is a lack of information on its dimensional stability, particularly concerning variations in transverse shrinkage within the tree, including radial and axial differences.
This study aimed to explore the variation in transverse shrinkage properties of M azedarach trees planted in northern Vietnam, focusing on three key objectives: first, to analyze the radial and axial patterns of wood shrinkage from pith to bark at various heights within the trees; second, to examine the correlation between transverse shrinkage and basic density (BD); and third, to identify nondestructive parameters that can predict transverse shrinkage for effective lumber sorting Additionally, the research assessed the physical properties of M azedarach timber to evaluate its quality for grading purposes.
Material and methods
Sample trees for this study were collected from state-owned plantations in Tuyen Quang (Northeast) and Son La (Northwest) in northern Vietnam Six trees, aged approximately 17–19 years and selected for their straight trunks, normal branching, and absence of diseases or pests, were chosen These trees were planted at a density of 830 trees per hectare, spaced 4 m × 3 m apart, from seedlings sourced from nearby natural forests The trees were felled at 15 cm above soil level, and 50-cm-long logs were extracted from various heights (0.3, 1.3, 3.3, 5.3, and 7.3 m) Prior to felling, the north and south sides of each tree were marked The selected trees were consistent with those used in previous studies.
4.3.2 Dynamic modulus of elasticity of log (DMOE log )
The DMOElog was assessed using the stress wave method in the green condition for each log, employing the Fakopp microsecond timer from Hungary to measure the acoustic wave propagation time through the log's axis This device features two electrodes, a transducer and a receiver, which are positioned at the midpoint between the pith and bark on either side of the log An acoustic signal is generated by striking the start transducer with a hammer, and the propagation time is measured five times for each log, with the average value serving as the experimental result The velocity of acoustic wave propagation is calculated as the length of the log divided by the propagation time, while the green density of the logs is determined by the ratio of green weight to green volume Finally, DMOElog in the direction parallel to the grain is derived from a straightforward relation.
DMOElog – dynamic modulus of elasticity of log (GPa), ρ – green density of log (kg/m 3 ),
Fig 4.1 Method of measuring dynamic modulus of elasticity of log (DMOElog) and cutting specimens from each tree
4.3.3 Basic density and transverse shrinkage
In the conducted experiment, discs with a thickness of 3 cm were prepared to assess transverse shrinkage and bulk density (BD) Measurements of tangential shrinkage (αT), radial shrinkage (αR), and BD were performed following the Japanese Industrial Standards (JIS 2000) at various radial and height positions Specimens measuring 30 × 30 × 5 mm (Tangential × Radial × Longitudinal) were extracted at a rate of 10 for analysis.
The study involved 180 small clear wood specimens, measured at various heights (0.3, 1.3, 3.3, 5.3, and 7.3 m) from the pith, covering 50% and 90% of the radial length on both the North and South sides To mitigate drying stress effects on shrinkage, the samples were conditioned at 20 °C and 60% relative humidity until they reached constant weight Subsequently, they were oven-dried at 60 °C overnight and at 103 °C for two nights to determine their oven-dry weight and dimensions Each specimen had centerlines drawn for length measurements, with dimensions in both radial and tangential directions measured using a Mitutoyo digital micrometer in both green and oven-dried states The tangential/radial shrinkage ratio (αT/αR) was calculated by dividing αT by αR, and the basic density (BD) was determined as the oven-dry weight per green volume Shrinkage calculations were performed based on these measurements.
𝑙 , ×100 where: l g: length of centerline which was measured at green condition (mm), l o: length of centerline which was measured at oven-dry condition (mm), α: shrinkage from green to oven-dry (%)
An analysis of variance (ANOVA) was conducted on wood shrinkage properties (α T, α R, and α T/α R) to evaluate the effects of site, tree, height level, and radial position, utilizing the model outlined in Table 4.1 In this analysis, trees were treated as random effects while other variations were fixed effects, with variance components estimated accordingly Additionally, differences in radial and height positions within the stem were assessed using the Tukey-Kramer HSD test All statistical evaluations were carried out using R software version 3.2.3.
The grade yield for the specimens was evaluated according to the grading standards for the physical properties of timbers from Southeast Asia and the Pacific regions, as established by the Forestry and Forest Products Research Institute (FFPRI) in Japan (1975).
