| Artificial forest is an important part of forest resources in China.Although its area ranks first in the world for a long time,the area of young and middle-aged forest is too large.Therefore,there are still problems of shortage of wood resources,uneven wood quality and unequal wood distribution in China.Therefore,because of protecting natural forest,rational management and utilization of artificial forest has great significances.As the basis of forest management planning,stand yield model is used to simulate and predict stand growth in the natural state,so accurate prediction of stand growth dynamics can provide important data support for reasonable management planning.In this study,artificial Larix gmelinii forest and Pinus sylvestris forest in Mengjiagang Forest Farm were taken as the research objects,the stand volume growth model with dummy variables was constructed based on Richards equation,and based on the optimal stand volume growth model,four forest planning models with different constraints were established,and the effects of different spatial constraints on the planning results were analyzed.The main research contents and results are as follows:(1)Construction of stand volume dummy variable model.The optimal basic model,i.e.Richards model,was fitted and selected among five alternative theoretical growth equations,and the stand volume growth basic model with the average age of stand as independent variable was established.On this basis,dumb variables of stand type and related site factors were introduced successively to construct a unified stand volume dumb variable model.The results showed that introducing dummy variables of stand type could improve the fitting precision of the model.The goodness of fit of Pinus sylvestris plantation was greatly improved,and its coefficients of determination was increased by 14%.Among the site factors,slope position and altitude factors had a significant effect on stand volume growth,so it was introduced into dummy variables model with stand type.The results show that the introduction of slope position and altitude dummy variables can further improve the goodness of fit and prediction effect of the model to a certain extent.The determination coefficient of the model is increased to 0.6676,and the prediction accuracy can reach 96.17%.(2)Establishment of plantation management planning model.Taking the maximization of wood yield during the planning period as the management goal,four planning models are constructed.The models contain constraints such as minimum harvest age,volume equilibrium harvest,cumulative harvest quota of each stage,cutting times and[0-1]type decision variables.In the planning model constructed,the non-spatial model only includes the above non spatial constraints.In addition to the above non spatial constraints,the other three spatial models add adjacency constraints,T_m=1 green-up constraints and T_m=2 green-up constraints to construct adjacency model,T_m=1 model and T_m=2 model respectively.(3)Solution and comparison of plantation management planning models.Four different planning models are solved based on Lingo11.0 software.The results show that compared with the non-spatial model,the objective function values of the three spatial planning models are reduced in varying degrees,among which the reduction of the adjacent model is minimal and negligible.The other spatial planning models are reduced by about 2.33%and 17.13%respectively compared with the non-spatial model.The student’s t test shows that there is no significant difference among non-spatial model,adjacency model and T_m=1 model,but there is significant difference between T_m=2 model and the other three planning models.The area distribution of age class at the end of the planning period has been improved under the four constraints,and the age class adjustment effect of T_m=1 model is the best.Although the addition of spatial information in the process of forest planning will reduce wood yield,the temporal and spatial distribution of management measures will be more reasonable.Considering different green periods in spatial constraints will have varying degrees of impact on the planning results,therefore,forest decision-makers should make corresponding choices according to the actual situation and management needs. |
