Non-invasive assessment of bamboo leaf optical properties and microstructural changes under light stress using time-domain OCT Authors: Jumar G. Cadondon, Hiyam A. Halapit, Maria Cecilia Galvez, Tatsuo Shiina, Edgar Vallar Published: 2026 ABSTRACT Bamboo is a vital renewable resource in tropical Asia, yet its physiological responses to environmental Bamboo is an ecologically and economically important renewable resource in tropical regions, yet its physiological response to light-induced stress remains poorly characterized under in vivo conditions. In this study, time-domain optical coherence tomography (TD-OCT) was employed as a non-invasive imaging technique to monitor microstructural and optical changes in Bambusa spinosa leaves subjected to controlled light stress. A portable TD-OCT system was used to quantify extinction coefficients and palisade layer thickness following repeated exposure to high-intensity illumination over several days. These measurements were complemented by Vegetation Ratio (VR) index analysis derived from optical microscopy to assess pigment-related changes. The results reveal a systematic decrease in extinction coefficients in both epidermal and mesophyll boundary layers with increasing exposure duration, indicating progressive tissue degradation. Mean extinction coefficients declined from approximately 1.03 mm⁻1 to 0.72 mm⁻1 under prolonged illumination, while palisade layer thickness showed a decreasing trend consistent with stress-induced structural alteration. Variations in VR index values correlated with OCT-derived optical parameters, reflecting concurrent changes in leaf optical properties and physiological condition. The ability of TD-OCT to perform repeated, label-free measurements on intact leaves enables longitudinal monitoring of stress responses. These findings highlight the potential of TD-OCT combined with VR index analysis as a sensitive tool for early stress detection and precision monitoring in sustainable bamboo cultivation and environmental assessment.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-51400-6. Introduction Bamboo, a woody grass belonging to the Poaceae family, is renowned for its extraordinary growth rates and unique structural properties. Globally, species such as Phyllostachys edulis (Moso bamboo) have demonstrated rapid growth—up to 114.5 cm per day—driven by intercalary meristems and complex hormonal regulation involving gibberellins, auxins, and cytokinins1. This rapid development is further modulated by environmental factors such as light intensity, soil nutrients, and moisture availability. In the Philippines, bamboo species thrive under diverse environmental conditions and exhibit adaptive growth behaviors aligned with tropical climate variations. Recent studies emphasize light intensity as a critical determinant of bamboo development. For instance, Bambusa blumeana cultivated in open, high-light environments produced more robust culms and extensive rhizome networks than those grown under shaded conditions. Similarly, Dendrocalamus asper demonstrated enhanced culm diameter and internode length when grown under full sunlight, underscoring the significance of optimizing light exposure in agroforestry practices2. Because of its wide ecological distribution, rapid physiological responses, and sensitivity to changing environmental conditions, bamboo has emerged as a potential indicator plant for environmental stress. Its growth dynamics, leaf coloration, and structural characteristics respond measurably to variations in light intensity, temperature, water availability, and soil quality. These attributes make bamboo a valuable biological sensor for assessing ecosystem health and monitoring the impacts of climate variability, pollution, and land-use change in tropical environments. In vivo monitoring of bamboo microstructure under natural conditions is essential for understanding how Philippine bamboo species respond to such environmental stressors. Cadondon et al.3 utilized time-domain optical coherence tomography (TD-OCT) to compare healthy and diseased leaves of Bambusa spinosa (Bambusa sp.), a native Philippine bamboo species. Their findings revealed distinct microstructural changes and measurable reductions in chlorophyll-a content and extinction coefficients in diseased leaves. This demonstrates the potential of TD-OCT as a non-destructive technique for assessing bamboo leaf physiology in situ—supporting its application as an environmental stress indicator in ecological monitoring and plant health diagnostics. Bamboo growth typically progresses through three major phases: shoot emergence, elongation, and maturation. During the elongation phase, bamboo operates heterotrophically, reallocating carbon from mature culms via rhizomes to support new shoot development4. Light intensity not only drives photosynthesis but also influences morphogenetic responses. In Phyllostachys pubescens, variations in light regimes significantly impact carbon allocation patterns, biomass accumulation, and shoot morphology5. Similar observations have been reported for tropical bamboo species, where increased light availability correlates with improved growth performance Pioquinto-Laguardia et al.6. Furthermore, Fan et al.7 demonstrated