Evaluation of the carbon sink potential of wooden foundation piles embedded in oxygen-depleted soils using density profile analysis

Wooden foundation piles are a renewable resource that contributes to mitigating climate change because of the minimal processing requirements and remarkably long service life. Previous studies have shown that while these piles can have a long service life, the buried wood is also susceptible to bacterial activity; however, the extent to which this affects the carbon stock in wooden piles remains uncertain. This study addressed this by estimating the decrease in the mass of a pile buried in soil using the density profile, because bacterially induced deterioration was observed to initiate on the pile surface. Piles extracted from various locations in Japan were prepared as thin specimen strips. Then, these specimens were exposed to soft X-rays, and the resulting images were acquired on film. The images were analyzed to calculate the mean wood density at 0.5 cm intervals from the pile surface. The results revealed that no significant decrease in density was observed near the surface of a pile, indicating that the bacterial effect on the carbon stock was negligible when a pile was installed in waterlogged conditions. By contrast, the possibility remains that piles exposed to oxygen-depleted conditions, such as fluctuating groundwater levels, might have experienced a reduction of approximately 10% of their initial mass after 60–84 years of exposure to soil conditions. This reduction in mass is significantly lower than the 50% reduction observed in lumber after 35 years and wooden panels after 25 years; therefore, wooden foundation piles have promising potential as carbon sinks.

Wood is a renewable resource that can contribute to mitigating climate change [1, 2]. Wooden products that assist in mitigating climate change should possess two characteristics. First, they should be prepared using low-energy consumption processes. Conventional wooden products, such as construction lumber, require substantial energy to dry the unseasoned wood [3]. In addition, laminated lumber requires additional energy inputs to prepare glue from fossil fuel-based resources. Optimal wooden products should be prepared without using these energy-intensive processes [4]. Second, they should possess a long lifespan [5]. Lumber and laminated lumber are used in above-ground conditions, where fungi and termites are active. These wooden products, especially those used in wet conditions, degrade because of fungal and/or termite infestations without preservative treatments [6, 7]. Furthermore, the majority of wooden products are demolished and burned before being degraded by biological agents because of functional obsolescence. Consequently, wooden construction products are considered to have half-lives of 35 years for lumber and 25 years for wooden panels [8].

Compared with conventional wooden products, the wooden products used in piles have preferable characteristics as carbon sinks in terms of energy consumption and lifespan [9]. For example, piles can be used without drying, significantly reducing energy consumption compared with conventional lumber. Regarding lifespan, the piles driven into the soil below the water table are safe from biological agents, such as wood-degrading fungi and termites. Thus, these piles are expected to have minimum lifespans of several decades.

In general, piles are exposed to conditions free from wood-degrading fungi and termites; however, the presence of wood-degrading bacteria under oxygen-limited conditions has been reported [10]. In addition, these bacteria are reportedly tolerant to low oxygen concentrations [11], such that minor differences in oxygen concentration can restrict the bacterial deterioration of wood [12]. The degradation of wood under oxygen-depleted conditions was studied extensively by a research group in Europe, which found that several types of bacteria are involved in degrading wood, including tunneling and erosion bacteria [13, 14]. In addition, this group revealed that bacterial deterioration of wood was initiated on the pile surface and progressed slowly toward the heartwood of the pile [15, 16]. Similar bacterial wood deterioration has been found in North America and Asia; thus, it is a global phenomenon [17].

New construction methods using softwood piles have been developed in Japan to assist in mitigating climate change [18, 19]. These methods involve installing piles into anaerobic soils as carbon sinks, where wood-degrading fungi and termites cannot survive; however, the influence of wood-degrading bacteria under these conditions remains unclear because of limited information regarding bacterially induced wood decay in Japan. In addition, a plethora of studies have been conducted to elucidate the impact of bacterial wood degradation in oxygen-depleted conditions, as mentioned in [10,11,12]; however, no reliable approach focusing on quantifying the decrease in mass of piles in these conditions has been identified. The lack of quantitative analyses stems from the difficulty associated with comparing a pile's final mass after undergoing bacterial deterioration without a known initial mass when installed several decades or hundreds of years previously. A novel approach was devised to overcome this issue that evaluated the decrease in mass resulting from bacterial activity comparing the variations in wood density at the pile surface and the interior and statistically analyzed this variation against the position within the pile. This approach is effective, because bacterial deterioration of piles progresses from the surface toward the heartwood. Here, the potential of wooden foundation piles as carbon sinks is discussed, demonstrating the utility of this approach for assessing bacterial degradation in these environments.

