Abstract:
Eucalyptus trees are an important source of timber in South Africa. Unfortunately, the trees are susceptible to a number of pests and pathogens. Chrysoporthe austroafricana is a significant pathogen of Eucalyptus trees in South Africa. The pathogen causes stem cankers which lead to wilting and eventually death. Clonal and hybridization trials have been carried out to improve the genotype of these trees. In order to breed for pest and pathogen resistance, this procedure necessitates the screening of clones and hybrids. Current screening methods rely on inoculation trials and natural infection, both of which are time consuming and, in the case of inoculation trials, destructive. Here I show that resistance phenotyping can be conducted, rapidly and non-destructively, by way of infrared (IR) spectroscopy, which generates chemical fingerprints that can be used to predict and identify resistant and susceptible trees when combined with chemometric analyses. Specifically, I used Fourier transform IR (FT-IR) spectroscopy in combination with machine learning algorithms to distinguish between resistant and susceptible Eucalyptus hybrid clones against C. austroafricana. The results from the study is a proof of concept on the potential of FT-IR as a tool to screen Eucalyptus clones for resistance to C. austroafricana.
