Survival modeling of branch–knot developmental dynamics in Korean Pine plantations: Linking climate, region, and artificial pruning

Understanding the developmental dynamics of branches and knots is critical for managing wood quality and crown structure for Korean pine (Pinus koraiensis) plantations. In this study, Cox models and accelerated failure time (AFT) models were employed to analyze branch growth cessation ($\mathrm{TBC}$), death ($\mathrm{TCD}$), and occlusion ($\mathrm{TDO}$) using data from 2706 knots on 101 destructively sampled trees across two representative plantation regions in northeastern China. Time-dependent Cox models revealed that drought ($\mathrm{CMD}$) and high summer temperatures (${T\max }_{\mathrm{JJA}}$) significantly accelerated branch growth cessation and death, with increasing severity as branches aged. Conversely, higher site quality ($\mathrm{SI}$) and relative diameter ($\mathrm{Rd}$) initially delayed branch mortality, though these effects gradually weakened over time. Artificial pruning ($\mathrm{IFP}$) had a significant and progressively increasing effect on knot occlusion over time. Factors such as tree age ($\mathrm{TSB}$), growth rate (${\mathrm{iDBH}}_{\mathrm{averDO}}$) and knot height ($\mathrm{KH}$), also showed significant yet consistent effects on branch–knot development. AFT models provided direct predictions of developmental timing, with $\mathrm{RMSE}$s of 3.5, 2.9 and 4.7 years for $\mathrm{TBC}$, $\mathrm{TCD}$ and $\mathrm{TDO}$, respectively. Model validation further demonstrated good predictive accuracy, with global $\mathrm{MAE}$s of 2.8, 2.2, and 3.4 years, though errors were slightly higher at early and late time intervals. Incorporating climate variables accounted for most interregional variation, rendering geographic indicators insignificant and enhancing model generality. Overall, the models developed in this study offer potential for integration into growth and yield simulation systems to support optimized pruning and thinning schedules, thereby enhancing wood quality and informing adaptive silvicultural management under varying regional climates.