Table 4.1 Model used in the analysis of variance
Shrinkage from green to oven dry Tangential (%) Radial (%)
Table 4.2 Grading standard of physical properties of timber from Southeast Asia and Pacific regions by Forestry and Forest Products Research Institute (1975)
Results and discussion
4.4.1 Basic density and transverse shrinkage
Table 4.3 presents the statistics for bulk density (BD) and transverse shrinkage from green to oven-dry conditions for M azedarach trees in northern Vietnam, showing average values of 0.43 g/cm³ for BD, 7.05% for tangential shrinkage (αT), 4.38% for radial shrinkage (αR), and a ratio of αT to αR of 1.64 across six trees at various stem heights For comparative analysis, equivalent properties of M azedarach trees from different countries, as reported by Botero (1956), Coronel (1989), Pramana (1998), and Venson et al (2008), are listed in Table 4.4.
Radial variation is the most studied aspect of within-tree variability in wood, significantly influencing shrinkage characteristics Research indicates that radial position affects shrinkage (P < 0.001), accounting for 42.38% of total variation in αT and 45.54% in αR As shown in Fig 4.2, αT and αR increase from pith to bark, while the ratio αT/αR decreases outward This consistent radial pattern of transverse shrinkage is maintained across different tree heights and sites, with no significant differences observed The contrasting trends of αR and αT/αR suggest that the variation in αR is greater than in αT Previous studies support these findings, showing that transverse shrinkage increases from pith to bark in species like Cedrela odorata and Populus euramericana However, discrepancies exist among other hardwoods, such as Swietenia macrophylla, which showed no significant differences, and Acacia species that exhibited varying trends This study confirms that in M azedarach from northern Vietnam, αT and αR increase radially, indicating that transverse shrinkage patterns differ by species due to variations in wood density, microfibril angle, and anatomical structures The subsequent section will explore the relationship between transverse shrinkage and bulk density (BD) to assess its impact on αT and αR.
Table 4.3 Variation in basic density (BD), tangential shrinkage (αT), radial shrinkage (αR), and tangential/radial shrinkage ratio (αT/αR) within stem and between sites of Melia azedarach
Mean 180 0.43 ± 0.01 7.05 ± 0.06 4.38 ± 0.06 1.64 ± 0.02 n number of wood specimen
Mean values are followed by standard errors a,b,c Means with different superscript within a column significantly differ (P < 0.05)
Table 4.4 Shrinkage parameter of Melia azedarach planted in northern
Vietnam compared with corresponding data of Melia azedarach planted from other provenances
Shrinkage from green to oven dry α T (%) α R (%) α T /α R This study (17-19) 6.73 – 7.37 4.10 – 4.67 1.68 – 1.60
Age of tree was given in parentheses; a Pramana (1998); b Venson et al (2008); c Botero (1956); d Coronel (1989)
Table 4.5 Variance components for tangential shrinkage (α T ), radial shrinkage (α R ), and tangential/radial shrinkage ratio (α T /α R ) of Melia azedarach
% Site (S) 1 0.001 15.39 0.001 13.93 0.017 3.17 Tree/Site (T/S) 4 0.326 2.21 0.358 2.12 0.044 5.26 Height level (L) 4 0.163 3.64 0.712 1.20 0.092 4.43
Residuals 90 19.02 12.80 24.71 df degrees of freedom, Var variance (%)
Fig 4.2 Radial variation in tangential shrinkage (α T ), radial shrinkage (α R ), and tangential/radial shrinkage ratio (αT/αR) for two different sites of Melia azedarach (Bars mean standard deviation)
MFA significantly influences wood shrinkage, with the angle between microfibrils and the wood's longitudinal axis affecting shrinkage rates (Barber and Meylan 1964) Strong, moisture-resistant crystalline cellulose leads to minimal shrinkage parallel to microfibrils, while decreasing MFA results in increased transverse shrinkage and reduced longitudinal shrinkage A previous study by Duong et al (2017) revealed that MFA in M azedarach decreases from the pith (16°) to the bark (12°), contrasting with the observed trends in transverse shrinkage This within-tree variation in MFA likely contributes to greater transverse shrinkage in outer wood compared to inner wood Additionally, the earlywood-latewood ratio, characterized by larger vessels and thinner fiber walls in earlywood compared to latewood, may further influence wood shrinkage variability Further research is necessary to explore cell morphology and earlywood-latewood ratios from pith to bark in M azedarach wood.