that adequate light exposure significantly affects the number of emerging shoots and culm vigor, further validating light as a key determinant of bamboo growth. Structurally, bamboo culms consist of vascular bundles embedded in a parenchymatous matrix, contributing to both nutrient transport and mechanical strength. High-resolution imaging studies, such as those on Qiongzhuea tumidinoda, quantified node tissue composition as approximately 65.7% parenchyma, 30.5% fibers, and 3.8% conducting tissues8. Variations in vascular bundle arrangement between nodes and internodes influence both hydraulic efficiency and mechanical performance9. While conventional microscopy techniques provide valuable anatomical insights, they often require invasive sample preparation. Non-destructive approaches such as X-ray microcomputed tomography (μCT) have emerged as powerful tools for visualizing 3D internal structures, revealing the spatial organization of vascular bundles and their relationship to mechanical properties such as fracture toughness and flexibility10. Optical coherence tomography (OCT), a label-free, non-invasive imaging modality based on low-coherence interferometry, enables real-time, micrometer-resolution imaging of tissue microstructure11–14. Although widely applied in biomedical research, OCT remains underutilized in bamboo studies. Findings from other plant systems suggest that OCT can effectively visualize vascular patterns, monitor water transport, and assess structural differentiation in vivo15,16. Integrating OCT into bamboo research holds great promise for non-destructive, longitudinal monitoring of environmental stress responses, providing a foundation for using bamboo as a bioindicator species in both ecological and agricultural contexts. In this study, the leaf optical properties—specifically the optical thickness and extinction coefficient—were measured using a portable TD-OCT system. The effects of light stress were examined based on changes in these optical properties, supplemented by Vegetation Ratio (VR) Index analysis and visual observations. The results aim to advance the understanding of how bamboo responds optically and structurally to light-induced stress, reinforcing its role as an indicator plant for environmental monitoring. Materials and methods Theoretical perspective Beer lambert law for OCT The Beer-Lambert law, also known as the Beer–Lambert–Bouguer law or simply Beer’s law, illustrates the relationship between the attenuation of light and to the properties of the material the light travels through. Since our TD-OCT only measures the backscattered light, the detected backscattered OCT signal depends on how much light is attenuated as it enters and leaves the tissue. Hence, the OCT intensity is measured as:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${I}_{OCT}\left(z\right)= {I}_{o} R\left(z\right){e}^{-2\mu z}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${I}_{0}$$\end{document} is the incident optical intensity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R(z)$$\end{document} represents the local reflectivity or scattering strength at depth \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$z$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mu$$\end{document} is the extinction coefficient of the plant tissue. The factor of 2 in the exponential accounts for the round-trip propagation of light—once during penetration into the tissue and once during its return to the detector. The local reflectivity of scattering strength at depth z is given by R(z), and μ is the extinction coefficient of the plant tissue. The OCT A-scan signal in logarithmic scale becomes linear, where as2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{ln}{I}_{OCT}\left(z\right)=\mathrm{ln}\left({I}_{o}R\left(z\right)\right)-2\mu z$$\end{document} A straight line is fitted in the logarithmic OCT intensity which is related to the slope given by3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mu = -\frac{1}{2}(slope)$$\end{document} Development of TD-OCT Optical coherence tomography (OCT) is fundamentally based on low-coherence interferometry, which employs a Michelson interferometer illuminated by a broadband light source (Fig. 1). In the developed TD-OCT system, a superluminescent diode (SLD) is used as the broadband light source to generate a low-coherence optical beam. This light then travels through the 50:50 beam splitter and is divided into two beams: a sample beam and a reference beam. The sample beam travels through the sample arm and exits the probe, which consists of a focusing lens and a collimator. When the beam penetrates the sample, the light is backscattered. A portion of those light travels back into the probe and is collected by the photodiode detector. The probe that illuminates the light into the sample is designed to be compact, flexible, and long depth of field.Fig. 1(A) Schematic Diagram of the TD-OCT system17 and (B) actual OCT set-up and measurement. (A) Schematic Diagram of the TD-OCT system17 and (B) actual OCT set-up and measurement. On the other hand, the reference beam travels through the reference arm, exits through a collimator, and directed toward a reflector or reference mirror. In standard Michelson interferometers, the