The morphology features of the C. tumidissinoda BN were shown in Figure 3A . The BN had a culm and a branch, the branch was between the sheath scar and the nodal ridge, the medullary cavity appeared under the nodal ridge. The junction of the branch and the culm was in a similar position as the diaphragm in height. The shape of unbranched side was more exoconvex than that of the culm side.

The morphological variations of the VB observed in the longitudinal section of the BN can be attributed to its transforming orientation and size ( Figure 3A ). In contrast, the internodal VBs appear to align nearly parallel to each other (Yu et al., 2002; Zhang et al., 2017b). Slices 0 - 1292 corresponded to the cross-section of the BN from the bottom to the top. Slices 50, 110, 300, 385, 600, and 872 were on the lower end of the diaphragm, near the sheath scar, the upper end of the diaphragm, medullary cavity, the lower end of the nodal ridge, and nodal ridge, respectively. The obvious morphological changes of the VB in these slices allowed further analyses.

In the same line, the boundary of each tissue in the cross-section of the BN was clear, allowing the distinct observation of parenchyma, fibers, and conducting tissue. This finding suggested that µCT could achieve good imaging results of BN in C. tumidissinoda. The shape of the VB changed significantly and differed from the typical VB types, including the metaxylem vessels, phloem with sieve tubes, and protoxylem ( Figures 3A, B ).

The anatomical features at different heights in the BN was shown in Figure 4 . Notable morphological variations were observed in the VB in cross-section along the vertical direction of the BN. Some VBs were present at the bottom of the BN but lacked the characteristic features of typical VBs based on the shape of their fibers. The metaxylem vessels and sieve tubes displayed an absence of a clearly demarcated boundary and amalgamated into a larger conducting tissue in VBs. ( Figure 4A , white circle). Some VBs changed direction on the lower end of the diaphragm, appearing as a transverse vascular bundle (TVB) in the cross-section ( Figure 4B , white circle). Notably, the image of the diaphragm at this location is relatively bright ( Figure 4B , yellow arrow). In the same line, some small VBs were found near the sheath scar in the cortex ( Figure 4C , magenta circle), while others were located in the culm of the BN, bifurcated from the VB ( Figure 4C , white circle). Branching emerged at the node on the upper end of the diaphragm ( Figure 4D , yellow arrow), while the TVB was predominantly located on the side of the culm close to the branch ( Figure 4D , white circle). However, a significant amount of TVBs were distributed at the diaphragm of the unbranched bamboo node (Li et al., 2021). This indicates that the branch had an impact on the distribution of TVBs in the BN. The diaphragm disappeared under the nodal ridge, but numerous TVBs remained interconnected through rotation ( Figure 4E ). The boundary between the culm and the branch was apparent and was mainly characterized by abundant VBs ( Figure 4E , white circle). The diaphragm disappeared, the culm and branch of the bamboo node separated, and the fibers near the branch side were particularly developed and connected together, which is different from the result of the previous study that the fibers of unbranched bamboo nodes were distributed more evenly in the cross-section under the nodal ridge (Li et al., 2021). Further research is needed to explore the impact of this uneven fiber distribution on the BN properties. The cortex was not very pronounced on the sides of the culm closer to the branch ( Figure 4F , white circle), and some starch granules appeared near the medullary cavity at the lower end of the nodal ridge ( Figure 4F , magenta circle). The sudden enlargement of the nodal ridge resulted in cracking, which could be categorized into two: shorter cracks found at the interface between fibers and conducting tissue ( Figure 4G , white circle) and longer cracks at the interface between fibers and parenchyma ( Figure 4G , magenta circle). These observations suggested that the interface between different tissues and the point of abrupt node enlargement were the areas where the BN was prone to cracking. There was an inverse relationship among the radial variations of VB size within the same cross-section on the upper part of the BN. Specifically, on the swollen side of the BN, the size of the VB gradually decreased from the outer to the inner layer ( Figure 4H , white box) but increased gradually from the outer to the inner layer on the other side ( Figure 4H , magenta box). This phenomenon resembles the size distribution pattern of VBs at the top of unbranched bamboo nodes (Li et al., 2021). These findings added the conclusions drawn by other researchers, who only suggested that the VB size of the bamboo culm increases progressively from the outer to the inner layer (Kanzawa et al., 2011; Chen et al., 2018; Wang and Shao, 2020). The development of reaction wood is typically concomitant with eccentric growth (Liu et al., 2023). However, bamboo lacks vascular cambium and therefore does not exhibit reaction tissues associated with secondary growth (Ahmad and Kamke, 2005); as a result, this phenomenon in bamboo is not related to stress. By observing a large number of cross-sections of BN, the variations in anatomical characteristics of VB were captured at different heights. These results were difficult to achieve through microscopy techniques, which offer restricted spatial information (Grosser and Liese, 1971; Xiong et al., 1980; Ding and Liese, 1995; Ding et al., 2000), especially when dealing with the large sizes of TVBs and significant deviations in the transverse orientation of VBs.

