Terpenes associated with resistance against the gall wasp, Leptocybe invasa, in Eucalyptus grandis

Naidoo, Sanushka; Christie, Nanette; Acosta, Juan J.; Mphahlele, Makobatjatji M.; Payn, Kitt G.; Myburg, Alexander Andrew; Külheim, Carsten

Terpenes associated with resistance against the gall wasp, Leptocybe invasa, in Eucalyptus grandis

Naidoo, Sanushka; Christie, Nanette; Acosta, Juan J.; Mphahlele, Makobatjatji M.; Payn, Kitt G.; Myburg, Alexander Andrew; Külheim, Carsten

URI: http://hdl.handle.net/2263/66066

Date: 2018-08

Abstract:

Leptocybe invasa is an insect pest causing gall formation on oviposited shoot tips and leaves of Eucalyptus trees leading to leaf deformation, stunting, and death in severe cases. We previously observed different constitutive and induced terpenes, plant specialized metabolites that may act as attractants or repellents to insects, in a resistant and susceptible clone of Eucalyptus challenged with L. invasa. We tested the hypothesis that specific terpenes are associated with pest resistance in a Eucalyptus grandis half‐sib population. Insect damage was scored over 2 infestation cycles, and leaves were harvested for near‐infrared reflectance (NIR) and terpene measurements. We used Bayesian model averaging for terpene selection and obtained partial least squares NIR models to predict terpene content and L. invasa infestation damage. In our optimal model, 29% of the phenotypic variation could be explained by 7 terpenes, and the monoterpene combination, limonene, α‐terpineol, and 1,8‐cineole, could be predicted with an NIR prediction ability of .67. Bayesian model averaging supported α‐pinene, γ‐terpinene, and iso‐pinocarveol as important for predicting L. invasa infestation. Susceptibility was associated with increased γ‐terpinene and α‐pinene, which may act as a pest attractant, whereas reduced susceptibility was associated with iso‐pinocarveol, which may act to recruit parasitoids or have direct toxic effects.

Description:

Supplementary Table 1. Environmental and phenotype data (full population) for the three Eucalyptus grandis sites surveyed for Leptocybe invasa infestation. [Excel file]

Supplementary Table 2. Predictor variable datasets and outlier detection prior to partial least squares modeling. (A) Predictor variable datasets used for outlier detection and partial least squares modeling. (B) The proportion of samples classified as outliers (and thus trimmed) for each set of models, prior to partial least squares modeling. [Excel file]

Supplementary Table 3. The 48 measured terpenes. (A) The name, major ions and retention time of the 48 measured terpenes. (B) The motivation for combining groups of terpenes. (C) The correlation of terpenes with the Leptocybe invasa screenings (LS1, LS2) and individual breeding values (IBV) for the Siya Qubeka (SQF) site. [Excel fie]

Supplementary Table 4. Leptocybe invasa heritability estimates for L. invasa screening 1 (LS1) and L. invasa screening 2 (LS2) for the Eucalyptus grandis population across sites. [Excel file]

Supplementary Table 5. Summary of the best partial least squares models, based on near-infrared reflectance (NIR) data, for Leptocybe invasa screenings (LS1, LS2) and individual breeding values (IBV) at the Mtunzini (MTZ) and Nyalazi (NYL) sites. [Excel file] Supplementary Table 6. Bayesian model selection results to identify the most important terpenes for predicting Leptocybe invasa infestation. (A) Bayesian model selection results at the Siya Qubeka (SQF) site. (B) Bayesian model selection results at the Mtunzini (MTZ) site. (C) Bayesian model selection results at the Nyalazi (NYL) site. (D) Bayesian model selection results across all three sites. [Excel file]

Show full item record