Hyperspectral signatures and betalain indicator for beet mosaic virus infection in sugar beet

Remote sensing technologies provide the potential to support the breeding process of crop cultivars. The presented work suggests a new phenotyping routine for sugar beet genotypes, resistant or tolerant to beet mosaic virus (BtMV). The use of non-invasive sensors could potentially eliminate the need for time-consuming and expensive laboratory analysis during resistance breeding for BtMV. Sugar beet plants were inoculated with two variants of BtMV (i.e., wild type and genetically modified) and grown in a climate chamber. The genetically modified virus induces betalain expression in sugar beet leaves and petioles, thus, a red pigmentation appears during disease progression. It was assessed a) whether this coloration could be harnessed to improve the discrimination between diseased and healthy plants; and b) whether it is useful to quantify virus content. Thus, plants were measured with a non-imaging field spectrometer during disease development to assess their spectral signatures, which were analyzed using machine learning models. We demonstrate that the spectra of plants, inoculated with the genetically modified virus, therefore expressing betalain, had a higher discriminative power compared to the spectra of those inoculated with the wild-type virus. However, the expected superior performance of combining the genetically modified virus and a non-imaging sensor could not be proven. Further studies should investigate the use of imaging sensors for improved performance. Furthermore, a potential offset of the betalain expression must be validated for consistency across individual infections.