Cutting logs into thin discs facilitates the understanding of drying stresses by enabling comparisons of drying strains in tangential and radial directions (Fu et al 2016) The shrinkage ratio is a crucial parameter for assessing wood drying performance, with significant differences between tangential and radial shrinkage impacting internal stresses during the drying process (Dahlblom et al 1999) For M azedarach, tangential shrinkage is generally 1.5 to 2.5 times greater than radial shrinkage In this study, the αT/αR ratio decreased significantly from 10% to 50% of the radial length from the pith before stabilizing toward the outer regions, with values of 1.76 near the pith compared to 1.59 and 1.56 in the middle and outer parts, respectively This disparity in radial and tangential shrinkage is linked to the differing wood densities of earlywood and latewood, where latewood cells, having higher density, exhibit greater shrinkage In the radial direction, earlywood and latewood shrink independently, resulting in total shrinkage that reflects the weighted mean of both components, while tangential shrinkage is predominantly influenced by latewood changes.
Anagnost et al (2005) found that the microfibril angle (MFA) in the radial walls of Acer saccharum and Prunus serotina is comparable, while it is significantly larger in Drimys winteri compared to the tangential walls This variation is attributed to the presence of bordered pits in the radial walls, which cause the microfibrils to deviate around them.
ANOVA results showed that height level was not a significant source of variation, while site significantly influenced α T, α R, and α T/α R (P < 0.05) Notably, α T and α R values were higher at site 2 compared to site 1 Variations in wood properties among the same species arise from different genotypes and growth conditions Although no previous studies have addressed the impact of these factors on transverse shrinkage in M azedarach, research on other hardwood species has explored the effects of seed source and growth conditions on this aspect Montes et al (2007) reported significant genetic variation in wood shrinkage.
Calycophyllum spruceanum exhibits significant variations in shrinkage properties influenced by site conditions, as highlighted by Yang et al (2002) in Eucalyptus globulus Labill This study reveals that differences in transverse shrinkage between two locations may stem from factors such as altitude, annual rainfall, and soil types Future research should investigate the impact of seed source on wood shrinkage in M azedarach, as previous studies have noted varying shrinkage results for this species across different countries Such discrepancies may arise from the unique characteristics of forest sites in each study, indicating that site conditions play a crucial role in determining shrinkage, drying stress, and overall wood properties.
68 incidence of drying defects such as cracks or deformations might differ among lumber sawn from different sites
4.4.2 Relationships between transverse shrinkage and basic density
Linear regression analysis was conducted to evaluate the correlation between transverse shrinkage and bulk density (BD), with results detailed in Table 4.6 and Fig 4.3 The analysis revealed significant positive linear correlations between BD and both α T and α R at a confidence level of 0.001 for each site as well as when combined For the combined sites, the correlation coefficients were r = 0.72 for α T and r = 0.82 for α R Additionally, a negative correlation was observed between the ratio α T /α R and BD (r = -0.53, P < 0.001), indicating an inverse relationship compared to α T and α R.
Wood density is widely acknowledged to reflect shrinkage property Istikowati et al (2014) found strong positive correlations of BD with αT (r = 0.83), and αR (r = 0.83) in Artocarpus elasticus,
Research by Pliura et al (2005), Wu et al (2006), Kord et al (2010), and Sadegh et al (2012) indicates significant positive correlations between wood density (BD) and both α T and α R in various species, including Poplar hybrid crosses, Eucalypt species, Populus euramericana, and Tamarix aphylia These findings imply that selecting for higher wood density in M azedarach, particularly in northern Vietnam, may result in increased wood transverse shrinkage.
Table 4.6 Relationship between transverse shrinkage and basic density (BD) for each site and combined sites of Melia azedarach
180 12.63 1.56 0.72 *** 13.31 - 1.41 0.82 *** - 2.43 2.69 - 0.53 *** n number of wood specimen α T tangential shrinkage, α R radial shrinkage, α T /α R tangential/radial shrinkage ratio r correlation coefficient; *** P < 0.001
Fig 4.3 Relationships between transverse shrinkage (αT, αR, and αT/αR) and basic density
(BD) for combined sites of Melia azedarach (Triple asterisk: P < 0.001) y = 12.63x + 1.56 r = 0.72 ***
The results of linear regression analysis for the transverse shrinkage (α T , α R , and α T /α R ) and
In green condition, the acoustic wave measurements of V L, as shown in Table 4.7 and Fig 4.4, reveal positive correlations between αT and αR with V L, with respective r values of 0.47 and 0.45 This suggests that acoustic wave measurement can serve as an effective nondestructive method for predicting the transverse shrinkage of M azedarach in northern Vietnam Previous studies by Dundar et al (2013, 2016) have also indicated that ultrasonic velocity is a strong predictor of transverse shrinkage, demonstrating significant correlations in both softwood and hardwood species.
