Anatomical and Chemical Properties Analysis of Korean Broussonetia kazinoki Bast Fibers and Selection of Superior Individuals

In this study, one-year-old stems of Broussonetia kazinoki were collected annually from 2015 to 2019 for experiments. Samples used in this study were transplanted and managed at the Bioenergy Crop Center of the National Institute of Crop Science (Muan-gun, Jeollanam-do, South Korea) after they were collected from 43 regions across Korea (Table 1). Transplanted samples showed successful root establishment and active growth. When collecting 1-year-old stems, individuals with good growth were selected. To ensure uniformity and to accurately assess fibrous properties, stems were harvested between November and February when the growth ended.

Five selected B. kazinoki stems were divided into upper, middle, and lower segments and bast fibers measuring 3 cm and 1.5 cm in length were prepared from each segment for analysis. For delignification, fiber samples were soaked in Schultze’s solution (100 mL of 35% HNO₃ with 6 g of KClO₃) at room temperature for 7 days. Afterward, they were washed with distilled water and used for measurement.

For staining, delignified bast fibers were immersed in a double-stain solution prepared by mixing 0.5% Astra blue solution (0.5 g Astra blue and 2 mL glacial acetic acid dissolved in 98 mL distilled water) and 1% Safranin solution (1 g Safranin dissolved in 99 mL distilled water) at a 1:1 ratio for 4 minutes. Stained fibers were then observed using an optical microscope (CX23, Olympus, Tokyo, Japan). An eXcope XCAM UHD SE (Sony, Tokyo, Japan) was used to analyze their anatomical characteristics. Fiber length was measured at 40× magnification, while fiber width and lumen width were measured at 400× magnification. Measurements were taken 20 times for each segment (upper, middle, and lower) from five stems to ensure accuracy and reliability. These measurement values were used to calculate various fiber indices, including the Runkel ratio (RR), aspect ratio (AR), flexibility coefficient (FC), and rigidity coefficient (RC; Sadiku et al., 2016).

Climate data for Muan (Station 699) from 2014 to 2018 were obtained from the Korea Meteorological Administration's data portal. Parameters collected included average temperature, average maximum temperature, average minimum temperature, and total precipitation.

The bark was peeled from collected B. kazinoki stems and the outer bark was carefully removed using a knife to isolate bast fibers. Samples were then cleaned, dried, and ground with a microfine grinder followed by sieving using a 40–80 mesh sieve. Chemical composition analysis was performed according to the standards of the Technical Association of the Pulp and Paper Industry (TAPPI). Samples were defatted using organic solvents (TAPPI T 204 cm-97) and used to determine 1% NaOH solubility (TAPPI T 212 om-02), Holocellulose content (TAPPI T 9 wd-75), Klason lignin content (TAPPI T 222 om-02), and ash content (TAPPI T 211 om-02).

Descriptive statistics, including minimum, maximum, mean, and SD, were calculated using R Studio. Multivariate analyses, including principal component analysis (PCA) and k-means clustering, were conducted to classify samples. The elbow method was used to determine the optimal number of clusters. Various R packages, including stats, factoextra, FactoMineR, and cluster, were used for statistical analysis.

Anatomical properties of plant fibers are influenced by various factors such as geographic origin, fiber position, climate, and harvest time, all of which can significantly affect the quality of bioproducts made from these fibers (Yan et al., 2014). Climate variability, particularly changes in temperature, can promote the growth of certain fiber crops. If the temperature exceeds optimal ranges, it might reduce fiber length and strength. Similarly, changes in precipitation can impact water availability, which in turn can affect fiber growth and quality (Mall et al., 2017).

Bast fibers of B. papyrifera, a species in the same genus, have bene reported to have an average fiber length of 10 mm (range, 6 to 20 mm) and an average fiber width of 30 μm (range, 25 to 35 μm; Ilvessalo-Pfäffli, 1995). Over five years of observation, the fiber length of B. kazinoki bast fibers peaked at 10.85 mm in 2017. It was the shortest at 8.34 mm in 2015 (Table 2). Fiber width was the largest in 2018 (23.25 μm). It was the smallest in 2015 (18.38 μm). Lumen width was the widest in 2018 (12.44 μm). It was the narrowest in 2015 (9.86 μm). These findings revealed variations of 2.51 mm in fiber length, 4.87 μm in fiber width, and 2.58 μm in lumen width. These variations might be attributed to climate conditions.

