NC State
BioResources
Kwon, S. L., Jung, J.-H., and Lee, H.-M. (2026). "Anatomical characteristics and natural durability of three Korean diffuse-porous hardwood species," BioResources 21(4), 11699–11718.

Abstract

Natural durability is an important wood property that influences utilization potential. This study investigated the anatomical characteristics and natural durability of three Korean diffuse-porous hardwood species, namely Betula platyphylla, Liriodendron tulipifera, and Quercus salicina, to assess whether measurable anatomical traits could serve as indicators of durability. Traits were measured using microscopic image analysis, and natural durability was evaluated using fungal decay and termite resistance tests. Associations between anatomical traits and durability were explored using multiple factor analysis (MFA). Among the species examined, L. tulipifera exhibited the lowest fungal weight loss and highest termite resistance, whereas B. platyphylla showed the greatest susceptibility to fungal decay and termite attack. The MFA suggested exploratory associations, based on species-level data from three species, between fungal decay resistance and vessel-related anatomical traits, whereas termite resistance showed limited correspondence with the anatomical variables examined. These results suggest an exploratory relationship between vessel dimensions and variation in fungal decay susceptibility among the three species; anatomical traits corresponded only weakly with overall natural durability. The findings may support future wood informatics research and facilitate the development of data-driven models that integrate anatomical, physical, and chemical datasets to predict wood properties.


Download PDF

Full Article

Anatomical Characteristics and Natural Durability of Three Korean Diffuse-porous Hardwood Species

Sun Lul Kwon , Jae-Hee Jung, and Hyun-Mi Lee *

Natural durability is an important wood property that influences utilization potential. This study investigated the anatomical characteristics and natural durability of three Korean diffuse-porous hardwood species, namely Betula platyphylla, Liriodendron tulipifera, and Quercus salicina, to assess whether measurable anatomical traits could serve as indicators of durability. Traits were measured using microscopic image analysis, and natural durability was evaluated using fungal decay and termite resistance tests. Associations between anatomical traits and durability were explored using multiple factor analysis (MFA). Among the species examined, L. tulipifera exhibited the lowest fungal weight loss and highest termite resistance, whereas B. platyphylla showed the greatest susceptibility to fungal decay and termite attack. The MFA suggested exploratory associations, based on species-level data from three species, between fungal decay resistance and vessel-related anatomical traits, whereas termite resistance showed limited correspondence with the anatomical variables examined. These results suggest an exploratory relationship between vessel dimensions and variation in fungal decay susceptibility among the three species; anatomical traits corresponded only weakly with overall natural durability. The findings may support future wood informatics research and facilitate the development of data-driven models that integrate anatomical, physical, and chemical datasets to predict wood properties.

DOI: 10.15376/biores.21.4.11699-11718

Keywords: Diffuse-porous hardwood; Fungal decay; Multiple factor analysis; Natural durability; Termite resistance; Vessel dimensions; Wood anatomy

Contact information: Department of Forest Products and Industry, National Institute of Forest Science, 57 Hoegi-ro, Dongdaemun-gu, Seoul, 02455, Republic of Korea;

* Corresponding author: [email protected]

INTRODUCTION

Diffuse-porous wood species are characterized by vessels that are relatively uniform in size throughout the annual growth ring. Such wood species are predominantly found in the tropical and temperate regions of America, Africa, and Asia (Wardle 1991). Because of their homogeneous vessel distribution, diffuse-porous wood species usually exhibit a relatively uniform appearance and more homogeneous mechanical features than ring-porous species (Zhan et al. 2015; Esteban et al. 2024). Such characteristics facilitate the prediction of mechanical properties and contribute to relatively consistent stress transfer within wood products (de Borst and Bader 2014; Zhan et al. 2015).

These features are advantageous for the manufacture of engineered wood products, such as glued laminated timber and laminated veneer lumber, where uniform adhesive performance and strength distribution are beneficial (Yelle and Stirgus 2016; Morin-Bernard et al. 2020). Furthermore, the relatively uniform anatomical structure of diffuse-porous wood species has recently attracted attention in the development of high-strength densified wood materials (Mania et al. 2024).

