BIOREFOREST: Reforestation of forests and orchards after disturbances – A transformative action plan balancing socioeconomic and biodiversity needs

Summary

The BIOREFOREST project aims to initiate a societal transformation towards increased awareness of biodiversity as a key factor in the planning of reforestation strategies for forests and orchards. The objectives are to: i) assess the impact of different reforestation strategies on biodiversity and economic outcomes across spatial and temporal scales; ii) assess the interaction of disturbances and reforestation successes; and iii) develop site-specific approaches that: a) balance site-specific ecological and economic needs, and b) represent customised reforestation strategy for stakeholders. BIOREFOREST will focus on European forest sites affected by natural disturbances, reforested as mono-species (including orchards) or mixed-species stand, compared with naturally regenerated sites with minimal management intervention. The project will combine state-of-the-art data collection, modelling, AI and participatory approaches to guide reforestation decisions. Reforestation sites and their spatial and temporal development will be identified using satellite data from the past 50 years. Policy makers, forest and orchard owners of identified reforested sites will then be engaged in a participatory activity network. Landowners’ engagement will allow BIOREFOREST to collect site-specific information via customized questionnaires and a systematic sampling campaign of soil, leaf, and deadwood samples, along with other biodiversity and forest structure data. Multitrophic biodiversity will be analyzed using metagenomics for each site, and the data will be functionally annotated. Remote sensing information, metadata from selected forest and orchard sites, and biodiversity data will be used to carbon sequestration, and ecosystem services via the LPJ-GUESS model, in relation to climate change as well as potential threats to plants and soil quality. Finally, all data will be used to train, validate, and test eXplainable AI (XAI) models including Random Forest, XGBoost, and Neural Networks to assess site-specific reforestation strategies that best balance biodiversity goals and economic needs. The XAI models trained with all features will extract important markers to serve as decision support systems. These markers will include taxonomic and functional indicators to monitor ongoing and future reforestation efforts on biodiversity and predict effects of reforestation strategies on biodiversity.

The outputs of the project will provide a decision support system on how to balance biodiversity and economic outcomes of reforestations. Moreover, forest and orchard owners, land managers, and policymakers will be engaged through workshops, training, and self-sampling campaigns throughout the project. This will ensure that reforestation strategies are practical, widely adopted, and aligned with the goals of the EU Forest Strategy, the Nature Restoration Law, and the Soil Mission.