522 



Fishery Bulletin 103(3) 



including a more comprehensive geographical survey 

 of these four species to confirm that the genetic differ- 

 ences observed are diagnostic. However, it is clear that 

 DNA sequence-based approaches appear more powerful 

 in discriminating closely related species pairs and less 

 likely to produce false positives than other DNA-based 

 assays. 



Although it was not our intent to conduct an ex- 

 haustive analysis of higher-order relationships among 

 western North Atlantic shark species, some interesting 

 results nonetheless deserve mention. First, the orders 

 Carcharhiniformes and Lamniformes were strongly 

 supported as monophyletic, as were the families Sphy- 

 rnidae, Triakidae, and Lamnidae that were included 

 in the study. The order Hexanchiformes was likewise 

 monophyletic, but bootstrap support for this grouping 

 was low. The family Carcharhinidae was poorly sup- 

 ported as monophyletic, which is consistent with previ- 

 ous studies (Nalyor, 1992; Nelson, 1994; Musick et al., 

 2004). Interestingly, our phylogenetic hypotheses place 

 the family Triakidae basal to all other families within 

 the Carcharhiniformes, following Compagno (1988), 

 but this position was not strongly supported and is 

 predicated on limited sampling of Carcharhiniform 

 familes (only four of eight were included in our analy- 

 sis). Clearly this gene region contains some phylogenet- 

 ically useful information regarding shark relationships, 

 confined principally to higher-level groupings. 



We are careful in judging the utility of a locus for 

 species identification on the basis of phylogenetic sig- 

 nal alone. Clearly, rapidly evolving molecular mark- 

 ers are valuable tools for species identification but 

 might not be appropriate for reconstructing phylo- 

 genetic relationships at certain scales. Conversely, 

 those regions containing sufficient signal to generate 

 reasonable phylogenetic reconstructions (i.e., general 

 concordance with accepted phylogenetic relationships 

 based on other independent characters) must be useful 

 (and appropriate) markers for species identification. 

 Further, these regions are amenable to the addition 

 of uncharacterized species and the inclusion of in- 

 traspecific diversity (e.g., diverged mtDNA lineages 

 within species). Importantly, however, DNA sequence- 

 based approaches offer the potential to assign at least 

 some level of taxonomic characterization to unknown 

 or unrepresented samples. Although the use of DNA 

 sequencing has historically been viewed as cost pro- 

 hibitive, the genomic revolution over recent years has 

 spawned cost-effective sequencing services, making 

 routine sequencing of samples for species identification 

 not only practical but optimal. 



The size of the amplification product in the present 

 study might place limitations on the application of this 

 method to the poor-quality tissue and DNA often en- 

 countered in forensic studies. It has been our experience 

 that the primers used in our study consistently have 

 generated strong amplification products with DNA iso- 

 lated from a variety of tissue types, including dried tis- 

 sue and fins; however, we have yet to explore the range 

 of amplifications possible using tissues more commonly 



encountered in forensic cases. To circumvent potential 

 problems with large amplifications on degraded DNA 

 samples, we have constructed a preliminary, search- 

 able DNA-sequence database using the FASTA program 

 (Univ. Virginia, Charlottesville, VA; Pearson, 1999) 

 and the 12s-16s sequences presented in the present 

 study. Our preliminary analyses indicate that all spe- 

 cies examined in the study can be uniquely identified 

 from approximately 400 bp of sequence generated by 

 the 12SA-5' primer. We are examining the limitations 

 of sequence length in combination with the search ac- 

 curacy of this informative fragment. 



We are mindful of the restriction placed on these 

 analyses due to limited within-taxon sampling (par- 

 ticularly within-family) and of the incomplete represen- 

 tation (notably the Pristophoriformes) of all orders of 

 sharks and are aware that the phylogenetic affinities 

 presented in this study could change with the addition 

 of characters and taxa. These caveats notwithstand- 

 ing, we believe that a taxonomically restricted DNA 

 sequence database offers certain advantages over per- 

 haps more rapid RFLP or multiplex PCR assays. DNA 

 databases 1) can be "curated" (additions and access to 

 the database can be selective) and distributed as an 

 alignment suitable for further subsequent statistical 

 or phylogenetic manipulation; 2) can be easily amended 

 to include additional taxa, genetic variation within 

 species, and additional gene loci more appropriate at 

 various taxonomic scales; 3) allow for unequivocal as- 

 signment (subject to limits of discrimination of those 

 loci included) of species identification while making 

 available the raw data necessary for the development 

 of more rapid assays (RFLP/Multiplex PCR) for select 

 taxa (note that the opposite is not necessarily true); 

 and, 4) facilitate the identification of those taxa not 

 currently represented in the database through phylo- 

 genetic analysis. 



In summary, we have found that the sequence of 

 the 12S-16S region of the mtDNA that we examined 

 contains ample information for discriminating between 

 the shark species studied and shows promise for the 

 placement of species not yet examined within the cor- 

 rect phylogenetic group (family). We are continuing 

 to examine geographic variation within and among 

 species and to assay genetic variability at nuclear loci 

 in an effort to resolve potential introgression and (or) 

 hybridization events. As information is added to our 

 database, either in the form of additional species or 

 loci, our species identification method will become more 

 robust. 



Acknowledgments 



Much of this work derived directly from forensic case 

 work conducted by Ann Colbert for the National Marine 

 Fisheries Service. Robert Chapman provided primer 

 sets and guidance. Laura Webster conducted initial sur- 

 veys of shark mtDNA variability and assisted in sample 

 acquisition. Shannon Leonard and David Carter assisted 



