Foliar Infrared Spectra Track Nematode Density and Symptom-Specific Phytobiome Signatures in Beech Leaf Disease

Abstract

Beech leaf disease (BLD), caused by the nematode Litylenchus crenatae ssp. mccannii (LC), poses a severe threat to American beech (Fagus grandifolia) across eastern North America. The disease causes wholesale anatomical, morphological, and physiological alterations, including symptoms such as leaf banding and bud abortion; the latter eventually leads to beech mortality, especially of younger trees. This outcome severely affects regeneration and diminishes the important ecosystem services provided by this keystone species. To advance our understanding of the disease, we hypothesized that significant links exist among LC abundance, foliar phytobiome dysbiosis, and phytochemical alterations detectable via NIR reflectance spectroscopy. To test this hypothesis, we applied molecular diagnostics to quantify LC in the tissues, bacterial and fungal foliar microbial community profiling, and NIR reflectance spectroscopy to characterize three distinct tissue phenotypes in BLD-infected trees: (1) asymptomatic tissues of asymptomatic leaves (AA); (2) asymptomatic tissues of symptomatic leaves (AS); and (3) symptomatic (galled) tissues of symptomatic leaves (GS). Overall, the three tissue types differed significantly in LC load (AA < AS < GS). Furthermore, NIR spectral profiles differed consistently among tissue types, with distinct wavelength regions associated with water and structural chemistry driving the separation. Machine learning and multivariate models of NIR spectra predicted both LC abundance and bacterial community composition by tissue type, enabling possible discrimination of dysbiotic foliar phytobiomes, with moderately high accuracy, but not so for fungal community composition. Bacterial taxa such as Pseudomonas, Wolbachia, Luteibacter, and Pedobacter were significantly associated with LC infection. Taken together, these results validate our hypothesis. This study establishes NIR technology as a platform for LC quantification and, more broadly, as a tool for assessing bacterial dysbiosis in plant systems.