Heist et al.: Mitochondrial DNA diversity and divergence among Rhizoprionodon terraenovae 



667 



sharks from the Gulf of Mexico to the Mid-Atlantic 

 Bight to prevent significant divergence in mitochon- 

 drial DNA haplotypes. The frequencies of the most 

 common alleles, as well as the occurrences of rare 

 alleles, were nearly identical in each of the three 

 samples. The nucleon and nucleotide sequence di- 

 versities were also similar among samples. The hy- 

 pothesis that Atlantic sharpnose sharks collected 

 from locations as distant as Veracruz and Virginia 

 were members of a single homogeneous gene pool 

 could not be rejected. 



The distribution of mtDNA haplotypes has been 

 used previously to infer patterns of gene flow in sev- 

 eral other commercially important shark species in 

 the Gulf of Mexico and southeastern (U.S.) Atlantic 

 coast (reviewed in Avise, 1992). Gene flow in fishes 

 is accommodated both by the active movement of ju- 

 veniles and adults, as well as by the passive move- 

 ment of eggs and larvae. Significant differences in 

 mtDNA haplotype frequencies have been detected in 

 species with limited adult migration and demersal 

 eggs (marine toadfishes, Opsanus spp. ; Avise et al., 

 1987) as well as in those with pelagic eggs and lar- 

 vae (black sea bass, Centropristis striata; Bowen and 

 Avise, 1990; and menhaden, Brevoortia spp. , Bowen 

 and Avise, 1990). 



Sharpnose sharks are nektonic from the moment 

 of parturition; therefore, gene flow is accommodated 

 only by the active movements of juveniles and adults. 

 Significant differences in haplotype frequencies were 

 detected in redfish {Scianops ocellatus; Gold et al., 

 1993) of the Gulf of Mexico and southeast U.S. At- 

 lantic coast but not in the hardhead catfish (Arius 

 felis; Avise et al., 1987), two species with large active 

 adults but with presumably little passive transport 

 of eggs and larvae. In addition, Heist et al. (1995) 

 detected no heterogeneity in mtDNA haplotype fre- 

 quencies in the sandbar shark, Carcharhinus 

 plumbeus, over the same geographic range. 



In menhaden, the presence of two groups of ge- 

 netically divergent mtDNA haplotypes in the Atlan- 

 tic was interpreted as indicating complete isolation 

 of these two groups in the past, followed by a mix- 

 ture of stocks with divergent mtDNA haplotypes 

 (Bowen and Avise, 1990). The close relationships 

 among all haplotypes detected in the Atlantic 

 sharpnose shark is consistent with the hypothesis 

 of a single evolutionary lineage with no historical 

 subdivision. 



The lack of genetic divergence among Atlantic and 

 Gulf of Mexico sharpnose sharks can not prove that 

 separate stocks do not exist. An exchange rate of a 

 small number (<20) of females per generation be- 

 tween isolated breeding populations is enough to 

 prevent drift from establishing significant heteroge- 



neity in allele frequencies (Allendorf and Phelps, 

 1981). Therefore fishery-relevant stocks can be main- 

 tained in the absence of statistically significant ge- 

 netic divergence. Furthermore, if a single population 

 has recently diverged into multiple stocks, there may 

 not have been sufficient time for a significant level 

 of genetic divergence to have become established. 

 Perhaps by examining genetic characters that evolve 

 more rapidly than whole mitochondrial DNA (such 

 as direct sequencing of the mitochondrial control re- 

 gion or microsatellite analysis) stock structure may 

 be eventually detected in this species. The only way, 

 however, to determine the current level of gene flow 

 in this species may be through a tag and recapture 

 program. This information is necessary to determine 

 whether regional exploitation of this species will be 

 compensated by immigration from other regions and 

 whether regional (state) regulations will be an effec- 

 tive means of conservation. 



In order to perform robust tests of hypotheses con- 

 cerning gene flow in organisms, the markers used 

 must have sufficient intraspecific variation so that 

 differences in the frequencies of alleles can be as- 

 sessed between regions. The level of intraspecific 

 variation in the Atlantic sharpnose shark (NSD= 

 0. 13%) is considerably higher than the NSD of 0.036% 

 reported by Heist et al. ( 1995) in the sandbar shark, 

 Carcharhinus plumbeus, although lower than that 

 detected by Heist et al. (1996) in the shortfin mako, 

 Isurus oxyrinchus (NSD=0.38%<). Although the num- 

 ber of individuals surveyed in this study is small, 

 the similar amount of variation detected within each 

 sample and the close agreement in frequencies be- 

 tween regions strongly suggest mtDNA haplotype 

 homogeneity between sharpnose sharks of the Mid- 

 Atlantic Bight and Gulf of Mexico. This study has 

 demonstrated that this small coastal shark, with no 

 passive larval transport, nevertheless exhibits 

 mtDNA haplotype homogeneity across a broad geo- 

 graphic range. 



Acknowledgments 



The authors wish to thank Shelton Applegate, 

 Fernando Farias, and Leonardo Castillo for help in 

 obtaining samples from Mexico. Rocky Ward, Randy 

 Blankenship, and Ivonne Blandon provided samples 

 from Texas, and Captain J. A. "Tony" Penello and 

 Steve Branstetter provided assistance in Virginia. 

 This project was supported by the Wallop Breaux pro- 

 gram of the U.S. Department of Interior, U.S. Fish 

 and Wildlife Service, with additional funds distrib- 

 uted by the Virginia Marine Resources Commission 

 of the Commonwealth of Virginia. 



