Background: Perennial ryegrass (Lolium perenne) is systemically infected by seed-transmitted fungal endophytes (Epichloë sp.). The presence of Epichloë endophytes alters the nutritive quality of its hosts by modifying several plant traits. The aim of this research was to develop a fast method based on near-infrared reflectance spectroscopy (NIRS) to discriminate between perennial ryegrass plants infected (E+) or not infected (E-) with two endophyte species, Epichloë festucae var. lolii, and Epichloë typhina, using a heterogonous set of perennial ryegrass samples collected from wild grasslands and cultivars. Epichloë festucae var. lolii cultures show two morphotypes, M1 and M3, and Epichloë typhina cultures have a different M2 morphotype.
Results: Near-infrared reflectance spectra from E+ and E- ryegrass plants were recorded. Applying the best NIRS model for the detection of Epichloë, 93.3% of E+ plants were classified correctly. The NIRS morphotype classification was correct for 92.9% of M1 morphotype and 100% of M2 morphotypes. The NIRS classification of M3 morphotypes was not as accurate, but it was in accordance with the fungal species classification, identifying some M3 as M1 morphotypes.
Conclusion: Near-infrared reflectance spectroscopy can detect the presence of Epichloë fungal endophytes directly in samples of perennial ryegrass, and it is adequate for discriminating among fungal species. © 2018 Society of Chemical Industry.
Keywords: Epichloë endophytes; NIRS; chemometrics; endophyte classification; fungal identification; perennial ryegrass.
© 2018 Society of Chemical Industry.