Balancing cage aquaculture and macroalgal cultivation: Scale-dependent management references for eutrophication mitigation in Sansha Bay, China

Cage aquaculture can increase nutrient loading and exacerbate eutrophication in weakly flushed, semi-enclosed bays, whereas macroalgal cultivation is often expected to enhance nutrient uptake and dissolved oxygen (DO). However, scale-dependent management benchmarks for coordinating these aquaculture types remain poorly constrained. In Sansha Bay, China, we combined seasonal monitoring (2016–2020) with Sentinel-2 mapping to quantify cage-aquaculture and macroalgal cultivation areas within 500-, 1000-, and 1500-m buffers around monitoring stations. Linear mixed-effects models, with season as a random effect and salinity as a proxy for mixing, were used to relate aquaculture area to DO, dissolved inorganic nitrogen (DIN), and soluble reactive phosphorus (SRP). Water quality exhibited a pronounced estuarine–marine gradient. Within 500 m, cage-aquaculture area was positively associated with DIN and SRP, whereas macroalgal area was positively associated with DO. At the 1500 m scale, cage-aquaculture area was negatively associated with DIN and SRP, and macroalgal area was negatively associated with DIN; these larger-scale negative associations likely reflect spatial covariation along the estuarine–marine gradient rather than direct mitigation effects. Under the fitted model, maintaining at least 2.12 km² of macroalgal cultivation within a 1500 m buffer was associated with model-predicted DIN below 0.40 mg·L⁻¹ , consistent with the Class III limit in China’s Seawater Quality Standard (GB 3097–1997). Under a simple proportional extrapolation to the bay’s 406 km² open-water area, this local threshold would imply a heuristic bay-wide reference of approximately 122 km² of macroalgal cultivation; this value should not be interpreted as a uniform target because practical implementation would need to account for aquaculture spatial planning, competing space uses, and the distribution of hydrodynamically suitable zones. In contrast, reducing cage-aquaculture area alone was unlikely to lower model-predicted SRP to ≤ 0.03 mg·L⁻¹ . These results provide a scale-explicit reference for aquaculture zoning and integrated multi-trophic aquaculture planning in semi-enclosed coastal bays.