Three-Dimensional Visualization of Major Anatomical Structural Features in Softwood

Abstract

Wood displays three-dimensional characteristics at both macroscopic and microscopic scales. Accurately reconstructing its 3D structure is vital for a deeper understanding of the relationship between its anatomical characteristics and its physical and mechanical properties. This study aims to apply X-ray micro-computed tomography (XμCT) for the high-resolution, non-destructive visualization and quantification of softwood anatomical features. Six typical softwood species—Picea asperata, Cupressus funebris, Pinus koraiensis, Pinus massoniana, Cedrus deodara, and Pseudotsuga menziesii—were selected to represent a range of structural characteristics. The results show that a scanning resolution of 1–2 μm is suitable for investigating the transition from earlywood to latewood and resin canals, while a resolution of 0.5 μm is required for finer structures such as bordered pits, ray tracheids, and cross-field pits. In Pinus koraiensis, a direct 3D connection between radial and axial resin canals was observed, forming an interconnected resin network. In contrast, wood rays were found to be distributed near the surface of axial resin canals but without forming interconnected structures. The three-dimensional reconstruction of bordered pit pairs in Pinus massoniana and Picea asperata clearly revealed interspecific differences in pit morphology, distribution, and volume. The average surface area and volume of bordered pit pairs in Pinus massoniana were 1151.60 μm² and 1715.35 μm³, respectively, compared to 290.43 μm² and 311.87 μm³ in Picea asperata. Furthermore, XμCT imaging effectively captured the morphology and spatial distribution of cross-field pits across species, demonstrating its advantage in comprehensive anatomical deconstruction. These findings highlight the potential of XμCT as a powerful tool for 3D analysis of wood anatomy, providing deeper insight into the structural complexity and interconnectivity of wood.

1. Introduction

As a vital plant resource, softwood possesses distinctive anatomical characteristics [1], which support its extensive application in industries such as construction, pulp production, and furniture manufacturing [2]. These anatomical features primarily include axial tracheids, wood rays, axial parenchyma, and resin canals [3,4,5], which significantly influence wood growth, development, and environmental adaptability [6,7]. They also play crucial roles in the physical and mechanical properties of wood, which certainly determines the overall utilization value of timber [7,8]. For instance, the morphology and spatial distribution of axial tracheids influence the water transport capacity of wood [9], while the arrangement of wood rays is related to its compressive strength [10]. The detailed analysis of the morphological and spatial distribution of these internal anatomical characteristics provides accurate and comprehensive 3D data, which can aid in the development of advanced biomimetic wood materials. Moreover, such research enables a better understanding of the relationships between different hierarchical structures within wood, allowing for the optimization of wood processing techniques and usage to improve properties such as strength, stiffness, and resistance to deformation.

Traditionally, wood anatomy research has primarily relied on two-dimensional sectioning combined with optical or electron microscopy [11]. While these methods generate high-resolution images for detailed cell characterization, they are inherently limited in capturing complex 3D structures. To overcome this limitation, early attempts at 3D visualization involved methods such as resin casting. For example, Mauseth and Fujii injected resin into plant tissues and subsequently digested organic material, enabling initial 3D modeling of tracheids [12]. This approach provided novel insights into wood anatomical characteristics. Kitin et al. further utilized this method to analyze the 3D shapes of cambial cells in Celtis and tracheid elements [13]. The results highlighted the potential of resin casts in revealing plant cell wall structures. Additionally, magnetic resonance imaging (MRI) was utilized to observe the process of xylem cavitation (the formation of gas bubbles due to water column breakage) and to investigate the dynamic development of embolism (the blockage of water transport caused by these bubbles) [14]. Oven et al. further employed 3D magnetic resonance microscopy (3D MRM) to visualize internal wood structures and moisture distribution non-destructively [15]. Although these pioneering studies have significantly advanced our understanding, the comprehensive characterization of wood’s 3D anatomy remains incomplete.

Recently, X-ray micro-computed tomography (XμCT), a non-destructive, high-resolution imaging technique, has emerged as a powerful tool in materials research [16,17,18].

