Identification and joint modeling of competition effects and environmental heterogeneity in three Douglas-fir (Pseudotsuga menziesii var. menziesii) trials

Abstract

(Co)variance matrices for the assumed model, and thus the specification of the dispersion parameters, should take into account both the negative competition and the positive spatial correlations. In this context, we applied several approaches to identify and quantify the genetic and environmental competition effects and/or environmental heterogeneity in three Douglas-fir genetic trials from the British Columbia tree improvement program in total height and diameter at breast height at ages 12 and 35. Then, we applied an individual-tree mixed model to account jointly for competition effects and environmental heterogeneity (competition + spatial mixed model, CSM). We also compared the resulting estimates of all dispersion parameters and breeding values (BVs) with corresponding estimates from three simpler mixed models. Our analysis revealed that strong spatial environmental variation (predominantly at large-scale) covered the effects of competition in the three Douglas-fir progeny trials. While diameter at breast height at age 35 revealed strong competition effects at both genetic and environmental levels, these effects were not as strong for total height. In general, with strong competition genetic effects, the CSM gave a better fit than the simpler models. Ignoring competition effects and environmental heterogeneity resulted in lower additive genetic variances and higher residual variances than those estimated from the CSM. Ignoring competition effects leads to overestimating environmental heterogeneity, while ignoring the environmental heterogeneity leads to underestimating competition effects. Spearman correlations between BVs predicted from the simplest model and total BVs from the CSM were moderate to high. The implications of all these findings for the genetic improvement of coastal Douglas-fir in British Columbia are discussed.