Assessment of tradeoffs between ecosystem services in large spatially constrained forest management planning problems

Abstract

Forests provide multiple ecosystem services, some of which are competitive, while others are complementary. Pareto frontier approaches are often used to assess the trade-offs among these ecosystem services. However, when dealing with spatial optimization problems, one is faced with problems that are computationally complex. In this paper, we study the sources of this complexity and propose an approach to address adjacency conflicts while analyzing trade-offs among wood production, cork, carbon stock, erosion, fire resistance and biodiversity. This approach starts by sub-dividing a large landscape-level problem into four smaller sub-problems that do not share border stands. Then, it uses a Pareto frontier method to get a solution to each. A fifth sub-problem included all remaining stands. The solution of the latter by the Pareto frontier method is constrained by the solutions of the four sub-problems. This approach is applied to a large forested landscape in Northwestern Portugal. The results obtained show the effectiveness of using Pareto frontier approaches to analyze the trade-offs between ecosystem services in large spatial optimization problems. They highlight the existence of important trade-offs, notably between carbon stock and wood production, alongside erosion, biodiversity and wildfire resistance. These trade-offs were particularly clear at higher levels of these optimized services, while spatial constraints primarily affected the magnitude of the services rather than the underlying trade-off patterns. Moreover, in this paper, we study the impact of the size and complexity of the spatial optimization problem on the accuracy of the Pareto frontiers. Results suggest that the number of stands, and the number of adjacency conflicts do not affect accuracy. They show that accuracy decreases in the case of spatial optimization problems but it is within an acceptable range of discrepancy, thus showing that our approach can effectively support the analysis of trade-offs between ecosystem services.