Tree-Level Decision Support Systems for Forest Management: a Systematic Review

Societal demands for forest biodiversity and ecosystem services (BES) are growing and diversifying, which necessitates careful decision-making in forest management. Decision support systems (DSS) are a valuable tool to compare different management strategies and model the trade-offs between BES objectives, and they are successfully applied for forest management at the resolution of forest stands and landscapes. However, there is a growing interest in developing DSS at an even finer resolution: the individual-tree level.

We present a systematic review of tree-level decision support systems in forest management, which take individual-tree data as input, apply an optimisation algorithm, and prescribe a management decision for every tree as the output. Tree-level DSS directly include relevant tree attributes in the planning process rather than relying on aggregated proxies at the stand level. This enables a greater flexibility and precision in forest management, which complements the developments in close-to-nature forestry, remote sensing and autonomous forest machines. Our review identified 47 studies that describe a tree-level DSS. These studies use diverse optimisation techniques such as heuristic algorithms, mathematical programming and machine learning to generate the decisions. Several management targets have been addressed in the studies, such as economic value, biodiversity, forest fire risk mitigation and the amenity of the landscape. Thanks to advances in remote sensing, rich information about individual trees can be derived, although the attributes typically gathered during field inventory, like species, tree height and diameter at breast height, are still the most commonly used in decision-making.

Important challenges for the further development of tree-level DSS are to include natural disturbance risk predisposition in the management decisions; to design generalisable approaches that accommodate diverse forest BES, rather than focusing only on specific case studies; to connect tree-level decisions with management plans at larger spatial scales; and to enable the real-world implementation of the optimised decisions. Informed by the findings of our review, we will present our ongoing work on a new tree-level DSS designed to address these challenges.