Advertisement

Marine Biology

, Volume 148, Issue 5, pp 1143–1155 | Cite as

Genetic evidence of cryptic speciation within hammerhead sharks (Genus Sphyrna)

  • J. M. QuattroEmail author
  • D. S. Stoner
  • W. B. Driggers
  • C. A. Anderson
  • K. A. Priede
  • E. C. Hoppmann
  • N. H. Campbell
  • K. M. Duncan
  • J. M. Grady
Research Article

Abstract

Surveys of genetic variation within cosmopolitan marine species often uncover deep divergences, indicating historical separation and potentially cryptic speciation. Based on broad geographic (coastal eastern North America, Gulf of Mexico, western Africa, Australia, and Hawaii) and temporal sampling (1991–2003), mitochondrial (control region [CR] and cytochrome oxidase I [COI]) and nuclear gene (lactate dehydrogenase A intron 6 [LDHA6]) variation among 76 individuals was used to test for cryptic speciation in the scalloped hammerhead, Sphyrna lewini (Griffith and Smith). CR and COI gene trees confirmed previous evidence of divergence between Atlantic and Indo-Pacific scalloped hammerhead populations; populations were reciprocally monophyletic. However, the between-basin divergence recorded in the mtDNA genome was not reflected in nuclear gene phylogenies; alleles for LDHA6 were shared between ocean basins, and Atlantic and Indo-Pacific populations were not reciprocally monophyletic. Unexpectedly, CR, COI, and LDHA6 gene trees recovered a deep phylogenetic partition within the Atlantic samples. For mtDNA haplotypes, which segregated by basin, average genetic distances were higher among Atlantic haplotypes (CR: DHKY=0.036, COI: DGTR=0.016) than among Indo-Pacific haplotypes (CR: DHKY=0.010, COI: DGTR=0.006) and approximated divergences between basins for CR (DHKY=0.036 within Atlantic; DHKY=0.042 between basins). Vertebral counts for eight specimens representing divergent lineages from the western north Atlantic were consistent with the genetic data. Coexistence of discrete lineages in the Atlantic, complete disequilibrium between nuclear and mitochondrial alleles within lineages and concordant partitions in genetic and morphological characters indicates reproductive isolation and thus the occurrence of a cryptic species of scalloped hammerhead in the western north Atlantic. Effective management of large coastal shark species should incorporate this important discovery and the inference from sampling that the cryptic scalloped hammerhead is less abundant than S. lewini, making it potentially more susceptible to fishery pressure.

Keywords

Cryptic Species Western North Vertebral Count Hammerhead Shark Cryptic Lineage 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

Application of molecular techniques to assay genetic variation within cosmopolitan marine species has revealed evidence of deep genetic partitions that suggests cryptic speciation within many taxa, including foraminiferans (de Vargas et al. 1999), cnidarians (Dawson and Jacobs 2001), crustaceans (Williams et al. 2001), copepods (Schizas et al. 1999; Lee 2000), gastropods (Etter et al. 1999; Quattro et al. 2001a), bony fishes (Colborn et al. 2001; Borsa 2002), birds (Friesen et al.1996), and mammals (Garcia-Rodrigues et al. 1998; Dalebout et al. 2002). Recent studies extend this trend to elasmobranchs, a relatively unstudied component of marine ecosystems. Newly discovered species of hound shark (Mustelus; Last and Stevens 1994; Heemstra 1997; Gardner and Ward 2002) and thresher shark (Alopias; Eitner 1995) were first recognized from studies of genetic variation. Unfortunately, too few comprehensive population genetic surveys have been completed to determine the extent of cryptic speciation among cosmopolitan elasmobranch species.

We used broad geographic and genetic sampling to investigate the possibility of cryptic evolutionary lineages among hammerhead sharks in the genus Sphyrna Rafinesque, a group of eight widely distributed species (Gilbert 1967; Compagno 1984). The scalloped hammerhead (S. lewini Griffith and Smith) was the focal species because of morphological and mtDNA indications of a partition between Atlantic and Indo-Pacific populations. Springer (1941) detected sufficient morphological divergence between basins to recognize Atlantic (S. diplana) and Indo-Pacific (S. lewini) scalloped hammerhead species. However, these two forms were synonymized after broader geographic representation and larger samples sizes indicated that diagnostic characters were distributed across basins (Fraser-Brunner 1950; Gilbert 1967). In contrast, the first application of molecular techniques to evaluate genetic variation among hammerhead sharks, Martin’s (1992) RFLP analysis of the mtDNA control region, indicated deep inter-basin divergence within S. lewini. Although based on two individuals per basin, subsequent analyses of sequence variation in the mitochondrial cytochrome b (Cytb) and cytochrome oxidase (COI) genes (Martin 1993) supported the partition.

Additional incentives to survey scalloped hammerheads for cryptic species included commercial and recreational fishing pressure and use of Sphyrna lewini as ‘utilized bycatch’ (i.e., ‘finning’; Bonfil 1997; Kotas 2002). Sphyrna lewini is an abundant coastal shark and, consequently, an important element of commercial fisheries worldwide. Also, scalloped hammerhead nursery grounds in shallow coastal bays and inlets are subject to high levels of commercial and recreational fisheries activity, and neonates and juveniles constitute a significant proportion of shark landings (e.g., Bonfil 1997; Kotas 2002). With a low intrinsic rate of increase (Smith et al. 1998) and increasing fishing pressure, prudent management of this fishery is warranted, especially if cryptic species are included.

To test for cryptic speciation in S. lewini, variation within two evolutionarily independent genomes (mitochondrial and nuclear loci) was assayed to reduce potential effects of bias in individual data sets. Specifically, taxonomic interpretations of mtDNA variation and gene trees are generally concordant for deeply divergent taxa (Weins and Penkrot 2002), such as bonefish (Colborn et al. 2001) and pygmy sunfishes (Quattro et al. 2001b), but can be misleading taxonomically at lower divergence levels, particularly for allopatric populations (e.g., Weins and Penkrot 2002). For example, mtDNA variation in Carcharodon carcharius is significantly structured between South Africa and Australia/New Zealand and could be interpreted as evidence for allopatric speciation. However, nuclear markers (microsatellite loci) did not support this fundamental genetic break, and the mtDNA data were interpreted as reflecting male-biased dispersal and female philopatry (Pardini et al. 2001). This dichotomy in mtDNA and nuclear inferences for C. carcharius emphasizes the importance of using independently evolving markers to test phylogeographic hypotheses (Slade et al. 1994; Quattro et al. 2001b; Weins and Penkrot 2002). Concordance among independent markers is strong support for species hypotheses, and a two-pronged strategy of mitochondrial and nuclear gene assessments is a valuable tool for detecting cryptic species (Avise and Ball 1990; Sites and Crandall 1997; Grady and Quattro 1999; Weins and Penkrot 2002).

