The Qigu Lagoon is protected by the last of seven remaining sandbars along Taiwan’s southern coast. Once about 150 km², now its area has shrunk to 15 km². Rising sea levels are eroding sandbars, pushing them deeper into an already shrinking lagoon. The sediment flows into the lagoon from the adjacent Zengwun River. The lagoon is vital to sustaining local aquaculture and oyster farming, forming the backbone of the region’s economy. Government interventions have largely focused on the edge, trying to reinforce the sand barrier. This research fundamentally inquires how coastal defence strategies and oyster farming can be integrated to protect the sandbar, rehabilitate the shallowing lagoon, and sustain its changing economy. The research is structured in two phases. The First Phase was developed along the eroding seaward edge of the sandbar and was organised around three research verticals: land-growing, material study, and spatial structure. Drawing on island accretion strategies, this phase developed sediment-deposition prediction models to support the organic regrowth of the depleting sandbar. Discarded oyster shells were processed into material systems to test sediment interaction and structural potential. Through the logic of topological interlocking, oyster-shell-based modules evolved into funicular shell systems and residential units that form a seasonal settlement for fisher people embedded within the regenerated sandbar.

The Second Phase turns the focus inward toward the lagoon, and more critically, toward its bed. Progressive siltation has shallowed the lagoon, forcing oyster racks to cluster within limited deep zones and intensifying ecological stress. This phase undertakes a detailed analysis of water flow patterns, velocities and bathymetric relationships to understand the consequences of sediment imbalance in the lagoon. The earlier research into predictive land accretion models is expanded and inverted to enable controlled land-scouring. This is coupled with machine-learned, fluid-dynamics-informed simulations to test targeted underwater-erosion modules, introducing a performative infrastructural system that recalibrates underwater depths while supporting aquaculture, access and movement above water. A reconfigurable architectural spine emerges as both an ecological regulator and a social framework for work and seasonal use, dynamically adapting to varying requirements across the farming cycles, typhoon seasons and eco-tourism peaks, allowing spatial layouts to follow shifting depths. Together, the two phases reposition architecture as a mediator across land modulation, material behaviour, water movement and spatial opportunities. Rather than resisting change, the project proposes a system that operates through the shifting conditions of the lagoon, aligning ecological processes with spatial occupation to support a more resilient coastal future.