In wood science, Lautner et al. utilized synchrotron radiation X-ray micro-computed tomography (SRμCT) to study microscopic structures of poplar (Populus trichocarpa) [19] and to non-destructively examine detailed morphology within phloem tissues [20]. Using this method, they successfully mapped the size, volume, and connectivity of various cells. Similarly, Tim Koddenberg employed XμCT to obtain morphological images of axial tissues in Fraxinus excelsior L., and the length, diameter, and volume of tracheids, wood fibers, and axial parenchyma cells were quantified [21].

As a non-destructive method, XμCT can accurately reveal internal structural changes in wood [22,23,24,25,26]. Thus, it is a great method to investigate wood degradation, including cell wall degradation and cavity formation [27]. Shi et al. systematically investigated the 3D structural and chemical composition changes in wood cells during cellulose purification, by combining high-resolution X-ray CT and spectroscopy, revealing the mechanisms underlying structural changes in these cells [28]. In addition to quantifying mass loss and visualizing structural changes, the combined technique provided important insights into the infection pathways and degradation mechanisms of fungi [29]. XμCT has also been employed to analyze the distribution and interconnectivity of different components in wood-based composites [30], as well as the distribution of moisture [31,32]. Therefore, XμCT is an effective method to investigate the 3D anatomical characteristics of softwood from a holistic perspective [33].

This study employed X-ray micro-computed tomography (XμCT) to achieve high-resolution three-dimensional (3D) visualization and quantitative analysis of the major anatomical features of softwood. Based on distinct anatomical characteristics, six representative softwood species were selected: Picea asperata, Cupressus funebris, Pinus koraiensis, Pinus massoniana, Cedrus deodara, and Pseudotsuga menziesii. For instance, Pinus koraiensis exhibits a gradual transition from earlywood to latewood, along with both radial and axial resin canals and pinoid cross-field pits. In contrast, Pinus massoniana features an abrupt earlywood–latewood transition, well-developed bordered pits, and ray tracheids. These species collectively represent the typical structural diversity found in softwoods.

Through 3D reconstruction, we obtained intuitive visualizations and quantitative data for five key anatomical features: the transition from earlywood to latewood, resin canals, bordered pit pairs, ray composition, and cross-field pits—features closely associated with wood functionality. By reconstructing and quantifying these structures, this research overcomes the limitations of traditional 2D methods, revealing previously inaccessible spatial complexities. By supplementing key knowledge gaps in softwood anatomy, this comprehensive and systematic approach also offers essential data to support improved wood utilization and material innovation.

2. Materials and Methods

2.1. Materials and Sample Preparation

Softwood specimens were selected from 6 species: Picea asperata, Cupressus funebris, Pinus koraiensis, Pinus massoniana, Cedrus deodara, and Pseudotsuga menziesii. These specimens were sourced from the sample repository of Nanjing Forestry University.

To optimize the 3D visualization of specific anatomical characteristics, the scanning resolution was adjusted accordingly. A resolution of 1–2 μm was employed for larger characteristics such as the earlywood-to-latewood transition and resin canals, while a finer resolution of 0.5 μm was used for smaller characteristics such as bordered pit pairs, ray tracheids, and cross-field pits.

The sample preparation procedure was as follows:

(i)

Sample Cutting and Smoothing: All wood specimens were cut to a size of 7 mm (radial, R) × 7 mm (tangential, T) × 10 mm (longitudinal, L). The surfaces of the blocks were then smoothed using a sliding microtome (TU-213, Yamato, Japan).

(ii)

Freeze-Drying: the blocks were placed in a freeze-dryer at −40 °C for 48 h to achieve an oven-dry state.

(iii)

Platinum Coating: after freeze-drying, the wood sample surfaces were coated with platinum for 60 s under a current of 30 mA using an ion sputter coater (ETD-2000III, Vision Precision Instruments, China).

(iv)

SEM Preparation: the areas of interest, containing features for observation, were selected and prepared using a scanning electron microscope (SEM, Quanta 200, FEI, Hillsboro, America).

(v)

Sample Fabrication: The blocks were then fabricated into rectangular specimens with standard orthogonal cut surfaces. Samples with two sets of dimensions were prepared:

-

2 mm (R) × 2 mm (T) × 10 mm (L) for micro-CT scanning;

-

1 mm (R) × 1 mm (T) × 10 mm (L) for nano-CT scanning.

