Multi-objective Forest Planning Research Articles

Multi-objective land allocation for zoning of ecosystem services in mountain forests

Abstract The paper proposes a system for zoning of mountain areas based on the level of provisioning of ecosystem services. Techniques of multi-objective land allocation were applied to allocate complementary and conflicting objectives. The zoning system consists of four phases: i) Identification of criteria for the evaluation of ecosystem services; ii) Quantification of criteria for three different forestland states; iii) Evaluation of potential and effect of the forest on providing the ecosystem services and iv) Zoning of ecosystem services with their prioritization and spatial allocation of support measures. The study was conducted in the Tatra Mountains (Slovakia). Erosion control, avalanche control, wood production and cultural services were evaluated. The greatest differences between potential and effect of the evaluated ecosystem services were identified for the avalanche control. A comparison of our results with the existing (control) map of ecosystem services has proved that the proposed system is a potent means for multi-objective forest planning.

Influence of size and shape of forest inventory units on the layout of harvest blocks in numerical forest planning

The purpose of this study was to assess the effect of using alternative types of forest inventory units (FIUs) in multi-objective forest planning. The research was carried out in a Mediterranean forest area in central Spain. The study area was divided, alternatively, into pixels (square cells) and segments of two different sizes (small and large), which represented the tested FIU types. Airborne laser scanning data (ALS) and field sample plots were combined using the area-based approach to estimate forest attributes for each FIU. Dynamic treatment units were created using cellular automaton optimization aiming at maximizing timber production during a 60-year plan with periodical even-flow cuttings both with and without the aim of creating aggregated harvest blocks. The hypothesis was that the use of segments would enhance the clustering of harvests, as compared to cells, and provide dynamic treatment units more suitable for forestry practice. The results showed that segment-based planning created compact harvest blocks even without the use of spatial objective variables in optimization. The spatial layout of the solution for large segments was the most efficient in the absence of spatial objective variables. The FIU type that performed the best in maximizing timber production was the small segments. For the three tested FIU types, the inclusion of spatial objective variables further improved the clustering of harvests, especially during the latter half of the 60-year planning period. Segmentation acted as a first-phase clustering that made spatial optimization easier and faster. In the case of square cells, the clustering of harvests was greatly improved by the inclusion of spatial goals. The forest planning system and the spatial optimization method proposed in this study maximize the utility of fine-grained ALS data.

Use of interactive data visualization in multi-objective forest planning

Common to multi-objective forest planning situations is that they all require comparisons, searches and evaluation among decision alternatives. Through these actions, the decision maker can learn from the information presented and thus make well-justified decisions. Interactive data visualization is an evolving approach that supports learning and decision making in multidimensional decision problems and planning processes. Data visualization contributes the formation of mental image data and this process is further boosted by allowing interaction with the data. In this study, we introduce a multi-objective forest planning decision problem framework and the corresponding characteristics of data. We utilize the framework with example planning data to illustrate and evaluate the potential of 14 interactive data visualization techniques to support multi-objective forest planning decisions. Furthermore, broader utilization possibilities of these techniques to incorporate the provisioning of ecosystem services into forest management and planning are discussed.

Modelling bilberry and cowberry yields in Finland: different approaches to develop models for forest planning calculations

The aim of this research was to create yield models for the two most significant wild berry species in Finland, namely, the bilberry and the cowberry. In particular, the aim was to develop models applicable to multi-objective forest planning calculations. The models were developed using approaches based on both expertise and empirical measurements. In expert-knowledge-based studies, different methods of data collection were used: experts evaluated berry production of different forest stands (i) from slides, (ii) visually in the field and (iii) using a questionnaire. Data for empirical studies were also collected using different methods: measurements (i) of systematically placed sample plots and (ii) of forest stands found to be advantageous with respect to berry production. Statistical analyses of data and modelling techniques varied from study to study. Thus, another aim was to test different methods of data collection and different modelling techniques and also assess their usability. According to the results, the best bilberry yield can be found in a mature stand which is not too dense and is located on a mineral soil site of medium or rather poor fertility. It is also obvious that pine-dominated heath forests of rather poor or poorer fertility have the highest cowberry yields. A stand suitable for cowberry picking should not be dense. These results are very much in line with previous studies on the effects of site and stand characteristics on berry production. However, it is concluded that there is still a need to for further research to explore the dependence of cowberry crops on stand characteristics on poor mineral soil sites. Comparisons of the results of this thesis with empirical berry yield studies and also with earlier berry models indicate that modelling expertise is a reasonable way to create production functions for different wild berry species. This finding encourages the utilization of expertise in other applications for the development of models for different non-wood forest products. In particular, the method based on visual assessments conducted in the field was found to be a promising one.

