Reconciling short- and long-term predictions for ecosystem management

Marta Magnani , Mara Baudena, Leonardo Ancillotto , Alessio Collalti, Daniela Dalmonech, Fasma Diele, Giulia Furlanetto, Silvana Beatriz Goirán, Carmela Marangi, Antonello Provenzale, Cesare Ravazzi, Roberta Sciascia, Gabriele Vissio, Paolo Fiorucci, Gianluigi Ottaviani

1. Management plans grounded in scientific evidence can be used to limit the impacts of ongoing global changes on socio-ecological systems. In this framework, modeling tools play a crucial role in informing and supporting management strategies.

2. While the urgency of implementing evidence-based actions directed most scientific efforts towards short-term ecological forecasting (ranging from days to decades), we argue that long-term projections (longer than a few decades) can be as important as short-term forecasts. Complex ecological feedbacks and long-term ecosystem dynamics can have effect over decades if not centuries, possibly leading to undesired management outcomes. In this viewpoint, we highlight the need to incorporate long-term ecosystem responses into decision-support studies and discuss the technical requirements and current limitations of state-of-the-art modeling frameworks and datasets.

3. We recommend defining the prediction horizon based on intrinsic ecosystem timescales and studying ecological legacies at biogeographical levels higher than the landscape, such as ecoregions. Combining information from different sources could provide complementary data layers with varying resolution, detail, and uncertainty. Integrating and leveraging these information layers across different spatiotemporal scales represents a key step towards reconciling short- and long-term predictions.

4. We emphasize the necessity of routinely integrating short- and long-term predictions. To this end, we envisage international communities that foster the convergence of transdisciplinary knowledge and expertise, also engaging with stakeholders, to generate timely and reliable ecological predictions aiming at assisting management planning through a mutual learning loop.