Heist et al : Mitochondrial DNA diversity and divergence among Rhizopnonodon terraenovae 



665 



because catch rates are assumed to be at or below 

 maximum sustainable yield and because life history 

 parameters predict a relatively high recruitment rate 

 for this species in comparison to the large coastal 

 species targeted by the U.S. Atlantic shark longline 

 fishery (NMFS 3 ). Recently Cortes ( 1995) challenged 

 this assessment on the basis of a reevaluation of life 

 history characteristics relevant to recruitment (age 

 at maturity, fecundity, and longevity) and suggested 

 that the Atlantic sharpnose shark may be more vul- 

 nerable to overfishing than was previously assumed. 

 Proper management of this species requires not 

 only accurate estimates of standing stock and recruit- 

 ment values but also an understanding of the repro- 

 ductive population structure of the species. The ob- 

 servation that pups of this species are found both in 

 the South Atlantic Bight as well as in the Gulf of 

 Mexico, coupled with the small size and apparent 

 lack of significant longshore migration, suggests that 

 there may be isolated breeding populations of this 

 species. Information on stock structure of marine 

 fishes has traditionally relied on two approaches: tag 

 and recapture studies and analyses of genetic varia- 

 tion. Although considerable information has been ob- 

 tained on the movements of large sharks by means 

 of tagging (Casey and Kohler, 1990), these tagging 

 studies have generally neglected smaller species of 

 sharks. Furthermore, to our knowledge the genetic 

 structure of any population of small coastal shark 

 has not been investigated. This study tests the hy- 

 pothesis of genetic homogeneity in allele frequency 

 in Atlantic sharpnose sharks between the Gulf of 

 Mexico and Mid-Atlantic Bight by using restriction 

 fragment haplotypes of mitochondrial DNA(mtDNA). 



Materials and methods 



Atlantic sharpnose sharks were collected with re- 

 search longlines in the Mid- Atlantic Bight (n =23) as 

 part of an ongoing shark research program of the 

 Virginia Institute of Marine Science, from the recre- 

 ational fishery of southern Texas (n=21) and from 

 artisanal longline vessels from Veracruz, Mexico 

 (n=8) (Fig. 1). Heart tissue samples from Atlantic 

 sharpnose sharks caught in the Mid-Atlantic Bight 

 were placed into cryovials and stored under liquid 

 nitrogen in the field. In Texas and Mexico, whole 

 hearts were collected on wet ice and stored frozen at 

 (-20°C) until shipped to Virginia. All samples were 

 stored at -70°C. 



Mitochondrial DNA was isolated from tissue by 

 following the rapid isolation protocol of Chapman and 

 Powers ( 1983 ). Aliquots of mtDNA were digested with 

 ten restriction enzymes (Ava I, Ava II, Ban I, Bel I, 



Summer 

 1990 

 t 1991 

 1992 



Figure 1 



Locations and dates for collection of sharpnose shark, Rhuo- 

 prionodon terraenovae. 



BstE II, Dra I, Hind III, Hpa I, Sea I, and Xho I) by 

 following the manufacturers' instructions. Restric- 

 tion fragments were separated on 1.0% horizontal 

 agarose gels run at 2V/cm overnight, then transferred 

 after electrophoresis to a nylon membrane by means 

 of Southern transfer according to the protocols of 

 Sambrook et al. ( 1989). Filters were hybridized with 

 highly purified mtDNA from tiger shark, Galeocerdo 

 cuvier, nick-translated with biotin-14-dATP, and vi- 

 sualized with the BRL BlueGene Nonradioactive 

 Nucleic Acid Detection System. 



Fragment patterns were scored for each restric- 

 tion enzyme and each individual was assigned a com- 

 posite genotype based on the fragment patterns for 

 all enzymes (Tables 1 and 2). The nucleon (haplo- 

 type) diversity was calculated for each sample and 

 for the composite of all samples following Nei (1987). 

 Nucleotide sequence diversity was calculated follow- 

 ing the site approach of Nei and Miller (1990). Chi- 

 square significance of the difference in genotypic fre- 

 quencies between samples was computed by using 

 the randomization protocol of Roff and Bentzen 

 (1989). Nucleotide sequence diversities and diver- 

 gences were calculated by using the REAP statisti- 

 cal analysis package (McElroy et al., 1991). 