In Korea, three diffuse-porous hardwood species, namely Betula platyphylla, Liriodendron tulipifera, and Quercus salicina, are recommended for afforestation (Korea Forest Service 2025). Previous studies have examined various properties of B. platyphylla and L. tulipifera, including wood anatomy, mechanical performance, durability, and acoustic behavior (Kang et al. 2011; Zhan et al. 2015; Erdene-Ochir et al. 2021; Esteban et al. 2024; Jung et al. 2025). However, quantitative information linking the anatomical characteristics with natural durability among these diffuse-porous hardwood species remains limited, particularly for Q. salicina.

The natural durability of wood varies considerably among species and is primarily influenced by the type and composition of its extractives present in the wood (Rodrigues et al. 2011; Amusant et al. 2014). Certain extractives possess antimicrobial and insecticidal properties that improve resistance to wood-decaying fungi and termites. Previous studies have demonstrated that the removal of extractives substantially decreases resistance to biological deterioration (Scheffer and Cowling 1966; Smith et al. 1989; Kirker et al. 2013). Furthermore, durability is determined by the quantity, chemical profile, and spatial allocation of extractives, leading to considerable variation even within the same species (Scheffer and Cowling 1966; Kirker et al. 2013).

In addition to chemical composition, anatomical characteristics, such as vessel structure, cell wall thickness (CWT), tissue arrangement, and wood density, have recently been recognized as important factors influencing wood durability (Dadzie and Amoah 2015; De Ligne et al. 2021). Previous studies have suggested that parenchyma content and the vessel-to-fiber ratio are associated with resistance to brown-rot fungi, whereas other anatomical traits are closely related to wood density and mechanical performance (De Ligne et al. 2021). Based on these findings, we expected differences in vessel and fiber characteristics, particularly vessel dimensions and cell wall thickness, to be associated with variation in natural durability.

Traditionally, the anatomical characteristics of wood have been extensively used for species identification and taxonomic studies. Recent advances in digital microscopy and image analysis have improved the accuracy and reproducibility of quantitative anatomical assessments (von Arx et al. 2016; Frigo et al. 2024). These developments have increased the availability of anatomical datasets that can be integrated with data-driven approaches to predict wood properties and performance (He et al. 2024; Keret et al. 2024). Consequently, understanding how the anatomical characteristics of wood relate to natural durability has become increasingly important in scientific research and functional applications.

Therefore, the anatomical characteristics and natural durability of three diffuse-porous hardwood species commonly used in Korean forestry were investigated. Specifically, this study examined whether variation in quantitative anatomical traits was associated with fungal decay and termite resistance, and we assessed the potential of these traits as practical indicators of natural durability.

EXPERIMENTAL

Wood Materials

Betula platyphylla Sukaczev, Liriodendron tulipifera Linnaeus, and Quercus salicina Blume were used in this study; the sampling sites were Hongcheon-gun in Gangwon-do, Gangjin-gun in Jeollanam-do, and Sacheon-si in Gyeongsangnam-do, respectively (Table 1). For each species, a single tree was sampled at the corresponding site. A wood disk was obtained from each stem at breast height (approximately 1.3 m), and its cross-sectional surface was polished for subsequent analyses. The collected disks contained 35, 28, and 16 growth rings for B. platyphylla, L. tulipifera, and Q. salicina, respectively. The cross-sections and the edge- and flat-grain surfaces of three wood species are shown in Fig. 1.

Cross-section, edge-grain, and flat-grain surfaces of three hardwood species

Fig. 1. Cross-section, edge-grain, and flat-grain surfaces of three hardwood species

Wood blocks for anatomical trait analysis were taken from the mature-wood zone of each disk. The exact number of blocks prepared per species and their precise radial positions within the mature-wood zone (near-pith, intermediate, or near-bark) were not documented. Because the Q. salicina disk contained no mature wood (defined as ≥ 20 growth rings from the pith), blocks for this species were obtained from juvenile wood instead. Specimens for the fungal decay test were taken from sapwood, whereas those for the termite test were taken from heartwood. These specimens were cut from the same breast-height disk, but their exact radial positions were not recorded.

