Spatial and competition models increase the progeny testing efficiency of Japanese larch
Leiming Dong,Yunhui Xie,Harry X. Wu,Xiaomei Sun
DOI: https://doi.org/10.1139/cjfr-2020-0007
2020-12-01
Canadian Journal of Forest Research
Abstract:The main purpose of this study was to examine spatial and competition effects on estimates of genetic parameters, as well as on selection options for growth traits, including height (H), diameter at breast height (DBH), and volume (V), in a progeny test of Japanese larch (Larix kaempferi (Lam.) Carrière) at age 20 years. We compared performances among the individual-tree additive genetic base model (B) with design factors only, the spatial effect model (AR1), the competition model (C), and the combined competition and spatial model (CS). We found that spatial heterogeneity had significant effects on growth traits and that plot variance decreased by more than 80% in the AR1 model relative to the B model. Competition had significant effects on DBH and V but a smaller effect on H. In the C model, direct additive genetic variances ([Formula: see text]) for DBH and V increased by 205% and 93%, respectively, whereas residual variances ([Formula: see text]) decreased by 8% and 6%, respectively. In the CS model, the correlations between direct and competitive genetic effects were 0.83, −0.97, and −0.98 for H, DBH, and V, respectively. Competition significantly affected the forward selection. The proportions of selected elite trees were only 39% and 25% common between the B and CS models for DBH and V, respectively, when selection intensity was 5%. For breeding selection, depending on thinning regimes planned, trees of high additive breeding values but low competitive breeding values are preferable for plantation.
forestry