reflector typically uses translational motion but the particular OCT used in the study contains a modified reflector that undergoes rotational motion. The rotating reference mirror provides a longer scanning range, as well as improved repeatability when scanning. The beam reflected through the reference mirror travels back through the probe and is collected by the photodiode detector. These two beams interact with each other, producing an interference signal. The OCT contains a receiver circuit that would convert the interference signal into an electrical signal which would then be displayed on the oscilloscope18. The interference pattern created by the combined beams is used to generate an A-scan, which is a one-dimensional scan that provides information about the depth of the sample at a single point. The specifications of the TC-OCT system are provided in Table 1.Table 1Specifications of the TD-OCT System3.SpecificationValueCenter wavelength1310 nmSpectral wavelength56 nmAxial resolution14.2 μmLateral resolution, spot size6 μmNumerical aperture0.14Scanning rate25 scans/sScanning depth in air12–14 mm Table 1Specifications of the TD-OCT System3.SpecificationValueCenter wavelength1310 nmSpectral wavelength56 nmAxial resolution14.2 μmLateral resolution, spot size6 μmNumerical aperture0.14Scanning rate25 scans/sScanning depth in air12–14 mm Specifications of the TD-OCT System3. Philippine bamboo cultivation parameters In this study, TD-OCT has been used to obtain cross-sectional images of Bambusa sp. leaves to illustrate microstructural differences in healthy and unhealthy part due to light stress. Water Bamboo cultivation thrives in soil with a pH range of moderately acidic to neutral, typically between 5.0 and 7.5. To support optimal growth, the bamboo plants were watered every other day, ensuring the seedlings received sufficient moisture for healthy development. The soil pH was monitored weekly using a pH meter to maintain the recommended range. The experimental setup utilized a shelf to house the bamboo seedlings, divided into three compartments. The top half of the shelf is further split into two sections, with each section containing high-intensity lights for irradiation. A central partition separates the two sections to prevent light from spilling between them. This arrangement allows for the simultaneous irradiation of two plants under controlled conditions. The bottom half of the shelf serves as a resting area for plants that require recovery time during or after irradiation. Each compartment was equipped with an ESP8266 Wi-Fi microcontroller connected to a DHT11 temperature and humidity sensor. The sensors transmit data to the Thingspeak platform for real-time monitoring of environmental conditions. The light intensities, namely Light A and Light B, in each compartment, respectively, was turned on separately for 75 min while monitoring the temperature and humidity levels of all compartments. In this experimental set-up, when Light A was turned on, compartments B and C showed minimal changes, with percentage differences of only 2.86% and 1.2%, respectively. Similarly, when Light B was activated, compartments A and C exhibited negligible increases of 2.94% and 1.25%. In contrast, the other compartments showed minimal changes in humidity, with differences remaining under 2%. The light irradiation process involved exposing bamboo plants to high-intensity light using 100W incandescent bulbs. A total light concentration of 4800 lx or 648 μmol m−2 s−1 were used for artificial sunlight. Each plant is subjected to one of three durations: 15 min, 45 min, and 75 min. The incandescent bulbs were selected for their ability to provide consistent and intense illumination, simulating conditions of light stress. Optical properties After the irradiation, TD-OCT was used to collect data. The OCT probe was fixed in a vertical position to direct the light downward onto the leaf surface. Each leaf was also carefully taped down to flatten and immobilize it during measurements. This technique minimized leaf curvature, ensuring accurate OCT readings. A micrometer was used to adjust the vertical stage the leaf was set on. Extinction coefficient measurement The raw interference intensity from the OCT’s A-scan was processed by background subtraction, focal length correction, and distance squared correction. After obtaining the A-scans from the OCT, the python program was implemented to execute the steps needed for the EC calculation. The selection of the smoothing factor σ was guided by the evaluation of two critical metrics: Signal-to-Noise Ratio (SNR) and Root Mean Square Error (RMSE). These metrics were used to assess the quality of the signal after applying smoothing. SNR measures the clarity of the signal relative to noise, while RMSE shows the degree of deviation from the original signal. Together, these metrics balance the trade-off between noise reduction and signal distortion, ensuring optimal signal quality. The SNR was calculated using Eq. 4.4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$SNR\left(dB\right)=20 {log}_{10}\left(\frac{{A}_{signal}}{{A}_{noise}}\right)$$\end{document}where Asignal is the difference between the highest signal peak and the average of the maximum