[U-04] Geospatial Applications for Societal Benefits

[U04-03] Ganoderma Diseases Detection on Oil Palm using Optical Remote Sensing

*Muhammad Arif Yusuf¹, Yukihiro Takahashi¹ (1.Hokkaido University)

Keywords:Oil Palm, Ganoderma, Reflectance, Classification

Basal stem rot (BSR) disease, which is caused by Ganoderma sp. infection, has become a serious problem in oil palm (Elaeis guineensis Jacq.) plantation in Indonesia and Malaysia (Santoso et al., 2013). A major concern regarding BSR is its detrimental effect on oil palm plantation density. Research indicates that in plantations with severe infections, tree density can decline to less than 50% of the initial population (Hushiarian, Yusof, & Dutse, 2013; Susanto et al., 2013; Priwiratama, Prasetyo, & Susanto, 2014). This reduction has profound implications for yield optimization and the long-term sustainability of oil palm cultivation. Numerous studies (Assis et al., 2015; Hushiarian et al., 2013; Ling-Chie et al., 2012; Priwiratama & Susanto, 2014; Susanto et al., 2013) have emphasized the absence of fully effective curative treatments for BSR. Existing treatment approaches are primarily aimed at prolonging the lifespan of infected oil palms rather than eradicating the disease. Consequently, early detection is a critical strategy for controlling its spread. As highlighted by Hushiarian, Yusof, and Dutse (2013), as well as Priwiratama, Prasetyo, and Susanto (2014), the timely identification of infections enables the implementation of management practices that extend the productive period of affected oil palms and mitigate economic losses. Remote sensing has emerged as a highly promising technology for large-scale disease monitoring in oil palm plantations. Through the analysis of spectral data from oil palm canopies, this technique enables the early detection of stress indicators in trees before the manifestation of visible symptoms. This capability positions remote sensing as an optimal tool for the comprehensive monitoring of BSR-infected trees across extensive plantation areas. A significant limitation of existing research is its reliance on single-time spectral measurements, which capture disease characteristics at a specific moment without tracking its progression over time (Azmi et al., 2020; Santoso et al., 2018; Santoso, 2020; Santoso, 2023; Kurihara, 2022; Wahyuni et al., 2023). This approach did not provide insights the temporal dynamics of infection, which are critical for effective early detection. Continuous monitoring time by time is essential to understanding the disease’s development and implementing timely interventions. Another drawback of previous studies is the lack of investigation into the differential manifestation of disease symptoms between older and younger leaves. Since symptom intensity may vary depending on leaf maturity, analyzing these variations could enhance detection accuracy and improve diagnostic precision. To address these limitations, this research will introduce a time-series spectral analysis approach, enabling continuous monitoring of disease progression. Additionally, differential leaf measurements will be incorporated to assess how Ganoderma symptoms develop across various leaf ages. By bridging these gaps, this study aims to establish a more comprehensive and dynamic framework for early disease detection in oil palm plantations. This study was conducted in an oil palm pre-nursery setting to evaluate spectral responses under different health conditions. Two treatment groups were established: one comprising oil palm seedlings inoculated with Ganoderma since the seed stage, and the other consisting of healthy (non-infected) plants. Each treatment group included five individual plants. Spectral measurements were performed on every leaf of each plant to assess variations in spectral characteristics associated with disease progression. The results indicate that during the early stages of infection, leaves from Ganoderma-infected plants (K+) exhibit higher reflectance in the green spectral region, suggesting a greener appearance compared to healthy leaves (K-). Additionally, in younger leaves, K+ plants demonstrate increased reflectance in the yellow spectral region, indicating a more pronounced yellowish hue relative to K- leaves.