The correlation analysis between fiber length and climatic factors showed weak positive correlations of fiber length with average minimum temperature (correlation coefficient: 0.2332) and total precipitation (correlation coefficient: 0.2921; Fig. 1). Previous studies have indicated that wood fiber length is greatly influenced by temperature and that fiber growth can be promoted up to a certain threshold as temperature rises (Jozsa and Middleton, 1994; Kilpeläinen et al., 2003). Bast fibers of Kenaf (Hibiscus cannabinus L.) are also utilized in various industries through cultivation. The optimal growth temperature for bast fibers of Kenaf (H. cannabinus L.) is known to be 22°C–30°C (Faruk and Sain, 2014; Lewin, 2006). The positive correlation between fiber length and minimum temperatures suggests that excessively high maximum temperatures beyond the optimal range may adversely affect growth. Increased total precipitation promotes growth by enhancing nutrient and water supply, positively influencing cell growth (Fritts et al., 1991; Honjo et al., 2005; Ryan, 2010). In this study, it appeared that an increase in precipitation positively affected cell length.

Fig. 1. Correlations between anatomical properties of Broussonetia kazinoki bast fibers and climate factors.

Download Original Figure

Fiber width showed a strong positive correlation with lumen width (correlation coefficient: 0.8648). This indicates that wider fibers tend to have larger lumens and thicker cell walls. Fiber width exhibited negative correlations with maximum temperature (correlation coefficient: –0.2919) and total precipitation (correlation coefficient: –0.2554), suggesting that excessive temperatures and precipitation might hinder fiber width development. Although increased precipitation generally promotes fiber growth, excessive or concentrated rainfall may hinder growth (Bakhshi et al., 2011).

Cell wall thickness was the only trait that showed significant correlations with all climatic factors. It decreased with increasing temperature and precipitation. This decrease in cell wall thickness might be attributed to increased transpiration caused by high temperatures and excessive water supply. While cell growth may increase, cell wall thickness tends to become thinner under such conditions (Mäkinen et al., 2002; Xu et al., 2015).

Anatomical and chemical properties of bast fibers from 43 regional samples were analyzed and grouped using K-means clustering. A total of twelve variables were examined, including eight related to anatomical characteristics and four related to chemical compositions (Table 3). To reduce dimensionality, PCA was performed before clustering (Table 4). The first three principal components (PC1, PC2, and PC3) had eigenvalues greater than 1, explaining 72.59% of the total variance. PC1 accounted for 34.57% of the variance. It was associated with high positive loadings for RR and RC. It was also associated with high negative loadings for FC and lumen width. PC2 explained 23.77% of the variance. It was positively associated with fiber length, fiber width, and cell wall thickness. PC3 accounted for 14.23% of the variance. It was negatively associated with NaOH solubility.

Based on PCA results, K-means clustering categorized samples into three clusters, with the optimal number of clusters determined using the elbow method (Fig. 2). The three clusters displayed distinct distributions of principal components. Cluster 1 characterized by high FC and lumen width values represented flexible but less strong fibers. Cluster 2 exhibited notable fiber length, width, and cell wall thickness, indicating structurally strong and thick fibers. Cluster 3 showed high RR and RC values, suggesting durable and strong fibers. However, geographic distributions of clusters did not correspond to specific regions. This indicates that traits of bast fibers are influenced more by human-mediated distribution than by geographic factors.

Fig. 2. K-means clustering analysis of anatomical and chemical properties of Broussonetia kazinoki bast fibers. (a) Optimal number of clusters determined using the Elbow method; (b) Cluster distribution based on principal component analysis; (c) Geographic distribution.

Download Original Figure

Hanji made with B. kazinoki as a raw material exhibits differences in anatomical properties, absorption, and ink dispersion depending on compositions and characteristics of the material (Jung, 2007). To standardize the quality of Hanji, which is influenced by properties of its raw materials, 43 samples of B. kazinoki bast fibers were analyzed to select superior individuals. During the process of removing non-cellulose components to produce bleached bark (Baekpi) during Hanji manufacturing, raw materials with low intermediate alkali content, low alcohol-benzene extractives, and high cellulose content were found to be suitable for producing high-quality Hanji (Go and Jeong, 2018).

Therefore, in this study, superior individuals were selected based on holocellulose content and fiber length. Holocellulose is a key component required for Hanji manufacturing. Thus, individuals having holocellulose contents within the top 10% were identified first. From these top 10% individuals, one individual each was selected based on fiber length: the shortest, the longest, and an individual with an average fiber length (Fig. 3). This approach ensured that individuals with excellent holocellulose content and diverse fiber length characteristics were secured, providing genetic resources suitable for various industrial applications such as paper production. According to our selection criteria, samples 6, 34, and 39 were identified as superior individuals.

Fig. 3. Selection of superior individuals based on holocellulose contents and fiber lengths of Broussonetia kazinoki bast fibers.

Download Original Figure