The size and quantity of AVBs at different height were quantitatively analyzed due to the structural changes of VBs along the BN height are obvious ( Table 1 ). The size of the AVBs enlarged steadily from the bottom of the BN to the sheath scar. This trend was intricately linked with the amplification of conducting tissue within the AVBs in the region. Consequently, this enlargement slows down transpirational flow in the sheath scar, allowing nutrients to remain within the node for a longer period before reaching the branch (Yamaji and Ma, 2017). However, the size of the AVBs declined along the node from the sheath scar to the upper end of the diaphragm. This phenomenon was attributed to the fact that the AVB bifurcated some small VBs, reducing the size of the original one. Notably, the size of AVBs significantly increased from the point of the medullary cavity up to the nodal ridge, but decreased towards the top of the BN. The former effect is attributed to the disappearance of TVBs, whereas the latter is driven by the pronounced nodal morphology. The changes in VB size from the sheath scar to the diaphragm were inconsistent with the observations reported by the previous researchers (Li et al., 2021). This deviation may be attributed to the significant amount of TVBs in the branch.

The number and frequency of VBs in the slice of the BN varied with height ( Table 2 ). The total count of VBs increased gradually from the bottom of the BN to the upper end of the diaphragm and decreased towards a stable level at the top of the BN. This phenomenon was attributed to the bifurcation of VBs between the lower end and the upper end of the diaphragm (Li et al., 2021). The number of AVBs decreased from the bottom of the node to the medullary cavity and then stabilized until the top of the BN. Notably, the part with reduced VBs corresponded with the emergence of the branch, whereas the position with stabilized VB number corresponded with the branch absence, indicating that the branch affected the number of VBs. The TVBs appeared in the lower end of the diaphragm and disappeared above the medullary cavity, with their number reaching the maximum at the upper end of the diaphragm. Moreover, the minimal size of the AVB in the upper end of the diaphragm indicated that an increased number of TVBs reduces the size of AVBs. The number of the branch VBs was less than those of the culm.

The frequency of the VB of the branch was significantly higher than that of the culm, confirming that the branching VBs were small and dense. The frequency of VBs in the culm decreased from the bottom of the BN to the nodal ridge, then increased towards the top of the BN. The frequency pattern of VBs was similar to that of the unbranched bamboo node (Li et al., 2021), with the minimum frequency of VBs located at the nodal ridge. The frequency of branch VBs increased from the bottom to the top in the BN, which was inconsistent with the culm regularity.

An advantage of µCT scanning is its ability to analyze the 3D interior tissue proportions, enabling the acquisition of more representative volume data for tissues ( Figure 5A ) and highlighting changes across continuous node slices ( Figure 5B ). The proportion of tissue in each slice is quantified as the difference between the highest and lowest values corresponding to the respective colors. The volume fraction of parenchyma, fibers, and conducting tissue in the BN of C. tumidissinoda were 61.3%, 35.3%, and 3.4%, respectively ( Figure 5A ). However, there were continuous changes in tissue proportion at different heights from the bottom to the top of the BN ( Figure 5B ). The maximum proportion of parenchyma, fibers, and conducting tissue in the node are 71.5%, 43.8%, and 14.8%, respectively. These slices are located at the top of the BN, nodal ridge, and lower end of the diaphragm, respectively. The fibers in the nodal ridge exhibited the highest proportion in the BN ( Figure 5B ), which experiences high tensile and shear stresses at the junction of different internodes. Consequently, there is a greater abundance of fibers in this area to prevent culm fracture or bending. However, fiber cells are more prone to shrinkage under the influence of moisture content (Yuan et al., 2023). The abundant fibers in the nodal ridge is an important cause of macroscopic cracking in BN, which further leads to the BN becoming the weak point of bamboo. The minimum proportion of parenchyma, fibers, and conducting tissue in the node are 48.1%, 25.8%, and 1.3%, respectively. These slices are appearing in the lower end of the diaphragm, top of the node, and near the nodal ridge, respectively. The patterns of continuous changes in tissue proportion along the bamboo culms observed in this study are similar to those reported in previous studies (Li et al., 2021). However, the proportion of fibers in this study was higher, while that of parenchyma was lower because of node branching. These results further confirmed that the enlargement and bifurcation of the VB could lead to an increase in the proportion of the VB and a decrease in the parenchyma. The bifurcation of the VB caused an increase in the fibers and conducting tissue near the sheath scar, while enlargement of the VB caused an increase in the fibers and a decrease in the conducting tissue in the nodal ridge. The change patterns in fibers and conducting tissue thickness reported in this study are consistent with those reported in earlier studies (Li et al., 2021), which demonstrated inverse patterns between the thickness of conducting tissue and fibers at the nodal ridge.

