Something fishy

Of the many victories wrought by DNA barcoding - the ability to place an unknown sample in a phylogenetic, and often taxonomic, context using short fragments of DNA sequence data - some of the most useful applications for management have come from the sea. One of the best citation-to-data ratios in this regard belongs to a 2004 study by Peter Marko. This project extended naturally from Marko's molecular ecology course: students purchased samples of "red snapper" from various fish markets, and sequence data from the mitochondrial cytochrome b gene region showed that most of these specimens were not, in fact, Lutjanus campechinus - they were often understudied and probably rare relatives, or in some cases not snapper at all. The conservation implications from this study were huge, and a number of papers have followed suit, looking at a variety of similar systems. If you aren't interested in adding to your list of papers to read, check out the short film based on work done in Steve Palumbi's lab that documents their work to identify shark fins.

In a paper by Lowenstein et al., published last week in PLoS ONE, the focus was on sushi, specifically tuna. The common labeling errors were caught again: there were mismatches between what restaurants called the fish, and what was actually being offered. In some cases, very phylo-distant species are being sold as "tuna", and these species can actually make consumers ill. The story is an interesting one of how fraud develops in samples of organisms that can no longer be visually tied to the species they came from, and the difficulties in protecting consumers from fraud under current regulations. Obviously, the perils of overfishing are becoming quite clear and interested readers should carry their Monterey Bay Aquarium seafood guides or similar (there's even an iPhone app for that!) with them before ordering at restaurants.

A particularly interesting advance in this study was taking the barcoding approach beyond the visual appeal of tree-building and similarity with databased sequence data. One concern about barcoding has been that even when a new clade appears in a phylogeny, taxonomy cannot be updated without some sort of diagnostic characters. It is uncommon for new species (especially of animals) to be described based on DNA sequence data alone, but it is nevertheless the norm to define the character states that uniquely define a species from its relatives. In Lowenstein et al.'s paper, they identified 14 diagnostic DNA substitutions that could be used to uniquely identify all species of Thunnus and suggested that focusing on particular characters within the "barcode gene" (mitochondrial cytochrome oxidase I) will also be necessary for new technologies to accelerate in-the-field identification.

This latter step is of interest for anybody interested in cryptic species, or identifications when other reference material is not available. I hassled one of my former Ph.D. students endlessly as she revised her dissertation because we had been comfortably using a phylogenetic tree to assign unknown individuals to one of three cryptic taxa (in the isopod genus Idotea), but prior to publication we knew that diagnostic characters would be necessary for subsequent work to be readily comparable. And, since the undergraduate evolution lab at the University of Georgia repeats Marko's work on red snapper every few years (the local Kroger now knows not to advertise their special on "red snapper from Indonesia"), perhaps the lab can be extended by having the students generate these characters for the genus Lutjanus as well. I don't seem to have any problem convincing students to do their homework when it involves going out for sushi.