A Decision Support System for Assessing Trade-Offs between Ecosystem Management Goals: An Application in Portugal

Cork oak (Quercus suber L.) and holm oak (Quercus rotundifolia) ecosystems are characteristic of Mediterranean forestry in Portugal. Even though cork is the most valuable product, these ecosystems provide multiple products and services. Assessing trade-offs between multiple goals is thus critical for the effectiveness of oak ecosystem management planning. This paper focuses on the development of a decision support system for oak ecosystems’ scenario analysis including multiple criteria. It includes an innovative decision support systems (DSS) functionality to assess trade-offs between the criteria that may support negotiation and consensus building between decision-makers and forest stakeholders. Specifically, a module that encapsulates the Feasible Goals Method/Interactive Decision Maps (FGM/IDM) technique is developed for interactive visualization of the Pareto frontier. The Pareto frontier illustrates the degree to which improving one particular criterion requires accepting sacrifices in the achievements of others. It thus provides information about trade-offs between competing decision-makers’ preferences. Results are discussed for a large-scale application encompassing over 1 million ha of cork and holm oak forest ecosystems in Southern Portugal. This study demonstrates the potential of the new DSS functionality to enhance multi-objective forest planning, namely by facilitating participation by stakeholders and providing transparency to the decision-making processes.

Importance of priority areas for multi-objective forest planning: a Central European perspective

SUMMARY In the framework of multi-objective forest management, ‘priority areas’ which are relatively more important for the selected management objectives are commonly designated. Using a comparative analysis of guided interviews, we examined the use and importance of priority areas in forest planning in nine Central European countries. In all countries, priority areas have been widely used, forest function areas and protected areas being the most common. According to management objectives, more than 20 types of priority areas were recognised, with priority areas for protection against natural hazards, nature conservation, recreation, welfare, and production being the most prevalent. Criteria for the designation differ among the countries; however, site conditions and infrastructure facilities are most often used. The scale of designation ranges from 1:10 000–1:50 000, and the size of priority areas varies from 0.1 ha to several hundreds of ha. The level of participation of stakeholders involved in the designation of priority areas differs among and within the countries. The effectiveness of priority areas for forest management can be improved by transparent designation criteria, objective oriented management measures, and efficient financial instruments.

Spatial genetic algorithm for multi-objective forest planning

A spatial operator for genetic algorithms is introduced herein, in order to improve their efficiency for multi-objective spatial forest planning. Constrained NSGA-II is used as a standard genetic algorithm for the integration and the evaluation of the proposed methodology. A typical harvest scheduling problem is examined with two objectives: a) maximization of timber volume and b) minimization of sediment levels. Two constraints are imposed: a) minimum timber yield and b) even‐flows. The proposed algorithm (Spatial NSGA), gives better results for both the constrained and the unconstrained problem. Moreover, it achieves old forest compactness, although it is neither a separate objective nor a constraint, but it renders compactness as an emergent result. The purpose of the suggested approach is to support forestry decision-making by generating a set of optimal management alternatives. The implementation of the method produced a Pareto front consisting of non-dominating solutions and showing the tradeoffs between timber harvest and sediment levels in water runoff. The present approach offers potential applications to a wide spectrum of spatial planning problems beyond the one examined in this paper.

Developing a Multi-Objective Forest Planning Process with Goal Programming: A Case Study

In this study, multi-objective forest management planning process and application were realized in a case study involving faculty research forest. Forest values were quantitatively determined. Goal programming was used in developing multi objective forest management planning model. Forest function considered as forest management objectives (goals) in the model include wood production, soil protection and water production. Four different models containing different goals combinations were developed. Goals were prioritized differently in these models. The goals were fully achieved. Moreover, for the tree goals, the optimum GP solutions presented wanted deviations with respect to the initial target values. Forest function maps were prepared according to models solutions in the end of this study.

