关键词: 竞争效应, 空间效应, 混合线性模型, 遗传参数, 杉木

Abstract:

Objective: Genetic testing is an important approach for genetic improvement of forest trees and can be used to estimate genetic parameters and select elite trees. This study is to explore the effects of different models by combining spatial and competition effects on the estimation of genetic parameters of growth traits in Chinese fir (Cunninghamia lanceolate) at different forest age stages, in order to provide a theoretical basis for improving the accuracy of breeding value prediction and selection of elite trees. Method: This study was conducted in a 40-year-old open-pollinated progeny trial forest in the Yangkou State Forest Farm in Fujian Province, and a mixed linear model was used to fit the effects of experimental design, spatial, competition, spatial-competition effects for diameter at breast height (DBH) and tree heights measured at different tree ages from the 40-year-old Chinese fir trial. The Akaike information criterion (AIC) was used to evaluate different models, and compare genetic parameters and breeding values of the different models. Result: 1) The spatial-competition effect of fit increased by an average of 39.72% of the narrow heritability estimates of growth traits. The additive genetic variance component of DBH in the mixed spatial-competition model increased 4.46%, 26.95% and 35.87% compared to that of the standard mixed linear model (MLM), spatial model, and competition model, respectively. While the residual variances of DBH in the mixed spatial-competition model decreased by 21.95%, 26.81% and 20.19% compared to those of the standard mixed linear model, spatial model, and competition model, respectively. The additive genetic variance of the tree height in the spatial-competition model was 10.73% and 1.22% higher than that in MLM in 1991 and 2020, respectively, and the residual variance decreased by 12.75% and 33.83%, respectively. 2) The intensity of competition was related to the distance between stand individuals. The residual correlation among individuals showed that the competition effect of growth traits in the same row and column was significantly higher than the competition effect of individuals in the diagonal direction. 3) There was a competitive effect on growth traits, and diameter growth was greater than tree height growth. The additive-competitive effect correlation of the same trait showed that the growth traits of Chinsese fir showed a competitive effect at different forest ages, and the competitive effect of diameter growth was greater than that of high growth at the young forest age stage; the growth traits at the middle forest age stage showed a strong competitive correlation, and combined with the residual correlation results, it showed that the growth traits of Chinsese fir showed a strong competitive effect in the middle forest age. The competitive influence among individual chest diameters was greater than the competitive influence between tree height, and DBH was more sensitive to competition than tree height. 4) The accuracy of the breeding values was 0.559 under 5% selection rate, and 23 individuals were selected, among them, 78.3% of individuals were replicated. The mean values of DBH of standard MLM, spatial model, competition model and spatial-competition were 37.3, 37.2, 37.3, 37.3 cm, respectively. and genetic gains of DBH for the selected individuals were 12.73%, 13.86%, 13.32% and 14.37%, respectively. Conclusion: Using competition-spatial model improves the accuracy of genetic parameter estimation. It is showed that the competition of individual plants in the diagonal direction is weaker than that of individual plants in the row and column directions. The growth traits at middle stand ages are most affected by competition, DBH is more sensitive to competition than tree height.

Key words: competition effects, spatial effects, mixed linear model, genetic parameters, Chinese fir