Preparation of specimens and soft X-ray irradiation imaging

In general, wooden piles installed in the ground were extracted by attaching a rope and removing them using heavy machinery (Table 1). Information regarding the piles is provided in Table 1, including extraction location, wood species, approximate residence time, and pile diameter and length. Further information regarding the locations and piles is available in reports by Numata et al. [20,21,22], and Hara [23].

A total of 19 softwood piles were sectioned at the top, middle, and bottom (Fig. 1) to prepare approximately 30 cm length logs, with care taken to avoid sectioning the damaged portion of the pile during extraction. Wood disks approximately 2 cm thick were prepared from these sections and dried in a conditioned room at 20 °C and 60% relative humidity (RH) until the equilibrium moisture content was attained. Four wood strips, each approximately 2 cm wide, were sectioned from each direction on each disk (Fig. 1). During this sampling process, with the tree pith positioned away from the center of the cross-sectional disk, the strips were sectioned from areas containing both wider and narrower annual ring spacings. In addition to these strips, small specimens were collected from the piles to determine the wood species using microscopic observation.

Each strip was sectioned using a circular saw (Model CS-800, Hisada Tekko, Ltd., Japan) in the end-grain direction to produce two specimens, each 2 mm thick (Fig. 1), for soft X-ray analysis. However, after sectioning, the thickness values of the specimens were not uniform and deviated from the initially planned 2 mm. To correct the specimen thickness, it was measured at 2 cm intervals from the pile surface using a Digital Linear Gauge DG-750 (Ono Sokki, Japan). In addition, the thickness at the interior end was recorded. The amended thickness was calculated based on the distance from each measured point and its respective thickness using the following equation:

where x is the distance from the pile surface at which the amended thickness should be determined, x0 and x1 represent the distances from the surface at the nearest points, where thicknesses were measured using the Linear Gauge, and t0 and t1 are the thicknesses at x0 and x1, respectively, obtained from the Linear Gauge measurements.

These samples were stored in a conditioned room (20 °C and 65%RH) until irradiated using a soft X-ray apparatus (custom EMB-W model, Softex Corp., Japan). The specimens and a density standard comprising cellulose acetate were placed directly on a film (Industrial X-ray film IXFR 20.3X25.4CM, Fujifilm, Japan) and irradiated with soft X-rays for 4 min at 20 kV and 14 mA to acquire images of the specimens and standard. Then, the films were developed and dried overnight. A total of 200 images, including the specimens and density standard, were scanned with a flatbed scanner (GT-X980, Seiko Epson Corp., Japan) in transparency mode to obtain 16-bit grayscale images at a resolution of 12,800 dpi for subsequent analysis.

Analysis of film images

Scanned image data were analyzed using a program written in Python [24]. Image data for the density standard, consisting of 14 different thickness areas (Fig. 2), were analyzed first. A circle with a radius of 1500 dots was set at the center of each area. The mean gray levels in the circles were calculated based on the gray level at each point within the circle, and they were plotted against the areal wood density at 2 mm thickness (g/cm²) obtained from the areal density of cellulose acetate and a conversion factor provided by the manufacturer of the soft X-ray apparatus. An example of such a plot is shown in Fig. 3. The solid line in the figure was obtained by approximating the points with a quadratic equation determined using NumPy’s polyfit [25].

The gray level for each film was converted to areal density at 2 mm thickness using the inverse function derived from the quadratic equation for each film (Eq. 3):

Finally, the apparent wood density (g/cm³) was obtained by multiplying the areal density at a thickness of 2 mm by a factor of 5 (Eq. 4):

The workflow outlining the data analysis procedure for the soft X-ray images is shown in Fig. 4. Image data (16-bit, 12,800 dpi) for the specimens were initially rotated such that the pile surface appeared on the left, and the interior end was positioned on the right. A new 16-bit gray image was created by copying a rectangular area of the original rotated image. The rectangular image comprising gray level data was then converted into a new image indicating the areal density, calculated using the inverse functions (Eq. 3). The apparent density profile (i.e., the density value before thickness adjustment) was determined by collecting apparent density values along a line traversing from the pile surface side of the specimen to the interior side to obtain the density profile of each specimen. Based on the thickness of the specimen obtained using Eq. 1, the apparent density profile was amended using the following equation:

The mean density of each specimen was calculated by integrating the density values over the specified range on the density profile.

