Discrimination of some red sea coral reef species based on hyperspectral signature field data

Abstract

Hyperspectral techniques offer significant advantages in monitoring coral reefs by providing detailed spectral information that supports precise identification and analysis of coral species. This study utilized hyperspectral field data to assess coral biodiversity and condition in the Gemsa coastal region of the Egyptian Red Sea. Coral spectral reflectance is primarily influenced by pigments from photosynthetic symbionts within the host cnidarian, offering a basis for species discrimination. In this research, in-situ hyperspectral data were collected using an HR-512i spectroradiometer from five sites, documenting the spectral signatures of 29 coral species across 11 families, including six soft coral species from four families and 23 hard coral species from seven families. Over a 20-day field campaign, underwater surveys and direct spectral acquisitions were performed using advanced optical techniques. Spectral analysis focused on identifying species-specific signatures influenced by pigment and structural variations. Second derivative analysis was applied to enhance differentiation among species. Results revealed distinct spectral patterns for each coral species. For instance, Acropora species such as A. austera, A. cytherea, A. pectinatus, A. pharaonis, and A. spicifera exhibited unique signatures in the blue and green regions, indicating differences in pigment composition or morphology. Favia stelligera and Pocillopora verrucosa showed notable reflectance in the red region, while Montipora informis and Goniopora planulata demonstrated subtler variations. Significant spectral variability was observed in soft corals like Sarcophyton glaucum and S. trocheliophorum across all spectral regions. Dead corals displayed distinct spectral signatures compared to living ones, highlighting changes associated with tissue and pigment loss. These findings underscore the value of hyperspectral imaging for coral reef conservation, offering a robust, non-invasive method for species identification and ecosystem monitoring. The study contributes meaningful insights to marine remote sensing, enhancing our ability to assess coral biodiversity and health under changing environmental conditions.