DNA Barcodes Reveal Species Identities of Larval Fishes

Genetic identification of species, using DNA sequences, has great promise for solving certain problems in fisheries management. This area of research focuses on finding DNA sequences that are unique for individual species, a "genetic barcode". Genetic barcoding can be used to identify the species of tiny fish eggs and larvae, as well as seafood products such as fish fillets, targets that were difficult or impossible to identify previously. The national Fish Barcode of Life (FishBOL) project has the goal of developing genetic barcodes for as many species as possible and building a depository of the information in a Barcode of Life Database (BOLD). Scientists can use the database to match DNA sequences from unidentified specimens to sequences of species already known.

Using a plankton net, University of Hawaii students collect fish larvae for a study comparing genetic 
            barcoding and traditional morphological identification techniques. Photo by Dr. Shai Sharif, UH 
            postdoctoral fellow.
Using a plankton net, University of Hawaii students collect fish larvae for a study comparing genetic barcoding and traditional morphological identification techniques. Photo by Dr. Shai Sharif, UH postdoctoral fellow.

DNA barcoding is based on the sequencing of a standard 650 base pair fragment of a particular mitochondrial DNA (mtDNA) gene, cytochrome c oxidase I (COI). The technique utilizes the unique mtDNA COI sequences that occur at the species level.

PIFSC scientists in the Life History Program, part of the Fishery Biology and Stock Assessment Division, have been collaborating with students in a University of Hawai'i at Mānoa Marine Ecology and Evolution class to investigate the utility of genetic barcodes for larval fish identification in Hawai`i. The class, taught by Dr. Anuschka Faucci, has collected larval fish samples near O'ahu yearly, beginning in 2007. PIFSC scientists used traditional morphological techniques to identify the larvae collected. The class then extracted DNA from the larvae, obtained COI sequences, and used BOLD to obtain genetic barcode identifications of the larvae.

The results so far have been informative in different ways. Many of the larvae, such as those of the slender ocean sunfish (Ranzania laevis), are identified as the same species by traditional morphological methods and genetic barcoding. This indicates the reliability of both techniques for those species. Other larvae, such as those of tunas in the genera Auxis and Thunnus cannot be reliably identified at small sizes by their morphology, but can be identified to species at all stages of development using the DNA barcode technique.

Despite the promise of genetic barcoding, further work is needed to improve reliability of the approach. One problem is an incomplete database. Some larvae, particularly those of species found only in Hawai'i, are identified by BOLD as closely related species not found here. However, further examination of BOLD reveals that the DNA barcodes for these Hawaiian endemic species are not yet available in the database. The finding of such gaps in the DNA barcoding database indicates the need to barcode these endemic species based on morphologically identifiable adult specimens. An example is the ebony blenny (Enchelyurus brunneolus), which has a unique larva in Hawai'i, but whose larvae were identified in BOLD searches as a South Pacific relative, the black blenny (E. ater). In some instances, highly distinctive larvae, such as those of the lanternfish Benthosema fibulatum, were identified in BOLD as different genera, indicating large gaps in the species and genera included in BOLD.

The specificity of genetic barcoding has helped to show where problems exist with traditional morphological techniques, and where more careful research is needed. For example, small larvae of jacks identified by morphological methods as 'opelu (Decapterus species) were found by genetic barcode identification to be a mixture of larval 'opelu and ulua (Caranx species).

The goal of PIFSC research with DNA barcode techniques is to identify larval and juvenile stages of the "Deep 7" bottomfish management unit species, including specimens archived at the Center. Better identifications will improve information on age and growth, pelagic larval duration, distributions, settlement, and habitat requirements of the species. Currently, early life history stages of most deepwater snapper species are very difficult to identify reliably. Eggs and larvae of ehu (Etelis carbunculus) and onaga (Etelis coruscans) can only be identified to genus using morphology, but not to species. The use of genetic barcoding will help us confidently link confirmed species identifications to life history data for these early stages. Better identifications will help the Life History Program acquire more biological information relevant to fisheries management. As an example, accurate modeling of early bottomfish dispersal will require better estimates of pelagic duration prior to settlement. Accurate species identification of pelagic juveniles and recently settled juveniles will enable researchers to identify a "settlement mark" on the fish's otolith that will allow the retrospective determination of pelagic duration from the otoliths of successfully recruited bottomfishes.

A 3.9 mm long larva of ehu (Etelis carbunculus) or onaga (E. coruscans) collected off windward 
            O'ahu. Small larvae of Etelis can be identified to genus using morphological characters, but 
            identification to the species level will require genetic barcoding. A 3.5 mm long carangid larva initially identified as 'opelu or mackerel scad (Decapterus 
            macarellus) but more likely an ulua or trevally jack (Caranx species). Genetic barcode identifications show 
            that a large percentage of small jack larvae are misidentified using morphological characters, with many 
            ulua larvae misidentified as 'opelu. This indicates the need for more careful research on this group.
Left: A 3.9 mm long larva of ehu (Etelis carbunculus) or onaga (E. coruscans) collected off windward O'ahu. Small larvae of Etelis can be identified to genus using morphological characters, but identification to the species level will require genetic barcoding.
Right: A 3.5 mm long carangid larva initially identified as 'opelu or mackerel scad (Decapterus macarellus) but more likely an ulua or trevally jack (Caranx species). Genetic barcode identifications show that a large percentage of small jack larvae are misidentified using morphological characters, with many ulua larvae misidentified as 'opelu. This indicates the need for more careful research on this group.