The peak positions and peak heights (i.e., densities) were determined by visual inspection with the assistance of SciPy's find_peaks function [26]. Peak heights with a notably high value resulting from adhered soil were removed from the peak list. The obtained peak positions and peak heights, representing density, were saved and used for further analysis. Density data between two adjacent peaks were examined to determine the lowest earlywood density. The mean density, based on a minimum of 50 data points, was calculated to determine the lowest earlywood density, because the density in the earlywood area did not show an obvious peak similar to that observed in the latewood area, and the degraded area must have a lower density. In addition, areas with low densities caused by vertical resin ducts and cracks were excluded from the density counting area. In addition, the mean lowest earlywood density and lowest earlywood position were collected for further analysis. The annual ring width was calculated by dividing the distance from the pile surface by the number of peaks obtained through visual observation.

The maximum distance from the pile surface was determined to ensure an adequate number of specimens to allow for the statistical significance of the differences between line segments to be evaluated. The range used for calculations was set such that the number of peaks in the innermost region was at least one-third of the number of peaks on the pile surface. Data in this range were first evaluated using the Kruskal–Wallis test, and if significant differences were found among each line segment, they were further evaluated using Dunn’s test. These tests were performed using the Kruskal function from the scipy.stats module [26] and the posthoc_dunn function from the scikit_posthocs module with Holm's p-adjustment [27].

Properties of the piles and specimens

Based on microscopic observation and the distribution status of log species, the 19 wooden piles were identified as three wood species: Japanese cedar (Cryptomeria japonica D. Don), Japanese larch (Larix kaempferi (Lamb.) Carr.), and Japanese red pine (Pinus densiflora Sieb. et Zucc). The residence times for the piles installed in anaerobic soils ranged from approximately 10 years for Japanese cedar and Japanese larch to greater than 60 years for Japanese red pine. The length and diameter of the piles ranged from 1 to 7 m and 11 cm to 25 cm, respectively. All piles appeared to be exposed to reducing environments because of the blue soil color or surrounding moisture conditions, except for piles denoted as "Saga" that had been embedded for creek bank stabilization, with their tops exposed to the atmosphere. Visual observation of the top of "Saga" piles suggested that they were heavily infested with termites; therefore, they were removed from further calculations.

Preparation of density profiles based on soft X-ray images

Density profiles were prepared based on the soft X-ray irradiation images according to the following steps:

• Step 1: Image scanning.

  The image on each film was scanned in transparency mode to obtain a positive 16-bit grayscale image at a resolution of 12,800 dpi.

• Step 2: Elucidation of the relationship between gray level and areal density.

  The image of the density standard was used to determine the relationship between the mean gray level and the areal density (at 2 mm thickness). An example of such a plot, shown in Fig. 3, revealed that the constants a, b, and c had values of − 14,083, 45,596, and 5144, respectively. These constants were used to convert gray level to areal density according to Eqs. 3 and 4.

• Step 3: Preparation of apparent density images from grayscale images.

  Image data associated with the specimens were rotated so that the surface side of the pile was positioned on the left. A new 16-bit, 12,800 dpi image was created by copying the central area of the specimen’s image, focusing on the pile surface toward the heartwood. Each dot in the new image represented the gray level and was then converted to an apparent density using Eqs. 3 and 4. Sample images before and after the conversion are shown in Fig. 5. This conversion changed the scale of the image from the gray level to apparent density. The term "apparent" indicates that the density was not corrected when adjusting a specimen’s thickness.

• Step 4: Preparation of plot profiles of the apparent densities against distance from the pile surface.

  The apparent densities, following a line along the center of the image, were obtained and plotted against the distance from the pile surface. A sample is shown in Fig. 6. Notably, the apparent density profiles obtained from these specimens had a mean thickness of 2.13 mm and a standard deviation of 0.083 mm and thus, correcting the apparent density to the density based on the thickness of a specimen was necessary.

• Step 5: Correction of the apparent density to the density.

  The apparent density was corrected using the values from a specimen with a thickness of 2 mm by dividing the apparent density value by the thickness of the specimen at each position to obtain the density profile. An example of a density profile is shown in Fig. 7. In this case, the density value in Fig. 7 decreased by approximately 10% from the apparent density value, because the specimen thickness was greater than 2 mm. The numbers at the peaks and x marks shown in Fig. 7 are explained later.