Table 1. Site Characteristics for the Three Sampling Locations

Site Characteristics for the Three Sampling Locations

Annual Ring Width and Anatomical Trait Analysis

The cross-section of the wood disc was divided into four sections with two perpendicular lines and scanned using an Epson Expression 10000XL scanner (EU-88, Seiko Epson Corp., Suwa, Japan) to obtain an image. The annual ring width (ARW) was measured from the pith to the bark using DeltaPix v5.3.11.

Anatomical traits were analyzed using the wood blocks (10 mm × 10 mm × 10 mm in the longitudinal (L), tangential (T), and radial (R) directions) prepared from the mature-wood zone of each disk (juvenile wood for Q. salicina). Following Park et al. (2024a, 2024b), the blocks were sectioned longitudinally with a razor blade. Subsequently, the sections were macerated in Franklin’s solution (a 1:1 (v/v) mixture of 30% H2O2 and 95% CH3COOH) and heated at 80 °C on a heating mantle for 48 h. The macerated vessel and fiber elements were stained with methylene blue to improve cellular contrast, and their lengths were measured using DeltaPix software v5.3.11. For measurements of cell wall thickness (CWT) and widths in the radial (Drad) and tangential (Dtan) directions, thin transverse sections (10 to 20 µm thick) were prepared using a sliding microtome, stained with 1% safranin, dehydrated through a graded ethanol series, and permanently mounted in Canada balsam for microscopic examination. The traits of vessels and fibers were measured on the stained sections with the same software. For each of the eight anatomical traits, 30 measurements were obtained from the single tree sampled per species, resulting in 240 anatomical measurements per species. These measurements therefore reflect within-tree variation rather than independent tree-level biological replicates.

Durability Test Against Wood-decaying Fungi

The durability test against wood-decaying fungi was conducted according to the guidelines outlined in the Assessment Manual for Domestic Wood Characteristics (Park et al. 2024c). The guidelines were based on KS F 2213 (2018). The fungus was pre-cultured on potato dextrose agar at 26 ± 2 °C for 2 weeks in the dark. Five 10-mm-diameter plugs were then excised from the actively growing margins of the pre-cultured mycelium and inoculated into 100 mL of liquid culture medium adjusted to pH 5.5 to 6.0. The medium consisted of glucose (40 g/L), malt extract (15 g/L), and peptone (3 g/L), supplemented with 0.3 g of 100-mesh pine sapwood powder. All culture media were sterilized prior to use. The inoculated cultures were incubated on a rotary shaker at 26 ± 2 °C for 1 week in the dark and subsequently used for the test.

Twelve sapwood specimens each of B. platyphylla and Q. salicina and nine of L. tulipifera (9 × 25 × 25 mm, L × T × R) were used for the fungal decay test. Three specimens were placed in each culture vessel, resulting in four culture vessels each for B. platyphylla and Q. salicina and three for L. tulipifera. In addition, 30 sapwood specimens of Carpinus laxiflora were prepared as controls to monitor the test endpoint.

For the test, cylindrical glass vessels with a bottom surface area of approximately 50 to 100 cm2 and a volume of 500 to 800 mL were used. Each culture vessel was filled with 250 g of sea sand and 80 mL of the same liquid culture medium used for fungal cultivation, and the vessels were sterilized prior to use. The fungus grown in shake culture was thoroughly suspended, and 3 mL of the suspension was inoculated into each culture vessel. The inoculated vessels were incubated at 26 ± 2 °C and 70% relative humidity in the dark until the mycelium completely covered the sand surface. Subsequently, three gas-sterilized wood specimens were placed in each culture vessel, and their fungal decay resistance was evaluated under the same incubation conditions.

After 8 weeks of incubation, three control specimens were randomly removed from the culture vessels, and their weight loss (WL) was measured to determine whether the test endpoint had been reached. Thereafter, three further control specimens were randomly removed at 1-week intervals, and the test was terminated when their mean WL reached 50%. To determine WL, the specimens were dried at 60 ± 2 °C for 48 h and then cooled in a desiccator for 30 min before being weighed to the nearest 0.01 g. The same procedure was applied before and after the decay test, and WL (%) was calculated according to Eq. (1).

WL (%) = (w1-w2)/w1 × 100     (1)

where w1 and w2 are the oven-dried weights (g) before and after the decay test, respectively.

