Marine landscape mapping for ecologically relevant habitat suitability modelling

• Published in: EARTH SCIENCE INFORMATICS
• Main Authors: Yong, Bryan; Ismail, Khaira; Ali, Azizi; Hasan, Rozaimi Che; Muhamad, Muhammad Abdul Hakim
• Format: Article
• Language: English
• Published: SPRINGER HEIDELBERG 2025
• Subjects: Computer Science; Geology
• Online Access: https://www-webofscience-com.uitm.idm.oclc.org/wos/woscc/full-record/WOS:001400109000002

Integrating marine landscape classification with Maximum Entropy (MaxEnt) habitat suitability modelling can potentially develop thematic marine habitat models but has yet to be extensively studied. For Marine Protected Area (MPA) spatial planning, the application of such frameworks remains limited, especially in the Coral Triangle region. The absence of a standardised marine habitat mapping framework in this biodiverse area hampers ecosystem-based management for its transboundary MPA networks. This study aims to create a thematic map of potential coral reef habitats in Taman Laut Tioman, Malaysia, combining oceanic data and multiscale high-resolution multibeam echosounder bathymetry and backscatter data for marine landscape classification. Clustered environmental inputs and presence-only data from field surveys and citizen science were then applied in MaxEnt modelling. Despite limitations in spatial resolution, oceanic data-hydrodynamics and ocean colour satellite imagery-notably enhanced the model performance, showcasing the value of these variables even in smaller study areas. The final model identified four distinct marine landscape classes, providing a detailed abiotic profile that surpasses the continuous data controls and establishes a vital baseline for assessing ecological boundaries, serving as a precursor to biodiversity mapping and informing ecosystem-based management. This study provides important insight for developing a standardised framework to establish a transboundary network of MPAs in the Coral Triangle region to conserve its marine biodiversity.