From heuristics to mechanistic understanding (and modelling) of ecological dynamics – a focus on the restoration of Posidonia oceanica meadows

A crucial challenge in ecosystem restoration, due to the ecological system complexities and the spatial and temporal scales involved, is the prediction of system evolution, the associated uncertainties and the final outcomes. A natural approach to this goal is the use of process-based models, but their requirements in terms of data and mechanistic understanding of system ecology still limit their adoption in the restoration of high complexity systems. Indeed, despite the large amount of data globally acquired, the lack of shared comprehensive strategies for what needs to be collected and how, promotes the adoption of heuristic approaches based on collections of failures and successes. A remarkable example in this context is provided by the restoration of Posidonia oceanica meadows, one of the most important Mediterranean marine coastal ecosystems in terms of productivity, biodiversity and control of local, regional and even global ecological dynamics. The limited resilience of these ecosystems, threatened by diverse anthropogenic pressures, forces the adoption of restoration approaches with variable and hardly predictable degree of effectiveness. With the aim to transitioning from heuristic to mechanistic restoration approaches, the research focused on creating an individual-based model of meadow evolution grounded in Dynamic Energy Budget theory, and a cured and harmonized information base on P. oceanica ecology, summarizing more than 6 decades of research in the form of an open geo-database. Results revealed striking imbalances in data type, quality and redundance, with surprising shortage of usable data for model development and parameterization. The coupling between the information base and the model, however, has the potential to form a feedback loop providing the much-needed strategy to move from heuristics to mechanistic approaches. Indeed, the model’s functional hypotheses can orient the collection of data embedded in clear theoretical processes, which in turn allow model development ensuring the effectiveness of restoration approaches.