(vi)

Final Freeze-Drying: prior to scanning, the specimens were again subjected to freeze-drying at −40 °C for 48 h to achieve an oven-dry state.

Additionally, since the anatomical features observed in this study are primarily located in the earlywood region, all samples, except those intended for the observation of the transition from earlywood to latewood, were taken exclusively from earlywood regions.

2.2. X-Ray Micro-Computed Tomography

Three-dimensional data acquisition was conducted using XμCT (NanoVoxel-3502E Tianjin Sanying, Sanying Precision Instruments Ltd., China) at Nanjing Forestry University. As illustrated in Figure 1, the specimens were vertically fixed on a cylindrical sample stage using hot melt glue (W9215, GZWW, China). For micron-level scanning, the parameters were set to a resolution of 1–2 μm, a scanning voltage of 80 kV, and a scanning current of 200 μA. For nano-level scanning, the parameters were adjusted to a resolution of 0.5 μm, a scanning voltage of 60 kV, and a scanning current of 15 μA.

2.3. Data Processing and Analysis

Image preprocessing, segmentation, visualization, and quantitative analysis were conducted in a commercial 3D imaging software package, Dragonfly version 2024.1 for Windows (Object Research Systems (ORS) Inc., Montreal, QC, Canada).

2.3.1. Image Preprocessing

The X-ray CT images were preprocessed using a combination of three image enhancement techniques: histogram equalization, median filtering, and unsharp masking. These methods were selected for their complementary capabilities in improving image quality while preserving fine anatomical characteristics. The specific explanation of these approaches is as follows:

(i)

Histogram Balancing: This method transforms the grayscale histogram of the image into an approximately uniform distribution using the cumulative distribution function. It enhances contrast across the image, particularly improving the visibility of subtle structural features.

(ii)

Median Filtering: chosen for its effectiveness in removing impulse noise without blurring edges, this method eliminates noise arising from imaging hardware and environmental disturbances, thereby preserving anatomical boundary integrity.

(iii)

Unsharp Masking: This method is applied to enhance edges and fine textures in the image. It increases the clarity of structural outlines—such as cell walls—without amplifying background noise, which is essential for distinguishing closely packed features in wood anatomy.

These techniques, when combined, ensure a balance between contrast enhancement, noise suppression, and edge preservation, facilitating more accurate segmentation and feature extraction in subsequent steps.

2.3.2. Image Segmentation

The average grayscale values of the images facilitated subsequent segmentation (Figure 2). The lower limit of Otsu’s threshold within the image grayscale range was selected to obtain the voids in the wood image, and the grayscale values were inverted to obtain the cell walls. The wood void regions were extracted as regions of interest (ROIs), and semantic segmentation was performed on the pore ROIs to label the corresponding feature regions of interest, such as axial tracheids, wood rays, resin canals, ray tracheids, and bordered pits. By combining morphological operations (dilation, erosion, smoothing, and internal filling) and watershed segmentation, the feature ROIs were refined. This process resulted in the rendering of 3D images of softwood anatomical characteristics. Through rendering, different anatomical characteristics were assigned distinct colors, effectively separating them from the overall structure to obtain the 3D morphological information and local detail information of the samples.

2.3.3. Quantitative Analysis

Using the Object Analysis module in Dragonfly, the volume and surface area of each anatomical feature within the ROIs were calculated. The volume was computed using Equation (1):

$$V=R^3\ast N$$

(1)

where V represents the volume of the feature ROI (µm³), R is the scanning resolution of the data (µm), and N is the number of voxels contained within the feature ROI.

The surface area was calculated using the weighted local configuration method proposed by Lindblad [34]. This technique estimates the surface area of a 3D object by evaluating the local neighborhood configurations of each voxel. Each configuration is assigned a specific weight based on geometric modeling of voxel connectivity. The method balances computational efficiency with geometric accuracy, allowing high-resolution surface measurement while reducing computational complexity. This approach is particularly suitable for anatomical characteristics with irregular boundaries, such as bordered pits or resin canals.

In addition to volume and surface area, the linear dimensions of anatomical characteristics were measured using the ruler tool, with care taken to select isolated and complete structures for accurate measurement.