The anatomical characteristics showed the 2D information of the BN internal structure, which lacked the spatial structure and the continuous size variation related to the number and angle of the BN slices. A morphometrical study of the conducting tissue in a BN is instrumental in understanding its 3D features. Thickness is an important morphological index calculated as the diameter of a hypothetical sphere that fits within each boundary point using the Dragonfly software (Silin et al., 2011), where red denotes a larger thickness while purple denotes a smaller thickness.

The thickness of the conducting tissue in the BN was shown in Figure 6A , with a maximum of 290.09 μm, revealing a slightly helical structure. The morphology of the conducting tissue in the segments was assessed ( Figures 6B–F ) to observe the variation of the conducting tissue at different heights of the BN. Parts of the conducting tissue under the sheath scar were bifurcated, forming the transverse conducting tissue (TCT). Notably, the thickness of the TCT was greater than that of the axial conducting tissue (ACT), that facilitates the transverse transport of water. The highest thickness of the conducting tissue was under the sheath scar ( Figure 6B ). The number of TCT increased while their thickness decreased ( Figure 6C ), the numerous bifurcated conducting tissue in the diaphragm facilitated the liquid movement in different directions. The TCT above the diaphragm was mainly congregated in the branch side of the BN, their thickness was less than that of the ACT ( Figure 6D ). In the lower end of the nodal ridge, the TCT gradually disappeared with the separation of the culm and branch, and some were distributed along the outer layer of the BN near the branch ( Figure 6E ). However, the TCT was not observed in the outer layer of the unbranched bamboo node (Li et al., 2021), which is related to the branch of the bamboo node. The unbranched conducting tissue above the branch changed slightly in direction, bending outward under the nodal ridge. The inward bend above the nodal ridge revealed a slightly helical structure (Palombini et al., 2016; Li et al., 2021). The thickness of the culm was greater than that of the branch ( Figure 6F ). The structure of the BN varied at different heights, unlike the internode structure of bamboo, which only exhibited a slight change (Yu et al., 2002; Zhang et al., 2017b). Unlike the unbranched bamboo nodes (Li et al., 2021), the TCT was not distributed in a concentric radial pattern. This work has revealed that the conducting tissue of the BN was more densely distributed on the branch side of the culm. Therefore, the conducting tissue structure of the culm would also be more biased toward new growth nutritional organs. In addition, although the TCT provided the possibility for transversal water transport, the thickness of the ACT decreased after the appearance of the TCT. This could reduce the axial water conductivity of the node.

The 3D thickness model of the conducting tissue achieved a good observation of its morphological variation at different heights but did not elaborate on the relationship between the culm and branch. Image labeling provided valuable and advanced 3D insights into the intricate conducting tissue of the BN, distinguishing the volume of units by different colors ( Figure 7A ). The conducting tissue cannot be easily traced back to its origins (Kraehmer, 2017), connectivity of the conducting tissue originating from the Euler number is a fundamental topologic measure (Odgaard and Gundersen, 1993), which allows the characterization of the redundancy in structure connections (Bouxsein et al., 2010). The volume of the connected conducting tissue was 41.57 mm³ ( Figure 7B ). The connectivity in the conducting tissue with 46.76 mm³ was 88.91% ( Figure 7A ). The high connectivity of the conducting tissue in the 3D space revealed why VBs reportedly enter the diaphragm through elusive movement trails (Shao et al., 2010b; Chen et al., 2023). The disconnected conducting tissue was mostly distributed in the outer layer of the BN ( Figure 7C ). They were the unrecognizable small-sized conducting tissue and tissue bent directly into the unidentified branch above slice 722 observed during image processing. The largest conducting tissue volume, with 34.51 mm³, composed 73.81% of the total volume ( Figure 7D ). The conducting tissue of the culm and the branch were connected, forming an intricate network structure. This structure was distinct from the unconnected VBs of the culm and branch reported by Palombini et al. (2020). Some interconnected conducting tissue did not establish a connection with the largest conducting tissue ( Figure 7E ). This specific type of conducting tissue was primarily located within the outer layer of the BN. Investigating this phenomenon required the use of higher-resolution equipment. Visualization of fluid pathways within the bamboo BN was achieved by reconstructing a 3D model and analyzing the connectivity of the intersecting conducting tissue in the model. This work provided real evidence of the pathways of multidirectional fluid transport and the fluid pathways between the culm and the branch.

