Heist and Gold: Genetic identification of sfiarks 



59 



eleven species of carcharhiniform sharks. Each spe- 

 cies possessed a unique set of restriction fragments, 

 and even in the species where polymorphism was 

 detected, misidentification was not a problem. This 

 technique can be performed on small pieces of tissue 

 collected at dockside and stored indefinitely at room 

 temperature and moreover can distinguish between 

 species whose carcasses may be difficult to discrimi- 

 nate (Fig. 1). Similar DNA technology has been used 

 for identifying carcasses in teleosts (Chow et al., 1993; 

 Bartlett and Davidson, 1991 ), and for identifying plank- 

 tonic fish eggs (Daniel and Graves, 1993) — cases where 



Figure 1 



Ten-percent polyacrylamide gel of Alul digests in three species that 

 produce very similar carcasses. Lane zero is a size standard, lanes one 

 to three are sandbar sharks, lanes four to six are bignose sharks, and 

 lanes seven to nine are dusky sharks. Numbers at left refer to sizes 

 (base pairs) of size standards. Predicted sizes of fragments in each spe- 

 cies are given in Table .3. 



morphological characters proved insufficient for spe- 

 cies level identification . 



Use of genetic characters to identify species can 

 be complicated by population structure. Although 

 most marine fishes that are distributed across vast 

 geographic stretches are open-ocean pelagics (Briggs, 

 1960), with presumed minor genetic differences 

 across the range of a species (but see Crosetti et al., 

 1994; Graves et al., 1992), many large coastal sharks 

 are distributed as multiple discrete populations 

 (Compagno, 1984). It is thus important to know 

 whether genetic differences among individuals are 

 diagnostic of species or populations. Some 

 questions, among others, are the follow- 

 ing: to what degree are published genetic 

 data useful in regions other than those 

 from which the original specimens were 

 collected, and can this method be used to 

 identify populations as well as species? 

 Baker et al. ( 1996), for example, were able 

 to determine geographic origin of marine 

 mammals they identified from flesh 

 samples on the basis of genetic data. In 

 our study, regional differences in restric- 

 tion patterns were observed in only one 

 species, spinner shark, whereas sequence 

 differences that did not affect restriction 

 patterns were observed in sandbar shark. 

 In four species ( dusky, silky, scalloped ham- 

 merhead, and tiger), sequences from one 

 individual accurately predicted restriction 

 sites in specimens collected thousands of 

 kilometers from where the original speci- 

 men was collected. 



The low level of intra- and interregional 

 polymorphism within species is not sur- 

 prising given the low genetic diversity typi- 

 cally reported for sharks (Smith, 1986). In 

 restriction fragment length polymorphism 

 (RFLP) studies of whole mtDNA molecules. 

 Heist et al. (1995, 1996) found a very low 

 nucleotide sequence diversity of 0.036% in 

 sandbar shark and 0.13% in Atlantic 

 sharpnose shark (Rhizoprionodon terraen- 

 ovae). The small numbers of substitutions 

 between geographically distant popula- 

 tions of sandbar and spinner shark ob- 

 served in this study indicate that intraspe- 

 cific diversity should not hinder species 

 identification of these eleven species. 



The erroneous tiger shark sequence in 

 the study of Martin and Palumbi (1993) 

 indicates a further strength of the PCR- 

 RFLP technique beyond use for forensic 

 identification; it can be used to evaluate 



