Genetic variation and age-age correlations of Eucalyptus grandis at Dongmen Forest Farm in southern China

Five seed source - family trials of Eucalyptus grandis were planted at Dongmen Forest Farm in southern Guangxi, China, between 1987 and 2001. Growth was assessed up to age 5 y on an annual basis for three trials and in 4 out of 5 y for two trials. In 2008 growth was assessed again, along with pilodyn penetration, at trial ages ranging from 7 to 21 y. These trials initially contained 386 families representing 49 seed sources, including a total of 97 families from four different improvement programmes. Two trials included small numbers of Dongmen second-generation families. Thinning was applied variously in all trials. Large significant differences between both seed sources and families within seed sources were observed at each assessment. At age 5 y, which approximates an average rotation for fibre plantations in southern China, the best seed sources included Ravenshoe, Windsor Tableland, Copperlode and Paluma areas of northern Queensland; Bellthorpe in southern Queensland; Aracruz, Brazil; and Florida, USA. Some families from local selections at Dongmen also performed very well. Sources from the Coffs Harbour region in New South Wales were markedly inferior for growth, as was a South African seed orchard source. Superiority of seed sources for growth at age 5 y was generally maintained through to the 2008 assessment. At the later-age assessments in 2008 (trial ages 7, 18 and 21 y), sources notable for combining superior volume growth with higher wood density (i.e. lower pilodyn penetration) included Ravenshoe, Queensland; Aracruz, Brazil; Florida, USA; and some from Dongmen, China. However, the 2008 growth results may have been biased by heavy selective thinning regimes. Within-seed-source heritabilities at single sites for height at age 1 y and then volume at ages 2, 3, 4 and 5 y ranged from 0.03 +/- 0.02 up to 0.45 +/- 0.10 and generally peaked at age 3-4 y in each trial. Genetic correlations between growth at early ages (height at 1 y, volume at ages 2, 3 and 4 y) with growth at rotation age (age 5 y volume) were moderate to very strong. Optimum selection efficiency for volume was generally reached at age 2 to 3 y but did not decline markedly between ages 3 and 5 y.