noise peak and the lowest noise value, and Anoise is the spread of the noise values. SNR is expressed in decibels (dB), with values above 20–25 dB generally considered indicative of good signal quality. A higher SNR corresponds to a clearer and more distinguishable signal relative to the background noise.5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$RMSE = \sqrt{\sum_{i =1}^{n}\frac{{(y1-y2)}^{2}}{n}}$$\end{document} Equation 5 was used to calculate the RMSE. where y1 and y2 represent the original and smoothed signal values, respectively, and n is the total number of data points. RMSE is used to measure the average magnitude of error introduced by the smoothing process. While the RMSE values across all smoothing factors are relatively low, the smoothing factor, σ = 3, was selected to minimize the RMSE as much as possible. This choice ensures a strong balance between achieving a high SNR and preventing excessive distortion of the signal. The lens peaks of the two A-scans occur at slightly different times. This discrepancy, referred to as jitter19, is caused by vibrations generated by the motor-driven rotation of the retroreflector20. Jitter can lead to errors because it prevents the baselines of the sample and background A-scans from fully aligning. To correct this, one of the lens peaks was selected as the reference, designated at time ​t0, while the other peak occurs at time t. The difference, Δt, is determined using the formula Δt = t0 − t. The second trigger was then adjusted based on Δt to align t with t0. This adjustment ensures that the second peak’s position matches the reference peak. The first peak corresponds to the epidermal surface layer, while the second marks the boundary layer of the mesophyll between the palisade and spongy layers21. A best-fit line was calculated for the logarithmic signal at each peak. Thickness measurement The thickness calculations are computed as shown in Fig. 2. The raw OCT A-signals were converted to a logarithmic signal for analysis. In the presented OCT A-scan signal, only the internal backscattered logarithmic signals are shown, corresponding to the epidermal and mesophyll layers of the tissue. Our program, selects the positions of the surface layer and palisade layer. The first peak corresponds to the surface layer, while the second peak represents the boundary between the palisade and spongy layers. The distance between these peaks is used to determine the thickness of the palisade layer21. The OCT may not detect the palisade layer effectively since this is the region where light is primarily absorbed, rather than reflected. This is why the second peak is the boundary layer, as OCT relies on light reflection to create images. A threshold value was used to prevent peaks below an acceptable value. The outliers of peak positions in average A-scans were removed and interpolated. The moving average was taken to enhance the peak detection results and fits smoothly with the OCT image22.Fig. 2(A) Cross section of the Bamboo Leaf and (B) corresponding peak detection for the different layers. (A) Cross section of the Bamboo Leaf and (B) corresponding peak detection for the different layers. Vegetation ratio index measurement An Olympus microscope with a high-resolution camera was used to collect the images. To measure specific frequency distribution of the specific intensities in Eq. 7, a region of interest (ROI) was chosen. This is also designated as the region where OCT imaging was conducted. A spectral probe (Ocean Optics USB2000) was further used simultaneously to measure the intensity from the 535-nm and 575-nm profiles. The raw intensity profiles were subjected to background subtraction, and smoothening method using FFT analysis to obtain the corrected intensity signals. These corrected profiles were normalized using Eq. 6 below. The comparison of spectral reflectance among leaf samples, the measured intensity values were normalized using a min–max scaling approach. This method reduces the influence of variations in absolute signal intensity arising from instrumental and measurement conditions and emphasizes the relative differences between the green and yellow wavelength regions. A Vegetation Ratio (VR) index, shown in Eq. 7, will then be used to compare the green-yellow profiles among the samples23,24.6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${I}_{normalized} = \frac{{I-I}_{min}}{{I}_{max}-{I}_{min}}$$\end{document}7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$VR Index = \frac{{I}_{535}}{{I}_{575}}$$\end{document} Data processing and analysis There are three main parts measured using Python, which includes (a) extinction coefficient determination, measurement of palisade layer and VR index investigation. After obtaining the A-scan, a denoising technique using Gaussian filtering was employed guided with the evaluation metrics from SNR and RMSE. One-way ANOVA was performed to evaluate differences among days for each exposure time. Significant effects (p < 0.05) were followed by Tukey HSD post-hoc tests. Results and discussion Optical extinction coefficient of bamboo leaves After the bamboos were subjected to daily exposures of 15, 45, and 75 min for four days, with measurements taken at each exposure duration. The