What question did this study set out to answer?

The aim is to evaluate microplastic contamination and its ecological risks in coastal sediments of Cox's Bazar.

March 14, 2026Open Access

A novel polymer-sensitive index coupled with multivariate and machine learning modeling for microplastic risk assessment in coastal sediments of the bay of Bengal

Key Points

The aim is to evaluate microplastic contamination and its ecological risks in coastal sediments of Cox's Bazar.
Conducted polymer analysis to identify microplastic types and sources.
Developed the Sediment Polymer Hazard Index (SPHI) and Microplastic Pollution Risk Index (MPRI).
Utilized multivariate analyses to assess microplastic composition based on site use.
Applied machine learning models, particularly Random Forest, to classify contamination risks.
Tourism areas showed significantly higher microplastic abundance with up to 111 items kg^-1.
Polyethylene and polypropylene were the most common polymers found.
Risk indices indicated tourism hotspots as high-risk zones due to toxic polymer contributions.
Machine learning classifications indicated polymer-specific compositions predict high-risk sites more accurately than abundance alone.

Abstract

Microplastic (MP) contamination in coastal sediments poses growing ecological and human health concerns, yet data for developing nations remain limited. This study provides a comprehensive assessment of MPs along the Cox's Bazar shoreline, the world's longest natural sea beach and a rapidly expanding tourism hub in Bangladesh. Tourism-dominated beaches showed significantly higher abundances (up to 111 items kg-1 dw) compared to rural low-use sites, with fibres and fragments representing the dominant morphotypes. Polymer analysis identified polyethylene (PE) and polypropylene (PP) as the most common constituents, reflecting consumer packaging waste and fishing-related debris as major sources. Risk evaluation demonstrated that abundance alone underestimates potential ecological hazard. Novel, hazard-weighted indices developed in this study, the Sediment Polymer Hazard Index (SPHI) and Microplastic Pollution Risk Index (MPRI), identified tourism hotspots as high-risk zones due to elevated contributions from toxic polymers (e.g., PS, PET) and ingestion-prone particle characteristics. Multivariate analyses further indicated that site-use category significantly constrain MP composition, confirming the influence of direct human pressure. Machine learning models, applied to classify MP contamination in coastal sediments, demonstrated that polymer-specific composition outperforms total abundance in predicting high-risk sites, with Random Forest achieving the highest classification accuracy. These results highlight the need for targeted coastal management prioritizing tourism-intensive zones, improved waste handling, and sustainable fishing practices. Integrating hazard-based indices and advanced predictive tools into long-term monitoring frameworks will be essential to protect the ecological and socioeconomic value of Cox's Bazar as coastal development accelerates.

A novel polymer-sensitive index coupled with multivariate and machine learning modeling for microplastic risk assessment in coastal sediments of the bay of Bengal

Key Points

Abstract

Cite This Study

Also Consider

Also Consider