Abstract
Using a hierarchical Bayesian model, we analyzed effects of suppression by woody vegetation and release after improvement cutting on growth (tree height and basal diameter) of Cryptomeria japonica clones. The model was constructed based on the differential of the Gompertz equation. The response variable was annual growth and the explanatory variables were annual initial size, suppression indices with coefficients, and growth decrease ratio after release for both traits. Each factor included a fixed effect represented by the average and a random effect of the clone. The suppression indices for evergreen and deciduous trees were the sum of the angles to the top of the competitors within a constant radius (R) from a fixed proportion (β) of the height of the target tree. The values of R and β used were 1.5 m and 0.75, respectively based on the Deviance Information Criterion. The estimated fixed effects indicated that basal diameter growth was more susceptible to suppression but recovered faster after release than height growth. The random effects of the clone suggest that the parameters of the initial growth rate and the maximum size were different among clones with respect to both traits. Clone differences were also observed in the response to deciduous-tree suppression and the recovery in height, but not basal diameter, growth after release. A weak positive correlation was found between the initial growth rate and the recovery after release of each clone. The characteristics of clones that were adapted to Cryptomeria japonica forestry with reduced weeding were discussed.
