Identification of spectral ranges that contribute to phytoplasma detection in apple trees - A step towards an on-site method

Abstract

'Candidatus Phytoplasma mali' is the bacterial agent associated with Apple Proliferation, a disease that causes high economic losses in affected commercial apple growing regions. The identification of the disease is carried out by visual inspection performed by skilled professionals in the orchards. To confirm an infection, costly molecular laboratory methods must be applied. Furthermore, both methods are very time-consuming. Here, we analysed the potential of a non-destructive method using in-field measurements to differentiate infected from non-infected apple trees (Malus domestica) based on spectral signatures of fresh leaves. By using multivariate statistics, we were able to distinguish infected from non-infected trees and identified the wavelengths relevant for the differentiation. Factors affecting the differentiation performance were the sampling date and bacterial colonization behaviour.

Keywords: Apple proliferation; Hyperspectral sensing; Non-destructive method; Phytoplasma; Plant disease detection; Spectroscopy.