optical μ were measured using the developed TD-OCT system as discussed previously. Epidermal layer Figure 3 presents a box plot summary of the μ data across the four days. The plot shows a general decrease in the mean μ values over time for all exposure durations, with the 15-min exposure demonstrating the smallest reduction in μ when compared to the longer exposure times. The R2 values for the best fit line of the μ data across the days were 0.93, 0.97, and 0.83 for the 15-, 45- and 75-min exposure times respectively. This temporal trend suggests that the epidermal layer undergoes changes that affect its optical properties. For the 15-min exposure, the relatively stable mean μ values across days might indicate a less pronounced or slower response in the epidermal layer compared to the longer exposures. This could imply that shorter exposure times may not induce as significant a change in the optical characteristics of the epidermis, or the changes may occur more gradually. For the 45-min and 75-min exposures, the more pronounced decrease in μ could indicate more permanent alteration to the epidermal layer’s structure. One-way ANOVA showed a significant effect of day at 15 min (F ≈ 18.6, p < 0.001), 45 min (F ≈ 15.2, p < 0.001), and a weaker but significant effect at 75 min (F ≈ 4.9, p ≈ 0.01).Fig. 3Box plot summary of μ in the epidermal layer at varying exposure time. Box plot summary of μ in the epidermal layer at varying exposure time. Mesophyll boundary layer Figure 4 provides a box plot summary of the μ data for the mesophyll boundary layer. Overall, there is a noticeable decrease in μ over time for all exposure durations. The R2 values for the best fit line of the μ data across the days were 0.88, 0.63 and 0.45 for the 15, 45, and 75-min exposure times, respectively. The decrease in μ values for the mesophyll boundary layer suggests structural changes that affect its optical properties, similar to the epidermal layer A similar decreasing trend is observed for 15-min and 75-min exposure time, suggesting that even prolonged exposure results in changes to the mesophyll boundary layer. When comparing the mesophyll boundary layer to the epidermal layer, the mesophyll boundary layer generally exhibits a lower μ. Furthermore, the mesophyll boundary layer also shows a smaller change in μ, suggesting that the epidermis is more responsive to external conditions.Fig. 4Box plot summary of μ in the mesophyll boundary layer at varying exposure time. Box plot summary of μ in the mesophyll boundary layer at varying exposure time. One-way ANOVA indicated a significant effect of day on mesophyll extinction coefficient at 15 min (F ≈ 11.4, p ≈ 0.001) and 45 min (F ≈ 4.2, p ≈ 0.02), while no significant difference was observed at 75 min (F ≈ 1.9, p ≈ 0.15). When compared to the epidermal layer, the mesophyll boundary layer shows a less marked reduction in μ, further validating that the epidermis may be more sensitive to light exposure than the mesophyll boundary25. This difference could be due to the distinct functions of these layers. The epidermis serves as a protective barrier, which may undergo more significant changes under light stress, while the mesophyll boundary, as the site of photosynthesis, should be more resilient to light-induced damage. Optical thickness of palisade layer For all exposure times, the mean thickness generally decreased over the days. Figure 5 presents the changes in the thickness of the palisade layer across the three days for the different exposure durations. One-way ANOVA indicated no significant effect of day on palisade layer thickness at 15 min (F ≈ 1.6, p ≈ 0.20), 45 min (F ≈ 0.9, p ≈ 0.45), or 75 min exposure (F ≈ 1.3, p ≈ 0.28). The decrease in thickness is predominantly observed in 15-min and 75-min exposure times as compared to the 45-min exposure time. This is due to the decrease of moisture content as it is exposed at a longer time. The mean palisade layer thickness ranges from 0.02 to 0.06 mm which is common to plant samples undergoing environmental stresses21. Bamboo leaf thickness can vary significantly depending on the species, but general measurements indicate that it typically ranges from 0.15 mm to 0.5 mm26. Since bamboo seedlings are being used in the study, the leaves should fall on the thinner end of the spectrum, as it is still growing.Fig. 5Box plot summary of the palisade layer’s thickness. Box plot summary of the palisade layer’s thickness. Microscopic and VR index analyses Color changes are observed in all bamboo leaves at varying exposure time and sampling days (Fig. 6). Excessive exposure to sunlight, especially under high heat, can cause damage to plant tissues, leading to sunburn. The most common symptoms of plant sunburn include pale or white leaves, which may also become scorched, turning brown27. Most of these burns can be seen through the naked eye but have been better examined under the microscope. The discoloration in bamboo leaves can be analyzed through VR Index analyses23.Fig. 6Microscopic images of bamboo leaves showing (a) bloating; (b) discoloration, and (c) burnt surfaces. Microscopic images of bamboo leaves showing (a) bloating; (b) discoloration, and (c) burnt surfaces. The changes in the intensity values of the