Samples of S. lewini and appropriate outgroups (Table 1) were first characterized for mitochondrial CR variation. To test the CR interpretation of scalloped hammerhead evolution within and across basins and to interpret mtDNA variation in a broader phylogenetic and taxonomic context, select samples were screened for variation in the COI subunit I locus of the mitochondrial genome. To expand taxonomic representation and eliminate the possibility of misidentified individuals, COI data for S. lewini were combined with Martin’s (1993) COI data set (available at http://spot.colorado.edu/~am/Cyb.COI.Data.1), which included sequences for all members of the genus Sphyrna, except S. zygaena, and several outgroups (Eusphyra blochii, Negaprion brevirostris, and Prionace glauca). The data set was expanded to include S. zygaena and a broader geographic representation of S. lewini. To test phylogenetic patterns in mtDNA, samples were then screened for variation in a nuclear gene, the sixth intron of the muscle-type lactate dehydrogenase-A locus (Stoner et al. 2003).
Table 1

Sphyrna lewini. Sampling locations, collection date, sex, fork length (FL), and haplotypes of specimens

Region Location

Haplotype (Allele)

Date

Sex

FL

CR

COI

LDHA6

Clade

Western Atlantic (WA)

Oregon Inlet, North Carolina

May 1990

120

1

 

2

Atlantic

May 1990

129

1

6

2

Atlantic

June 1990

169

1

7

2

Atlantic

June 1990

153

1

6

3

Atlantic

June 1990

198

1

6

2

Atlantic

June 1990

169

1

6

2

Atlantic

June 1990

170

1

 

2

Atlantic

June 1990

136

1

6

2

Atlantic

Folly River, North Carolina

30 Aug 2003

N/A

70

1

 

2

Atlantic

30 Aug 2003

N/A

36

5

 

6

Cryptic

Cape Romaine, South Carolina

12 July 1994

33.5

5

 

6

Cryptic

Bulls Bay, South Carolina

14 May 2002

34.8

5

 

6

Atlantic (189)

11 July 2002

N/A

N/A

5

 

6

Cryptic (173)

19 July 2001

N/A

N/A

5

 

6

Cryptic (179)

19 July 2001

N/A

N/A

5

 

6

Cryptic

Coastal South Carolina

27 July 1994

37

5

 

6

Cryptic

27 July 1994

35

5

8

6

Cryptic

27 July 1994

35

5

8

6

Cryptic

14 Aug 1995

103.8

5

8

6

Cryptic

29 Sept 1999

54.3

5

8

6

Cryptic

29 Sept 1999

54

6

8

6

Cryptic

29 Sept 1999

43

5

 

6

Cryptic

St. Helena Sound, South Carolina

26 Aug 2002

55.8

5

 

6

Cryptic

30 Aug 2002

50.3

5

 

6

Cryptic (162)

30 Aug 2002

41.8

1

 

1

Atlantic

30 Aug 2002

52.6

1

 

1

Atlantic (195)

30 Aug 2002

46.0

1

 

2

Atlantic

30 Aug 2002

52.0

1

 

2

Atlantic (199)

30 Aug 2002

54.0

1

 

2

Atlantic (191)

3 Sep 2002

43.8

5

 

6

Cryptic

3 Sep 2002

42.7

5

 

6

Cryptic (171)

3 Sep 2002

48.9

1

 

1

Atlantic

St. Augustine, Florida

12 July 1995

N/A

N/A

1

 

1

Atlantic

27 Feb 1995

42

1

 

2

Atlantic

Cocoa Beach, Florida

2 May 1995

62

1

 

1

Atlantic

2 May 1995

N/A

N/A

1

6

2

Atlantic

Fort Lauderdale, Florida

N/A

N/A

N/A

5

 

6

Cryptic

Gulf of Mexico (GM)

Panama City, Florida

12 June 2003

39

1

 

1

Atlantic

12 June 2003

36

1

 

2

Atlantic

12 June 2003

37

1

 

2

Atlantic

12 June 2003

38

1

 

2

Atlantic

12 June 2003

79

1

 

1

Atlantic

10 June 2003

39

1

 

2

Atlantic

10 June 2003

46

1

 

2

Atlantic

11 Sept 2003

48

1

 

1

Atlantic

11 Sep 2003

49

1

 

2

Atlantic

7 Oct 2003

53

1

 

2

Atlantic

23 Oct 2003

53

1

 

2

Atlantic

30 Oct 2003

31

1

 

2

Atlantic

S of New Orleans, Louisiana

5 Aug 1995

187

1

6

2

Atlantic

8 Aug 2001

N/A

N/A

1

 

4

Atlantic

9 Aug 2001

81

1

6

2

Atlantic

12 Aug 2001

87

1

6

1

Atlantic

13 Aug 2001

70

1

 

2

Atlantic

Southeastern Atlantic (AF)

Abidjam, Ivory Coast

14 Oct 1999

N/A

169

1

6

1

Atlantic

14 Oct 1999

N/A

163

1

6

1

Atlantic

14 Oct 1999

N/A

182

1

6

1

Atlantic

14 Oct 1999

N/A

172

1

 

2

Atlantic

14 Oct 1999

N/A

157

1

6

2

Atlantic

14 Oct 1999

N/A

195

1

6

1

Atlantic

14 Oct 1999

N/A

157

1

 

2

Atlantic

14 Oct 1999

N/A

163

1

 

2

Atlantic

14 Oct 1999

N/A

166

1

6

2

Atlantic

Hawaii (HW)

Kaneohe Bay, Hawaii

20 Feb 200

55.5

4

 

1

Indo-Pacific

20 Feb 200

57.1

4

2

2

Indo-Pacific

20 Feb 200

59.9

4

2

1

Indo-Pacific

20 Feb 200

68.5

4

 

2

Indo-Pacific

29 Mar 2000

51.7

4

 

2

Indo-Pacific

29 Mar 2000

47.8

4

 

2

Indo-Pacific

29 Mar 2000

49.1

4

 

2

Indo-Pacific

26 June 2000

N/A

N/A

4

2

2

Indo-Pacific

26 June 2000

N/A

4

4

2

Indo-Pacific

Australia (AU)

NW Australia

5 Sept 2001

129

2

2

1

Indo-Pacific

11 Sept 2001

125

3

2

1

Indo-Pacific

13 Sept 2001

121

2

3

5

Indo-Pacific

13 Sept 2001

161

2

3

5

Indo-Pacific

Acronyms for regions are used in other tables, figures, and text. Two CO1 hapotypes (1 and 5) were obtained from Martin (1993). For clarity and consistency with the text, specimens are identified to one of three phylogenetically divergent clades (Cryptic, Atlantic, and Indo-Pacific) recovered in the CR and COI trees (Fig. 1). Numbers in parentheses after ‘Clade’ designation are total vertebral counts for select individuals; see text for detail

N/A-data not available

Finally, as a comparison to the genetic data, a small subset of the S. lewini samples was evaluated for variation in total vertebrae. The impetus for this portion of the study was Gilbert’s (1967) analysis of morphological variation in S. lewini in which he reported wide variation in vertebral counts for nine scalloped hammerhead specimens. Total vertebrae ranged from 174 to 204, but variation was considerably narrower across eight of the nine specimens, with counts ranging from 192 to 204. As noted by Gilbert (1967), the broader range was due to substantially fewer total vertebrae (174) in one specimen from the western north Atlantic, notably coastal South Carolina.

