Abstract
Acoustic velocity (AV) data from 7,423 coastal Douglas-fir trees drawn from 347 wind-pollinated families on 14 sites, from four first-generation testing programs in the north Oregon Cascades, were analyzed. Families were measured on two or four sites at ages 23 to 41 years from seed using the Fakopp TreeSonic standingtree tool. Height (HT) and DBH data collected at ages 15 and 16 from seed, from all trees in the four programs (95,795 trees, 955 families), were used to calculate volume index (VOL = HT*DBH2) and stem taper (TAP = DBH/HT). All traits were analyzed using multivariate mixed model analyses.
Across-site individual narrow-sense heritabilities for AV2 ranged from 0.24 to 0.40 among first-generation programs, compared to 0.12 to 0.23 for HT, 0.10 to 0.16 for DBH, 0.11 to 0.20 for VOL and 0.14 to 0.17 for TAP.
Across-site type B correlations for AV2 ranged from 0.85 to 0.95, compared to 0.62 to 0.83 for HT, 0.60 to 0.74 for DBH, 0.67 to 0.78 for VOL and 0.66 to 0.79 for TAP. AV2 was negatively correlated with HT in three programs (rA = 0.17 to −0.28), and negatively correlated with DBH (−0.12 to −0.46), VOL (−0.05 to −0.44) and TAP (−0.09 to −0.40) in all four programs.
Selecting the top 10% of the families sampled based on AV2 gave predicted gains of 4.4% to 9.6% for AV2 and −9.3% to 10.6% for VOL. The adverse genetic correlations between AV2 and growth, and the losses in gain in AV2 from selection based on growth, may be overestimated by suppression of slower-growing families in these older tests.