Durability Test Against Termites

The durability test against termites was conducted in accordance with the guidelines outlined in the Assessment Manual for Domestic Wood Characteristics (Park et al. 2024c). The guidelines were based on AWPA E1-17 (2017). Subterranean termites (Reticulitermes speratus), which are distributed throughout South Korea, were used in this study. Five heartwood specimens each were prepared for B. platyphylla and L. tulipifera and seven for Q. salicina, all measuring 6.4 × 25.4 × 50.8 mm (L × T × R).

Prior to the test, the specimens were sterilized with ethylene oxide gas for at least 5 h at room temperature and then conditioned at 20 ± 2 °C and 65 ± 3% relative humidity until exposure to termites. After the specimens reached a constant weight, they were weighed to the nearest 0.01 g. Each specimen was then cut into two pieces measuring 6.4 × 25.4 × 25.4 mm (L × T × R): one for the termite resistance test and the other for moisture content (MC) determination. The pieces used for MC determination were oven-dried at 103 ± 2 °C for 24 h and weighed to calculate the MC and oven-dry mass of the wood.

Throughout the termite resistance test, the sea sand in the culture vessels was maintained at a constant moisture level corresponding to 7% below its water saturation level. The water saturation level of the sea sand was determined by oven-drying 100 g of the sand at 103 °C for 24 h and subsequently adding distilled water until water became visible above the sand surface. The water saturation level was calculated according to Eq. (2),

Water saturation level (%) = (Wwater / Wsand) × 100     (2)

where Wwater is the mass of distilled water added until water became visible above the sand surface (g), and Wsand is the oven-dry mass of the sea sand (g).

For the termite resistance test, 150 g of oven-dried sea sand was placed in each clean culture vessel, followed by the calculated amount of distilled water. The vessels were maintained at room temperature for 2 h to allow the water to distribute throughout the sand. One test specimen was then placed in each vessel with its cross-sectional surface on the sand and one edge against the side of the vessel. Additional culture vessels without wood specimens were prepared to monitor termite activity.

A total of 400 termites, consisting of 396 workers and four soldiers, were introduced into each culture vessel. The termites were placed on the side of the vessel opposite the wood specimen. Once the termites were introduced, the total mass of each culture vessel was recorded and used as a reference for replenishing water when necessary.

The culture vessels were maintained at 26 ± 2 °C and ≥ 70% relative humidity for 4 weeks. Termite activity was monitored throughout the exposure period, and water was replenished as needed to maintain adequate moisture. After 1 week of exposure, termite viability was confirmed in the monitoring vessels.

After 4 weeks of exposure, termite damage was visually evaluated according to AWPA E1-17 (AWPA 2017) using the following rating scale: 10, sound; 9.5, trace attack with surface nibbles permitted; 9, light attack with up to 3% of the cross-sectional area affected; 8, moderate attack with 3 to 10% affected; 7, moderate to severe attack with penetration and 10 to 30% affected; 6, severe attack with 30 to 50% affected; 4, very severe attack with 50 to 75% affected; and 0, failure. After visual assessment, the specimens were oven-dried at 103 ± 2 °C for 24 h, cooled in a desiccator for 30 min, and weighed to the nearest 0.01 g. WL (%) was then calculated according to Eq. (1) with w1 taken as the oven-dry mass estimated from the paired MC piece. Finally, the surviving termites in each culture vessel were counted, and termite mortality was calculated according to Eq. 3,

Termite mortality (%) = [(Ninitial − Nsurviving) / Ninitial] × 100    (3)

where Ninitial is the initial number of termites introduced into each vessel and Nsurviving is the number of surviving termites after the exposure period.

Statistical Analysis

All statistical analyses were conducted using R v4.5.1 (R Core Team 2025). Differences in anatomical traits and durability among the three wood species were tested at a significance level of α = 0.05 (95% confidence level). For each anatomical trait, 30 measurements were obtained from the sampled tree of each species and used in the statistical analyses. Prior to testing, the normality and homogeneity of variance of ARW, the anatomical traits, and the durability data were assessed using the Shapiro–Wilk and Levene’s tests, respectively.

