Assessing microplastics risk in marine habitats of Haizhou Bay based on an integrated approach

Current risk assessments often overlook the spatial overlap between microplastics (MPs) exposure and species-specific habitat distributions, which may bias ecological risk estimates. To address this gap, we developed an integrated approach combining MPs distribution maps with habitat suitability models for ten key fishery species in Haizhou Bay, China. Traditional risk quotients (RQ) were compared with a probabilistic method based on joint probability curves (JPC) to estimate overall risk probability (ORP). Mean MPs abundances were 2.34 ± 2.32 items/L in spring and 1.89 ± 2.38 items/L in autumn, with the highest concentrations found in the southern nearshore area. Random Forest identified dissolved iron, temperature, and velocity as the main factors influencing MPs distribution in both seasons. RQ-based assessment classified over 90% of Haizhou Bay as potential risk zones, and more than half of the suitable habitat for all ten species overlapped with these areas. By contrast, ORP values from JPC analysis remained approximately 1.0% across the entire bay and within individual habitats. This divergence suggests that concentrations above safety thresholds do not necessarily imply high ecological risk. By integrating spatial predictions, habitat modeling, and probabilistic risk estimation, the proposed framework can identify priority management areas and quantify overall risk probability and is adaptable to other coastal systems facing MPs pollution.