Determination of the mean, peak, and lowest earlywood densities and their positions

The density values for each specimen in each line segment were integrated to calculate the density distribution in 0.5 cm intervals. These results were grouped according to Table 1 and plotted in Fig. 8.

The peak position (the numbers in Fig. 7) and densities were determined using visual observation with the assistance of SciPy's find_peaks function. The lowest earlywood density (the x marks in Fig. 7) was determined based on the mean value of the lowest 50 density values between two peaks after removing negative peaks areas originating from vertical resin canals (Fig. 7, between peaks 4 and 5). The peak density and the lowest earlywood density values were extracted at 0.5 cm intervals from the pile surface based on the peak positions and lowest earlywood positions, respectively. Then, these values were grouped according to Table 1 and plotted to investigate the relationships between density and distance from the pile surface, as shown in Fig. 9.

Analysis of the density distribution against the distance from the pile surface

The Kruskal–Wallis test and Dunn’s test were applied to investigate whether significant differences exist in the mean density among the line segments at 0.5 cm intervals. The results showed no significant difference between the mean density near the pile surface and those in other areas (Table 2). In addition, the peak density and lowest earlywood density were evaluated using these tests (Table 2). When both tests indicated a significant difference, the results of Dunn’s test were used to visualize the line segments exhibiting statistically significant differences in density. As shown in Fig. 10 and Table 2, a significant difference is observed only in the lowest earlywood density, especially between the outer and inner line segments.

A comparison of the mean densities of pine-Fukui and pine-Tokyo found a significant difference (Fig. 11). The mean of the mean densities in pine-Tokyo was 12% lower than that of pine-Fukui, respectively. The mean densities of other piles were not compared, because they represented different wood species, and their initial densities likely differed. The comparison of densities from other pine piles, including peak density and lowest earlywood density, also showed significant differences. The mean peak density of pine-Tokyo was 15% lower than that of pine-Fukui, and the mean lowest earlywood density of pine-Tokyo was 21% lower than that of pine-Fukui. The Kruskal–Wallis test and Dunn's test showed that the mean, peak, and lowest earlywood density values of pine-Fukui and pine-Tokyo differed significantly.

The differences in density among the top, middle, and bottom specimens of pine-Fukui indicated that the top section had the highest density (Fig. 12), and the significance tests showed that the density at the top differs significantly from the middle and bottom densities. By contrast, pine-Tokyo did not exhibit a similar pattern (Fig. 12).

Wooden products used in pile foundations have preferable characteristics for mitigating climate change, such as low energy consumption and extremely long service life [10, 18]. The piles in anaerobic soils show such long lifespans, because the major biological agents that degrade wood in aerobic conditions degrade wood sparingly in anaerobic soil beneath the water table [28, 29].

Previous studies investigating the microorganism DNA in foundation piles driven into soil showed the presence of various bacterial DNA, while fungal DNA was limited [30, 31]. In addition, the importance of bacterial decay on foundation piles in soil environments was reported in a series of studies conducted by Klaassen et al. [13,14,15,16]. The findings revealed that bacteria rather than fungi caused the deterioration of wood in low oxygen conditions and that bacterial deterioration began at the pile surface and migrated gradually toward the heartwood. Initially, only a small proportion of cell walls near the wood surface were degraded by bacteria. Subsequently, the area undergoing deterioration expanded with bacterial migration. The estimated speeds of bacterial degradation of wooden piles were 0.18 mm/year for spruce and 0.32 mm/year for pine [15], and this deterioration ceased at the boundary between sapwood and heartwood. In contrast with these reports, another paper found that piles were largely unaffected by degradation [32]. As mentioned by Björdal et al. [12], the slight differences in oxygen concentration are presumed to be the mechanism underlying the differences observed in these papers.

As mentioned in previous papers, fungal deterioration can indeed be disregarded in oxygen-depleted conditions, leaving bacterial deterioration as the primary process affecting foundation piles in these conditions. However, the extent of mass loss due to bacterial activity remains unclear, as prior studies used largely qualitative microscopic observations with limited focus on quantitative analyses. A previous study tried to estimate the reduction in mass of wooden piles in anaerobic conditions [33], but it contained a fundamental flaw in estimating the decreases in mass in the absence of initial values. Thus, this study adopted a statistical approach to estimate the reduction in mass due to bacterial decay in pile surfaces under oxygen-depleted conditions. Because the initial pile densities were unknown, the quantitative changes that occurred in wood density were compared using measurements ranging from the pile surface to the interior across various regions in multiple piles.

