A novel collaborative planning framework for artificial forest harvesting and replanting

Abstract

Achieving precise and efficient selective harvesting-replanting management in artificial forest quality improvement projects faces significant challenges. Random search algorithms, such as Monte Carlo and simulated annealing, are often used to find near-optimal harvesting plans but are inefficient, costly, and poorly scalable, leading to unstable solutions. Additionally, replanting efficiency post-harvest is low and rule-based. The synergistic effects between harvesting and replanting are often overlooked, limiting the potential for forest structure improvement. To address this issue, this study develops a unified harvesting-replanting collaborative planning framework that integrates the novel monarch-based enhanced genetic algorithm (MEGA), a new Delaunay boundary force algorithm (DF-Boundary), and two multi-objective functions. The framework also incorporates 3D visualisation technology for simulation validation and has been applied to eight mingled plots in southern China. Simulation results show that the diameter distribution of the plots follows an inverted-J shape, with improved species diversity. After selective harvesting, the objective function values ($L_1^{\ast }$) of 8 plots increased by 38.51 %–198.22 %, outperforming other algorithms in performance and stability. Replanting layout reduced the objective function values ($L_2^{\ast }$) of 8 plots by 4.96 %–19.96 %. All spatial structure indicators significantly improved. The results demonstrate that the proposed framework excels in both accuracy and efficiency, enhancing forest structure and moving the forest toward a near-natural state. Further exploration of its application in larger-scale forest ecosystems is needed.