植被高解析反射光譜利用於評估水稻白葉枯病危害等級可行性

Potential of Using High-Resolution Canopy Reflectance Spectrum to Assess Disease Severity of Rice Bacterial Leaf Blight

楊純明(Chwen-Ming Yang)

《科儀新知》 164期 (2008/06) Pp. 89-96

https://doi.org/10.29662/IT.200806.0011

摘要

白葉枯病係水稻一重要維管束型病害，嚴重感染將可能導致明顯減產。本研究以田間試驗之人為感染方式產生不同感染等級，再量測兩參試稻種各等級近地面植被高解析反射光譜，分析、篩檢有關之光譜特徵，進而建立光譜特徵模式（組）於未來精準管理稻株病害用途。秈稻臺中秈10號係高感性，稉稻臺農67號則係中感性。結果顯示，感病徵狀與光譜特徵具高相關，因此所建立之光譜特徵模式（組）能夠適當地評估白葉枯病感染等級。

並列摘要

The bacterial leaf blight (BLB) is an important vascular disease to irrigated rice and may cause substantial yield loss under serious conditions. This study analyzed near-ground canopy hyperspectral reflectance of two rice cultivars (Oryza sativa L.), TNG 67 and TCS 10, infested with various levels of the pathogen so as to establish spectral models for assessing disease severity for future applications in site-specific management. Cultivar TCS 10 is an indica type of high susceptibility and cultivar TNG 67 is a japonica type with moderate susceptibility. Field experiment was conducted to produce various levels of BLB infestations for canopy reflectance measurements and spectral characteristics (SCs) associated with disease symptoms were identified for models establishment. Results suggest that the approaches used for hyperspectral analysis in this study are less complex and are conceptually easy to understand. Correlation of disease symptoms and SCs is high in BLB-infested plants so that disease severity may be properly assessed.