A total of 19 piles from seven distinct locations were obtained for the quantitative analysis. The wood species used in the piles and their respective residence times in soil were as follows: Japanese cedar, approximately 10 years; Japanese larch, approximately 10 years; Japanese red pine, approximately 60–84 years. The specimens used for the density analysis were selected from the top, middle, and bottom areas of the piles to elucidate the relationship between burial depth and its impact on pile deterioration. The top area was potentially most susceptible to bacterial decay, and the middle and bottom areas might have been less susceptible, because the bacterial degradation of wood proceeds at an accelerated rate in well-oxygenated soil [12], even though the bacteria may be tolerant of strictly anaerobic conditions. The differences in soil compositions may influence the rate of deterioration; however, the analysis was conducted without considering these factors.

In this investigation, 16-bit grayscale images were obtained at resolutions of 12,800 dpi (Fig. 5a). All initial images were converted step-by-step (Figs. 5b and 6) to obtain a density profile showing the relationship between density and distance from the pile surface (Fig. 7). Moreover, the peak density and lowest earlywood density of each annual ring were obtained from these density profile data. Moreover, the peak and lowest earlywood densities in each annual ring were obtained using density profile data. The latter was hypothesized to deteriorate most rapidly because of bacterial activity [32], while the former was expected to deteriorate more slowly. Therefore, this study focused on the lowest earlywood density and the peak density based on the understanding that the former would be most susceptible to bacterial degradation, and the latter least.

The statistical analysis of the mean, peak, and lowest earlywood densities was conducted for each of the following groups: (1) pine piles from the Fukui area; (2) pine piles from the Tokyo area; (3) cedar piles; and (4) larch piles (Table 1). The pine piles were separated into the Fukui and Tokyo groups, because there were differences in the size of these piles and the environments in which they were installed. The results of the statistical analysis showed no significant differences between the pile surface and other areas in terms of the mean and peak densities. This novel approach clearly indicates that the deterioration of wooden piles resulting from bacterial activity in waterlogged soil was minimal. By contrast, the lowest earlywood density changed based on distance from the pile surface (Table 2 and Fig. 10), suggesting that the earlywood area had begun to deteriorate because of bacterial activity. Although a significant difference was observed among the lowest earlywood densities, this difference was minimal, because the mean density was unaffected by distance from the pile surface despite the significant difference in the lowest earlywood density. While this difference could potentially affect the mean density with longer residence times, in this specific case, the variation in the lowest earlywood density did not substantially impact the mean density. These results are consistent with recent findings reported in a study by Horisawa et al. [31]. They examined piles that had been buried in anaerobic conditions for approximately 50 years and found that bacterial degradation of the cell walls was observed in only a small portion of the entire structure.

Conversely, the mean density of the piles installed in Tokyo area soils showed an approximately 12% lower value than in Fukui, explainable mainly by three factors. First, the differences in the initial masses resulted from differences in the sizes of the piles. In Japanese red pine, the density of juvenile wood is lower than that of mature wood, which can be estimated on the basis of the results shown in Fig. 12. The variation in density among the three sections of pine-Fukui reveals that the mean density of the top section is highest when the ratio of juvenile wood is lowest. As shown in Table 1, pine-Tokyo comprised smaller diameter sections, indicating a higher percentage of juvenile wood. Second, the differences in annual ring width. The narrow annual rings in pine-Fukui probably resulted in increased density compared with pine-Tokyo, whose annual rings were wider than pine-Fukui (Table 3). Third, the environmental conditions to which the piles are subjected impact the mean density. Large-scale groundwater extraction in Tokyo caused a significant drop in the groundwater level, reaching depths of several tens of meters, for several decades [34]. Subsequently, the anaerobic conditions surrounding the wooden piles were altered to nearly anaerobic conditions because of the reduced groundwater level, which may have led to the deterioration of the wooden piles through bacterial activity and other biological factors. The greater difference in the lowest earlywood density between pine-Fukui and pine-Tokyo (21%) compared with the difference in peak density (15%) suggests that microorganisms, including bacteria, preferentially decompose earlywood instead of latewood [32]. In addition, the variation in density among the three sections of pine-Tokyo was investigated, and no significant difference was found (Fig. 12). This is likely because of the short length of the pine-Tokyo and the significant reduction in the groundwater level in Tokyo, which is thought to exceed the burial depth of the piles, and thus investigating additional specimens that possess intermediate properties between pine-Fukui and pine-Tokyo is necessary.

