Modelling water use efficiency (WUE) for estimating the severity of Ganoderma boninense-derived basal stem rot disease in oil palm

• Published in: Journal of Plant Pathology
• Main Author: Baharim M.S.A.; Adnan N.A.; Izzuddin M.A.; Laurence A.L.; Karsimen M.K.; Arof H.
• Format: Article
• Language: English
• Published: Springer Science and Business Media Deutschland GmbH 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85206386758&doi=10.1007%2fs42161-024-01770-5&partnerID=40&md5=efffa4c04191b4a3668b87435ee091d7

Basal stem rot (BSR) disease, caused by the Ganoderma boninense pathogen, is a significant threat in oil palm-producing nations, particularly in Malaysia and Indonesia. The disease proliferates extensively within oil palm plantations and is anticipated to persist for an extended duration. Although various assessment methods have been deployed at different stages of BSR infection, none of them seem to be effective. Therefore, this research proposes a predictive modeling approach to evaluate plant stress conditions induced by BSR disease. This approach incorporates variables such as water use efficiency (WUE) alongside other leaf physiology parameters and hyperspectral imagery subjected to various digital image processing transformations to increase assessment accuracy. The results indicate that images denoised with CR transformation offer superior performance compared with alternative methods. Furthermore, five significant wavelengths (w717.4, w736.7, w741, w790.8, w860.7) were identified, which were strongly correlated with WUE and BSR disease severity through pairwise comparisons. Both Model 3 and Model 4 produced acceptable and relatively high accuracy results for WUE prediction. Ultimately, the study advocates for the adoption of Model 4 (integrating leaf physiology variables and hyperspectral data: Ci, Pr, gs, with w741) for WUE prediction in oil palm, as it shows lower training and validation model RMSE values (0.44, 0.37) with the highest regression plot values (R2: 0.92, 0.98), thereby offering a more robust analytical perspective than Models 1, 2, and 3 (individual leaf physiology variables: Ci, Pr and gs). © The Author(s) under exclusive licence to Società Italiana di Patologia Vegetale (S.I.Pa.V.) 2024.