Materials and methods

Specimens, DNA extraction, PCR amplification, sequencing

Blood, fin clip, muscle, or liver tissue was obtained from specimens taken during 1991–2003 collections and identified in the field as scalloped hammerhead sharks, S. lewini (Griffith and Smith). Collections were made primarily by the Marine Resources Division, South Carolina Department of Natural Resources Marine Forensics Branch, Center for Coastal Ecosystem Health and Biomolecular Research, National Ocean Service, Charleston, South Carolina, the National Marine Fisheries Service, North Carolina Division of Marine Fisheries, and local fishermen. Sample distribution included the western north Atlantic (coastal North Carolina to Florida), the Gulf of Mexico (western Florida to Louisiana), southeastern Atlantic (coastal Africa), central Pacific (Hawaii), western Pacific (eastern Australia), and eastern Indian Ocean (northwestern Australia) (Table 1). Tissues were also obtained from specimens of S. zygaena taken in the southern Atlantic (coastal Africa). Tissues were stored frozen or in 70% ethanol until total nucleic acids were extracted with QiAmp tissue extraction columns, following the manufacturer’s (Qiagen) protocol. Total nucleic acids were isolated from blood through standard phenol–chloroform extraction (Hillis et al. 1996). Tissue samples from all individuals sequenced and specimens used for vertebral counts were stored at the University of South Carolina (available through JMQ).

The complete mitochondrial control region (~1100 bp) was amplified from genomic DNA extracts, using the primers ElasmoCR15642F (5′ - TTG GCT CCC AAA GCC AAR ATT CTG - 3’) and ElasmoCR16638R (5′ - CCC TCG TTT TWG GGG TTT TTC GAG - 3′) designed by Stoner et al. (2003). Amplifications were conducted in 50-μl volumes, which included ~10 ng of total DNA, 10 mM Tris (pH 8.3), 2.5 mM MgCl2, 50 mM KCl, 0.01% Triton X-100, 10 pmol of each primer, 200 μM each dNTP, and 2 U of Taq DNA polymerase. PCR conditions were: 4 min at 94˚C; 40 cycles of 94°C for 1 min, 58°C for 1 min, 72°C for 2 min, and a final extension of 7 min at 72°C. A 750-bp portion of the- COI subunit I gene was amplified with the universal primers CO1e and CO1f (Palumbi 1996) and reaction ingredients and conditions described above for the CR. The sixth intron (~200 bp) of the LDHA locus (LDHA6) was amplified by hemi-nested PCR described by Stoner et al. (2003). The first round of PCR used primers ElasmoLDHA6F1 (5′ - GCT TAT GAR GTG ATW AAA CTG AA - 3′) and ElasmoLDHAR1 (5′ - GAA RAC CTC RTT YTY WAT RCC ATA - 3′) and the reaction mixture described above, with an annealing temperature of 52°C. The second PCR included a 2-μl aliquot of the first PCR product, primers ElasmoLDHA6F2 (5′ - GGG WTG TCT GTG GCA GAC CTC GC - 3′) and ElasmoLDHAR1, and the reaction mixture and cycling parameters described for the first PCR (Stoner et al. 2003).

Amplification products were sequenced on an ABI 377 automated sequencer. Approximately 460 and 400 bases, respectively, of the COI and CR amplicons were characterized in the forward direction. For LDHA6 products, heterozygotes were diagnosed as individuals with two equally intense peaks at single base positions on chromatograms. Suspected heterozygotes were rare; however, when encountered, both strands were sequenced to confirm peak intensity at individual positions and subsequently compared to homozygotes to infer the phase of mutations.

Sequence and phylogenetic analyses

Chromatograms were edited and aligned in Sequencher (version 4.1; Gene Codes Corp., Inc.) or BioEdit (version 5.09; T. Hall, North Carolina State University). Sequences for each gene were sufficiently homologous for ingroup and outgroup taxa to be aligned by eye. Aligned COI sequences were checked for correct reading frame by translation to amino acid sequence in Sequencher and BioEdit. McClade (Maddison and Maddison 1992) was used to determine the number of alleles/haplotypes and to identify and remove repeated occurrences of each in the COI, CR, and LDHA6 sequence datasets, each consisting of sequences for 76 individuals. Genetic distance calculations and phylogenetic analyses were conducted in PAUP* (version 4.0b10, D. L. Swofford, Florida State University). Data files are available from the corresponding author.

Parsimony trees were reconstructed using the exhaustive search (CR and LDHA6) or branch-and-bound (COI) algorithm. For parsimony analyses, COI, CR, and LDHA6 characters were weighted equally. Additional COI parsimony analyses weighted characters according to an empirically derived 10:1 transition to transversion (ti:tv) ratio. Bootstrap resampling (1,000 pseudoreplicates) was used to test support for hypothesized relationships (Felsenstein 1985).

Likelihood ratio tests (Goldman 1993), as implemented in MODELTEST (version 3.0; Posada and Crandall 1998), were used in conjunction with PAUP* to select sequence evolution models appropriate to each sequence data set. Maximum likelihood trees were reconstructed using the heuristic search routine, including likelihood bootstrapping (100 pseudoreplicates).

A partition homogeneity test (Farris et al. 1994, 1995) was conducted in PAUP* to assess congruence of phylogenetic signal across a combined CR/LDHA6 data set. Invariant positions and gaps were excluded for these analyses. Tree lengths for 1,000 random partitions of the combined data set were not significantly different from trees generated from the CR and LDHA6 components individually (P=0.28), and the data were pooled for subsequent analyses. Unweighted parsimony analyses on the pooled data set used the exhaustive search strategy implemented in PAUP*. Maximum likelihood reconstructions used a sequence evolution model obtained from MODELTEST (Posada and Crandall 1998). Bootstrapping (Felsenstein 1985) was used to estimate the reliability of parsimony and likelihood reconstructions (1,000 and 100 pseudoreplicates, respectively).

Vertebral counts

Eight juvenile scalloped hammerheads, three from Bulls Bay and five from St. Helena Sound, SC (Table 1), were characterized for pre-caudal and total vertebrae by X-radiography using a Summit generator (62 mA at a voltage of 2.5 kV). Specimens were positioned so that the left lateral surface of the body was perpendicular to the tube. The object-film distance was increased to facilitate magnification of the image and aid accuracy of vertebral counts. Prior to exposure a dissection pin was placed through the precaudal pit of each specimen to ensure a consistent point of reference for all counts. To be consistent with Gilbert (1967), only counts for total vertebrae are reported. Total number of vertebrae was counted on each radiograph twice, once each by independent researchers; if a difference between the two counts was observed, a third count was conducted.

Results

Sequence characteristics and variation

Sequence variation in mitochondrial and nuclear genes was assayed in 76 S. lewini, representing samples from all major ocean basins (Table 1). Both mitochondrial genes were considerably more variable (proportion of variable and phylogenetically informative sites) than the nuclear gene (Table 2). Variation in a 408-bp fragment of the CR defined six haplotypes in S. lewini (Table 3; GenBank accessions DQ168917–DQ168922), which were segregated by ocean basin (three each from the Atlantic and Indo-Pacific), two for S. tiburo (GenBank accession DQ168923–DQ168924), and one for S. mokarran (GenBank accession DQ168925). The COI data set encompassed 461 bp (Table 2) for the species included in Martin (1993) plus 29 additional S. lewini and three S. zygaena. The 11 taxa included in the COI dataset were represented by 19 haplotypes, including eight from the additional S. lewini samples (Table 4 and GeneBank accessions DQ68934–DQ168941) and one for S. zygaena (GenBank accession DQ168942). Variation among COI haplotypes was distributed across codon positions but was most common at third positions (10 first position changes, 1 second position, and 100 third positions). Like the CR variants, COI haplotypes recovered from scalloped hammerhead samples were segregated by ocean basin, with four haplotypes unique to each basin (Table 4).
Table 2

Sphyrna lewini. Mitochondrial and nuclear gene variation in Sphyrna lewini, including specimens of the Cryptic lineage. See Table 1 and text for explanation

Category

Locus

Length (bp)

% A+T

Ti/Tv

Variable sites (proportion)

Phylogenetically informative

Mitochondrial

Control Region

408

71.2

0.99

94 (23%)

60 (14.7%)

Cytochrome Oxidase - I

461

61.9

9.63

111 (24.1%)

85 (18.4%)

Nuclear Intron

LDHA6

171

61.5

1.07

9 (5.3%)

4 (2.3%)

Table 3

Sphyrna lewini. Characterization, geographic distribution, and abundance (numbers in table) of mtDNA control region (CR) haplotypes. Only variable sites (28) and haplotypes for S. lewini (haplotypes 1–4) and the Cryptic lineage (haplotypes 5 and 6) are shown

Distribution

Haplotype

Variable site

Atlantic

Indo-Pacific

WA

GM

AF

AU

HW

 

111111122222222222333333

     

4455234447822466666888155669

     

8912260145218002568348408167

     

1

TATAAATAATTTCGCA-CCCTACAACGA

19

17

9

  

2

.....C..TA.A.AT.T.T.C.T....G

   

3

 

3

.....C..TA.ATATGT..TC.......