3. Results and Discussion

The selection of XμCT resolution is of crucial significance for wood anatomy research. With an increasing spatial resolution, the observable field of view gradually narrows, and the microstructure of the specimen can be presented with higher clarity and detail. Recent studies indicate that medium resolution is optimal for identifying most hardwood anatomical characteristics [35]. However, softwood requires higher precision due to its smaller anatomical characteristics. Our data indicate that a resolution of 0.5–2 μm is particularly critical for the 3D visualization of the anatomical characteristics of softwood.

Figure 3 illustrates X-ray CT images of Picea asperata at different spatial resolutions (5.5 μm, 1.8 μm, and 0.5 μm). Growth rings are clear at a resolution of 5.5 μm; however, the morphology of the cell walls cannot be clearly discerned (Figure 3A). Typical softwood features, such as axial tracheids, wood rays, and resin canals, are prominently visible at a resolution of 1.8 μm (Figure 3B).

Figure 4 presents electron microscopy and X-ray CT images of the axial parenchyma in Cupressus funebris. The axial parenchyma of the softwood can be observed at a resolution of 1.8 μm using electron microscopy, with the lumens containing filling materials (Figure 4A). In the X-ray CT images, some axial parenchyma is clearly displayed in the radial sections (Figure 4C); its morphology, however, is hard to identify in the cross-sections (Figure 4B). This difficulty is mainly because axial parenchyma is relatively scarce and small in softwood, causing it to blend into the background noise at lower resolutions. Anatomical characteristics, such as bordered pits and ray parenchyma cells, are distinctly visible at high resolution (0.5 μm) using X-ray CT (Figure 3C). These observations indicate that high-resolution CT scanning technology offers significant advantages in capturing the complex 3D anatomical characteristics of wood. However, the increase in resolution is accompanied by a reduction in the field of view. Therefore, it is essential to select a proper resolution that is suitable for investigating specific anatomical characteristics.

3.1. Transition from Earlywood to Latewood

The 3D structure in the transition regions from earlywood to latewood in softwood was obtained using X-ray CT at a resolution of 1.8 μm (Figure 5). The lumen size of axial tracheids, shown in dark green, gradually decreases in the transition from earlywood to latewood, reflected in color intensity (Figure 5A). This gradual change indicates that Pinus koraiensis progressively adjusts the function and structure of its wood cells during the growing season [36]. On the contrary, the size of axial tracheid lumens from earlywood to latewood shows abrupt changes in a single growth ring of Pinus massoniana (Figure 5B). The thickness of the cell walls rapidly increases in latewood, shown in light green in the images. The rapid increase in the thickness of the cell walls in latewood reflects a swift response to environmental conditions at the end of the growing season, such as sudden changes in temperature, light, or moisture [37].

Figure 6 presents layer-by-layer porosity curves along the tangential sections of Pinus koraiensis and Pinus massoniana, derived from X-ray CT images at a resolution of 1.8 μm. These curves quantitatively reflect the structural differences in the transition from earlywood to latewood within a single growth ring.

In Pinus koraiensis, the porosity gradually decreases across the slices, with a mean value of 64.71% and a relatively low standard deviation (8.13), indicating a smooth transition in axial tracheid structure. This smooth gradient supports the 3D morphological observation that Pinus koraiensis undergoes a gradual adjustment in cell lumen size and wall thickness throughout the growing season. Such a pattern reflects an adaptive strategy to slowly shifting environmental conditions, allowing for the continuous functional modulation of hydraulic conductivity and mechanical strength.

In contrast, Pinus massoniana exhibits abrupt variation in porosity, with values fluctuating more significantly across the tangential plane (mean = 62.09%; standard deviation = 17.78). This sharp transition implies a sudden structural adjustment at the end of the growing season, possibly triggered by abrupt environmental changes such as reduced moisture or declining temperatures. The pronounced contrast between earlywood and latewood in Pinus massoniana is likely associated with a more compartmentalized growth strategy, optimizing its mechanical performance in response to seasonal constraints.

These findings demonstrate that quantitative porosity profiling, when combined with 3D anatomical reconstruction, provides deeper insight into species-specific growth patterns and ecological adaptation mechanisms.