Variables satisfying both assumptions were analyzed using one-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test. Those that violated either assumption were analyzed with the non-parametric Kruskal–Wallis test followed by Dunn’s post-hoc test with Benjamini–Hochberg correction for pairwise comparisons.

Because only one tree was sampled per species, the 30 measurements represent within-tree anatomical variation rather than independent tree-level biological replicates. Therefore, species, individual-tree, and site effects could not be separated, and the statistical comparisons were interpreted as exploratory.

Multiple factor analysis (MFA) was conducted with the FactoMineR package in R to investigate the multivariate associations among four predefined trait blocks: anatomical traits, ARW, WL by fungal decay, and WL by termite attack. Prior to MFA, measurements were averaged at the species level, and all variables were standardized to unit variance. The MFA was used to evaluate the relative contribution of each trait block to the major dimensions of multivariate variation and visualize associations among species and trait variables in a shared ordination space.

RESULTS AND DISCUSSION

Qualitative and Quantitative Anatomical Features

The three hardwood species exhibited diffuse-porous wood structures with distinct growth rings (Figs. 2 to 4). B. platyphylla had vessels that were mostly solitary or occasionally arranged in radial multiples (Fig. 2A and D). In the radial section, the rays were composed entirely of procumbent cells, and the perforation plates were scalariform (Fig. 2B, E, and F). Vessel-ray pits were rounded to angular, with much reduced to apparently simple borders (Fig. 2F).

In the tangential section, the inter-vessel pits were arranged in opposite to alternate patterns (Fig. 2C and G). L. tulipifera contained numerous vessels that were mostly arranged in radial multiples, although solitary vessels were also observed (Fig. 3A and D). In the radial section, the homocellular rays consisted of procumbent cells, whereas the heterocellular rays consisted of procumbent and square-marginal cells (Fig. 3B, E, F, and G). The perforation plates were either simple or scalariform (Fig. 3B and G).

Vessel-ray pits were rounded to angular, with much reduced to apparently simple borders (Fig. 3G). In the tangential section, inter-vessel pits were arranged in opposite patterns (Fig. 3C and H). Q. salicina contained mostly solitary vessels, with some arranged in a radial pattern (Fig. 4A and D). In the radial section, the rays were composed of procumbent cells, and the perforation plates were simple (Fig. 4B and E). Vessel-ray pits were rounded to angular and horizontally elongated, ranging from scalariform to gash-like, or vertically elongated to palisade-like, with much reduced to apparently simple borders (Fig. 4F). In the tangential section, inter-vessel pits were arranged in alternate patterns (Fig. 4C and G).

Anatomical structures of Betula platyphylla: (A to C) Light micrographs

Fig. 2. Anatomical structures of Betula platyphylla: (A to C) Light micrographs; (D to G) scanning electron micrographs; Cross-sections (A, D) show vessel elements arranged in radial alignment; Radial surfaces (B, E, and F) show rays composed of procumbent cells (E), scalariform perforation plates, and vessel-ray pits rounded to angular, with much reduced to apparently simple borders (F); Tangential surfaces (C, G) show vessel elements with opposite and alternate inter-vessel pits (G); Scale bars: (A to C) = 200 µm; (D) = 100 µm; (E to G) = 20 µm

Anatomical structures of Liriodendron tulipifera

Fig. 3. Anatomical structures of Liriodendron tulipifera: (A to C) Light micrographs; (D to H) SEM images; Cross-sections (A, D) show vessel elements arranged in radial alignment; Radial surfaces (B, E to G) show homocellular rays composed of procumbent cells (E), heterocellular rays composed of procumbent cells and square marginal cells (F), perforation plates, either simple or scalariform, and vessel-ray pits rounded to angular, with much reduced to apparently simple borders (G); Tangential surfaces (C, H) show vessel elements with opposite inter-vessel pits (H); Scale bars: (A to C) = 200 µm; (D to H) = 100 µm

