Identification Of Barley Grain Mycoflora By Next Generation Sequencing And Videometer Multispectral Imaging

Seeds of Barley (Hordeum vulgare) are infected by a high number of fungi, including pathogens such as Fusarium graminearum, F. culmorum, F. poae, F. avenaceum and Pyrenophora teres. Fusarium spp. is a widely distributed fungus causing yield reduction in a range of agricultural crops and many species in the genus produce mycotoxins responsible for serious quality deterioration.

In malting barley, Fusarium also has a negative effect by causing gushing in beer. A number of barley seeds (app. 200) assumed to be infected by fungal from different origins and years of cultivation were tested by NGS sequencing the ITS (Internal Transcribed Spacer) region from total DNA. Approximately 2-4000 sequences were obtained from each seed and these were subsequently identified to species level in order to give an exact identification of fungal genera on each seed.

The main fungal genera identified were Fusarium, Pyrenophora, Epicoccum, Didymella, Alternaria, Bipolaris and Microdochium. The fungal composition and quantities on each seed varied significantly. Some were infected mainly by a single fungus and some were infected by multiple fungi. All seeds were prior to this evaluated by multispectral imaging on the dorsal and ventral sides by the VideometerLab multispectral imaging system (Videometer A/S, Hørsholm, Denmark).

This system is an instrument equipped with 19 different light emitting diodes at wavelengths ranging from 375 to 970nm (ultraviolet, visual and lower wavelength of the near-infrared region) in the reflectance mode (5 Mpix per band, pixel size app. 45 μm x 45 μm). Spectral information over the surface of seeds may be combined with information about size, shape, and texture of the seeds.

This information links detection of fungal infection with other seed characteristics known from general seed testing. Analytical separation of the identified fungi was based on mean pixel intensity and a normalized Canonical Discriminant Analysis (nCDA) using the images of infected and healthy seeds.

The potential of using spectral characteristics of the fungal species as a way to provide a fast optical screening method for fungal contamination of barley on the fungal species level was investigated by comparing results from the next generation sequencing and multispectral imaging.