Early hyperspectral detection of Carlavirus vignae in common bean under field conditions

Abstract

Abstract Virus-associated diseases are among the biological stresses that affect common bean yield, such as Carlavirus vignae ( Cowpea mild mottle virus , CPMMV), transmitted by the whitefly Bemisia tabaci . CPMMV is present in different continents, and recent outbreaks have concerned Brazilian farmers and researchers. Integrated Pest Management routines for field monitoring of viral spread are laborious and may be limited to visible symptoms. We hypothesized that CPMMV infection can be detected in asymptomatic plants by differences in the plant canopy reflectance. To test this hypothesis, we used a hyperspectral sensor mounted on a drone to capture images of CPMMV-inoculated and non-inoculated field plots in 2022 and 2023, across a tolerant common bean cultivar (BRS FC420 RMD) and a susceptible one (BRS FC401 RMD). Results showed that CPMMV was detected in common bean plants by hyperspectral imaging at early infection stages (~ 6 DAI) before symptom onset and at an advanced infection stage (~ 22 DAI). Reflectance within the visible light spectrum was affected by soil cover on all flights, and in most of these, also in the near-infrared region. The main differences between CPMMV-inoculated and control plants were consistent across two years of experiments, regardless of the common bean phenological stage and genotype. Fit statistics using the sum of squared errors, R 2 and AIC indicated that reflectance from 401 to 425 nm, especially near 415 nm, differed significantly between infected and healthy plants. Such changes are associated with chlorophyll degradation and disruption of the photosynthetic apparatus, and are detectable even before symptom onset. Progress in the disease severity index also differentiated the tolerant cultivar from the susceptible one. CPMMV infection significantly reduced common bean yield by ~ 21% compared with healthy plants. CPMMV detection by hyperspectral imaging enables early scouting to optimize disease management.

Plant phenotyping relevance

ハイパースペクトル画像を用いて、症状発現前の感染植物の反射特性と病害状態を検出し、複数年・品種で技術性能を検証しているため、植物フェノタイピング手法が中心である。

Code and data availability

The paper's hyperspectral imagery, reflectance data, and analysis are not publicly deposited; the authors state data are available only upon reasonable request. No public code, dataset, or model repository is provided.

No evidence-backed public reproduction asset is currently recorded.