   

1

 

4

.....C..TA.A.ATGT..TC.......

    

9

5

ATGTCCATCACATAT.CT...T.TGTA.

17

    

6

ATGTCCATCACA.AT.CT...T.TGTA.

1

    

Geographic codes are as follows: WA western Atlantic, GM Gulf of Mexico, AF western Africa, AU Australia, and HW Hawaii

Table 4

Sphyrna lewini. Characterization, geographic distribution, and abundance (numbers in table) of COI haplotypes. Only variable sites (26) and haplotypes for S. lewini (haplotypes 1–7) and the cryptic lineage (haplotype 8) are illustrated. Geographic codes follow Table 3, with the addition of PN (Panama) and BJ (Baja California), which were included in Martin’s (1993) data set

Distribution

Haplotype

Position

Atlantic

Indo-Pacific

WA

GM

PN

AF

AU

BJ

HW

 

111112222223333333444

       

35559456770244682235699016

       

50392657393158616926208731

       

1

CCCACTCGTATTTTTCTTGAATTCTA

      

1

2

.................A.T.C...A

    

2

1

3

3

.................AAT.C...A

    

2

  

4

.................A.T.C...T

      

1

5

TTT.TCTAC.C..CC.CAAT.C...A

  

1

    

6

TTT.TCTAC.C..CC.CAAT.CC..A

6

3

 

6

   

7

TTT.TCTAC.C..CCTCAAT.CC..A

1

      

8

.T.GTC.A.GCCC.C.CAATTC.T.A

  

5

    
Taxa characterized for mitochondrial genes were subsequently examined for variation in a 171-bp portion of LDHA6. Six alleles (GenBank accessions DQ168926–DQ168931) were recovered from samples of S. lewini; three were common and distributed across ocean basins (Atlantic–Indo-Pacific), two were singletons from the Atlantic Ocean, and one was restricted to samples from the western Atlantic (Table 5). One LDHA6 allele was recovered for the two outgroup species, S. tiburo (GenBank accession DQ168933) and S. mokarran (GenBank accession DQ168932).
Table 5

Sphyrna lewini. Characterization, geographic distribution, and abundance (numbers in table) of alleles for the nuclear LDHA6 locus. Only variable sites (6) and alleles for S. lewini (allele 1–5) and the cryptic lineage (allele 6) are shown. Geographic codes follow Table 1

Distribution

Allele

Position

Atlantic

Indo-Pacific

WA

GM

AF

AU

HW

 

1

     

456671

     

980260

     

1

TGAGAA

5

4

4

2

2

2

....C.

13

12

5

7

 

3

..T.C.

1

    

4

.C.T..

 

1

   

5

.....-

   

2

 

6

C.....

18

    
Comparisons of genetic distance estimates within and between basins revealed trends that were consistent across mitochondrial and nuclear genes (Table 6). Average sequence divergence was lowest among Indo-Pacific haplotypes and highest between basins (Atlantic–Indo-Pacific). However, divergence among Atlantic haplotypes was substantially higher than within the Indo-Pacific, and comparable to between basin estimates for CR. This pattern reflected two suites of Atlantic haplotypes for each gene. One group was widely distributed within the basin, and a second was restricted to the northwestern Atlantic (Cryptic in Table 6), notably from coastal North Carolina, South Carolina, and Florida. Haplotypes within these groups differed minimally, but divergence between groups was substantial.
Table 6

Sphyrna lewini. Genetic distances within (diagonal elements) and between (off-diagonal elements) scalloped hammerhead lineages calculated as HKY corrected distances for CR, GTR corrected distances for COI, and F81 corrected distances for LDHA6. Lineage designations correspond to those in Table 1 and Fig. 1. Atlantic and Indo-Pacific locations were pooled for LDHA6 since only two well-supported lineages were recovered at this locus

Locus

Lineage

Genetic Distance

Cryptic

Atlantic

Indo-Pacific

CR

Cryptic

0.003

  

Atlantic

0.053

0.000

 

Indo-Pacific

0.051

0.025

0.010

COI

Cryptic

0.000

  

Atlantic

0.030

0.003

 

Indo-Pacific

0.036

0.032

0.005

LDHA6

Cryptic

0.000

  

Atlantic/Indo-Pacific

0.018

0.011

 

Haplotype divergence within the Atlantic was consistent across genes and across individuals. Specimens with the divergent northwestern Atlantic haplotype for CR also had the northwestern Atlantic COI and LDHA6 haplotypes.

Gene trees

Control region

Parsimony analysis of partial CR sequences recovered one shortest tree (L: 116, CI: 0.91, RI: 0.89; Fig. 1a) that recognized three lineages within a monophyletic S. lewini. One lineage included the divergent northwestern Atlantic haplotypes (referred to as the Cryptic lineage), the Indo-Pacific lineage encompassed haplotypes that are widely distributed but restricted to the Indo-Pacific basin, and the Atlantic lineage included CR haplotypes that were widely distributed in the Atlantic basin. CR sequences and bootstrap analysis supported early divergence of the Cryptic lineage, with a subsequent Atlantic–Indo-Pacific split. Based on the HKY85 model of sequence evolution (Hasegawa et al. 1985), empirically derived base frequencies (A: 0.372, C: 0.197, G: 0.072, T: 0.359), and a gamma distribution with a shape of α = 0.248, the best maximum likelihood reconstruction (−lnl = 1059.134) recovered relationships supported by parsimony analyses. Likelihood bootstrapping indicated support for basal isolation of the Cryptic lineage but did not distinguish the Atlantic–Indo-Pacific partition due to ambiguous placement of one Australian haplotype (Fig. 1a).
Fig. 1

Sphyrna lewini. a Phylogenetic relationships among mitochondrial control region and b cytochrome oxidase I haplotypes recovered from within S. lewini and among species of hammerhead sharks (Sphyrna). Parsimony reconstructions are depicted. Numbers near nodes are maximum parsimony/maximum likelihood bootstrap values; only nodes supported by >50% shown. Population codes for S. lewini haplotypes (in parentheses after terminal taxa) follow Table 1. CR lineages within S. lewini are designated Cryptic, Atlantic, and Indo-Pacific and noted in the COI tree. Haplotype numbers (Tables 3, 4) within S. lewini are indicated to the right of terminal nodes