3.2. Resin Canals

Three-dimensional images of resin canals in Pinus koraiensis at a resolution of 1.8 μm are shown in Figure 6. Radial resin canals and fusiform rays can be clearly distinguished (Figure 7A–D). Radial resin canals distribute in the radial direction of the wood and interweave with fusiform rays to form complex network structures. The detailed 3D images contribute to the understanding of the role of resin canals in the defense mechanisms of wood, such as resisting pests and mechanical damage. The secretory cells of the axial resin canals can hardly be observed at the current resolution (Figure 7E,F). We inferred that the small size and tight connection of the secretory cells may cause their edges to be less distinguishable, making them appear as a bright area during image segmentation.

It is possible to observe the 3D interactions between resin canals and wood rays at a resolution of 1.8 μm (Figure 8). Axial resin canals connect with radial resin canals (Figure 8A), creating regions of interaction (Figure 8B). It can be observed from a 3D perspective that portions of the radial resin canals penetrate through the axial resin canals (Figure 8C). The radial resin canals and axial resin canals are interconnected, forming a complex network of resin channels. Similar results have been demonstrated by performing a continuous sectioning experiment on Pinus densiflora [38]. The mutual interconnectivity structure of resin canals facilitates efficient resin transport in multiple directions, meanwhile enhancing the wood’s defensive capabilities and healing mechanisms [39]. The 3D structure of the resin canal allows a more accurate and convenient analysis of resin canal connectivity, which contributes to a more intuitive understanding of the complex network.

Unlike the interactions described above, there are few interconnections between wood rays and axial resin canals (Figure 8D–G). Distinct concave marks are clearly visible on the surface of the axial resin canals after rendering the wood rays transparent in Figure 8D. The spatial interaction between the wood rays and the axial resin canals is visually demonstrated: the wood rays curve along the surface of the axial resin canals and do not intersect with them. During the process of “bypassing” the resin canals, the wood rays exhibit a tightly enveloping arrangement, which may result from spatial constraints and cellular compression during development [40]. Ray parenchyma cells grow closely along the surfaces of resin canals, forming a continuous and functionally complementary microscopic system. When trees are attacked by insects or fungi, or suffer mechanical damage, both constitutive and traumatic resin canals, along with ray tissues, participate in the secretion and transport of resin to surrounding tissues, thereby serving as a defensive mechanism to block and seal off the affected areas [41]. The wood rays that adhere tightly and bend along the surfaces of resin canals can disperse or counteract stress at the microscopic level if trees are subjected to external forces. The wood rays contribute to the support and reinforcement of the wood structure by preventing the widespread propagation of cracks [42]. According to the above discussion, the results for 3D structure can effectively illustrate the adaptation and synergy between resin canals and wood rays in aspects such as defense, mechanical support, and developmental formation.

3.3. Bordered Pit Pairs

The bordered pit pairs on the cell wall of axial tracheids in Pinus massoniana were observed at a resolution of 0.5 μm (Figure 9). The bordered pit pairs in two softwood species exhibit similar morphologies, and connections between bordered pits among axial tracheids are clearly visible.

Figure 9 displays 3D images of the standard morphologies of bordered pit pairs on the axial tracheid walls of Pinus massoniana and Picea asperata, along with the specific measurement directions for pit pair dimensions. In Pinus massoniana, bordered pit pairs on the walls of axial tracheids exhibit regular circular or elliptical shapes, with an axial diameter of 24.2 μm, a radial diameter of 27.11 μm, and a depth of 5.91 μm (Figure 9A–D). The relative largeness and orderly morphology of these pits contribute to increasing water transportation and enhancing the mechanical strength of trees, thereby adapting to their growth requirements in humid environments [43]. For Picea asperata, bordered pit pairs on the walls of axial tracheids display polygonal or irregular shapes, with an axial diameter of 10.26 μm, a radial diameter of 11.10 μm, and a depth of 4.56 μm (Figure 9E–H). Bordered pit pairs with small inter-pit distances are helpful for enhancing the water transportation and mechanical strength of trees [44].

