Understanding and predicting the causes and the consequences of seagrass fragmentation in the lagoonal environment of Reunion Island
Researchers
Research Areas
Project Brief
Seagrasses are vital ecosystem engineers supporting biodiversity, coastal protection, and carbon trapping. However, they face significant global decline, particularly in tropical regions with high anthropogenic pressures. On Reunion Island, once-resilient seagrass meadows have rapidly declined since 2017. Thus, understanding fragmentation processes and resilience thresholds in this vulnerable ecosystem is critical for biodiversity conservation, climate adaptation, and sustainable management.
Overall, seagrass ecosystems are complex systems comprising multiple interacting components, including biological, growth, reproductive, environmental, hydrodynamic, and sediment effects, which together can produce emergent and difficult-to-predict behaviour under uncertainty. Ultimately, the research project will aim at integrating these factors into a comprehensive numerical model, in order to study and understand the resilience of seagrass ecosystems under various environmental conditions, with a focus on the region of Reunion Island.
Both applied and methodological approaches will be used to better understand fragmentation processes and their uncertainties in real-world systems. Since probabilistic models of ecosystem resilience using Dynamic Bayesian Networks (DBNs) have been successfully applied to time anthropogenic activities in order to maximize ecosystem resilience, this will be the primary modelling approach used in the research project. These DBNs have been extended and integrated with downsampled regional climate models in a whole-of-systems approach, which has proven useful in predicting resilience impacts under different climate scenarios, focusing on heatwaves and temperature (Hatum et al., 2024; Wu et al., 2017).
When applied to Reunion Island’s case study, a similar modelling procedure will benefit from existing cover maps for the past 20 years and will incorporate a hydrodynamic model to reconstruct past physical forcings. The results from this model configuration and calibration will be validated using data collected during a field campaign planned for 2025.