关键词: 长白落叶松人工林, 多目标, 随机多准则可接受性分析, 经营决策

Abstract:

Objective: Heavy reliance on subjective preferences, combined with substantial uncertainty in management objectives weights, often results in inaccurate decision-making in multi-objective management of plantations. In this study, Larix olgensis plantations were taken as the research object to comprehensively explore the multidimensional weight space, providing a sound basis for multi-objective management of plantations. Method: Stochastic multi-criteria acceptability analysis (SMAA) integrated with stand growth simulations was used to determine the optimal thinning schemes for L. olgensis plantations under different site conditions (SI=16, 18, 20, and 22 m). With the annual mean wood production, net present value (NPV), and the average carbon storage over the rotation as the criteria, this study explored the dynamic effects of 12 alternative thinning schemes (upper or lower thinning, interval periods of 5 and 10 years, thinning intensity of 10%, 20%, and 30%) under different site conditions. By analyzing the overall acceptability index ($ a_{i}^{h} $) and decision-making risk degree ($ \text{DR}{\mathrm{D}}^{r} $) of different thinning schemes, each scheme was comprehensively evaluated and ranked from multiple dimensions (comprehensive performance and risk) to identify the optimal thinning scheme. Result: Under the same thinning scheme, management objectives all increased to different degrees with the increase of site index (every 2 meters increase). The mean annual wood production increased by 10%–52%, NPV increased by 10%–67%, and the average carbon storage increased by 1%–47% during the rotation period for the 12 schemes. The upper thinning resulted in higher mean annual wood production and NPV but lower average carbon storage than lower thinning. For a plantation with a SI of 18 m, the increases were 7%?17% in mean annual wood production and 3%?45% in NPV, while the decrease was 19%?47% in average carbon storage over the rotation. For plantations with site indices of 16 and 18 m, the optimal thinning scheme was Ⅷ (i.e. upper thinning over a 10-year interval with a thinning intensity of 10%) as its overall acceptability index reached up to 79% and 80% and its decision-making risk degree was 1%. For plantations on more productive sites with site indices of 20 or 22 m, thinning scheme X (i.e. upper thinning over a 10-year interval with a thinning intensity of 20%) was optimal, with an overall acceptability index reaching 85% and 83% and the decision-making risk degree were 2% and 1%, respectively. Conclusion: Stochastic multi-criteria acceptability analysis (SMAA) can provide strong support for decision-making in multi-objective forest management through a comprehensive exploration of objective preferences and the multidimensional objective weight space. This study demonstrates its usefulness in determining the optimal thinning schemes for L. olgensis plantations across four site quality classes, providing a sound basis for multi-objective management decision of these plantations in northeast China.

Key words: Larix olgensis plantation, multi-objective, stochastic multi-criteria acceptability analysis (SMAA), management decision-making