leaves (VR Index) can be linked to the plant’s stress and deteriorating health (Fig. 7). Plants change color because as they deteriorate, their chlorophyll breaks down and they can no longer absorb the light needed for photosynthesis. VR index served as proxy for NDVI, which was extracted from RGB micrographs to quantify temporal changes in chlorophyll-related reflectance. Although lacking a NIR band, this index reliably captured progressive tissue degradation and pigment loss28. The VR index showed positive correlation with the μ (R2 = 0.68, RMSE = 9, N = 30) after exposure. The decrease in green intensity is also notable. Normally, green light is reflected by the chlorophyll in healthy plants, and this decrease could indicate a reduction in the amount of chlorophyll3. Studies have shown that a decrease in the green reflectance of leaves correlates with reduced chlorophyll content and photosynthetic activity29.Fig. 7Correlation between the μ and the VR Index after 75 min exposure time. Correlation between the μ and the VR Index after 75 min exposure time. TD-OCT scanner for sustainable precision agriculture The integration of time-domain optical coherence tomography (TD-OCT) into plant science research, particularly for bamboo, represents a significant advancement in the tools available for precision agriculture. As demonstrated in this study, TD-OCT provides a non-invasive, high-resolution imaging modality capable of capturing in vivo structural changes in bamboo leaves under different light stress conditions. This technology offers an alternative method for early stress detection, physiological monitoring, and quantification of anatomical changes in plant tissues—key pillars of precision agriculture11,22. Unlike traditional destructive sampling techniques, TD-OCT allows for repeated measurements on the same plant, enabling longitudinal studies that are essential for understanding dynamic physiological processes. In this research, OCT was successfully used to measure both extinction coefficients (μ) and palisade layer thickness, parameters that are directly affected by light-induced stress (Figs. 2 and 3). The reduction in EC values across both the epidermal and mesophyll layers over time, especially under prolonged light exposure, provides quantifiable evidence of microstructural degradation. These findings align with typical stress responses involving chlorophyll degradation and cellular damage, reinforcing the value of EC as a potential marker for environmental stress3,11 . The ability to accurately detect and quantify these changes before visible symptoms manifest offers an invaluable tool for farmers and agricultural researchers alike. For instance, early detection of stress via OCT could allow for timely irrigation adjustments, shade provisioning, or nutrient supplementation—interventions that are central to sustainable farming practices. Additionally, the integration of OCT data with environmental sensors (e.g., DHT11 for temperature and humidity monitoring) and cloud-based platforms (e.g., Thingspeak) demonstrates the feasibility of deploying TD-OCT within an IoT-enabled precision agriculture system. The use of green intensity profiling further adds to the precision monitoring capabilities of this system. Changes in green and green-yellow reflectance due to pigment degradation—specifically, chlorophyll loss—were captured and quantified, offering an indirect but rapid proxy for plant vitality23,24. This approach can be integrated into imaging platforms, complementing OCT readings and providing a more holistic view of plant health. The adaptation of a rotational reference mirror in the OCT setup enhanced the scanning range and system repeatability, making it more suitable for agricultural applications where uniformity and coverage are vital. This level of engineering innovation ensures that TD-OCT systems can be optimized for field deployment, even in rugged or non-laboratory environments. Conclusions This study demonstrates the viability and effectiveness of time-domain optical coherence tomography (TD-OCT) as a non-invasive, high-resolution imaging modality for monitoring the physiological responses of Bambusa sp. leaves to light-induced stress. By combining TD-OCT with VR Index and environmental sensor integration, we captured quantifiable changes in leaf microstructure, particularly in extinction coefficients and palisade layer thickness, across varying light exposure durations. The results show a consistent decline in optical parameters and chlorophyll-related indices under prolonged light stress, correlating with observable visual symptoms and pigment degradation. These findings not only validate the sensitivity of TD-OCT in detecting early microstructural changes but also underscore its potential in longitudinal plant health monitoring. The integration of TD-OCT data with IoT-based environmental monitoring platforms offers a powerful, real-time diagnostic framework for precision agriculture. In tropical agricultural contexts such as the Philippines, this approach holds significant promise for enhancing sustainable bamboo cultivation and optimizing resource use through early stress detection and adaptive management. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1