Three-dimensional anatomical analysis provides the spatial distribution, surface area, volume, and other quantitative data of bordered pit pairs (Figure 10). The surface area values and the spatial distribution of bordered pit pairs on the walls of axial tracheids are analyzed by excluding the incomplete and spatially interconnected pit pairs. In total, 677 and 688 isolated and complete bordered pit pairs are selected from Pinus massoniana and Picea asperata specimens, respectively (Figure 10A,B). The surface area and volume of bordered pit pairs on the axial tracheid walls of Pinus massoniana are 1151.60 ± 211.66 μm² and 1715.35 ± 351.37 μm³, respectively. For Picea asperata, they are 290.43 ± 36.31 μm² and 311.87 ± 58.16 μm³ (Figure 10C,D). The surface area and volume of bordered pit pairs on axial tracheid walls in Pinus massoniana are larger than in Picea asperata. The surface area and volume variation in Pinus massoniana is also larger than in Picea asperata. The axial diameter, radial diameter, and depth of the bordered pit pairs on the axial tracheid walls of Pinus massoniana are 25.60 ± 3.28 μm, 21.31 ± 1.78 μm, and 7.22 ± 0.62 μm, respectively. For Picea asperata, they are 11.79 ± 1.92 μm, 12.04 ± 0.64 μm, and 8.47 ± 1.17 μm (Figure 10E,F). The size of bordered pit diameters in Pinus massoniana is significantly larger than that in Picea asperata. The larger size of bordered pit diameters in Pinus massoniana is in concordance with the fact that Pinus massoniana is suited to survival in a moist environment thanks to its wide water pathways. Picea asperata exhibits great depth in the tangential direction of bordered pit pairs (Figure 10G), which helps enhance the mechanical strength and durability of the cell wall, thereby improving the wood’s resistance to environmental stress [45].

These interspecific differences in bordered pit pair morphology may reflect adaptations to different ecological conditions. The relatively larger surface areas and diameters of pit pairs in Pinus massoniana likely enhance water conductivity, which is advantageous in the humid environments where this species typically grows. In contrast, the greater pit depth observed in Picea asperata may strengthen the mechanical integrity of the cell wall and improve resistance to freeze–thaw cycles or mechanical stress, which aligns with the species’ adaptation to colder, higher-altitude habitats. Such structural variations suggest species-specific strategies for balancing hydraulic efficiency with mechanical safety in response to their native environments.

3.4. Ray Composition

The uniseriate rays in Pinus massoniana can be clearly visualized at a resolution of 0.5 μm (Figure 11). The growth pattern of the axial tracheids is influenced by the presence of wood rays in the radial section (Figure 11A). The axial tracheids exhibit a curved line instead of a straight line. This curve allows the wood rays to tightly intercalate between the axial tracheids, forming a complex 3D network structure. This structure probably enhances the mechanical properties and structural integrity of the wood at the microscopic level. Meanwhile, the curve helps to optimize the transportation of water and nutrients within the trees, improving the trees’ adaptability in different environmental conditions [45,46].

The shape of uniseriate rays in Pinus massoniana is irregular or rectangular (Figure 11B,C). The rays with an irregular shape have distinct end wall edges, with portions extending outward. Despite the resolution being as high as 0.5 μm, the images based on virtual slices still exhibit some blurry boundaries. Generally speaking, the uniseriate rays in Pinus massoniana uniformly arrange in the radial direction, forming a tight network structure. This structure contributes to water and nutrient transportation along the radial direction, and to the mechanical stability and bending resistance of the wood [47].

3.5. Cross-Field Pits

The cross-field pits between ray parenchyma cells and axial tracheids in softwood can be clearly visualized at a resolution of 0.5 μm (Figure 12). There are five observed types of cross-field pits that display obvious differences in both structural morphology and spatial distribution. The window-like cross-field pits in Pinus massoniana (Figure 12A–C) are large, and there are 1–2 pits present in most cross-fields. The pinoid cross-field pits in Pinus koraiensis (Figure 12D–F) are slightly smaller than the window-like cross-field pits, and the shape of pinoid cross-field pits is rather consistent. The taxodioid cross-field pits in Cedrus deodara (Figure 12G–I) exhibit a unique wave-like morphology, and their shape is nearly round with oval-shaped pit apertures. The cupressoid cross-field pits in Cupressus funebris (Figure 12J–L) are of a round shape and evenly arranged, and there are two pits in most cases. The piceoid cross-field pits in Pseudotsuga menziesii (Figure 12M–O) are tightly packed, and there is a large number of small pits. The piceoid cross-field pits have narrow pit apertures that are slightly extruded or inwardly contained. The variation in the shape, distribution, and size of the different types of cross-field pits reflects the trees’ adaptive strategies to different environmental conditions during growth [48]. For instance, larger and more sparsely distributed pits (e.g., window-like cross-field pits in Pinus massoniana) may enhance radial transport efficiency, which is beneficial in environments requiring rapid water redistribution. In contrast, smaller and denser pits (e.g., in Pseudotsuga menziesii or Cupressus funebris) may provide better control over water movement, offering increased resistance to embolism under drought stress. These structural differences likely represent trade-offs between transport efficiency and hydraulic safety, contributing to the species’ overall ecological strategies.

