Coral Genus Differentiation Based on Direct Analysis in Real Time-High Resolution Mass Spectrometry-Derived Chemical Fingerprints.

Abstract

Coral reefs are one of the most biologically diverse ecosystems, and the accurate identification of the species is essential for diversity assessment and conservation. Current genus determination approaches are time-consuming and resource-intensive and can be highly subjective. To explore the hypothesis that the small-molecule profiles of coral are genus-specific and can be used as a rapid tool to catalogue and distinguish between coral genera, the small-molecule chemical fingerprints of the species Acanthastrea echinata, Catalaphyllia jardinei, Duncanopsammia axifuga, Echinopora lamellosa, Euphyllia divisa, Euphyllia paraancora, Euphyllia paradivisa, Galaxea fascicularis, Herpolitha limax, Montipora confusa, Monitpora digitata, Montipora setosa, Pachyseris rugosa, Pavona cactus, Plerogyra sinuosa, Pocillopora acuta, Seriatopora hystrix, Sinularia dura, Turbinaria peltata, Turbinaria reniformis, Xenia elongata, and Xenia umbellata were generated using direct analysis in real time-high resolution mass spectrometry (DART-HRMS). It is demonstrated here that the mass spectrum-derived small-molecule profiles for coral of different genera are distinct. Multivariate statistical analysis processing of the DART-HRMS data enabled rapid genus-level differentiation based on the chemical composition of the coral. Coral samples were analyzed with no sample preparation required, making the approach rapid and efficient. The resulting spectra were subjected to kernel discriminant analysis (KDA), which furnished accurate genus differentiation of the coral. Leave-one-out cross-validation (LOOCV) was carried out to determine the classification accuracy of each model and confirm that this approach can be used for coral genus attribution with prediction accuracies ranging from 86.67 to 97.33%. The advantages and application of the statistical analysis to DART-HRMS-derived coral chemical signatures for genus-level differentiation are discussed.