Multi-objective Forest Planning. Managing Forest Ecosystems, Vol. 6

Multi-objective Forest Planning. Managing Forest Ecosystems, Vol. 6

The performance of alternative spatial objective types in forest planning calculations: a case for flying squirrel and moose

This study examined the characteristics and performance of three different types of spatial objectives in multi-objective forest planning. These types were tested with a case study problem that aimed at improvements in the spatial pattern of the habitats of flying squirrel ( Pteromys volans) and moose ( Alces alces). In the first step, the habitats were determined by using a stand-level habitat suitability index (HSI). Stands in which the HSI exceeded a user-specified limit were considered as habitats. In the forest level problem formulations, the spatial objective types were included in the multi-objective forest-planning model, which was solved by using a heuristic optimization technique. The first type of spatial objective was the proportion of a specified type of stand boundary of the total boundary length. For flying squirrel, the aim was to increase the boundaries between neighboring stands, which were both suitable for breeding and foraging. For moose, the aim was to increase the amount of edge, i.e. the boundary between winter-foraging habitats and other areas. The second spatial objective type was the spatial autocorrelation of HSI. It was maximized for flying squirrel and minimized for moose. Whereas maximization of spatial autocorrelation tends to cluster similar stands, its minimization produces a pattern where neighboring stands often have different characteristics. The third spatial objective type was a weighted mean of the stand-level HSI. The weight depended on the stand location being highest for current habitat patches and decreased as the distance to them increased. This objective, which tends to cluster habitat patches, was applied to flying squirrel only. The spatial objective types were tested in six forest areas located in north Karelia, Finland, ranging from 404.6 to 984.8 ha in area. The first spatial objective resulted in a large area of habitat exceeding the threshold HSI value with a good spatial pattern. The second spatial objective yielded a clearly lower total habitat area but a good spatial pattern, and a good connectivity of the flying squirrels’ habitats. The third objective resulted in a slightly improved spatial pattern of the flying squirrels’ habitats.

Multiple risks in multi-objective forest planning: integration and importance

Multiple risks were integrated in multi-objective forest planning with the scenario technique and with an application of multi-attribute utility theory. The sources of risks were divided into three categories: risks related to inventory data, risks arising from unknown future states of nature, and risks associated with the decision maker's preferences. Stochastic models and other procedures for producing scenarios of inventory data, growth rate, mortality and natural regeneration of trees, success of tree planting, timber prices, cost level, and importance of management objectives are described. Some of the models for the scenarios were based on empirical data and obtained from earlier investigations, while others were specified by parameters given by the decision maker. The decision maker's attitude toward risk was integrated in planning via the distribution of a utility index. The utility index was computed from an additive utility function. Cumulative points of the distribution were weighed in a manner reflecting decision maker's risk preferences. The weighed utility index was used as the ultimate criterion when searching for the optimal plan. Optimization was done using a heuristic algorithm. The case study conducted indicated that both risks and attitude toward risks affect the choice of plan; optimal plans produced by a deterministic system were different from the stochastic optima, and a risk-neutral decision maker had optima different from those of risk-averse and risk-seeking decision makers. Depending on the decision maker's attitude toward risk, utility losses ranging from 0.1% to 4.5% were to be expected when the deterministic optimum was selected. Implementing a risk-neutral decision maker's optimal plan caused utility losses less than 3.5% for risk-averse and risk-seeking decision makers.

Operationalization of biological diversity as a decision objective in tactical forest planning

An approach for including biological diversity in calculations of multiobjective forest planning is described. This makes it possible to consider biodiversity as a decision objective among multiple objectives in tactical forest planning. Biodiversity is operationalized by decomposing it into measurable environmental components that describe its dimensions. The relative importance of the components is assessed via pairwise comparisons. The relationship between the value of the component and the overall biodiversity is estimated and described for each component as a subpriority function. The relationship can be nonlinear. A biodiversity index is calculated for each alternative forest plan based on the components and their weighting, the subpriority functions, and the predicted qualities of the forest area when implementing the plans. The approach is illustrated by a case study. Using this approach, the relationship between biodiversity and the other objective variables, as well as the effects of the components of biodiversity, can be analyzed numerically.

Interactive Resolution of Multi-objective Forest Planning Problems with Shadow Price and Parametric Analysis

Abstract Forest management planning is a joint process of decision synthesis and tradeoff analysis. We present an interactive method of multi-objective linear programming for supporting the planning process. In the proposed procedure, emphasis is placed on reducing uncertainty that surrounds the resolution of goal conflicts. Parametric analysis is iteratively performed to shed insights into the question of what happens to goal tradeoffs if a particular criterion is achieved at alternative levels. The composite objective function guiding the parametric analysis is updated as decision maker replaces a previous choice with an improved one. Central to this step is the induction of criterion weights implied by a decision instance, through the use of shadow prices. Finally, ordinal goal priority is employed to generate initial solution and decompose a multi-objective planning problem into more tractable subproblems. The interactive procedure is illustrated by a land-use planning model containing five forest-related outputs, including noncommercial grazing, commercial grazing, timber, camping, and profit. For. Sci. 41(3):452-469.