It can be observed that the distribution of cross-field pits on two sides of the same ray parenchyma cell shows notable differences (Figure 12C,F,I,L,O). These differences are not only in terms of the number and size of the pits but, more clearly, in the asymmetric distribution pattern. This distribution pattern is difficult to detect through two-dimensional anatomical observation. These pits can be clearly visualized in three dimensions and quantified based on the XμCT results.

4. Conclusions and Outlook

This study aimed to visualize and quantitatively analyze the typical anatomical characteristics of softwood using high-resolution X-ray micro-computed tomography (XμCT). These characteristics were reconstructed in three dimensions, offering comprehensive and non-destructive observations of the transition from earlywood to latewood (both abrupt and gradual), resin canals, wood rays, bordered pit pairs, and cross-field pits. The results indicate that the optimal scanning resolution for capturing the transition from earlywood to latewood and resin canals is 1–2 μm, while a finer resolution of 0.5 μm is necessary to effectively visualize bordered pit pairs, ray tracheids, and cross-field pits. Compared with traditional two-dimensional sectioning, XμCT provides a more intuitive, accurate, and holistic representation of spatial interactions—particularly, between wood rays and resin canals—thus overcoming the limitations of conventional methods.

The morphology, spatial distribution, and volume differences in bordered pit pairs on the cell walls of axial tracheids in Pinus massoniana and Picea asperata were intuitively displayed in the 3D reconstruction, offering a more precise spatial analysis. Additionally, XμCT technology with high resolution can reconstruct the 3D morphology of cross-field pits between ray parenchyma cells and axial tracheids, revealing differences in their spatial morphology, distribution, and size. The spatial distribution and asymmetrical arrangement of the pits can be observed with high precision. Data analysis showed that the average surface area and volume of the bordered pit pairs in Pinus massoniana are 1151.60 μm² and 1715.35 μm³, respectively, with larger and more regular pit pairs. The average surface area and volume of the bordered pit pairs in Picea asperata are 290.43 μm² and 311.87 μm³, respectively, with smaller and more densely distributed pit pairs. These details are difficult to observe with two-dimensional techniques, while XμCT technology significantly enhances the accuracy and comprehensiveness of anatomical feature analysis in softwood.

This research offers a new perspective for understanding the biological structure and mechanical function of wood. The 3D anatomical datasets provide foundational information for studying growth patterns, microstructural organization, and water or metabolite transport pathways. In particular, the data can guide the optimization of wood modification techniques such as thermal treatment and resin impregnation.

While XμCT provides significant advantages for the non-destructive 3D visualization of wood anatomical characteristics, several limitations should be acknowledged. The inherent trade-off between spatial resolution and field of view constrains the simultaneous observation of both micro-scale and macro-scale anatomical features. For example, a resolution of 0.5 μm is necessary to capture detailed structures such as bordered pit pairs and cross-field pits, but this often limits the observable sample size due to hardware constraints. Moreover, high-resolution scans require prolonged scanning times and generate large datasets, increasing the burden on postprocessing and computational resources. Imaging dense or resin-rich tissues may also introduce artifacts or reduce contrast under certain conditions.

Despite these challenges, XμCT remains a powerful tool for wood anatomy research, offering spatial precision and structural continuity at levels unattainable with traditional 2D methods. Future advancements in imaging hardware, scanning speed, and data processing algorithms are expected to alleviate current technical limitations. Additionally, future research could benefit from integrating XμCT-derived anatomical data with chemical composition analysis to provide a more comprehensive understanding of wood functionality across species, contributing to both fundamental research and applied wood science.