Evaluation of variation within the barcode region of Cytochrome c Oxidase I (COI) for the detection of commercial Callinectes sapidus Rathbun, 1896 (blue crab) products of non-US origin

Callinectes sapidus Rathbun, 1896 is a western Atlantic species with a disjointed natural geographic range from Massachusetts, USA to Venezuela (distribution area 1) and from Alagoas, Brazil to northern Argentina (distribution area 2). It is the only species of portunid crab commercially harvested in the continental United States but is also imported into the US from several Latin American countries, Venezuela and Mexico in particular. In the United States, crab products labeled as “blue crab” and “Product of the USA” may not legally contain other species of crab or C. sapidus not harvested in the United States. The present study documents nucleotide variation within the barcode region of cytochrome c oxidase I (COI) in 417 reference specimens of C. sapidus collected from throughout its natural range. The goal of this study is to determine if this variation can be utilized to detect mislabeled C. sapidus products sold in interstate commerce by comparing genetic signatures in reference specimens to those observed in commercial crabmeat labeled as “Product of the USA” and “Product of Venezuela.” In reference specimens, we observed high levels of genetic variation in the barcode region. However, three lineages were consistently observed with significant pairwise F st values between the lineages. Lineage 1 was observed throughout the natural geographic range but predominated in the continental US and was the only lineage observed in the major crabmeat-producing states (MD, LA, VA, NC). Lineage 2 primarily occurred in the Caribbean region of distribution area 1 but was also infrequently encountered in the South Atlantic Bight region of the US coast. Finally, Lineage 3 was only observed in Brazilian waters and had the lowest haplotype and nucleotide diversity values. Lineages 1 and 2 were separated by a mean pairwise distance ( p -distance) of 3.15%, whereas Lineage 3 had a mean p -distance of 2.55% and 1.35% to Lineages 1 and 2, respectively. Within lineage mean p -distances were 0.45%, 0.19%, and 0.07% for Lineages 1, 2, and 3, respectively. Among all vouchered reference specimens collected from the continental United States, Mexico, Puerto Rico, and Venezuela, we identified 22 phylogenetically informative sites that drive observed lineage divergences. Haplotypes identified from barcode COI sequences from commercial C. sapidus products labeled as originating from the US all aligned with haplotypes from Lineage 1 reference specimens and haplotypes from commerci