Detecting Brown Planthopper, Nilaparvata lugens (Stål) Damage in Rice Using Hyperspectral Remote Sensing

Abstract

Remote sensing is being increasingly used in agricultural stress management, offering real-time analysis of crop stress beyond what can be observed visually alone. Hopper burn, induced by brown planthopper (BPH, Nilaparvata lugens) is a significant challenge to rice production in India. This study aims to characterize the spectral reflectance of rice plants with different infestation levels of 2^nd instar BPH nymphs across three varieties namely Pusa Basamti-1509, Pusa Basmati-1121, and TN-1 (susceptible variety to BPH). The differential infestation included 0, 5, 10, 20, 40, 80, 100, 200 number of nymphs. The spectral signatures of the crop were collected at 20 and 40 days after infestation (DAI). The results indicated a distinct change in reflectance pattern between healthy and infested rice plants across all three varieties in green (490–559 nm), yellow (560–584 nm), orange (585–639 nm), and red (640–699 nm) region of visible portion as affected by chlorophyll pigments and in NIR region (700–1800 nm) as affected by cell structure and in water absorption (1915 nm) region of SWIR portion. The specific wavelength band (470, 660,750, 1800, 1915 nm) showing a correlation above 0.8 with pest severity level, was identified as sensitive for assessing BPH damage. Analysis of reflectance changes across wavelengths highlighted that the first derivative has a strong correlation between BPH severity and reflectance in the green region (500–540 nm) and red edge position (680–760 nm). The amplitude of the red edge value decreased with an increase in the severity level of insects. This study identifies the sensitivity of visible and near-infrared regions in detecting and assessing BPH infestation severity.