In the quantitative analysis, the three diffuse-porous species showed significant differences in anatomical traits (Table 2). B. platyphylla exhibited the narrowest ARW, with an average of 2.33 ± 0.96 mm. However, this species had the longest vessel elements, with an average length of 931.61 ± 57.35 μm. L. tulipifera exhibited the widest fibers in the tangential direction (16.91 ± 2.51 μm) and the thickest fiber cell walls (4.32 ± 0.89 μm). Its vessel dimensions, including radial diameter (Drad; 71.59 ± 7.30 μm) and tangential diameter (Dtan; 53.80 ± 7.21 μm), were smaller than those of the other species. Q. salicina showed the widest ARW, with an average of 6.98 ± 3.75 mm, and the thickest vessel cell walls (3.20 ± 0.24 μm). Its vessel element length was shorter than that of the other species, with an average value of 320.42 ± 32.96 μm. Additionally, Q. salicina showed relatively small fiber dimensions, including radial diameter (Drad; 10.83 ± 1.74 μm) and tangential diameter (Dtan; 7.72 ± 1.12 μm), as well as relatively thin fiber cell walls (3.27 ± 0.38 μm).

Anatomical structures of Quercus salicina

Fig. 4. Anatomical structures of Quercus salicina. (A to C) Light micrographs; (D to G) scanning electron micrographs (SEM). Cross-sections (A, D) show vessel elements arranged. Radial surfaces (B, E, and F) show rays composed of procumbent cells (E), simple perforation plates, and vessel-ray pits rounded to angular, horizontal (scalariform, gash-like) to vertical (palisade), with much reduced to apparently simple borders (F). Tangential surfaces (C, G) show vessel elements with alternate inter-vessel pits (G); Scale bars: (A to C) = 200 µm; (D, F) = 100 µm; (E, G) = 20 µm

The predominance of radial multiples in L. tulipifera indicates a potentially higher degree of vessel-to-vessel contact compared with that of the other species. In contrast, Q. salicina is mainly characterized by solitary vessels, whereas B. platyphylla exhibits an intermediate condition with mostly solitary vessels and occasional radial multiples. Carlquist (2009) has emphasized that vessel distribution in wood is often nonrandom and may have functional significance, particularly in terms of conductive safety, mechanical organization, and water and photosynthate storage capabilities. Conductive safety is closely related to preventing or limiting xylem cavitation and the spread of embolism; thus, vessel arrangement may provide anatomical information relevant to hydraulic safety. However, because hydraulic conductivity and cavitation resistance were not measured in this study, the functional implications of these vessel arrangements should be interpreted with caution. Previous studies have reported contrasting interpretations of vessel groupings. Loepfe et al. (2007) predicted that enhanced vessel connectivity increased conductivity but reduced cavitation resistance by increasing the probability of embolism spreading. In contrast, Lens et al. (2011) found that enhanced vessel grouping in Acer was associated with increased cavitation resistance and reduced xylem area-specific conductivity. Therefore, the vessel arrangement patterns observed in this study are best interpreted as anatomical evidence of potential differences in xylem network organization rather than as direct evidence of hydraulic performance.

Table 2. Anatomical Traits of Three Diffuse-porous Hardwood Species

Anatomical Traits of Three Diffuse-porous Hardwood Species

Natural Durability of Three Diffuse-porous Wood Species

The natural resistance of B. platyphylla and L. tulipifera against wood-decaying fungi and termites has been examined in previous studies using the same samples (Jung et al. 2025). Therefore, the corresponding data were referenced in this study. The durability of Q. salicina was investigated in the present study, following the same procedures used in previous studies.

In the durability test against wood-decaying fungi, B. platyphylla exhibited the lowest resistance among the three diffuse-porous wood species, with the highest WL of 45.53 ± 4.68% (Fig. 5A). Although Q. salicina showed a greater WL (40.24 ± 4.12%) than L. tulipifera (35.95 ± 5.73%), the difference was not statistically significant (Tukey’s HSD, adjusted p = 0.126; Fig. 5A).

In the termite resistance test, clear differences in termite resistance were observed among the three wood species (Fig. 5B). B. platyphylla showed the lowest resistance, with the highest WL of 40.32 ± 1.60%, which was consistent with its low resistance to wood-decaying fungi. In contrast, L. tulipifera exhibited the highest resistance to termite attack, resulting in a significantly low WL of 1.76 ± 0.33% (Fig. 5B). Q. salicina showed moderate termite resistance among the species tested, with a WL of 11.48 ± 2.93% (Fig. 5B).