Genetic variation in Eucalyptus camaldulensis and E-tereticornis for early growth and susceptibility to the gall wasp Leptocybe invasa in China

The gall wasp Leptocybe invasa is a major insect pest on plantation Eucalyptus in many countries. Since appearing in China in 2007 it has had major impacts on commercial plantations-some commonly planted Eucalyptus varieties have proven particularly susceptible, including hybrids involving Eucalyptus camaldulensis and E. tereticornis parent species. Intra-specific variation within each species for L. invasa susceptibility was examined in two seed source and family trials in southwest of Guangdong. The E. camaldulensis trial included 101 seedlots representing five natural stand and six seed orchard sources which also represented three sub-specific taxa. The E. tereticornis trial included 143 seedlots representing 11 natural stand and four seed orchard sources, including breeding seed orchards in China. Both trials were assessed for susceptibility to L. invasa along with height at age 9 months. Sub-specific taxa within E. camaldulensis differed significantly (P < 0.01) in L. invasa susceptibility but not height or survival; subsp. acuta had the lowest average susceptibility and subsp. simulata and obtusa were of intermediate susceptibility whilst material of uncertain sub-specific taxa from India had the highest average susceptibility. In E. tereticornis regions of origin and seed sources within regions differed significantly (P < 0.01) for both L. invasa susceptibility and height; the region China (all somewhat improved sources) had both the best average height growth and lowest susceptibility whilst the region Australia (all natural stand sources) proved inferior to China for both average susceptibility and height. A strong significant correlation was found between seed source average L. invasa susceptibility and annual rainfall at seed source geographic origin in E. tereticornis (r = -0.873; P < 0.01), implying that seedlots from higher rainfall environments are markedly less susceptible to L. invasa. The equivalent parameter in E. camaldulensis was also strong (r = -0.730) though not significant, perhaps due to having only five data points available. Differences between families within seed sources for both L. invasa susceptibility and height growth were also highly significant (P < 0.01) with the former trait proving moderately to strongly heritable (h (i) (2) = 0.54 +/- A 0.40 in E. camaldulensis; 0.52 +/- A 0.50 in E. tereticornis). Height had low to moderate heritability in both species (h (i) (2) = 0.11 +/- A 0.15 and 0.26 +/- A 0.08 respectively). Tree height and L. invasa susceptibility showed a moderate negative (favourable) genetic correlation in E. camaldulensis (-0.33 +/- A 0.64) and a moderate to strong negative (favourable) genetic correlation in E. tereticornis (-0.47 +/- A 0.31). Corresponding phenotypic correlations, though significant, were somewhat weaker (-0.16 and -0.29 respectively), indicating a trend for taller trees to have lower levels of L. invasa susceptibility.