Cytochrome oxidase-I

Unweighted and weighted parsimony analyses of partial COI sequences recovered one tree (L: 229, CI: 0.528, RI: 0.666; Fig. 1b) that differed from Martin’s (1993) combined COI and Cytochrome b phylogeny only in the placement of S. lewini. In the COI parsimony tree, S. lewini diverges deeper relative to Martin’s (1993) tree and is sister to a lineage consisting of S. corona ([S. tudes, S. media] S. tiburo). The COI parsimony analyses recognized three haplotype lineages within a monophyletic S. lewini that corresponded to the Cryptic, Atlantic, and Indo-Pacific lineages in the CR trees. However, relationships among these clades were not well resolved. Using a general time reversible model with rate heterogeneity (GTP+G; MODELTEST), empirically derived base frequencies (A: 0.273, C: 0.205, G: 0.170, T: 0.353), and a gamma distribution with a shape of α = 0.156, likelihood analyses recovered a single shortest tree (−lnl = 1,653.063) that recognized the Cryptic, Atlantic, and Indo-Pacific lineages. However, likelihood analyses did not recover a monophyletic S. lewini. The Cryptic and Indo-Pacific lineages were placed in a clade with four other hammerhead species, but relationships among these clades were unresolved (tree not shown). The Atlantic lineage was basal to other clades within S. lewini.

Lactate dehydrogenase-A intron six

The shortest tree (L: 9, CI: 1.000, RI: 1.000) recovered in parsimony analyses of the LDHA6 sequences recognized two clades within a monophyletic S. lewini (Fig. 2a). One corresponded to the Cryptic lineage in mitochondrial trees and included a single allele in 18 individuals. A second clade encompassed five alleles that were distributed across the Atlantic and Indo-Pacific basins. The best likelihood tree (−lnl = 293.430), reconstructed under the F81 (Felsenstein 1981) model with empirically derived nucleotide frequencies (A: 0.283, C: 0.151, G: 0.234, T: 0.332), repeated relationships portrayed in the parsimony tree (Fig. 2a). Likelihood and parsimony bootstrap analyses indicated moderate support for the CrypticAtlantic/Indo-Pacific partition, but relationships among haplotypes in the Atlantic/Indo-Pacific clade generally were not supported.
Fig. 2

Sphyrnalewini. a Phylogenetic relationships among alleles of Lactate Dehydrogenase-A Intron Six (LDHA6) and b as recovered from a combined analysis of mitochondrial CR and nuclear LDHA6 sequence data (2B). Parsimony reconstructions are depicted. CR and LDHA6 sequences were combined by individual. Numbers near nodes are maximum parsimony/maximum likelihood bootstrap values; only nodes supported by >50% shown. Population codes for S. lewini haplotypes (in parentheses after certain terminal taxa) follow Table 1. For illustration, CR lineages, Cryptic, Atlantic, Indo-Pacific, are applied to the trees. Allele/haplotype numbers indicated to the right of the terminal nodes

Combined control region and lactate dehydrogenase

A data set comprising mitochondrial and nuclear gene sequences was constructed by pairing observed control region and nuclear intron data for each individual. Unweighted parsimony analyses of the pooled data set recovered one tree (L: 126, CI: 0.910, RI: 0.910; Fig. 2b) that differentiated the three lineages recognized in all mitochondrial gene trees. Combining data sets produced trees that were better resolved and in which nodes received greater support than in single gene trees. Seven nodes in the combined tree were supported in 50% or more bootstrap replicates, compared to five and three nodes for the CR and LDHA6 gene trees, respectively. The Cryptic, Atlantic, and Indo-Pacific lineages in Sphyrna lewini also were strongly supported by bootstrapping. Likelihood analysis of the pooled data set recovered a single tree (-lnL = 1414.027) under the HKY+G model (Hasegawa et al. 1985); this tree supported monophyly of S. lewini and the Cryptic clade as basal to the AtlanticIndo-Pacific partition.

Vertebral variation

Eight juvenile S. lewini, three from Bulls Bay and five from St. Helena Sound, South Carolina, were characterized for total vertebrae and CR and LDHA6 haplotypes. Sharks from St. Helena Sound represented individuals within the Cryptic and Atlantic lineages that were captured in late August and early September 2002. Counts for total vertebrae segregated according to mitochondrial and nuclear haplotype lineages. Control region and LDHA6 haplotypes identified four each of the Atlantic and Cryptic clades among the specimens sampled for vertebrae. Specimens of the Cryptic lineage had significantly (two-tailed t test, P<0.002) fewer total vertebrae (mean/SD 171.3/7.0, range 162 to 179) than samples with Atlantic haplotypes (mean/SD 193.5/4.4, range 189 to 199).

Discussion and conclusions

Uplift of the Isthmus of Panama has partitioned variation in many marine, freshwater, and terrestrial species (reviewed by Bermingham et al. 1999), as first recognized by Jordan’s (1908) ‘law of geminate species’. The scalloped hammerhead, S. lewini, is circumtropical to subtropical; thus divergence between allopatric Atlantic and Indo-Pacific populations of S. lewini is predicted from geological history and was indicated in limited geographic sampling of the mtDNA genome (two specimens per basin; Martin1993). A primary concern was whether inter-basin divergence was sufficient to warrant taxonomic recognition, prompting our evaluation of evolutionarily independent markers. Mitochondrial and nuclear gene trees and allelic distributions support cryptic speciation among hammerhead sharks, but the event recorded in these genes was within the Atlantic basin rather than between basins as predicted by geological history and Martin’s (1993) preliminary data.

Genetic sampling recovered three mtDNA lineages among S. lewini samples. Two lineages correspond to the predicted divergence between Atlantic and Indo-Pacific populations. However, a third, deeper mtDNA lineage was recovered and was restricted in our sampling to the western north Atlantic (coastal North Carolina to Florida). The second Atlantic lineage was first recorded in CR sequences and gene trees but was confirmed with broader taxonomic sampling of a second mitochondrial gene, COI. Inclusion of all recognized hammerhead species in the COI data set confirmed the occurrence of two divergent lineages in the western Atlantic.

Interestingly, sequence divergence estimates for the mitochondrial genes provide different impressions of population subdivision and scalloped hammerhead lineage diversification. COI sequence divergence (GTR corrected) is comparable among the Atlantic, Cryptic, and Indo-Pacific scalloped hammerhead lineages, suggesting coincident separation (Table 6). Whereas, CR sequences record earlier isolation of the Cryptic lineage, based on 5.0% divergence (HKY corrected) relative to Atlantic and Indo-Pacific haplotypes, followed by a more recent separation of Atlantic and Indo-Pacific clades (2.5% divergence). Also, the lowest divergence estimate among lineages for COI is between the coexisting Atlantic haplotypes (Cryptic and Atlantic), which, curiously, record the highest CR sequence divergence. Potentially, these differences reflect variation in substitution patterns and saturation effects between coding (COI) and noncoding (CR regions) regions of the mtDNA genome.

Likelihood and parsimony reconstructions of mitochondrial gene trees generally recognize monophyly of the Atlantic, Cryptic, and Indo-Pacific lineages, but relationships among lineages vary across trees and bootstrap support for some nodes is weak. Assuming a roughly uniform mtDNA clock, disparity between divergence estimates for COI and CR haplotypes might reflect saturation effects. Variation among COI haplotypes in S. lewini was limited to third positions as anticipated of coding genes, whereas differences among CR haplotypes were distributed across the fragment. Similarly, base frequencies varied between genes (Table 2), with COI third positions having greater disparity in the proportion of purines. Divergent A and G frequencies might be contributing to saturation effects, even at low divergence levels (Kocher and Carleton 1999).

