Eucalypts are used globally on a larger scale for pulpwood than for sawn timber. They are known for high growth stresses which have limited its value and productivity as sawn timber. Previous studies have addressed the growth stress problem in some species but not at a large scale, as they are time consuming and costly to measure. The demand for durable wood is increasing, both in domestic and international markets, and there is a need to grow durable wood for heavy structural applications such as post and poles. Some eucalypts produce naturally durable heartwood.

This thesis focused on the assessment of early screening of three eucalyptus species for properties such as growth, growth strain, checking/collapse, heartwood diameter, extractive content, volumetric shrinkage, acoustic velocity and stiffness. Chapter 1 gives an overview of growth strain, different methods of assessing durability, collapse in wood, acoustic velocity assessment methods, and genetic parameters in a breeding programme.

Chapter 2 investigates the accuracy of growth strain assessments using the ‘splitting’ and ‘quartering’ test method as well as other characteristic properties (acoustic velocity, diameter, air- dry density, dynamic MoE and volumetric shrinkage) of 22 families of E. bosistoana at age 2 years old.

Positive phenotypic and genetic correlations (rp = 0.78, rg = 0.96) were found between the ‘quartering’ and the ‘splitting’ test. In light of the high genetic correlation, there is no need to consider the more time consuming ‘quartering’ test for assessing growth strain as the genetic change was insignificant.

Relevant properties revealed promising genetic control (h2 = 0.63 for diameter, h2 = 0.16 to 0.33 for growth strain, and h2 = 0.83 for volumetric shrinkage). Significant variability was also observed for diameter (CGV = 23%), growth strain (CGV = 15 to 23%) and volumetric shrinkage (CGV = 18%) indicating that wood quality improvement through genetic selection is feasible. There was a non-significant favourable negative correlation between dynamic MoE and growth strain (rg = −0.22), and a significant favourable correlation between the volumetric shrinkage and growth strain (rg = 0.34) was reported.

Chapter 3 describes the genetic parameters of wood properties in 83 families of 2-year-old E. quadrangulata with the following traits measured: diameter, growth strain, acoustic velocity, dynamic MoE, air-dry density and volumetric shrinkage. The growth strain was assessed in the stem using the ‘splitting’ test method in the green state by sawing along the length via the pith to measure the distortion and the large end diameter. The growth strain varied from 458 to 4742 µɛ with average strain of 1784 µɛ and a coefficient of phenotypic variation (CPV) of 26%, while volumetric shrinkage ranged from 17.1% to 36.1% with an average of 19.0% and a CPV of 21%. The traits revealed narrow sense heritability estimates that varied from 0.20 to 0.92 with substantial genetic gain. Phenotypic and genetic variability ranging from 5% to 26% and from 4% to 20% was observed for all the wood properties, respectively.

Wood properties, namely diameter, growth strain, checking, acoustic velocity, dynamic MoE, volumetric shrinkage and air-dry density of 2-, 7- and 8-year-old E. globoidea grown from seed or as coppice at two sites in New Zealand were assessed in Chapter 4. Noticeable levels of growth strain (means 2406 µɛ to 3084 µɛ) were present. Therefore, growth strain should be considered in a breeding programme for this species.

Growth strain, volumetric shrinkage and air-dry density were higher in the coppice than at the top of seed-grown trees. As the effects of coppice and stem height were confounded in this study further work is needed to confirm if trees from coppice suffer from higher growth strain.

Checking was observed in discs, indicating challenges for drying E. globoidea timber. Heartwood was more prone to checking (4.01%) than sapwood (2.14%). Checking was significant and negatively correlated with growth strain (rp = −0.19) but showed significant positive correlation with growth (rp = 0.22) suggesting that bigger trees are more likely to have a checking problem.

Substantial variability was revealed for growth strain (CPV = 19% to 24%) and checking (CPV = 83% to 102%), the traits most likely causing wood quality issues for E. globoidea, suggesting possibilities of improvement in a breeding programme if the trait is heritable.

Chapter 5 describes genetic variation in wood properties of mid-rotation age of 141 families of E. globoidea by assessing the following traits: heartwood diameter, core length, combined sapwood diameter, heartwood collapse, sapwood collapse, standing tree acoustic velocity and extractive content in the heartwood. Heartwood diameter ranged from 0 to 190 mm with heritability (h2) of 0.51. Predicted extractive content ranged from −4.4% to 31.7% and had a value for h2 of 1.16.

Collapse was higher in the heartwood than in the sapwood and heartwood collapse revealed genetic control of h2 = 0.30) while lower heritability was found for sapwood collapse (h2 = 0.12). Heritability for acoustic velocity was h2 = 0.36. There was significant positive genetic correlation between the heartwood diameter and the core length (rg = 0.88), that is, large trees also having the most heartwood. However, a significant negative correlation was revealed between the heartwood diameter and extractive content (rg = −0.45), indicating that a compromise is required for simultaneous genetic selection to be feasible. Genetic gain, especially for heartwood diameter, growth and extractives, can be realised in this species. However, improving acoustic velocity for this species might be challenging, as low genetic variation (CGV = 6%) was observed.