The thickness of the conducting tissue and fibers in the VB of the bamboo node were inversely correlated (Li et al., 2021). The fibers thickness was calculated similarly to that of the conducting tissue. The morphology of the fibers in the segments was examined to investigate the variations in their characteristics at various heights of the BN ( Figures 8B–F ).

The maximum thickness of fibers in the BN was 591.88 μm, which was greater than that of the conducting tissue. Here, the fibers were more dispersed while the conducting tissue was more concentrated. ( Figure 8A ). The thickness of the transverse fibers under the sheath scar was less than that of the axial fibers and was distinct from the regular pattern of the conducting tissue thickness ( Figure 8B ). The branching of fibers in the diaphragm was increased, forming a complex network structure ( Figure 8C ), which improved the node’s mechanical properties, including bending strength and splitting resistance (Zou et al., 2016; Huang et al., 2017). The range of fibers thickness above the diaphragm increased, especially on the branch side of the culm. This phenomenon was not observed in previous studies of unbranched bamboo nodes (Li et al., 2021), so it is inferred that the branch is accompanied by the appearance of a large number of fibers on the branch side of the culm, these fibers added structural support to the branch. The transverse fibers extended twistedly, possibly due to the squeezing between transverse and axial fibers (Chen et al., 2023). Notably, the thickness of the transverse fibers was less than that of the axial fibers ( Figure 8D ). This disparity was attributed to the requirement of axial fibers to support the weight of the entire bamboo plant, while transverse fibers are only responsible for supporting branches. Fibers in the lower end of the nodal ridge were mainly congregated in the branch side of the culm, similar to the distribution of conducting tissue ( Figure 8E ). However, fibers thickness from Slices 386 to Slices 600 increased, while that of the conducting tissue decreased, indicating that thin conducting tissue required more fibers for mechanical support and protection. Fibers in the Slices 601-1292 had the maximum thickness of the BN ( Figure 8F ), which was consistent with the location where the maximum proportion of fibers occurs. Nevertheless, conducting tissue and the fibers had similar shapes but varying thicknesses. The functional integration of mechanical reinforcement and liquid flow of the bamboo node was achieved through the thickness variations of their conducting tissue and fibers, showing the structural optimization potential of the bamboo nodes.

Image labeling provided valuable 3D insights into the complex fibers of the BN, allowing for the differentiation of volume units through the use of distinct colors ( Figure 9A ). Notably, 483.56 mm³ of the connected fibers accounted for 99.95% of 483.79 mm³ of the total fibers in the BN ( Figure 9B ). The connectivity of the fibers was higher than that of the conducting tissue because of the larger area of fibers in the cross-section, which enhanced segmentation during image processing. Some branching of the VB under the bottom of the BN got into the sheath scar and stopped growing, possibly giving rise to the disconnected fibers distributed near the sheath scar. Small-sized fibers in the branch possibly also led to a disconnection of the fibers ( Figure 9C ). The largest fibers, with 481.22 mm³ comprised 99.47% of the fibers in the BN ( Figure 9D ), with the fibers of the culm and branch twined and connected to form whole fibers. However, 0.53% of the fibers were unconnected with the largest fibers. Most of them were like disconnected fibers ( Figure 9E ) and possibly connected to other bamboo parts. While the study unveiled the interrelation of VBs between the branch and the culm, the high connectivity of fibers and conducting tissue within the BN necessitates further observation to understand the spatial relationships among VBs exhibiting diverse anatomical characteristics in the cross-sections.