With all possible phylogenetic arrangements (three-taxon statements) portrayed in various mitochondrial gene trees, the origin and factors contributing to isolation and divergence of Sphyrna lewini lineages are unresolved. The (Cryptic (Atlantic + Indo-Pacific)) arrangement is more tenable than alternatives in terms of conventional biogeography and is consistent with a monophyletic S. lewini. Of course, any phylogenetic arrangement of scalloped hammerhead lineages requires sympatric divergence of the Atlantic and Cryptic lineages. More extensive sampling is required to evaluate alternative scenarios for speciation in S. lewini. Similarly, variable sequence divergence estimates for COI and CR haplotypes prevents application of a general mtDNA molecular clock to date separations and identify corresponding geological events and ecological factors.

Even with well-resolved mtDNA relationships, determining the taxonomic status of recently diverged mitochondrial lineages is problematic (Moritz 1994; Sites and Crandall 1997; Weins and Penkrot 2002). The Atlantic and Indo-Pacific clades of S. lewini have divergent COI and CR haplotypes, do not share haplotypes, and are exclusive (following Weins and Penkrot 2002). However, divergence is anticipated among allopatric populations. Conversely, the Atlantic and Cryptic hammerhead lineages are sympatric and divergent for mtDNA. While mtDNA divergence estimates and gene trees suggest speciation, co-occurrence of divergent haplotypes could reflect retention of ancestral polymorphisms (Campton et al. 2000).

Despite the power of mtDNA to recover population divergences, the strongest evidence of speciation is concordant partitioning of evolutionarily independent characters (Avise and Ball 1990; Sites and Crandall 1997; Grady and Quattro 1999; Weins and Penkrot 2002). Allelic distributions and trees for the nuclear encoded LDHA6 gene confirm the evolutionary independence of two scalloped hammerhead lineages recognized in the mtDNA data and trees. Individuals with mtDNA haplotypes corresponding to the Cryptic lineage were fixed for an LDHA6 allele that was not recovered from Indo-Pacific or other Atlantic samples. Absence of heterozygotes for LDHA6 and disequilibrium across mitochondrial and nuclear loci indicates that the sympatric Atlantic and Cryptic clades do not share a gene pool.

LDHA6 trees also support evolutionary independence of the Cryptic lineage. Likelihood and parsimony reconstructions for LDHA6 consistently recover basal divergence of the Cryptic allele relative to an Atlantic–Indo-Pacific assemblage. Parsimony and likelihood analyses on the pooled LDHA6 and CR data (Fig. 2b) yielded well-supported relationships among lineages within S. lewini. As in the individual datasets, three distinct phylogenetic partitions are apparent. However, unlike single gene analyses, partitions and inter-relationships in the combined tree received strong bootstrap support, particularly a monophyletic S. lewini and a sister group relationship between the Cryptic and an Atlantic + Indo-Pacific clade.

Like genetic data, variation in total vertebrae for western north Atlantic specimens of S. lewini is strongly partitioned, and the division is consistent with genetic lineages. Although the meristic data are preliminary due to small sample size and limited geographic representation, concordant morphological and genetic (mitochondrial and nuclear) variation strongly supports the occurrence of two scalloped hammerhead shark species in the Atlantic. Moreover, Gilbert’s (1967) inability to discriminate morphologically between lineages indicates that acquisition of reproductive isolation (as supported by genetic data) preceded morphological differentiation, i.e., speciation was cryptic. The sampling bias of this study, i.e., small samples largely from the western north Atlantic and primarily juveniles, precludes confident assessment of the geographic distribution of the cryptic hammerhead species. Available samples indicate that the cryptic species occurs in the northwestern Atlantic, specifically in coastal areas from North Carolina to Florida. Within this range, the two Atlantic scalloped hammerhead lineages occurred sympatrically in North and South Carolina, where they were taken syntopically, but other samples of the cryptic species also were taken within the range of S. lewini. An independent, global survey of S. lewini based on ITS2 variation recorded three specimens of the cryptic species, all from eastern coastal Florida (Abercrombie et al. in press). A phylogeographic assessment of S. lewini is underway (KMD) and might illuminate the geographic distribution of the cryptic species, particularly whether it is limited to the Atlantic.

In addition to the intrinsic significance of cryptic speciation among cartilaginous fishes and among cosmopolitan species, thorough geographic and genetic surveys of broadly distributed shark species are fundamental to comprehensive management plans. The rapid expansion of commercial shark fisheries in conjunction with marine habitat degradation has prompted international concern for the sustainability of these fisheries and persistence of target species (Compagno and Cook 1995; Walker 1998; Castro et al. 1999; Stevens et al. 2000). The undetected presence of cryptic species within commercially exploited sharks, such as the hound shark (genus Mustelus; Last and Stevens 1994; Heemstra 1997; Gardner and Ward 2002), points to a growing conservation crisis.

The scalloped hammerhead is a common and abundant element of the large coastal shark fishery, which is currently considered over-fished (NMFS 2001). The tendency for scalloped hammerheads to aggregate makes this species vulnerable to increasing fishing efficiency. For example, reports of scalloped hammerhead catches include estimates of nearly 35 tons taken in individual purse seine hauls in the northwestern Atlantic (Bonfil 1997). Limited catch data for the scalloped hammerhead indicate decreasing CPUE in the western Atlantic (Brown 1998; Cramer 1998) and substantial declines in many areas of the northwestern Atlantic (Baum et al. 2002). Recognition of two sympatric scalloped hammerheads species in the western Atlantic should prompt careful re-evaluation of the current management plan. Intense coastal fishing pressure on scalloped hammerheads places at least two species at greater risk for over fishing.

Data from this study and related efforts suggest a lower abundance of the cryptic species relative to its sister species S. lewini. Sampling for this study was more intense in coastal South Carolina, where 16 of 22 specimens were of the cryptic species. However, only two of the other 54 specimens screened for genetic variation and taken from across the range of S. lewini were identified as the cryptic species. Similarly, an independent assessment of genetic variation in S. lewini sampled more extensively than this study, yet only three specimens of the cryptic species, also from the western north Atlantic, were detected among 140 samples (Abercrombie et al. in press).

The apparent high relative abundance of the cryptic species in coastal South Carolina could be an artifact of sampling but also might highlight a conservation focus. Most specimens screened in this study were neonates to juveniles, including those from coastal South Carolina. High relative abundance of juveniles of the cryptic species in South Carolina estuaries and its rarity in other coastal areas (i.e., Gulf of Mexico) suggests that South Carolina bays are among the more important nursery grounds for the cryptic species. Protecting prime nursery habitat is vital to the persistence of the cryptic species, since species with narrow geographic distributions, overall or during critical life history stages, inherently are at higher risk of extinction. Concentrated reproduction in South Carolina costal waters also could increase the risk of extinction of the cryptic species. Population declines due to intense coastal fisheries could be exacerbated by a gender-biased harvest of the cryptic species as female density increases during the reproductive season. If South Carolina coastal waters are the primary nursery grounds for the cryptic species and females aggregate during reproductive season, these areas are conservation priorities. Data on the geographic distribution and relative abundance of both scalloped hammerhead species is critical at this juncture and should be used to evaluate current management plans.

