Multi-Objective Management Models Of Larch(Larix Olgensis A.Henry)Plantations In Heilongjiang Province

As the main body of terrestrial ecosystem,forest provides economic,ecological and social value for the development of human society.However,with the continuous progress of human industrialization,forest ecological environment has been seriously damaged.At the same time,problems such as low forest productivity and weak ecological service function which caused by unscientific and unreasonable management activities in the forest management process have seriously restricted sustainable development.In this context,how to adopt effective management measures and formulate scientific forest management plan to improve forest quality has become the urgent problem for forest managers to solve.In this study,Larix olgensis plantations in Heilongjiang province were taken as the research objects,multi-attribute utility theory was applied to construct a multi-objective function including net present value,large log production and forest carbon storage during the rotation and compromise analysis,in which the optimal weight ratio of each object is determined by compromise analysis.The particle swarm optimization algorithm was used to solve the optimal management schedules of different standard stands,while the multi-objective management models of Larix olgensis plantations was constructed based on the optimization results of real stands.The main research contents and results are as follows:（1）Weight analysis of objective function and simulation optimization results of standard stands.The synergistic relationship and trade-off curves were analyzed between the objectives when the weight of forest carbon storage was fixed at 0,0.1 and 0.2.Finally,the weight ratios of net present value,large log production and forest carbon storage were determined to be 7:2:1 and6:2:2,respectively.At the same time,in order to compare the management performance between different management objectives,a single objective equation with the maximum net present value as the management objective was constructed.The simulation optimization of single objective equation and multi-objective equation was carried out to analyze the influence of different management objectives on the optimization results.The results show that:The weight of carbon storage significantly affects the optimization results of management objectives.With the increase of the weight of carbon storage,the mean annual large log production was increased by 20%-110%,the average forest carbon storage was increased by 9%-66%during rotation,The rotation lengths were extended by 4-30 years,the number of remaining trees during thinning was increased by 26%-69%,but the net present value was decreased by 6%-26%.The multi-objective utility function MOF₁ is a compromise between different management objectives.Management optimization results were related to stand site quality and afforestation density.Under the same management objective and afforestation density,with the improvement of site quality,the thinning time was advanced by 1-14 years,the number of remaining trees was reduced by 3%-19%,and the rotation lengths were shortened by 3-7 years for single objective function and 5-12and 15-36 years for multi-objective functions,respectively.Under the same management objective and the same site quality,with the increase of afforestation density,the thinning time was advanced 3 years on average to improve the growth of remaining trees.Finally,the best management schedules of different standard stands under different management objectives was proposed by summarizing the growth and cutting process of different management objectives.（2）The fitting and testing of management models.The multi-objective management models of Larix olgensis plantations was fitted according to the simulation optimization results of each sample plots when the management objective was MOF₁.The predicting accuracy of the management models was evaluated by testing the management models and comparing the simulation optimization results and the prediction results of management models.The results showed that,the stand basal area before and after thinning and the clear-cutting diameter were positively correlated with site index,but negatively correlated with discounting rate in the management models.In the thinning models,the effect of discounting rate on the stand basal area before and after thinning increased with the increasing of stand age.The stand basal area before and after thinning also decreased with the increasing of the number of thinning.Compared with the first thinning,the stand basal area before and after the second thinning decreased by1.078 m²/hm² and 1.192 m²/hm²,respectively,and the stand basal area before and after the third thinning decreased by 3.403 m²/hm² and 2.519 m²/hm²,respectively.The thinning intensity could be calculated by the basal area before and after thinning,which was inversely proportional to site index and directly proportional to the discounting rate.In the three thinning,the intensity of the second thinning（19%-20%）was greater than that of the first thinning（18%-19%）,which was greater than that of the third thinning（15%-18%）.In the clear-cutting model,under the same site index,the clear-cutting diameter decreased by 0.754 cm when discounting rate increased from 2%to 4%.Under the same discounting rate,the clear-cutting diameter increased by 1.23cm when site index increased from 18m to 24m.According to the test results of the models,all the test indexes of the management models constructed in this study were all small,indicating that the fitting error of the model is small and the prediction effect is good.（3）Application and evaluation of management models.This study introduces the specific steps of the application of the models,that is,by comparing diameter at breast height and stand basal area of real stands with the prediction results of the management models to determine whether the stands should be clear-cut,thinned or no management measures should be taken.Three young larch plantations with different stand densities and site indexes were selected in the study areas,and the prediction accuracy of the management models was evaluated by comparing simulation optimization results of the stands with prediction results of the management models.The results showed that the stand basal area before and after thinning of the three plots based on simulation optimization was generally distributed within the prediction range of the thinning models,and the difference between the rotation predicted by the clear-cutting model and the simulation optimization results was 1-3 years,which proved that the prediction accuracy of the management models constructed in this study was good and could effectively guide the multifunctional management of Larix olgensis plantations.