A relevant number of ecological questions raised by policymakers, managers, and citizens often pertain to the short-term future (e.g., the coming days or weeks). In this sense, short-term ecological and biological forecasts can make substantial and practical contributions to achieving policy objectives and benefit society broadly. Specifically, short-term forecasts of Essential Biodiversity Variables (EBVs) and Essential Ecosystem Service Variables (EESVs) can support decision-making by stakeholders from multiple sectors, enabling to anticipate ecological transformations and support proactive, informed decisions that promote conservation, economic activities, and human well-being.

The aim of this task was to demonstrate how a European Biodiversity Observation Network can support the generation of short-term spatial forecasts of ecological and biological phenomena relevant to the Bioeconomy Strategy and to EU citizens at large. Our specific objectives included showcasing 1) a computational workflow that enables the production of days-ahead forecasts for distinct ecological or biological phenomena and 2) a specialized computational workflow for days-ahead forecasts of bird aerial biomass. The first, (‘generic’) workflow, is exemplified using two case studies: i) forecasting the fruiting of a wild mushroom of commercial and recreational relevance, and ii) forecasting the life stage of relevance for surveillance of an invasive pest species important for agriculture. These case studies aim to demonstrate specific, tangible contributions that short-term ecological forecasting can make towards the sustainable use of bio-based economy sectors, ecosystem protection, and anticipation of ecological risks. Beyond aligning with the EU Bioeconomy Strategy, our three forecasting targets also offer relevant contributions to a wider range of EU strategies and policies.

We actively involved stakeholders in defining the end-products and in the development of computational modelling approaches of the workflows. This process entailed two distinct approaches. For the generic forecasting workflow, we engaged in a participatory process from the project's start, focusing on stakeholders involved in mushroom foraging and experts in mycology and ecological modelling. For the bird aerial biomass forecasting workflow, we built upon substantial developments that predated the project, with our engagement primarily drawing on insights and input from earlier initiatives. The two workflows serve complementary purposes in terms of the primary data they use. While the first (generic) workflow is based on the growing body of opportunistic biodiversity observation data, particularly from citizen science initiatives, the second workflow requires highly specialized radar data from weather stations. However, both workflows use predictor data from weather observations and forecasts and employ machine learning algorithms to correlate these data with observed variations in the phenomena being forecasted.