Mitochondrial haplotype and nuclear allele distributions and phylogenies support the existence of two sympatric scalloped hammerhead species in the northwestern Atlantic. The diagnosis of a cryptic species of hammerhead shark is based on sympatric occurrence of two deep genetic lineages and complete disequilibrium between mitochondrial haplotypes and nuclear alleles within lineages. Coexistence of these lineages in the north Atlantic and the lack of genetic exchange inferred from disequilibrium strongly suggest independent gene pools and reproductive isolation, two properties of species integrity in some species concepts (e.g., see Mayr 1940; Dobzhansky 1950). Preliminary morphological data, notably vertebral counts, support these genetic findings.

Notes

Acknowledgements

This study was made possible through the generous contribution of hammerhead shark samples by C. Woodley, National Ocean Service, Charleston, SC; M. Grace, L. Jones, National Marine Fisheries Service, Pascagoula, MS Laboratory; J. Carlson, Southeast Fisheries Science Center, National Marine Fisheries Service, Panama City, FL; G. Ulrich, D. Oakley, South Carolina Department of Natural Resources, Marine Resources Division, Charleston, SC; E. Heist, Southern Illinois University, Carbondale, IL; F. Rohde, North Carolina Division of Marine Fisheries; R. McAuley, Department of Fisheries, Government of Western Australia; M. Shivji, Nova University, and B. Stequert, Centre ORSTOM de Brest. T. Burgess, K. Oswald, and M. Roberts provided invaluable assistance with laboratory and analytical techniques. Funding for this project was provided by the Cooperative Institute for Fisheries Molecular Biology (FISHTEC; NOAA/NMFS (RT/F-1)), the National Science Foundation (OCE-9814172), and SC SeaGrant (R/MT-5) to JMQ. Additional funding was provided by the Louisiana Board of Regents Support Fund, LEQSF ((2000-2001)-ENH-TR-67) and the Graduate School, College of Sciences, and Department of Biological Sciences, University of New Orleans to JMG.