Finally, the results were compared with the confidential values presented in a study prepared for the Intergovernmental Panel on Climate Change [35]. In this article, the mean densities and their standard deviations were 451 ± 66 kg/m³ (P. densiflora), 314 ± 32 kg/m³ (C. japonica), and 404 ± 39 kg/m³ (L. kaempferi), respectively. These values represent the basic density determined by dividing the oven-dried mass by the green volume, while the densities in this study were obtained by dividing the air-dried mass by the air-dried volume. Considering the difference between the basic and air-dried densities, the density observed in this study mostly aligns with the range considered to be the typical density of non-deteriorated wood.

The decrease in density near the pile surface in the lowest earlywood might suggest an early stage of bacterial decay beginning in a section of the earlywood; however, the degree of deterioration is insufficient to cause a significant decrease in the mean density near the pile surface. The results showed no significant difference between the mean density near the pile surface and the interior area, and a comparison of the density values in this study with literature values clearly suggests that the decrease in density induced by bacterial activity was negligible under anaerobic conditions for approximately 85 years. Compared with piles in anaerobic conditions over similar periods, those in nearly anaerobic conditions showed a 12% reduction in mean density. This difference likely results from three factors: (1) the difference in the initial mass resulting from the ratio of mature to juvenile wood; (2) the wider annual ring width; and (3) the activity of wood-degrading biological agents in nearly anaerobic environments. In addition, even under the assumption that all the difference is entirely due to biological activity, it is considerably lower than either lumber or wooden panels, which are 50% at 35 years and 15 years, respectively. These results clearly indicate the excellent potential of wooden piles to function as carbon sinks, even in nearly anaerobic conditions.

Data will be made available on request.

This research was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number 22H02410. We would like to express our gratitude to Mr. Shigeru Takahara (Japan International Forestry Promotion and Cooperation Center), Mr. Osamu Yamaguchi (Saga Prefectural Forestry Experiment Station), Mr. Koji Suzuki (Obayashi-Tobu Yanagida JV Skytree Station Project Office), and Forestry Agency-subsidized project entitled "Survey on the Evaluation of CO₂ Accumulation in Underground-utilized Wood as Soft Ground Countermeasures) for their assistance in extracting the wooden piles used in our experiments. We deeply appreciate Jiro Funyu's significant contributions to the specimen preparation process.

This study was funded by Japan Society for the Promotion of Science (22H02410).

Authors and Affiliations

1. Department of Wood Improvement, Forestry and Forest Products Research Institute, 1 Matsunosato, Tsukuba, Ibaraki, 305-8687, Japan

   Ikuo Momohara

2. Department of Wood Properties and Processing, Forestry and Forest Products Research Institute, 1 Matsunosato, Tsukuba, Ibaraki, 305-8687, Japan

   Kana Yamashita & Yuka Miyoshi

3. Civil Engineering Headquarters, Tobishima Corporation, 1-8-15 Konan, Minato-Ku, Tokyo, 108-0075, Japan

   Takumi Murata

4. Soilwood, 1-18-1-2-506 Sengendai-Nishi, Kosigaya, Saitama, 343-0041, Japan

   Atsunori Numata

Contributions

I Momohara: conceptualization, data curation, formal analysis, funding acquisition, investigation, resources, software, visualization, writing—original draft, and writing—review and editing. K Yamashita: data curation, investigation, resources, and writing—review and editing. Y Miyoshi: funding acquisition, investigation, resources, and writing—review and editing. T Murata: investigation, resources, and writing—review and editing. A Numata: conceptualization, funding acquisition, investigation, resources, and writing—review and editing.

Corresponding author

Correspondence to Ikuo Momohara.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this manuscript, the authors used ChatGPT and Gemini to polish the writing. After using these services, the authors reviewed and edited the content as needed and accepted full responsibility for the content of the published article.

Competing interests

Dr. Numata A is an independent consultant promoting the use of wooden piles for in-ground applications. Dr. Murata T is an employee of Tobishima Corporation, a company specializing in construction methods using wooden piles. These relationships have been disclosed to the other authors and do not affect the integrity or objectivity of the research presented in this paper. The other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions