How does marker choice affect your diet analysis: comparing genetic markers and digestion levels for diet metabarcoding of tropical-reef piscivores

<p>Tropical reefs are highly diverse ecosystems, and reliable biomonitoring, through diet metabarcoding, is needed to understand present and future trophic relationships in this changing habitat. Several studies have assessed the reliability and effectiveness of single molecular markers; however, a cross-marker validation has rarely been performed. This study identified crucial properties for <i>12S</i> rDNA, <i>16S</i> rDNA and <i>COI</i> metabarcoding in tropical-reef piscivores (<i>Plectropomus</i> spp.). In addition, three new versatile primer sets for <i>16S</i> were designed <i>in silico</i> for metabarcoding of reef fish. Results showed that <i>COI</i> was overall better at recovering true diversity because of a well-supported database. Second, optimal <i>16S</i> amplicon sizes ranged between 160 and 440 base pairs for full diversity recovery, with increased species detection for the 270-base pairs region. Finally, blocking of predator-specific <i>COI</i> sequences was not equally effective in all host species, potentially introducing bias when diet compositions are directly compared. In conclusion, this novel study showed that marker success for prey identification is highly dependent on the reference database, taxonomic scope, DNA quality, amplicon length and sequencing platform. Results suggest that <i>COI</i>, complemented with <i>16S</i>, yields the best outcome for diet metabarcoding in reef piscivores. Findings in this paper are relevant to other piscivores and other metabarcoding applications.</p>