References

  1. Abercrombie DL, Clarke SC, Shivji MS (2005) Global-scale genetic identification of hammerhead sharks: application to assessment of the international fin trade and law enforcement. Conserv Genet (in press)Google Scholar
  2. Avise JC, Ball RM (1990) Principles of genealogical concordance in species concepts and biological taxonomy. Oxford Surv Evol Biol 7:45–67Google Scholar
  3. Baum JK, Myers RA, Kehler D, Gerber L, Blanchard W, Harley SJ (2002) Preliminary standardized catch rates for pelagic and large coastal sharks from logbook and observer data from the Northwest Atlantic. Col Vol Sci Pap ICCAT 54:1294–1313Google Scholar
  4. Bermingham E, McCafferty SS, Martin AP (1999) Fish biogeography and molecular clocks: Perspectives from the Panamanian isthmus. In: Kocher TD, Stepien CA (eds) Molecular systematics of fishes. Academic, NewYork, pp 113–128Google Scholar
  5. Bonfil R (1997) Status of shark resources in the Southern Gulf of Mexico and Caribbean: implications for management. Fish Res Amsterdam, 29:101–117CrossRefGoogle Scholar
  6. Borsa P (2002) Allozyme, mitochondrial-DNA, and morphometric variability indicate cryptic species of anchovy (Engraulis encrasicolus). Biol J Linnean Soc 75:261–269Google Scholar
  7. Brown CA (1998) Standardized catch rates of four shark species in the Virginia–Massachusetts (U.S.) rod and reel fishery 1986–1997. SB-IV-5, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Southeast Fisheries Center, MiamiGoogle Scholar
  8. Campton DE, Bass AL, Chapman FA, Bowen BW (2000) Genetic distinction of pallid, shovelnose, and Alabama sturgeon: emerging species and the US Endangered Species Act. Conser Gen 1:17–32CrossRefGoogle Scholar
  9. Castro JI, Woodley CM, Brudek RL (1999) A preliminary evaluation of the status of shark species. FAO Tech Pap 0 (380): I–iv;1–72Google Scholar
  10. Colborn J, Crabtree RE, Shaklee JB, Pfeiler E, Bowen BW (2001) The evolutionary enigma of bonefishes (Albula spp.): cryptic species and ancient separations in a globally distributed shorefish. Evolution 55: 807–820CrossRefGoogle Scholar
  11. Compagno LJV (1984) FAO species catalogue, vol 4. Sharks of the world. An annotated and illustrated catalogue of shark species known to date. Part 1. Hexanchiformes to Lamniformes. United Nations Development Programme Food and Agriculture Organization of the United Nations, Rome, ItalyGoogle Scholar
  12. Compagno LJV, Cook SF (1995) The exploitation and conservation of freshwater elasmobranchs: status of taxa and prospects for the future. J Aquariculture Aquat Sci 7:62–90Google Scholar
  13. Cramer J (1998) Large pelagic logbook catch rates for sharks. SB-IV-II, Southeast Fisheries Science Center, National Marine Fisheries Service, MiamiGoogle Scholar
  14. Dalebout ML, Mead JG, Baker CS, Baker AN, van Heldene AL (2002) A new species of beaked whale Mesoplodon perrini sp. N. (Cetacea: Ziphiidae) through phylogenetic analysis of DNA sequences. Mar Mammal Sci 18:577–608CrossRefGoogle Scholar
  15. Dawson MN, Jacobs DK (2001) Molecular evidence for cryptic species of Aurelia aurita (Cnidaria, Scyphozoa). Biol Bull 200:92–96CrossRefGoogle Scholar
  16. de Vargas C, Norris R, Zaninetti L, Gibb SW, Pawlowski J (1999) Molecular evidence of cryptic speciation in planktonic foraminifers and their relation to oceanic provinces. Proc Natl Acad Sci USA 96:2864–2868CrossRefGoogle Scholar
  17. Dobzhansky T (1950) Mendelian populations and their evolution. Am Nat 84:401–418CrossRefGoogle Scholar
  18. Eitner BJ (1995) Systematics of the genus Alopias (Lamniformes: Alopiidae) with evidence for the existence of an unrecognized species. Copeia 1995:562–571CrossRefGoogle Scholar
  19. Etter RJ, Rex MA, Chase MC, Quatro JM (1999) A genetic dimension to deep-sea biodiversity. Deep Sea Res Part I Oceanog Res Pap 6:1095–1099CrossRefGoogle Scholar
  20. Farris JS, Källersjö M, Kluge AG, Bult C (1994) Testing significance of incongruence. Cladistics 10:315–319CrossRefGoogle Scholar
  21. Farris JS, Källersjö M, Kluge AG, Bult C (1995) Constructing a significance test for incongruence. Syst Biol 44:570–572CrossRefGoogle Scholar
  22. Felsenstein J (1981) Evolutionary trees from DNA sequences: a maximum likelihood approach. J Mol Evol 17:368–376CrossRefGoogle Scholar
  23. Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783–791CrossRefGoogle Scholar
  24. Fraser-Brunner A (1950) A synopsis of the hammerhead sharks (Sphyrna), with description of a new species. Rec Austral Mus 22(3):231–219CrossRefGoogle Scholar
  25. Friesen VL, Piatt JF, Baker AJ (1996) Evidence from cytochrome b sequences and allozymes for a ‘new’ species of alcid: the longbilled murrelet (Brachyramphus perdix). Condor 98:681–690CrossRefGoogle Scholar
  26. Garcia-Rodriguez AI, Bowen BW, Domning D, Mignucci-Giannoni A, Marmontel M, Montoya-Ospina A, Morales Vela B, Rudin M, Bonde RK, McGuire PM (1998) Phylogeography of the West Indian manatee (Trichechus manatus): how many populations and how many taxa? Mol Ecol 7:1137–1149CrossRefGoogle Scholar
  27. Gardner MG, Ward RD (2002) Taxonomic affinities within Australian and New Zealand Mustelus sharks (Chondrichthyes: Triakidae) inferred from allozymes, mitochondrial DNA and precaudal vertebrae counts. Copeia 2002:356–363CrossRefGoogle Scholar
  28. Gilbert CR (1967) A revision of the hammerhead sharks (family Sphyrnidae). Proc US Nat Mus 119(3539)CrossRefGoogle Scholar
  29. Goldman N, (1993) Statistical tests of models of DNA substitution. J Mol Evol 36:182–198CrossRefGoogle Scholar
  30. Grady JM, Quattro JM (1999) Using character concordance to define taxonomic and conservation units. Conserv Biol 13:1004–1007CrossRefGoogle Scholar
  31. Hasegawa M, Kishino M, Yano T (1985) Dating the human-ape splitting by a molecular clock of mitochondrial DNA. J Mol Evol 22:160–174CrossRefGoogle Scholar
  32. Heemstra PC (1997) A review of the smooth-hound sharks (genus Mustelus, family Triakidae) of the western Atlantic Ocean, with description of two new species and a new subspecies. Bull Mar Sci 60:894–928Google Scholar
  33. Hillis DM, Mable BK, Larson A, Davis SK, Zimmer EA (1996) Nucleic acids IV: sequencing and cloning. In: Hillis DM, Moritz C, Mable BK (eds) Molecular systematics. Sinauer Assoc, Inc., Sunderland, pp 321–381Google Scholar
  34. Jordan DS (1908) The law of geminate species. Am Nat XLII(494):73–80CrossRefGoogle Scholar
  35. Kocher TD, Carleton KL (1999) Base substitution in fish mitochondrial DNA: patterns and rates. In: Kocher TD, Stepien CA (eds) Molecular systematics of fishes. Academic, NewYork, pp 13–24Google Scholar
  36. Kotas JE (2002) IUCN red list of threatened species—Sphyrna lewini. In: 2002 IUCN Red List of Threatened SpeciesGoogle Scholar
  37. Last PR, Stevens JD (1994) Sharks and rays of Australia. CSIRO, HobartGoogle Scholar
  38. Lee CE (2000) Global phylogeography of a cryptic copepod species complex and reproductive isolation between genetically proximate “populations”. Evolution 54:2014–2027CrossRefGoogle Scholar
  39. Maddison WP, Maddison DR (1992) McClade, version 3.0. Sinauer, SunderlandGoogle Scholar
  40. Martin AP (1992) Application of mitochondrial DNA sequence analysis to the problem of species identification of sharks. NOAA Technical Report NMFS 115. National Technical Information Service, Silver SpringsGoogle Scholar
  41. Martin A (1993) Hammerhead shark origins. Nature 364:494CrossRefGoogle Scholar
  42. Mayr E (1940) Speciation phenomena in birds. Amer Nat 74:249–278CrossRefGoogle Scholar
  43. Moritz C (1994) Application of mtDNA in conservation: a critical review. Mol Ecol 3:401–411CrossRefGoogle Scholar
  44. NMFS (2001) Final United States plan of action for the conservation and management of sharks. NOAA/NMFS/US Department of Commerce, Feb 2001Google Scholar
  45. Palumbi SR (1996) Nucleic acids II: The polymerase chain reaction. In: Hillis DM, Moritz C, Mable BK (eds) Molecular systematics. Sinauer Assoc, Inc., Sunderland, pp 205–247Google Scholar
  46. Pardini AT, Jones CS, Noble LR, Kreiser B, Malcolm H, Bruce BD, Stevens JD, Cliff G, Scholl MC, Francis M, Duffy CAJ, Martin AP (2001) Sex-biased dispersal of great white sharks. Nature 412:139–140CrossRefGoogle Scholar
  47. Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics 14:817–818CrossRefGoogle Scholar
  48. Quattro JM, Chase MR, Rex MA, Greig TW, Etter RJ (2001a) Extreme mitochondrial DNA divergence within populations of the deep-sea gastropod Frigidoalvania brychia. Mar Biol 139:1107–1113CrossRefGoogle Scholar
  49. Quattro JM, Jones WJ, Grady JM, Rohde FC (2001b) Gene-gene concordance and the phylogenetic relationships among rare and widespread pygmy sunfishes (genus Elassoma). Mol Phylogenet Evol 18:217–26CrossRefGoogle Scholar
  50. Schizas NV, Street GT, Coull BC, Chandler GT, Quattro JM (1999) Molecular population structure of the marine benthic copepod Microarthridion littorale along the southeastern and Gulf coasts of the United States. Mar Biol 135:399–405CrossRefGoogle Scholar
  51. Sites JW, Crandall KA (1997) Testing species boundaries in biodiversity studies. Conserv Biol 11:1289–1297CrossRefGoogle Scholar
  52. Slade RW, Moritz C, Heideman A (1994) Multiple nuclear-gene phylogenies: applications to pinnipeds and comparison with a mitochondrial DNA gene phylogeny. Mol Biol Evol 11:341–356PubMedGoogle Scholar
  53. Smith SE, Au DW, Show C (1998) Intrinsic rebound potentials of 26 species of Indo-Pacific sharks. Mar Freshw Res 49:663–678CrossRefGoogle Scholar
  54. Springer S (1941) A new species of hammerhead shark of the genus Sphyrna. Proc Florida Acad Sci (1940) 5:45–62Google Scholar
  55. Stevens JD, Bonfil R, Dunlvy NK, Walker PA. (2000) The effects of fishing on sharks, rays, and chimeras (chondrichthyans), and the implications for marine ecosystems. ICES J Mar Sci 57:476–494CrossRefGoogle Scholar
  56. Stoner DS, Grady JM, Priede KA, Quattro JM (2003) Amplification primers for the mitochondrial control region and sixth intron of the nuclear-encoded lactate dehydrogenase A gene in elasmobranch fishes. Conserv Genet 4:805–808CrossRefGoogle Scholar
  57. Walker TI (1998) Can shark resources be harvested sustainably? A question revisited with a review of shark fisheries. Mar Freshw Res 49:553–572CrossRefGoogle Scholar
  58. Weins JL, Penkrot TA (2002) Deliniting species using DNA and morphological variation and discordant species limits in spiny lizards (Sceloporus). Syst Biol 51:69–91CrossRefGoogle Scholar
  59. Williams ST, Knowlton N, Wiegt LA, Jara JA (2001) Evidence for three major clades within the snapping shrimp genus Alpheus inferred from nuclear and mitochondrial gene sequence data. Mol Phyl Evol 20:375–89CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2005

Authors and Affiliations

  • J. M. Quattro
    • 1
    Email author
  • D. S. Stoner
    • 1
  • W. B. Driggers
    • 1
  • C. A. Anderson
    • 1
  • K. A. Priede
    • 1
  • E. C. Hoppmann
    • 1
  • N. H. Campbell
    • 1
  • K. M. Duncan
    • 2
  • J. M. Grady
    • 3
  1. 1.Department of Biological Sciences, Marine Science Program, and School of the EnvironmentUniversity of South CarolinaColumbiaUSA
  2. 2.Department of ZoologyUniversity of HawaiiHonoluluUSA
  3. 3.Department of Biological SciencesUniversity of New OrleansNew OrleansUSA

Personalised recommendations