P-518 - Phenomics-based approaches to study Dutch Elm Disease resistance in American elm: linking spectroscopy, metabolite profiling, and gas exchange

American elm (Ulmus americana) breeding for Dutch elm disease (DED) resistance has been ongoing for decades but remains costly and time-intensive, requiring extensive screening of resistant clones. Recent advances in phenomics, particularly reflectance spectroscopy, offer rapid, cost-effective methods to assess disease resistance by monitoring leaf chemical composition. In this study, we investigated whether spectroscopy combined with untargeted metabolomics could identify early signatures distinguishing DED-resistant and susceptible cultivars. Clonally propagated one- and two-year-old elms with varying resistance were inoculated with Ophiostoma novo-ulmi, and distal leaves were collected at 0 and 96 hours post-inoculation for spectral and metabolomic analysis. To further assess vascular responses, gas exchange measurements were taken at 20 days post-inoculation, while canopy decline was monitored until 100 days post-inoculation, with vascular staining conducted at harvest. Machine learning models classified presymptomatic trees based on spectral and metabolite profiles, though cultivar-specific responses emerged. Gas exchange measurements and safranin staining revealed that resistant cultivars maintained vascular function despite infection, though variability among clones suggests additional resistance mechanisms. These findings highlight the potential of spectroscopy and metabolomics as high-throughput tools for early DED resistance screening, ultimately accelerating breeding programs and aiding the restoration of American elms in urban and forest ecosystems.