Lieblein and Castanea sativa Mill.), respectively
The correlation coefficients between DMOElog and transverse shrinkage (r values of 0.62 for αT and 0.59 for αR) were stronger than those between V L and transverse shrinkage When combining acoustic wave velocity (V L) and green density (ρ) of the log, the coefficients of determination significantly increased, with V L explaining 21.8% of αT and 19.9% of αR alone, which rose to 39.1% for αT and 34.3% for αR when used together This indicates that better predictions of transverse shrinkage can be achieved using both V L and ρ in calculating DMOElog under green conditions Additionally, the αT/αR ratio was negatively correlated with both V L and DMOElog, although not significantly Previous studies, such as those by Yamashita et al (2009b) and Dundar et al (2016), also support the notion that combining modulus of elasticity measurements with other parameters enhances the prediction of transverse shrinkage in various wood species.
Table 4.7 The coefficients of correlation (r) and coefficients of determination
(R 2 ) between transverse shrinkage (α T , α R , and α T /α R ) and acoustic wave velocity (V L), dynamic modulus of elasticity of log (DMOElog) for the combined sites
Fig 4.4 Relationships between transverse shrinkage (α T , α R , and α T /α R ) and acoustic wave velocity (V L); transverse shrinkage and dynamic modulus of elasticity of log (DMOElog) for combined sites of
Melia azedarach exhibits relationships at 10%, 50%, and 90% radial positions from the pith, represented by long dash dot, round dot, and long dash lines The thick solid lines, along with the coefficients of determination (R²), illustrate the relationships for combined samples across three positions.
4.4.4 Grade yield of shrinkage properties
The grade yield analysis for α T and α R, based on the FFPRI grading standard for physical properties of timbers from Southeast Asia and the Pacific regions, reveals that site 1 exhibited the highest yield for grade II at 68%, with grade I and III at 28% and 4%, respectively Similarly, site 2 showed a predominant yield for grade II at 78%, followed by grade III.
In a study of M azedarach wood quality in northern Vietnam, it was found that αR specimens predominantly fell into grade III, with frequencies of 62% and 59% for site 1 and site 2, respectively Notably, there were no grade yields for αT in grades IV and V at either site The research indicates that the wood can be classified as grade II for tangential shrinkage and grade III for radial shrinkage, according to the grading standards for Southeast Asia and Pacific timber properties These findings offer valuable insights for wood processors in the drying and furniture industries and promote the sustainable utilization of M azedarach trees in the region.
Figure 4.5 illustrates the allocation of specimen grades based on tangential shrinkage (αT) and radial shrinkage (αR) for both sites, in accordance with the grading standards for the physical properties of timbers from Southeast Asia and Pacific regions established by FFPRI.
Conclusions
This study revealed that radial position significantly affects shrinkage variations (P < 0.001), with αT and αR increasing from pith to bark in a consistent pattern across both sites Notably, αT/αR decreased from 10% to 50% of the radial length from the pith before stabilizing outward The transverse shrinkage variation with height was minimal and not statistically significant A strong correlation between bulk density (BD) and transverse shrinkage indicates that BD is a reliable predictor of dimensional stability Both αT and αR positively correlate with longitudinal velocity (V L) and dynamic modulus of elasticity (DMOElog) The best predictions for M azedarach shrinkage were achieved by combining V L and density (ρ) through DMOElog calculations, suggesting that acoustic wave measurements in green conditions effectively predict transverse shrinkage in M azedarach in northern Vietnam Consequently, this study establishes a basis for sorting lumber with significant drying shrinkage using the stress wave method, allowing M azedarach wood from northern Vietnam to be classified into grade II for tangential shrinkage and grade III for radial shrinkage according to Southeast Asian and Pacific timber grading standards.