Abstract
Introduction Coastal ecosystems are vital blue carbon sinks that are increasingly threatened by climate change. Their vulnerability and sensitivity are strongly influenced by ecological structures and local conditions. Methods Using the Ecopath model, we evaluated the responses of three bays in Fujian, China (Sansha Bay, Fuqing Bay, and Xinghua Bay) under scenarios of increased precipitation, ocean deoxygenation, and warming. The analysis focused on how differences in food web structures shape ecosystem responses to climate pressures and determine their sensitivity and vulnerability. Results The results revealed differentiated response patterns dictated by baseline food web characteristics, including constant, linear, and non-linear threshold collapse, in which the baseline ecological structure of a bay dictates its degree of vulnerability. Fuqing Bay, despite having the lowest total system throughput (2405 t/km²/year), showed the highest resilience and bivalve ecological carrying capacity (22.80 t/km²). In contrast, Xinghua Bay, a high biomass system (Total Biomass: 39.93 t/km²), exhibited the highest sensitivity, with its food web structure collapsing even under low deoxygenation stress (shrimp EE > 1). Under severe warming, bivalve ecological carrying capacity declined linearly by up to 50% across all the bays, with absolute losses being the greatest in the most productive systems. Discussion Our findings underscore the critical role of baseline ecosystem structure in shaping divergent climate responses and provide a scientific basis for site-specific adaptive management and blue carbon conservation strategies.