Phenotyping of <em>Cercospora beticola</em> resistance of sugar beet genotypes by hyperspectral imaging

Abstract

Cultivation of disease resistant crops is an important strategy in the integrated pest management which is a guiding principle for good agricultural practice. Therefore, high yielding cultivars with resistance to important plant diseases are needed. The integration of resistance sources such as wild relatives or compatible subspecies might help to enhance the resistance of crops and thus to reduce the need for chemical crop protection measures. For the selection of plants with a specific trait, such as resistance to a plant pathogen, precise determination of the genotype and a reliable characterization of the phenotype are necessary. The rather rapid development of molecular methods and knowledge about genes enhanced the genotyping in plant breeding greatly. The phenotyping, however, is still the bottleneck in resistance breeding. As the phenotype is the result of the interaction of the genotype and the environment phenotyping must be reliable, reproducible and non-invasive. The implementation of sensors in phenotyping systems provides many advantages. Hyperspectral imaging sensors are well suited to characterize different plant traits. <br /><em>Cercospora</em> leaf spot (CLS) is the most important foliar disease of sugar beets and is mainly controlled by fungicide applications. The aim of this study was to characterize the resistance of sugar beet genotypes against <em>Cercospora beticola </em>and the development of a hyperspectral imaging system for phenotyping this disease resistance.<br /> A hyperspectral microscope that measures reflection in the visible and near-infrared range from 400 to 1000 nm with a magnification of up to 7.3x was established to determine spectral changes on the plant tissue level. Disease development on five genotypes infected with CLS was evaluated and compared under controlled conditions. Two additional genotypes were used to validate the results of the hyperspectral measurement of CLS dynamics.<br /> Resistant genotypes had a lower percentage of diseased leaf area, a reduced number of lesions, lesion size and growth rate and a decreased spore production. Apart from the quantitative difference between highly susceptible and more resistant sugar beets, the lesion phenotype varied in size and spatial composition depending on the host genotype. Using the hyperspectral microscope, lesions could be differentiated into subareas based on their spectral characteristics. Sugar beet genotypes with lower disease severity typically had lesions with smaller centers and produced fewer spores in comparison to highly susceptible genotypes. The differences in number of spores per lesion were closely associated to the spectral difference calculated as area between spectral signatures before and after sporulation. The CLS development, analyzed by hyperspectral imaging over ten days, differed depending on the host genotype and the resistance source. For example, lesion development on a resistant genotype carrying two quantitative trait loci (QTL) was characterized by a fast and abrupt change in spectral reflectance, whereas it was slower and ultimately more severe on the closely related genotype lacking the QTL. The analysis of reflectance and transmittance images by calculating spectral vegetation indices and extracting spectral signatures revealed a potential benefit of transmission measurements. Depending on the topic and analysis method, effects were sometimes stronger pronounced in the transmittance data. <br /> The resistance against <em>C. beticola</em> was not complete, instead, it can be described as quantitative and rate-reducing. Some resistance parameters such as a decreased sporulation matter particularly with regards to disease epidemics in the field. Based on the hyperspectral images, a detailed analysis of the lesions was possible. The presented method allowed a reliable differentiation of CLS dynamics and the characterization of even subtle differences in resistance. Hyperspectral imaging is a promising tool with the potential to improve the screening process in breeding for CLS resistance.